Handbooks of Vitamins
Knowledge of vitamins in health and disease has increased markedly in recent years due
to extraordinary advances in analytical and separation methodologies (1,8,9). Although
each of the past six decades has been exciting from the standpoint of improved approaches
in the isolation, identification, and synthesis of vitamins, the past 10 years have been
particularly important. In addition, improved methodology has never been so available to
the typical researcher. With improved methodologies and sources of inexpensive vitamins,
it is now possible to study the functions of vitamins in clinical settings. Improved methods
have led, in turn, to a better understanding of the roles that vitamins play as catalysts,
cellular regulators, and co-substrates. Moreover, there is now a better appreciation of vitaminlike
accessory factors, which may be beneficial to overall health or required on a
conditional basis.
As a result of the continued interest in vitamin-related research, as might be expected,
a large body of information has accumulated in the past 10 years. One goal for
the Handbook of Vitamins: Third Edition, Revised and Expanded, is to provide basic and
fundamental background material to aid the reader in assessing the importance of new
findings regarding vitamin function. In addition to the Handbook of Vitamins, the reader
is directed to other books published by Marcel Dekker, Inc. (5–7, 10–13), to the Methods
in Enzymology series (e.g., Ref. 8), edited by two of the editors of the Handbook of
Vitamins, and to other related textbooks (2–4).
Much of the interest in vitamins stems from an appreciation that there remain sizable
populations at risk for vitamin deficiencies. In addition, the need for vitamins in some
individuals is known to be influenced by genetic polymorphisms and related genetic factors.
Some of the material in the Handbook of Vitamins addresses such points and relationships
that have significant public health impact, e.g., the relationship of B12 and folate
metabolism to homocysteine regulation and, in turn, the connections that homocysteine
may have to vascular diseases and developmental defects.
vi Preface
Indeed, for each of the vitamins, there now exists a broader understanding of function
and requirement. In the chapters dedicated to fat-soluble vitamins, new roles for vitamin
A, K, and D are described. An overview of vitamin E’s role as an antioxidant is
included, as well as a separate chapter on various flavinoids that play roles in oxidant
defense. The section on water-soluble vitamins begins with a brief discussion and review
of bioorganic mechanisms. Each of the subsequent chapters focusing on water-soluble
vitamins was written to provide a general understanding of chemical, physiological, and
nutritional relationships. Some sections highlight methods for evaluating deficiencies or
nutritional requirements and the interaction of vitamins with environmental factors. Also,
the efficacy and safety of vitamin use at high levels is addressed.
The chapters were written with a varied audience in mind: the clinician, the biochemist
whose physiological background may be limited, the advanced nutrition student, and
the dietician. The Handbook of Vitamins is a single source, a summary of vitamin functions,
that is complete, yet accessible. In this kind of endeavor, the challenge is to deal
with an expanding factual information base. The contributors have often chosen to develop
the material in a conceptual, rather than in a highly detailed manner. The Handbook of
Vitamins serves as an authoritative and comprehensive source.
We are indebted to the staff at Marcel Dekker, Inc., for their patience and assistance.
The Handbook of Vitamins is part of a serious and significant effort by Dekker to provide
seminal works that speak broadly to micronutrient function. In addition to those referenced
below, other offerings that complement this book may be found at http://www.dekker.com.
We also wish to acknowledge three colleagues who died before this edition was
completed. Graham Garratt, Executive Vice President and Publisher, was very helpful in
the early stages of assembling the handbook. James Olson, one of the eminent leaders in
the field of vitamin A metabolism, and Lawrence Machlin, who edited the first two editions
of Handbook of Vitamins, passed away during the past year. Each man was a dear friend,
truly distinguished in his field, and will be missed.
Robert B. Rucker
John W. Suttie
Donald B. McCormick
Lawrence J. Machlin
1. G. F. M. Ball, Bioavailability and Analysis of Vitamins in Foods, Chapman & Hall, New
York, 1998, pp. 1–569.
2. J. Basu and T. Kumar, Vitamins in Human Health and Disease, Wallingford: CAB International,
New York, 1996, pp. 1–345.
3. G. F. Combs, The Vitamins: Fundamental Aspects in Nutrition and Health, Academic Press,
San Diego, CA, pp. 1–618.
4. P. J. Quinn and V. E. Kagan, eds., Fat-Soluble Vitamins, Plenum, New York, 1998, pp. 1–
5. C. Rice-Evans and L. Packer, eds., Flavonoids in health and disease, Marcel Dekker, New
York, 1997.
6. L. B. Bailey, ed., Folate in Health and Disease, Marcel Dekker, New York, 1994.
7. A. Bendich, C. E. Butterworth, Micronutrients in Health and in Disease Prevention, Marcel
Dekker, New York, 1991.
Preface vii
8. D. B. McCormick, J. W. Suttie, C. Wagner, eds., Vitamins and Coenzymes, Parts I, J, K, and
L, Academic Press, San Diego, CA, Methods in Enzymology series, Vol. 279–283, 1997.
9. Sauberlich, H. E., Laboratory Tests for the Assessment of Nutritional Status, 2nd ed., CRC
Press, Boca Raton, 1999, pp. 1–486.
10. R. Blomhoff, Vitamin A in Health and Disease, Marcel Dekker, New York, 1994.
11. L. Packer, J. Fuchs, eds., Vitamin C in Health and Disease, Marcel Dekker, New York, 1997.
12. L. Packer, J. Fuchs, eds., Vitamin E in Health and Disease: Biochemistry and Clinical Applications,
Marcel Dekker, New York, 1992.
13. S. K. Gaby, A. Bendich, V. N. Singh, L. J. Machlin, eds., Vitamin Intake and Health: A
Scientific Review, Marcel Dekker, New York, 1990.
This Page Intentionally Left Blank
Preface v
Contributors xi
1. Vitamin A 1
James Allen Olson
2. Vitamin D 51
Elaine D. Collins and Anthony W. Norman
3. Vitamin K 115
John W. Suttie
4. Vitamin E 165
Ching K. Chow
5. Bioorganic Mechanisms Important to Coenzyme Functions 199
Donald B. McCormick
6. Niacin 213
James B. Kirkland and Jean M. Rawling
7. Riboflavin (Vitamin B2) 255
Richard S. Rivlin and John Thomas Pinto
x Contents
8. Thiamine 275
Vichai Tanphaichitr
9. Pantothenic Acid 317
Nora S. Plesofsky
10. Vitamin B6 339
James E. Leklem
11. Biotin 397
Donald M. Mock
12. Folic Acid 427
Tom Brody and Barry Shane
13. Cobalamin (Vitamin B12) 463
William S. Beck
14. Choline 513
Steven H. Zeisel and Minnie Holmes-McNary
15. Ascorbic Acid 529
Carol S. Johnston, Francene M. Steinberg, and Robert B. Rucker
16. Ascorbic Acid Regulation of Extracellular Matrix Expression 555
Jeffrey C. Geesin and Richard A. Berg
17. Nutrients and Oxidation: Actions, Transport, and Metabolism of
Dietary Antioxidants 569
J. Bruce German and Maret G. Traber
Index 589
William S. Beck, M.D. Department of Medicine and Department of Biochemistry and
Cell Biology, Harvard University, Cambridge, Massachusetts
Richard A. Berg, Ph.D. Biomaterials Research, Collagen Corporation, Palo Alto, California
Tom Brody, Ph.D. Department of Nutritional Sciences and Toxicology, University of
California, Berkeley, California
Ching K. Chow, Ph.D. Graduate Center for Nutritional Sciences, University of Kentucky,
Lexington, Kentucky
Elaine D. Collins, Ph.D. Department of Chemistry, San Jose? State University, San Jose?,
Jeffrey C. Geesin, Ph.D. Wound Healing Technology Resource Center, Johnson and
Johnson, Skillman, New Jersey
J. Bruce German, Ph.D. Department of Food Science and Technology, University of
California, Davis, California
Minnie Holmes-McNary, Ph.D. Lineberger Comprehensive Cancer Center, School of
Medicine, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina
xii Contributors
Carol S. Johnston, Ph.D. Department of Nutrition, Arizona State University East,
Mesa, Arizona
James B. Kirkland, Ph.D. Department of Human Biology and Nutritional Sciences,
University of Guelph, Guelph, Ontario, Canada
James E. Leklem, Ph.D. Department of Nutrition and Food Management, Oregon State
University, Corvallis, Oregon
Donald B. McCormick, Ph.D. Department of Biochemistry, School of Medicine, Emory
University, Atlanta, Georgia
Donald M. Mock, M.D., Ph.D. Departments of Biochemistry and Molecular Biology,
University of Arkansas for Medical Sciences, Little Rock, Arkansas
Anthony W. Norman, Ph.D. Division of Biomedical Sciences, Department of Biochemistry,
University of California, Riverside, California
James Allen Olson, Ph.D.† Department of Biochemistry, Biophysics, and Molecular
Biology, Iowa State University, Ames, Iowa
John Thomas Pinto, Ph.D. Nutrition Research Laboratory, Department of Medicine,
Memorial Sloan-Kettering Cancer Center, New York, New York
Nora S. Plesofsky, Ph.D. Department of Plant Biology, University of Minnesota, St.
Paul, Minnesota
Jean M. Rawling, M.D., Ph.D. Department of Family Medicine, Faculty of Medicine,
University of Calgary, Calgary, Alberta, Canada
Richard S. Rivlin, M.D. Weill Medical College of Cornell University, and Department
of Medicine, Clinical Nutrition Research Unit, Memorial Sloan-Kettering Cancer Center,
New York, New York
Robert B. Rucker, Ph.D. Department of Nutrition, University of California, Davis,
Barry Shane, Ph.D. Department of Nutritional Sciences and Toxicology, University of
California, Berkeley, California
Francene M. Steinberg, Ph.D., R.D. Department of Nutrition, University of California,
Davis, California
John W. Suttie, Ph.D. Department of Biochemistry, University of Wisconsin, Madison,
† Deceased.
Contributors xiii
Vichai Tanphaichitr, M.D., Ph.D., F.A.C.P., F.R.A.C.P., F.R.I. Department of Medicine,
Ramathibodi Hospital, and Faculty of Medicine, Mahidol University, Bangkok, Thailand
Maret G. Traber, Ph.D. Linus Pauling Institute, Oregon State University, Corvallis,
Steven H. Zeisel, M.D., Ph.D. Department of Nutrition, School of Public Health and
School of Medicine, University of North Carolina at Chapel Hill, Chapel Hill, North
This Page Intentionally Left Blank
Vitamin A
Iowa State University, Ames, Iowa
Probably the first nutritional deficiency disease to be clearly recognized was night blindness.
The ancient Egyptians, as indicated in the Papyrus Ebers and later in the London
Medical Papyrus, recommended that juice squeezed from cooked liver be topically applied
to the eye to cure night blindness. These writings date from 1500 bc, but the observations
probably are of much earlier origin. The Greeks, who depended heavily on Egyptian medicine,
recommended both the ingestion of cooked liver and its topical application as a cure
for night blindness, a tradition that has persisted in many societies to this day (1).
Although interesting references to vitamin A deficiency diseases and their cure can
be found throughout history, the modern science of nutrition is only about a century old.
The observation that experimental animals lose weight and die on purified diets was noted
by many investigators toward the end of the nineteenth century. In the early part of this
century, specific factors necessary for growth and survival were beginning to be identified.
Frederick Gowland Hopkins in England, for example, during the period 1906–1912 found
that a growth-stimulating principle from milk was present in an alcoholic extract of milk
rather than in the ash.
During the same period, Stepp in Germany identified one of these ‘‘minimal qualitative
factors’’ as a lipid. Soon thereafter, E. V. McCollum and Marguerite Davis in Wisconsin
showed that butter or egg yolk, but not lard, contained a lipid-soluble factor necessary
for the growth of rats. In 1913 they coined the term ‘‘fat-soluble A’’ and thereby
attributed for the first time the growth-stimulating property of these extracts to a single
compound. Approaching the problem in a very different way, Osborne and Mendel at
Yale concomitantly found that cod liver oil or butter was an essential growth-promoting
food for rats. The year 1913, therefore, was the beginning of the modern age of vitamin
A exploration.
2 Olson
Many outstanding findings have been made during the past eight decades, of which
only the most notable will be mentioned here. Inasmuch as both colorless extracts of
liver as well as colored plant lipids showed biological activity, Steenbock in Wisconsin
postulated in 1919 the interconversion of two forms of the vitamin. A decade later, Moore
in England showed that ?-carotene, the plant pigment, was converted to the colorless form
of the vitamin in liver tissue. In the early 1920s, experimental vitamin A deficiency was
well characterized by Wolbach and Howe in Boston, and the existence of vitamin A defi-
ciency in children was noted by Bloch in Denmark. In 1930, Karrer and his colleagues
in Switzerland determined the structure both of ?-carotene and of vitamin A and shortly
thereafter synthesized some of its derivatives. One of these derivatives was retinaldehyde,
and in 1935 George Wald of Harvard, while working in Germany, proved that the retinene
found in visual pigments of the eye was identical to Karrer’s chemical compound retinaldehyde.
Of great importance for applied nutrition was the elegant synthesis of all-trans vitamin
A from the inexpensive precursor ?-ionone by Otto Isler and his collaborators in
Basel in the late 1940s. Within a few years the price of vitamin A, which earlier had been
painstakingly isolated from fish liver oils by molecular distillation, fell 10-fold, and the
possibility of using vitamin A more generally in foods at a reasonable cost was assured.
With the availability of radioactive isotopes in the 1950s, many studies were initiated
on the metabolism of vitamin A, and much of our present knowledge concerning it was
defined. The details of the visual cycle involving rhodopsin in the eye were in large part
worked out, and attention began to focus on the possible somatic function of vitamin A in
growth and cellular differentiation. In subsequent sections, the present state of knowledge
relative to vitamin A will be summarized. Much of the early history of vitamin A is
discussed in Moore’s highly readable monograph (2).
A. Isolation
In nature vitamin A is largely found as an ester and, consequently, is highly soluble in
organic solvents but not in aqueous solutions. The major provitamin carotenoid, ?-carotene,
has similar solvent properties. One of the richest sources of vitamin A is liver tissue,
in particular the liver oils of marine fish and mammals. The esters can be directly isolated
from these oils by molecular distillation at very low pressure, a procedure that has been
used extensively for the commercial preparation of vitamin A–rich oils. Alternatively,
vitamin A might be directly extracted with chloroform or with some other solvent combination,
such as hexane together with ethanol, followed by purification of vitamin A by
chromatographic means. To hydrolyze esters, not only of vitamin A and carotenoids but
also of triglycerides and other lipids, saponification with KOH is commonly used, followed
by extraction with organic solvents. Retinol or its esters can be readily crystallized at low
temperature from a variety of organic solvents, including ethyl formate, propylene oxide,
and methanol.
B. Structure and Nomenclature
The structure and recommended names of major compounds in the vitamin A group of
substances are depicted in Fig. 1. The acknowledged reference compound is called alltrans
retinol, whose formula and numbering system are given in Fig. 1a (3).
Vitamin A 3
Fig. 1 Formulas for retinol and its derivatives. All are all-trans isomers except as noted. (a) retinol,
(b) retinal, (c) retinoic acid, (d) 3,4-didehydroretinol, (e) 11-cis retinal, (f) 9-cis retinoic acid, (g)
5,6-epoxyretinol, (h) retro anhydroretinol, (i) 4-oxoretinol, (j) retinoyl ?-glucuronide, (k) 14-
hydroxy-4,14-retro retinol, (l) retinyl palmitate.
In addition to the all-trans form of vitamin A, all 15 of the other possible isomers
have been prepared and characterized (4–6). The most interesting and important isomers
(4–6) are 11-cis retinal, depicted in Fig. 1e, which is the chromophore of the visual pigments
rhodopsin and iodopsin; 9-cis retinoic acid (Fig. 1f), which is the ligand for the
retinoid X receptor (RXR) transcription factors; and 13-cis retinal, which serves lightdependent
functions in halobacteria. As the number of cis bonds increase, both the absorption
maximum of the isomer and its absorbance tend to decrease. Thus, whereas all-trans
retinaldehyde has an absorption maximum and molecular extinction coefficient of 368 nm
and 48,000 in hexane, respectively, the 9,11,13-cis isomer shows corresponding values
of 302 nm and 15,500 (5,6). The tetra-cis isomer, 7,9,11,13-cis, of ethyl retinoate, retinal,
and retinol show similar hypsochromic and hypochromic shifts relative to their all-trans
analogs (4,5). Although the terms ‘‘cis’’ and ‘‘trans’’ have traditionally been used to
denote various isomers of vitamin A, the current chemical notation is ‘‘Z’’ for ‘‘cis’’
and ‘‘E’’ for ‘‘trans.’’ Thus, 11-cis retinal might equally well be termed 7E,9E,11Z,13Eretinal.
4 Olson
The major commercial forms of vitamin A are all-trans retinyl acetate and all-trans
retinyl palmitate (Fig. 1l). The esters are generally produced because of their increased
stability and better solubility in oils and other commercial preparations. The esters can
also be incorporated into a gelatin matrix that protects them from oxidation over reasonable
periods of time, even when subjected to cooking. These latter forms have been extensively
used in the fortification of foods and animal feeds.
Carotenoids not only serve as precursors of vitamin A but also can quench singlet
oxygen and act as both antioxidants and prooxidants. Of over 600 carotenoids that have
been isolated from nature, only about 50 possess provitamin A activity. An excellent
summary of many aspects of carotenoid isolation, structure, and synthesis is given in the
fine treatise edited by Isler (7) and in a recent comprehensive work (8). Thus, the term
provitamin A is used as a generic indicator for all carotenoids that show the biological
activity of vitamin A. Carotenoids commonly found in ingested fruits and vegetables are
depicted in Fig. 2. The most active and quantitatively the most important of these provitamins
is all-trans ?-carotene (Fig. 2:5). Generally, carotenoids must contain at least one
?-ionone ring that is not hydroxylated to show vitamin A activity.
C. Structures of Important Synthetic Analogs of Vitamin A
A very large number of analogs of vitamin A have been synthesized during the past 15
years. The procedures for synthesizing these retinoids have been comprehensively reviewed
(9). The initial impetus for this vast effort was to identify analogs of all-trans
retinoic acid with a better efficacy/toxicity ratio (‘‘therapeutic index’’) in the prevention
of chemically induced cancers in experimental animals (10). With the discovery of the
retinoid receptors, RAR and RXR, the focus has shifted to the synthesis of analogs that
bind strongly and specifically to various forms of these receptors, thereby either activating
them or inhibiting their activation by retinoic acid (9). Some retinoids with interesting
therapeutic or binding properties are given in Fig. 3. In brief, 13-cis retinoic acid (Accutane)
(Fig. 3a) has been used effectively in the treatment of acne, etretinate (Fig. 3b) and
acitretin (Fig. 3c) in the treatment of psoriasis, and 4-hydroxyphenylretinamide (fenretinide)
(Fig. 3d) in preventive studies of breast cancer. TTNPB (Fig. 3e) is one of the most
toxic retinoids as well as one of the most physiologically active (11). TTNPB interacts
well with RARs but poorly with RXRs. Its 3-methyl derivative (Fig. 3f), in contrast, reacts
well with RXRs but not with RARs. The naphthylamide derivative Am-580 (Fig. 3g) is
specific for RAR?, CD-417 (Fig. 3h) for RAR?, CD-666 (Fig. 3i) for RAR?, and SR-
11237 (Fig. 3j) for RXR (12,13). The latter compound has a fixed cisoid configuration at
the position equivalent to 9-cis retinoic acid.
D. Synthesis
Since the complete elucidation of the structure of vitamin A by Karrer in 1930, intense
effort has been expended in synthesizing vitamin A from a variety of precursors. Of many
possibilities, the two major synthetic procedures used commercially are those of Hoffmann-
La Roche and of Badische Anilin- und Sod-Fabrik (BASF). The Roche procedure
involves as a key intermediate a C14 aldehyde and further requires the efficient reduction
of acetylenic to olefinic bonds near the end of the synthesis. The BASF procedure, on the
other hand, depends heavily on the Wittig reaction, by which a phosphonium ylid reacts
with an aldehyde or ketone to give an olefin and phosphine oxide. The major steps in
Fig. 2 Polyenes and carotenoids in foods that may also be found in animal tissues. 1, phytoene;
2, phytofluene; 3, lycopene; 4, ?-carotene; 5, ?-carotene; 6, ?-cryptoxanthin; 7, zeaxanthin; 8, lutein;
9, canthaxanthin; 10, violaxanthin; 11, neoxanthin; 12, astaxanthin. (From Ref. 139.)
6 Olson
Fig. 3 Some synthetic retinoids with specific therapeutic or protein-binding properties. (a) 13-
cis retinoic acid, (b) etretinate, (c) acitretin, (d) 4-hydroxyphenylretinamide (HPR, Fenretinide), (e)
tetrahydrotetramethylnaphthalenylpropenylbenzoic acid (TTNPB), (f) 3-methyl TTNPB, (g) AM-
580, (h) CD-417, (i) CD-666, (j) SR-11237.
Fig. 4 Major commercial routes to the synthesis of retinyl acetate (B) from ?-ionone (A). The
Isler (Roche) procedure is on the left, the Wittig (BASF) method on the right.
Vitamin A 7
Table 1 Physical Properties of All-trans Retinol and Its Esters
Property Retinol Retinyl acetate Retinyl palmitate
Formula C20H30O C22H32O2 C36H60O2
Formula weight 286.46 328.50 524.88
Melting point (°C) 63–64 57–59 28–29
UV absorptiona
?max 325 326 326
1cm 1820 1530 960
 52,140 50,260 50,390
Excitation ?max 325 325 325
Emission ?max 470 470 470
aIn isopropanol. Values are similar in ethanol but differ in chloroform and other solvents. Absorbance values
for retinyl esters in hexane are about 3% higher than for retinol in hexane.
these two procedures are presented in Fig. 4. Major synthetic routes to vitamin A and the
retinoids have been exhaustively reviewed (7,9,14).
E. Chemical and Physical Properties
In concentrated solution, retinol and its esters are light yellow to reddish hued oils that
solidify when cooled and have a mild pleasant odor. As already mentioned, they are insoluble
in water or glycerol but are readily miscible with most organic solvents. The formula,
formula weights, melting points, and characteristics of the absorption and fluorescent spectra
are summarized in Table 1. Indeed, many other physicochemical properties, including
infrared spectra, polarographic characteristics, proton magnetic resonance, and the like,
have also been studied (7–9,14). Nuclear magnetic resonance (NMR) spectra are particularly
useful in characterizing various isomers of vitamin A (5,14). Vitamin A in crystalline
form or in oil, if kept under a dry nitrogen atmosphere in a dark cool place, is stable for
long periods. In contrast, vitamin A and its analogs are particularly sensitive to oxidation
by air in the presence of light, particularly when spread as a thin surface film. In its natural
form, whether present in liver or bound to protein in the plasma, vitamin A also is quite
stable when stored in the frozen state, preferably at 70°C, in a hermetically closed container
in the dark (15). On the other hand, upon extraction from biological materials, care
must be taken to prevent both oxidation and isomerization of vitamin A (15).
Commercial firms have been successful in preparing several stabilized forms of
vitamin A. For example, retinyl acetate, propionate, or palmitate can be coated in the
presence of antioxidants into a gelatin-carbohydrate matrix. The beadlets thus formed,
upon being mixed in animal feeds or in human food, retain 90% or more of their activity
for at least 6-months if kept under good storage conditions. However, in the presence of
high humidity, heat, and oxygen, the loss is considerably greater.
The most common method for analyzing vitamin A and its analogs in pharmaceutical
preparations, feedstuffs, and tissues is high-performance liquid chromatography (HPLC)
combined with a UV detector, usually set at 325 nm (15). Two major types of chromato8
graphic columns are employed, so-called straight-phase and reverse-phase supports. In
the former, hydrophobic compounds are eluted first and more polar compounds later,
whereas in the latter the reverse order of elution occurs. Isomers of a given retinoid are
best separated on the former (16), whereas different retinyl esters are best resolved on
the latter. Carotenoids can also be well resolved by HPLC, usually in conjunction with a
UV-visible (UV-Vis) detector set at 450 nm (17–19). Other retinoids are usually measured
at their maximal absorption wavelengths.
One of the most sensitive methods for measuring vitamin A is by its intensive greenish
yellow fluorescence (Table 1). Because many other natural compounds fluoresce, such
as the polyene pigment phytofluene, chromatographic separation by HPLC or by other
procedures is usually necessary to ensure adequate specificity in the response. The intense
fluorescence of vitamin A has recently been used as well in separating liver cells containing
large amounts of retinyl ester (stellate cells) from those containing smaller amounts by
means of laser-activated flow cytometry (20).
Mass spectroscopy (MS) is also being increasingly used in the analysis of retinoids
and carotenoids (15). The combination of gas chromatography on selected capillary columns
with MS has been used to study the in vivo kinetics and the equilibration of deuterated
retinol with endogenous reserves of vitamin A in humans (21,22). By using ?-carotene
richly labeled with carbon 13, isotope ratio mass spectrometry has provided important
new information about its metabolism in humans (23,24).
The older conventional methods, such as the colorimetric measurement of the transient
blue complex formed at 620 nm when vitamin A is dehydrated in the presence of
Lewis acids and the decrease in absorbance at 325 nm when vitamin A in serum extracts
is inactivated by UV light, remain of value. Currently, however, these procedures are the
methods of choice only in field surveys in which more sophisticated instrumentation is
not available (25).
Biological tests have a utility that cannot be replaced by specific chemical or physical
methods. The physiological response of a species to a given provitamin, a mixture of
provitamins, or a vitamin A preparation can only be assessed by biological procedures.
Indeed, the complexities of absorption, metabolism, storage, transport, and uptake by tissues
is integrated into a meaningful whole only by biological tests.
Biological methods for vitamin A include the classical growth response tests in
vitamin A–deficient rats, liver storage assays in rats and chicks, and the vaginal smear
technique. The requisite design and pitfalls of using the classical rat growth test have been
well reviewed (26). These classical procedures are rarely used now.
Cell culture systems are now commonly employed for measuring the biological
activity, including the toxicity, of retinoids (27–29). A major focus has been the effects
of retinoids in inducing the differentiation of transformed cells in vitro (28–30).
A. Foods
Common dietary sources of preformed vitamin A are various dairy products, such as milk,
cheese, butter, ice cream, and eggs; liver; other internal organs, such as kidney and heart;
Vitamin A 9
and many fish, such as herring, sardines, and tuna. The very richest sources are liver oils
of shark; of marine fish, such as halibut; and of marine mammals, such as polar bear.
With respect to carotenoids, carrots and green leafy vegetables, such as spinach and
amaranth, generally contain large amounts. Although tomato contains some vitamin A–
active carotenoids, the major pigment is lycopene, which has no nutritional activity. Fruits
like papaya and orange have appreciable quantities of carotenoids. The cereal grains generally
contain very little vitamin A, particularly when milled (an important consideration
in dealing with the problem of vitamin A deficiency among very young children throughout
the world). Yellow maize does have vitamin A–active carotenoids, whereas white
maize, which is popular in many parts of Africa, does not. The richest source of carotenoids
is red palm oil, which contains about 0.5 mg of mixed ?- and ?-carotene per
milliliter. Thus, about 7 mL of red palm oil per day should meet the nutritional needs of
a preschool child.
Food composition tables are commonly used to determine the content of vitamin
A, including carotenoid sources, in the diet. Although the adequacy of vitamin A and
carotenoid intake can usually be estimated by their use, such tables are much less helpful
for quantifying mean intakes of vitamin A.
B. Assessment of Vitamin A Status
Vitamin A status can be classified into five categories: deficient, marginal, satisfactory,
excessive, and toxic. The deficient and toxic states are characterized by clinical signs,
whereas the other three states are not. The most commonly used indicator of clinical
deficiency in surveys is the presence of xerophthalmia (25). Of various eye signs that
appear, Bitot’s spots (XlB) in preschool children have proven to be the most useful. Corneal
involvement, although quite specific, is a relatively rare sign, and conjunctival xerosis
(XlA) is fairly nonspecific. Of course, the disappearance of any reversible sign in response
to vitamin A treatment strengthens its validity. Night blindness, which is difficult to evaluate
in young children by direct measurement, can often be satisfactorily assessed by interviewing
the mother (25).
Other methods of assessing vitamin A status include dietary, physiological, histological,
and biochemical procedures. Dietary methods provide useful information about food
habits but, as already indicated, are less useful for quantifying mean intakes. The current
availability of better values for various carotenoids in foods (31), however, enhances the
utility of this approach. Several physiological techniques exist for measuring the impaired
sensitivity of rod cells in the eye (25). The major histological method is conjunctival
impression cytology. In vitamin A–depleted individuals, goblet cells in the conjunctiva
of the eye tend to disappear, and epithelial cells often assume abnormal shapes (25). However,
the subjectivity of the technique and the existence of confounding variables, e.g.,
trachoma, limit its specificity somewhat. Biochemical methods include the measurement
of concentrations of retinol and retinol-binding protein (RBP) in plasma and tears (25).
Because plasma retinol is maintained homeostatically and is affected by other nutrients
and infections, its measurement is most useful when individuals are depleted of the vitamin.
Two response tests, the relative dose response (RDR) and modified relative dose
response (MRDR) methods, provide more specific information about the vitamin A statuses
of individuals and populations from deficiency through marginal status (25). These
latter tests are increasingly being used to assess vitamin A status. Finally, isotope dilution
10 Olson
techniques, employing vitamin A labeled with deuterium or 13C, are starting to be used
to determine total body stores of vitamin A in populations at risk (21–23,25,32).
C. Requirements and Recommendations
A nutritional requirement, whether for animals or humans, must be defined operationally
in terms of nutrient-specific indicators. If a different indicator is selected, the value of the
requirement may well change. Thus, the presumed requirement will progressively increase
as survival, prevention of clinical signs of deficiency, reproductive performance, and longevity
are used as end points. As a consequence, several tiers of requirements might be
defined, based on the physiological level of satisfactory performance that is desired. Many
vitamins, including vitamin A, also show biological actions, some beneficial and some
adverse at yet higher levels, e.g., the adjuvant effect of large doses of vitamin A in stimulating
the immune system. These effects, which are pharmacological in nature, should not
be considered in setting the nutritional requirement. However, some responses to nutrients
that fall between clearly nutritional and clearly pharmacological effects can be difficult
to classify.
Strictly speaking, the nutritional requirement for a population, as defined by a given
selected indicator, should be expressed as the mean requirement, i.e., when 50% of those
in the population show satisfactory values and 50% show unsatisfactory values of a given
indicator. However, because the objective of nutritional analysis is the maintenance of a
satisfactory status in most individuals in a population, the recommended dietary intake
(RDI), a term used worldwide (33), or recommended dietary allowance (RDA), as used
in the United States and in a few other countries, has been defined as the average intake
over time that theoretically would meet or exceed the nutritional needs of 97.7% (mean
 2 SD) of a selected healthy population group, or, more pragmatically, of ‘‘nearly all’’
members of the group. Although reliable data on the frequency distribution of requirements
for many species are not available, a coefficient of variation of 20% seems to apply
to several vitamins in humans and other species.
Often, the RDI of a nutrient is set at a level generously above that which provides
for any possible physiological need but significantly lower than that which causes toxic
effects. In such instances, a ‘‘safety factor,’’ usually selected somewhat arbitrarily, is
included in the calculation of the desired intake level.
Because all of these procedures have been used in setting recommended intakes
of vitamin A, a range of recommended values exists. In selecting an appropriate level
for a given application, the basis of various recommendations must consequently be
Three sets of recommended dietary intakes for vitamin A are given in Table 2: the
1988 RDI of the Food and Agriculture Organization and the World Health Organization
(FAO/WHO) (34); the 1989 RDAs of the Food and Nutrition Board, National Research
Council, U.S. National Academy of Sciences (35); and the 1991 dietary reference values
(DRV) for the United Kingdom (36). The age categories cited in the table (37) are similar
but not identical in the three recommendations. Three different systems have been used:
a single value in the RDA (USA), a two-tier system in the RDI (FAO/WHO), and a threetier
system in the DRV (UK). In a conceptual sense, the higher tier in the FAO/WHO
system, termed ‘‘safe’’ intake, the top tier in the UK system, termed ‘‘reference nutrient
intake’’ (RNI), and the single RDA values in the U.S. system are roughly equivalent (37).
All presume the presence of an adequate total body reserve of vitamin A for most individuVitamin
A 11


12 Olson
als but also include the requirements of ‘‘those few members of the community with
particularly high needs’’ (36).
The lowest tiers in the FAO/WHO system and in the DRV system show similar
values but are defined differently. The FAO/WHO defines basal requirements as the ‘‘(average
daily) amount needed to prevent clinically demonstrable impairment of function’’
(34). The lower reference nutrient intake (LRNI) is defined as the ‘‘(average daily) amount
of the nutrient that is enough for only the few people in a group who have low needs’’
The estimated average requirement (EAR) in the UK system is the (average daily)
amount of a nutrient that meets the needs of 50% of the analyzed group of people. Thus,
it is truly a mean requirement. In the UK system, the LRNI is 2 SD below the EAR and
the RNI is 2 SD above the EAR. The assigned values for the LRNI in the UK system,
which in large part is an inadequate intake for most people, and for the basal requirement
in the FAO/WHO classification, which is a minimally adequate intake for most people,
are similar. This confounding fact may in part be explained by the different reference
standards for weight that are used. The reference weights for adult men and women, respectively,
in the FAO/WHO report are 65 and 55 kg, in the UK report are 74 and 60
kg, and in the U.S. report are 76 and 62 kg. Despite these differences, the RDA in the
United States is still very generous relative to those of many other countries (37).
Recommended nutrient intakes are established for healthy individuals. Febrile conditions
and lipid malabsorption can markedly increase needs. Genetic defects in the handling
of vitamin A also can significantly affect requirements. Homozygous defects have not been
identified for vitamin A, however, probably because they are inconsistent with survival.
Nonetheless, cases of vitamin A intolerance, in which toxic signs appear in some individuals
who ingest moderate amounts of vitamin A, have been reported (38). These instances,
though rare, seem to have a genetic basis (37).
Another aspect of nutritional requirements that merits comment is the concept of a
single ‘‘optimal’’ intake. Little evidence favors such a view. Rather, most studies indicate
that a broad range of satisfactory intakes exists between a low level that prevents defi-
ciency and a high level that causes toxicity. The ‘‘satisfactory range’’ viewpoint is inherent
in the establishment of a ‘‘safe and adequate range’’ for several nutrients (35).
The most suitable indicator of vitamin A nutriture, as indicated earlier, is its total
body reserve. The status of many other nutrients, such as iron, folic acid, vitamin B12, and
vitamin C, can also be related to their total body reserves. The relation of various indicators
of nutritional status to the total body reserve and the selection of a total body reserve that
meets specified operational criteria of adequacy are essential for the future definition of
more scientifically based nutritional requirements and recommendations.
A. Metabolism
1. Absorption
After foods are ingested, preformed vitamin A of animal tissues and the provitamin carotenoids
of vegetables and fruits are released from proteins by the action of pepsin in the
stomach and of proteolytic enzymes in the small intestine. In the stomach the free carotenoids
and retinyl esters tend to aggregate in fatty globules, which then enter the duodenum.
In the presence of bile salts, the globules are broken up into smaller lipid aggregates,
Vitamin A 13
which can be more easily digested by pancreatic lipase, retinyl ester hydrolase, cholesteryl
ester hydrolase, and the like. The resultant mixed micelles, which contain retinol, the
carotenoids, sterols, some phospholipids, mono- and diglycerides, and fatty acids, then
diffuse into the glycoprotein layer surrounding the microvillus and make contact with the
cell membranes. Various components of the micelles, except for the bile salts, are then
readily absorbed into mucosal cells, mainly in the upper half of the intestine (39). A
specific transporter exists for vitamin A (40) but evidently not for carotenoids. Nonetheless,
carotenoids do interact with each other during absorption (41,42).
The bioavailability and the digestion of vitamin A and carotenoids, needless to say,
are affected by the overall nutritional status of the individual and the integrity of the
intestinal mucosa. Nutritional factors of importance are protein, fat, vitamin E, zinc, and
probably iron. Some types of fiber, e.g., highly methoxylated pectins, markedly reduce
carotenoid absorption (43).
Bile salts, which are detergents, promote the rapid cleavage of retinyl and carotenoid
esters and assist in the transfer of these lipids into mucosal cells. Interestingly, carotenoid
absorption has an absolute requirement for bile salts, independent of its dispersion in a
suitable micelle, whereas vitamin A in any properly solubilized form is readily absorbed
The overall absorption of dietary vitamin A is approximately 80–90%, with somewhat
less efficient absorption at very high doses. The efficiency of absorption of carotenoids
from foods is 50–60%, depending on their bioavailability. The absorption effi-
ciency of carotenoids decreases markedly at high intakes or doses (24,39,44).
2. Transport
Retinol in intestinal mucosal cells, whether derived directly from the diet or formed from
carotenoids, is largely converted to retinyl esters in these cells. Retinyl esters, together
with some unesterified retinol, phospholipids, triglycerides, and apolipoproteins, are incorporated
into chylomicra, which are released into the lymph.
During the transport of chylomicra from the lymph into the general circulation, the
triglycerides are degraded by lipoprotein lipase of the plasma. The latter is activated by
divalent metal ions and requires the presence of albumin as a fatty acid acceptor. After
much of the triglyceride is digested and removed from the chylomicron, the resultant
chylomicron remnant, which contains retinyl ester, cholesteryl ester, some other lipids,
and several apolipoproteins, is taken up primarily by the liver but also to some extent by
other tissues (39). Uptake by liver involves interaction between apolipoproteins E and
B48 in the chylomicron remnant and high-affinity receptors on the cell surface of parenchymal
cells (39).
Whereas dietary retinol is transported to the liver largely as an ester in lipoproteins
formed in the intestine, the mobilization of vitamin A from the liver stores and its delivery
to peripheral tissues is a highly regulated process. A precursor (pre-RBP) of a specific
retinol-binding protein is synthesized in liver parenchymal cells (45). After the removal
of a 3500-d polypeptide, the resultant apo-RBP binds all-trans retinol, and the complex
(holo-RBP), together with transthyretin, is secreted into the plasma. Human RBP, first
isolated in the late 1960s, is a single polypeptide chain, has a molecular weight of 21,000,
and possesses a single binding site for retinol. Under normal conditions of vitamin A
nutriture, approximately 90% of RBP is saturated with retinol. In children the usual level
of RBP in plasma is approximately 1–1.5 µmol/L (20–30 µg/mL), which rises at puberty
to adult levels of 2–2.5 µmol/L (40–50 µg/mL). In human plasma RBP is largely com14
plexed with transthyretin, a tetramer that also binds thyroid hormones, in a 1:1 molar
complex. The association constant (Ka) between transthyretin and RBP is approximately
106 M/L. The amino acid sequence and three-dimensional structure of RBP have been
determined, and a cDNA clone for it has been identified and studied (45).
Within RBP, retinol resides in a hydrophobic ? barrel (46), which protects it from
oxidation or destruction during transport (Fig. 5). RBP has been isolated from the plasma
of many other species, including rat, monkey, rabbit, cow, dog, and chicken. In all of
these species, RBP is similar in size to that found in the human, and in most instances it
forms a complex with transthyretin.
The formation of the larger transthyretin–RBP complex may minimize the loss of
RBP in the urine during its passage through the kidney (45). In chronic renal disease, the
levels of RBP and of plasma retinol are greatly elevated, whereas in severe protein calorie
malnutrition, the amounts of both are reduced to approximately 50%. Thus, the steadystate
concentrations of both vitamin A and RBP in the plasma are dependent on factors
other than vitamin A status alone.
Under normal physiological conditions, the turnover of holo-RBP in the plasma is
quite rapid. When associated with transthyretin, its half-life is approximately 11 to 16 h,
Fig. 5 The structure of human holo-RBP, depicted as a ribbon. The loop that changes conformation
between the holo and apo forms of RBP is depicted in black. The N-terminal end is at the
upper left, the C-terminal end at the lower right. Retinol is also depicted in black within a eightstranded
? barrel. (Graciously provided by Dr. Marcia Newcomer, Vanderbilt University, Nashville,
TN. See also Ref. 46.)
Vitamin A 15
whereas in the apo form it is removed much more rapidly. These half-life values increase,
i.e., turnover decreases, about 50% in severe protein calorie malnutrition.
The retinol ligand of holo-RBP in the plasma not only is taken up by peripheral
target tissues but also is recycled back to the liver. By use of in vivo kinetic analysis,
Green and Green (47) found that an average retinol molecule in the plasma recycles 7–
13 times before being irreversibly lost. As another expression of the efficiency of recycling,
the turnover rate of plasma retinol is 10 times the irreversible disposal rate of wholebody
vitamin A (47). The kidney is the major organ involved in recycling. If its function
is adversely affected by disease, infection, and high fever, large amounts of holo-RBP
are excreted in the urine (48).
Most of the retinol recycled from tissues back to the liver, on kinetic grounds, is
in the form of holo-RBP. However, other transport agents, e.g., retinyl ester on lipoproteins
and retinyl ?-glucuronide, may play some role in recycling. The key point is that retinol
is carefully conserved by the body and is not indiscriminately lost through excretion.
Indeed, the irreversible daily loss of vitamin A in the feces and urine of vitamin A–
depleted animals, and presumably of humans, is much reduced relative to the rate of loss
in vitamin A–sufficient animals (47).
Of all processes relative to the uptake, storage, and transport of vitamin A, the
combination of retinol with RBP in the liver seems to be one of the most specific. Thus,
the only ligand found under normal physiological conditions is all-trans retinol. Other
vitamin A analogs will combine with apo-RBP in vitro, however, and 3-dehydroretinol,
15-methylretinol, and 15-dimethylretinol form holo-RBP analogs in vivo (49,50). Indeed,
the abilities of specific analogs to saturate RBP in vivo can be closely related to their
biological activities (49).
Retinoic acid is primarily transported on plasma albumin and retinyl esters in
low-density lipoproteins (39). The ?-glucuronides of retinol and retinoic acid, although
water-soluble, are bound to plasma albumin and possibly to other plasma proteins in vivo
3. Uptake, Storage, and Release by Cells
The major circulating form of vitamin A in plasma is holo-RBP. Two mechanisms have
been explored for its uptake by cells, one of which is dependent on cell surface receptors
and the other of which is not (45,52).
Cell surface receptors for holo-RBP have been best characterized from the retinal
pigment epithelium of the vertebrate eye. The receptor on the external surface of these
cells, i.e., that in contact with plasma, binds holo-RBP strongly, i.e., with a Ka of 1–3 
107 M/L (45). The estimated size of the receptor is 63 kd. Transthyretin is not bound to
the receptor, nor is an RBP–transthyretin complex necessary for binding. Each cell of the
retinal pigment epithelium has about 50,000 cell surface receptors for RBP. Cells of other
tissues, such as kidney, lung, and muscle, do not possess this receptor, although other
proteins that facilitate retinol uptake may be present on them (45,52). Receptors for holo-
RBP may also be present on cells of the liver, skin, placenta, and the barriers between
the blood and testes and the blood and brain (52). After complexing with the cell surface
receptor, holo-RBP, in all likelihood, is internalized within target cells by receptor-mediated
endocytosis. Thereafter, retinol is released into the cytosol, and the apo-RBP is either
degraded or secreted from cells in a modified form (45,52). Apo-RBP is bound to surface
receptors less tightly than holo-RBP.
16 Olson
Receptor-independent uptake of retinol has been demonstrated in keratinocytes and
in some liver cell lines cultured in vitro (45). Retinol can dissociate from RBP at a rate
sufficient to explain its direct uptake by lipid membranes (52). In all probability, both
mechanisms operate with different cells and under different conditions.
Under normal conditions, more than 90% of the vitamin A in the animal body is
stored in the liver, although small quantities are also found in all other tissues. Thus, most
of the attention has been given to elucidating the mechanism of its storage and release
from the liver.
Endocytosed chylomicron remnants are initially located in low-density endosomes
of parenchymal cells (52,53), where retinyl esters are hydrolyzed. Retinol is then translocated,
presumably in a complex with CRBP-I (cellular retinol-binding protein I), to the
endoplastic reticulum, which is rich in apo-RBP and contains both lecithin : retinol acyltransferase
(LRAT) and acyl-CoA: retinol acyltransferase (ARAT). Retinol might continue
to circulate as holo-CRBP-I, be transferred to apo-RBP for release into the plasma or to
stellate cells, be transferred to stellate cells by some other mechanism, or be esterified
to retinyl ester. Retinyl esters, which predominantly consist of palmitate with smaller
amounts of stearate, oleate, and linoleate, are then stored in vitamin A–containing globules
(VAGs) within the parenchymal cells. When liver reserves of vitamin A are low, parenchymal
cells are the major storage site of vitamin A (39,52,53).
When liver reserves are adequate (20 µg retinol per gram), approximately 80%
of the newly absorbed vitamin A is transferred from parenchymal cells to a specialized
type of perisinusoidal cell, termed the stellate cell (fat-storing cell, lipocyte, Ito cell) (52–
54). The mechanism of intercellular transport is not clear, although complexes of retinol
with CRBP and RBP have been postulated to serve as possible carriers. Within stellate
cells, retinol is rapidly esterified to fatty acids in a pattern similar to that found in parenchymal
cells. The resultant retinyl esters are stored in vitamin A–containing globules, which
contain up to 60% retinyl esters, a significant amount of triglyceride, less phospholipid,
and small amounts of cholesterol, cholesteryl esters, and ?-tocopherol. Several proteins,
including retinyl ester synthase and hydrolase, are associated with these globules. Stellate
cells also contain appreciable amounts of CRBP-I and cellular retinoic acid–binding protein
(CRABP). Whether stellate cells can synthesize RBP is controversial (52,53). Circulating
radiolabeled RBP, however, can be taken up by stellate cells. All other cells of the
liver, including Kupffer’s cells, endothelial cells, and other nonparenchymal cells, contain
very small amounts (5%) of the total liver reserve of vitamin A.
In the storage of dietary vitamin A, the hydrolysis and formation of retinyl esters
is important. Similarly, in the release of vitamin A from the liver as holo-RBP, hydrolysis
of retinyl esters is a key step. In regard to the control of these processes, it is interesting
that several forms of retinyl ester hydrolase exist in the liver (53) and that their activity
is enhanced by a poor vitamin A status and by vitamin E deficiency. On the other hand,
the esterifying enzymes are enhanced when large amounts of vitamin A are ingested. Thus,
these enzymes should be included among factors that control vitamin A metabolism in
vivo. Another important factor may be the ratio of apo-CRBP-I to holo-CRBP-I (55).
When apo-CRBP-I predominates, the hydrolysis of vitamin A is favored, but when holo-
CRBP-I is the major form, the storage of retinyl esters is enhanced.
Stellate cells are found not only in rat liver but in many other tissues and in many
other species (54). Besides storing vitamin A, stellate cells are highly active in synthesizing
collagen and other structural proteins. The factors underlying the association of these two
important processes in stellate cells have not been elucidated.
Vitamin A 17
As already mentioned, retinol is released from parenchymal cells of the liver as a
1:1 complex of holo-RBP and transthyretin. In mice that cannot make transthyretin, holo-
RBP concentrations in the plasma are very low (56). Stellate cells of the liver may also
release holo-RBP directly or may transfer retinol to circulating apo-RBP (52). Cells of
other tissues, and particularly of kidney, seem to release holo-RBP for transport back to
the liver. No structural differences have been noted between RBP synthesized in the liver
and that formed in peripheral tissues (45). Retinyl esters are released from the intestine as
chylomicra, as already noted, and from the liver as very low-density lipoproteins (VLDLs)
(52,53). Although retinoic acid, retinyl ?-glucuronide, and retinoyl ?-glucuronide are present
in plasma in low concentrations, the mode of their release from tissues has not been
4. Transformations
Dietary forms of vitamin A are retinyl esters, derived entirely from animals sources, and
provitamin A carotenoids, derived largely from plant foods. Of the total carotenoids present
in nature, however, less than 10% serve as precursors of vitamin A. Retinol is released
from its dietary esters by the action of pancreatic hydrolases, which act in the presence
of bile salts within the intestinal lumen. Within intestinal mucosal cells, retinol combines
with CRBP-II (cellular retinol-binding protein II) to give a complex that can undergo
either esterification to retinyl ester or oxidation to retinal (39,40,53,55). The major source
of the acyl moiety is the usually saturated fatty acid in the ? position of phosphatidylcholine,
but a transfer from acyl-CoA can also occur (39,40,53,55).
The major reactions of vitamin A metabolism, which are summarized in Fig. 6,
are esterification, oxidation at C-15, oxidation at C-4, conjugation, isomerization, other
miscellaneous oxidative reactions, and chain cleavage (53,55). Retinol and retinal, as well
Fig. 6 Major metabolic transformations of vitamin A. ROL, retinol; RE, retinyl ester; RAL, retinal;
RA, retinoic acid; RG, retinyl ?-glucuronide; RAG, retinoyl ?-glucuronide; 5,6-EROL, 5,6-
epoxyretinol; 5,6-ERE, 5,6-epoxyretinyl ester; 14-HrROL, 14-hydroxy-4,14-retro retinol; 4-HRA,
4-hydroxyretinoic acid; 4-ORA, 4-oxoretinoic acid; 5,6-ERA, 5,6-epoxyretinoic acid; C-19 and C-
16, chain-shortened, oxidized products with the indicated number of carbon atoms. Double arrows
indicate reversible reactions; single arrows irreversible changes. (From Ref. 39.)
18 Olson
as other metabolites reversibly converted to them, all possess significant biological activity.
Retinoic acid and its glucuronide are active in growth but not in vision or, in most
species, in reproduction. Except for 14-hydroxy-4,14-retro retinol, more oxidized products,
such as 4-hydroxyretinoic acid, 5,6-epoxyretinoic acid, and C-19 metabolites, are
largely devoid of biological activity. Retinoyl ?-glucuronide, retinyl ?-glucuronide, and
retinoic acid are normally present in small amounts (3–11 nmol/L, or 1–5 µg/L) in human
plasma. Retinoyl ?-glucuronide is not hydrolyzed in some cells and slowly hydrolyzed
in vivo. Retinoic acid can also be covalently bound to proteins, possibly by means of a
coenzyme A intermediate (39,57).
The enzymology of these reactions in most instances is conventional. Retinol is
reversibly oxidized to retinal by a nicotinamide adenine dinucleotide (NAD)–dependent
pathway in many tissues. Although alcohol dehydrogenase can catalyze this reaction, specific
retinol dehydrogenases (retinal reductases) also exist.
Retinal is then irreversibly converted to retinoic acid in many tissues by aldehyde
dehydrogenases and oxidases (53). The rate of conversion of retinal to retinoic
acid is several fold faster than the oxidation of retinol to retinal. Retinoic acid is inactivated
biologically either by hydroxylation at the C-4 position or by epoxidation at
the C-5/6 positions. In animal tissues, both of these oxidative reactions are irreversible.
4-Hydroxyretinoic acid can be further oxidized to the 4-keto derivative, followed by
a variety of oxidative, chain-cleaving, and conjugative reactions. One of the major
conjugated cleavage products is retinotaurine, which appears in the bile in significant
amounts (58).
Vitamin A also forms several ?-glucuronides by interacting with UDP-glucuronic
acid in the presence of glucuronyltransferases. The most interesting of these conjugates
are retinyl ?-glucuronide and retinoyl ?-glucuronide, both of which retain high biological
activity (53). Both are synthesized in the intestine, in the liver, and probably in other
tissues. In addition, both are endogenous components of human blood and, together with
the glucuronides of 4-ketoretinoic acid and of other oxidized metabolites, appear in the
bile (53). The rate of formation of ?-glucuronides of 9-cis and of 13-cis retinoic acid by
rat liver microsomes in vitro is significantly faster than that of the all-trans isomer (59).
When retinyl acetate or retinyl ?-glucuronide is administered to rats, 5,6-epoxyretinyl
ester is found in discrete amounts in the liver (60). Thus, epoxidation occurs at the
level of retinol as well as of retinoic acid.
?-Carotene and several other provitamin A carotenoids are converted by oxidative
cleavage at the 15,15? double bond to yield two molecules of retinal. The enzyme catalyzing
this reaction, 15,15?-carotenoid dioxygenase, is found in the intestine, liver, kidney,
and some other tissues. This enzyme will also convert ?-apocarotenoids found in plants
and in some animal tissues to retinal (53). 9-cis ?-Carotene is also cleaved by the enzyme
to yield 9-cis retinal, which can be oxidized to 9-cis retinoic acid (61,62), the ligand for
RXR receptors. ?-Carotene can also be cleaved eccentrically to ?-apocarotenals, which
can be further degraded oxidatively to retinal (63). Although the relative importance of
the two pathways in vivo in mammals has not been determined, cell-free preparations of
pig and guinea pig intestines yield retinal as the major, if not sole, product of ?-carotene
cleavage (64,65).
5. Retinoid-Binding Proteins
Retinoid-binding proteins play key roles in the metabolism and function of their ligands
(39,45,53,55,66,67). Major well-characterized binding proteins for retinoids are listed in
Vitamin A 19
Table 3 Properties of Major Retinoid-Binding Proteinsa
Name Abbreviation weight (kd) Major ligand
Retinol-binding protein RBP 21.2 All-trans retinol
Cellular retinol-binding protein, CRBP-I 15.7 All-trans retinol
type I
Cellular retinol-binding protein, CRBP-II 15.6 All-trans retinol
type II
Cellular retinoic acid–binding CRABP-I 15.5 All-trans retinoic acid
protein, type I
Cellular retinoic acid–binding CRABP-II 15.0 All-trans retinoic acid
protein, type II
Cellular retinaldehyde-binding CRALBP 36.0 11-cis retinol and retinal
Interphotoceptor retinol-binding IRBP 135.0 11-cis retinal and all-trans
protein retinol
Epididymal retinoic acid–bind- E-RABP 18.5 All-trans and 9-cis retinoic
ing protein acid
aBased on Refs. 46, 55, and 68.
Table 3. Several are coligands for enzymatic transformations of retinoids, namely, CRBPII,
CRBP-I, CRABP-I, and possibly CRABP-II (66,67). RBP (as already mentioned), interphotoreceptor
retinal-binding protein (IRBP) and epididymal retinoic acid–binding protein
(E-RABP) primarily serve transport functions, the first in the plasma, and the second
in the eye, and the third in the testes (39,45,68). Cellular retinaldehyde-binding protein
(CRALBP) probably also serves a transport function in retinal pigment epithelial cells and
possibly in Mu?ller cells of the eye, but it may also have some role during morphogenesis of
the eye (68). The ratio of apo to holo forms of some of these proteins, as already mentioned,
may regulate metabolic processes (66,67). Apart from these major functions, all
binding proteins sequester retinoids in hydrophobic cavities, thereby protecting the ligand
from degradation and membrane systems of the cell from unwanted interaction with these
amphiphilic ligands. Finally, the RBPs render the hydrophobic ligands water-soluble,
much like glucuronidation does. Some redundancy in their functions may exist in that
transgenic mice without CRABP-II and/or CRABP-I developed normally in one study
(69) but showed polydactyly in another (70).
Other proteins, such as several fatty acid–binding proteins, serum albumin, and ?-
lactoglobulin, also bind retinoids. These interactions, although relatively nonspecific,
might play some physiological role.
6. Excretion
In a quantitative sense, ingested vitamin A is generally metabolized in the following way:
(a) 10–20% is not absorbed and, hence, is excreted within 1–2 days into the feces; (b)
of the 80–90% absorbed, 20–60% is either conjugated or oxidized to products that are
excreted within approximately 1 week in the feces or urine, with a small amount in expired
CO2; and (c) the remainder (30–60%) of the absorbed vitamin A is stored, primarily in
the liver. However, when the initial liver reserves are depleted, the relative amount stored
in the liver is much lower (39).
20 Olson
Stored vitamin A is metabolized much more slowly in the liver and peripheral tissues
to conjugated and oxidized forms of vitamin A, which then are excreted. The half-life for
the overall depletion rate in humans is 128–156 days (71). As a general rule, derivatives
of vitamin A with an intact carbon chain are excreted in the feces, whereas acidic chain–
shortened products tend to be excreted in the urine. In the steady state, approximately
equal amounts of metabolites are excreted in the feces and the urine. During periods of
severe infection, however, large amounts of holo-RBP, often exceeding the daily requirement
for vitamin A, are excreted in the urine (48).
B. Functions
1. Vision
When a photon of light strikes the dark-adapted retina of the eye, 11-cis retinal, present
as a protonated Schiff base of lysine residue 296 in rhodopsin, is converted to a highly
strained transoid form in bathorhodopsin. The more stable all-trans form of bathorhodopsin
is then converted to metarhodopsin I, and subsequently by deprotonation to metarhodopsin
II (68).
Rhodopsin contains both hydrophilic and hydrophobic regions; has a molecular
weight of approximately 38,000 in the cow; is asymmetrical, with a folded length of
approximately 70A° ; and absorbs maximally at 498–500 nm. Like several other transmembrane
proteins, rhodopsin contains seven helical segments that extend back and forth
across the disk membrane. Approximately 60% of the total amino acid structure is found
in these helical proteins. In addition to loops between helical portions, less structured Cterminal
and N-terminal portions extend into the cytosol and the intradisk space, respectively.
The chromophore resides in a hydrophobic pocket formed by several transmembrane
segments near the cytosolic side of the disk membrane. Residues close to the
chromophore include Phe115, Ala117, Glu122, Trp126, Ser127, Trp265, and Tyr268 (68).
Rhodopsin contains an acetylated N terminus and two oligosaccharides on asparagines 2
and 15. The complete amino acid sequence of rhodopsin is known.
Three color pigments exist in the cone cells of the human eye, with maximal absorption
at 420 nm (blue cones), 534–540 nm (green cones), and 563–570 nm (red
cones). The amino acid sequences of these iodopsins are similar, but not identical, to
each other and to that of rhodopsin. All adopt a similar conformation within the disk
membrane, being composed of seven transmembrane segments. Only seven amino acids
differ among the transmembrane segments of the cone pigments. Thus, the differences in
absorption maxima must reside in subtle changes in the ambience surrounding the chromophore
Metarhodopsin II, the penultimate conformational state of the light-activated visual
pigment, reacts with transducin, a G protein attached to the disk membrane that contains
three subunits. In response, the ? subunit of transducin binds GTP in place of endogenously
associated GDP, thereby activating cGMP phosphodiesterase, which hydrolyzes
cGMP to GMP. Because cGMP specifically maintains in an open state the sodium pore
in the plasmalemma of the rod cell, a decrease in its concentration causes a marked reduction
in the influx of sodium ions into the rod outer segment. The membrane consequently
becomes hyperpolarized, which triggers the nerve impulse to other cells of the retina
through the synaptic terminal of the rod cell (68). This sequence of events is summarized
in Fig. 7.
Vitamin A 21
Fig. 7 Possible sequence of steps between the light-induced activation of rhodopsin and hyperpolarization
of the rod cell membrane. R., Rhodopsin; R: light-activated metarhodopsin II; R-PO4,
phosphorylated rhodopsin; ATP, adenosine triphosphate; GTP, guanosine triphosphate; GDP, guanosine
diphosphate; cGMP, cyclic 3?–5?-guanosine monophosphate; GMP, guanosine monophosphate;
T., transducin; T.-GDP, a complex of transducin with GDP; T.-GTP, a complex of transducin
with GTP; PDE, phosphodiesterase. (From Ref. 39.)
Recovery from this activated state occurs in three ways: (a) The ? subunit of transducin,
which also shows GTPase activity, hydrolyzes bound GTP to GDP, thereby leading
to subunit reassociation to inactive transducin. As a consequence, cGMP phosphodiesterase
activity falls and the cGMP level increases back to a normal level. (b) Metarhodopsin
II is phosphorylated at sites in the C-terminal portion by ATP, which then reacts with
arrestin to form a complex that no longer activates transducin. (c) Metarhodopsin II dissociates
through metarhodopsin III to yield all-trans retinal and opsin, which also does not
activate transducin. These events are also summarized in Fig. 7.
Retinol is very much involved in the process of vision, as shown in Fig. 8 (68).
All-trans retinol bound to RBP interacts with a cell surface receptor and is internalized
into retinal pigment epithelial cells (RPEs) (reaction 8). Thereafter, probably as a CRBP
complex, retinol is transacylated from phosphatidylcholine to yield retinyl ester (reaction
3), which in turn undergoes a concerted hydrolysis and isomerization to 11-cis retinol
(reaction 4) (72). The latter, probably as a CRALBP complex, can either be esterified by
transacylation (reaction 3) or be oxidized to 11-cis retinal (reaction 5). The latter compound
is transferred to IRBP, which ferries it to the rod outer segment in the neural retina.
11-cis Retinal, probably in association with a protein but not with CRALBP, then combines
with opsin in the disk membranes to form rhodopsin (reaction 7). Another source
of both all-trans and 11-cis retinal in the RPEs is the discarded tips of the constantly
regenerating rod outer segments, which are phagocytized by the RPE (not shown in Fig. 8).
Light activation of rhodopsin, as already noted, ultimately gives rise to all-trans
retinal (reaction 1), which is reduced in the rod outer segment to all-trans retinol (reaction
2). The latter is transported by IRBP back to the RPE, where it is stored as all-trans retinyl
ester (reaction 3). Thus, the cycle is complete.
During continued exposure of the retina to strong light, rhodopsin is largely converted
to opsin plus all-trans retinal, which tends to accumulate as all-trans retinyl ester
in the RPE. During dark adaptation, the reverse process occurs. As the rhodopsin concentration
increases in the rod outer segment, the retina becomes increasingly sensitive to
light of very low intensity.
22 Olson
Fig. 8 Reactions of the mammalian visual cycle. Horizontal lines separate cellular and extracellular
compartments in the retina involved in retinoid metabolism. Most of the enzymatic reactions of
the cycle occur in RPE, including the generation of 11-cis retinol. The reactions shown below the
double line occur within the photoreceptor cell. Retinoid-binding proteins found within the indicated
compartments are shown on the left. CRALBP, cellular retinaldehyde-binding protein; CRBP, cellular
retinol-binding protein; IRBP, interphotoreceptor retinoid-binding protein; RBP, retinol-binding
protein; at-, all-trans-; 11-, 11-cis-; Rol, retinol; Ral, retinaldehyde; RE, retinyl ester; gNa, sodium
conductance; RPE, retinal pigment epithelium. Reactions shown are: 1, photoisomerization; 2, alltrans
retinol dehydrogenase; 3, lecithin:retinol acyltransferase (LRAT); 4, retinyl ester isomerohydrolase;
5, 11-cis retinol dehydrogenase; 6, retinyl ester hydrolase; 7, regeneration of rhodopsin
(nonenzymatic); 8, uptake of plasma retinol. (From Ref. 68.)
Most of the processes involved in the visual cycle have been extensively studied,
including the enzymes that are involved (68).
2. Growth
The earliest assays for vitamin A were based on the growth response of rats fed a purified
diet. The fact that maintenance of normal vision and enhancement of growth are two
separate properties of the vitamin A molecule was dramatically demonstrated by the observation
that retinoic acid stimulated growth but could not maintain vision (73). In nutritional
studies, the onset of vitamin A deficiency has often been detected by a so-called growth
plateau, which after several days is followed by a rapid loss of weight and, ultimately,
death. However, in animals cycled on retinoic acid, i.e., given retinoic acid for 18 days
followed by 10 days of deprivation, rats become exquisitely sensitive to the removal of
Vitamin A 23
retinoic acid from the diet (74). In these animals loss of appetite occurs within 1–2 days
after the withdrawal of retinoic acid, which is closely followed by a depression in growth.
Loss of appetite is therefore one of the first symptoms noted in all vitamin A–deficient
animals. In humans, however, night blindness and mild xerophthalmia seem to be the
earliest signs of clinical deficiency.
Decreased food intake is not due in this case to impaired taste function or to the
poor palatability of the deficient diet. Because many factors affect appetite, it has not been
possible to define specifically the molecular effect of vitamin A on this process. Inasmuch
as distortions in nitrogen metabolism and in amino acid balances within tissues and in
the plasma occur concomitantly, the effect of vitamin A deficiency on appetite may be
related to these latter abnormalities as well as to disturbances in cell differentiation.
3. Cellular Differentiation
In vitamin A deficiency, mucus-secreting cells are replaced by keratin-producing cells in
many tissues of the body. Conversely, the addition of vitamin A to vitamin A–deficient
keratinizing cells in tissue culture induces a shift to mucus-producing cells. Retinoids also
rapidly induce F-9 teratocarcinoma cells, as well as many other cell lines, to differentiate.
In this process, many new proteins appear in the newly differentiated cells. Thus, vitamin
A and its analogs, both in vivo and in vitro, markedly influence the way in which cells
differentiate (30,39).
The mechanism by which retinoids induce cellular differentiation is becoming clear
(Fig. 9). Within tissue cells, all-trans retinol, in association with CRBP, can be oxidized
to all-trans retinoic acid and presumably can also be isomerized to 9-cis retinol, which
in turn can be oxidized to 9-cis retinoic acid. The latter can also arise by isomerization
of all-trans retinoic acid or by central cleavage of 9-cis ?-carotene (61,62). All-trans or
9-cis retinoic acid is transported on CRABP or on other RBPs to the nucleus, where it is
tightly bound to one or more of the three (?, ?, ?) retinoic acid receptors (RAR) or to
one or more of the three (?, ?, ?) retinoid X receptors (RXR), respectively (39,75–79).
The RAR and RXR receptors, like other nuclear hormone receptors, possess six
protein domains with specific functions. At the N-terminal end, domains A and B serve
as physiological activators of the receptor; domain C, which is highly conserved, contains
zinc–sulfhydryl interactions (‘‘zinc fingers’’) that bind to DNA; domain D is a hinge
region that provides the necessary conformation of the receptor; domain E binds the ligand;
and domain F, at the C-terminal end, enhances dimerization. All nuclear retinoid receptors
contain 410–467 amino acids and have molecular weights of 45–51 kd (39,75–79). RAR?
and RXR? show very limited homology to each other (only 61% in the DNA binding
domain and 27% in the ligand binding domain). Each of the six human retinoid receptors
shows different chromosomal locations, developmental expression, and organ localization.
Thus, RAR? and RXR? are widespread in their occurrence; RAR? and RXR? are both
present in adult muscle and heart but differ in their embryonic appearance and localization;
RAR? is found in adult skin and lung but has a variable embryonic pattern; and RXR?
is found in liver, skin, and kidney of both adults and embryos. Thus, each receptor seems
to show its own pattern of expression and function. Furthermore, no functional relationship
exists between the respective ?, ?, and ? forms of the two receptors (75–79).
Of the two retinoid receptor families, RXR shows the broadest actions. In the absence
of 9-cis retinoic acid, RXR forms heterodimers with the vitamin D receptor (VDR),
the thyroid receptor (TR), and RAR but only when the latter are charged with their respective
ligands, namely, 1?, 25-calcitriol, triiodothyronine, and all-trans or 9-cis retinoic
24 Olson
Fig. 9 Roles of retinoids in cellular differentiation at, All-trans; 9c, 9-cis, ROL, retinol; HrROL,
14-hydroxy-4,14-retro retinol; RAL, retinal; RA, retinoic acid; RAG, retinoyl ?-glucuronide; RR,
cell surface receptor for holo-RBP; RBP, plasma retinol-binding protein; CRBP, cellular retinolbinding
protein; CRABP, cellular retinoic acid–binding protein; RAR?, ?, ?, nuclear retinoic acid
receptors, forms ?, ?, and ?; RXR?, ?, ?, nuclear 9-cis retinoic acid receptors, forms ?, ?, ?; TR,
triiodothyronine bound to the nuclear thyroid receptor; VDR, calcitriol bound to the nuclear vitamin
D receptor; X, unknown nuclear receptors; solid arrows, known transformations or effects; dashed
arrows, postulated transformations or effects; ?, unknown occupancy. (From Ref. 39.)
acid (75–79). In the presence of 9-cis retinoic acid, RXR will bind its ligand and form a
homodimer. Thus, heterodimer and homodimer formation are competitive actions that are
dependent on the ratio of all-trans retinoic acid to 9-cis retinoic acid and the latter’s
concentration in the nucleus.
The hormone response element (HRE) in DNA for both RXR and RAR is the
consensus sequence AGGTCA. For gene activation, a direct repeat, i.e., AGGTCA–
other bases–AGGTCA, rather than a palindromic arrangement, is most common. The
spacing between the two sequences and the number of tandem repeats largely defines
the specificity of the interaction. This observation has given rise to the DR 1-2-3-4-5
rule, in which the numbers refer to the number of intervening deoxynucleotide base pairs
between the directly repeating (DR) consensus sequences (75–79). The RXR response
element usually employs DR-1 spacing, the RAR response element uses DR-2 or DR-5
spacing, and the VDR and TR response elements use DR-3 and DR-4 spacing, respectively.
The sequence of intervening bases often is important as well, but not always. The
RXR homodimer is demonstrably formed in the presence of 9-cis retinoic acid, whereas
the RAR homodimer only seems to form under special circumstances within cells
In addition to the interaction with RAR, VDR, and TR, RXR also forms heterodimers
with two orphan receptors, the chicken ovalbumin upstream promoter transcription
Vitamin A 25
factor (COUP-TF) and the peroxisomal proliferator activated receptor (PPAR). The ligands
for these two receptors are not known. Although most heterodimers of RXR, including
that with PPAR, activate gene expression, RXR-COUP-TF, either as such or as a
COUP-TF homodimer, inhibits the expression mediated by the RXR homodimer of the
CRABP-II gene. Similarly, RAR, probably by combining with cJun in a nonproductive
complex, also inhibits the activation of the AP-1 site, which is important for cell proliferation.
The AP-1 site normally is activated by the cJun-cFos heterodimer. Thus, the retinoid
receptors can show both activation and suppression of gene expression, depending on the
nature of the heterodimers formed. Genes are activated by retinoids as a result of the
binding of an appropriate homo- or heterodimer to a hormone response element in DNA,
whereas gene expression seems to be suppressed by the competition between a retinoid
receptor and some other transcription factor for the latter’s activating partner protein
Many genes contain response elements for the retinoid receptors, as shown in Table
4 (75–80). Although all of these effects cannot be placed in a cohesive physiological
framework, some deserve special mention, namely: (a) the stimulation of both certain
cytosolic and nuclear binding proteins for retinoids by retinoic acid via retinoic acid response
elements; (b) the stimulation of Hox a-1 (Hox 1.6) and Hox b-1 (Hox 2.9), initiating
genes of embryonic development in several cells and species and (c) enhancement of class
I alcohol dehydrogenase type 3, which may well induce the synthesis of more retinoic
acid from retinol.
Some retinoids may stimulate differentiation by a different pathway; for example,
retinoyl ?-glucuronide does not bind to CRBP, CRABP, or nuclear RAR but nonetheless
is highly active biologically (39,53). Similarly, B lymphocytes differentiate in response
to 14-hydroxy-4,14-retro retinol but not to all-trans retinoic acid (81). Furthermore, the
binding to RAR of various acidic retinoids, both natural and synthetic, relates closely to
their invoked cellular responses but not to their binding affinities for CRABP, at least in
some cellular systems (39).
4. Morphogenesis
Both a deficiency and an excess of vitamin A and of most other retinoids adversely affect
embryogenesis (79,82). Studies on the role of vitamin A in embryogenesis were greatly
stimulated by the demonstration that an implant containing all-trans retinoic acid, when
placed in the anterior part of the developing chick limb bud, mimics the activity of the
naturally occurring zone of polarizing activity (ZPA) (82,83). Thus, the hypothesis arose
Table 4 Some Proteins Whose Genes Contain Response
Elements for Retinoid Receptors
RAR?2 Hox a-1 (Hox 1.6)
RAR?2 Hox b-1 (Hox 2.9)
RAR?2 Phosphoenolpyruvate carboxykinase
CRBP-I Apolipoprotein A-I
CRBP-II Ovalbumin
CRABP-II Complement factor H
Laminin B1 Alcohol dehydrogenase (class I, type 3)
Acyl coenzyme A oxidase
26 Olson
that all-trans retinoic acid might well be one of a presumed host of morphogens that
control embryologic development (39).
The initial hypothesis was refined to postulate that a posterior-to-anterior gradient
of all-trans retinoic acid, but not of retinol, was primarily responsible for pattern formation
in the chick limb bud (82,83). In a historical context, pattern formation in the skin was
postulated to depend on gradients of vitamin A more than 20 years ago (84). A significant
body of experimental data accord with the gradient hypothesis. On the other hand, many
other observations do not, including the observation that the gene product of Sonic hedgehog
induces the polarizing activity of the ZPA, independent of the presence of RA (85).
Thus, the gradient hypothesis is not as attractive today as it was when first formulated
(80,82,84). Nonetheless, retinoic acid, some of its metabolites, and its receptors clearly
play crucial roles in development (79,80,82,83).
Other retinoids also induce limb bud duplication: tetrahydrotetramethylnaphthalenylpropenylbenzoic
acid (TTNPB), CD-367, a structurally related compound, and 9-cis
retinoic acid are much more active than all-trans retinoic acid; all-trans 3,4-didehydroretinoic
acid and Am-580 are as active; and 13-cis retinoic acid and many other retinoids are
less active (82).
Retinoid receptors and cellular retinoid-binding proteins are expressed in a temporally
dependent manner in various parts of the developing embryo. In 9.5- to 10-d mouse
embryos, for example, RAR? and RAR? are uniformly expressed throughout the mesenchyme,
whereas RAR? is restricted to the proximal part of the limb bud. Concomitantly,
CRABP-I is expressed uniformly in the mesenchyme as well as in the apical ectodermal
ridge. As the embryo develops, the distribution and concentrations of both retinoid-binding
proteins and retinoid receptors change.
To define the possible roles of specific retinoid receptors in development, mice lacking
one allele of various isoforms of the RAR family, of all RAR? isoforms, and of all
RAR? isoforms were created (79). Null mice homozygous for single RAR isoforms
showed no deformities, whereas mice homozygous for either all RAR? or all RAR? isoforms
showed an altered phenotype. Although indistinguishable at birth from their wildtype
littermates, mice null for RAR? were rejected by their mothers, grew slowly, and
eventually died, even though no malformations could be detected (79). In contrast, most
RAR? null mice showed abnormalities of the axial skeleton but not of the limbs. Null
mice for both RAR? and RAR? showed more severe abnormalities and died soon after
birth (79). The RXR? null mutation was embryolethal. These interesting findings strongly
implicate the retinoid receptors in embryogenesis but also indicate that considerable redundancy
exists in the functions of specific isoforms. The observation that limbs develop
normally in the absence of RAR? or RAR? is particularly interesting in view of the marked
actions of retinoic acid on that process (79).
Genes that clearly play direct roles in development are the four Hox gene clusters
(80,82). As already indicated, Hox a-1 (Hox 1.6) contains a retinoic acid response element
(RARE) at its 3? end. Activated Hox a-1 could set in motion sequential activation of other
5?-located genes in its cluster (Hox a-2 to a-13). Hox b-1 (Hox 2.9) of the Hox b family,
which is involved in hindbrain development, contains an enhancing RARE at its 3? end
and a suppressing RARE at its 5? end (86). Genes of the Hox d family, which are known
to be involved in limb development, are also activated by retinoic acid (80,82). Other
genes are also activated or suppressed by treatment with retinoic acid (80). Their interplay
in development is slowly being clarified.
Vitamin A 27
Retinoids not only influence the normal physiological development of the embryo
but also, in larger amounts, induce fetal abnormalities. Thus, the regulation of retinoic
acid metabolism in the developing embryo is a crucial aspect of the overall complex
5. The Immune Response
Vitamin A was early termed the ‘‘antiinfective’’ vitamin, based on the increased number
of infections noted in vitamin A–deficient animals and humans (81). In vitamin A defi-
ciency, both specific and nonspecific protective mechanisms are impaired: namely, the
humoral response to bacterial, parasitic, and viral infections; cell-mediated immunity; mucosal
immunity; natural killer cell activity; and phagocytosis (39). Large doses of vitamin
A can also serve as an adjuvant. When vitamin A–deficient animals are supplemented
with vitamin A, immune responses generally improve. The immune responses to certain
antigens in vitamin A–depleted children are also enhanced by vitamin A supplementation
(87). However, increased responsiveness in children is less marked than in animals, probably
because of the presence of multiple nutritional deficiencies in malnourished populations
and a poor, but not acutely deficient, vitamin A status (39).
The primary immune response to protein antigens, such as tetanus toxoid, is markedly
reduced in vitamin A deficiency (81). On the other hand, the process of immunological
memory, essential for a marked secondary response, does not seem to be adversely
affected. In a similar manner, the primary immune response to membrane polysaccharides
of bacteria is severely depressed by vitamin A depletion but quickly recovers with vitamin
A supplementation (81). In contrast, the immune response to bacterial lipopolysaccharides
is unaffected by vitamin A status. Interestingly, the injection of lipopolysaccharide into
vitamin A–deficient rats greatly enhances their immune response to tetanus toxoid and
to bacterial polysaccharides (81). The nature of this stimulatory effect of lipopolysaccharide
has not been clarified. Nonetheless, the observation shows that the antibody-forming
process in vitamin A–deficient animals is functionally intact (81).
The humoral response to many viruses, including measles, herpes simplex, and the
Newcastle disease virus, is impaired in vitamin A deficiency and improved by repletion.
Similarly, the host response to some parasitic infections is reduced by vitamin A depletion
In most instances, except as noted below, retinoic acid was more potent than retinol
in restoring the humoral response in vitamin A–depleted animals (88). At least one
mechanism of action for retinoic acid in the stimulation of antibody formation is now
becoming clear (88). In vitamin A–sufficient animals, an antigen is phagocytized by an
antigen-presenting cell (APC), often a macrophage, which metabolizes the antigen to a
fragment that is presented on the cell surface bound to class II molecules of the major
histocompatibility complex (MHC). T-helper lymphocytes are stimulated by contact with
APC to form interleukin-2 (IL-2), which in turn enhances T-cell proliferation as well
as B-cell growth and differentiation. Activated B cells (plasma cells) initially produce
IgM antibodies but switch to the production of high-affinity IgG antibodies upon maturation
Two types of T-helper cells exist, TH1 and TH2. TH1 cells secrete cytokines that
stimulate cell-mediated immunity (CMI) and TH2 cells secrete cytokines that enhance
antibody production (88). Thus, the balance between TH1 and TH2 will determine the
nature of the response to a given antigen. TH1 cells produce interferon ? (IFN-?) and TH2
28 Olson
cells produce IL-10, IL-5, and IL-4. Cross-regulation exists, inasmuch as IFN-? inhibits
TH2 cell proliferation and IL-10 and IL-4, via different mechanisms, inhibit both TH1 cell
development and IFN-? production (88).
In mesenteric lymph node cells of vitamin A–deficient mice, IFN-? is overproduced
and IL-10 and IL-5 are underproduced (88). Thus, the ratio of TH1 to TH2 cells is increased,
presumably leading to an enhancement of the CMI response and to a decrease in the
antibody response (88). These interesting observations explain the decrease in humoral
immunity in vitamin A deficiency but do not accord with the general observation that
CMI responses are usually depressed in vitamin A deficiency (81). Nonetheless, these
valuable studies pinpoint the T-helper cell as a major site of vitamin A action in the
immune response.
Much interest is now being shown in ?-14-hydroxy-retro retinol (HRR). B-Lymphoblastoid
cells transformed with Epstein–Barr virus die unless HRR, which is derived
from retinol, is present in the medium (81). Retinoic acid is inactive in this system. Retinol,
presumably via HRR, also is involved in the proliferation of normal B cells and T cells.
Various cytokines can modulate the process but cannot replace HRR in it. HRR has also
been identified in many other types of cells. These studies have stimulated much interest
for two reason: (a) retinoic acid, which in most instances is presumed to be the active
form of vitamin A, is inactive in these systems, and (b) other retro derivatives of vitamin
A, such as anhydroretinol and retroretinol, do not enhance animal growth. As yet, however,
the enzyme responsible for the conversion of retinol to HRR has not been studied. These
observations indicate that another signaling pathway for retinoids, albeit specialized in
given cells, may exist.
Phagocytosis, particularly the ‘‘oxygen burst’’ following the ingestion of a foreign
body, and IgA secretion are also depressed in vitamin A deficiency (81). The synthesis
of goblet cell mucins is reduced as well by vitamin A depletion, both in the intestinal
mucosa and in the conjunctiva of the eye. Because protein malnutrition also adversely
affects the immune response, children afflicted with both deficiencies are clearly at increased
risk of developing severe infections (39).
In vitamin A–sufficient animals, carotenoids also enhance the immune response.
Both nutritionally inactive (canthaxanthin) and nutritionally active (?-carotene) carotenoids
have similar effects (39). Their mechanism of action is not known.
6. Transmembrane Proton Transfer
The past generalization that vitamin A was found only in the animal kingdom was upset
in 1971 by the dramatic discovery of a new retinaldehyde-containing pigment, bacteriorhodopsin,
in the membrane of the purple bacterium Halobacterium halobium (89). Soon
thereafter, its basic function was discovered, namely, that under the influence of light,
protons are pumped from the inside to the outside of the bacterial cell. The chemiosmotic
gradient thereby established could be used for the active transport of nutrients into the
cell and for the formation of ATP and other energy storage compounds.
Bacteriorhodopsin and rhodopsin are similar in many ways: Bacteriorhodopsin contains
248 amino acids and has a molecular weight of 26,000, 70% of which is in the form
of ? helices. Seven helical segments stretch back and forth across the membrane, each
of which contains approximately 23 amino acid residues (90). The two chromophores of
bacteriorhodopsin are 13-cis and all-trans retinaldehyde, both of which are bound to the
protein as a Schiff base at Lys216 in approximately a 1:1 ratio at thermal equilibrium.
The chromophore lies in the middle of the membrane at a tilted angle of 27 degrees from
Vitamin A 29
the plane of the membrane. In the light, the 13-cis isomer is converted to the all-trans
form. In the dark-adapted pigment, as in the case of rhodopsin, the Schiff base is protected
from reagents like hydroxylamine; however, after exposure to light, the reaction sites are
exposed. After the absorption of a photon, a sequence of conformational changes occur
in bacteriorhodopsin that are analogous to those seen with rhodopsin.
In the photocycle of bacteriorhodopsin, the Schiff base is deprotonated and then
reprotonated. Because only one proton is transferred across the membrane in one photocycle,
the Schiff base proton released probably plays a role in proton transfer.
In addition to bacteriorhodopsin, halobacteria contain four other related pigments:
halorhodopsin, involved in chloride transport; sensory photosystems SR-I and SR-II,
which control phototaxis; and slow-cycling rhodopsin (SCR), the function of which is
unclear (91). SR-I (?max 587 nm) provides a light-attractant signal and SR-II (?max 373) a
light-repellant signal. SR-I forms a molecular complex with a membrane-bound transducer
protein, Htrl (92). The phototactic action of the SR-I–Htrl complex does not involve deprotonation
and reprotonation. However, SR-I, when separated from Htrl, pumps protons
across the membrane just like bacteriorhodopsin. Thus, the phototactic function of SR-I, as
contrasted with the proton-pumping, energy-yielding action of bacteriorhodopsin, depends
crucially on its interaction with the transducer protein Htrl (92). The actions of SR-I in
photosynthetic bacteria and of rhodopsin in mammalian physiology clearly are highly
In animals, retinoids do not seem to play similar roles, although they can inhibit
some processes involving active transport.
7. Gap Junction Communication
All-trans retinoic acid and some carotenoids elevate mRNA in cells in vitro for connexin
43, a gap junctional protein involved in intercellular communication (93). Interestingly,
the action of canthaxanthin in enhancing connexin 43 synthesis is mediated by its cleavage
product, 4-oxoretinoic acid (94). A transcription factor for 4-oxoretinoic acid, however,
has not as yet been identified.
Vitamin A and various retinoids have been used to treat nutritional inadequacy, some skin
disorders, and certain forms of cancer.
A. Nutrition
Oral doses of vitamin A of 200,000 IUa (60 mg or 2.1 mmol) in oil have been used as a
prophylactic measure in many less industrialized countries, but particularly in Asia
(95,96). Such doses are usually given one to three times a year to children 2–6 years of
age. Smaller doses (25,000–100,000 IUa) have been used in younger children. Because
vitamin A is generally well absorbed and is stored in the liver and other tissues, the procedure
has been effective in reducing the incidence of acute vitamin A deficiency in children.
Although vitamin A is relatively inexpensive, the logistics of maintaining a public health
measure involving the individual dosing of children has not always proved easy.
Lactating women in less industrialized countries are also at risk. Thus, the administration
of 200,000 IUa of vitamin A to a mother soon after the birth of her child is helpful
in enhancing her vitamin A reserves as well as in increasing the concentration of vitamin
30 Olson
A in her milk. The only danger is that a lactating woman may again become pregnant
and, as a consequence, the dose may have teratogenic effects.
Other intervention strategies for improving the vitamin A status of groups at risk
include the fortification of foods with vitamin A, horticultural programs featuring carotenoid-
rich foods, and nutrition education. All of these interventions have been applied
successfully in various countries. Needless to say, each approach has its own set of advantages
and drawbacks (95,96).
In general, the vitamin A status of children in most less industrialized countries is
improving. This encouraging positive development may be attributed to a variety of factors:
the above-cited intervention strategies, other government-sponsored programs in nutrition
and health, improved communication, a better standard of living, and political stability
B. Skin Disorders
Various retinoids, and particularly 13-cis retinoic acid (Fig. 3a), etretinate (Fig. 3b), and
acitretin (Fig. 3c), are used to treat acne, psoriasis, and other skin disorders (97,98). Although
highly effective, retinoic acid and etretinate are teratogenic at high doses when
given orally, and retinoic acid is a skin irritant when applied topically. Although most
efficacious retinoids are also toxic, some conjugated forms, such as retinoyl ?-glucuronide
and hydroxyphenylretinamide, retain their therapeutic actions with less, if any, toxicity
(99). Topical all-trans retinoic acid can also reduce wrinkling and hyperpigmentation
caused by photo-induced aging (39,97,98).
The mechanism of these effects is probably multifactorial. On the physiological
level, retinoic acid can alter epidermal differentiation pathways, induce epidermal hyperplasia,
inhibit transglutaminase, and, as a result, reduce the cross-linking of proteins in the
cornified envelope, change the pattern of keratins formed by keratinocytes, alter membrane
viscosity, stimulate gap junction formation, influence various facets of the immune system
(including the inflammatory response), inhibit collagenase formation, and reduce sebum
production (97,98).
On the molecular level, skin cells contain the usual cytosolic binding proteins and
some nuclear receptors for retinoids (98,100). The distribution of the latter, however,
seems unique for skin: RXRs predominate over RARs by a ratio of 5: 1. Of the RARs,
RAR? represents 87% and RAR? 13% of the total. RAR? is not detected. RXR? is the
only detected receptor of the RXR family. The major heterodimer present, as expected,
is RXR?/RAR? (100). Topically applied all-trans retinoic acid induces CRABP-II
strongly, but not CRABP-I, in human skin. Thus, all of the elements for the control of
differentiation by retinoids are present in skin cells, but the key factors involved in the
therapeutic effects of RA on abnormal skin are only being identified slowly (98,101).
C. Cancer
Retinoids have been used both as therapeutic and as chemopreventive agents against a
large variety of tumors in both experimental animals (102) and humans (103).
In experimental animals, retinoids have been tested, in most instances with positive
results, as chemopreventive agents against chemically induced cancers of the mammary
gland, skin, lung, bladder, pancreas, liver, digestive tract, and prostate gland (102). Combination
therapy, e.g., chemopreventive use of hydroxyphenylretinamide and tamoxifen
against N-methyl-N-nitrosourea-induced mammary cancers, has often proved more effecVitamin
A 31
tive, and sometimes synergistically so, than the use of only one agent (102). Retinoids
alone show the best chemopreventive effects with tumors that follow discrete promotional
stages of carcinogenesis, e.g., mammary and skin tumors.
In humans, trials with retinoids or carotenoids, often together with other nutrients,
have been conducted to prevent oral premalignancies, bronchial metaplasia, laryngeal papillomatosis,
Barrett’s esophagus, actinic keratosis, and cervical dysplasia—all considered
to be premalignant lesions (103). Chemoprevention trials against secondary primary tumors
of the head and neck, skin, breast, lung, and bladder have also been initiated or
conducted, as have several primary prevention trials for epithelial cancer. Therapeutic
trials have been conducted with a variety of hematological and epithelial malignancies
The most promising results have been found with all-trans retinoic acid for acute
promyelocytic leukemia, but drug resistance developed rapidly. Other positive responses
were obtained with 13-cis retinoic acid for squamous cell cancers of the skin; with 13-
cis retinoic acid for mycosis fungoides; with 13-cis retinoic acid, etretinate or ?-carotene
for oral leukoplakia; with 13-cis retinoic acid for laryngeal papillomatosis; with etretinate
for actinic keratosis; with all-trans retinoic acid for cervical dysplasia; and with 13-cis
retinoic acid for secondary primary tumors in patients with head and neck cancer (103).
Responses to retinoids were poor or unconvincing in a large number of other cancers.
Responses to retinoids tended understandably to be better in diseases that were not too
far advanced. The major problem encountered was the toxicity caused by the most efficacious
doses (103).
In only one disease, acute promyelocytic leukemia (APL), is the mechanism of action
of retinoic acid fairly well defined (30,75,103). In essence, the RAR? gene and the
PML gene form a hybrid because of a chromosomal translocation. Presumably higher
concentrations of retinoic acid are therefore required to activate the modified RAR? transcription
factor and to induce cellular differentiation. After several months, resistance to
retinoic acid treatment develops, which may be due to an increased rate of retinoic acid
metabolism, increased sequestration of retinoic acid by induced CRABP, or reduced permeability
of the cell to retinoic acid. Thus, the treatment of APL by retinoic acid, which
initially was heralded with enthusiasm, is clearly less effective as long-term therapy than
initially hoped.
Cancer is such a complex and variegated group of diseases that generalizations about
possible mechanisms of action of retinoic acid have not been very fruitful. Nonetheless,
retinoids probably act by stimulating the differentiation of precancerous stem cells,
whereas carotenoids probably are involved in a network of antioxidants that include vitamin
E, vitamin C, sulfhydryl groups, and a variety of other enzymatic and nonenzymatic
D. Other Diseases
Retinoids and carotenoids have been implicated as protective agents in a variety of other
diseases, including heart disease, cortical cataract, and age-related macular degeneration
(104,105). Epidemiological and intervention studies with carotenoids are considered later.
Three types of toxicity to retinoids exist: acute, chronic, and teratogenic (106–111). Acute
toxicity is due to a single dose or a limited number of large doses taken during a short
32 Olson
period, whereas chronic toxicity is caused by moderately high doses taken frequently,
usually daily, over a span of months or years. Teratogenicity is caused by the ingestion
of moderate to high doses during the first trimester of pregnancy in humans and at selected
‘‘sensitive’’ periods during gestation in animals.
A. Acute Vitamin A Toxicity
When a single dose of more than 0.7 mmol of vitamin A (200 mg, or 660,000 IUa)
is ingested by adults or when a dose larger than 0.35 mmol (100 mg or 330,000 IUa)
is ingested by children, nausea, vomiting, headache, increased cerebrospinal fluid pressure,
vertigo, blurred (double) vision, muscular incoordination, and (in infants) bulging of the
fontanelle may occur (39,106). Some infants can be adversely affected by single doses
of only 0.1 mmol. These signs are generally transient and subside within 1 to 2 days.
When the dose is extremely large, drowsiness, malaise, inappetence, reduced physical
activity, skin exfoliation, itching around the eyes, and recurrent vomiting soon follow.
Finally, when lethal doses are given to young monkeys, an excellent model for vitamin
A toxicity in the human, the animals have deepening coma, convulsions, and respiratory
irregularities, and they finally die of either respiratory failure or convulsions (107). The
median lethal dose (LD50 value) of vitamin A injected intramuscularly in a water-miscible
form in the young monkey is 0.6 mmol (168 mg) retinol per kilogram body weight. Extrapolated
to a 3-kg child and a 70-kg adult, the total LD50 dose would be 1.8 mmol (500
mg) and 41 mmol (11.8 g), respectively. A newborn child, who mistakenly was given
0.09 mmol (25 mg) daily, or 28 µmol/kg for 11 days, died of vitamin A toxicity (108).
The total dose received was 0.31 mmol/kg, 50% of the LD50 value for young monkeys.
Such enormous amounts of vitamin A are present only in high-potency preparations of
vitamin A or in large amounts (500 g) of livers particularly rich in vitamin A (0.035
mmol/g or 10 mg/g) (39).
The LD50 values of all-trans retinyl ester, 13-cis retinoic acid, and all-trans retinoic
acid in the adult rat are 27.7 mmol/kg body weight, 13.3 mmol/kg body weight, and 6.7
mmol/kg body weight, respectively (109). These LD50 values reflect the general observation
that the relative toxicities of these retinoids are all-trans retinoic acid 13-cis retinoic
acid  all-trans retinol and its esters (109,110). Retinoids injected intraperitoneally are
much more toxic than those administered orally, which in turn are more toxic than those
applied topically to the skin. Species differences in toxicity also exist.
B. Chronic Vitamin A Toxicity
Chronic toxicity is induced in humans by the recurrent intake of vitamin A in amounts
at least 10 times the RDA, i.e., 13 µmol (3.75 mg retinol equivalents or 12,500 IUa) for
an infant or 35 µmol (10 mg retinol equivalents or 33,300 IUa) for an adult. A healthfood
enthusiast who ingested 26 µmol (25,000 IUa) of vitamin A as a supplement daily
plus a similar amount in food showed severe signs of toxicity (39,109,110). Approximately
50 signs of chronic toxicity have been reported, of which the most common are alopecia,
ataxia, bone and muscle pain, cheilitis, conjunctivitis, headache, hepatotoxicity, hyperlipemia,
hyperostosis, membrane dryness, pruritis, pseudotumor cerebri, various skin disorders,
and visual impairment (39,109,110). Particular attention has recently been paid to
serious adverse effects on the liver caused by daily doses of 26–52 µmol (25,000–50,000
IUa) of vitamin A (111,112). When the supplemental intake of vitamin A is eliminated,
these signs usually, but not always, disappear over a period of weeks to months (39).
Vitamin A 33
In chronic hypervitaminosis A, holo-RBP in the plasma is not much elevated,
whereas retinyl esters are usually increased markedly. Factors that enhance toxicity include
alcohol ingestion, low protein intake, viral hepatitis, other diseases of the liver and kidney,
and possibly tetracycline use. Elderly individuals may be more sensitive because of a
slower rate of storage in the liver and a reduced plasma clearance of administered vitamin
A. Tocopherol, taurine, and zinc are protective in tissue culture cells, but they may or
may not be effective in vivo (39,110).
Some individuals seem to suffer from vitamin A intolerance, i.e., the appearance of
signs of toxicity upon routinely ingesting moderate amounts of vitamin A. This relatively
rare condition, which seems to be genetic, mainly affects males (38,39,110).
C. Teratogenicity
Vitamin A and other retinoids are powerful teratogens both in experimental animals and
in women (39,109,110,113). A single extremely large dose, exposure for as short as a
week on high daily doses (0.1–0.3 mmol, or 30–90 mg), or long-term daily intakes of
26 µmol (25,000 IUa or 7500 retinol equivalents) of vitamin A during early pregnancy
can induce spontaneous abortions or major fetal malformations. Common defects are craniofacial
abnormalities, including microcephaly, microtia, and harelip; congenital heart
disease; kidney defects; thymic abnormalities; and central nervous system disorders
Permanent learning disabilities have been noted in otherwise normal rat pups whose
dams received nonteratogenic doses of vitamin A (114), and similar effects have been
noted in children whose mothers received large doses of 13-cis retinoic acid (115,116).
Most of the children with severely impaired intelligence quotients in this study, however,
also had major physical malformations. In general, the doses that induce demonstrable
learning disabilities in rats are approximately 40% (range 10–70%) of those that cause
terata (114). Many procedures have been employed to measure behavior, and usually the
measurement has been conducted within weeks of birth. Whether the observed effects
become worse or are ameliorated at a later time is still unclear.
Synergism between vitamin A and other teratogens, such as alcohol and drugs, at
nonteratogenic doses of each is probable. Thus, women who are pregnant, or who might
become so, should carefully control their intake of vitamin A, both in regard to rich food
sources, such as liver, and to vitamin A supplements (39).
D. Relative Toxic Effects of Different Retinoids
Most attention has focused on the relative teratogenic effects of natural and synthetic
retinoids that have been, or might be, used therapeutically. As shown in Table 5, a 10,000-
fold range exists, with TTNPB as the most teratogenic and retinoyl ?-glucuronide as the
least (117). The reasons for this wide range of teratogenic activities are only partly understood.
All-trans retinoyl ?-glucuronide, for example, when given in large doses, seems
to be absorbed from the intestinal tract less well than all-trans retinoic acid, is transferred
across the placenta much less well than all-trans retinoic acid, is converted to a significant
extent to the less toxic 13-cis isomer, is taken up by the embryo less well than all-trans
retinoic acid, and is inherently much less teratogenic than all-trans retinoic acid (118).
Indeed, the formation of glucuronides of retinoids may well be a protective metabolic
process to prevent toxic effects (113). Isomerization may also serve a protective action,
e.g., 9-cis retinoic acid is rapidly converted to the much less active 9,13-cis retinoic acid
34 Olson
Table 5 Relative Teratogenicity of Retinoids in Hamstersa
Retinoid TDb
50 (mg/kg)
TTNPB 0.02
Etretinate 6
All-trans retinoic acid 11
13-cis retinoic acid 22
All-trans retinal 23
All-trans retinol 23
All-trans hydroxyphenylretinamide 139
All-trans retinoyl ?-glucuronide 200c
aSelected data from Refs. 117 and 118.
bTeratogenic dose to the dam that induced malformations in 50% of the
cExtrapolated value.
during pregnancy (119,120). Nonetheless, 9,13-cis retinoic acid might possibly play some
functional role in pregnancy (120).
TTNPB, as a stable aromatic molecule, might be metabolized less rapidly than retinoic
acid. Thus, its high activity both physiologically and as a teratogen might be attributed
to a lack of its conversion to less active products. Inasmuch as the rate of glucuronidation
of TTNPB and of all-trans retinoic acid by rat liver microsomes is similar (G. Genchi
and J. A. Olson, unpublished observations), some other metabolic step might be affected.
E. Mechanisms of Toxic Action
Acute toxicity is probably caused by the presence of significant amounts of ‘‘free’’ retinol
and retinyl ester in blood and tissues. Retinol bound to RBP is much less cytotoxic than
unbound retinol. Similarly, the activity of all-trans retinoic acid in cell differentiation is
reduced by the presence of large amounts of CRABP in the cells (30). When large doses
of retinol are given to monkeys, the amount of retinol in the brain markedly increases
(107), consistent with the many signs of central nervous system involvement in acute
Chronic toxicity and teratogenicity seem to involve primarily the nuclear retinoid
receptors (80,109,113). The teratogenic effects of AM-580 (Fig. 3g), a RAR?-specific
agonist, in pregnant mice are offset by a sulfur-containing analog of TTNBP that serves
as a RAR? antagonist (109). As already mentioned, retinoic acid activates via its receptors
several families of Hox genes that are closely implicated with specific developmental
structures. Thus, the concentration of retinoic acid at various locations in the embryo must
be very carefully regulated so that a proper spacial and temporal sequence of activations
can occur. By flooding the embryo with retinoic acid, regulation will be lost. Thus, in
early mouse embryos treated with all-trans retinoic acid, Hox a-1 (Hox 1.6) expression
at 7.5 d is much more extensive and less localized than in control embryos (109).
Other possible actions of retinoids include effects on cell proliferation, cell differentiation,
cell migration, programmed cell death (apoptosis), and membrane integrity. Many
of these effects also involve known retinoid receptors, but others do not (109,113).
Vitamin A 35
The structural characteristics of highly teratogenic retinoids comprise the following
1. A polar terminus with an acidic pKa
2. A lipophilic polyene side chain with good ? electron delocalization, e.g., cis
isomers are less active than the all-trans isomer
3. A fairly lipophilic ring with no set nature or dimensions opposite the polar
terminus, and
4. Conformational restriction (for acidic retinoids only)
Interestingly, the two least teratogenic compounds, 4-hydroxyphenylretinamide
(Fig. 3d) and retinoyl ?-glucuronide (Fig. 1j) are conjugated, more polar molecules.
F. Safe Levels of Vitamin A Intake in Humans
This issue of crucial public health importance can be approached in two ways: (a) by
assessing the no-effect level of vitamin A intake experimentally in various populations
at risk and (b) by reviewing the formal recommendations made by expert groups. The
major groups at risk are infants, young children, pregnant women, and lactating women.
The issue is complicated in part by whether or not a given population will clearly benefit
as well as be put at-risk by the ingestion of a large dose of vitamin A. In the following
discussion, however, only the issue of toxicity will be considered.
Statistical issues are also crucial. It clearly is easier to define an appropriate maximal
intake for a population in which an allowable percentage of that group, e.g., 1%, 3%, or
5%, is adversely affected than to specify an intake that presumably will affect nobody.
As already mentioned, genetic vitamin A intolerance is known to exist in a handful of
persons (38,110,121). However, recommendations for a healthy population certainly
should not be based on the adverse reactions of a few hypersensitive individuals. Finally,
some relatively nonspecific reactions to moderate doses might not be causally related to
the vitamin A in the dose.
Nonetheless, despite these caveats, some useful guidelines can be defined for vitamin
A intake. RDIs of vitamin A are completely safe insofar as we know. RDA values for
infants, children 1–6 years, and pregnant women are 375, 400–500, and 800 µg retinol
equivalents, respectively (35); (Table 2). These values, however, assume that intakes are a
mixture of preformed vitamin A and corotenoids. Thus, the estimated amount of preformed
vitamin A in the diet of infants, children 1–6 years, and pregnant women are approximately
340 µg (1133 IUa), 300–375 µg (1000–1250 IUa), and 600 µg (2000 IUa), respectively.
1. Pregnant Women
Although additional vitamin A is not usually needed by healthy pregnant women, daily
multivitamin tablets containing 5000 IUa (1500 µg) of vitamin A are ingested by large
numbers of women. Toxic reactions to their ingestion have not been reported.
Toxic reactions have been noted, however, in women, both pregnant and nonpregnant,
ingesting daily doses of 18,000 IUa (5400 µg) (106,110–113), whereas only
sporadic claims, often poorly documented, have been made for toxic effects at lower daily
doses. In a study of retinoid embryopathy in pregnant women, the lowest dose of 13-cis
retinoic acid that was associated with birth defects was 10 mg/day (122). Thus, 1500 µg
of vitamin A seems safe for essentially all pregnant women, whereas 5400 µg probably
is not.
36 Olson
Responding to concerns about the teratogenicity of vitamin A, the Teratology Society
recommended that women who might become pregnant limit their daily intake of
preformed vitamin A to 8000 IUa (2400 µg) and ingest provitamin A carotenoids as a
primary source of dietary vitamin A (123). Furthermore, they recommended that the unit
dose of commercially available vitamin A be limited to 5000–8000 IUa and that the hazards
of excessive intakes of vitamin A be indicated on the labels of such products. Similarly,
the American Institute of Nutrition, the American Society for Clinical Nutrition,
and the American Dietetic Association issued a joint formal statement that supplements of
vitamins and minerals were not needed by well-nourished, healthy individuals, including
pregnant women, except in some specific instances (124). The Council for Responsible
Nutrition, a group sponsored by industry, has also advised that pregnant women, while
needing to ensure an adequate intake of vitamin A, should prudently limit their intake
of nutritional supplements of vitamin A to 5000–10,000 IUa (125). Subsequently, they
recommended that the unit dosage of retinol in commercial vitamin A preparations be
limited to 10,000 IUa.
In recognition of the fact that a deficiency of vitamin A in the mother can also
cause abortion and fetal abnormalities, the International Vitamin A Consultative Group
(IVACG) has recommended that the average daily diet of pregnant women should supply
620 µg retinol equivalents, in keeping with FAO/WHO recommendations (34). However,
in areas of the world where this level of intake does not occur and little opportunity exists
for dietary improvement, or in emergency situations in which food supplies are disrupted,
IVACG recommends that daily supplements of 3000 µg retinol equivalents (10,000 IUa)
can be given safely anytime during pregnancy. They do not suggest, by the way, that
well-nourished women take supplements.
To summarize, well-nourished healthy women of reproductive potential should include
carotenoid-rich fruits and vegetables in their diet. They should also avoid taking
supplements of preformed vitamin A during the first trimester of pregnancy, during which
increased nutritional demands are small and the risk of fetal abnormalities is high. If
supplements of vitamin A are subsequently taken, the daily dose should be carefully limited
to 5000–10,000 IUa.
2. Infants and Young Children
As already indicated, the probable daily amounts of preformed vitamin A in recommended
diets of infants and young children are 1133 IUa and 1250 IUa, respectively. As a general
public health measure, oral doses (200,000 IUa) of all-trans retinyl palmitate in oil have
been administered one to three times a year to preschool children, usually 1–6 years of
age, in less industrialized countries (95). Side effects, e.g., nausea, vomiting, and bulging
of the fontanelle in infants, have usually been reported in 5% of the treated children
and have been transient (1–3 days) in nature. Because vitamin A toxicity is a function
of weight, younger children tend to be most affected.
Because of international public health interest in combining vitamin A supplementation
with the expanded program of immunization (EPI), the effects of dosing infants with
vitamin A from 6 weeks to 9 months of age has been explored. When 50,000 IUa of
vitamin A was given orally in oil to Bangladesh infants at 6, 11, and 16 weeks of age,
11% showed transient bulging of the fontanelle. In a subsequent study in which 25,000
IUa was given orally at approximately the same three times to Bangladesh infants, 8%,
corrected for the 2.5% incidence in the placebo group, showed the same effect (126). In
Vitamin A 37
this study, nearly all of the affected children had received three doses of vitamin A; none
showed any toxicity after one dose. If we assume that the infants’ daily intake from breast
milk was approximately 100 µg (333 µg IUa), the calculated mean daily increment from
the dose is only an additional 715 IUa, giving a total of 1048 IUa, essentially the same
as the recommended daily intake. A single oral dose of 50,000 IUa given to Indonesian
infants showed much less toxicity, i.e., 4.5% in the vitamin A–treated group and 2.5%
in the control group (127). Furthermore, the cerebral fluid volume, but not the pressure,
was transiently increased in these infants, and no lasting side effects were noted. Bulging
of the fontanelle, in consequence, may be more of a transient physiological response to
a dose of this magnitude than an indicator of toxicity.
As yet, the relationship between an acceptably safe dose and the age of infants has
not been defined. Quite possibly, infants suffering from inadequate intakes of protein and
calories may be more susceptible to vitamin A toxicity than better nourished infants. If
so, the dose of vitamin A that is selected for a given country might well be based on
anthropometric indices within that country.
Thus, a generally safe single dose for most infants, although currently undefined,
probably will fall in the range of 4000–5000 IUa per kilogram body weight with an interval
between doses of 5 weeks.
Unlike retinoids, including vitamin A, carotenoids are generally nontoxic. However, individuals
who routinely ingest large amounts of carotenoids, either in tomato or carrot juice
or in commercial supplements of ?-carotene, can develop hypercarotenosis, characterized
by a yellowish coloration of the skin and a very high concentration of carotenoids in the
plasma. This benign condition, although resembling jaundice, gradually disappears upon
correcting the excessive intake of carotenoids. The only known toxic manifestation of
carotenoid intake is canthaxanthin retinopathy, which can develop in patients with erythropoietic
porphyria and related disorders who are treated with large daily doses (50–100
mg) of canthaxanthin, the 4,4?-diketo derivative of ?-carotene, for long periods (128). In
most instances, however, these deposits of canthaxanthin disappear slowly upon termination
of treatment (128). Canthaxanthin-containing supplements are not currently available
in the United States. ?-Carotene at similar doses is not known to cause retinopathy. Carotenoids,
even when ingested in large amounts, are not known to cause birth defects or
hypervitaminosis A, primarily because the efficiency of their absorption from the intestine
falls rapidly as the dose increases and because their conversion to vitamin A is not suffi-
ciently rapid to induce toxicity (39).
Quite apart from their function as precursors of vitamin A, carotenoids are distributed
widely in mammalian tissues, can quench singlet oxygen, can serve as an antioxidant
in tissues (particularly under conditions of low oxygen tension), and can stimulate the
immune response (39,104).
Thus, by using provitamin A activity as the nutritional function and singlet oxygenquenching
and antioxidant activity as the biological action, four classes of carotenoids
might be defined: those that are both nutritionally and biologically active, such as ?-
carotene; those that are nutritionally active and biologically inactive, such as 14?-?-apocarotenal;
those that are nutritionally inactive but biologically active, such as lycopene and
violaxanthin; and those that are both nutritionally and biologically inactive, such as phytoene.
38 Olson
Because over 90% of the 600 characterized carotenoids in nature are not precursors
of vitamin A, their biological effects in mammalian physiology, independent of their provitamin
A activity, are being followed with interest.
A. Retinoids
The requirements for vitamin A and safe levels of intake have already been discussed. A
positive response to chemopreventive treatment with large doses of vitamin A has been
shown in leukoplakia and actinic keratosis (103). The retinoids most commonly used for
therapy and chemoprevention, however, are 13-cis retinoic acid, all-trans retinoic acid,
hydroxyphenylretinamide, etretinate, and acitretin (103). As indicated earlier, these agents
have shown promising results in the prevention or treatment of some carcinomas (103).
The major drawbacks in their use are their toxicity at highly efficacious doses and, in the
case of APL, a rapidly developing resistance to the drug. Of the retinoids listed, all-trans
retinoic acid and etretinate, because of the latter’s slow turnover in the body, are the
most toxic and hydroxyphenylretinamide the least. The search consequently continues to
identify new retinoids with high efficacy but low toxicity. Retinoyl ?-glucuronide, a naturally
occurring metabolite of retinoic acid, shows these properties (99,118). It is active in
treating acne but has not been tested against other diseases. In epidemiological surveys,
the dietary intake of preformed vitamin A and the plasma concentration of retinol are
rarely associated with a reduced incidence of chronic diseases.
B. Carotenoids
1. Cancer
One of the most dramatic and consistent observations in epidemiological studies is the
inverse association between ?-carotene intake and the incidence of lung cancer (104,129).
These findings have stimulated intervention trials in two high-risk groups: asbestos workers
in Tyler, Texas and middle-aged male smokers in Finland. The results of these intervention
trials are disappointing. In asbestos workers, no differences in the prevalence of sputum
atypia was noted between treated (50 mg ?-carotene  25,000 IUa retinol every other
day) and control groups over a 5-year period. In the Finnish study, the group treated daily
with ?-carotene (20 mg) showed a significantly higher incidence of lung cancer [relative
risk (RR)  1.18, 95% confidence interval (CI)  1.03–1.36] and total mortality (RR 
1.08, 95% CI  1.01–1.16) than did the placebo group. Supplemental ?-carotene did not
affect the incidence of other major cancers found in this population (129).
The unexpected negative finding in the Finnish study has several possible explanations:
(a) Supplemental ?-carotene is interfering with the intestinal absorption of other
possible chemopreventive nutrients. In this regard, ?-carotene inhibits the absorption in
humans of lutein, which shows good antioxidant activity (42). In that same vein, ?-carotene,
which shows chemopreventive properties, might be similarly affected. (b) Supplemental
?-carotene may be serving as a pro-oxidant in the well-oxygenated ambient of the
lung (130). (c) The population of middle-aged male smokers is not representative of other
groups, who might well benefit from a higher intake of carotenoids. (d) A comparison
between treated and control subjects that fall only in the lowest quartile of initial plasma
Vitamin A 39
carotene values might yield different results. (e) Vitamin C, which is low in the plasma
of most Finns, may have played some role in the outcome. (f) Alcohol intake may also
play an important role in the outcome.
Several other trials have provided results that support the findings in the Finnish
study, namely, the Carotene and Retinal Efficacy Trial (CARET) and the Physician’s
Health Study (129).
The development of head and neck cancers, including those of the oral cavity, pharynx,
and larynx, are influenced by many factors, including smoking, other uses of tobacco,
alcohol, and diet (129). Serum carotene concentrations, adjusted for smoking, are inversely
related to the incidence of these carcinomas. Supplements of ?-carotene can markedly
reduce leukoplakia, although the lesion returns upon cessation of treatment (129). A chemoprevention
trial designed to assess the effect of daily supplements of ?-carotene (50 mg)
on the recurrence of head and neck cancer is currently under way (129).
The effects of various nutrient combinations on esophageal cancer and on stomach
cancer was evaluated in Linxian, China, where the incidence of esophageal cancer is 100-
fold higher than in the United States (129). Of four nutrient treatments, only one, involving
supplements of ?-carotene, selenium, and ?-tocopherol, showed a positive effect: the reductions
in total deaths, cancer deaths, esophageal cancer deaths, and gastric cancer deaths
were 9% (RR  0.91, 95% CI  0.84–0.99), 13% (RR  0.87, 95% CI  0.75–1.00),
4% (RR  0.96, 95% CI  0.78–1.18), and 21% (RR  0.79, 95% CI  0.64–0.99),
respectively (129). Although these results support the concept that diet influences cancer
incidence, the general nutritional status of the population was poor. Thus, whether the
mixed supplement, or one component of it, was protective as a result of generally improved
health or of a more specific anticancer effect is not clear (129).
The dietary intake and serum concentrations of carotenoids are often inversely associated
with the risk of colorectal cancer (129). However, by using adenomas as an indicator,
supplemental ?-carotene (25 mg/day) was found to be ineffective (RR  1.01, 95%
CI  0.85–1.20) in preventing the recurrence of this lesion (129).
?-Carotene intake has been associated with an improved survival rate in breast cancer
patients (129). Whether supplements of carotenoids reduce the incidence of breast
cancer in a well-designed clinical trial is not known. In an ongoing trial, the effect of
hydroxyphenylretinamide on the recurrence of breast cancer is being explored (103).
The risk of cervical cancer has been correlated with the prediagnostic serum levels
of ?-, ?-, and total carotenoids (RR  2.7–3.1, 95% CI  1.1–	8.1) (129). On the
other hand, invasive cervical cancer among white women in the United States did not
relate to any specific food group of the diet or to the use of supplements of vitamins A,
C, and E and folic acid (131). However, cervical dysplasia, considered to be a precancerous
lesion, did respond to ?-carotene supplements (30 mg/day) (129).
The recurrence of skin cancer was not affected by ?-carotene supplements (50 mg/
day) over a 5-year period (RR 1.05, 95% CI  0.91–1.22) (129).
Thus, a dichotomy exists. Most of the associations found between diseases in dietary
or plasma level studies do not agree with the results of intervention trials. The former
tend to show strong significant correlations and the latter, in large part, do not. Possible
explanations are as follows: (a) ?-Carotene, which is only one of approximately 600 known
carotenoids, might not be the most active one, or indeed, might inhibit the absorption of
other more chemopreventive carotenoids and other nutrients. (b) Carotenoids might be
only one of a group of chemopreventive agents in foods that act synergistically in preventing
carcinogenesis. The fact that the relative risk values for colored fruits and vegeta40
bles usually are less (more protective) than those for carotenoids or for any other component
of the food supports this viewpoint. (c) Carotenoids may serve solely as a useful
marker for a healthful lifestyle. (d) The preventive action of carotenoids might occur very
early in disease progression but be ineffective later. Thus, subjects in identified high-risk
groups, who often have had a primary tumor, may be resistant to nutritional supplements.
(e) The associations found in observational epidemiology are not causal and can be confounded
by a variety of unanticipated and unmeasured factors. In essence, the intervention
trials may well be providing more valid answers (129).
2. Photosensitivity Disorders
Patients with erythropoietic porphyria and similar diseases benefit by ingesting supplements
(180 mg/day) of ?-carotene. Canthaxanthin, though also protective, is no longer
used because of the reversible retinopathy that results (128). Although the concentrations
of ?-carotene and vitamin A are elevated in the livers of these patients, the side effects
of ?-carotene ingestion over a period of years are minimal (129).
3. Cardiovascular Disease
Epidemiologic studies suggest protective effects of carotenoid intake against both coronary
events and stroke (104,129). In a European study (WHO/MONICA, i.e., monitoring cardiovascular
disease), mortality from ischemic heart disease correlated inversely with serum
vitamin E concentrations (r2  0.63) but not with ?-carotene levels (r2  0.04) (132).
If the 3 Finnish sites, which were outliers, of the 16 examined were excluded, however,
the inverse correlation with ?-carotene concentrations improved markedly (r2  0.50). A
mean serum ?-carotene concentration in populations of 0.4 µmol/L or higher was associated
with good health in the European studies, whereas a concentration 0.25 µmol/L
in populations was related to an increased risk of coronary disease, stroke, and cancer
(132). The risk of myocardial infarction was inversely related to adipose ?-carotene content
in smokers (RR  2.62, 95% CI  1.79–3.83) but not in nonsmokers (RR  1.07)
(129). Furthermore, physicians with stable angina or prior coronary revascularizations,
who were supplemented with ?-carotene for 5 years, showed a 51% reduction in the risk
of major coronary events (129). ?-Carotene did not show beneficial effects, however,
in the total population enrolled in the Physicians Health Study (129). The incidence of
cardiovascular deaths in the Finnish lung cancer study also was not affected by ?-carotene
The overall results, therefore, are somewhat mixed. The most likely mechanism of
action of carotenoids, but by no means the only one, is a reduction in the oxidation of
low-density lipoproteins, which seem to play a key role in atherogenesis (104). Of various
antioxidants studied both in vivo and in vitro, however, ?-carotene does not seem to be
very protective, if at all. Thus, the relationship among dietary intakes of carotenoids, their
plasma and tissue concentrations, and cardiovascular disease remains unclear.
4. Age-Related Macular Degeneration
The macular of the eye predominantly contains two pigments, lutein and zeaxanthin
(133,134). Because these two pigments account for less than 25% of plasma carotenoids,
their uptake from plasma and deposition in the macula show specificity. These pigments
might consequently play a role in protecting the macula from damage caused by light and
particularly by blue light. In a recent study with patients suffering from age-related macular
degeneration (ARMD) vs. matched controls, subjects in the highest quintile of carotVitamin
A 41
enoid intake had a 43% lower risk (RR  0.57, 95% CI  0.35–0.92) of suffering from
ARMD than those in the lowest quintile (129). Of various carotenoids, the intake of spinach
and collard greens, rich in these two carotenoids, was most strongly associated with
reduced risk. However, not all studies support this finding (105). Nonetheless, higher
plasma concentrations of lutein and zeaxanthin as well as ?-carotene showed a significant
trend toward a lower risk of developing ARMD. These interesting findings are currently
being investigated.
5. Senile Cataract
Cataract consists of gradual opacification of the lens with aging, which may in part result
from oxidative stress. Carotenoid intake, as well as that of vitamins C and E, has been
associated with a reduced risk of cataract (105,129). In the main Linxian, China trial,
however, combined supplements of ?-carotene, selenium, and ?-tocopherol were not associated
with a reduction in the incidence of cataracts, and inconclusive results have been
reported by others (105,129). Thus, whereas the concept that antioxidant nutrients might
prevent oxidative damage to a fairly exposed structure, such as the lens, is highly feasible,
the data supporting a protective role of dietary components in the process are mixed.
6. HIV Infection
In HIV infection T-helper (CD4) cells are destroyed, thereby impairing the immune response.
In humans as well as in experimental animals, both ?-carotene, which is a provitamin
A carotenoid, and canthaxanthin, which is not, can enhance the immune response
(135). Indeed, in HIV-infected patients, large doses of ?-carotene increased the CD4/CD8
ratio, which is usually depressed in HIV infection, and improved the response to vaccines
(135). UV light tends both to activate human HIV expression, at least in transgenic mice,
and to reduce plasma carotenoid concentrations in humans. In phase II HIV-infected subjects,
plasma carotenoid concentrations are reduced by 50%. AIDS patients treated daily
with a combination of ?-carotene supplementation (120 mg) and whole-body hyperthermia
(42°C, 1 h) showed a better and longer lasting response than either treatment separately
(136). Thus, carotenoids seem to ameliorate the condition of AIDS patients, probably, at
least in part, by enhancing the immune response.
A quite different effect of a carotenoid has also been reported, namely, that halocynthiaxanthin
one) strongly and rather specifically inhibits RNA-dependent DNA polymerase of the HIV
virus (135). The use of carotenoids in the treatment of subjects with HIV infections clearly
merits further attention.
Revising a chapter that was written approximately 9 years ago (137) is a valuable but
humbling experience. The facts cited earlier have not changed, and many of the concepts
have been modified only slightly; however, interests have shifted markedly and a whole
new body of information and hypotheses—some quite clear, some conflicting—has arisen.
Old observations are viewed in new ways, and a completely new set of research questions
are being asked. Thus, in revising the chapter, the addition of a paragraph here and a
reference there was just not feasible. As a consequence, the chapter is largely rewritten,
and the reference list is in large part new. This new chapter and the previous one are,
therefore, complementary to each other.
42 Olson
A major recent advance has been the discovery of the nuclear retinoid receptors and
their impact on embryogenesis, cell differentiation, disease, and pharmacology. In truth,
the paradigm of interpreting vitamin A actions has markedly changed. A new chemistry
has arisen, with the focus on finding compounds that serve as specific agonists and antagonists
for given retinoid receptors. The linkage between nutrition and molecular biology,
seemingly such diverse fields, has been strengthened by the observation that the RXR
receptors for vitamin A interact meaningfully with the thyroid receptor, dependent on
iodine for its activity, and with the vitamin D receptor, dependent of course on another
fat-soluble vitamin. The hint (not the demonstration) that ?-tocopherol may also have
specific nuclear effects adds further interest to this linkage.
Diet, admittedly along with a variety of other factors, is known to affect the onset
and possibly the severity of major chronic diseases. In the past several years, an explosion
of information about specific nutrients that may play roles in these processes has appeared.
Despite the great care with which most of these studies have been done, these surveys
have inherent constraints. Thus, the findings in large part have tantalized us rather than
presented a coherent picture. Carotenoids, primarily together with vitamin E, have played
a central role in these surveys.
Although these studies have great potential impact, some biases have arisen: namely,
that carotenoids, vitamin A, vitamin E, vitamin C, and selenium all act solely as antioxidants
in these processes. Indeed, the term ‘‘antioxidant vitamins’’ has become common
parlance. To stimulate a broader, less constrained view of their potential actions, these
nutrients, as well as many other naturally occurring compounds, both of endogenous as
well as of dietary origin, might better be called ‘‘physiological modulators’’ (138). By so
doing, the mechanism of action is not automatically inferred from the outset for whatever
beneficial or adverse effect that they might show (138,139).
To keep the reference list within bounds, references are largely made to reviews,
which in turn can serve as a guide to the primary research literature. Recent reviews,
monographs, and articles of particular interest are cited in references 140–162. I regret
the necessary omission of many specific research papers that have enriched our knowledge
in this dynamic field.
This study has been supported in part by grants from the National Institutes of Health
(DK-39733), the USDA (NRICGP 94-37200-0490 and ISU/CDFIN/CSRS 94-34115-
2835), and the W. S. Martin Fund. This is Journal Paper J-16524 of the Iowa Agriculture
and Home Economics Experiment Station, Ames, IA (Iowa Project No. 3335). The author
is indebted to Ms. Margaret Haaland for outstanding administrative and secretarial assistance.
1. G. Wolf, Historical note on mode of administration of vitamin A for cure of night blindness,
Am. J. Clin. Nutr. 31:290–292 (1978).
2. T. Moore, Vitamin A, Elsevier, Amsterdam, 1957.
3. Nomenclature policy: Generic descriptors and trivial names for vitamins and related compounds,
J. Nutr. 120:12–19 (1990).
4. A. E. Asato, A. Kim, M. Denny, and R. S. H. Liu, 7-cis, 9-cis, 11-cis Retinal, all-cis vitamin
Vitamin A 43
A, and 7-cis, 9-cis, 11-cis, 12-fluororetinal. New geometric isomers of vitamin A and carotenoids.
J. Am. Chem. Soc. 105:2923–2924 (1983).
5. R. S. H. Liu and A. E. Asato, Photochemistry and synthesis of stereoisomers of vitamin A.
Tetrahedron 40:1931–1969 (1984).
6. C. G. Knudsen, S. C. Carey, and W. H. Okamura, [1,5]-Sigmatropic rearrangement of vinyl
allenes: a novel route to geometric isomers of the retinoids possessing 11-cis linkages including
9-cis, 11-cis, 13-cis retinal. J. Am. Chem. Soc. 102:6355–6356 (1980).
7. O. Isler (ed.), Carotenoids, Birkha?user Verlag, Basel, 1971.
8. G. Britton, S. Liaaen-Jensen, and H. Pfander (eds.), Carotenoids, Vols. 1A and 1B, Birkha
?user Verlag, Basel, 1995.
9. M. I. Dawson and P. D. Hobbs, The synthetic chemistry of retinoids, in The Retinoids: Biology,
Chemistry, and Medicine, 2nd ed. (M. B. Sporn, A. B. Roberts, and D. S. Goodman,
eds.), Raven Press, New York, 1994, pp. 5–178.
10. W. Bollag and A. Matter, From vitamin A to retinoids in experimental and clinical oncology:
achievements, failures, and outlook, Ann. N.Y. Acad. Sci. 359:9–23 (1981).
11. A. Stephens-Jarnegan, D. A. Miller, and H. F. DeLuca, The growth supporting activity of
a retinoidal benzoic acid derivative and 4,4-difluororetinoic acid. Arch. Biochem. Biophys.
237:11–16 (1985).
12. B. A. Bernard, J. M. Bernardon, C. Delescluse, B. Martin, M. C. Lenoir, J. Maignan, B.
Charpentier, W. R. Pilgrim, U. Reichert, and B. Shroot, Identification of synthetic retinoids
with selectivity for human nuclear retinoic acid receptor ?, Biochem. Biophys. Res. Commun.
186:977–983 (1992).
13. J. M. Lehmann, L. Jong, A. Fanjul, J. F. Cameron, X. P. Lu, P. Haefner, M. I. Dawson, and
M. Pfahl, Retinoids selective for retinoid X receptor response pathways. Science 258:1944–
1946 (1992).
14. F. Frickel, Chemistry and physical properties of retinoids, The Retinoids (B. Sporn, A. B.
Roberts, and D. S. Goodman, eds.), Academic Press, Orlando, FL, 1984, pp. 7–145.
15. H. C. Furr, A. B. Barua, and J. A. Olson, Analytical methods, in The Retinoids: Biology,
Chemistry, and Medicine, 2nd ed. (M. B. Sporn, A. B. Roberts, and D. S. Goodman, eds.),
Raven Press, New York, 1994, pp. 179–209.
16. G. M. Landers and J. A. Olson, Rapid simultaneous determination of isomers of retinaldehyde,
retinal oxime, and retinol by high-performance liquid chromatography, J. Chromatogr.
438:383–392 (1988).
17. L. Packer (ed.), Carotenoids Part A, Meth. Enzymol. 213:1–538 (1992).
18. L. Packer (ed.), Carotenoids Part B, Meth. Enzymol. 214:1–468 (1993).
19. A. B. Barua, D. Kostic, and J. A. Olson, New simplified procedures for the extraction and
simultaneous high-performance liquid chromatographic analysis of retinol, tocopherols and
carotenoids in human serum, J. Chromatogr. 617:257–264 (1993).
20. Y. Tanaka, R. Hirata, Y. Minato, Y. Hasumura, and J. Takeuchi, Isolation and higher purifi-
cation of fat-storing cells from rat liver with flow cytometry, in Cells of the Hepatic Sinusoids,
Vol. 1 (A. Kien, D. L. Knook, and E. Wisse, eds.), The Kupffer Cell Foundation, Rijswijk,
The Netherlands, 1986, pp. 473–477.
21. H. C. Furr, O. Amedee-Manesme, A. J. Clifford, H. R. Bergen III, A. D. Jones, D. P. Anderson,
and J. A. Olson, Relationship between liver vitamin A concentration determined by
isotope dilution assay with tetradeuterated vitamin A and by biopsy in generally healthy adult
humans. Am. J. Clin. Nutr. 49:713–716 (1989).
22. A. J. Clifford, A. D. Jones, and H. C. Furr. Stable isotope dilution mass spectrometry to
assess vitamin A status, Meth. Enzymol. 189:94–104 (1990).
23. R. S. Parker, J. E. Swanson, B. Marmor, K. J. Goodman, A. B. Spielman, J. T. Brenna, S.
M. Viereck, and W. K. Canfield, Study of ?-carotene metabolism in humans using 13C-?-
carotene and high precision isotope ratio mass spectrometry, Ann. N.Y. Acad. Sci. 691:86–
95 (1993).
44 Olson
24. R. S. Parker, Absorption, metabolism and transport of carotenoids, FASEB J. 10:542–551
25. B. A. Underwood and J. A. Olson (eds.), A Brief Guide to Current Methods of Assessing
Vitamin A Status, Int. Vitamin A Consult. Group, ILSI Press, Washington, D.C., 1993.
26. P. L. Harris, Bioassay of vitamin A compounds, Vitam. Horm. (U.S.) 18:341–370 (1960).
27. M. B. Sporn, N. M. Dunlop, D. L. Newton, and W. R. Henderson, Relationship between
structure and activity of retinoids, Nature 263:110–113 (1976).
28. A. M. Jetten, Induction of differentiation of embryonal carcinoma cells by retinoids, in Retinoids
and Cell Differentiation (M. I. Sherman, ed.), CRC Press, Boca Raton, FL, 1986,
pp. 105–136.
29. M. Imaizumi and T. R. Breitman, Retinoic acid-induced differentiation of the human
promyelocytic leukemia cell line HL-60 and fresh human leukemia cells in primary
culture, a model for differentiation-inducing therapy of leukemia, Eur. J. Haematol. 38:
289–302 (1987).
30. L. J. Gudas, M. B. Sporn, and A. B. Roberts, Cellular biology and biochemistry of the retinoids,
in The Retinoids: Biology, Chemistry, and Medicine, 2nd ed. (M. B. Sporn, A. B.
Roberts, and DeWitt S. Goodman, eds.), Raven Press, New York, 1994, pp. 443–520.
31. A. R. Mangels, J. M. Holden, G. R. Beecher, M. R. Forman, and E. Lanza, Carotenoid content
of fruits and vegetables: an evaluation of analytic data, J. Am. Diet. Assoc. 93:284–296
32. A. J. Clifford, A. D. Jones, Y. Tondeur, H. C. Furr, H. R. Bergen III, and J. A. Olson,
Assessment of vitamin A status of humans by isotope dilution GC/MS, 34th Annual Conference
on Mass Spectrometry and Allied Topics, Cincinnati, OH, 1986, pp. 327–328.
33. A. S. Truswell, Recommended dietary intakes around the world, Nutr. Abstr. Rev. 53:939–
1015, 1075–1119 (1983).
34. Food and Agriculture Organization/World Health Organization, Requirements of vitamin A,
iron, folate, and vitamin B12. Report of a joint FAO/WHO Expert Committee. FAO Food
and Nutrition Series 23, FAO, Rome, 1989, pp. 1–107.
35. National Research Council, Recommended Dietary Allowances, 10th ed., National Academy
of Sciences Press, Washington, D.C., 1989, pp. 1–284.
36. Department of Health, Dietary reference values for food energy and nutrients for the United
Kingdom, Report No. 41 on Health and Social Subjects, HMSO, London, 1991, pp. 1–210.
37. J. A. Olson, Vitamin A, in Present Knowledge in Nutrition, 7th ed. (E.E. Ziegler and L.J.
Filer Jr, eds.), ILSI Press, Washington, D.C., 1996, pp. 109–119.
38. J. A. Olson, Upper limits of vitamin A in infant formulas, with some comments on vitamin
K, Am. J. Clin. Nutr. 119:1820–1824 (1989).
39. J. A. Olson, Vitamin A, retinoids and carotenoids, in Modern Nutrition in Health and Disease,
8th ed. (M.E. Shils, J.A. Olson, and M. Shike, eds.), Lea & Febiger, Philadelphia, 1994,
pp. 287–307.
40. D. E. Ong, Absorption of vitamin A, in Vitamin A in Health and Disease (R. Blomhoff, ed.),
Marcel Dekker, New York, 1994, pp. 37–72.
41. W. S. White, M. Stacewicz-Sapuntzakis, J. W. Erdman, Jr., and P. E. Bowen, Pharmacokinetics
of ?-carotene and canthaxanthin after ingestion of individual and combined doses by
human subjects, J. Am. Coll. Nutr. 13:665–671 (1994).
42. D. Kostic, W. S. White, and J. A. Olson, Intestinal absorption, serum clearance, and interactions
between lutein and ?-carotene when administered to human adults in separate or combined
oral doses, Am. J. Clin. Nutr. 62:604–610 (1995).
43. J. W. Erdman, G. C. Fahey, and C. B. White, Effects of purified dietary fiber sources on ?-
carotene utilization by the chick, J. Nutr. 116:2415–2423 (1986).
44. G. Brubacher and H. Weiser, The vitamin A activity of beta-carotene, Intern. J. Vitam. Nutr.
Res. 55:5–15 (1985).
45. D. R. Soprano and W. S. Blaner, Plasma retinol-binding protein, in The Retinoids: Biology,
Vitamin A 45
Chemistry, and Medicine, 2nd ed. (M.B. Sporn, A.B. Roberts, and D.S. Goodman, eds.),
Raven Press, New York, 1994, pp. 257–281.
46. M. E. Newcomer, Retinoid-binding proteins: structural determinants important for function,
FASEB J. 9:229–239 (1995).
47. M. H. Green and J. B. Green, Dynamics and control of plasma retinol, in Vitamin A in Health
and Disease (R. Blomhoff, ed.), Marcel Dekker, New York, 1994, pp. 119–133.
48. C. B. Stephenson, J. O. Alvarez, J. Kohatsu, R. Hardmeier, J. I. Kennedy, Jr., and R. B.
Gammon, Jr., Vitamin A is excreted in the urine during acute infection, Am. J. Clin. Nutr.
60:388–392 (1994).
49. P. Tosukhowong and J. A. Olson, The syntheses, biological activity and metabolism of 15-
methyl retinone, 15-methyl retinol and 15-dimethyl retinol in rats, Biochem. Biophys. Acta.
529:438–453 (1978).
50. T. C. M. Wilson and G. A. J. Pitt, 3,4-Didehydroretinol (vitamin A2) has vitamin A activity
in the rat without conversion to retinol, Biochem. Soc. Trans. 14:950–951 (1986).
51. B. Becker, A. B. Barua, M. Barua, and J. A. Olson, Relative effects of various carbohydrate
conjugates of retinoic acid on the differentiation and viability of HL-60 cells, FASEB J. 7:
A304 (1993).
52. R. Blomhoff, Transport and metabolism of vitamin A, Nutr. Rev. 52:13–23 (1994).
53. W. S. Blaner and J. A. Olson, Retinol and retinoic acid metabolism, in The Retinoids: Biology,
Chemistry, and Medicine, 2nd ed. (M. B. Sporn, A. B. Roberts, D. S. Goodman, eds.), Raven
Press, New York, 1994, pp. 229–255.
54. K. Wake, Role of perisinusoidal stellate cells in vitamin A storage, in Vitamin A in Health
and Disease (R. Blomhoff, ed.), Marcel Dekker, New York, 1994, pp. 73–86.
55. D. E. Ong, M. E. Newcomer, and F. Chytil, Cellular retinoid-binding proteins, in The Retinoids:
Biology, Chemistry, and Medicine, 2nd ed. (M. B. Sporn, A. B. Roberts, and D. S.
Goodman, eds.) Raven Press, New York, 1994, pp. 283–317.
56. S. H. Wei, V. Episkopou, R. Piantedosi, S. Maeda, K. Shimada, M. E. Gottesman, and W.
S. Blaner, Studies on the metabolism of retinol and retinol-binding protein in transthyretindeficient
mice produced by homologous recombination, J. Biol. Chem. 270:866–870 (1995).
57. N. Takahashi and T. R. Breitman, Retinoylation of proteins in mammalian cells, in Vitamin
A in Health and Disease (R. Blomhoff, ed.), Marcel Dekker, New York, 1994, pp. 257–
58. K. L. Skare, H. K. Schnoes, and H. F. DeLuca, Biliary metabolites of all-trans retinoic acid:
isolation and identification of a novel polar metabolite, Biochemistry 21:3308–3317 (1982).
59. G. Genchi, W. Wang, A. Barua, W. R. Bidlack, and J. A. Olson, Formation of ?-glucuronides
and of ?-galacturonides of various retinoids catalyzed by induced and noninduced microsomal
UDP-glucuronosyltransferases of rat liver. Biochim. Biophys. Acta 1289:284–290
60. A. B. Barua, R. O. Batres, and J. A. Olson, Synthesis and metabolism of all-trans [11-3H]retinyl
?-glucuronide in rats in vivo, Biochem. J. 252:415–420 (1988).
61. A. Nagao and J. A. Olson, Enzymatic formation of 9-cis, 13-cis, and all-trans retinals from
isomers of ?-carotene, FASEB J. 8:968–973 (1994).
62. X.-D. Wang, N. I. Krinsky, P. N. Benotti, and R. M. Russell, Biosynthesis of 9-cis-retinoic
acid from 9-cis-?-carotene in human intestinal mucosa in vitro, Arch. Biochem. Biophys.
313:150–155 (1994).
63. N. I. Krinsky, X.-D. Wang, G. Tang, and R. M. Russell, Mechanism of carotenoid cleavage
to retinoids, Ann. N.Y. Acad. Sci. 691:167–176 (1993).
64. J. Devery and B. V. Milborrow, ?-Carotene-15,15?-dioxygenase (EC isolation
reaction mechanism and an improved assay procedure, Br. J. Nutr. 72:397–414 (1994).
65. A. Nagao, A. During, C. Hoshino, J. Terao, and J. A. Olson, Stoichiometric conversion of
all-trans ?-carotene to retinal by pig intestinal extract. Arch. Biochem. Biophys. 328:57–63
46 Olson
66. J. L. Napoli, Retinoic acid homeostasis: prospective roles of ?-carotene, retinol, CRBP, and
CRABP, in Vitamin A in Health and Disease (R. Blomhoff, ed.), Marcel Dekker, New York,
1994, pp. 135–188.
67. D. E. Ong, Cellular transport and metabolism of vitamin A: roles of the cellular retinoidbinding
proteins, Nutr. Rev. 52:24–31 (1994).
68. J. C. Saari, Retinoids in photosensitive systems, in The Retinoids: Biology, Chemistry, and
Medicine, 2nd ed. (M. B. Sporn, A. B. Roberts, and D. S. Goodman, eds.), Raven Press,
New York, 1994, pp. 351–385.
69. C. Lampron, C. Rochette-Egly, P. Gorry, P. Dolle, M. Mark, T. Lufkin, M. Lemeur, and
P. Chambon, Mice deficient in cellular retinoic acid-binding protein II (CRABP-II) or
in both CRABP-I and CRABP-II are essentially normal, Development 121:539–548
70. D. Fawcett, P. Pasceri, R. Fraser, M. Colbert, J. Rossant, and V. Giguere, Postaxial polydactyly
in forelimbs of CRABP-II mutant mice, Development 121:671–679 (1995).
71. J. A. Olson, Recommended dietary intakes (RDI) of vitamin A in humans, Am. J. Clin. Nutr.
45:704–716 (1987).
72. R. R. Rando, Isomerization reactions of retinoids in the visual system, Pure Appl. Chem.
66:989–994 (1994).
73. J. E. Dowling and G. Wald, The role of vitamin A acid, Vitam. Horm. (U.S.) 18:515–541
74. A. J. Lamb, P. Apiwatanaporn, and J. A. Olson, Induction of rapid synchronous vitamin A
deficiency in the rat, J. Nutr. 104:1140–1148 (1974).
75. D. J. Mangelsdorf, K. Umesono, and R. M. Evans, The retinoid receptors, in The Retinoids:
Biology, Chemistry, and Medicine, 2nd ed. (M. B. Sporn, A. B. Roberts, and D. S. Goodman,
eds.), Raven Press, New York, 1994, pp. 319–349.
76. D. J. Mangelsdorf, Vitamin A receptors, Nutr. Rev. 52:32–44 (1994).
77. S. M. Pemrick, D. A. Lucas, and J. F. Grippo, The retinoid receptors, Leukemia 8:1797–
1806 (1994).
78. M. Pfahl, R. Apfel, I. Bendik, A. Fanjul, G. Graupner, M.-O. Lee, N. La-Vista, X. P. Lu,
J. Piedrafita, M. A. Ortiz, G. Salbert, and X.-K. Zhang, Nuclear retinoid receptors and their
mechanism of action, Vitam. Horm. (U.S.) 49:327–382 (1994).
79. P. Chambon, The molecular and genetic dissection of the retinoid signaling pathway, Rec
Progr Hormone Res 50:317–332 (1995).
80. L. J. Gudas, Retinoids and vertebrate development, J. Biol. Chem. 269:15399–15402 (1994).
81. A. C. Ross and U. G. Ha?mmerling, Retinoids and the immune system, in The Retinoids:
Biology, Chemistry, and Medicine, 2nd ed. (M. B. Sporn, A. B. Roberts, and D. S. Goodman,
eds.), Raven Press, New York, 1994, pp. 521–543.
82. C. Hofmann and G. Eichele, Retinoids in development, in The Retinoids: Biology, Chemistry,
and Medicine, 2nd ed. (M. B. Sporn, A. B. Roberts, and D. S. Goodman, eds.), Raven Press,
New York, 1994, pp. 387–441.
83. M. Maden, Vitamin A in embryonic development, Nutr. Rev. 52:3–12 (1994).
84. J. A. Olson, The biological role of vitamin A in maintaining epithelial tissues, Isr. J. Med.
Sci. 8:1170–1178 (1972).
85. Y.-P. Chen, D. Dong, M. Solursh, and M. H. Zile, Direct evidence that retinoic acid is not
a morphogen in the chick limb bud, FASEB J. 9:A833 (1995).
86. M. Studer, H. Po?pperl, H. Marshall, A. Kuroiwa, and R. Krumlauf, Role of a conserved
retinoic acid response element in rhombomere restriction of Hoxb-1, Science 265:1728–1732
87. R. D. Semba, Vitamin A, immunity, and infection, Clin. Infect. Dis. 19:489–499 (1994).
88. M. T. Cantorna, F. E. Nashold, and C. E. Hayes, In vitamin A deficiency multiple mechanisms
establish a regulatory T helper cell imbalance with excess Th1 and insufficient Th2
function, J. Immunol. 152:1515–1522 (1994).
Vitamin A 47
89. W. Stoeckenius and R. A. Bogomolni, Bacteriorhodopsin and related pigments of halobacteria,
Annu. Rev. Biochem. 51:587–616 (1982).
90. J. K. Lanyi, Proton translocation mechanism and energetics in the light-driven pump bacteriorhopsin,
Biochim. Biophys. Acta 1183:241–261 (1990).
91. N. A. Dencher, The five retinal-protein pigments of halobacteria: bacteriorhodopsin, halorhodopsin,
P565, P370, and slow-cycling rhodopsin, Photochem. Photobiol. 38:753–757 (1983).
92. J. L. Spudich, Protein-protein interaction converts a proton pump into a sensory receptor,
Cell 79:747–750 (1994).
93. J. S. Bertram, Cancer prevention by carotenoids: mechanistic studies in cultured cells, Ann.
N.Y. Acad. Sci. 691:177–191 (1993).
94. M. Hanusch, W. Stahl, W. A. Schulz, and H. Sies, Induction of gap junctional communication
by 4-oxoretinoic acid generated from its precursor canthaxanthin, Arch. Biochem. Biophys.
317:423–428 (1995).
95. B. A. Underwood, Vitamin A in animal and human nutrition, in The Retinoids, Vol. 1 (M.
B. Sporn, A. B. Roberts, and D. S. Goodman, eds.), Academic Press, New York, 1984, pp.
96. B. A. Underwood, Vitamin A in human nutrition: public health considerations, in The Retinoids:
Biology, Chemistry, and Medicine, 2nd ed. (M.B. Sporn, A.B. Roberts, and D.S. Goodman,
eds.), Raven Press, New York, 1994, pp. 211–227.
97. G. L. Peck and J. J. DiGiovanna, Synthetic retinoids in dermatology, in The Retinoids: Biology,
Chemistry, and Medicine, 2nd ed. (M. B. Sporn, A. B. Roberts, and D. S. Goodman,
eds.), Raven Press, New York, 1994, pp. 631–658.
98. A. Vahlquist, Role of retinoids in normal and diseased skin, in Vitamin A in Health and
Disease (R. Blomhoff, ed.), Marcel Dekker, New York, 1994, pp. 365–424.
99. D. B. Gunning, A. B. Barua, R. A. Lloyd, and J. A. Olson, Retinoyl ?-glucuronide: a nontoxic
retinoid for the topical treatment of acne, J. Dermatol. Treat. 5:181–185 (1994).
100. G. J. Fisher, H. S. Talwar, J.-H. Xiao, S. C. Datta, A. P. Reddy, M.-P. Gaubs, C. Rochette-
Egly, P. Chambon, and J. J. Voorhees, Immunological identification and functional quantitation
of retinoic acid and retinoid X receptor proteins in human skin, J. Biol. Chem. 269:
20629–20635 (1994).
101. C. E. Griffiths and J. J. Voorhees, Human in vivo pharmacology of topical retinoids, Arch.
Dermatol. Res. 287:53–60 (1994).
102. R. C. Moon, R. G. Mehta, and K. V. N. Rao, Retinoids and cancer in experimental animals,
in The Retinoids: Biology, Chemistry, and Medicine, 2nd ed. (M. B. Sporn, A. B. Roberts,
and D. S. Goodman, eds.), Raven Press, New York, 1994, pp. 573–595.
103. W. K. Hong and L. M. Itri, Retinoids and human cancer, in The Retinoids: Biology, Chemistry,
and Medicine, 2nd ed. (M. B. Sporn, A. B. Roberts, and D. S. Goodman, eds.), Raven
Press, New York, 1994, pp. 597–630.
104. L. M. Canfield, N. I. Krinsky, and J. A. Olson (eds.), Carotenoids in human health, Ann.
N.Y. Acad. Sci. 691:1–300 (1993).
105. W. Schalch and P. Weber, Vitamins and carotenoids—a promising approach to reducing the
risk of coronary heart disease, cancer and eye diseases, in Free Radicals in Diagnostic Medicine
(D. Armstrong, ed.), Plenum Press, New York, 1994, pp. 335–350.
106. J. C. Bauernfeind, The Safe Use of Vitamin A, Int. Vitamin A Consult. Group, Nutrition
Foundation, Washington, D.C., 1980, pp. 1–44.
107. M. P. Macapinlac and J. A. Olson, A lethal hypervitaminosis A syndrome in young monkeys
following a single intramuscular dose of a water-miscible preparation containing vitamins
A, D2 and E, Int. J. Vitam. Nutr. Res. 51:331–341 (1981).
108. M. E. Bush and B. B. Dahms, Fatal hypervitaminosis in a neonate, Arch. Pathol. Lab. Med.
108:838–842 (1984).
109. R. B. Armstrong, K. O. Ashenfelter, C. Eckhoff, A. A. Levin, and S. S. Shapiro, General
and reproductive toxicology of retinoids, in The Retinoids: Biology, Chemistry, and Medicine,
48 Olson
2nd ed. (M. B. Sporn, A. B. Roberts, and D. S. Goodman, eds.), Raven Press, New York,
1994, pp. 545–572.
110. J. N. Hathcock, D. G. Hattan, M. Y. Jenkins, J. T. McDonald, P. R. Sundaresan, and V. L.
Wilkening, Evaluation of vitamin A toxicity, Am. J. Clin. Nutr. 52:183–202 (1990).
111. A. P. Geubel, C. De Galocsy, N. Alves, J. Rahier, and C. Dive, Liver damage caused by
therapeutic vitamin A administration: estimate of dose-related toxicity in 41 cases, Gastroenterology
100:1701–1709 (1991).
112. T. E. Kowalski, M. Falestiny, E. Furth, and P. F. Malet, Vitamin A hepatotoxicity: a cautionary
note regarding 25,000 IU supplements, Am. J. Med. 97:523–528 (1994).
113. H. Nau, I. Chahoud, L. Dencker, E. J. Lammer, and W. J. Scott, Teratogenicity of vitamin
A and retinoids, in Vitamin A in Health and Disease (R. Blomhoff, ed.), Marcel Dekker,
New York, 1994, pp. 615–663.
114. J. Adams, Structure–activity and dose–response relationships in the neural and behavioral
teratogenesis of retinoids, Neurotoxicol. Teratol. 15:193–202 (1993).
115. J. Adams, Neural and behavioral pathology following prenatal exposure to retinoids, in Retinoids
in Clinical Practice (G. Koren, ed.), Marcel Dekker, New York, 1993, pp. 111–115.
116. J. Adams and E. J. Lammer, Neurobehavioral teratology of isotretinoin, Reprod. Toxicol. 7:
175–177 (1993).
117. C. C. Willhite, Molecular correlates in retinoid pharmacology and toxicology, in Chemistry
and Biology of Synthetic Retinoids (M. L. Dawson and W. H. Okamura, eds.), CRC Press,
Boca Raton, FL, 1990, pp. 539–573.
118. D. B. Gunning, A. B. Barua, and J. A. Olson, Comparative teratogenicity and metabolism
of all-trans retinoic acid, all-trans retinoyl ?-glucose, and all-trans retinoyl ?-glucuronide in
pregnant Sprague-Dawley rats, Teratology 47:29–36 (1993).
119. G. Tzimas, J. O. Sass, W. Wittfoht, M. M. A. Elmazar, K. Ehlers, and H. Nau, Identification
of 9,13-di-cis-retinoic acid as a major plasma metabolite of 9-cis-retinoic acid and limited
transfer of 9-cis-retinoic acid and 9-13-di-cis-retinoic acid to the mouse and rat embryos,
Drug Metab. Dispos. 22:928–936 (1994).
120. R. L. Horst, T. A. Reinhardt, J. P. Goff, B. J. Nonnecke, V. K. Gambhir, P. D. Fiorella, and
J. L. Napoli, Identification of 9-cis, 13-cis-retinoic acid as a major circulating retinoid in
plasma, Biochemistry 34:1203–1209 (1995).
121. T. O. Carpenter, J. M. Pettifor, R. M. Russell, J. Pitha, S. Mobarhan, M. S. Ossip, S. Wainer,
and C. S. Anast, Severe hypervitaminosis A in siblings: evidence of variable tolerance to
retinol intake, J. Pediatr. 111:507–512 (1987).
122. F. W. Rosa, Retinoid embryopathy in humans, in Retinoids in Clinical Practice (G. Koren,
ed.), Marcel Dekker, New York, 1993, pp. 77–109.
123. Recommendations for vitamin A use during pregnancy: Teratology Society position paper,
Teratology 35:269–275 (1987).
124. C. W. Callaway, K. W. McNutt, R. S. Rivlin, A. C. Ross, H. H. Sandstead, and A. P. Simopoulos,
Statement on vitamin and mineral supplements, The Joint Public Information Committee
of the American Institute of Nutrition and the American Society for Clinical Nutrition,
J. Nutr. 117:1649 (1987).
125. Council for Responsible Nutrition, Safety of Vitamins and Minerals: A Summary of Findings
of Key Reviews, Washington, D.C., 1986, p. 12.
126. A. H. Baqui, A. de Francisco, S. E. Arifeen, A. K. Siddique, and R. B. Sack, Bulging fontanelle
after supplementation with 25,000 IU vitamin A in infancy using EPI contacts, Proc.
16th Int. Vitam. A Consult. Group, 24–28 Oct. 1994, Chiangrai, Thailand, p. 74.
127. J. H. Humphrey, T. Agoestina, G. Taylor, A. Usman, K. P. West, Jr., and A. Sommer, Acute
and long term risks and benefits to neonates of 50,000 IU oral vitamin A, Proc. XVI Int.
Vitam. A Consult. Group, 24–28 Oct. 1994, Chiangrai, Thailand, p. 74.
128. U. Weber, G. Georz, H. Baseler, and L. Michaelis, Canthaxanthin retinopathy: follow-up of
over 6 years, Klin. Monatsbl. Augenheilkd, 201:174–177 (1992).
Vitamin A 49
129. S. T. Mayne, ?-Carotene, carotenoids and disease prevention in humans, FASEB J. 10:690–
701 (1996).
130. G. W. Burton, Antioxidant action of carotenoids, J. Nutr. 119:109–111 (1989).
131. R. G. Ziegler, L. A. Brinton, R. F. Hamman, H. F. Lehman, R. S. Levine, K. Mallin, S. A.
Norman, J. F. Rosenthal, A. C. Trumble, and R. N. Hoover, Diet and the risk of invasive
cervical cancer among white women in the United States, Am. J. Epidemiol. 132:432–445
132. K. F. Gey, U. K. Moser, P. Jordan, H. B. Sta?helin, M. Eichholzer, and E. Lu?din, Increased
risk of cardiovascular disease at suboptimal plasma concentrations of essential antioxidants:
an epidemiological update with special attention to carotene and vitamin C, Am. J. Clin.
Nutr. 57:787S–797S (1993).
133. R. A. Bone, J. T. Landrum, G. W. Hime, A. Cains, and J. Zamor, Stereochemistry of the
human macular carotenoids, Invest. Ophthalmol. Vis. Sci. 34:2033–2040 (1993).
134. G. J. Handelman, E. A. Dratz, C. C. Reay, and F. J. G. M. van Kuijk, Carotenoids in the
human macula and whole retina, Invest. Ophthalmol. Vis. Sci. 29:850–855 (1988).
135. A. Bendich, Recent advances in clinical research involving carotenoids, Pure Appl. Chem.
66:1017–1024 (1994).
136. P. Pontiggia, A. B. Santamaria, K. Alonso, and L. Santamaria, Whole body hyperthermia
associated with beta-carotene supplementation in patients with AIDS, Biomed. Pharmacother.
49:263–265 (1995).
137. J. A. Olson, Vitamin A, in Handbook of Vitamins, 2nd ed. (L. Machlin, ed.), Marcel Dekker,
New York, 1991, pp. 1–57.
138. J. A. Olson, Benefits and liabilities of vitamin A and carotenoids, J. Nutr. 126:1208S–1212S
139. A. Bendich and J. A. Olson, Biological actions of carotenoids, FASEB J. 3:1927–1932 (1989).
140. P. Chambon, J. A. Olson, and A. C. Ross, coordinators, The retinoid revolution, FASEB J.
10:939–1107 (1996).
141. A. Sommer and K. P. West, Jr., Vitamin A deficiency, in Health, Survival and Vision, Oxford
University Press, New York, 1996, pp. 1–438.
142. A. B. Barua, Retinoyl beta-glucuronide: a biologically active form of vitamin A, Nutr. Rev.
55:259–267 (1997).
143. R. K. Miller, A. G. Hendricks, J. L. Mills, H. Hummler, and U. W. Wiegand, Periconceptional
vitamin A use: how much is teratogenic? Reprod. Toxicol. 12:75–88 (1998).
144. J. E. Thurman and A. D. Mooradian, Vitamin supplementation therapy in the elderly, Drugs
Aging 11:433–449 (1997).
145. R. D. Semba, Overview of the potential role of vitamin A in mother-to-child transmission
of HIV-1, Acta Paediatr. Suppl. 421:107–112 (1997).
146. R. W. Evans, B. J. Shaten, B. W. Day, and L. H. Kuller, Prospective association between
lipid soluble antioxidants and coronary heart disease in men. The Multiple Risk Factor Intervention
Trial, Am. J. Epidemiol. 147:180–186 (1998).
147. C. E. Orfanos, C. C. Zouboulis, B. Almond-Roesler, et al., Current use and future potential
role of retinoids in dermatology, Drugs 53:358–388 (1997).
148. A. C. Ross, Vitamin A and retinoids, in Modern Nutrition in Health and Disease, 9th ed.
(M. E. Shils, J. A. Olson, M. Shike, and A. C. Ross, eds.), Williams and Wilkins, Baltimore,
1999, pp. 305–327.
149. K. D. Lawson, S. J. Middleton, and C. D. Hassall, Olestra, a nonabsorbed, noncaloric replacement
for dietary fat: a review, Drug Metab. Rev. 29:651–703 (1997).
150. G. Britton, ed., Eleventh international symposium on carotenoids, Pure Appl. Chem. 69:
2027–2173 (1997).
151. H. C. Furr, and R. M. Clark, Intestinal absorption and tissue distribution of carotenoids, J.
Nutr. Biochem. 8:364–377 (1997).
152. R. S. Parker, Bioavailability of carotenoids, Eur. J. Clin. Nutr. 51(Suppl.): S86–S90 (1997).
50 Olson
153. C. A. Rice-Evans, J. Sampson, P. M. Bramley, and D. K. Holloway, Why do we expect
carotenoids to be antioxidants in vivo? Free Radic. Res. 26:381–398 (1997).
154. J. A. Olson, Carotenoids, in Modern Nutrition in Health and Disease, 9th ed. (M. E. Shils,
J. A. Olson, M. Shike, and A. C. Ross, eds.), Williams & Wilkins, Baltimore, 1999, pp. 525–
155. J. A. Olson, N. Loveridge, G. G. Duthie, and M. J. Shearer, Fat soluble vitamins, in Human
Nutrition and Dietetics, 10th ed. (J. S. Garrow, W. P. T. James, and A. Ralph, eds.), Churchill
Livingstone, Edinburgh, 2000, pp. 211–247.
156. M. I. Dawson, The importance of vitamin A in nutrition, Curr. Pharm. Des. 6:311–325
157. J. P. Shenai, Vitamin A supplementation in very low birth weight neonates: rational and
evidence, Pediatrics 104:1369–1374 (1999).
158. A. Coutsoudis, K. Pillay, E. Spooner, L. Kuhn, and H. M. Coovadia, Influence of infantfeeding
patterns on early mother-to-child transmission of HIV-1 in Durban, South Africa: a
prospective cohort study, Lancet 354:471–476 (1999).
159. J. J. Castenmiller and C. E. West, Bioavailability and bioconversion of carotenoids, Ann.
Rev. Nutr. 18:19–38 (1998).
160. IARC Handbooks of Cancer Prevention, Vol. 2. Carotenoids, Intern. Agency for Research
on Cancer Press, Lyon, 1998, pp. 1–326.
161. IARC Handbooks of Cancer Prevention, Vol. 3. Vitamin A, Intern. Agency for Research on
Cancer Press, Lyon, 1999, pp. 1–261.
162. IARC Handbooks of Cancer Prevention, Vol. 4. Retinoids, Intern. Agency for Research on
Cancer Press, Lyon, 1999, pp. 1–331.
Vitamin D
San Jose? State University, San Jose?, California
University of California, Riverside, California
Vitamin D designates a group of closely related compounds that possess antirachitic activity.
The two most prominent members of this group are ergocalciferol (vitamin D2) and
cholecalciferol (vitamin D3). Ergocalciferol is derived from a common plant steroid, ergosterol,
and is the form that was employed for vitamin D fortification of foods from the
1940s to the 1960s. Cholecalciferol is the form of vitamin D obtained when radiant energy
from the sun strikes the skin and converts the precursor 7-dehydrocholesterol into vitamin
D3. Since the body is capable of producing cholecalciferol, vitamin D does not meet the
classical definition of a vitamin. It is more accurate to call vitamin D a prohormone; thus,
vitamin D is metabolized to a biologically active form that functions as a steroid hormone
(1–4). However, since vitamin D was first recognized as an essential nutrient, it has historically
been classified among the lipid-soluble vitamins. Even today it is thought of by many
as a vitamin, although it is now known that there exists a vitamin D endocrine system
that generates the steroid hormone 1?,25-dihydroxyvitamin D3 [1?,25(OH)2D3].
Vitamin D functions to maintain calcium homeostasis together with two peptide
hormones, calcitonin and parathyroid hormone (PTH). Vitamin D is also important for
phosphorus homeostasis (5,6). Calcium and phosphorus are required for a wide variety
of biological processes (Table 1). Calcium is necessary for muscle contraction, nerve pulse
transmission, blood clotting, and membrane structure. It also serves as a cofactor for such
enzymes as lipases and ATPases and is needed for eggshell formation in birds. It is an
important intracellular signaling molecule for signal transduction pathways, such as those
involving calmodulin and protein kinase C. Phosphorus is an important component of
52 Collins and Norman
Table 1 Biological Calcium and Phosphorus
Calcium Phosphorus
Body content: 70-kg man has 1200 g Ca2 Body content: 70-kg man has 770 g P
Structural: bone has 95% of body Ca Structural: Bone has 90% of body P
Plasma [Ca2] is 2.5 mM, 10 mg % Plasma [Pi] is 2.3 mM, 2.5–4.3 mg %
Muscle contraction Intermediary metabolism (phosphorylated intermediates)
Nerve pulse transmission
Blood clotting Genetic information (DNA and RNA)
Membrane structure Phospholipids
Enzyme cofactors (amylase, trypsinogen, Enzyme/protein components (phosphohistilipases,
ATPases) dine, phosphoserine)
Eggshell (birds) Membrane structure
Daily requirements (70-kg man)
Dietary intake: 700a Dietary intake: 1200a
Fecal excretion: 300–600a,b Fecal excretion: 350–370a,b
Urinary excretion: 100–400a,b Urinary excretion: 200–600a,b
aValues in milligrams per day.
bBased on the indicated level of dietary intake.
DNA, RNA, membrane lipids, and the intracellular energy–transferring ATP system. The
phosphorylation of proteins is important for the regulation of many metabolic pathways.
Furthermore, the maintenance of serum calcium and phosphorus levels within narrow
limits is important for normal bone mineralization. Any perturbation in these levels results
in bone calcium accretion or resorption. Disease states, such as rickets, can develop if the
serum ion product is not maintained at a level consistent with that required for normal bone
mineralization. Maintaining a homeostatic state for these two elements is of considerable
importance to a living organism.
Recently, 1?,25(OH)2D3 has been shown to act on novel target tissues not related
to calcium homeostasis. There have been reports characterizing receptors for the hormonal
form of vitamin D and activities in such diverse tissues as brain, pancreas, pituitary, skin,
muscle, immune cells, and parathyroid (Table 2). These studies suggest that vitamin D
status is important for insulin and prolactin secretion, muscle function, immune and stress
response, melanin synthesis, and cellular differentiation of skin and blood cells.
There are a number of recent books (7–9) and comprehensive reviews (10–19) that
cover many aspects of vitamin D, including its endocrinological aspects.
Rickets, a deficiency disease of vitamin D, appears to have been a problem in ancient
times. There is evidence that rickets occurred in Neanderthal man about 50,000 bc (20).
The first scientific descriptions of rickets were written by Dr. Daniel Whistler (21) in 1645
and by Professor Francis Glisson (22) in 1650. Rickets became a health problem in northern
Europe, England, and the United States during the Industrial Revolution when many
people lived in urban areas with air pollution and little sunlight. Prior to the discovery of
vitamin D, theories on the causative factors of rickets ranged from heredity to syphilis
Vitamin D 53
Table 2 Distribution of 1,25(OH)2D3 Actionsa
Tissue Distribution of Nuclear 1,25(OH)2D3 Receptor
Adipose Intestine Pituitary
Adrenal Kidney Placenta
Bone Liver (fetal) Prostrate
Bone marrow Lung Retina
Brain Muscle, cardiac Skin
Breast Muscle, embryonic Stomach
Cancer cells Muscle, smooth Testis
Cartilage Osteoblast Thymus
Colon Ovary Thyroid
Eggshell gland Pancreas ? cell Uterus
Epididymus Parathyroid Yolk sac (bird)
Hair follicle Parotid
Distribution of Nongenomic Responses
Intestine Transcaltachia
Osteoblast Ca2 channel opening
Osteoclast Ca2 channel opening
Liver Lipid metabolism
Muscle A variety
aSummary of the tissue location of the nuclear receptor for 1?,25(OH)2D3 (nVDR) (top) and
tissues displaying ‘‘rapid’’ or membrane-initiated biological responses (bottom).
Some of the important scientific discoveries leading to the understanding of rickets
were dependent on discoveries about bone. As reviewed by Hess (23), the first formal
descriptions of bone were made by Marchand (1842), Bibard (1844), and Friedleben
(1860). In 1885, Pommer wrote the first description of the pathological process taking
place in the rachitic skeleton. In 1849, Trousseau and Lasque recognized that osteomalacia
and rickets were different manifestations of the same disorder. In 1886 and 1890,
Hirsch and Palm did a quantitative geographical study of the worldwide distribution of
rickets and found that the incidence of rickets paralleled the incidence of lack of sunlight
(23). This was substantiated in 1919 when Huldschinsky demonstrated that ultraviolet
(UV) rays were effective in healing rickets (24).
In the early 1900s, the concept of vitamins was developed and nutrition emerged
as an experimental science, allowing for further advances in understanding rickets. In
1919, Sir Edward Mellanby (25,26) was able to experimentally produce rickets in puppies
by feeding synthetic diets to over 400 dogs. He further showed that rickets could be prevented
by the addition of cod liver oil or butterfat to the feed. He postulated that the
nutritional factor preventing rickets was vitamin A since butterfat and cod liver oil were
known to contain vitamin A (26). Similar studies were also conducted by McCollum et
al. (27).
The distinction between the antixerophthalmic factor, vitamin A, and the antirachitic
factor, vitamin D, was made in 1922 when McCollum’s laboratory showed that the antirachitic
factor in cod liver oil could survive both aeration and heating to 100°C for 14 h,
whereas the activity of vitamin A was destroyed by this treatment. McCollum named the
new substance vitamin D (28).
Although it was known that UV light and vitamin D are equally effective in preventing
and curing rickets, the close interdependence of the two factors was not immedi54
Collins and Norman
ately recognized. Then, in 1923, Goldblatt and Soames (29) discovered that food that was
irradiated and fed to rats could cure rickets; food that was not irradiated could not cure
rickets. In 1925, Hess and Weinstock demonstrated that a factor with antirachitic activity
was produced in the skin upon UV irradiation (30,31). Both groups demonstrated that the
antirachitic agent was in the lipid fraction. The action of the light appeared to produce a
permanent chemical change in some component of the diet and the skin. They postulated
that a provitamin D existed that could be converted to vitamin D by UV light absorption.
Much more work ultimately demonstrated that the antirachitic activity resulted from the
irradiation of 7-dehydrocholesterol.
The isolation and characterization of vitamin D was now possible. In 1932, the
structure of vitamin D2 was simultaneously determined by Windaus in Germany, who
named it vitamin D2 (32), and by Askew in England, who named it ergocalciferol (33).
In 1936, Windaus identified the structure of vitamin D found in cod liver oil, vitamin D3
(34). Thus, the ‘‘naturally’’ occurring vitamin is vitamin D3, or cholecalciferol. The structure
of vitamin D was determined to be that of a steroid or, more correctly, a seco-steroid.
However, the relationship between its structure and its mode of action was not realized
for an additional 30 years.
Vitamin D was believed for many years to be the active agent in preventing rickets.
It was assumed that vitamin D was a cofactor for reactions that served to maintain calcium
and phosphorus homeostasis. However, when radioisotopes became available, more precise
measurements of metabolism could be made. Using radioactive 45Ca2, Linquist found
that there was a lag period between the administration of vitamin D and the initiation of its
biological response (35). Stimulation of intestinal calcium absorption required 36–48 h for
a maximal response. Other investigators found delays in bone calcium mobilization and
serum calcium level increases after treatment with vitamin D (36–40). The duration of the
lag and the magnitude of the response were proportional to the dose of vitamin D used (37).
One explanation for the time lag was that vitamin D had to be further metabolized
before it was active. With the development of radioactively labeled vitamin D, it became
possible to study the metabolism of vitamin D. Norman et al. were able to detect three
metabolites that possessed antirachitic activity (41). One of these metabolites was subsequently
identified as the 25-hydroxy derivative of vitamin D3 [25(OH)D3] (42). Because
25(OH)D3 was found to have 1.5 times more activity than vitamin D in curing rickets in
the rat, it was thought that this metabolite was the biologically active form of vitamin D
(43). However, in 1968, Haussler et al. reported a more polar metabolite that was found
in the nuclear fraction of the intestine from chicks given tritiated vitamin D3 (44). Biological
studies demonstrated that this new metabolite was 13–15 times more effective
than vitamin D in stimulating intestinal calcium absorption and 5–6 times more effective in
elevating serum calcium levels (45). The new metabolite was also as effective as vitamin D
in increasing growth rate and bone ash (45). In 1971, the structural identity of this metabolite
was reported to be the 1?,25-dihydroxy derivative of vitamin D[1?,25(OH)2D3] (46–
48), the biologically active metabolite of vitamin D.
In 1970, the site of production of 1?,25(OH)2D3 was demonstrated to be the kidney
(49). This discovery, together with the finding that 1?,25(OH)2D3 is found in the nuclei
of intestinal cells, suggested that vitamin D was functioning as a steroid hormone (44,50).
Subsequently, a nuclear receptor protein for 1?,25(OH)2D3 was identified and characterized
(50,51). Since the cDNA for the 1?,25(OH)2D3 nuclear receptor from several species
has now been cloned and sequenced (52–55), the relationship between vitamin D and the
other steroid hormones has been clearly established (56). The discovery that the biological
Vitamin D 55
actions of vitamin D could be explained by the classical model of steroid hormone action
marked the beginning of the modern era of vitamin D.
A. Structure
As previously mentioned, vitamin D refers to a family of compounds that possess antirachitic
activity. Members of the family are derived from the cyclopentanoperhydrophenanthrene
ring system (Fig. 1), which is common to other steroids, such as cholesterol.
However, vitamin D has only three intact rings; the B ring has undergone fission of the 9,10
carbon bond, resulting in the conjugated triene system of double bonds that is possessed by
all D vitamins. The structure of vitamin D3 is shown in Fig. 1. Naturally occurring mem-
Fig. 1 Chemistry and irradiation pathway for production of vitamin D3 (a natural process) and
vitamin D2 (a commercial process). In each instance the provitamin, with a ?5,?7 conjugated doublebond
system in the B ring, is converted to the seco-B previtamin, with the 9,10 carbon–carbon bond
broken. Then the previtamin D thermally isomerizes to the ‘‘vitamin’’ form, which contains a system
of three conjugated double bonds. In solution vitamin D is capable of assuming a large number of
conformations due to rotation about the 6,7 carbon–carbon bond of the B ring. The 6-s-cis conformer
(the steroid-like shape) and the 6-s-trans conformer (the extended shape) are presented for both
vitamin D2 and vitamin D3.
56 Collins and Norman
Table 3 Side Chains of Provitamin D
bers of the vitamin D family differ from each other only in the structure of their side
chains; the side chain structures of the various members of the vitamin D family are given
in Table 3.
From the x-ray crystallographic work of Nobel laureate Crowfoot-Hodgkin et al.,
it is now known that the diene system of vitamin D that extends from C-5 to C-8 is transoid
and nearly planar (57,58). However, the C-6 to C-19 diene system is cisoid and not planar.
The C-10 to C-19 double bond is twisted out of the plane by 60°. As a result, the A ring
exists in one of two possible chair conformations. It is also known that the C and D rings
are rigid and that the side chain prefers an extended configuration. In 1974, Okamura et
al. reported that vitamin D and its metabolites have a high degree of conformational mobility
(59). Using nuclear magnetic resonance (NMR) spectroscopy, they were able to detect
that the A ring undergoes rapid interconversion between the two chair conformations, as
shown in Fig. 2. This conformational mobility is unique to the vitamin D molecule and
is not observed for other steroid hormones. It is a direct consequence of the breakage of
the 9,10 carbon bond of the B ring, which serves to ‘‘free’’ the A ring. As a result of this
mobility, substituents on the A ring are rapidly and continually alternating between the
axial and equatorial positions.
B. Nomenclature
Vitamin D is named according to the new revised rules of the International Union of Pure
and Applied Chemists (IUPAC). Since vitamin D is derived from a steroid, the structure
retains its numbering from the parent steroid compound. Vitamin D is designated ‘‘seco’’
because its B ring has undergone fission. Asymmetrical centers are named using R, S
Vitamin D 57
Fig. 2 The dynamic behavior of 1?,25(OH)2D3. The topological features of the hormone
1?,25(OH)2D3 undergo significant changes as a consequence of rapid conformational changes (due
to single-bond rotation) or, in one case, as a consequence of a hydrogen shift (resulting in the
transformation of 1?,25(OH)2D3 to pre-1?,25(OH)2D3). (top) The dynamic changes occurring
within the seco-B conjugated triene framework of the hormone (C5, 6, 7, 8, 9, 10, 19). All of the
carbon atoms of the 6-s-trans conformer of 1?,25(OH)2D3 are numbered using standard steroid
notation for the convenience of the reader. Selected carbon atoms of the 6-s-cis conformer are also
numbered as are those of pre-1?,25(OH)2D3. (middle) The rapid chair–chair inversion of the A ring
of the secosteroid. (bottom) The dynamic single-bond conformational rotation of the cholesterollike
side chain of the hormone. The C/D trans-hydrindane moiety is assumed to serve as a rigid
anchor about which the A-ring, seco-B triene, and side chain are in dynamic equilibrium.
notation and Cahn’s rules of priority. The configuration of the double bonds is notated
E, Z; E for ‘‘trans,’’ Z for ‘‘cis.’’ The formal name for vitamin D3 is 9,10-seco(5Z,7E)-
5,7,10(19)-cholestatriene-3?-ol and for vitamin D2 is 9,10-seco(5Z,7E)-5,7,10(19),21-
C. Chemical Properties
1. Vitamin D3(C27H44O)
Three double bonds
Melting point, 84–85°C
UV absorption maximum at 264–265 nm with a molar extinction coefficient of
18,300 in alcohol or hexane, ?D20 84.8° in acetone
58 Collins and Norman
Molecular weight, 384.65
Insoluble in H2O
Soluble in benzene, chloroform, ethanol, and acetone
Unstable in light
Will undergo oxidation if exposed to air at 24°C for 72 h
Best stored at 0°C
2. Vitamin D2(C28H44O)
Four double bonds
Melting point, 121°C
UV absorption maximum at 265 nm with a molar extinction coefficient of 19,400
in alcohol or hexane, ?D20  106° in acetone
Same solubility and stability properties as D3
D. Isolation
Many of the studies that have led to our understanding of the mode of action of vitamin
D have involved tissue localization and identification of vitamin D and its various metabolites.
Since vitamin D is a steroid, it is isolated from tissue by methods that extract total
lipids. The technique most frequently used for this extraction is that of Bligh and Dyer
Over the years, a wide variety of chromatographic techniques have been used to
separate vitamin D and its metabolites. These include paper, thin-layer, column, and gas
chromatographic methods. Paper and thin-layer chromatography usually require long development
times with unsatisfactory resolutions and have limited capacity. Column chromatography,
with alumina, Floridin, Celite, silicic acid, and Sephadex LH-20 as supports,
has been used to rapidly separate many closely related vitamin D compounds (2). However,
none of the above methods are capable of resolving and distinguishing vitamin D2
from vitamin D3. Gas chromatography can separate these two compounds, but in the process
vitamin D is thermally converted to pyrocalciferol and isopyrocalciferol, resulting in
two peaks. High-performance liquid chromatography (HPLC) has become the method of
choice for the separation of vitamin D and its metabolites (61,62). This powerful technique
is rapid and gives good recovery with high resolution.
E. Synthesis of Vitamin D
1. Photochemical Production
In the 1920s, it was recognized that provitamins D were converted to vitamins D upon
treatment with UV radiation (Fig. 1). The primary structural requirement for a provitamin
D is a sterol with a C-5 to C-7 diene double-bond system in ring B. The conjugated
double-bond system is a chromophore which upon UV irradiation initiates a series of
transformations resulting in the production of the vitamin D seco-steroid structure. The
two most abundant provitamins D are ergosterol (provitamin D2) and 7-dehydrocholesterol
(provitamin D3).
2. Chemical Synthesis
There are two basic approaches to the synthesis of vitamin D. The first involves the chemical
synthesis of a provitamin that can be converted to vitamin D by UV irradiation. The
second is a total chemical synthesis.
Vitamin D 59
Since vitamin D is derived from cholesterol, the first synthesis of vitamin D resulted
from the first chemical synthesis of cholesterol. Cholesterol was first synthesized by two
groups in the 1950s. The first method involves a 20-step conversion of 4-methoxy-2,5-
toluquinone to a progesterone derivative, which is then converted in several steps to progesterone,
testosterone, cortisone, and cholesterol (63). The other method uses the starting
material 1,6-dihydroxynaphthalene. This is converted to the B and C rings of the steroid.
A further series of chemical transformations leads to the attachment of the A ring and
then the D ring. The final product of the synthesis was epiandrosterone, which could be
converted to cholesterol (64). The cholesterol was then converted to 7-dehydrocholesterol
and UV-irradiated to give vitamin D. The yield of vitamin D from photochemical conversion
is normally 10–20%.
The first pure chemical synthesis of vitamin D, without any photochemical irradiation
steps, was accomplished in 1967 (65). This continuing area of investigation allows
for the production of many vitamin D metabolites and analogs without the necessity of
a photochemical step. Pure chemical synthesis also allows for the synthesis of radioactive
vitamin D and metabolites for the study of the metabolism of vitamin D.
Figure 3 summarizes some of the currently used synthetic strategies (14). Method
A involves the photochemical ring opening of a 1-hydroxylated side-chain-modified deriv-
Fig. 3 Summary of approaches to the chemical synthesis of 1?,25(OH)2D3. The general synthetic
approaches A–H, which are discussed in the text, represent some of the major synthetic approaches
used in recent years to synthesize the hormone 1?,25(OH)2D3 and analogs of 1?,25(OH)2D3.
60 Collins and Norman
ative of 7-dehydrocholesterol 1 producing a provitamin that is thermolyzed to vitamin D
(66,67). Method B is useful for producing side chain and other analogs. In this method,
the phosphine oxide 2 is coupled to a Grundmann’s ketone derivative 3, producing the
1?,25(OH)2D3 skeleton (68,69). In method C, dienynes such as 4 are semihydrogenated
to a previtamin structure that undergoes rearrangement to the vitamin D analog (70,71).
Method D involves the production of the vinylallene 6 from compound 5 and the subsequent
rearrangement with heat- or metal-catalyzed isomerization followed by sensitized
photoisomerization (72). Method E starts with an acyclic A ring precursor 7 that is intramolecularly
cross-coupled to bromoenyne 8, resulting in the 1,25(OH)2D3 skeleton
(73,74). Method F starts with the tosylate of 11, which is isomerized to the i-steroid 10.
This structure can be modified at C-1 and then reisomerized under sovolytic conditions
to 1?,25(OH)2D3 or analogs (75,76). In method G, vitamin D derivatives 11 are converted
to 1-oxygenated 5,6-trans vitamin D derivatives 12 (77). Finally, method H involves the
direct modification of 1?,25(OH)2D3 or an analog 13 through the use of protecting groups,
such as transition metal derivatives, or by other direct chemical transformations on 13
(78). These synthetic approaches have allowed the synthesis of more than 300 analogs
of 1?,25(OH)2D3. For an article that gives an extensive review of all the synthetic
approaches, see (14).
The elucidation of the metabolic pathway by which vitamin D is transformed into its
biologically active form is one of the most important advances in our understanding of
how vitamin D functions and the development of the vitamin D endocrine system. It is
now known that both vitamin D2 and vitamin D3 must be hydroxylated at the C-1 and C-
25 positions (Fig. 4) before they can produce their biological effects. The activation of
vitamin D2 occurs via the same metabolic pathway as does the activation of vitamin D3,
and the biological activities of both vitamin D2 and vitamin D3 have been shown to be
identical in all animals except birds and the New World monkey. Apparently, these animals
have the ability to discriminate against vitamin D2 (79).
A. Absorption
Vitamin D can be obtained from the diet, in which case it is absorbed in the small intestine
with the aid of bile salts (80,81). In rat, baboon, and human, the specific mode of vitamin
D absorption is via the lymphatic system and its associated chylomicrons (82,83). It has
been reported that only about 50% of a dose of vitamin D is absorbed (83,84). However,
considering that sufficient amounts of vitamin D can be produced daily by exposure to
sunlight, it is not surprising that the body has not evolved a more efficient mechanism
for vitamin D absorption from the diet.
Although the body can obtain vitamin D from the diet, the major source of this
prohormone is its production in the skin from 7-dehydrocholesterol. The 7-dehydrocholesterol
is located primarily in the malpighian layer of the skin. Upon exposure to UV
light, it is photochemically converted to previtamin D, which then isomerizes to vitamin
D over a period of several days (85). Once formed, vitamin D is preferentially removed
from the skin into the circulatory system by the blood transport protein for vitamin D,
the vitamin D–binding protein (DBP).
Vitamin D 61
Fig. 4 Overview of the vitamin D endocrine and paracrine system. Target organs and cells for
1?,25(OH)2D3 by definition contain receptors for the hormone. Biological effects are generated by
both genomic and nongenomic signaling pathways.
B. Transport
The actual site of transfer of vitamin D from the chylomicrons to its specific plasma carrier
protein, DBP, is unknown. After an oral dose of radioactive vitamin D, the radioactivity
becomes associated with the lipoprotein fraction of the plasma (86). As time passes, there
is a progressive shift from this fraction to the ?-globulin fraction (87,88). It has been
shown that the electrophoretic mobility of DBP is identical to that of ?2-globulins and
albumins (89,90).
In mammals, vitamin D, 25(OH)D3, 24R,25(OH)2D3, and 1?,25(OH)2D3 are transported
on the same protein, DBP (91,92). DBP, also known as group-specific protein (Gc
protein), is a globulin protein with a molecular weight in humans of 58,000. It is a bifunctional
protein, responsible both for the transport of vitamin D and its metabolites as well
as functioning as a scavenger for actin which may be inappropriately present in the plasma.
DBP possesses a high-affinity binding site for monomeric actin and forms a high molecular
weight complex with it (93). DBP also possesses a high-affinity binding site for 25(OH)D3
and binds other vitamin D metabolites with somewhat lower affinity (92). Sequence analysis
of the cDNA for DBP indicates that it shares homology with serum albumin and ?-
fetoprotein (94). DBP has been proposed to help in the cellular internalization of vitamin
D sterols, and levels of DBP influence the concentration of ‘‘bound’’ and ‘free’ hormone
62 Collins and Norman
in the plasma (95). The concentration of the free hormone may be important in determining
the biological activity of the hormone (95–99). Several review articles on DBP are available
C. Storage
Vitamin D is taken up rapidly by the liver. Since it was known that the liver serves as a
storage site for retinol, another fat-soluble vitamin, it was thought that the liver also functioned
as a storage site for vitamin D. However, it has since been shown that blood has
the highest concentration of vitamin D, in comparison with other tissues (100,101). From
studies in rats it was concluded that no rat tissue can store vitamin D or its metabolites
against a concentration gradient (82). The persistence of vitamin D in animals during
periods of vitamin D deprivation may be explained by the slow turnover rate of vitamin
D in certain tissues, such as skin and adipose tissue. During times of deprivation, the
vitamin D in these tissues is released slowly, thus meeting the vitamin D needs of the
animal over a period of time. In contrast, it was found that in pig tissue concentrations
of 1?,25(OH)2D3, especially in adipose tissue, are three- to seven-fold higher than plasma
levels (102).
Similarly, Mawer et al. carried out studies in humans on the distribution and storage
of vitamin D and its metabolites (103). In human tissue, adipose tissue and muscle were
found to be major storage sites for vitamin D. Their studies also indicated that adipose
tissue serves predominantly as the storage site for vitamin D3 and that muscle serves as
the storage site for 25(OH)D3.
D. Metabolism
The parent vitamin D is largely biologically inert; before vitamin D can exhibit any biological
activity, it must be metabolized by the body to its active forms. 1?,25(OH)2D3 is the
most active metabolite known, but there is evidence that 24R,25(OH)2D3 is required for
some of the biological responses attributed to vitamin D (16,104,105). Both of these metabolites
are produced in vivo following carbon-25 hydroxylation of the parent vitamin
D molecule.
1. 25(OH)D3
In the liver, vitamin D undergoes its initial transformation, which involves the addition
of a hydroxyl group to the 25-carbon. The metabolite thus formed is 25(OH)D3, which
is the major circulating form of vitamin D. Although there is some evidence that this
metabolite can be formed in other tissues, such as intestine and kidney, it is generally
accepted that the formation of 25(OH)D3 occurs predominantly in the liver.
The production of 25(OH)D3 is catalyzed by the enzyme vitamin D3 25-hydroxylase.
The 25-hydroxylase is found in liver microsomes and mitochondria (106–109). It is a
P450-like enzyme that is poorly regulated (110). Therefore, circulating levels of 25(OH)D3
are a good index of vitamin D status, i.e., they reflect the body content of the parent
vitamin D3 (111,112). Recent studies have suggested that the 25-hydroxylation of vitamin
D is partially regulated by 1?,25(OH)2D3. Other studies suggest that 25-hydroxylation is
dependent on intracellular calcium levels (113). However, the extent, nature, and physiological
significance of any regulatory mechanism of this step in the metabolism of vitamin
D remains uncertain. The 25-hydroxylase has been cloned (114) and expressed in yeast
cells (115).
Vitamin D 63
2. 1?,25(OH)2D3
From the liver, 25(OH)D3 is returned to the circulatory system where it is transported
via DBP to the kidney. In the kidney, a second hydroxyl group can be added at the
C-1 position. The enzyme responsible for the 1?-hydroxylation of 25(OH)D3 is the 25-
hydroxyvitamin D3-1?-hydroxylase (1-hydroxylase) (116).
1-Hydroxylase is located in the mitochondria of the proximal tubules in the kidney.
The enzyme belongs to a class of enzymes known as mitochondrial mixed-function oxidases.
Mixed-function oxidases use molecular oxygen as the oxygen source instead of
water. 1-Hydroxylase is composed of three proteins that are integral components of the
mitochondrial membrane: they are renal ferredoxin reductase, renal ferredoxin, and a cytochrome
The most important point of regulation of the vitamin D endocrine system occurs
through the stringent control of the activity of the renal 1-hydroxylase (117). In this way,
the production of the hormone 1?,25(OH)2D3 can be modulated according to the calcium
needs of the organism. Although extrarenal production of 1?,25(OH)2D3 has been demonstrated
in placenta (118,119), cultured pulmonary alveolar and bone macrophages (120–
122), cultured embryonic calvarial cells (123), and cultured keratinocytes (124,125), which
can provide the hormone to adjacent cells in a paracrine fashion,the kidney is considered
the primary source of circulating 1?,25(OH)2D3. Several regulatory factors have been
identified that modulate 1-hydroxylase activity, but some are functional only in certain
species and under certain experimental conditions. The major factors are 1?,25(OH)2D3
itself, PTH, and the serum concentrations of calcium and phosphate (126).
Probably the most important determinant of 1-hydroxylase activity is the vitamin
D status of the animal. When circulating levels of 1?,25(OH)2D3 are low, the production
of 1?,25(OH)2D3 in the kidney is high, and when circulating levels of 1?,25(OH)2D3 are
high, synthesis of 1?,25(OH)2D3 is low (117). The changes of enzyme activity induced
by 1?,25(OH)2D3 can be inhibited by cycloheximide and actinomycin D (127), which
suggests that 1?,25(OH)2D3 is acting at the level of transcription. Another modulator of
renal 1?,25(OH)2D3 production is PTH. PTH is released when plasma calcium levels are
low, and in the kidney it stimulates the activity of the 1-hydroxylase and decreases the
activity of the 24-hydroxylase. 1?,25(OH)2D3 and 24R,25(OH)2D3 also operate in a feedback
loop to modulate and/or reduce the secretion of PTH. Other modulators of renal
1?,25(OH)2D3 production are shown in Fig. 4.
3. 24R,25(OH)2D3
A second dihydroxylated metabolite of vitamin D is produced in the kidney, namely,
24R,25(OH)2D3. Also, virtually all other tissues that have receptors for 1?,25(OH)2D3
can also produce 24R,25(OH)2D3. There is some controversy concerning the possible
unique biological actions of 24R,25(OH)2D3. However, there is some evidence that
24R,25(OH)2D3 plays a role in the suppression of PTH secretion (128,129), in the mineralization
of bone (130,131), and in fracture healing (132,133). Other studies demonstrated
that the combined presence of 24R,25(OH)2D3 and 1?,25(OH)2D3 are required for normal
egg production, fertility, and hatchability in chickens (104) and quail (134). From these
studies it was apparent that only combination doses of both compounds were capable of
eliciting the same response as the parent vitamin D. Thus, it appears that both
1?,25(OH)2D3 and 24R,25(OH)2D3 may be required for some of the known biological
responses to vitamin D.
64 Collins and Norman
The enzyme responsible for the production of 24R,25(OH)2D3 is the 25-hydroxyvitamin
D3-24R-hydroxylase (24-hydroxylase). Experimental evidence suggests that this
enzyme is also a mixed-function oxidase. The activity of this enzyme is regulated so that
when 1?,25(OH)2D3 levels are low, the activity of the 24R-hydroxylase is also low, but
when 1?,25(OH)2D3 levels are high, the activity of the 24R-hydroxylase is high. Under
normal physiological conditions both 1?,25(OH)2D3 and 24R,25(OH)2D3 are secreted
from the kidney and circulated in the plasma of all classes of vertebrates.
In addition to these three metabolites, many other vitamin D3 metabolites have been
chemically characterized, and the existence of others appears likely. The chemical struc-
Fig. 5 Summary of the metabolic transformations of vitamin D3. Shown here are the structures
of all known chemically characterized vitamin D3 metabolites.
Vitamin D 65
tures of the 37 known metabolites are shown in Fig. 5. Most of these metabolites appear
to be intermediates in degradation pathways of 1?,25(OH)2D3. None of these other metabolites
have been shown to have biological activity except for the 1?,25(OH)2D3-26,23-
lactone. The lactone is produced by the kidney when the plasma levels of 1?,25(OH)2D3
are very high. The metabolite appears to be antagonistic to 1?,25(OH)2D3 because it mediates
a decrease in serum calcium levels in the rat. Other experiments suggest that lactone
inhibits bone resorption and blocks the resorptive action of 1?,25(OH)2D3 on the bone
(135), perhaps functioning as a natural antagonist of 1?,25(OH)2D3 to prevent toxic effects
from overproduction of 1?,25(OH)2D3.
E. Catabolism and Excretion
Several pathways exist in humans and animals to further metabolize 1?,25(OH)2D3. These
include oxidative cleavage of the side chain, hydroxylation of C-24 to produce
1?,24,25(OH)3D3, formation of 24-oxo-1?,25(OH)2D3, formation of 1?,25(OH)2D3-
26,23-lactone, and formation of 1?,25,26(OH)3D3 (Fig. 5). It is not known which of these
pathways are involved in the breakdown or clearance of 1?,25(OH)2D3 in humans.
The catabolic pathway for vitamin D is obscure, but it is known that the excretion
of vitamin D and its metabolites occurs primarily in the feces with the aid of bile salts.
Very little appears in the urine. Studies in which radioactively labeled 1?,25(OH)2D3 was
administered to humans have shown that 60–70% of the 1?,25(OH)2D3 was eliminated
in the feces as more polar metabolites, glucuronides, and sulfates of 1?,25(OH)2D3. The
half-life of 1?,25(OH)2D3 in plasma has two components. Within 5 min, only half of an
administered dose of radioactive 1?,25(OH)2D3 remains in the plasma. A slower component
of elimination has a half-life of about 10 h. 1?,25(OH)2D3 is catabolized by a number
of pathways that result in its rapid removal from the organism (136).
The major classical physiological effects of vitamin D are to increase the active absorption
of Ca2 from the proximal intestine and to increase the mineralization of bone. This is
achieved via two major signal transduction pathways: genomic and nongenomic.
A. Genomic
Vitamin D, through its daughter metabolite 1?,25(OH)2D3, functions in a manner homologous
to that of steroid hormones. A model for steroid hormone action is shown in
Fig. 6. In the general model, the hormone is produced in an endocrine gland in response
to a physiological stimulus and then circulates in the blood, usually bound to a protein
carrier (i.e. DBP), to target tissues where the hormone interacts with specific,
high-affinity, intracellular receptors. The receptor–hormone complex localizes in the
nucleus, undergoes some type of ‘‘activation’’ perhaps involving phosphorylation
(137–140), and binds to a hormone response element (HRE) on the DNA to modulate
the expression of hormone-sensitive genes. The modulation of gene transcription results
in either the induction or repression of specific mRNAs, ultimately resulting in
changes in protein expression needed to produce the required biological response. Highaffinity
receptors for 1?,25(OH)2D3 have been identified in at least 26 target tissues
(12,16,141) and more than 50 genes are known to be regulated by 1?,25(OH)2D3 (142).
66 Collins and Norman
Fig. 6 General model for the mode of action of steroid hormones. Target tissues contain receptors
for the steroid which confer on them the ability to modulate gene transcription. S, steroid; R, receptor
protein, which may be present inside the cell in either the cytosol or nuclear compartment; SR,
steroid–receptor complex; DBP, serum vitamin D–binding protein, which functions to transport the
steroid hormone from the endocrine gland to its various target tissues.
Genes that have been shown to be transcriptionally regulated by 1?,25(OH)2D3 are listed
in Table 4.
1. Nuclear Receptor
The 1?,25(OH)2D3 receptor was originally discovered in the intestine of vitamin-D defi-
cient chicks (50,51). It has been extensively characterized and the cDNA for the nuclear
receptor has been cloned and sequenced (52–55). The 1?,25(OH)2D3 receptor is a DNAbinding
protein with a molecular weight of about 50,000 da. It binds 1?,25(OH)2D3 with
high affinity with a KD in the range of 1–50  1010 M (143–145). The ligand specificity
of the nuclear 1?,25(OH)2D3 receptor is illustrated in Table 5. The 1?,25(OH)2D3 receptor
protein belongs to a superfamily of homologous nuclear receptors (52). To date only a
single form of the receptor has been identified.
The superfamily of ligand-dependent nuclear receptors includes receptors for glucocorticoids
(GR), progesterone (PR), estrogen (ER), aldosterone, androgens, thyroid hormone
(T3R), hormonal forms of vitamins A (RAR, RXR) and D (VDR), and several
orphan receptors (141,146,147). Comparative studies of these receptors reveal that they
have the common structural organization consisting of five domains (148), shown in Fig.
7. The different domains act as distinct modules that can function independently of each
other (149–151).
The DNA binding domain, C, is the most conserved domain throughout the family.
About 70 amino acids fold into two zinc finger–like motifs. Conserved cysteines
coordinate a zinc ion in a tetrahedral arrangement. The first finger, which contains four
cysteines and several hydrophobic amino acids, determines the DNA response element
specificity. The second zinc finger, which contains five cysteines and many basic amino
Vitamin D 67
Table 4 Genes Regulated by 1?,25(OH)2D3
Gene Reg. Evidence Tissue/cell
?-Tubulin Down mRNA Chick intestine
Aldolase subunit B Up mRNA Chick kidney
Alkaline phosphatase Up mRNA Rat intestine
Chick intestine
TE-85 cells
ATP synthase Up mRNA Rat intestine
Chick intestine
Down mRNA Chick kidney
c-FMS Up mRNA HL-60 cells
c-FOS Up mRNA MG-63 cells
HL-60 cells
c-KI-RAS Up mRNA BALB-3T3 cells
c-MYB Down mRNA HL-60 cells
c-MYC Up mRNA MG-63
Down mRNA U937 cells
HL-60 cells Transcription
HL-60 cells
Calbindin28K Up mRNA Chick intestine
Mouse kidney Transcription
Chick intestine
Calbindin9K Up mRNA Mouse kidney
Carbonic anhydrase Up mRNA Marrow cells
Transcription Myelomonocytes
CD-23 Down mRNA PBMC
Collagen type I Down mRNA/VDRE Rat
Cytochrome oxidase sub- Up mRNA Rat intestine
unit I
Chick intestine
Down mRNA Chick kidney
Cytochrome oxidase sub- Up mRNA Chick intestine
unit II
Down mRNA Chick kidney
Cytochrome oxidase sub- Up mRNA Rat intestine
unit III
Chick intestine
Down mRNA Chick kidney
Cytochrome B Down mRNA Chick kidney
Fatty acid–binding pro- Down mRNA Chick intestine
Ferridoxin Down mRNA Chick kidney
Fibronectin Up mRNA MG-63
HL-60 cells
?-Interferon Down mRNA T lymphocytes
Glyceraldehyde-3- Up mRNA BT-20 cells
phosphate dehydrogenase
68 Collins and Norman
Table 4 Continued
Gene Reg. Evidence Tissue/cell
GM-colony-stimulating Down mRNA T lymphocytes
Heat shock protein 70 Up mRNA PBMC
Histone H4 Down mRNA/ HL-60 cells
1-Hydroxyvitamin D-24- Up mRNA Rat kidney
hydroxylase mRNA/ Rat kidney
Integrin???3 Up mRNA/ Avian osteoclast precursor cells
Interleukin-6 Up mRNA U937
Interleukin-4 Up mRNA U937 cells
Interleukin-2 Down mRNA T lymphocytes
Interleukin-3 receptor Up mRNA MC3T3 cells
Matrix gla protein Up mRNA UMR106-01, ROS 25/1, 25/4 cells
Metallothionein Up mRNA Rat kertinocytes
Mouse liver/kidney/skin
Chick kidney
Monocyte-derived neutro- Up mRNA/transcription HL-60 cells
phil-activating peptide
NADH DH subunit I Down mRNA Chick kidney
NADH DH subunit III Up mRNA Chick intestine
NADH DH subunit IV Up mRNA Chick intestine
Nerve growth factor Up mRNA L-929 cells
Osteocalcin Up mRNA ROS 17/2.8
ROS 25/1
VDRE ROS 17/2.8
Osteopontin Up mRNA ROS 17/2.8
VDRE ROS 17/2.8
Plasma membrane cal- Up mRNA Chick intestine
cium pump
Pre-pro-PTH Down mRNA Rat
mRNA/transcription Bovine parathyroid
Prolactin Up mRNA GH4C1 cells
Protein kinase inhibitor Down mRNA Chick kidney
Protein kinase C Up mRNA/transcription HL-60 cells
PTH Down mRNA Rat parathyroid
PTH-related protein Down mRNA/transcription TT cells
Transferrin receptor Down mRNA PBMC
Tumor necrosis factor ? Up mRNA U937 cells
Transcription HL-60 cells
VDR Up mRNA Rat intestine
Rat pituitary
MG-60 cells
Source: Ref. 142.
Vitamin D 69
Table 5 Ligand Specificity of the Nuclear 1?,25(OH)2D3 Receptor
Ligand Structural modification (%)a
1?,25(OH)2D3 100
1?,25(OH)2-24-nor-D3 Shorten side chain by 1 carbon 67
1?,25(OH)2-3-epi-D3 Orientation of 3?-OH altered 24
1?,25(OH)2-24a-dihomo-D3 Lengthen side chain by 2 carbons 24
1?,25(OH)2D3 Orientation of 1?-OH changed 0.8
1?(OH)D3 Lacks 25-OH 0.15
25(OH)D3 Lacks 1?-OH 0.15
1?,25(OH)2-7-dehydrocholesterol Lacks a broken B ring; is not a seco steroid 0.10
Vitamin D3 Lacks 1? and 25-OH 0.0001
aThe Relative Competitive Index (RCI) is a measure of the ability of a nonradioactive ligand to compete, under
in vitro conditions, with radioactive 1?,25(OH)2D3 for binding to the nuclear 1?,25(OH)2D3 receptor (VDR)
Source: Ref. 14.
acids, is also necessary for DNA binding and is involved in receptor dimerization
The next conserved region is the steroid binding domain (region E). This region
contains a hydrophobic pocket for ligand binding and also contains signals for several
other functions, including dimerization (154–157), nuclear translocation, and hormonedependent
transcriptional activation (149,150,158).
The A/B domain is also known as the immuno- or transactivation domain. This
region is poorly conserved in amino acids and in size, and its function has not been clearly
defined. The VDR has the smallest A/B domain (25 amino acids) of the known receptors;
mineralocorticoid receptor has the largest (603 amino acids). An independent transcriptional
activation function is located within the A/B region (146,150,151) that is constitutive
in receptor constructs lacking the ligand binding domain (region E). The relative
importance of the transcriptional activation by this domain depends on the receptor, the
context of the target gene promoter, and the target cell type (159).
Region D is the hinge region between the DNA binding domain and the ligand
bonding domain. The hinge region in the VDR contains 156 amino acids and has immunogenic
properties. The VDR has the longest hinge region of the known receptors (160).
Human GR and PR have hinge regions of 92 and 101 amino acids, respectively.
The VDR belongs to a subgroup of the receptors designated group II, which includes
T3R, RAR, RXR, and several orphan receptors. All of the group II receptors can form
heterodimers with RXR (161,162), and other heterodimeric interactions have also been
Fig. 7 Schematic representation of the human nuclear VDR. The DNA binding domain (C) and
ligand binding domain (E) are boxed.
70 Collins and Norman
reported (163). T3R lacking the DNA binding domain can inhibit the transactivation of
RAR (156) and VDR (152), but not a chimeric receptor containing the ligand binding
domain of GR. The VDR can also form heterodimers with RAR (163,164). The ability
to form heterodimers with other receptors allows for enhanced affinity for distinct DNA
targets, generating the diverse range of physiological effects.
2. Calbindin D
One of the major effects of 1?,25(OH)2D3 in many of its target tissues is the induction
of the calcium binding protein, calbindin D. In the mammalian kidney and brain and
in avians, a larger form of the protein is expressed, calbindin D28K (165), whereas in the
mammalian intestine and placenta a smaller form is expressed, calbindin D9K (166). The
expression of calbindins in various tissues and species appears to be regulated to differing
degrees by 1?,25(OH)2D3 (167).
Early experiments showed that actinomycin D and ?-amanitin, transcriptional inhibitors,
could block the induction of calbindin D28K by 1?,25(OH)2D3 (168). Later experiments
showed that 1?,25(OH)2D3 was able to stimulate total RNA synthesis in the chick
intestine (169) in addition to specifically inducing the mRNA for calbindin D28K (170).
Nuclear transcription assays have shown that transcription of calbindin D28K mRNA is
directly induced by 1?,25(OH)2D3 in the chick intestine and is correlated to the level of
occupied 1?,25(OH)2D3 receptors (171). The gene for calbindin D28K has now been cloned
and sequenced, but there is still much to learn about how 1?,25(OH)2D3 induces this gene
B. Nongenomic Actions
Recent studies (173) suggest that not all of the actions of 1?,25(OH)2D3 can be explained
by receptor–hormone interactions with the genome. 1?,25(OH)2D3 can stimulate the intestinal
transport of calcium within 4–6 min, i.e., too quickly to involve genome activation.
The rapid transport of calcium mediated by 1?,25(OH)2D3 in the intestine has been termed
‘‘transcaltachia’’ (‘‘trans’’  across; ‘‘cal’’  calcium; ‘‘tachia’’  swiftly). Transcaltachia
is not inhibited by actinomycin D but is inhibited by colchicine, an antimicrotubule
agent, and by leupeptin, an antagonist of lysosomal cathepsin B. Transcaltachia induced
by 1?,25(OH)2D3 in the intestine appears to involve the internalization of calcium in
endocytic vesicles at the brush-border membrane, which then fuse with lysosomes and
travel along microtubules to the basal lateral membrane where exocytosis occurs. Therefore,
some of the actions of 1?,25(OH)2D3 may be mediated at the cell membrane or by
extranuclear subcellular components.
Other effects of 1?,25(OH)2D3 that do not appear to be mediated by the nuclear
receptor are phosphoinositide breakdown (174), enzymatic activity in osteoblast-derived
matrix vesicles (175), certain secretion events in osteoblasts (176), rapid changes in cytosolic
Ca2 levels in primary cultures of osteoblasts and osteosarcoma cells (177–179),
and increases in cyclic guanosine monophosphate levels in fibroblasts (180). These rapid
effects appear to be mediated by a membrane receptor–like protein for 1?,25(OH)2D3
(181); a candidate membrane receptor for 1?,25(OH)2D3 has been proposed (182). Other
steroid hormones, i.e., estrogen (183), progesterone (184–187), testosterone (188), glucocorticoids
(189,190), corticosteroid (191), and thyroid (192,193), have been shown to have
similar membrane effects (181). A model for the nongenomic signal transduction pathway
is shown in Fig. 8.
Vitamin D 71
Fig. 8 Model describing the signal transduction pathways associated with the nongenomic response
of transcaltachia. The general model of vesicular Ca2 transport includes formation of Ca2-
containing endocytic vesicles at the brush-border membrane, fusion of endocytic vesicle with lysosomes,
movement of lysosomes along microtubules, and exocytotic extrusion of Ca2 via fusion of
the lysosomes with the basal lateral membrane of the intestinal enterocyte. The binding of
1?,25(OH)2D3 to a membrane receptor results in an increase of several second messengers, including
IP3, cAMP, activation of PKC or intracellular Ca2, which may result in the transient opening of
Ca2 channels. The increased Ca2 concentration may then initiate the exocytosis of the lysosomal
A. 1,25(OH)2D3 and Mineral Metabolism
The classical target tissues for 1?,25(OH)2D3 are those tissues that have been found to
be directly involved in the regulation of mineral homeostasis. In humans, serum calcium
levels are normally maintained between 9.5 and 10.5 mg/100 mL, whereas the phosphorus
concentration is between 2.5 and 4.3 mg/100 mL (2). Together with PTH and calcitonin,
1?,25(OH)2D3 maintains serum calcium and phosphate levels by its actions on the intestine,
kidney, bone, and parathyroid gland.
In the intestine, one of the best characterized effects of 1?,25(OH)2D3 is the stimulation
of intestinal lumen–to–plasma flux of calcium and phosphate (39,194,195). Although
extensive evidence exists showing that 1?,25(OH)2D3 interacting with its receptor upregulates
calbindin D in a genome-mediated fashion, the relationship between calbindin
D and calcium transport is not clear (196). In the vitamin D–deficient state, both mammals
and birds have severely decreased intestinal absorption of calcium with no detectable
72 Collins and Norman
levels of calbindin. There is a linear correlation between the increased cellular levels of
calbindin D and calcium transport. When 1?,25(OH)2D3 is given to vitamin D–deficient
chicks, the transport of calcium reaches maximal rates at 12–14 h, whereas calbindin D
does not reach its maximal levels until 48 h (173). In one study employing immunohistochemical
techniques, it was demonstrated that the cellular location of calbindin D28K
changed with the onset of calcium transport (197).
1?,25(OH)2D3 treatment also is known to alter the biochemical and morphological
characteristics of the intestinal cells (198,199). The size of the villus and the size of the
microvilli increase upon 1?,25(OH)2D3 treatment (200). The brush border undergoes noticeable
alterations of structure and composition of cell surface proteins and lipids, occurring
in a time frame corresponding to the increase in Ca2 transport mediated by
1?,25(OH)2D3 (201). However, despite extensive work, the exact mechanisms involved
in the vitamin D–dependent intestinal absorption of calcium remain unknown (202–204).
The kidney is the major site of synthesis of 1?,25(OH)2D3 and of several other
hydroxylated vitamin D derivatives. Probably the most important effect of 1?,25(OH)2D3
on the kidney is the inhibition of 25(OH)D3-1?-hydroxylase activity, which results in a
decrease in the synthesis of 1?,25(OH)2D3 (205,206). Simultaneously, the activity of the
25(OH)D3-24-hydroxylase is stimulated. The actions of vitamin D on calcium and phosphorus
metabolism in the kidney has been controversial, and more research is needed to
clearly define the actions of 1?,25(OH)2D3 on the kidney.
Although vitamin D is a powerful antirachitic agent, its primary effect on bone is
the stimulation of bone resorption leading to an increase in serum calcium and phosphorus
levels (207). With even slight decreases in serum calcium levels, PTH is synthesized,
which then stimulates the synthesis of 1?,25(OH)2D3 in the kidney. Both of these hormones
stimulate bone resorption. Maintaining constant levels of calcium in the blood is
crucial, whether calcium is available from the diet or not. Therefore, the ability to release
calcium from its largest body store—bone—is vital. Bone is a dynamic tissue that is
constantly being remodeled. Under normal physiological conditions, bone formation and
bone resorption are tightly balanced (208). The stimulation of bone growth and mineralization
by 1?,25(OH)2D3 appears to be an indirect effect of the provision of minerals for
bone matrix incorporation through an increase of intestinal absorption of calcium and
phosphorus. In bone, nuclear receptors for 1?,25(OH)2D3 have been detected in normal
osteoblasts (209), osteoblast-like osteosarcoma cells, but not in osteoclasts. In addition,
1?,25(OH)2D3 can induce rapid changes in cytosolic Ca2 levels in osteoblast and osteosarcoma
cells by opening voltage-gated Ca2 channels via a nongenomic signal transduction
pathway (178,179).
Some of the actions of 1?,25(OH)2D3 in bone are related to changes in bone cell
differentiation. 1?,25(OH)2D3 is known to affect a number of osteoblast-related functions.
For example, 1?,25(OH)2D3 decreases type I collagen production (210), and increases
alkaline phosphatase production and the proliferation of cultured osteoblasts (211);
1?,25(OH)2D3 increases the production of osteocalcin (212) and matrix Gla protein (213),
and decreases the production of type I collagen by fetal rat calvaria (214).
1?,25(OH)2D3 also affects the growth and differentiation of osteoclasts and osteoclast-
like cells in vivo in rats (215) and in primate bone marrow cell cultures (216). Since
the osteoclast does not have a nuclear VDR, 1?,25(OH)2D3 must affect osteoclasts indirectly
or by nongenomic mechanisms. There is some evidence that a factor produced by
osteoblasts promotes the formation of osteoclasts (217,218). It is possible that the only
effect of 1?,25(OH)2D3 on the osteoclasts is to stimulate its generation from progenitor
Vitamin D 73
PTH is an important tropic stimulator of 1?,25(OH)2D3 synthesis by the kidney.
High circulating levels of 1?,25(OH)2D3 have been shown to decrease the levels of PTH
by two different mechanisms: an indirect mechanism due to the resulting increase in serum
calcium levels, which is an inhibitory signal for PTH production, and a direct mechanism
involving the interaction of 1?,25(OH)2D3 and its receptor, which directly suppresses the
expression of the prepro-PTH gene.
During pregnancy and lactation, large amounts of calcium are needed for the developing
fetus and for milk production. Hormonal adjustments in the vitamin D endocrine
system are critical to prevent depletion of minerals leading to serious bone damage for
the mother. Although receptors for 1?,25(OH)2D3 have been found in placental tissue and
in the mammary gland, the role of vitamin D is not clear.
B. Vitamin D in Nonclassical Systems
In the 1970s and 1980s, nuclear receptors for 1?,25(OH)2D3 were discovered in a variety
of tissues and cells not directly involved in calcium homeostasis. Thus, the role of the
vitamin D endocrine system has expanded to include general effects on cell regulation
and differentiation (12,18). Nuclear VDRs are present in muscle, hematolymphopoietic,
reproductive, and nervous tissue, as well as in other endocrine tissues and skin. More than
50 proteins are known to be regulated by 1?,25(OH)2D3, including several oncogenes
(56,142) (Table 2), which extend by far the classical limits of vitamin D actions on calcium
homeostasis. In many of these systems it is not yet clear what the effect of vitamin D is
on the tissue or its mode of action.
Skeletal muscle is a target organ for 1?,25(OH)2D3. Clinical studies have shown
the presence of muscle weakness or myopathy during metabolic bone diseases related to
vitamin D deficiency (19,22,219). These abnormalities can be reversed with vitamin D
therapy. Experimental evidence has shown that 1?,25(OH)2D3 has a direct effect on Ca2
transport in cultured myoblasts and skeletal muscle tissue. Furthermore, there is evidence
that the action of 1?,25(OH)2D3 on skeletal muscle may be important for the calcium
homeostasis of the entire organism because the hormone induces a rapid release of calcium
from muscle into the serum of hypocalcemic animals. 1?,25(OH)2D3 receptors have been
detected in myoblast cultures, and the changes in calcium uptake have been shown to be
RNA- and protein synthesis–dependent, suggesting a genomic mechanism. 1?,25(OH)2D3
has also been shown to be important for cardiac muscle function (220–223).
In the skin, 1?,25(OH)2D3 appears to exert effects on cellular growth and differentiation.
Receptors for 1?,25(OH)2D3 have been found in human (224) and mouse skin (225).
1?,25(OH)2D3 inhibits the synthesis of DNA in mouse epidermal cells (225). The hormone
induces changes in cultured keratinocytes, which are consistent for terminal differentiation
of nonadherent cornified squamous cells (226). Additional experiments have shown that
human neonatal foreskin keratinocytes produce 1?,25(OH)2D3 from 25(OH)D3 under in
vitro conditions (227), suggesting that keratinocyte-derived 1?,25(OH)2D3 may affect epidermal
differentiation locally. Psoriasis is a chronic hyperproliferative skin disease. Some
forms of psoriasis have been shown to improve significantly when treated topically with
calcipotriol, a nonhypercalcemic analog of 1?,25(OH)2D3 (228–230). In mouse skin carcinogenesis,
1?,25(OH)2D3 blocks the production of tumors induced by 12-O-tetradecanoylphorbol-
12-acetate (231).
In the pancreas, 1?,25(OH)2D3 has been found to be essential for normal insulin
secretion. Experiments with rats have shown that vitamin D increases insulin release from
the isolated perfused pancreas, in both the presence and the absence of normal serum
74 Collins and Norman
calcium levels (232–236). Human patients with vitamin D deficiency, even under conditions
of normal calcemia, exhibit impaired insulin secretion but normal glucagon secretion,
suggesting that 1?,25(OH)2D3 directly affects ?-cell function (237).
Receptors for 1?,25(OH)2D3 have been found in some sections of the brain
(238,239). However, the role of 1?,25(OH)2D3 in the brain is not well understood. Both
calbindins D have been found in the brain, but neither the expression of calbindin D28K
nor that of calbindin D9K appears to be directly modulated by vitamin D (238,239). In the
rat, 1?,25(OH)2D3 appears to increase the activity of the choline acetyltransferase in specific
regions of the brain (238). Other steroid hormones have also been shown to affect
the metabolism of specific brain regions (240,241).
Also, normal, benign, hyperplastic and malignant prostatic epithelial and fibroblastic
cells contain receptors for 1?,25(OH)2D3 (242). In hematopoietic tissue, 1?,25(OH)2D3
promotes the differentiation and inhibits proliferation of both malignant and nonmalignant
hematopoietic cells. Human promyelocytic leukemia cells, HL-60, have been shown to
have receptors for 1?,25(OH)2D3 and to differentiate toward macrophages upon treatment
with 1?,25(OH)2D3 (243,244). Other effects of 1?,25(OH)2D3 on the immune system will
be discussed in the next section.
C. Immunoregulatory Roles
In 1979, when the VDR was discovered in several neoplastic hematopoietic cell lines as
well as in normal human peripheral blood mononuclear cells, monocytes, and activated
lymphocytes (245,246), a role for 1?,25(OH)2D3 in immune function was suggested. Since
then, 1?,25(OH)2D3 has been shown to affect cells of the immune system in a variety of
ways. 1?,25(OH)2D3 reduces the proliferation of HL-60 cells and induces their differentiation
to monocytes (243) and macrophages (244,247,248). The actions of 1?,25(OH)2D3
on normal monocytes are controversial, but it appears that the molecule may enhance
monocyte function. 1?,25(OH)2D3 appears to reduce levels of HLA-DR and CD4 class
II antigens on monocytes or macrophages with no effect on the expression of class I
antigens (249). The enhancement of class II antigen expression is a common feature of
autoimmunity and often precedes the onset of autoimmune diseases.
1?,25(OH)2D3 also promotes the differentiation of leukemic myeloid precursor cells
toward cells with the characteristics of macrophages (243). Subsequent experiments have
shown that 1?,25(OH)2D3 does not alter the clonal growth of normal myeloid precursors
but does induce the formation of macrophage colonies preferentially over the formation
of granulocyte colonies (247). In addition, macrophages derived from different tissues can
synthesize 1?,25(OH)2D3 when activated by ?-interferon (121). Also, 1?,25(OH)2D3 can
suppress immunoglobulin production by activated B lymphocytes (250) and inhibit DNA
synthesis and proliferation of both activated B and T lymphocytes (251–253). These findings
suggest that a vitamin D paracrine system exists that involves activated macrophages
and activated lymphocytes (Fig. 4).
1?,25(OH)2D3 also affects some functions of T and B lymphocytes and natural killer
(NK) cells. In T lymphocytes, the mitogen activation of lymphocyte proliferation is
blocked in the presence of 1?,25(OH)2D3 (251,254,255), apparently by interference with
cell cycle progression from early G1 to late G1 phase (254). 1?,25(OH)2D3 exhibits a
permissive or enhancing effect on T-cell suppressor activity. In an in vitro model of transplant
compatibility, the mixed lymphocyte reaction, 1?,25(OH)2D3 significantly enhanced
T-cell suppressor activity (256). 1?,25(OH)2D3 also affects the cytotoxicity of NK and
T-cytotoxic cells probably by interfering with their generation from precursor cells.
Vitamin D 75
1?,25(OH)2D3 has been shown to decrease mRNA levels for IL-2 (257), ?-interferon
(258) and granulocyte-macrophase colony-stimulating factor (GM-CSF) (259,260).
1?,25(OH)2D3 also attenuates the inducing effect of T-helper cells on IgG synthesis by
B cells (261).
Despite the wide range of actions of 1?,25(OH)2D3 on various immune cells, no
general immunomodulatory role for 1?,25(OH)2D3 has been defined. In contrast to the
serum calcium elevation observed in patients with sarcoidosis, lymphoma, and an anephric
patients with end-stage renal disease, no systemic immunosuppressive activity of
1?,25(OH)2D3 has been described in these disease states to date, suggesting that
1?,25(OH)2D3 acts in an autocrine or paracrine fashion to modulate local immune function
(261,262). 1?,25(OH)2D3 and cyclosporine, a potent immunosuppressive drug, appear to
affect the immune system in a similar fashion. They both affect T lymphocytes during
initial activation by antigen, select the generation of T-helper cells by inhibiting lymphokine
production at a genomic level, and inhibit the generation of T-cytotoxic and NK
cells. Both are involved in the enhancement of T-suppressor function, a key element in
the efficacy of cyclosporine as a drug that reduces allograft tissue rejection (263).
1?,25(OH)2D3 appears to work synergistically with cyclosporine when the two compounds
are used in combination (264,265).
The use of nonhypercalcemic 1?,25(OH)2D3 analogs can result in enhanced immunosuppressive
effects without the toxicity risks of 1?,25(OH)2D3. Because of the synergistic
effect when 1?,25(OH)2D3 is used with cyclosporine, synthetic 1?,25(OH)2D3 analogs
may be used in the treatment of autoimmune diseases (266) or for transplantation
(267) in combination with cyclosporine to reduce the toxicity of both compounds.
D. Structures of Important Analogs
In nephrectomized animals, vitamin D compounds cannot be hydroxylated at the C-1
position because the kidney is the site where this hydroxylation occurs. Researchers
found that neither vitamin D nor 25(OH)D was able to elicit a significant biological response
when administered in physiological doses to nephrectomized animals (48,268).
Also, it was noted in the 1940s and 1950s that dihydrotachysterol3 (a 5,6-trans analog
of vitamin D3) was biologically active under circumstances where the parent vitamin D
demonstrated little or no biological activity. These findings raised the question of the
functional importance of the various structural elements of the vitamin D molecule.
Studies using analogs of vitamin D have been used to address this question. The ability
of analogs to bind to the nuclear receptor for 1?,25(OH)2D3, to increase intestinal calcium
absorption (ICA) and bone calcium mobilization (BCM), and to promote cellular
differentiation are then determined. Because of recent advances in new vitamin D syntheses
described above and in Fig. 3, analogs have been synthesized with modifications
in the A ring, seco-B ring, C ring, C/D ring junction, D ring, and/or side chain (14,
The importance of the configuration of the A ring has been studied by synthesizing
5,6-trans analogs. Because of the rotation of the A ring, these analogs cannot undergo 1-
hydroxylation and have been found to be only 1/1000 as biologically effective as
1?,25(OH)2D3. The relative significance of the 3?-hydroxyl group has been assessed by
preparing analogs such as 3-deoxy-1?,25(OH)2D3. Although this analog is active in vivo,
it is interesting in that it preferentially stimulates intestinal calcium absorption over bone
calcium mobilization (270). Of all the analogs synthesized, only a few show such selective
biological activity.
76 Collins and Norman
The effect of altering the length of the side chain has been studied. The 27-nor-
25(OH)D3 and 26,27-bis-nor-25(OH)D3 are reportedly able to stimulate intestinal calcium
absorption and bone calcium mobilization in both normal and anephric rats but are 10–
100 times less active than 25(OH)D3 (271). The 24-nor-25(OH)D3 was found to have no
biological activity (272), although it was able to block the biological response to vitamin
D but not to 25(OH)D3 or 1?,25(OH)2D3. This suggests that it might have anti–vitamin
D activity.
One of the most interesting side-chain analogs of 1?,25(OH)2D3 is 1?(OH)D3. This
metabolite appears to have the same biological activity in the chick as 1?,25(OH)2D3
(273) and is approximately half as active in the rat (274). In an attempt to determine if
the biological activity of 1?(OH)D3 is the result of in vivo 25-hydroxylation, the 25-fluoro-
1?(OH)D3 derivative was prepared (275). The fluorine on C-25 prevents hydroxylation of
this carbon. The fluoro compound was found to be one-fiftieth as active as 1?,25(OH)2D3,
suggesting that 1?(OH)D3 has some activity even without 25-hydroxylation.
From such studies, the particular attributes of the structure of 1?,25(OH)2D3 that
enables it to elicit its biological responses are being defined. It is now known that the 3?-
hydroxy group does not appear to be as important for biological activity as the 1?- or
25-hydroxyl groups; the cis configuration of the A ring is preferred over the trans configuration;
and the length of the side chain appears critical, as apparently there is little tolerance
for its being shortened or lengthened.
Analogs of 1?,25(OH)2D3 have been used to study the in vivo metabolism and mode
of action of vitamin D compounds. There has also been widespread interest in developing
1?,25(OH)2D3 analogs to use as therapeutic agents in the treatment of osteoporosis, renal
osteodystrophy, cancer, immunodeficiency syndromes, autoimmune diseases, and some
skin disorders. Of particular interest are analogs that separate the calcemic effects from
the proliferation and differentiation effects of 1?,25(OH)2D3.
One of the most successful analogs in terms of separating biological activities is a
cyclopropyl derivative of 1?,25(OH)2D3, 1?,24S(OH)2-22ene-26,27-dehydrovitamin D3,
designated calcipotriol; this analog has weak systemic effects on calcium metabolism but
potent effects on cell proliferation and differentiation (276,277). It is rapidly converted
to inactive metabolites in vivo (278,279) and is 200-fold less potent than 1?,25(OH)2D3
in causing hypercalciuria and hypercalcemia in rats (276). It is as effective in binding to
the nuclear receptor as 1?,25(OH)2D3 and has similar effects on the growth and differentiation
of keratinocytes (280,281). It is currently marketed as a topical treatment for psoriasis,
a proliferative disorder of the skin (228,282–285).
Another analog that has potential as a therapeutic agent is 22-oxa-1?,25(OH)2D3.
This analog has been shown to suppress the secretion of PTH and may be useful in the
treatment of secondary hyperparathyroidism (286). It is 10 times more potent in suppressing
proliferation and inducing differentiation than 1?,25(OH)2D3, with only 1/50 to
1/100 of the in vitro bone-resorbing activity of 1?,25(OH)2D3 (287).
Still another set of analogs of 1?,25(OH)2D3 with potential therapeutic applications
are the compounds with a double bond at C-16 and/or a triple bond at C-23. The best
characterized of these compounds is 1?,25(OH)2-16ene-23yne-D3 (288–290). This analog
is 300-fold less active in intestinal calcium absorption (ICA) and bone calcium mobilization
(BCM) and 10 to 15 times less active in inducing hypercalcemia in vivo in mice than
1?,25(OH)2D3. In three leukemia models, therapy with the analog resulted in a significant
increase in survival (288,289). All of the 16-ene and or 23-yne analogs that have been
tested are equivalent or more potent than 1?,25(OH)2D3 in the induction of HL-60 cell
Vitamin D 77
differentiation and inhibition of clonal proliferation (289,291), and ten to two hundred
fold less active in ICA and BCM (291,292).
Fluorinated analogs of 1?,25(OH)2D3 have been especially useful for studying the
in vivo metabolism of 1?,25(OH)2D3. Fluorine groups have been substituted for the hydroxyls
at C-25, C-1, and C-3 to study the importance of these hydroxylations for the
biological activity of 1?,25(OH)2D3. Also, fluorine groups have been substituted for hydrogens
at C-23, C-24, and C-26 to facilitate the study of 1?,25(OH)2D3 catabolism. The
analog 1?,25(OH)2-26,26,26,27,27,27-hexafluoro-D3 has been shown to be 10 times more
potent than 1?,25(OH)2D3 in calcium mobilization, with longer lasting effects due to its
slower rate of catabolism and metabolic clearance (293). This analog is also 10 times
more potent than 1?,25(OH)2D3 in suppressing proliferation and inducing differentiation
of HL-60 cells (247,294,295).
With the exception of vitamin B12, vitamin D is the most potent of the vitamins (as defined
by the amount of vitamin required to elicit a biological response). Consequently, biological
samples and animal tissues usually contain very low concentrations of vitamin D. For
example, the circulating plasma level of vitamin D3 in humans is only 10–20 ng/mL, or
2–5  108 M (296). In order to detect such low concentrations of vitamin D, assays
that are specific for and sensitive to vitamin D and its biologically active metabolites are
A. Rat Line Test
From 1922 to 1958, the only official assay for determination of the vitamin D content of
pharmaceutical products or food was the rat line test. The term ‘‘official’’ indicates that
the reproducibility and accuracy of the assay are high enough that the results of the test
can be accepted legally. This assay, which is capable of detecting 1–12 IU (25–300 ng)
of vitamin D, is still widely used today to determine the vitamin D content of many foods,
particularly milk (297–299). The rat line test for vitamin D employs recently weaned
rachitic rats; these rats are fed a rachitogenic diet for 19–25 days until severe rickets
develops. The rats are then divided into groups of 7–10 animals and are fed diets that
have been supplemented either with a graded series of known amounts of vitamin D3 as
standards or with the unknown test sample. (Although vitamin D oils can be directly
assayed, milk, vitamin tablets, and vitamin D–fortified foods must be saponified and the
residue taken up into a suitable oil vehicle prior to assay.) The rats are maintained on
their respective diets for 7 days. The animals are sacrificed and their radii and ulnae dissected
out and stained with a silver nitrate solution. Silver is deposited in areas of bone
where new calcium has been recently deposited. The regions turn dark when exposed to
light. Thus, the effects of the unknown sample on calcium deposition in the bone can be
determined by visual comparison with the standards. Typical results for the rat line test
are shown in Fig. 9.
B. AOAC Chick Assay
Since the rat line test is done in rats, it cannot discriminate between vitamin D2 and vitamin
D3. In the chick, vitamin D3 is 10 times more potent than vitamin D2, so it is important
to accurately determine the amount of vitamin D3 in poultry feeds. The Association of
78 Collins and Norman
Fig. 9 Rat line test chart is shown in panel A. Photographs of radii sections scored according to
the line test chart are shown in panel B.
Official Analytical Chemists (AOAC) chick test was developed to specifically measure
vitamin D3 (300,301).
Groups of 20 newly hatched chicks are placed on vitamin D–deficient diets containing
added levels of vitamin D3 (1–10 IU) or the test substance. After 3 weeks on the
diet, the birds are sacrificed and the percentage of bone ash of their tibia is determined.
A rachitic bird typically has 25–27% bone ash, whereas a vitamin D–supplemented has
40–45% bone ash. This assay is not used frequently because it is time consuming and
C. Intestinal Calcium Absorption
Other biological assays have been developed that make use of the ability of vitamin D
to stimulate the absorption of calcium across the small intestine. Two basic types of assays
measure this phenomenon: those that measure the effect of the test substance on intestinal
calcium uptake in vivo (302) and those that employ in vitro methods (39,303). Each is
capable of detecting physiological quantities, i.e., 2–50 IU (50–1250 ng; 0.13–3.2 nmol),
of vitamin D.
1. In Vivo Technique
The in vivo technique for measuring intestinal calcium absorption uses rachitic chicks
that have been raised on a low-calcium (0.6%), rachitogenic diet for 3 weeks. The birds
are then given one dose of the test compound orally, intraperitoneally, or intracardially.
Twelve to 48 h later, the chicks are anesthetized and 4.0 mg of 40Ca2 and approximately
6  106 dpm 45Ca2 are placed in the duodenal loop. Thirty minutes later, the chicks are
killed by decapitation and serum is collected. Aliquots of serum are measured for 45Ca2
in a liquid scintillation counter (302).
2. In Vitro Technique
The general design of this technique is the same as that of the in vivo technique because
vitamin D activity is measured in terms of intestinal calcium transport. In these assays,
Vitamin D 79
a vitamin D standard or test compound is given orally or intraperitoneally 24–48 h before
the assay. At the time of the assay, the animals are killed and a 10-cm length of duodenum
is removed and turned inside out. A gut sac is formed by tying off the ends of the segment
so that the mucosal surface is on the outside and the serosal surface on the inside. The
everted intestinal loop is incubated with solutions of 45Ca2. The mucosal surface of the
intestine actively transports the calcium through the tissue to the serosal side. The ratio
of calcium concentration on the serosal vs. the mucosal side of the intestine is a measure
of the ‘‘active’’ transport of calcium (39,304,305). In a vitamin D–deficient animal this
ratio is 1–2.5; in a vitamin D–dosed animal it can be as high as 6–7. The chick in vivo
assay is usually preferred because of the tedious nature of preparing the everted gut sacs.
The in vitro technique is used primarily for studies with mammals rather than birds.
D. Bone Calcium Mobilization
Another assay for vitamin D activity that often is performed simultaneously with the chick
in vivo intestinal calcium absorption assay is measurement of the vitamin D–mediated
elevation of serum calcium levels. If 3-week-old rachitic chicks are raised on a zerocalcium
diet for at least 3 days before the assay and then are given a compound containing
vitamin D, their serum calcium levels will rise in a highly characteristic manner, proportional
to the amount of steroid given (302). Since there is no dietary calcium available,
the only calcium source for elevation of serum calcium is bone. By carrying out this assay
simultaneously with the intestinal calcium absorption assay, it is possible to measure two
different aspects of the animal’s response to vitamin D at the same time.
E. Growth Rate
The administration of vitamin D to animals leads to an enhanced rate of whole-body
growth. An assay for vitamin D was developed in the chick using the growth-promoting
properties of the steroid (45,306). One-day-old chicks are placed on a rachitogenic diet
and given standard doses of vitamin D3 or the test compound three times weekly. The
birds are weighed periodically, and their weight is plotted vs. age. In the absence of vitamin
D, the rate of growth essentially plateaus by the fourth week, whereas 5–10 IU of vitamin
D3 per day is sufficient to maintain a maximal growth rate in the chick. The disadvantage
of this assay is the 3- to 4-week time period needed to accurately determine the growth
F. Radioimmunoassay and Enzyme-Linked Immunosorbent Assay
for Calbindin D28K
Additional biological assays utilize the presence of calbindin D28K protein as an indication
of vitamin D activity. Calbindin D28K is not present in the intestine of vitamin D–deficient
chicks and is only synthesized in response to the administration of vitamin D. Therefore,
it is possible to use the presence of calbindin D28K to determine vitamin D activity. A
radioimmunoassay (RIA) and an enzyme-linked immunosorbent assay (ELISA), both capable
of detecting nanogram quantities of calbindin D28K, have been developed for this
purpose (307).
A comparison of the sensitivity and working range of the biological assays for vitamin
D is given in Table 6.
80 Collins and Norman
Table 6 Comparison of Sensitivity and Working Range of Biological Assays for Vitamin D
Minimal level
detectable in
Time Usual assay
required working
Assay for assay ng nmol range
Rat line test 7 d 12 0.03 25–300 ng
AOAC chick 21 d 50 0.113 50–1250 ng
Intestinal Ca2 absorption
In vivo
45Ca2 1 d 125 0.33 0.125–25 g
47Ca2 1 d 125 0.33 0.125–25 g
In vitro
Everted sacs 1 d 250 0.65 250–1000 ng
Duodenal uptake of 45Ca2 1 d 250 0.65 250–1000 ng
Bone Ca2 mobilization
In vivo 24 h 125 0.32 0.125–25 g
Body growth 21–28 d 50 0.06 50–1250 ng
Immunoassays for calcium-binding 1 d 1 0.0025 1 ng
Although considerable progress has been made in the development of chemical or physical
means to measure vitamin D, these methods at present generally lack the sensitivity and
selectivity of the biological assays. Thus, they are not adequate for measuring samples
that contain low concentrations of vitamin D. However, these physical and chemical means
of vitamin D determination have the advantage of not being as time consuming as the
biological assays and so are frequently used on samples known to contain high levels of
vitamin D.
A. Ultraviolet Absorption
The first techniques available for quantitation of vitamin D were based on the measurement
of the UV absorption at 264 nm. The conjugated triene system of double bonds in the
vitamin D seco-steroids produces a highly characteristic absorption spectra (Fig. 10). The
absorption maxima for vitamin D occurs at 264 nm, and at this wavelength the molar
extinction coefficient for both vitamins D2 and D3 is 18,300. Thus, the concentration of
an unknown solution of vitamin D can be calculated once its absorption at 264 nm is
known. Although this technique is both quick and easy, it suffers from the disadvantage
that the sample must be scrupulously purified prior to assay in order to remove potential
UV-absorbing contaminants.
B. Colorimetric Methods
Several colorimetric methods for the quantitation of vitamin D have been developed over
the years. Among these various colorimetric assays is a method based on the isomerization
of vitamin D to isotachysterol. This procedure, which employs antimony trichloride, can
detect vitamin D in the range of 1–1000 µg. Because it can detect such large amounts
Vitamin D 81
Fig. 10 Ultraviolet spectrum of provitamin D and vitamin D. Panel A illustrates the characteristic
UV absorption spectrum of provitamin D. The wavelengths of the several absorption maxima are
262, 271, 282, and 293 nm. The molar extinction coefficient at 282 nm is 11,500. Panel B illustrates
the characteristic UV spectrum of vitamin D. The molar extinction coefficient at the 265- to 265-
nm absorption maxima is 18,300.
of vitamin D, this assay is now used primarily to determine the vitamin D content of
pharmaceutical preparations and has become the official U.S. Pharmacopoeia (USP) colorimetric
assay for vitamin D3.
In this assay, three types of tubes are normally prepared: one containing the standard
vitamin D or unknown sample plus the color reagent; one containing only the solvent,
ethylene chloride; and a third containing ethylene chloride, acetic anyhdride, and the color
reagent. The absorbance at 500 nm is measured 45 s after addition of the color reagent.
82 Collins and Norman
The concentration of vitamin D in the assay tube is proportional to its absorbance, which
is corrected for the solvent blank and for the tube containing the acetic anhydride. This
procedure has been found to follow Beer’s law for solutions containing 3.25–6.5 nmol
vitamin D per milliliter assay solution. The major disadvantage to this assay is that the
copresence of vitamin A, which is often present in pharmaceutical samples along with
vitamin D, interferes with the assay. Thus, purification procedures that include adsorption
chromatography and partition chromatography are required, making this assay procedure
rather laborious and time consuming. Another disadvantage is the necessity for careful
timing of the reaction because of the short time required for the appearance of maximal
intensity of color. However, since this is the only direct chemical method routinely available,
it has widespread industrial application (308).
C. Fluorescence Spectroscopy
There have been two reports in the literature describing assays that depend on the reaction
of vitamin D with a substance capable of fluorescing (309,310). Both of these procedures
are based on the fact that an acetic anhydride–sulfuric acid solution containing vitamin
D is capable of fluorescence if the solution is activated by light of the correct wavelength.
The lower limits of detectability of this assay are the same as those of the antimony
trichloride colorimetric assay, and the same compounds that interfere with the colorimetric
assay also interfere with this fluorescence assay. As a result, this assay is not normally
used for the analytical determination of vitamin D concentration.
D. Gas Chromatography–Mass Spectrometry
One of the most powerful modern techniques available to steroid chemists for the analytical
determination of samples containing mixtures of steroids is mass spectrometry (MS),
or mass spectrometry coupled with prior separation by gas chromatography (GC). The
gas chromatography–mass spectrometry (GC-MS) technique can be coupled to an on-line
computer that collects information on the fragmentation patterns of steroids in the mass
spectrometer. In this way a sophisticated quantitative assay can be developed with a sensitivity
and selectivity approaching that of RIAs. There have recently appeared several GCMS
procedures applicable to vitamin D seco-steroids (311), but they are not yet widely
E. High-Performance Liquid Chromatography
Several papers that describe the separation of vitamin D and its various metabolites by
HPLC have appeared (312,313). This separation process has an exceedingly high resolving
capability due to the large number of theoretical plates present in a typical column. Of
equal importance to this technique is the sensitivity of the detector used for observing the
separated compounds. All of the published procedures for the separation of vitamin D by
HPLC have used an UV detector, and so their sensitivity is limited to approximately 5
ng. The chief advantage of using HPLC is the reduction in labor and time required to
separate vitamin D and its metabolites. In the current official USP method for the determination
of vitamin D, two prepurification steps, requiring up to 8 h, are necessary before
the colorimetric analysis can be performed (314). However, with HPLC, reproducible
separation of closely related compounds can be achieved in less than 1 h. HPLC has great
Vitamin D 83
potential once a more sensitive detection method is developed. There are now available
several official AOAC procedures for the HPLC determination of vitamin D in a variety
of sample types, including vitamin preparations, multivitamin preparations, vitamin D oil
concentrates, and fortified milk and milk powder (308).
F. Competitive Binding Assays
Various competition assays that can specifically quantitate the levels of 25(OH)D3,
24,25(OH)2D3, or 1?,25(OH)2D3 in a sample are now available. Such assays were developed
as a consequence of the discovery of specific vitamin D–binding proteins in the
serum and tissues of mammals and birds, along with the availability of high specific activity
tritiated 25(OH)D3 and 1?,25(OH)2D3. Since these steroid competition assays are sensitive
(they are capable of detecting picogram quantities), they are now routinely used to
measure vitamin D metabolite levels in plasma.
Two different types of steroid competition assays have been developed for the detection
of 1?,25(OH)2D3. The first employs incubation of intestinal mucosal cytosol plus the
nuclear chromatin fractions with standardized amounts of tritiated 1?,25(OH)2D3. The
1?,25(OH)2D3 in the sample competes with the tritiated hormone for the binding sites of
the 1?,25(OH)2D3 receptors present in the cytosol and chromatin fractions. By measuring
the amount of tritiated 1?,25(OH)2D3 bound to the receptor, the amount of 1?,25(OH)2D3
in the sample can be determined. The first such assay developed that could be used to
measure 1?,25(OH)2D3 levels in plasma was that of Brumbaugh and Haussler (315). Their
technique requires a minimum of 10 mL plasma and involves a laborious three-stage
chromatographic procedure. The final 1?,25(OH)2D3 peak is then assayed. Separation of
bound from free steroid is achieved by filtration of the incubation media through glass
filters. The steroid associated with the chromatin-cytosol receptor is specifically bound to
these filters. A similar assay has been described by Procsal et al., except that separation
of the bound from free steroid is achieved by high-speed differential centrifugation (316).
The second type of competition assay involves the use of calf thymus cytosol as
the source of binding protein. Reinhardt et al. developed a radioreceptor assay for vitamin
D2, vitamin D3, and their metabolites that does not require HPLC (317). Their technique
includes the use of a stable 1?,25(OH)2D3 receptor preparation from calf thymus, nonequilibrium
assay conditions, and solid-phase extraction of vitamin D metabolites from serum
or plasma samples. This procedure requires 0.2–1.0 mL plasma. 1?,25(OH)2D3 is removed
from the plasma on a C18-silica cartridge. The cartridge is first reverse-phase-eluted and
then switched to normal-phase elution (318). The 1?,25(OH)2D3 is recovered and incubated
with [3H]1?,25(OH)2D3 and reconstituted thymus receptor. Separation of receptorbound
hormone from free hormone is achieved by the addition of dextran-coated charcoal.
Similar assays have been developed for 25(OH)D3 and 24,25(OH)2D3 (146,319,320). This
assay has a sensitivity of 0.7 pg.
A. Humans
The vitamin D requirement for healthy adults has never been precisely defined. Since
vitamin D3 is produced in the skin upon exposure to sunlight, a human being does not
84 Collins and Norman
have a requirement for vitamin D when sufficient sunlight is available. However, vitamin
D does become an important nutritional factor in the absence of sunlight. In addition to
geographical and seasonal factors, UV light from the sun may be blocked by factors such
as air pollution and sun screens. In fact, as air pollution became prevalent during the
industrial revolution, the incidence of rickets became widespread in industrial cities. It is
now known that the rickets epidemic was partly caused by lack of sunlight due to air
pollution. Thus, rickets has been called the first air pollution disease. Any condition that
blocks sunlight from skin, such as the wearing of clothes, use of sunscreens, living indoors
and in cities with tall buildings, or living in geographical regions of the world that do not
receive adequate sunlight, can contribute to the inability of the skin to biosynthesize suffi-
cient amounts of vitamin D. Under these conditions, vitamin D becomes a true vitamin
in that it must be supplied in the diet on a regular basis.
Since vitamin D can be endogenously produced by the body and since it is retained
for long periods of time by vertebrate tissue, it is difficult to determine minimum daily
requirements for this substance. The requirement for vitamin D is also dependent on the
concentration of calcium and phosphorus in the diet, the physiological stage of development,
age, gender, degree of exposure to the sun, and the amount of pigment in the
The current allowance of vitamin D recommended in 1989 by the National Research
Council is 300 IU/day [1 IU  0.025 µg vitamin D (321)] for infants from birth to 6
months in age; 400 IU/day for children, adolescents, and pregnant and lactating women;
and 200 IU/day for other adults (322). Because rickets is more prevalent in preschool
children, the Food and Agricultural Organization/World Health Organization (FAO/
WHO) committee recommended that children receive 400 IU/day until the age of 6 years,
after which the recommended daily allowance (RDA) is 100 IU/day.
In the United States, adequate amounts of vitamin D can readily be obtained from
the diet and from casual exposure to sunlight. However, in some parts of the world where
food is not routinely fortified and sunlight is often limited, obtaining adequate amounts
of vitamin D becomes more of a problem. As a result, the incidence of rickets in these
countries is higher than in the United States. Rickets was practically eradicated from the
United States in the mid-1920s, but it has reappeared in the past two decades. The increase
is predominantly due to changes in infant feeding and dietary preferences. Of 27 cases
of infant rickets reported within the first year of life, 26 were found in breast-fed infants
(323). Of 62 cases involving older children, 56 were from families following strict vegetarian
diets that included no meat products, milk products, fish, or eggs.
B. Animals
The task of assessing the minimum daily vitamin D requirement for animals is no easier
than it is for humans. Such factors as the dietary calcium/phosphorus ratio, physiological
stage of development, gender, amount of fur or hair, color, and perhaps even breed, all
affect the daily requirement for vitamin D in animals. Also, some animals, such as chicken
and turkey, do not respond as well to vitamin D2 as to vitamin D3. As with humans,
animals that are maintained in sunlight can produce their own vitamin D, so that dietary
supplementation is not really necessary. For animals that are kept indoors or that live in
climates where the sunlight is not adequate for vitamin D production, the vitamin D content
of food becomes important. Sun-cured hays are fairly good sources of vitamin D, but
Vitamin D 85
Table 7 Vitamin D Requirements of Animals
Animal Daily requirements (IU)
Chickens, growing 90a
Dairy cattle:
Calves 660b
Pregnant, lactating 5000–6000c
Growing puppies 22d
Adult maintenance 11d
Ducks 100d
Monkey, growing
rhesus 25d
Mouse, growing 167d
Lambs 300e
Adults 250e
Breed sows 550c
Lactating sows 1210c
Young boars 690c
Adult boars 550c
Turkeys 400a
aIU required per pound of feed.
bIU required for 100 kg body weight.
cIU required per animal.
dIU required per kg body weight.
eIU required per 45 kg body weight.
Source: Published information by the Committee of Animal
Nutrition, Agricultural Board (National Research Council).
dehydrated hays, green feeds, and seeds are poor sources. A brief list of the RDAs for
animals is given in Table 7.
For the most part, vitamin D is present in unfortified foods in only very small and variable
quantities (Table 8). The vitamin D that occurs naturally in unfortified foods is
generally derived from animal products. Saltwater fish, such as herring, salmon, and sardine,
contain substantial amounts of vitamin D, and fish liver oils are extremely rich
sources. However, eggs, veal, beef, unfortified milk, and butter supply only small quantities
of the vitamin. Plants are extremely poor sources of vitamin D; fruits and nuts contain
no vitamin D, and vegetable oils contain only negligible amounts of the provitamin. As
a consequence, in the United Stated dietary requirements for vitamin D are met by the
artificial fortification of suitable foods. Among these fortified foods are milk, both fresh
and evaporated; margarine and butter; cereals; and chocolate mixes. Milk is fortified to
supply 400 IU vitamin D per quart, and margarine usually contains 2000 IU or more per
pound. A more complete listing of the vitamin D values of food is given by Booher et
al. (324).
86 Collins and Norman
Table 8 Vitamin D Content of Unfortified Foods
Vitamin D
Food source (IU/100 g)
Beef steak 13
Beet greens 0.2
Butter 35
Cabbage 0.2
Cheese 12
Cod 85
Cod liver oil 10,000
Corn oil 9
Cream 50
Egg yolk 25
Herring (canned) 330
Herring liver oil 140,000
Beef (raw) 8–40
Calf (raw) 0–15
Pork (raw) 40
Chicken (raw) 50–65
Lamb (raw) 20
Mackerel 120
Cow (100 mL) 0.3–4
Human (100 mL) 0–10
Salmon (canned) 220–440
Sardines (canned) 1500
Shrimp 1150
Spinach 0.2
Source: Refs. 2 and 324.
A. Humans
A deficiency of vitamin D results in inadequate intestinal absorption and renal reabsorption
of calcium and phosphate. As a consequence, serum calcium and phosphate levels fall
and serum alkaline phosphatase activity increases. In response to these low serum calcium
levels, hyperparathyroidism occurs. The result of increased levels of PTH, along with
whatever 1?,25(OH)2D3 is still present at the onset of the deficiency, is the demineralization
of bone. This ultimately leads to rickets in children and osteomalacia in adults. The
classical skeletal symptoms associated with rickets, i.e., bowlegs, knock-knees, curvature
of the spine, and pelvic and thoracic deformities (Fig. 11), result from the application of
normal mechanical stress to demineralized bone. Enlargement of the bones, especially in
the knees, wrists, and ankles, and changes in the costochondral junctions also occur. Since
in children bone growth is still occurring, rickets can result in epiphysial abnormalities
not seen in adult osteomalacia. Rickets also results in inadequate mineralization of tooth
Vitamin D 87
Fig. 11 Classic appearance of rickets in a child.
enamel and dentin. If the disease occurs during the first 6 months of life, convulsions and
tetany can occur. Few adults with osteomalacia develop tetany.
Low serum calcium levels in the range of 5–7 mg per 100 mL and high serum
alkaline phosphatase activity can be used to diagnose rickets and osteomalacia. Also, a
marked reduction in circulating 1?,25(OH)2D3 levels in individuals with osteomalacia or
rickets has been reported (325–327). Radiographic changes are also evident and can be
used in diagnosis.
B. Animals
The response to vitamin D deficiency in animals closely resembles that in humans. Among
the first symptoms of the deficiency is a decline in the plasma concentration of calcium
and phosphorus. This is followed by an abnormally low growth rate and the characteristic
alteration of bones, including faulty calcification of the bone matrix. As the disease progresses,
the forelegs bend sideways and the joints become swollen. In laying birds, the
eggs are thin-shelled, egg production declines, and hatchability is markedly reduced (328);
classic symptoms of rickets develop, followed by tetany and death.
88 Collins and Norman
Excessive amounts of vitamin D are not available from natural sources. However, vitamin
D intoxication is a concern in patients being treated with vitamin D or vitamin D analogs
for hypoparathyroidism, vitamin D–resistant rickets, renal osteodystrophy, osteoporosis,
psoriasis, some cancers, or in those who are taking supplemental vitamins. Hypervitaminosis
D is a serious problem because it can result in irreversible calcification of the heart,
lungs, kidneys, and other soft tissues. Therefore, care should be taken to detect early
signs of vitamin D intoxication in patients receiving pharmacological doses. Symptoms
of intoxication include hypercalcemia, hypercalciuria, anorexia, nausea, vomiting, thirst,
polyuria, muscular weakness, joint pains, diffuse demineralization of bones, and disorientation.
If allowed to go unchecked, death will eventually occur.
Vitamin D intoxication is thought to occur as a result of high 25(OH)D levels rather
than high 1?,25(OH)2D levels (329,330). Patients suffering from hypervitaminosis D have
been shown to exhibit a 15-fold increase in plasma 25(OH)D concentration as compared
to normal individuals. However, their 1?,25(OH)2D levels are not substantially altered
(331). Furthermore, anephric patients can still suffer from hypervitaminosis D even though
they are for the most part incapable of producing 1?,25(OH)2D. It has also been shown
that large concentrations of 25(OH)D can mimic the actions of 1?,25(OH)2D at the level
of the receptor (315,329,332,333).
In the early stages of intoxication, the effects are usually reversible. Treatment consists
of merely withdrawing vitamin D and perhaps reducing dietary calcium intake until
serum calcium levels fall. In more severe cases, treatment with glucocorticoids, which
are thought to antagonize some of the actions of vitamin D, may be required to facilitate
the correction of hypercalcemia. Since calcitonin can bring about a decline in serum calcium
levels, it may also be used in treatment.
A. Disease
In view of the complexities of the vitamin D endocrine system, it is not surprising that
many disease states are vitamin D–related. Figure 12 classifies some of the human disease
states that are believed to be associated with vitamin D metabolism according to the metabolic
step where the disorder occurs.
1. Intestinal Disorders
The intestine functions as the site of dietary vitamin D absorption and is also a primary
target tissue for the hormonally active 1?,25(OH)2D3. Impairment of intestinal absorption
of vitamin D can occur in those intestinal disorders that result in the malabsorption of
fat. Patients suffering from such disorders as tropical sprue, regional enteritis, and multiple
jejunal diverticulosis often develop osteomalacia because of what appears to be a malabsorption
of vitamin D from the diet (334). Surgical conditions, such as gastric resection
and jejunoileal bypass surgery for obesity, may also impair vitamin D absorption. Also,
patients receiving total parenteral nutrition in the treatment of the malnutrition caused by
profound gastrointestinal disease often develop bone disease (335).
On the other hand, intestinal response to vitamin D can be affected by certain disease
states. Patients suffering from idiopathic hypercalciuria exhibit an increased intestinal
Vitamin D 89
Fig. 12 Human disease states related to vitamin D. PTH, parathyroid hormone; CT, calcitonin;
VDRR, vitamin D–resistant rickets; Pi, inorganic phosphate; Ca2, calcium.
absorption of calcium that may result from an enhanced intestinal sensitivity to
1?,25(OH)2D3 or from an overproduction of 1?,25(OH)2D3. The disease sarcoidosis also
results in enhanced sensitivity to vitamin D. Sarcoidosis is characterized by hypercalcemia
and hypercalciuria in patients receiving only modest amounts of vitamin D. Experiments
have shown that these patients have elevated levels of serum 1?,25(OH)2D3. The excess
1?,25(OH)2D3 is likely of extrarenal origin and therefore not regulated by circulating
levels of PTH (336). Other experiments have shown that macrophages from patients with
sarcoidosis can produce 1?,25(OH)2D3 (120).
Other disease states that can result in extrarenal production of 1?,25(OH)2D3 are
tuberculosis (337), leprosy (338), and some lymphomas (339).
2. Liver Disorders
The liver plays an important role in the vitamin D endocrine system; not only is it the
primary site for the production of 25(OH)D, but it is also the source of bile salts that aid
in the intestinal absorption of vitamin D. Furthermore, it is likely that the liver is the site
where binding of 25(OH)D by vitamin D-binding protein occurs, and it may even be the
site at which this binding protein is synthesized. Hence, malfunctions of the liver can
possibly interfere with the absorption, transport, and metabolism of vitamin D. Malabsorption
of calcium and the appearance of bone disease have been reported in patients suffering
from either primary biliary cirrhosis or prolonged obstructive jaundice. The disappearance
of radioactive vitamin D from the plasma of these patients is much slower than in normal
humans (340), and their plasma 25(OH)D levels are reduced (341). Although these patients
respond poorly to vitamin D treatment, they immediately respond if treated with
90 Collins and Norman
25(OH)D3. Thus, it appears that the bone disease experienced by these patients results
from their inability to produce 25(OH)D.
3. Renal Disorders
The kidney functions as the endocrine gland for 1?,25(OH)2D3. Thus, disease states that
affect the kidney can concomitantly alter the production of this calcium homeostatic hormone.
It is well known that patients suffering from renal failure also often suffer from
skeletal abnormalities. Termed renal osteodystrophy, these skeletal abnormalities include
growth retardation, osteitis fibrosa, osteomalacia, and osteosclerosis. It became apparent
with the discovery that under normal conditions 1?,25(OH)2D3 is produced in the kidney
that these skeletal abnormalities result from the failure of patients to produce
1?,25(OH)2D3. Support for this theory came from studies on the metabolism of radioactively
labeled vitamin D in normal persons vs. patients with chronic renal failure.
From these studies, anephric or uremic individuals appeared incapable of producing
1?,25(OH)2D3. Direct evidence for this came from the observation that circulating level
of 1?,25(OH)2D3 in the normal subject is in the range of 30–35 pg/mL, whereas in chronic
renal failure the levels have been reported as low as 3–6 pg/mL (342,343). However,
after a successful renal transplant, 1?,25(OH)2D3 levels return to the normal range. Also,
the administration of 1?,25(OH)2D3 to these patients results in the stimulation of intestinal
calcium absorption and an elevation of serum calcium levels (344).
4. Parathyroid Disorders
As previously outlined, PTH influences the production of 1?,25(OH)2D3, so that any disease
state that affects the secretion of PTH may, in turn, have an effect on the metabolism
of vitamin D. Hyperactivity of the parathyroid glands, as in primary hyperparathyroidism,
results in the appearance of bone disease resembling osteomalacia. Circulating
1?,25(OH)2D3 levels in these subjects have been reported to be significantly elevated
(345), as is their intestinal calcium transport (346). On the other hand, in hypoparathyroidism,
hypocalcemia occurs. In these patients, a slight reduction in circulating
1?,25(OH)2D3 levels has been reported (347). When these patients are treated with
1?,25(OH)2D3 their serum levels are increased to normal.
There are several published reviews on the role of vitamin D in disease (18,334,348).
B. Genetics
Vitamin D–resistant rickets (hypophosphatemic rickets) appears to be an X-linked, dominant
genetic disorder. Winters et al. presented evidence that this disease is almost always
inherited and is usually congenital (349). Males are usually more severely affected by this
disease than females. Associated with the disease are skeletal abnormalities, such as rickets
or osteomalacia, and a diminished renal tubular reabsorption of phosphate that results in
hypophosphatemia. Individuals with this disease do not respond to physiological doses
of vitamin D; treatments with 25(OH)D3 and 1?,25(OH)2D3 are also ineffective, although
an increase in intestinal calcium absorption does occur (350). These patients have also
been reported to have normal serum 1?,25(OH)2D3 levels. Thus, it appears that this disorder
does not result from an alteration in the metabolism of vitamin D or from an impaired
intestinal response to 1?,25(OH)2D3, but rather from a specific defect in renal tubular
reabsorption of phosphate.
Vitamin D 91
A genetic defect that interferes with vitamin D metabolism has also been suggested
in vitamin D–dependent rickets type I. This ailment differs from rickets in that it appears
in children who are receiving adequate amounts of vitamin D and requires pharmacological
doses of vitamin D or 25(OH)D3 to reverse the harmful effect of disease on bone. However,
the disease is responsive to physiological amounts of 1?,25(OH)2D3, which suggests that
the defect occurs in the metabolism of 25(OH)D3 to 1?,25(OH)2D3. This disease state
appears to be the result of an autosomal recessively inherited genetic defect (351). It is
not known how this defect affects the metabolism of 25(OH)D3.
Vitamin D–dependent rickets type II also has a genetic basis. This ailment is similar
to vitamin D–dependent rickets type I except that children do not respond to large doses
of vitamin D, 25(OH)D3, or 1?,25(OH)2D3. The combination of symptoms, i.e., defective
bone mineralization, decreased intestinal calcium absorption, hypocalcemia, and increased
serum levels of 1?,25(OH)2D3, suggest end-organ resistance to the action of
1?,25(OH)2D3. Experiments have shown that these children have a single-point mutation
in the nuclear receptor for 1?,25(OH)2D3 (352–355).
C. Drugs
Recent evidence suggests that prolonged use of anticonvulsant drugs, such as diphenylhydantoin
or phenobarbital, can result in an impaired response to vitamin D; this results in
an alteration of calcium metabolism and the appearance of rickets or osteomalacia. Serum
25(OH)D3 levels in patients receiving these drugs have been reported to be markedly
reduced (356). Also, studies in animals suggest that these drugs stimulate the hepatic
microsomal cytochrome P450 enzymes, which could lead to an increased catabolism of
25(OH)D3 (357). However, 1?,25(OH)2D3 levels have been shown to be normal or even
increased after drug treatment (358). It appears that this drug-induced osteomalacia may
not be the result of an effect of the drug on vitamin D metabolism. Studies on rat and
chick duodena in organ culture indicate that anticonvulsant drugs may act on the gastrointestinal
tract and affect the absorption of calcium (359). Anticonvulsant drugs have also
been shown to inhibit calcium reabsorption in organ culture mouse calvaria (360). Further
research is needed to determine the mechanism by which these anticonvulsant drugs affect
calcium metabolism.
D. Alcohol
Persons suffering from chronic alcoholism exhibit a decrease in plasma 25(OH)D3 levels
and in intestinal calcium absorption and bone mineral content. This is observed in patients
with and without cirrhosis of the liver. Current evidence indicates that the impairment of
intestinal calcium absorption is the result of low 25(OH)D3 levels (361,362). However,
how chronic alcoholism results in low 25(OH)D3 levels is at present not understood.
E. Age
The fact that changes in the metabolism of vitamin D may occur with aging has been
suggested by the observation that the ability to absorb dietary calcium decreases with age
(363). In addition, loss of bone increases in the elderly and an age-related hypoplasia of
bone cells occurs. 1?,25(OH)2D3 levels in the plasma decrease with age, possibly due to
an age-related reduction of epidermal concentrations of 7-dehydrocholesterol, resulting in
less photochemical production of 1?,25(OH)2D3 by the skin. Also, the 1?-hydroxylase
enzyme is less responsive to induction by PTH due to the decrease of glomerular filtration
with age (364).

Gray et al. demonstrated that men and women differ in their metabolism of vitamin D in
response to various physiological stimuli (365). In women, they observed that dietary
phosphate deprivation resulted in a decrease in serum phosphorus levels with a concomitant
increase in plasma 1?,25(OH)2D3 concentrations. However, no change of either of
these parameters was noted in men. Thus, the mechanism by which men and women
respond to dietary phosphate deprivation seems to differ.
Several ailments are known to respond to massive doses of vitamin D. For example, the
intestinal malabsorption of calcium that results from chronic renal failure and the subsequent
development of rickets or osteomalacia can be overcome by administration of
100,000–300,000 IU vitamin D per day (364). Patients suffering from hypoparathyroidism
can usually be treated by giving 80,000–100,000 IU vitamin D per day (366). Also, children
afflicted with vitamin D–dependent rickets type I can be treated with 10,000–100,000
IU per day (2). The therapeutic effect of such massive doses can be explained by the fact
that 25(OH)D in sufficiently high concentrations will mimic the action of 1?,25(OH)2D3
at the receptor. However, as mentioned earlier, the administration of such pharmacological
doses of vitamin D to patients over a prolonged period of time carries with it the danger
of vitamin D toxicity.
Table 9 shows the drug forms of vitamin D metabolites that are currently available
for the treatment of several disease states. Experiments are in progress to develop additional
vitamin D analogs that will be useful pharmacological agents without the hypercalcemic
side effect of 1?,25(OH)2D3.
Recently, the x-ray crystal structure of the ligand-binding domain (LBD) of VDR (367)
and several other nuclear receptors have been determined. These LBDs are composed
primarily of 11–13 ?-helices that are folded to form three layers. The interior of the LBD
is composed primarily of hydrophobic residues designed to bind lipophilic ligands like
1,25(OH)D3. Ligand binding to its cognate receptor induces conformational changes in
the receptor, which increases its ability to activate gene transcription. The major structural
difference between unoccupied and ligand-bound nuclear receptors is the repositioning of
the C-terminal helix 12 containing the ligand-dependent transactivation domain, AF-2.
Upon ligand binding, helix 12 moves from projecting away from the LBD to a position
tightly packed against helix 3 of the LBD (368,369). Repositioning of the AF-2 domain
results in the formation of new surfaces, including a hydrophobic cleft (370), needed to
interact with other nuclear factors called coactivators. Coactivators are adapter proteins
that link transcriptional activators such as nuclear receptors to basal transcription machinery
resulting in increased gene transcription (371).
Three families of coactivator proteins are known to interact with nuclear receptors:
SRC-1 (372)/NcoA-1 (373), TIF2 (374)/GRIP-1 (375), and ACTR/pCIP (376). These
94 Collins and Norman
proteins contain a conserved leucine-containing motif (LXXLL), important for interactions
between coactivators and nuclear receptors (377). The coactivator complexes have activities
associated with them such as histone acetyltransferase activity (378) or methyl transferase
activity (379), which aid in the decondensation of chromatin to facilitate binding
of RNA polymerase II transcription complex (RNA Pol II) to DNA. Some of the proteins
in the coactivator complex can also interact with other proteins that are part of the RNA
Pol II core complex. In this way, the coactivators can link upstream activators such as
nuclear receptors to RNA Pol II and modulate gene transcription (380). Identifying these
intermediary factors involved in transcriptional control of specific target genes is essential
for understanding the biological actions of vitamin D and will continue to be an exciting
area of research for several years.
Vitamin D, through its hormonally active metabolite 1?,25(OH)2D3, is known to act on
bone and intestine to maintain calcium homeostasis. However, the actions of
1?,25(OH)2D3 are much broader than was originally thought. There is evidence that
1?,25(OH)2D3 is involved in the physiology of tissues not related to calcium homeostasis,
such as skin, pancreas, pituitary, muscle, and hematopoietic cells. Figure 4 demonstrates
the complexity of the vitamin D endocrine system as it is understood today. Although
many advances have been made in the past decade, there is still much to learn about the
detailed cellular and molecular mode of action of vitamin D and its metabolites.
1. A. W. Norman, 1,25-(OH)2-D3 as a steroid hormone, in Vitamin D: Molecular Biology and
Clinical Nutrition (A. W. Norman, ed.), Marcel Dekker, New York, 1980, pp. 197–250.
2. A. W. Norman, Vitamin D: The Calcium Homeostatic Steroid Hormone, Academic Press,
New York, 1979, pp. 1–490.
3. D. E. M. Lawson (ed.), Vitamin D, Academic Press, New York, 1978, pp. 1–433.
4. H. F. DeLuca, Vitamin D metabolism and function, Arch Intern Med 138:836 (1978).
5. H. E. Harrison, Vitamin D and calcium and phosphate transport, Pediatrics 28:531 (1961).
6. D. B. N. Lee, M. W. Walling, G. V. Silis, and J. W. Coburn, Calcium and inorganic phosphate
transport in rat colon, J Clin Invest 65:1326 (1980).
7. A. W. Norman, K. Schaefer, H.-G. Grigoleit, and D. V. Herrath (eds.), Vitamin D: Molecular,
Cellular and Clinical Endocrinology, 7th ed., Walter de Gruyter, Berlin, 1988, pp. 1–
8. A. W. Norman, R. Bouillon, and M. Thomasset (eds.), Vitamin D: Gene Regulation, Structure
Function Analysis and Clinical Application, 8th ed., Walter de Gruyter, Berlin, 1991, pp. 1–
9. A. W. Norman, R. Bouillon, and M. Thomasset (eds.), Vitamin D, A Pluripotent Steroid
Hormone: Structural Studies, Molecular Endocrinology and Clinical Applications, 9th ed.,
Walter de Gruyter, Berlin, 1994, pp. 1–973.
10. M. R. Haussler, Vitamin D receptors: nature and function, Ann Rev Nutr 6:527 (1986).
11. M. R. Haussler, D. J. Mangelsdorf, B. S. Komm, C. M. Terpenning, K. Yamaoka, E. A.
Allegretto, A. R. Baker, J. Shine, D. P. McDonnell, M. Hughes, N. L. Weigel, B. W. O’Malley,
and J. W. Pike, Molecular biology of the vitamin D hormone, in Recent Progress in
Hormone Research (R. O. Greep, ed.), Academic Press, New York, 1988, pp. 263–305.
12. M. R. Walters, Newly identified actions of the vitamin D endocrine system, Endocr Rev 3:
719 (1992).
Vitamin D 95
13. D. D. Bikle and S. Pillai, Vitamin D, calcium, and epidermal differentiation, Endocr Rev
14:3 (1993).
14. R. Bouillon, W. H. Okamura, and A. W. Norman, Structure–function relationships in the
vitamin D endocrine system, Endocr Rev 16:200 (1995).
15. S. Christakos, C. Gabrielides, and W. B. Rhoten, Vitamin D–dependent calcium binding
proteins: chemistry, distribution, functional considerations, and molecular biology, Endocr
Rev 10:3 (1989).
16. A. W. Norman, J. Roth, and L. Orci, The vitamin D endocrine system: steroid metabolism,
hormone receptors and biological response (calcium binding proteins), Endocr Rev 3:331
17. R. E. Weishaar and R. U. Simpson, The involvement of the endocrine system in regulating
cardiovascular function: emphasis on vitamin D3. Endocr Rev 10:351 (1989).
18. H. Reichel, H. P. Koeffler, and A. W. Norman, The role of the vitamin D endocrine system
in health and disease, N Engl J Med 320:980 (1989).
19. R. L. Boland, Role of vitamin D in skeletal-muscle function, Endocr Rev 7:434 (1986).
20. R. S. Soleki, Shanidar: The Humanity of Neanderthal Man, Knopf, New York, 1971, pp. 1–220.
21. D. Whistler, Morbo puerli Anglorum, quem patrio idiomate indigenae voacant ‘‘the rickets,’’
1645, pp. 1–13.
22. F. Glisson, A treatise of the rickets being a disease common to children, 1660, pp. 1–373.
23. A. F. Hess, Rickets Including Osteomalacia and Tetany, Lea & Febiger, Philadelphia, 1929,
pp. 1–485.
24. K. Huldschinsky, Heilung von Rachitis durch kuenstliche Hohensonne, Dtsch Med Wschr
45:712 (1919).
25. E. Mellanby and M. D. Cantab, Experimental investigation on rickets, Lancet 196:407 (1919).
26. E. Mellanby, Experimental rickets, Spec Rep Ser Med Res Council (GB) SRS-61:1–78
27. E. V. McCollum, N. Simmonds, H. T. Parsons, P. G. Shipley, and E. A. Park, Studies on
experimental rickets. I. The production of rachitis and similar disease in the rat by deficient
diets, J Biol Chem 45:333 (1920).
28. E. V. McCollum, N. Simmonds, J. E. Becker, and P. G. Shipley, Studies on experimental
rickets. XXI. An experimental demonstration of the existence of a vitamin which promotes
calcium deposition, J Biol Chem 53:293 (1922).
29. H. Goldblatt and K. N. Soames, A study of rats on a normal diet irradiated daily by the
mercury vapor quartz lamp or kept in darkness, Biochem J 17:294 (1923).
30. A. F. Hess and M. Weinstock, The antirachitic value of irradiated cholesterol and phytosterol.
II. Further evidence of change in biological activity, Meth Enzymol 64:181 (1925).
31. A. F. Hess and M. Weinstock, The antirachitic value of irradiated cholesterol and phytosterol.
III. Evidence of chemical change as shown by absorption spectra, Meth Enzymol 64:193
32. A. Windaus, O. Linsert, A. Luttringhaus, and G. Weidlinch, Uber das krystallistierte Vitamin
D2, Justus Liebigs Ann Chem 492:226 (1932).
33. T. C. Angus, F. A. Askew, R. B. Bourdillon, H. M. Bruce, R. Callow, C. Fischmann, L.
Philpot, and T. A. Webster, A crystalline antirachitic substance, Proc R Soc Ser. B 108:340
34. A. Windaus, F. Schenck, and v.f. Werder, Uber das antirachitisch wirksame Bestrahlungsprodukt
aus 7-dehydro-cholesterin, H -S Zeit Physiol Chem 241:100 (1936).
35. A. Carlsson and B. Lindquist, Comparison of intestinal and skeletal effects of vitamin D in
relation to dosage, Acta Physiol Scand 35:53 (1955).
36. B. B. Migicovsky, Influence of vitamin D on calcium resorption and accretion, Can J Biochem
Physiol 35:1267 (1957).
37. A. W. Norman, Actinomycin D effect on lag in vitamin D–mediated calcium absorption in
the chick, Am J Physiol 211:829 (1966).
96 Collins and Norman
38. J. E. Zull, E. Czarnowska-Misztal, and H. F. DeLuca, On the relationship between vitamin
D action and actinomycin-sensitive processes, Proc Natl Acad Sci USA 55:177 (1966).
39. D. Schachter, D. V. Kimberg, and H. Schenker, Active transport of calcium by intestine:
action and bio-assay of vitamin D, Am J Physiol 200:1263 (1961).
40. V. W. Thompson and H. F. DeLuca, Vitamin D and phospholipid metabolism, J Biol Chem
239:984 (1964).
41. A. W. Norman, J. Lund, and H. F. DeLuca, Biologically active forms of vitamin D3 in kidney
and intestine, Arch Biochem Biophys 108:12 (1964).
42. J. W. Blunt, H. F. DeLuca, and H. K. Schnoes, 25-Hydroxycholecalciferol: a biologically
active metabolite of vitamin D3, Biochemistry 6:3317 (1968).
43. H. F. DeLuca, 25-Hydroxycholecalciferol, the probable metabolically active form of vitamin
D3: its identification and subcellular site of action, Arch Intern Med 124:442 (1969).
44. M. R. Haussler, J. F. Myrtle, and A. W. Norman, The association of a metabolite of vitamin
D3 with intestinal mucosa chromatin, in vivo, J Biol Chem 243:4055 (1968).
45. A. W. Norman and R. G. Wong, The biological activity of the vitamin D metabolite 1,25-
dihydroxycholecalciferol in chickens and rats, J Nutr 102:1709 (1972).
46. D. E. M. Lawson, D. R. Fraser, E. Kodicek, H. R. Morris, and D. H. Williams, Identification
of 1,25-dihydroxycholecalciferol, a new kidney hormone controlling calcium metabolism,
Nature 230:228 (1971).
47. A. W. Norman, J. F. Myrtle, R. J. Midgett, H. G. Nowicki, V. Williams, and G. Popjak,
1,25-Dihydroxycholecalciferol: identification of the proposed active form of vitamin D3 in
the intestine, Science 173:51 (1971).
48. M. F. Holick, H. K. Schnoes, H. F. Deluca, T. Suda, and R. J. Cousins, Isolation and identifi-
cation of 1,25-dihydroxycholecalciferol: a metabolite of vitamin D active in intestine, Biochemistry
10:2799 (1971).
49. D. R. Fraser and E. Kodicek, Unique biosynthesis by kidney of a biologically active vitamin
D metabolite, Nature 288:764 (1970).
50. M. R. Haussler and A. W. Norman, Chromosomal receptor for a vitamin D metabolite, Proc
Natl Acad Sci USA 62:155 (1969).
51. H. C. Tsai and A. W. Norman, Studies on calciferol metabolism. VIII. Evidence for a cytoplasmic
receptor for 1,25-dihydroxyvitamin D3 in the intestinal mucosa, J Biol Chem 248:
5967 (1972).
52. A. R. Baker, D. P. McDonnell, M. Hughes, T. M. Crisp, D. J. Mangelsdorf, M. R. Haussler,
J. W. Pike, J. Shine, and B. W. O’Malley, Cloning and expression of full-length cDNA
encoding human vitamin D receptor, Proc Natl Acad Sci USA 85:3294 (1988).
53. H. Goto, K. Chen, J. M. Prahl, and H. F. DeLuca, A single receptor identical with that from
intestine/T47D cells mediates the action of 1,25-dihydroxyvitamin D3 in HL-60 cells, Biochim
Biophys Acta 1132:103 (1992).
54. J. K. Burmester, N. Maeda, and H. F. DeLuca, Isolation and expression of rat 1,25-dihydroxyvitamin
D3 receptor cDNA, Proc Natl Acad Sci USA 85:1005 (1988).
55. M. A. Elaroussi, J. M. Prahl, and H. F. DeLuca, The avian vitamin D receptors: primary
structures and their origins, Proc Natl Acad Sci USA 91:11596 (1994).
56. P. P. Minghetti and A. W. Norman, 1,25(OH)2-vitamin D3 receptors: gene regulation and
genetic circuitry, FASEB J 2:3043 (1988).
57. D. Crowfoot and J. D. Dunitz, Structure of calciferol, Nature 162:608 (1948).
58. D. C. Hodgkin, B. M. Rimmer, J. D. Dunitz, and K. N. Trueblood, The crystal structure of
a calciferol derivative, J Chem Soc 947:4945 (1963).
59. W. H. Okamura, A. W. Norman, and R. M. Wing, Vitamin D: concerning the relationship
between molecular topology and biological function, Proc Natl Acad Sci USA 71:4194
60. E. G. Bligh and W. J. Dyer, An extraction procedure for removing total lipids from tissue,
Can J Biochem 37:911 (1958).
Vitamin D 97
61. N. Ikekawa and N. Koizumi, Separation of vitamin D3 metabolites and their analogues by
high-pressure liquid chromatography, J Chromatogr 119:227 (1976).
62. B. P. Halloran, D. D. Bikle, and J. O. Whitney, Separation of isotopically labeled vitamin-
D metabolites by high-performance liquid-chromatography, J Chromatogr 303:229 (1984).
63. R. B. Woodward, F. Sondheimer, D. Taub, K. Heusler, and W. M. McLamore, The total
synthesis of steroids, J Am Chem Soc 74:4223 (1952).
64. H. M. E. Cardwell, J. W. Cornforth, S. R. Duff, H. Holtermann, and R. Robinson, Experiments
on the synthesis of substances related to the sterols. Ll. Completion of the syntheses
of androgenic hormones and of the cholesterol group of sterols, Chem Soc London J:361
65. P. R. Bruck, R. D. Clark, R. S. Davidson, W. H. Gunther, P. S. Littlewood, and B. Lythgoe,
Calciferol and its relatives. VIII. Ring A intermediates for the synthesis of tachysterol. J
Chem Soc 23:2529 (1967).
66. D. H. R. Barton, R. H. Hesse, M. M. Pechet, and E. Rizzardo, A convenient synthesis of
1?-hydroxy-vitamin D3, J Am Chem Soc 95:2748 (1973).
67. D. H. R. Barton, R. H. Hesse, M. M. Pechet, and E. Rizzardo, Convenient synthesis of
crystalline 1?,25-dihydroxyvitamin D3, J Chem Soc Chem Commun 203 (1974)
68. E. G. Baggiolini, J. A. lacobelli, B. M. Hennessy, A. D. Batcho, J. F. Sereno, and M. R.
Uskokovic, Stereocontrolled total synthesis of 1?,25-dihydroxycholecalciferol and 1?,25-
dihydroxyergocalciferol, J Org Chem 51:3098 (1986).
69. J. DeSchrijver and P. J. Declerco, A novel synthesis of an A-ring precursor to 1?-hydroxyvitamin
D, Tetrahedron Lett 34:4369 (1993).
70. R. G. Harrison, B. Lythgoe, and P. W. Wright, Calciferol and its relatives. XVIII. Total
synthesis of 1?-hydroxyvitamin D3, J Chem Soc Perkin Trans 1:2654 (1974).
71. S. A. Barrack, R. A. Gibbs, and W. H. Okamura, Potential inhibitors of vitamin D metabolism:
an oxa analogue of vitamin D, J Org Chem 53:1790 (1988).
72. W. H. Okamura, J. M. Aurrecoechea, R. A. Gibbs, and A. W. Norman, Synthesis and biological
activity of 9, 11-dehydrovitamin D3 analogues: stereoselective preparation of 6b-vitamin
D vinylallenes and a concise enynol synthesis for preparing the A-ring, J Org Chem 54:4072
73. B. M. Trost, J. Dumas, and M. Villa, New strategies for the synthesis of vitamin D metabolites
via Pd-catalyzed reactions, J Am Chem Soc 114:9836 (1992).
74. K. Nagasawa, Y. Zako, H. Ishihara, and I. Shimizu, Stereoselective synthesis of 1?-hydroxyvitamin
D3 A-ring synthons by palladium-catalyzed cyclization, Tetrahed Lett 32:4937
75. M. Sheves and Y. Mazur, The vitamin D-3, 5-cyclovitamin D rearrangement, J Am Chem
Soc 97:6249 (1975).
76. H. E. Paaren, H. F. DeLuca, and H. K. Schnoes, Direct C(1) hydroxylation of vitamin D3
and related compounds, J Org Chem 45:3253 (1980).
77. D. R. Andrews, D. H. R. Barton, K. P. Cheng, J. P. Finet, R. H. Hesse, G. Johnson, and M. M.
Pechet, A direct, regioselective and stereoselective 1-alpha-hydroxylation of (5E)-calciferol
derivatives, J Org Chem 51:1635 (1986).
78. L. Vanmaele, P. J. Declercq, and M. Vandewalle, An efficient synthesis of 1-?,25-dihydroxy
vitamin-D3, Tetrahedron 41:141 (1985).
79. A. W. M. Hay and G. Watson, Binding of 25-hydroxy vitamin D2 to plasma protein in New
World monkeys, Nature 256:150 (1975).
80. W. Heymann, Metabolism and mode of action of vitamin D. V. Intestinal excretion of vitamin
D, J Biol Chem 122:257 (1938).
81. W. Heymann, Metabolism and mode of action of vitamin D. IV. Importance of bile in the
absorption and excretion of vitamin D, J Biol Chem 12:249 (1938).
82. S. J. Rosenstreich, C. Rich, and W. Volwiler, Depostition in and release of vitamin D3 from
body fat: evidence for a storage site in the rat. J Clin Invest 50:679 (1971).
98 Collins and Norman
83. D. Schachter, J. D. Finelstein, and S. Kowarski, Metabolism of vitamin D. I. Preparation of
radioactive vitamin D and its intestinal absorption in the rat, J Clin Invest 43:787 (1964).
84. A. W. Norman and H. F. DeLuca, The preparation of H3-vitamins D2 and D3 and their localization
in the rat, Biochemistry 2:1160 (1963).
85. M. F. Holick, The cutaneous photosynthesis of previtamin D3: a unique photoendocrine system,
J Invest Dermatol 76:51 (1981).
86. H. F. DeLuca, Mechanism of action and metabolic fate of vitamin D, Vitam Horm 25:315
87. P. G. Walsh and J. G. Haddad, Rocket immunoelectrophoresis assay of vitamin D–binding
protein (Gc-globulin), Clin Chem 28:1781 (1982).
88. D. Dykes, H. Polesky, and E. Cox, Isoelectric focusing of Gc (vitamin D binding globulin)
in parentage testing, Hum Genet 58:174 (1981).
89. S. P. Daiger, M. S. Schanfield, and L. L. Cavalli-Sforza, Group-specific component
(Gc) proteins bind vitamin D and 25-hydroxyvitamin D, Proc Natl Acad Sci USA 72:2076
90. R. Bouillon, H. van Baelen, W. Rombauts, and P. De Moor, The purification and characterization
of the human-serum binding protein for the 25-hydroxycholecalciferol (transcalciferin),
Eur J Biochem 66:285 (1976).
91. J. Silver and M. Fainaru, Transport of vitamin D sterols in human plasma: effect of excess
vitamin D, 25-hydroxyvitamin D and 1,25-dihydroxyvitamin D, Eur J Clin Invest 9:433
92. J. P. Mallon, D. Matuskewski, and H. Sheppard, Binding specificity of the rat serum vitamin
D transport protein, J Steroid Biochem Molec Biol 13:409 (1980).
93. H. van Baelen, R. Bouillon, and P. Demoor, Vitamin D–binding protein (Gc-globulin) binds
actin, J Biol Chem 255:2270 (1980).
94. N. E. Cooke and E. V. David, Serum vitamin D–binding protein is a third member of the
albumin and alpha-fetoprotein gene family, J Clin Invest 76:2420 (1985).
95. R. Bouillon, F. A. van Assche, H. van Baelen, W. Heyns, and P. De Moor, Influence of the
vitamin D–binding protein on the serum concentration of 1,25-dihydroxyvitamin D3. J Clin
Invest 67:589 (1981).
96. D. D. Bikle, P. K. Siiteri, E. Ryzen, and J. G. Haddad, Serum protein binding of 1,25-diyhydroxyvitamin
D: a reevaluation by direct measurement of free metabolite levels, J Clin Endocrinol
Metab 61:969 (1985).
97. H. van Baelen, K. Allewaert, and R. Bouillon, New aspects of the plasma carrier protein for
25-hydroxycholecalciferol in vertebrates, Ann NY Acad Sci 538:60 (1988).
98. R. Bouillon and H. van Baelen, Transport of vitamin D: significance of free and total concentrations
of the vitamin D metabolites, Calcif Tissue Int 33:451 (1981).
99. N. E. Cooke and J. G. Haddad, Vitamin D binding protein (Gc- globulin). Endocr Rev 10:
294 (1989).
100. A. W. M. Hay and G. Watson, The binding of 25-hydroxycholecalciferol and 25-hydroxyergocalciferol
to receptor proteins in a New World and an Old World primate, Comp Biochem
Physiol 56B:131 (1977).
101. J. E. Compston, L. W. L. Horton, and M. F. Laker, Vitamin-D analogues and renal function,
Lancet 1:386 (1979).
102. J. Rungby, L. Mortensen, K. Jakobsen, A. Brock, and L. Mosekilde, Distribution of hydroxylated
vitamin D metabolities [25(OH)D3 and 1,25(OH)2D3] in domestic pigs: evidence that
1, 25(OH)2D3 is stored outside the blood circulation, Comp Biochem Physiol 104A:483
103. E. B. Mawer, J. Backhouse, C. A. Holman, G. A. Lumb, and S. W. Stanbury, The distribution
and storage of vitamin D and its metabolites in human tissues, Clin Sci 43:413 (1972).
104. H. L. Henry and A. W. Norman, Vitamin D: two dihydroxylated metabolites are required
for normal chicken egg hatchability, Science 201:835 (1978).
Vitamin D 99
105. A. W. Norman, V. L. Leathers, and J. E. Bishop, Studies on the mode of action of calciferol.
XLVIII. Normal egg hatchability requires the simultaneous administration to the hen of
1?,25-dihydroxyvitamin D3 and 24R,25-dihydroxyvitamin D3, J Nutr 113:2505 (1983).
106. K. Saarem, S. Bergseth, H. Oftebro, and J. I. Pedersen, Subcellular localization of vitamin
D3 25-hydroxylase in human liver, J Biol Chem 259:10936 (1984).
107. E. Axe?n, T. Bergman, and K. Wikvall, Microsomal 25-hydroxylation of vitamin D2 and
vitamin D3 in pig liver, J Steroid Biochem Mol Biol 51:97 (1994).
108. M. H. Bhattacharyya and H. F. DeLuca, Subcellular location of rat liver calciferol-25-hydroxylase,
Arch Biochem Biophys 160:58 (1974).
109. T. Bergman and H. Postlind, Characterization of mitochondrial cytochromes P-450 from
pig kidney and liver catalysing 26-hydroxylation of 25-hydroxyvitamin D3 and C27 steroids,
Biochem J 276:427 (1991).
110. G. Tucker, R. E. Gagnon, and M. R. Haussler, Vitamin D3-25-hydroxylase: tissue occurrence
and apparent lack of regulation, Arch Biochem Biophys 155:47 (1973).
111. J. M. Pettifor, F. P. Ross, and J. Wang, Serum levels of 25-hydroxycholecalciferol as a
diagnostic aid in vitamin D deficiency states, South Afr Med J 51:580 (1977).
112. E. B. Mawer, Clinical implication of measurements of circulating vitamin D metabolites, J
Clin Endocrinol Metab 9:63 (1980).
113. D. T. Baran and M. L. Milne, 1,25 Dihydroxyvitamin-D increases hepatocyte cytosolic
calcium levels: a potential regulator of vitamin-D-25-hydroxylase, J Clin Invest 77:1622
114. E. Usui, M. Noshiro, and K. Okuda, Molecular cloning of cDNA for vitamin D3 25-hydroxylase
from rat liver mitochondria. FEBS Lett 262:135 (1990).
115. M. Akiyoshi-Shibata, E. Usui, T. Sakaki, Y. Yabusaki, M. Noshiro, K. Okuda, and H. Ohkawa,
Expression of rat liver vitamin D3 25-hydroxylase cDNA in Saccharomyces cerevisiae,
FEBS Lett 280:367 (1991).
116. H. L. Henry and A. W. Norman, Studies on calciferol metabolism. IX. Renal 25-hydroxyvitamin
D3-1-hydroxylase. Involvement of cytochrome P-450 and other properties, J Biol
Chem 249:7529 (1974).
117. H. L. Henry, Vitamin D hydroxylases, J Cell Biochem 49:4 (1992).
118. Y. Weisman, A. Harell, S. Edelstein, M. David, Z. Spirer, and A. Golander, 1?,25-Dihydroxyvitamin
D3 and 24,25-dihydroxyvitamin D3 in vitro synthesis by human decidua and placenta,
Nature 281:317 (1979).
119. Y. Tanaka, B. P. Halloran, H. K. Schnoes, and H. F. DeLuca, In vitro production of 1,25-
dihydroxyvitamin D3 by rat placental tissue, Proc Natl Acad Sci USA 76:5033 (1979).
120. J. S. Adams, F. R. Singer, M. A. Gacad, O. P. Sharma, M. J. Hayes, P. Vouros, and M. F.
Holick, Isolation and structural identification of 1,25-dihydroxyvitamin D3 produced by cultured
alveolar macrophages in sarcoidosis, J Clin Endocrinol Metab 60:960 (1985).
121. H. Reichel, H. P. Koeffler, and A. W. Norman, Synthesis in vitro of 1,25-dihydroxyvitamin
D3 and 24,25-dihydroxyvitamin D3 by interferon-gamma-stimulated normal human bone
marrow and alveolar macrophages, J Biol Chem 262:10931 (1987).
122. H. Reichel, J. E. Bishop, H. P. Koeffler, and A. W. Norman, Evidence for 1,25-dihydroxyvitamin
D3 production by cultured porcine alveolar macrophages, Mol Cell Endocrinol 75:163
123. J. E. Puzas, R. T. Turner, G. A. Howard, J. S. Brand, and D. J. Baylink, Synthesis of 1,25-
dihydroxycholecalciferol and 24,25-dihydroxycholecalciferol by calvarial cells: characterization
of the enzyme-systems, Biochem J 245:333 (1987).
124. S. Pillai, D. D. Bikle, and P. M. Elias, 1,25-Dihydroxyvitamin D production and receptor
binding in human keratinocytes varies with differentiation, J Biol Chem 263:5390 (1988).
125. K. Matsumoto, Y. Azuma, M. Kiyoki, H. Okumura, K. Hashimoto, and K. Yoshikawa,
Involvement of endogenously produced 1,25-dihydroxyvitamin D3 in the growth and differentiation
of human keratinocytes, Biochim Biophys Acta 1092:311 (1991).
100 Collins and Norman
126. H. L. Henry, C. Dutta, N. Cunningham, R. Blanchard, R. Penny, C. Tang, G. Marchetto,
and S.-Y. Chou, The cellular and molecular regulation of 1,25(OH)2D3 production, J Steroid
Biochem Mol Biol 41:401 (1992).
127. K. W. Colston, I. M. Evans, T. C. Spelsberg, and I. Macintyre, Feedback regulation of vitamin
D metabolism by 1,25-dihydroxycholecalciferol. Biochem J 164:83 (1977).
128. H. L. Henry, A. N. Taylor, and A. W. Norman, Response of chick parathyroid glands to the
vitamin D metabolites, 1,25-dihydroxycholecalciferol and 24,25-dihydroxycholecalciferol, J
Nutr 107:1918 (1977).
129. J. M. Canterbury, G. Gavellas, J. J. Bourgoignie, and E. Reise, Metabolic consequences of
oral administration of 24,25-dihydroxycholecalciferol to uremic dogs, J Clin Invest 65:571
130. H. Yamato, R. Okazaki, T. Ishii, E. Ogata, T. Sato, M. Kumegawa, K. Akaogi, N. Taniguchi,
and T. Matsumoto, Effect of 24R,25-dihydroxyvitamin D3 on the formation and function of
osteoclastic cells, Calcif Tissue Int 52:255 (1993).
131. R. A. Evans, Hills E., S. Y. P. Wong, Dunstan C. R., and A. W. Norman, The use of 24,25-
dihydroxycholecalciferol alone and in combination with 1,25-dihydroxycholecalciferol in
chronic renal failure, in Vitamin D: Chemical, Biochemical and Clinical Endocrinology of
Calcium Metabolism (A. W. Norman, K. Schaefer, H.-G. Grigoleit, and D. v. Herrath, eds.),
Walter de Gruyter, Berlin, 1982, pp. 835–840.
132. C. Lidor, I. Atkin, A. Ornoy, S. Dekel, and S. Edelstein, Healing of rachitic lesions in chicks
by 24R,25-dihydroxycholecalciferol administered locally into bone, J Bone Miner Res 2:91
133. C. Lidor, S. Dekel, and S. Edelstein, The metabolism of vitamin D3 during fracture-healing
in chicks, Endocrinology 120:389 (1987).
134. A. W. Norman, V. L. Leathers, J. E. Bishop, Kadowaki S., and B. E. Miller, 24R,25-Dihydroxyvitamin
D3 has unique receptors (parathyroid gland) and biological responses (egg
hatchability), in Vitamin D: Chemical, Biochemical, and Clinical Endocrinology of Calcium
Metabolism (A. W. Norman, K. Schaefer, H.-G. Grigoleit, and D. Herrath, eds.), Walter de
Gruyter, Berlin, 1982, pp. 147–151.
135. S. Ishizuka, T. Ohba, and A. W. Norman, 1?,25(OH)2D3-26,23-Lactone is a major metabolite
of 1?,25(OH)2D3 under physiological conditions, in Vitamin D: Molecular Cellular and Clinical
Endocrinology (A. W. Norman, K. Schaefer, H. G. Grigoleit, and D. v. Herrath, eds.),
Walter de Gruyter, Berlin, 1988, pp. 143–144.
136. R. Kumar, The metabolism and mechanism of action of 1,25-dihydroxyvitamin D3, Kidney
Int 30:793 (1986).
137. E. Orti, J. E. Bodwell, and A. Munck, Phosphorylation of steroid hormone receptors, Endocr
Rev 13:105 (1992).
138. H. M. Darwish, J. K. Burmester, V. E. Moss, and H. F. DeLuca, Phosphorylation is involved
in transcriptional activation by the 1,25-dihydroxyvitamin D3 receptor, Biochim Biophys Acta
1167:29 (1993).
139. P. W. Jurutka, J.-C. Hsieh, and M. R. Haussler, Phosphorylation of the human 1,25-dihydroxyvitamin
D3 receptor by cAMP-dependent protein kinase, in vitro, and in transfected COS-
7 cells, Biochem Biophys Res Commun 191:1089 (1993).
140. J.-C. Hsieh, P. W. Jurutka, S. Nakajima, M. A. Galligan, C. A. Haussler, Y. Shimizu, N.
Shimizu, G. K. Whitfield, and M. R. Haussler, Phosphorylation of the human vitamin D
receptor by protein kinase C: biochemical and functional evaluation of the serine 51 recognition
site, J Biol Chem 268:15118 (1993).
141. K. E. Lowe, A. C. Maiyar, and A. W. Norman, Vitamin D–mediated gene expression, Crit
Rev Eukar Gene Exp 2:65 (1992).
142. S. S. Hannah and A. W. Norman, 1,25-Dihydroxyvitamin D3 regulated expression of the
eukaryotic genome, Nutr Rev 52:376 (1994).
143. W. R. Wecksler and A. W. Norman, A kinetic and equilibrium binding study of 1?,25-
Vitamin D 101
dihydroxyvitamin D3 with its cytosol receptor from chick intestinal mucosa, J Biol Chem
255:3571 (1980).
144. W. R. Wecksler, F. P. Ross, R. S. Mason, S. Posen, and A. W. Norman, Studies on the mode
of action of calciferol. XXIV. Biochemical properties of the 1?, 25-dihydroxyvitamin D3
cytoplasmic receptors from human and chick parathyroid glands, Arch Biochem Biophys 201:
95 (1980).
145. W. R. Wecksler, F. P. Ross, R. S. Mason, and A. W. Norman, Studies on the mode of
action of calciferol. XXI. Biochemical properties of the 1?,25-dihydroxyvitamin D3 cytosol
receptors from human and chicken intestinal mucosa, J Clin Endocrinol Metab 50:152
146. R. M. Evans, The steroid and thyroid hormone receptor superfamily, Science 240:889 (1988).
147. M. G. Parker, Nuclear Hormone Receptors: Molecular Mechanisms, Cellular Functions,
Clinical Abnormalities, Academic Press, London, 1991.
148. A. Krust, S. Green, P. Argos, V. Kumar, P. Walter, J. M. Bornert, and P. Chambon, The
chicken oestrogen receptor sequence: homology with v-erbA and the human oestrogen and
glucocorticoid receptors, Embo J 5:891 (1986).
149. M. Beato, Gene regulation by steroid hormones, Cell 56:335 (1989).
150. S. Green and P. Chambon, Nuclear receptors enhance our understanding of transcription
regulation, Trends Genet 4:309 (1988).
151. J. Ham and M. G. Parker, Regulation of gene expression by nuclear hormone receptors, Curr
Opin Cell Biol 1:503 (1989).
152. B. M. Forman and H. H. Samuels, Interactions among a subfamily of nuclear hormone receptors:
the regulatory zipper model, Mol Endocrinol 4:1293 (1990).
153. F. Rastinejad, T. Perlmann, R. M. Evans, and P. B. Sigler, Structural determinants of nuclear
receptor assembly on DNA direct repeats, Nature 375:203 (1995).
154. S. E. Fawell, J. A. Lees, R. White, and M. G. Parker, Characterization and colocalization
of steroid binding and dimerization activities in the mouse estrogen receptor. Cell 60:953
155. C. K. Glass, S. M. Lipkin, O. V. Devary, and M. G. Rosenfeld, Positive and negative regulation
of gene transcription by a retinoic acid–thyroid hormone receptor heterodimer, Cell 59:
697 (1989).
156. B. M. Forman, C. Yan, M. Au, J. Casanova, J. Ghysdael, and H. H. Samuels, A domain
containing leucine zipper like motifs may mediate novel in vivo interactions between the
thyroid hormone and retinoic acid receptors, Mol Endocrinol 3:1610 (1989).
157. W. Bourguet, M. Ruff, P. Chambon, H. Gronemeyer, and D. Moras, Crystal structure of the
ligand-binding domain of the human nuclear receptor RXR-?, Nature 375:377 (1995).
158. D. Picard, B. Khursheed, M. J. Garabedian, M. G. Fortin, S. Lindquist, and K. R. Yamamoto,
Reduced levels of hsp90 compromise steroid receptor action in vivo, Nature 348:166 (1990).
159. L. Tora, J. White, C. Brou, D. Tasset, N. Webster, E. Scheer, and P. Chambon, The human
estrogen receptor has two independent nonacidic transcriptional activation functions, Cell
59:477 (1989).
160. D. P. McDonnell, J. W. Pike, and B. W. O’Malley, The vitamin D receptor: a primitive
steroid receptor related to thyroid hormone receptor, J Steroid Biochem 30:41 (1988).
161. V. C. Yu, C. Delsert, B. Andersen, J. M. Holloway, O. V. Devary, A. M. Na?a?r, S. Y. Kim,
J.-M. Boutin, C. K. Glass, and M. G. Rosenfeld, RXR?: a coregulator that enhances binding
of retinoic acid, thyroid hormone, and vitamin D receptors to their cognate response elements,
Cell 67:1251 (1991).
162. S. A. Kliewer, K. Umesono, D. J. Mangelsdorf, and R. M. Evans, Retinoid X receptor interacts
with nuclear receptors in retinoic acid, thyroid hormone and vitamin D3 signalling, Nature
355:446 (1992).
163. C. Carlberg, RXR-independent action of the receptors for thyroid hormone, retinoid acid and
vitamin D on inverted palindromes, Biochem Biophys Res Commun 195:1345 (1993).
102 Collins and Norman
164. M. Schra?der, I. Bendik, M. Becker-Andre?, and C. Carlberg, Interaction between retinoic acid
and vitamin D signaling pathways, J Biol Chem 268:17830 (1993).
165. R. H. Wasserman, R. A. Corradino, and A. N. Taylor, Vitamin D–dependent calcium-binding
protein: purification and some properties, J Biol Chem 243:3978 (1968).
166. C. Perret, C. Desplan, A. Brehier, and M. Thomasset, Characterization of rat 9-kDa cholecalcin
(CaBP) messenger RNA using a complementary DNA: absence of homology with 28-
kDa cholecalcin mRNA. Eur J Biochem 148:61 (1985).
167. L. Cancela, H. Ishida, J. E. Bishop, and A. W. Norman, Local chromatin changes accompany
the expression of the calbindin-D28K gene: tissue specificity and effect of vitamin D activation,
Mol Endocrinol 6:468 (1992).
168. R. A. Corradino, 1,25-Dihydroxycholecalciferol: inhibition of action in organ-cultured intestine
by actinomycin D and ?-amanitin, Nature 243:41 (1973).
169. H. C. Tsai and A. W. Norman, Studies on the mode of action of calciferol. VI. Effect of
1,25-dihydroxyvitamin D3 on RNA synthesis in the intestinal mucosa, Biochem Biophys Res
Commun 54:622 (1973).
170. P. Siebert, W. Hunziker, and A. W. Norman, Studies on the mode of action of calciferol.
XLVIII. Cell-free translation analysis of vitamin D–dependent calcium binding protein
mRNA activity present in total RNA and polysomal extracts from chick intestine, Arch Biochem
Biophys 219:286 (1982).
171. G. Theofan, A. P. Nguyen, and A. W. Norman, Regulation of calbindin-D28k gene expression
by 1,25-dihydroxyvitamin D3 is correlated to receptor occupancy, J Biol Chem 261:16943
172. P. P. Minghetti, L. Cancela, Y. Fujisawa, G. Theofan, and A. W. Norman, Molecular structure
of the chicken vitamin D-induced calbindin-D28K gene reveals eleven exons, six Ca2-binding
domains, and numerous promoter regulatory elements. Mol Endocrinol 2:355 (1988).
173. I. Nemere and A. W. Norman, Studies on the mode of action of calciferol. LII. Rapid action
of 1,25-dihydroxyvitmin D3 on calcium transport in perfused chick duodenum: effect of inhibitors,
J Bone Miner Res 2:99 (1987).
174. M. Lieberherr, B. Grosse, P. Duchambon, and T. Dru?eke, A functional cell surface type
receptor is required for the early action of 1,25-dihydroxyvitamin D3 on the phosphoinositide
metabolism in rat enterocytes, J Biol Chem 264:20403 (1989).
175. B. D. Boyan, Z. Schwartz, L. Bonewald, and L. Swain, Localization of 1,25(OH)2D3-responsive
alkaline phosphatase in osteoblast-like cells (ROS 17/2.8, MG 63, and MC 3T3) and
growth cartilage cells in culture, J Biol Chem 264:11879 (1989).
176. M. C. Meikle, Hypercalcaemia of malignancy, Nature 336:311 (1988).
177. L. Cancela, I. Nemere, and A. W. Norman, 1?,25(OH)2 Vitamin D3: a steroid hormone
capable of producing pleiotropic receptor-mediated biological responses by both genomic
and nongenomic mechanisms, J Steroid Biochem Mol Biol 30:33 (1988).
178. J. M. Caffrey and M. C. Farach-Carson, Vitamin D3 metabolites modulate dihydropyridine-
sensitive calcium currents in clonal rat osteosarcoma cells, J Biol Chem 264:20265
179. M. Lieberher, Effects of vitamin-D3 metabolites on cytosolic free calcium in confluent mouse
osteoblasts, J Biol Chem 262:13168 (1987).
180. J. Barsony and S. J. Marx, Rapid accumulation of cyclic GMP near activated vitamin D
receptors, Proc Natl Acad Sci USA 88:1436 (1991).
181. I. Nemere, L.-X. Zhou, and A. W. Norman, Nontranscriptional effects of steroid hormones,
Receptor 3:277 (1993).
182. I. Nemere, M. C. Dormanen, M. W. Hammond, W. H. Okamura, and A. W. Norman, Identi-
fication of a specific binding protein for 1?,25-dihydroxyvitamin D3 in basal-lateral membranes
of chick intestinal epithelium and relationship to transcaltachia, J Biol Chem 269:
23750 (1994).
183. P. Morley, J. F. Whitfield, B. C. Vanderhyden, B. K. Tsang, and J. L. Schwartz, A new,
Vitamin D 103
nongenomic estrogen action: the rapid release of intracellular calcium, Endocrinology 131:
1305 (1992).
184. C. Mendoza and J. Tesarik, A plasma-membrane progesterone receptor in human sperm is
switched on by increasing intracellular free calcium, FEBS Lett 330:57 (1993).
185. C. Aurell Wistrom and S. Meize, Evidence suggesting involvement of a unique human sperm
steroid receptor/Cl channel complex in the progesterone-initiated acrosome reaction, Dev
Biol 159:679 (1993).
186. P. F. Blackmore, S. J. Beebe, D. R. Danforth, and N. Alexander, Progesterone and 17?-
progesterone: novel stimulators of calcium influx in human sperm, J Biol Chem 265:1376
187. M. D. Majewska and D. B. Vaupel, Steroid control of uterine motility via gamma-aminobutyric
acidA receptors in the rabbit: a novel mechanism, J Endocrinol 131:427 (1991).
188. H. Koenig, C.-C. Fan, A. D. Goldstone, C. Y. Lu, and J. J. Trout, Polyamines mediate androgenic
stimulation of calcium fluxes and membrane transport in rat myocytes, Circ Res 64:
415 (1989).
189. P. Rehberger, M. Rexin, and U. Gehring, Heterotetrameric structure of the human progesterone
receptor, Proc Natl Acad Sci USA 89:8001 (1992).
190. B. Gametchu, C. S. Watson, and S. Wu, Use of receptor antibodies to demonstrate membrane
glucocorticoid receptor in cells from human leukemic patients, FASEB J 7:1283 (1993).
191. M. Orchinik, T. F. Murray, and F. L. Moore, A corticosteroid receptor in neuronal membranes,
Science 252:1848 (1991).
192. T. J. Smith, F. B. Davis, and P. J. Davis, Stereochemical requirements for the modulation
by retinoic acid of thyroid hormone activation of Ca2-ATPase and binding at the human
erythrocyte membrane, Biochem J 284:583 (1992).
193. J. Segal, Thyroid hormone action at the level of the plasma membrane, Thyroid 1:83 (1990).
194. D. Schachter, E. B. Dowdle, and H. Schenker, Active transport of calcium by the small
intestine of the rat, Am J Physiol 198:263 (1960).
195. H. E. Harrison and H. C. Harrison, Vitamin D and permeability of intestinal mucosa to
calcium, Am J Physiol 208:370 (1965).
196. V. L. Leathers, Biophysical characterization and functional studies on calbindin-D28K: a vitamin
D–induced calcium binding protein, Ph.D. dissertation, University of California, Riverside,
197. I. Nemere, V. L. Leathers, B. S. Thompson, R. A. Luben, and A. W. Norman, Redistribution
of calbindin-D28K in chick intestine in response to calcium transport, Endocrinology 129:
2972 (1991).
198. J. A. Putkey, A. M. Spielvogel, R. D. Sauerheber, Dunlap C. S., and A. W. Norman, Studies
on the mode of action of calciferol. XXXIX. Vitamin D–mediated intestinal calcium transport:
effect of essential fatty acid deficiency and spin label studies of enterocyte membrane
lipid fluidity, Biochim Biophys Acta 688:177 (1982).
199. J. A. Putkey and A. W. Norman, Studies on the mode of action of calciferol. XLV. Vitamin
D: its effect on the protein composition and core material structure of the chick intestinal
brush border membrane, J Biol Chem 258:8971 (1983).
200. A. M. Spielvogel, R. D. Farley, and A. W. Norman, Studies on the mechanism of action of
calciferol. V. Turnover time of chick intestinal epithelial cells in relation to the intestinal
action of vitamin D, Exp Cell Res 74:359 (1972).
201. J. T. McCarthy, S. S. Barham, and R. Kumar, 1,25-Dihydroxyvitamin D3 rapidly alters the
morphology of the duodenal mucosa of rachitic chicks: evidence for novel effects of 1,25-
dihydroxyvitamin D3, J Steroid Biochem Mol Biol 21:253 (1984).
202. R. H. Wasserman, M. E. Brindak, M. M. Buddle, Q. Cai, F. C. Davis, C. S. Fullmer, R. F.
Gilmour, Jr., C. Hu, H. M. Mykkanen, and D. N. Tapper, Recent studies on the biological
actions of vitamin D on intestinal transport and the electrophysiology of peripheral nerve
and cardiac muscle, Prog Clin Biol Res 332:99 (1990).
104 Collins and Norman
203. I. Nemere and A. W. Norman, Transport of calcium, in Handbook of Physiology (M. Field
and R. A. Frizzell, eds.), Am. Physiol. Soc., Bethesda, MD, 1991, pp. 337–360.
204. I. Nemere and A. W. Norman, Transcaltachia, vesicular calcium transport, and microtubuleassociated
calbindin-D28K: emerging views of 1,25-dihydroxyvitamin D3–mediated intestinal
calcium absorption, Min Elect Metab 16:109 (1990).
205. M. R. Clements, L. Johnson, and D. R. Fraser, A new mechanism for induced vitamin-D
deficiency in calcium deprivation, Nature 325:62 (1987).
206. H. L. Henry and A. W. Norman, Vitamin D: metabolism and biological action, Annu Rev
Nutr 4:493 (1984).
207. J. L. Underwood and H. F. DeLuca, Vitamin-D is not directly necessary for bone-growth
and mineralization, Am J Physiol 246:E493 (1984).
208. A. W. Norman and S. Hurwitz, The role of the vitamin D endocrine system in avian bone
biology, J Nutr 123:S30 (1993).
209. M. R. Walters, D. M. Rosen, A. W. Norman, and R. A. Luben, 1,25-Dihydroxyvitamin D
receptors in an established bone cell line: correlation with biochemical responses, J Biol
Chem 257:7481 (1982).
210. J. R. Harrison, D. N. Petersen, A. C. Lichtler, A. T. Mador, D. W. Rowe, and B. E. Kream,
1,25-Dihydroxyvitamin D3 inhibits transcription of type I collagen genes in the rat osteosarcoma
cell line ROS 17/2.8, Endocrinology 125:327 (1989).
211. N. Kurihara, K. Ikeda, Y. Hakeda, M. Tsunoi, N. Maeda, and M. Kumegawa, Effect of 1,25-
dihydroxyvitamin D3 on alkaline phosphatase activity and collagen synthesis in osteoblastic
cells, clone MC3T3-E1, Biochem Biophys Res Commun 119:767 (1984).
212. L. C. Pan and P. A. Price, The effect of transcriptional inhibitors on the bone gamma-carboxyglutamic
acid protein response to 1,25-dihydroxyvitamin D3 in osteosarcoma cells, J Biol
Chem 259:5844 (1984).
213. J. D. Fraser and P. A. Price, Induction of matrix Gla protein synthesis during prolonged
1,25-dihydroxyvitamin D3 treatment of osteosarcoma cells. Calcif Tissue Int 46:270 (1990).
214. D. W. Rowe and B. E. Kream, Regulation of collagen synthesis in fetal rat calvaria by 1,25-
dihydroxyvitamin D3, J Biol Chem 257:8009 (1982).
215. M. E. Holtrop, K. A. Cox, M. B. Clark, M. F. Holick, and C. S. Anast, 1,25-Dihydroxycholecalciferol
stimulates osteoclasts in rat bones in the absence of parathyroid hormone, Endocrinology
108:2293 (1981).
216. G. D. Roodman, K. J. Ibbotson, B. R. MacDonald, T. J. Kuehl, and G. R. Mundy, 1,25-
Dihydroxyvitamin D3 causes formation of multinucleated cells with several osteoclast characteristics
in cultures of primate marrow, Proc Natl Acad Sci USA 82:8213 (1985).
217. N. Takahashi, T. Akatsu, N. Udagawa, T. Sasaki, A. Yamaguchi, J. M. Moseley, T. J. Martin,
and T. Suda, Osteoblastic cells are involved in osteoclast formation, Endocrinology 123:
2600 (1988).
218. E. Abe, Y. Ishimi, N. Takahashi, T. Akatsu, H. Ozawa, H. Yamana, S. Yoshiki, and T. Suda,
A differentiation-inducing factor produced by the osteoblastic cell line MC3T3-E1 stimulates
bone resorption by promoting osteoclast formation, J Bone Miner Res 3:635 (1988).
219. M. Peacock and P. J. Heyburn, Effect of vitamin D metabolites on proximal muscle weakness,
Calcif Tissue Res 24:78 (1977).
220. M. R. Walters, D. C. Wicker, and P. C. Riggle, 1,25-Dihydroxyvitamin D3 receptors identi-
fied in the rat heart, J Mol Cell Cardiol 18:67 (1986).
221. A. Rabie, A. Brehier, S. Intrator, M. C. Clavel, C. O. Parkes, C. Legrand, and M. Thomasset,
Thyroid state and cholecalcin (calcium-binding protein) in cerebellum of the developing rat,
Dev Brain Res 29:253 (1986).
222. J. Merke, W. Hofmann, D. Goldschm., and E. Ritz, Demonstration of 1,25(OH)2 vitamin-
D3 receptors and actions in vascular smooth-muscle cells in vitro, Calcif Tissue Res 41:112
223. R. Boland, A. W. Norman, E. Ritz, and W. Hasselbach, Presence of a 1,25-dihydroxyvitamin
Vitamin D 105
D3 receptor in chick skeletal muscle myoblasts, Biochem Biophys Res Commun 128:305
224. W. E. Stumpf, M. Sar, F. A. Reid, Y. Tanaka, and H. F. DeLuca, Target cells for 1,25-
dihydroxyvitamin D3 in intestinal tract, stomach, kidney, skin, pituitary, and parathyroid,
Science 206:1188 (1979).
225. J. Hosomi, J. Hosoi, E. Abe, T. Suda, and T. Kuroki, Regulation of terminal differentiation
of cultured mouse epidermal cells by 1-?,25-dihydroxyvitamin D3, Endocrinology 113:1950
226. M. Regnier and M. Darmon, 1,25-Dihydroxyvitamin D3 stimulates specifically the last steps
of epidermal differentiation of cultured human keratinocytes, Differentiation 47:173 (1991).
227. D. D. Bikle, M. K. Nemanic, J. O. Whitney, and P. W. Elias, Neonatal human foreskin
keratinocytes produce 1,25-dihydroxyvitamin D3, Biochemistry 25:1545 (1986).
228. K. Kragballe, H. I. Beck, and H. Sogaard, Improvement of psoriasis by a topical vitamin D3
analogue (MC 903) in a double-blind study, Br J Dermatol 119:223 (1988).
229. K. Kragballe, Vitamin D analogues in the treatment of psoriasis, J Cell Biochem 49:46
230. M. F. Holick, Will 1,25-dihydroxyvitamin D3, MC 903, and their analogues herald a new
pharmacologic era for the treatment of psoriasis, Arch Dermatol 125:1692 (1989).
231. K. Chida, H. Hashiba, M. Fukushim, T. Suda, and T. Kuroki, Inhibition of tumor promotion
in mouse skin by 1 ?,25-dihydroxyvitamin D3, Cancer Res 45:5426 (1985).
232. C. Cade and A. W. Norman, Vitamin D3 improves impaired glucose-tolerance and insulinsecretion
in the vitamin D-deficient rat in vivo, Endocrinology 119:84 (1986).
233. S. Kadowaki and A. W. Norman, Studies on the mode of action of calciferol. XLIX. Dietary
vitamin D is essential for normal insulin secretion from the perfused rat pancreas, J Clin
Invest 73:759 (1984).
234. C. Cade and A. W. Norman, Rapid normalization/stimulation by 1,25(OH)2-vitamin D3 of
insulin secretion and glucose tolerance in the vitamin D-deficient rat, Endocrinology 120:
1490 (1987).
235. S. Kadowaki and A. W. Norman, Studies on the mode of action of calciferol. LIX. Timecourse
study of the insulin secretion after 1,25-dihydroxyvitamin D3 administration, Endocrinology
117:1765 (1985).
236. S. Kadowaki and A. W. Norman, Demonstration that the vitamin D metabolite 1,25(OH)2-
vitamin D3 and not 24R,25(OH)2-vitamin D3 is essential for normal insulin secretion in the
perfused rat pancreas, Diabetes 34:315 (1985).
237. O. Gedik and S. Akalin, Effects of vitamin D deficiency and repletion on insulin and glucagon
secretion in man, Diabetologia 29:142 (1986).
238. J. Sonnenberg, V. N. Luine, L. C. Krey, and S. Christakos, 1,25-Dihydroxyvitamin D3 treatment
results in increased choline acetyltransferase activity in specific brain nuclei, Endocrinology
118:1433 (1986).
239. M. Thomasset, C. O. Parkes, and P. Cuisinier-Gleizes, Rat calcium-binding proteins: distribution,
development, and vitamin D dependence, Am J Physiol 243:E483 (1982).
240. A. L. Morrow, J. R. Pace, R. H. Purdy, and S. M. Paul, Characterization of steroid interactions
with gamma-aminobutyric acid receptor–gated chloride ion channels: evidence for multiple
steroid recognition sites. Mol Pharmacol 37:263 (1990).
241. B. J. Bowers and J. M. Wehner, Biochemical and behavioral effects of steroids on GABAA
receptor function in long- and short-sleep mice, Brain Res Bull 29:57 (1992).
242. D. M. Peehl, R. J. Skowronski, G. K. Leung, S. T. Wong, T. A. Stamey, and D. Feldman,
Antiproliferative effects of 1,25-dihydroxyvitamin D3 on primary cultures of human prostatic
cells, Cancer Res 54:805 (1994).
243. E. Abe, C. Miyaura, H. Sakagami, M. Takeda, K. Konno, T. Yamazaki, S. Yoshiki, and T.
Suda, Differentiation of mouse myeloid leukemia cells induced by 1?,25-dihydroxyvitamin
D3, Proc Natl Acad Sci USA 78:4990 (1981).
106 Collins and Norman
244. D. M. McCarthy, J. F. San Miguel, H. C. Freake, P. M. Green, H. Zola, D. Catovsky, and
J. M. Goldman, 1,25-Dihydroxyvitamin D3 inhibits proliferation of human promyelocytic
leukaemia (HL60) cells and induces monocyte-macrophage differentiation in HL60 and normal
human bone marrow cells, Leuk Res 7:51 (1983).
245. D. M. Provvedini, C. D. Tsoukas, L. J. Deftos, and S. C. Manolagas, 1,25-Dihydroxyvitamin
D3 receptors in human leukocytes. Science 221:1181 (1983).
246. A. K. Bhalla, E. P. Amento, T. L. Clemens, M. F. Holick, and S. M. Krane, Specific highaffinity
receptors for 1,25-dihydroxyvitamin D3 in human peripheral blood mononuclear
cells: presence in monocytes and induction in T lymphocytes following activation, J Clin
Endocrinol Metab 57:1308 (1983).
247. H. P. Koeffler, T. Amatruda, N. Ikekawa, and Y. Kobayashi, Induction of macrophage differentiation
of human normal and leukemic myeloid stem-cells by 1,25-dihydroxyvitamin D3
and its fluorinated analogs, Cancer Res 44:5624 (1984).
248. S. Murao, M. A. Gemmell, M. F. Callaham, N. L. Anderson, and E. Huberman, Control
of macrophage cell differentiation in human promyelocytic HL-60 leukemia cells by 1,25-
dihydroxyvitamin D3 and phorbol-12-myristate-13-acetate, Cancer Res 43:4989 (1983).
249. W. F. C. Rigby, M. Waugh, and R. F. Graziano, Regulation of human monocyte HLA-DR
and CD4 antigen expression, and antigen presentation by 1,25-dihydroxyvitamin D3. Blood
76:189 (1990).
250. S. Iho, T. Takahashi, F. Kura, H. Sugiyama, and T. Hoshino, The effect of 1,25-dihydroxyvitamin
D3 on in vitro immunoglobulin production in human B cells. J Immunol 136:4427
251. C. D. Tsoukas, D. M. Provvedini, and S. C. Manolagas, 1,25-Dihydroxyvitamin D3 a novel
immunoregulatory hormone, Science 224:1438 (1984).
252. I. Nemere and C. M. Szego, Early actions of parathyroid hormone and 1,25-dihydroxycholecalciferol
on isolated epithelial cells from rat intestine. I. Limited lysosomal enzyme release
and calcium uptake, Endocrinology 108:1450 (1981).
253. W. F. Rigby, S. Denome, and M. W. Fanger, Regulation of lymphokine production and
human T lymphocyte activation by 1,25-dihydroxyvitamin D3: specific inhibition at the level
of messenger RNA, J Clin Invest 79:1659 (1987).
254. W. F. Rigby, R. J. Nowlle, K. Krause, and M. W. Fanger, The effects of 1,25-dihydroxyvitamin
D3 on human T lymphocyte activation and proliferation: a cell cycle analysis, J Immunol
135:2279 (1985).
255. A. K. Bhalla, E. P. Amento, B. Serog, and L. H. Glimcher, 1,25-Dihydroxyvitamin D3 inhibits
antigen-induced T cell activation, J Immunol 133:1748 (1984).
256. M. A. Meehan, R. H. Kerman, and J. M. Lemire, 1,25-Dihydroxyvitamin D3 enhances the
generation of nonspecific suppressor cells while inhibiting the induction of cytotoxic cells
in a human MLR, Cell Immunol 140:400 (1992).
257. W. F. C. Rigby, S. Denome, and M. W. Fanger, Regulation of lymphokine production and
human T lymphocyte activation by 1,25-dihydroxyvitamin D3, J Clin Invest 79:1659 (1987).
258. H. Reichel, H. P. Koeffler, A. Tobler, and A. W. Norman, 1?,25-Dihydroxyvitamin D3 inhibits
gamma-interferon synthesis by normal human peripheral blood lymphocytes, Proc Natl
Acad Sci USA 84:3385 (1987).
259. A. Tobler, J. Gasson, H. Reichel, A. W. Norman, and H. P. Koeffler, Granulocyte macrophage
colony simulating factor: sensitive and receptor mediated regulation by 1,25-dihydroxyvitamin
D3 in normal human peripheral blood lymphocytes, J Clin Invest 79:1700 (1987).
260. A. Tobler, C. W. Miller, A. W. Norman, and H. P. Koeffler, 1,25-Dihydroxyvitamin D3
modulates the expression of a lymphokine (granulocyte-macrophage colony-stimulating factor)
posttranscriptionally, J Clin Invest 81:1819 (1988).
261. J. M. Lemire, Immunomodulatory role of 1,25-dihydroxyvitamin D3, J Cell Biochem 49:26
262. M. Hewison, Vitamin D and the immune system, J Endocrinol 132:173 (1992).
Vitamin D 107
263. A. D. Hess, P. M. Colobani, and A. Esa, Kidney Transplantation: Diagnosis and Treatment,
Marcel Dekker, New York, 1986, pp. 353–382.
264. M.-C. Boissier, G. Chiocchia, and C. Fournier, Combination of cyclosporine A and calcitriol
in the treatment of adjuvant arthritis, J Rheumatol 19:754 (1992).
265. C. Fournier, P. Gepner, M. Sadouk, and J. Charreire, In vivo beneficial effects of cyclosporin
A and 1,25-dihydroxyvitamin D3 on the induction of experimental autoimmune thyroiditis,
Clin Immunol Immunopathol 54:53 (1990).
266. S. T. Lillevang, J. Rosenkvist, C. B. Andersen, S. Larsen, E. Kemp, and T. Kristensen, Single
and combined effects of the vitamin D analogue KH1060 and cyclosporin A on mercuricchloride-
induced autoimmune disease in the BN rat, Clin Exp Immunol 88:301 (1992).
267. E. Lewin and K. Olgaard, The in vivo effect of a new, in vitro, extremely potent vitamin
D3 analog KH1060 on the suppression of renal allograft rejection in the rat, Calcif Tissue
Int 54:150 (1994).
268. R. G. Wong, A. W. Norman, C. R. Reddy, and J. W. Coburn, Biologic effects of 1,25-
dihydroxycholecalciferol (a highly active vitamin D metabolite) in acutely uremic rats, J
Clin Invest 51:1287 (1972).
269. R. A. Gibbs and W. H. Okamura, Studies on vitamin D (calciferol) and its analogs. XXXII.
Synthesis of 3-deoxy-1-?,25-dihydroxy-9, 11-dehydroxyvitamin-D3–selective formation of
6-?-vitamin-D vinylallenes and their thermal 1,5?-sigmatropic hydrogen shift, Tetrahedron
Lett 28: 6021 (1987).
270. W. H. Okamura, M. N. Mitra, D. A. Procsal, and A. W. Norman, Studies on vitamin D and
its analogs. VIII. 3-Deoxy-1,25-dihydroxy-vitamin D3, a potent new analog of 1,25-(OH)2-
D3, Biochem Biophys Res Commun 65:24 (1975).
271. M. F. Holick, M. Garabedian, H. K. Schnoes, and H. F. DeLuca, Relationship of 25-hydroxyvitamin
D3 side chain structure to biological activity, J Biol Chem 260: 226 (1975).
272. R. L. Johnson, W. H. Okamura, and A. W. Norman, Studies on the mode of action of
calciferol. X. 24-Nor-25-hydroxyvitamin D3, an analog of 25-hydroxyvitamin D3 having
‘‘anti-vitamin’’ activity, Biochem Biophys Res Commun 67:797 (1975).
273. M. R. Haussler, J. E. Zerwekh, R. H. Hesse, E. Rizzardo, and M. M. Peche, Biological
activity of 1?-hydroxycholecalciferol, a synthetic analog of the hormonal form of vitamin
D3, Proc Natl Acad Sci USA 70:2248 (1973).
274. M. F. Holick, P. Kasten-Schraufrogel, T. Tavela, and H. F. DeLuca, Biological activity of
1 ?-hydroxyvitamin D3 in the rat, Arch Biochem Biophys 166:63 (1975).
275. J. L. Napoli, M. A. Fivizzani, H. K. Schnoes, and H. F. DeLuca, 1Alpha-hydroxy-25-fluorovitamin
D3: a potent analogue of 1alpha, 25-dihydroxyvitamin D3, Biochemistry 17:2387
276. L. Binderup and E. Bramm, Effects of a novel vitamin D analogue MC-903 on cell proliferation
and differentiation in vitro and on calcium metabolism in vivo. Biochem Pharmacol 37:
889 (1988).
277. D. D. Bikle, E. Gee, and S. Pillai, Regulation of keratinocyte growth, differentiation, and
vitamin D metabolism by analogs of 1,25-dihydroxyvitamin D, J Invest Dermatol 101:713
278. H. Sorensen, L. Binderup, M. J. Calverley, L. Hoffmeyer, and N. R. Andersen, In vitro
metabolism of calcipotriol (MC 903), a vitamin D analogue. Biochem Pharmacol 39:391
279. S. Masuda, S. Strugnell, M. J. Calverley, H. L. J. Makin, R. Kremer, and G. Jones, In vitro
metabolism of the anti-psoriatic vitamin D analog, calcipotriol, in two cultured human keratinocyte
models, J Biol Chem 269:4794 (1994).
280. J. W. Pike, C. A. Donaldson, S. L. Marion, and M. R. Haussler, Development of hybridomas
secreting monoclonal antibodies to the chicken intestinal 1?-25-dihydroxyvitamin D3 receptor,
Proc Natl Acad Sci USA 79:7719 (1982).
281. P. H. Itin, M. R. Pittelkow, and R. Kumar, Effects of vitamin D metabolites on proliferation
108 Collins and Norman
and differentiation of cultured human epidermal keratinocytes grown in serum-free or defined
culture medium, Endocrinology 135:1793 (1994).
282. R. A. El-Azhary, M. S. Peters, M. R. Pittelkow, P. C. Kao, and S. A. Muller, Efficacy of
vitamin D3 derivatives in the treatment of psoriasis vulgaris: a preliminary report, Mayo Clin
Proc 68:835 (1993).
283. C. Nieboer and C. A. Verburgh, Psoriasis treatment with vitamin D3 analogue MC 903, Br
J Dermatol 126:302 (1992).
284. P. C. M. van de Kerkhof and E. M. G. J. De Jong, Topical treatment with the vitamin D3
analogue MC903 improves pityriasis rubra pilaris: clinical and immunohistochemcial observations,
Br J Dermatol 125:293 (1991).
285. T. Trydal, J. R. Lillehaug, L. Aksnes, and D. Aarskog, Regulation of cell growth, c-myc
mRNA, and 1,25-(OH)2 vitamin D3 receptor in C3H/10T1/2 mouse embryo fibroblasts by
calcipotriol and 1,25-(OH)2 vitamin D3, Acta Endocrinol 126:75 (1992).
286. A. J. Brown, C. R. Ritter, J. L. Finch, J. Morrissey, K. J. Martin, E. Murayama, Y. Nishii,
and E. Slatopolsky, The noncalcemic analogue of vitamin D, 22-oxacalcitriol, suppresses
parathyroid hormone synthesis and secretion. J Clin Invest 84:728 (1989).
287. J. Abe, M. Morikawa, K. Miyamoto, S. Kaiho, M. Fukushima, C. Miyaura, E. Abe, T. Suda,
and Y. Nishi, Synthetic analogues of vitamin D3 with an oxygen atom in the side chain
skeleton, FEBS Lett 226:58 (1987).
288. J. Y. Zhou, A. W. Norman, D. Chen, G. Sun, M. R. Uskokovic, and H. P. Koeffler, 1,25-
Dihydroxy-16-ene-23-yne-vitamin D3 prolongs time of leukemic mice. Proc Natl Acad Sci
USA 87: 3929 (1990).
289. A. W. Norman, J. Y. Zhou, H. L. Henry, M. R. Uskokovic, and H. P. Koeffler, Structure–
function studies on analogs of 1,25-dihydroxyvitamin D3 differential effects on leukemic cell
growth, differentiation, and intestinal calcium absorption, Cancer Res 50:6857 (1990).
290. S. J. Jung, Y. Y. Lee, S. Pakkala, S. De Vos, E. Elstner, A. W. Norman, J. Green, M. Uskokovic,
and H. P. Koeffler, 1,25(OH)2-16ene-Vitamin D3 is a potent antileukemic agent with
low potential to cause hypercalcemia, Leuk Res 18:453 (1994).
291. J. Y. Zhou, A. W. Norman, M. Lubbert, E. D. Collins, M. R. Uskokovic, and H. P. Koeffler,
Novel vitamin D analogs that modulate leukemic cell growth and differentiation with little
effect on either intestinal calcium absorption or bone calcium mobilization. Blood 74:82
292. P. M. Wovkulich, A. D. Batcho, E. G. Baggioloni, A. Boris, G. Truitt, and M. R. Uskokovic,
Synthesis of 1 ?,25S,26-trihydroxy-22-ene-cholecalciferol, a potent inducer of cell differentiation,
in Vitamin D: Chemical, Biochemical and Clinical Update (A. W. Norman, K.
Schaefer, H.-G. Grigoleit, and D. v. Herrath, eds.), Walter de Gruyter, Berlin, 1985, pp. 755–
293. Y. Tanaka, H. F. DeLuca, Y. Kobayashi, and N. Ikekawa, 26,26,26,27,27,27-Hexafluoro-
1,25-dihydroxyvitamin-D3: a highly potent, long-lasting analog of 1,25-dihydroxyvitamin-
D3, Arch Biochem Biophys 229:348 (1984).
294. M. Inaba, S. Okuno, Y. Nishizawa, K. Yukioko, S. Otani, I. Matsui-Yuasa, S. Morisawa,
H. F. DeLuca, and H. Morii, Biological activity of fluorinated vitamin D analogs at C-26
and C-27 on human promyelocytic leukemia cells, HL-60, Arch Biochem Biophys 258:421
295. K. Yukioka, S. Otani, I. Matsui-Yuasa, H. Goto, S. Morisawa, S. Okuno, M. Inaba, Y. Nishizawa,
and H. Morii, Biological activity of 26,26,26,27,27,27-hexafluorinated analogs of vitamin
D3 in inhibiting interleukin-2 production by peripheral blood mononuclear cells stimulated
by phytohemagglutinin, Arch Biochem Biophys 260:45 (1988).
296. R. Belsey, H. F. DeLuca, and J. T. Potts, Competitive binding assay for vitamin D and 25-
OH vitamin D, J Clin Endocrinol Metab 33:554 (1971).
297. Vitamin D assay, in The United States Pharmacopeia, U.S. Pharmacopeial Convention, 1995,
pp. 1756–1760.
Vitamin D 109
298. Association of Official Analytical Chemists, Conduct of the rat line test for measurement of
vitamin D activity, in Official Methods of Analyses, Washington DC, 1975, pp. 885–888.
299. Vitamins and other nutrients: AOAC official method for determination of vitamin D in milk,
vitamin preparations and feed concentrates via the rat bioassay, in Official Methods of Analysis
of AOAC International (M. J. Deutsch, ed.), AOAC International, Arlington, VA, 1995,
pp. 53–57.
300. Determination of vitamin D in vitamin D/AD concentrates, in multivitamin preparations, in
fortified milk and milk powder, in mixed feeds, pre mixes and pet foods: liquid chromatographic
method, Official Methods of Analysis of the Association of Official Analytical Chemists
(K. Helrich, ed.), Association of Official Analytical Chemists, Inc., Washington, DC,
1990, pp. 1067–1070.
301. Vitamins and other nutrients: AOAC official method of determination of vitamin D3 in poultry
feed supplements via the chick bioassay, in Official Methods of Analysis of AOAC International,
AOAC International, Arlington, VA, 1995, pp. 57–59.
302. K. A. Hibberd and A. W. Norman, Comparative biological effects of vitamins D2 and D3
and dihydrotachysterol2 and dihydrotachysterol3 in the chick, Biochem Pharmacol 18:2347
303. E. B. Olson and H. F. DeLuca, 25-Hydroxycholecalciferol: direct effect on calcium transport,
Science 165:405 (1979).
304. D. Schachter and S. M. Rosen, Active transport of Ca45 by the small intestine and its dependence
on vitamin D, Am J Physiol 196:357 (1959).
305. D. V. Kimberg, D. Schachter, and H. Schenker, Active transport of calcium by intestine:
effects of dietary calcium, Am J Physiol 200:1256 (1961).
306. H. L. Henry, A. W. Norman, A. N. Taylor, D. L. Hartenbower, and J. W. Coburn, Biological
activity of 24,25-dihydroxycholecalciferol in chicks and rats, J Nutr 106:724 (1976).
307. B. E. Miller and A. W. Norman, Enzyme-linked immunoabsorbent assay (ELISA) and radioimmunoassay
(RIA) for the vitamin D–dependent 28,000 dalton calcium-binding protein,
Meth Enzymol 102:291 (1983).
308. Vitamins and other nutrients: AOAC official methods for vitamin D in vitamin preparations,
in Official Methods of Analysis of AOAC International (M. J. Deutsch, ed.), AOAC International,
Arlington, VA, 1995, pp. 19–30.
309. C. H. Nield, W. C. Russell, and A. Zimmerli, The spectrophotometric determination of vitamin
D2 and D3, J Biol Chem 136:73 (1940).
310. A. J. Passannante and L. V. Avioli, Studies on the ultraviolet fluorescence of vitamin D and
related compounds in acid-alcohol solutions, Anal Biochem 15:287 (1966).
311. R. D. Coldwell, C. E. Porteous, D. J. Trafford, and H. L. Makin, Gas chromatography–mass
spectrometry and the measurement of vitamin D metabolites in human serum or plasma.
Steroids 49:155 (1987).
312. K. A. Tartivita, J. P. Sciarello, and B. C. Rudy, High-performance liquid chromatographic
analysis of vitamins. I. Quantitation of cholecalciferol or ergocalciferol in presence of photochemical
isomers of the provitamin and application to cholecalciferol resins. J Pharmacol
Sci 65:1024 (1976).
313. G. Jones and H. F. DeLuca, High-pressure liquid chromatography: separation of the metabolites
of vitamins D2 and D3 on small-particle silica columns, J Lipid Res 16:448 (1975).
314. F. J. Mulder, E. De Vries, and B. Borsje, Analysis of fat-soluble vitamins. XXII. High performance
liquid chromatographic determination of vitamin D in concentrates: a collaborative
study, J Assoc Off Anal Chem 62:1031 (1979).
315. P. F. Brumbaugh and M. R. Haussler, 1 ?,25-Dihydroxyvitamin D3 receptor: competitive
binding of vitamin D analogs, Life Sci 13:1737 (1973).
316. D. A. Procsal, W. H. Okamura, and A. W. Norman, Structural requirements for the interaction
of 1 ?,25-(OH)2-vitamin D3 with its chick intestinal receptor system, J Biol Chem 250:8382
110 Collins and Norman
317. T. A. Reinhardt, R. L. Horst, J. W. Orf, and B. W. Hollis, A microassay for 1,25-dihydroxyvitamin
D not requiring high performance liquid chromatography: application to clinical studies.
J Clin Endocrinol Metab 58:91 (1984).
318. B. W. Hollis and T. Kilbo, The assay of circulating 1,25(OH)2D using non-end-capped C-18
silica (C18OH): performance and validation, in Vitamin D: Molecular, Cellular and Clinical
Endocrinology (A. W. Norman, K. Schaefer, H.-G. Grigoleit, and D. v. Herrath, eds.), Walter
de Gruyter, Berlin, 1988, pp. 710–719.
319. N. Horiuchi, T. Shinki, S. Suda, N. Takahashi, S. Yamada, H. Takayama, and T. Suda, A
rapid and sensitive in vitro assay of 25-hydroxyvitamin D3-1?-hydroxylase and 24-hydroxylase
using rat kidney homogenates, Biochem Biophys Res Commun 121:174 (1984).
320. T. A. Reinhardt and R. L. Horst, Simplified assays for the determination of 25(OH)D,
24,25(OH)2D and 1,25(OH)2D, Vitamin D: Molecular, Cellular and Clinical Endocrinology
(A. W. Norman, K. Schaefer, H.-G. Grigoleit, and D. v. Herrath, eds.), Walter de Gruyter,
Berlin, 1988, pp. 720–726.
321. Report of the subcommittee on fat-soluble vitamins, WHO Bull 3:7–11 (1950).
322. Subcommittee on the Tenth Edition of the RDAs, Food and Nutrition Board, Commission
on Life Sciences, National Research Council, Recommended Dietary Allowances, 10th ed.,
National Academy Press, Washington, DC, 1989, pp. 1–285.
323. S. K. Bhowmick, K. R. Johnson, and K. R. Rettig, Rickets caused by vitamin D deficiency
in breast-fed infants in the southern United States, Am J Dis Child 145:127 (1991).
324. L. E. Booher, E. R. Hartzler, and E. M. Hewston, A compilation of the vitamin values of
foods in relation to processing and other variants, U.S. Department of Agriculture Circular
638: (1942).
325. F. R. Greer and S. Marshall, Bone mineral content, serum vitamin D metabolite concentrations,
and ultraviolet B light exposure in infants fed human milk with and without vitamin
D2 supplements, J Pediatr 114:204 (1989).
326. P. Wilton, Cod-liver oil, vitamin D and the fight against rickets, Can Med Assoc J 152:1516
327. P. Lips, F. C. van Ginkel, M. J. M. Jongen, F. Rubertus, W. J. F. van der Vijgh, and J. C.
Netelenbos, Determinants of vitamin D status in patients with hip fracture and in elderly
control subjects, Am J Clin Nutr 46:1005 (1987).
328. T. Suda, H. F. DeLuca, H. K. Schnoes, G. Ponchon, Y. Tanaka, and M. F. Holick, 21,25-
Dihydroxycholecalciferol: a metabolite of vitamin D3 perferentially active on bone, Biochemistry
9:2917 (1970).
329. S. J. Counts, D. J. Baylink, F. Shen, D. J. Sherrard, and R. O. Hickman, Vitamin D intoxication
in an anephric child, Ann Intern Med 82:196 (1975).
330. J. M. Pettifor, D. D. Bikle, M. Cavaleros, D. Zachen, M. C. Kamdar, and F. P. Ross, Serum
levels of free 1,25-dihydroxyvitamin D in vitamin D toxicity, Ann Intern Med 122:511
331. M. R. Hughes, D. J. Baylink, P. G. Jones, and M. R. Haussler, Radioligand receptor assay for
25-hydroxyvitamin D2/D3 and 1 ?,25-dihydroxyvitamin D2/D3, J Clin Invest 58:61 (1976).
332. J. M. Gertner and M. Domenech, 25-Hydroxyvitamin D levels in patients treated with highdosage
ergo- and cholecalciferol, Clin Pathol 30:144 (1977).
333. R. L. Morrissey, R. M. Cohn, R. N. Empson, H. L. Greene, O. D. Taunton, and Z. Z. Ziporin,
Relative toxicity and metabolic effects of cholecalciferol and 25-hydroxycholecalciferol in
chicks, J Nutr 107:1027 (1977).
334. J. W. Coburn and N. Brautbar, Disease states in man related to vitamin D, in Vitamin D:
Molecular Biology and Clinical Nutrition (A. W. Norman, ed.), Marcel Dekker, New York,
1980, pp. 515–577.
335. G. L. Klein, C. M. Targoff, M. E. Ament, D. J. Sherrard, R. Bluestone, J. H. Young, A. W.
Norman, and J. W. Coburn, Bone disease associated with total parenteral nutrition, Lancet
15:1041 (1980).
Vitamin D 111
336. J. K. Maesaka, V. Batuman, N. C. Pablo, and S. Shakamuri, Elevated 1,25-dihydroxyvitamin
D levels: occurrence with sarcoidosis with end-stage renal disease. Arch Intern Med 142:
1206 (1982).
337. S. Epstein, P. H. Stern, N. H. Bell, I. Dowdeswell, and R. T. Turner, Evidence for abnormal
regulation of circulating 1?,25-dihydroxyvitamin D in patients with pulmonary tuberculosis
and normal calcium metabolism, Calcif Tissue Int 36:541 (1984).
338. V. N. Hoffman and O. M. Korzenio, Leprosy, hypercalcemia, and elevated serum calcitriol
levels, Ann Intern Med 105:890 (1986).
339. A. Mudde, H., H. van den Berg, P. G. Boshuis, F. C. Breedveld, H. M. Markusse, P. M.
Kluin, O. L. Bijvoet, and S. E. Papapoulos, Ectopic production of 1,25-dihydroxyvitamin D
by B-cell lymphoma as a cause of hypercalcemia, Cancer 59:1543 (1987).
340. L. V. Avioli, S. W. Lee, J. E. McDonald, J. Lurid, and H. F. DeLuca, Metabolism of vitamin
D3-3H in human subjects: distribution in blood, bile, feces, and urine. J Clin Invest 46:983
341. J. G. Haddad and K. J. Chyu, Competitive protein-binding radioassay for 25-hydroxycholecalciferol,
J Clin Endocrinol Metab 33:992 (1971).
342. C. Christiansen, M. S. Christiansen, F. Melsen, P. Rodbro, and H. F. DeLuca, Mineral metabolism
in chronic renal failure with special reference to serum concentrations of 1,25(OH)2D
and 24,25(OH)2D, Clin Nephrol 15:18 (1981).
343. E. B. Mawer, C. M. Taylor, J. Backhouse, G. A. Lumb, and S. W. Stanbury, Failure of
formation of 1,25-dihydroxycholecalciferol in chronic renal insufficiency, Lancet 1:626
344. A. S. Brickman, J. W. Coburn, A. W. Norman, and S. G. Massry, Short-term effects of 1,25-
dihydroxycholecalciferol on disordered calcium metabolism of renal failure, Am J Med 57:
28 (1974).
345. M. R. Haussler, D. J. Baylink, M. R. Hughes, P. F. Brumbaugh, J. E. Wergedal, F. H. Shen,
R. L. Nielsen, S. J. Counts, K. M. Bursac, and T. A. McCain, The assay of 1 alpha,25-
dihydroxyvitamin D3: physiologic and pathologic modulation of circulating hormone levels,
Clin Endocrinol 5:151S (1976).
346. A. S. Brickman, J. Jowsey, D. J. Sherrard, G. Friedman, F. R. Singer, D. J. Baylink, N.
Maloney, S. G. Massry, A. W. Norman, and J. W. Coburn, Therapy with 1,25-dihydroxyvitamin
D3 in the management of renal osteodystrophy, in Vitamin D and Problems Related
to Uremic Bone Disease (A. W. Norman, K. Schaefer, H. G. Grigoleit, D. v. Herrath, and
E. Ritz, eds.), Walter de Gruyter, Berlin, 1975, pp. 241–247.
347. B. Lund, O. H. Sorensen, B. Lund, J. E. Bishop, and A. W. Norman, Vitamin D metabolism
in hypoparathyroidism, J Clin Endocrinol Metab 51:606 (1980).
348. H. Reichel and A. W. Norman, Systemic effects of vitamin D, Annu Rev Med 40:71 (1989).
349. R. W. Winters, J. B. Graham, T. F. Williams, V. W. McFalls, and C. H. Burnett, A genetic
study of familial hypophosphatemia and vitamin D resistant rickets with a review of the
literature, Medicine 37:97 (1958).
350. A. S. Brickman, J. W. Coburn, K. Kurokawa, J. E. Bethune, H. E. Harrison, and A. W.
Norman, Actions of 1,25-dihydroxycholecalciferol in patients with hypophosphatemic vitamin
D–resistant rickets, N Engl J Med 289:495 (1973).
351. C. Arnaud, R. Maijer, T. Reade, C. R. Scriver, and D. T. Whelan, Vitamin D dependency: an
inherited postnatal syndrome with secondary hyperparathyroidism, Padiatrics 46:871 (1970).
352. H. H. Ritchie, M. R. Hughes, E. T. Thompson, P. J. Malloy, Z. Hochberg, D. Feldman, J.
W. Pike, and B. W. O’Malley, An ochre mutation in the vitamin D receptor gene causes
hereditary 1,25-dihydroxyvitamin D3–resistant rickets in three families, Proc Natl Acad Sci
USA 86:9783 (1989).
353. T. Sone, R. A. Scott, M. R. Hughes, P. J. Malloy, D. Feldman, B. W. O’Malley, and J. W.
Pike, Mutant vitamin D receptors which confer hereditary resistance to 1,25-dihydroxyvitamin
D3 in humans are transcriptionally inactive in vitro, J Biol Chem 264:20230 (1989).
112 Collins and Norman
354. A. R. Rut, M. Hewison, K. Kristjansson, B. Luisi, M. R. Hughes, and J. L. H. O’Riordan,
Two mutations causing vitamin D resistant rickets: modelling on the basis of steroid hormone
receptor DNA-binding domain crystal structures, Clin Endocrinol 41:581 (1994).
355. K. Kristjansson, A. R. Rut, M. Hewison, J. L. H. O’Riordan, and M. R. Hughes, Two mutations
in the hormone binding domain of the vitamin D receptor cause tissue resistance to
1,25 dihydroxyvitamin D3, J Clin Invest 92:12 (1993).
356. T. J. Hahn, B. A. Hendin, C. R. Scharp, and J. G. Haddad, Effect of chronic anticonvulsant
therapy on serum 25-hydroxycalciferol levels in adults, N Engl J Med 287:900 (1972).
357. A. W. Norman, J. D. Bayless, and H. C. Tsai, Biologic effects of short term phenobarbital
treatment on the response to vitamin D and its metabolites in the chick, Biochem Pharmacol
25:163 (1976).
358. W. Jubiz, M. R. Haussler, T. A. McCain, and K. G. Tolman, Plasma 1,25-dihydroxyvitamin
D levels in patients receiving anticonvulsant drugs, J Clin Endocrinol Metab 44:617 (1977).
359. R. A. Corradino, Diphenylhydantoin: direct inhibition of the vitamin D3–mediated calcium
absorptive mechanism in organ-cultured duodenum. J Clin Invest 74:1451 (1976).
360. M. V. Jenkins, M. Harris, and M. R. Wills, The effect of phenytoin on parathyroid extract
and 25-hydroxycholecalciferol-induced bone resorption: adenosine 3,5 cyclic monophosphate
production, Calcif Tissue Res 16:163 (1974).
361. S. A. Mobarhan, R. M. Russell, R. R. Recker, D. B. Posner, F. L. Iber, and P. Miller, Metabolic
bone-disease in alcoholic cirrhosis: a comparison of the effect of vitamin-D2, 25-hydroxyvitamin-
D, or supportive treatment, Hepatology 4:266 (1984).
362. J. M. Barragry, R. G. Long, M. W. France, M. R. Wills, B. J. Boucher, and S. Sherlock,
Intestinal absorption of cholecalciferol in alcoholic liver disease and primary biliary cirrhosis,
Gut 20:559 (1979).
363. J. R. Bullamore, R. Wilkinson, J. C. Gallagher, B. E. Nordin, and D. H. Marshall, Effect of
age on calcium absorption, Lancet 2:535 (1970).
364. R. D. Lindeman, J. Tobin, and N. W. Shock, Longitudinal studies on the rate of decline in
renal function with age, J Am Geriatr Soc 33:278 (1985).
365. R. W. Gray, D. R. Wilz, A. E. Caldas, and J. Lemann, The importance of phosphate in
regulating plasma 1,25-(OH)2-vitamin D levels in humans: studies in healthy subjects, in
calcium-stone formers and in patients with primary hyperparathyroidism, J Clin Endocrinol
Metab 45:299 (1977).
366. A. W. Ireland, J. S. Clubb, F. C. Neale, S. Posen, and T. S. Reeve, The calciferol requirements
of patients with surgical hypoparathyroidism, Ann Intern Med 69:81 (1968).
367. N. Rochel, J. M. Wurtz, A. Mitschler, B. Klaholz, and D. Moras, The crystal structure of
the nuclear receptor for vitamin D bound to its natural ligand, Mol Cell 5:175–179 (2000).
368. J.-P. Renaud, N. Rochel, M. Ruff, V. Vivat, P. Chambon, H. Gronemeyer, and D. Moras,
Crystal structure of the RAR-y ligand-binding domain bound to all-trans retinoic acid, Nature
378:681–689 (1995).
369. W. Bourguet, M. Ruff, P. Chambon, H. Gronemeyer, and D. Moras, Crystal structure of
the ligand-binding domain of the human nuclear receptor RXR-a, Nature 375:377–382
370. W. Feng, R. C. J. Ribeiro, R. L. Wagner, H. Nguyen, J. W. Apriletti, R. J. Fletterick, J. D.
Baxter, P. J. Kushner, and B. L. West, Hormone-dependent coactivator binding to a hydrophobic
cleft on nuclear receptors, Science 280:1747–1749 (1998).
371. D. Moras and H. Gronemeyer, The nuclear receptor ligand-binding domain: structure and
function, Curr Opin Cell Biol 10:384–391 (1998).
372. S. A. Onate, V. Boonyaratanakornkit, T. E. Spencer, S. Y. Tsai, M. J. Tsai, D. P. Edwards,
and B. W. O’Malley, The steroid receptor coactivator-1 contains multiple receptor interacting
and activation domains that cooperatively enhance the activation function 1 (AF1) and AF2
domains of steroid receptors, J Biol Chem 273:12101–12108 (1998).
373. R. M. Lavinsky, K. Jepsen, T. Heinzel, J. Torchia, T. M. Mullen, R. Schiff, A. L. Del-Rio,
Vitamin D 113
M. Ricote, S. Ngo, J. Gemsch, S. G. Hilsenbeck, C. K. Osborne, C. K. Glass, M. G. Rosenfeld,
and D. W. Rose, Diverse signaling pathways modulate nuclear receptor recruitment of
N-CoR and SMRT complexes, Proc Natl Acad Sci USA 95:2920–2925 (1998).
374. J. J. Voegel, M. J. S. Heine, C. Zechel, P. Chambon, and H. Gronemeyer, TIF2, a 160 kDa
transcriptional mediator for the ligand-dependent activation function AF-2 of nuclear receptors,
EMBO J 15:3667–3675 (1996).
375. H. Hong, K. Kohli, M. J. Garabedian, and M. R. Stallcup, GRIP1, a transcriptional coactivator
for the AF-2 transactivation domain of steroid, thyroid, retinoid, and vitamin D receptors,
Mol Cell Biol 17:2735–2744 (1997).
376. J. Torchia, D. W. Rose, J. Inostroza, Y. Kamei, S. Westin, C. K. Glass, and M. G. Rosenfeld,
The transcriptional co-activator p/CIP binds CBP and mediates nuclear-receptor function,
Nature 387:677–684 (1997).
377. M. Mannervik, Y. Nibu, H. Zhang, and M. Levine, Transcriptional coregulators in development,
Science 284:606–609 (1999).
378. V. V. Ogryzko, R. L. Schiltz, V. Russanova, B. H. Howard, and Y. Nakatani, The transcriptional
coactivators p300 and CBP are histone acetyltransferases, Cell 87:953–959 (1996).
379. D. Chen, H. Ma, H. Hong, S. S. Koh, S-M. Huang, B. T. Schurter, D. W. Aswad, and M.
R. Stallcup, Regulation of transcription by a protein methyltransferase, Science 284:2174–
2177 (1999).
380. R. E. Kingston, A shared but complex bridge, Nature 399:199–200 (1999).
This Page Intentionally Left Blank
Vitamin K
University of Wisconsin, Madison, Wisconsin
The discovery of vitamin K was the result of a series of experiments by Henrik Dam on
the possible essential role of cholesterol in the diet of the chick. Dam (1) noted that chicks
ingesting diets that had been extracted with nonpolar solvents to remove the sterols developed
subdural or muscular hemorrhages and that blood taken from these animals clotted
slowly. The disease was subsequently observed by McFarlene et al. (2), who described
a clotting defect seen when chicks were fed ether-extracted fish or meat meal, and also
by Holst and Halbrook (3). Studies in a number of laboratories soon demonstrated that
this disease could not be cured by the administration of any of the known vitamins or
other known physiologically active lipids. Dam continued to study the distribution and
lipid solubility of the active component in vegetable and animal sources and in 1935
proposed (4,5) that the antihemorrhagic vitamin of the chick was a new fat-soluble vitamin,
which he called vitamin K. Not only was K the first letter of the alphabet that was not
used to described an existing or postulated vitamin activity at that time, but it was also
the first letter of the German word Koagulation. Dam’s reported discovery of a new vitamin
was followed by an independent report of Almquist and Stokstad (6,7), describing
their success in curing the hemorrhagic disease with ether extracts of alfalfa and clearly
pointing out that microbial action in fish meal and bran preparations could lead to the
development of antihemorrhagic activity.
The only plasma proteins involved in blood coagulation that were clearly defined
at that time were prothrombin and fibrinogen, and Dam et al. (8) succeeded in preparing
a crude plasma prothrombin fraction and demonstrating that its activity was decreased
when it was obtained from vitamin K–deficient chick plasma. At about the same period
of time, the hemorrhagic condition resulting from obstructive jaundice or biliary problems
was shown to be due to poor utilization of vitamin K by these patients, and the bleeding
116 Suttie
episodes were attributed to a lack of plasma prothrombin. The prothrombin assays used
at that time were not specific for prothrombin, and it was widely believed that the defect
in the plasma of animals fed vitamin K–deficient diets was due solely to a lack of prothrombin.
A real understanding of the various factors involved in regulating the generation
of thrombin from prothrombin did not begin until the mid-1950s, and during the next 10
years, factors VII, IX, and X were discovered and shown to be dependent on vitamin K
for synthesis.
A number of groups were involved in the attempts to isolate and characterize this
new vitamin, and Dam’s collaboration with Karrer of the University of Zurich resulted
in the isolation of the vitamin from alfalfa as a yellow oil. Subsequent studies soon established
that the active principle was a quinone and vitamin K1 was characterized as 2-
methyl-3-phytyl-1,4-naphthoquinone (9) and synthesized by Doisy’s group in St. Louis.
Their identification was confirmed by independent synthesis of this compound by Karrer
et al. (10), Almquist and Klose (11), and Fieser (12). The Doisy group also isolated a
form of the vitamin from putrified fish meal, which in contrast to the oil isolated from
alfalfa was a crystalline product. Subsequent studies demonstrated that this compound
called vitamin K2, contained an unsaturated side chain at the 3-position of the naphthoquinone
ring. Early investigators recognized that sources of the vitamin, such as putrified
fish meal, contained a number of different vitamins of the K2 series with differing chainlength
polyprenyl groups at the 3-position. Early observations suggested that the alkylated
forms of vitamin K could be formed in animal tissues from the parent compound, menadione.
This was not definitely established until Martius and Esser (13) demonstrated that
they could isolate a radioactive polyprenylated form of the vitamin from tissues of rats
fed radioactive menadione. Much of the early history of the discovery of vitamin K has
been reviewed by Almquist (14), and he and others (15–17) have reviewed the literature
in this field shortly after discovery of the vitamin.
A. Isolation
Vitamin K can be isolated from biological material by standard methods used to obtain
physiologically active lipids. The isolation is always complicated by the small amount of
desired product in the initial extracts. Initial extractions are usually made with the use of
some type of dehydrating conditions, such as chloroform-methanol, or by first grinding
the wet tissue with anhydrous sodium sulfate and then extracting it with acetone followed
by hexane or ether. Large samples (kilogram quantities) of tissues can be extracted with
acetone alone, and this extract is partitioned between water and hexane to obtain the crude
vitamin. Small samples, such as in vitro incubation mixtures or buffered subcellular fractions,
can be effectively extracted by shaking the aqueous suspension with a mixture of
isopropanol and hexane. The phases can be separated by centrifugation and the upper
layer analyzed directly. Methods for the efficient extraction of vitamin K from various
food matrices have been developed (18) and will be useful in developing accurate food
composition data.
Crude nonpolar solvent extracts of tissues contain large amounts of contaminating
lipid in addition to the desired vitamin. Further purification and identification of vitamin
K in this extract can be facilitated by a preliminary fractionation of the crude lipid extract
on hydrated silicic acid (19). A number of the forms of the vitamin can be separated from
each other and from other lipids by reversed-phase partition chromatography, as described
Vitamin K 117
by Matschiner and Taggart (20). These general procedures appear to extract the majority
of vitamin K from tissues. Following separation of the total vitamin K fraction from much
of the contaminated lipid, the various forms of the vitamin can be separated by the procedures
described in Sec. III.
B. Structure and Nomenclature
The nomenclature of compounds possessing vitamin K activity has been modified a number
of times since the discovery of the vitamin. The nomenclature in general use at the
present time is that most recently adopted by the IUPAC-IUB Subcommittee on Nomenclature
of Quinones (21). The term vitamin K is used as a generic descriptor of 2-
methyl-1,4-naphthoquinone (I) and all derivatives of this compound that exhibit an
antihemorrhagic activity in animals fed a vitamin K–deficient diet. The compound 2-
methyl-3-phytyl-1,4-naphthoquinone (II) is generally called vitamin K1, but is preferably
called phylloquinone. The USP nomenclature for phylloquinone is phytonadione. The
compound, first isolated from putrified fish meal and called at that time vitamin K2, is
one of a series of vitamin K compounds with unsaturated side chains, called multiprenylmenaquinones,
that are found in animal tissues and bacteria. This particular menaquinone
(2-methyl-3-farnesylgeranylgeranyl-1,4-naphthoquinone) had 7 isoprenoid units, or 35
carbons in the side chain; it was once called vitamin K2 (35) but now is called menaquinone-
7 (MK-7) (III). Vitamins of the menaquinone series with up to 13 prenyl groups
have been identified, as well as several partially saturated members of this series. The
parent compound of the vitamin K series, 2-methyl-1,4-naphthoquinone, has often been
called vitamin K3 but is more commonly and correctly designated as menadione.
C. Structures of Important Analogs, Commercial Forms, and
1. Analogs and Their Biological Activity
Following the discovery of vitamin K, a number of related compounds were synthesized
in various laboratories and their biological activity compared with that of the isolated
forms. A large number of compounds were synthesized by the Fieser group (22), and data
on the biological activity of these and other compounds have been reviewed and summarized
elsewhere (23,24). The data from some of the early studies are somewhat difficult
to compare because of variations in methods of assay, but a number of generalities were
apparent rather early. Although there were early suggestions that menadione (I) might be
118 Suttie
functioning as a vitamin, it is now usually assumed that the compound is alkylated to a
biologically active menaquinone either by intestinal microorganisms or by tissue alkylating
enzymes. The range of compounds that can be utilized by animals (or intestinal bacteria)
is wide, and compounds such as 2-methyl-4-amino-1-naphthol or 2-methyl-1-naphthol
have biological activity similar to that of menadione when fed to animals. Compounds
such as the diphosphate, the disulfate, the diacetate, and the dibenzoate of reduced vitamin
K series have been prepared and have been shown to have full biological activity. Most
early studies compared the activities of various compounds with that of menadione. The
2-methyl group is usually considered essential for activity, and alterations at this position,
such as the 2-ethyl derivative (IV), result in inactive compounds. This is not due to the
inability of the 2-ethyl derivative to be alkylated, as 2-ethyl-3-phytyl-1,4-naphthoquionone
is also inactive. There is some evidence from feeding experiments that the methyl group
may not be absolutely essential. 2-Phytyl-1,4-naphthoquionone (V) has biological activity,
but the available data do not establish that V functions as a biologically active desmethylphylloquinone
rather than being methylated to form phylloquinone.
Studies with substituted 2-methyl-1,4-naphthoquinones have revealed that polyisoprenoid
side chains are the most effective substituents at the 3-position. The biological
activity of phylloquinone, 2-methyl-3-phytyl-1,4-naphthoquinone (II), is reduced by saturation
of the double bond to form 2-methyl-3-(?,?-dihydrophytyl)-1,4-naphthoquinone
(VI). This compound is, however, considerably more active than 2-methyl-3-octadecyl-
1,4-naphthoquinone (VII), which has an unbranched alkyl side chain of similar size. Natural
phylloquinone is the trans isomer, and although there has been some confusion in the
past, Matschiner and Bell (25) have shown that the cis isomer of phylloquinone (VIII) is
essentially inactive. The naphthoquinone nucleus cannot be altered appreciably, as methylation
to form 2,6-dimethyl-3-phytyl-1,4-naphthoquinone (IX) results in loss of activity,
and the benzoquinone most closely corresponding to phylloquinone, 2,3,5-trimethyl-6-
phytyl-1,4-benzoquinone, has been reported (22) to have no activity.
Vitamin K 119
The activity of various structural analogs of vitamin K in whole-animal assay systems
is, of course, a summation of the relative absorption, transport, metabolism, and
effectiveness of this compound at the active site as compared with that of the reference
compound. Much of the data on biological activity of various compounds were obtained
by the use of an 18-h oral dose curative test utilizing vitamin K–deficient chickens. This
type of assay allows metabolic alterations of the administered form to an active form of
the vitamin, and significant activity of a compound in this assay may result from bioconversion
to an active form. Activity varies with length of the isoprenoid side chain, and
isoprenalogs with three to five isoprenoid groups have maximum activity (24) when administered
orally. The lack of effectiveness of higher isoprenalogs in this type of assay
may be due to the relatively poor absorptions of these compounds. Matschiner and Taggart
(26) have shown that when the intracardial injection of vitamin K in deficient rats is used
as a criterion, the very high molecular weight isoprenalogs of the menaquinone series are
the most active; maximum activity was observed with MK-9 (Table 1). Structure–function
relationships of vitamin K analogs have also been studied utilizing in vitro assays of the
vitamin K–dependent ?-glutamylcarboxylase, and these will be discussed in Sec. VII.
2. Commercial Form of Vitamin K
Only a few forms of the vitamin are commercially important. The major use of vitamin
K in the animal industry is in poultry diets. Chicks are very sensitive to vitamin K restriction,
and antibiotics that decrease intestinal vitamin synthesis are often added to poultry
diets. Supplementation is therefore required to ensure an adequate supply. Phylloquinone
is too expensive for this purpose, and different forms of menadione have been used. Menadione
itself possesses high biological activity in a deficient chick, but its effectiveness
depends on the presence of lipids in the diet to promote absorption. There are also problems
of its stability in feed products, and because of this, water-soluble forms are used.
Menadione forms a water-soluble sodium bisulfite addition product, menadione sodium
bisulfite (MSB) (X), which has been used commercially but which is also somewhat unstable
in mixed feeds. In the presence of excess sodium bisulfite, MSB crystallizes as a
complex with an additional mole of sodium bisulfite; this complex, known as menadione
Table 1 Effect of Route of Administration on Biological Activitya
Relative biological activity
series Menaquinone series
Number of C atoms
in side chain Oral (chick) Oral (chick) Intracardial (rat)
10 10 15 2
15 30 40 —
20 100 100 13
25 80 120 15
30 50 100 170
35 — 70 1700
40 — 68 —
45 — 60 2500
50 — 25 1700
a Data are expressed on a molar basis with phylloquinone assigned a value of 100.
Source: From Refs. 24 and 26.
120 Suttie
sodium bisulfite complex (MSBC) (XI), has increased stability and it is widely used in
the poultry industry. A third water-soluble compound is a salt formed by the addition
of dimethylpyridionol to MSB; it is called menadione pyridionol bisulfite (MPB) (XII).
Comparisons of the relative biopotency of these compounds have often been made on the
basis of the weight of the salts rather than on the basis of menadione content, and this
has caused some confusion in assessing their value in animal feeds. However, a number
of studies (27–29) have indicated that MPB is somewhat more effective in chick rations
than in MSBC. This form of the vitamin has also been demonstrated to be effective in
swine rations (30).
The clinical use of vitamin K is largely limited to two forms. A water-soluble form
of menadione, menadiol sodium diphosphate, which is sold as Kappadione or Synkayvite,
is still used in some circumstances, but the danger of hyperbilirubinemia associated with
menadione usage (see Sec. XI) had led to the use of phylloquinone as the desired form of
the vitamin. Phylloquinone (USP phytonadione) is sold as AquaMEPHYTON, Konakion,
Mephyton, and Mono-Kay.
3. Antagonists of Vitamin Action
The history of the discovery of the first antagonists of vitamin K, the coumarin derivatives,
has been documented and discussed by Link (31). A hemorrhagic disease of cattle, traced
to the consumption of improperly cured sweet clover hay, was described in Canada and
the U.S. Midwest in the 1920s. If serious hemorrhages did not develop, animals could be
aided by transfusion with whole blood from healthy animals, and by the early 1930s it
was established that the cause of the prolonged clotting times was a decrease in the concentration
of prothrombin in the blood. The compound present in spoiled sweet clover that
was responsible for this disease had been studied by a number of investigators but was
finally isolated and characterized as 3,3?-methylbis-(4-hydroxycoumarin) by Link’s group
(32,33) during the period from 1933 to 1941 and was called dicumarol (XIII). Dicumarol
was successfully used as a clinical agent for anticoagulant therapy in some early studies,
and a large number of substituted 4-hydroxycoumarins were synthesized both in Link’s
laboratory and elsewhere. The most successful of these, both clinically for long-term lowering
of the vitamin K–dependent clotting factors and subsequently as a rodenticide, have
been warfarin, 3-(?-acetonylbenzyl)-4-hydroxycoumarin (XIV), or its sodium salt; phenprocoumon,
3-(1-phenylpropyl)-4-hydroxycoumarin (XV); and ethyl biscovmacetate,
3,3?-carboxymethylenebis-(4-hydroxycoumarin) ethyl ester (XVI). The various drugs that
Vitamin K 121
have been used differ in the degree to which they are absorbed from the intestine, in their
plasma half-life, and presumably in their effectiveness as a vitamin K antagonist in the
active site. Because of this, their clinical use differs. Much of the information on the
structure–activity relationships of the 4-hydroxycoumarins has been reviewed by Renk
and Stoll (34). These drugs are synthetic compounds, and although the clinically used
compound is the racemic mixture, studies of the two optical isomers of warfarin have
shown that they differ both in their effectiveness as anticoagulants and in the influence
of other drugs on their metabolism. The clinical use of these compounds and many of
their pharmacodynamic interactions have been reviewed by O’Reilly (35).
Warfarin is widely used as a rodenticide, and concern has been expressed in recent
years because of the identification of anticoagulant-resistant rat populations. These were
first observed in northern Europe (39) and subsequently in the United States (37). Resistance
is now a significant problem in both North America (38) and Europe (39), and
concern over the spread of resistance has led to the synthesis of more effective coumarin
derivatives. Two of the most promising appear to be 3-(3-p-diphenyl-1,2,3,4-tetrahydronaphth-
1-yl)-4-hydroxycoumarin, difenacoum (XVII) and 3-(3-[4?-bromobiphenyl-4-
yl]-1,2,3,4-tetrahydronaphth-1-yl)-4 hydroxycoumarin, bromodifenacoum (XVIII). The
genetics of resistance and the differences between various resistant strains are now better
understood (40), and it appears that the problems can be brought under control.
A second class of chemical compounds having anticoagulant activity that can be
reversed by vitamin K administration (41) are the 2-substituted 1,3-indandiones. Many of
122 Suttie
these compounds also have been synthesized, and two of the more commonly used members
of the series have been 2-phenyl-1,3-indandione (XIX) and 2-pivalyl-1, 3-indandione
(XX). These compounds have had some commercial use as rodenticides, but because of
the potential for hepatic toxicity (35), they are no longer used clinically. Studies on the
mechanism of action of these compounds have not been as extensive as those on the 4-
hydroxycoumarins, but the observations that warfarin-resistant rats are also resistant to
the indandiones and their effects on vitamin K metabolism (42) suggest that the mechanism
of action of the indandiones is similar to that of the 4-hydroxycoumarins.
During the course of a series of investigations into the structural requirements for
vitamin K activity, it was shown (43) that replacement of the 2-methyl group of phylloquinones
by a chlorine atom to form 2-chloro-3-phytyl-1,4-naphthoquinone (XXI) or by a
bromine atom to form 2-bromo-3-phytyl-1,4-naphthoquinone resulted in compounds that
were potent antagonists of vitamin K. The most active of these two compounds is the
chloro derivative (commonly called chloro-K. Lowenthal (44) has shown that, in contrast
to the coumarin and indandione derivatives, chloro-K acts like a true competitive inhibitor
of the vitamin at its active site(s). Because its mechanism of action is distinctly different
from that of the commonly used anticoagulants, chloro-K has been used to probe the
mechanism of action of vitamin K, and, as it is an effective anticoagulant in coumarin
anticoagulant–resistant rats (45), it has been suggested as a possible rodenticide. Lowenthal’s
studies of possible agonists and antagonists of vitamin K indicated (46) that, in
contrast to earlier reports (22), some of the para-benzoquinones do have biological activity.
The benzoquinone analog of vitamin K1, 2,5,6-trimethyl-3-phytyl-1,4-benzoquinone,
was found to have weak vitamin K activity, and 2-chloro-5,6-dimethyl-3-phytyl-1,4-
benzoquinone (XXII) is an antagonist of the vitamin. When these compounds were modi-
fied to contain shorter isoprenoid side chains at the 3-position, they were neither agonists
nor antagonists.
Two compounds that appear to be rather unrelated to either vitamin K or the coumarins
and that have anticoagulant activity have recently been described. Marshall (47) has
shown that 2,3,5,6-tetrachloro-4-pyridinol (XXIII) has anticoagulant activity, and on the
basis of its action in warfarin-resistant rats (42), it would appear that it is functioning as
a direct antagonist of the vitamin, as does chloro-K. A second series of compounds even
less structurally related to the vitamin are the 6-substituted imidazole-[4-5-b]-pyrimidines.
These compounds were described by Bang et al. (48) as antagonists of the vitamin, and
Vitamin K 123
the action of 6-chloro-2-trifluoromethylimidazo-[4,5-b]-pyrimidine (XXIV) in warfarinresistant
rats (48,49) suggests that they function in the same way as a coumarin or indandione
type of compound. As is the case with potentially active forms of the vitamin,
studies of vitamin K antagonists have more recently been studied utilizing in vitro assays
and will be discussed in Sec. VII.
A hypoprothrombinemia can also be produced in some species by feeding animals
sulfa drugs and antibiotics. There is little evidence that these compounds are doing anything
other than decreasing intestinal synthesis of the vitamin by altering the intestinal
flora. They should, therefore, not be considered antagonists of the vitamin.
D. Synthesis of Vitamin K
The methods used in the synthesis of vitamin K have remained essentially those originally
described by Doisy’s group (50), Almquist and Klose (11,51), and Fieser (12,52) in 1939.
Those procedures involved the condensation of phytol or its bromide with menadiol or
its salt to form the reduced addition compound, which is then oxidized to the quinone.
Purification of the desired product from unreacted reagents and side products occurred
either at the quinol stage or after oxidation. These reactions have been reviewed in considerable
detail, as have methods to produce the specific menaquinones rather than phylloquionone
(53,54). The major side reactions in this general scheme is the formation of the cis
rather than the trans isomer at the ?2 position and alkylation at the 2- rather than the 3-
position to form the 2-methyl-2-phytyl derivative. The use of monoesters of menadiol and
newer acid catalysts for the condensation step (55) is the basis for the general method of
industrial preparation used at the present time. Naruta and Maruyama (56) have described
a new method for the synthesis of compounds of the vitamin K series based on the coupling
of polyprenyltrimethyltins to menadione. This method is a regio- and stereocontrolled
synthesis that gives a high yield of the desired product. It is likely that this method may
have particular utility in the synthesis of radiolabeled vitamin K for metabolic studies, as
the purification of the desired product appears to be somewhat simpler than with the synthesis
currently in use.
E. Physical and Chemical Properties
Compounds with vitamin K activity are substituted 1,4-naphthoquinones and, therefore,
have the general chemical properties expected of all quinones. The chemistry of quinoids
Fig. 1 Physical properties of phylloquinone. (A) Ultraviolet absorption spectra in petroleum ether.
(B) Infrared absorption spectra. (C) Nuclear magnetic resonance spectra in CDCl3 at 60 Mc. (D)
Mass fragmentation spectrum of the parent molecular ion is seen at m/e 450.
Vitamin K 125
has been reviewed in a book edited by Patai (57), and much of the data on the spectral
and other physical characteristics of phylloquinone and the menaquinones have been summarized
by Sommer and Kofler (58) and Dunphy and Brodie (59). The oxidized form of
the K vitamins exhibits an ultraviolet (UV) spectrum that is characteristic of the naphthoquinone
nucleus, with four distinct peaks between 240 and 280 nm and a less sharp absorption
at around 320–330 nm. The extinction coefficient (E1%
1cm) decreases with chain length
and has been used as a means of determining the length of the side chain. The molar
extinction value  for both phylloquinone and the various menaquinones is about 19,000.
The absorption spectrum changes drastically upon reduction to the hydroquinone, with
an enhancement of the 245-nm peak and disappearance of the 270-nm peak. These compounds
also exhibit characteristic infrared and nuclear magnetic resonance (NMR) absorption
spectra that are again largely those of the naphthoquinone ring. NMR analysis of
phylloquinone has been used to establish firmly that natural phylloquinone is the trans
isomer and can be used to establish the cis–trans ratio in synthetic mixtures of the vitamin.
Mass spectroscopy has been useful in determining the length of the side chain and the
degree of saturation of vitamins of the menaquinone series isolated from natural sources.
The UV, infrared, NMR and mass fragmentation spectra of phylloquinone are shown in
Fig. 1. Phylloquinone is an oil at room temperature; the various menaquinones can easily
be crystallized from organic solvents and have melting points from 35 to 60°C, depending
on the length of the isoprenoid chain.
Vitamin K can be analyzed by a variety of color reactions or by direct spectroscopy (58–
60). Chemical reactivity of the vitamin is a function of the naphthoquinone nucleus, and
as other quinones also react with many of the colorimetric assays, they lack specificity
for the vitamin. The number of interfering substances present in crude extracts is such
that a significant amount of separation is required before UV absorption spectra can be
used to quantitate the vitamin in anything other than reagent grade preparations. These
methods are therefore of little value in the determination of the small amount of vitamin
present in natural sources. Early quantitation determinations depended on a biological
assay, and analytical methods suitable for the small amounts of vitamin K present in tissues
and most food sources have been available only recently. The separation of the extensive
mixtures of menaquinones in bacteria and animal sources was first achieved with various
thin-layer or paper chromatographic systems (53,58,59,61). All separations involving concentrated
extracts of vitamin K should be carried out in subdued light to minimize UV
decomposition of the vitamin. Compounds with vitamin K activity are also sensitive to
alkali, but they are relatively stable to an oxidizing atmosphere and to heat and can be
vacuum-distilled with little decomposition.
Interest in the quantitation of vitamin K in serum and animal tissues has led to an
increasing emphasis on the use of high-performance liquid chromatography (HPLC) as
an analytical tool to investigate vitamin K metabolism. This method was first demonstrated
to be applicable to the separation of vitamin K by Williams et al. (62). The general development
of HPLC techniques for the quantitation of phylloquinone and menaquinones in
biological materials has been reviewed (63), and specific methods will be discussed in
Secs. V and VI.
126 Suttie
The classical assay for the vitamin K content of an unknown source was based on determination
of the whole-blood clotting time of the chick (64) and at the present time is largely
of historical interest. A small amount of material to be assayed was placed in the crop of
vitamin K–deficient chicks, and the response at 20 h was compared with that of known
amounts of vitamin K. When larger amounts of material were available, they were fed
for 2 weeks and the degree of vitamin sufficiency compared with known amounts of
the vitamin. Utilization of more sensitive plasma clotting factor assays (65,66) improves
sensitivity, as does better standardization of the degree of deficiency (67,68). Some investigators
have modified the degree of hypoprothrombinemia of the test chicks by anticoagulant
administration (29,65,66), but this modification makes the assay somewhat insensitive
to menadione and other unalkylated forms of the vitamin.
Because of the ease in producing a vitamin K deficiency in this species, chicks have
most often been used in biological assays; however, rats have also been used (26). All
oral bioassay procedures are complicated by the effects of different rates and extents of
absorption of the desired nutrients from the various products being assayed. They have
been superseded by HPLC techniques and will be used in the future only to compare the
biological activity of different forms of the vitamin or to investigate nutrient interactions
that modify biological activity.
Satisfactory tables of the vitamin K content of various commonly consumed foods have
not been made available until recently. Many of the values commonly quoted have apparently
been recalculated in an unspecified way from data obtained by a chick bioassay
(69,70) that was not intended to be more than qualitative and should not in any sense be
used to give absolute values. Tables in various texts and reviews (71–73) may also contain
data from this source, as well as considerable amounts of unpublished data. A major
problem has been the lack of suitable bioassays or chemical assays for the vitamin. The
range of values reported for the same foods has therefore been wide and is illustrated in
Table 2. The significant variation observed suggests that reporting a single value for the
vitamin K content of different foods gives an erroneous impression of the confidence that
should be given such a value.
Current methodology utilizes HPLC analysis of lipid extracts, and has been reported
(18) to have a within-sample coefficient of variation for different foods in the range of
7–14% and a between-sample coefficient of variation of 9–45%. Although green leafy
vegetables have been known for some time to be the major source of vitamin K in the
diet, it is now apparent that cooking oils, particularly soybean oil and rapeseed oil (79),
are major contributors. Interest in hemorrhagic diseases of the newborn has led to a determination
of phylloquinone in cow (80) and human milk (80–82) and in infant formulas
(80). Human milk contains only 1–2 ng of phylloquinone per milliliter, which is somewhat
less than that found in cow’s milk. Infant formulas are currently supplemented with vitamin
K, providing a much higher intake than that provided by breast milk.
The data in Table 3 are taken from a survey of literature (83), which considered
most of the reported HPLC derived values for various food items and from analyses of
the FDA total diet study (83a). In general, green and/or leafy vegetables are the best
sources of the vitamin, with cooking oils being the next major sources. Due to the low
Vitamin K 127
Table 2 Determination of Vitamin K by Bioassay and Chemical Methods
Vitamin K (µg per 100 g fresh)
Bioassay Chemical assay
Food Ia IIb IIIc IVd Ve
Beans, green 14 22 46 — —
Broccoli 20 65 147 230 —
Cabbage 95 37 110 110 61
Cauliflower 136 10 27 — —
Carrots 20 — 5 — —
Peas 19 — 39 50 —
Potatoes, white 4 — 1 — —
Spinach 177 130 415 260 231
Tomato, ripe 11 — 6 — —
a Doisy (75), chick bioassay of fresh food with phylloquinone as a standard.
b Richardson et al. (74), chick bioassay of frozen food with menadione as a standard.
c Shearer et al. (76), HPLC assay for phylloquinone.
d Kodaka et al. (77), HPLC assay for phylloquinone.
e Seifert (78), gas chromatography assay of phylloquinone; data have been recalculated on a dry weight basis.
Table 3 Vitamin K Content of Ordinary Foodsa
µg Phylloquinone/100 g of edible portion
Vegetables Nuts, Oils, Seeds Fruits
Kale 817 Soybean oil 193 Avocado 40
Parsley 540 Rapeseed oil 141 Grapes 3
Spinach 400 Olive oil 49 Cantaloupe 1
Endive 231 Walnut oil 15 Bananas 0.5
Green onions 207 Safflower oil 11 Apples 0.1
Broccoli 205 Sunflower oil 9 Oranges 0.1
Brussels sprouts 177 Corn oil 3
Cabbage 147 Dry soybeans 47 Meat & Dairy
Lettuce 122 Dry kidney beans 19 Ground beef 0.5
Green beans 47 Sesame seeds 8 Chicken 0.1
Peas 36 Dry navy beans 2 Pork 0.1
Cucumbers 19 Raw peanuts 0.2 Turkey 0.1
Tomatoes 6 Tuna 0.1
Carrots 5 Grains Butter 7
Cauliflower 5 Bread 3 Cheddar cheese 3
Beets 3 Oat meal 3 3.5% Milk 0.3
Onions 2 White rice 1 Yogurt 0.3
Potatoes 0.8 Wheat flour 0.6 Skim milk 0.1
Sweet corn 0.5 Dry spaghetti 0.2 Mayonnaise 81
Mushrooms 0.1 Shredded wheat 0.7 Egg yolk 2
Corn flakes 0.1 Egg white 0.1
a Values are taken from a provisional table (83) and are median values from a compilation of reported assays.
Only HPLC assays are included in the database.
128 Suttie
dietary requirement for vitamin K, a few servings of foods with high vitamin K content
contribute a significant amount toward satisfying this requirement, and an uncomplicated
dietary deficiency is unlikely. Utilizing current food composition data and 14-day food
diaries obtained from 2000 households, the mean adult phylloquinone intake in the United
States has been determined to be 80 µg for men and 73 µg for women (83b). Dihydrophylloquinone
formed during hydrogenation of cooking oils has lower biological activity and
is present in the U.S. diet at about 20% of the amount of phylloquinone. Because most
dietary vitamin K is present in a few foods, it is possible (84) to construct palatable research
diets that meet the recommended dietary allowances for other nutrients but that
have a phylloquinone content of less than 10 µg/day.
The vitamin K content of individual food items is probably more variable than that
of many other nutrients. Significant changes in the vitamin K content of plants during
growth and maturity and at various geographical locations have been reported (85). The
effect of food processing and cooking on vitamin K content has not been carefully considered.
Richardson et al. (74) have investigated the effect of heat processing (canning) or
of sterilization by ionizing radiation on the vitamin K content of six vegetables. The vitamin
K content of these foods, as determined by chick bioassay, did not differ consistently
from that of the fresh frozen food, nor did storage of any of the products for 15 months
have a significant effect. Phylloquinone in cooking oils has been shown to be relatively
stable to heat but rapidly destroyed by both daylight and fluorescent light (79). Other than
being affected by light, vitamin K present in food is most likely relatively stable; however,
only limited data are available.
A. Absorption
The absorption of nonpolar lipids, such as vitamin K, into the lymphatic system depends on
their incorporation into mixed micelles, and optimal formation of these micellar structures
requires the presence of both bile and pancreatic juice (86,87). Normal subjects have been
reported (88) to excrete less than 20% of a radiolabeled 1-mg dose of phylloquinone in
the feces, but as much as 70–80% of the ingested phylloquinone was excreted unaltered
in the feces of patients with impaired fat absorption caused by obstructive jaundice, pancreatic
insufficiency, or adult celiac disease. Much of the fecal radioactivity found in studies
such as these is due to polar biliary excretion products rather than unmetabolized phylloquinone.
Very little is known about the absorption of vitamin K from different food sources.
Vitamin K in spinach (89) was found to be absorbed 13.3% as well as from a commercial
detergent-solubilized preparation (Konakion) of phylloquinone when 25 g of butter was
included with the meal, but only 1.4% as well in the absence of butter. A second study
(89a) also indicates the phylloquinone in food sources is only 15–20% as bioavailable as
added phylloquinone. These limited data suggest that the absorption of vitamin K is probably
low but may be highly dependent on food sources and total dietary composition.
The potential for absorption of the large-bowel menaquinone pool has been reviewed
(90,91). Early, rather nonphysiological, experiments (92) demonstrated that MK-9 disappeared
from the lumen of the isolated rat ileum or large intestine but did not document
its appearance in plasma or lymph. Ichihashi et al. (93) have shown that in the presence
Vitamin K 129
of bile MK-9 is absorbed via the lymphatic pathway from rat jejunum but that in the
absence of bile no uptake of MK-9 from the colon to lymph or blood occurred within 6
h. The oral administration of 1 mg mixed long-chain menaquinones to anticoagulated
human subjects has been shown (94) to effectively decrease the extent of the acquired
hypoprothrombinemia. This demonstrates that the human digestive tract can absorb these
more hydrophobic forms of the vitamin when they are presented to the small intestine
but does not address their absorption from the large bowel. In addition, a small but nutritionally
significant portion of the intestinal content of the vitamin is located not in the
large bowel but in a region where bile acid–mediated absorption could occur (95).
Menadione is widely used in poultry, swine, and laboratory animal diets as a source
of vitamin K. It can be absorbed from both the small intestine and the colon by a passive
process (96,97), and after absorption it can be alkylated to MK-4, a biologically active
form of the vitamin.
B. Transport
The lymphatic system has been demonstrated to be the major route of transport of absorbed
phylloquinone from the intestine. There is no evidence that the vitamin in the lymph is
in any way modified. Phylloquinone is transported in chylomicrons and other low-density
lipoprotein particles (91), and its clearance is affected by apolipoprotein E polymorphism
(98). Hyperlipidemic patients have elevated phylloquinone concentrations, and this increase
is associated with the triglyceride-rich very low-density lipoprotein (VLDL) fraction
of the lipoprotein pool (91,98). The clearance of an injected dose of radioactive phylloquinone
from plasma has been investigated in humans (99,100) and shown to consist
of a two-phase exponential decline in radioactivity. The first phase had a half-life of 20–
30 min and the second a half-life of 120–165 min. Although the body pool of vitamin K
cannot be calculated from such data, it can be calculated (99) that the total body pool of
vitamin K is replaced approximately every 2.5 h.
C. Plasma Concentrations
The resolution provided by the development of HPLC system (63) and the increased sensitivity
provided by newer methods of detection have made it possible to obtain a quantitative
measure of the amount of vitamin K in serum or plasma. The clinical significance
of these measurements is not yet fully established, and little information on the relationship
between dietary intake and serum levels is available. Measurements of endogenous serum
phylloquinone concentrations (0.5–2 ng/mL) require a preliminary semipreparative column
to rid the sample of contaminating lipids followed by an analytical column. The chief
alterations and improvement in methodology in recent years have been associated with
the use of different methods of detection. Early methods utilized UV detectors, which
lack sensitivity, and electrochemical detection or fluorescence detection of the vitamin
following chemical or electrochemical reduction have replaced this methodology. These
techniques have been recently reviewed (101,102).
Initial reports of plasma or serum phylloquinone concentrations were probably too
high, and it now appears (103,104) that normal fasting values are in the region of 0.5 ng/
mL (1.1 nmol/L). This is a lower concentration than is found for the other fat-soluble
vitamins. Although concentrations of individual menaquinones were at one time thought
too low to be measured, this does not appear to be the case. The combined plasma concen130
tration of MK-7 and MK-8 in three studies (105–107) are in the same range as the concentration
of phylloquinone (108), and substantial amounts of plasma MK-6 have been reported
by other investigators (109,110). The significance of these observations is not yet
known. Circulating concentrations of phylloquinone vary with dietary intake, and in animal
(111,112) or human (113–115) studies where vitamin K intake has been restricted
or in patients with decreased food intake (116) there is a rapid decrease in plasma vitamin
K concentrations.
D. Tissue Deposition and Storage
Early studies of the distribution of dietary vitamin K in tissues were hampered by the low
specific radioactivity of the vitamin available. These studies, which utilized milligram
quantities of vitamin K injected into rats, indicated that phylloquinone was specifically
concentrated and retained in the liver but that menadione was poorly retained in this organ.
Menadione was found to be widely distributed in all tissues and to be very rapidly excreted.
The metabolism of more physiological doses of the vitamin has also been studied, and
significant differences in the distribution of phylloquinone and menadione have been seen
(117,118). About 50% of a 10-µg dose of phylloquinone administered to a rat was found
in liver at 3 h, but only 2% of a 2-µg dose of menadiol diphosphate was found in the
liver at that time. The half-life of phylloquinone in rat liver is in the range of 10–15 h
(111,112,119), so hepatic concentrations are greatly dependent on recent intake. Vitamin
K distribution has also been studied (120) by the technique of whole-body radiography
following administration of radioactive menadione, phylloquinone, or MK-4. Essentially
equal body distribution of phylloquinone and MK-4 were found at 24 h. Radioactive menadione
was spread over the whole body much faster than the other two compounds, but
the amount retained in the tissues was low. Whole-body radiography also confirmed that
vitamin K is concentrated by organs other than liver, and the highest activity was seen
in the adrenal glands, lungs, bone marrow, kidneys, and lymph nodes.
Early studies of vitamin K distribution in the liver (121) depended on bioassay of
isolated fractions of normal liver and indicated that the vitamin was distributed in all
subcellular organelles. When 0.02 or 3µg of phylloquinone was injected (122) into vitamin
K–deficient rats, more than 50% of the liver radioactivity was recovered in the microsomal
fraction and substantial amounts were found in the mitochondria and cellular debris fractions.
The specific activity (picomoles vitamin K per milligram protein) of injected radioactive
phylloquinone has been studied (118), and only the mitochondrial and microsomal
fractions had a specific activity that was enriched over that of the entire homogenate, with
the highest activity in the microsomal fraction. The less biologically active cis isomer of
phylloquinone was found to be concentrated in liver microsomes (25), but the mitochondrial
fraction had the highest specific activity. Nyquist et al. (123) found the highest specific
activity (dpm/mg protein) of radioactive phylloquinone to be in the Golgi and smooth
microsomal membrane fractions. Only limited data on the distribution of menaquinones
are available, and MK-9 has been reported (124) to be preferentially localized in a mitochondrial
rather than a microsomal subcellular fraction. Factors influencing intracellular
distribution of the vitamin are not well understood, and only preliminary evidence of an
intracellular vitamin K–binding protein that might facilitate intraorganelle movement has
been presented (125). The vitamin does appear to be lost more rapidly from the cytosol
than from membrane fractions as a deficiency develops (126).
Vitamin K 131
Because of the small amounts of vitamin K in animal tissues, it has been difficult
to determine which of the vitamers are present in tissue from different species. Only
limited data are available, and they have been reviewed by Matschiner (127,128). These
data, obtained largely by thin-layer chromatography, indicate that phylloquinone is found
in the liver of those species ingesting plant material and that, in addition to this, menaquinones
containing 6–13 prenyl units in the alkyl chain are found in the liver of most species.
Some of these long-chain vitamers appear to be partially saturated. More recently, analysis
of a limited number of human liver specimens has shown that phylloquinone represents
only about 10% of the total vitamin K pool and that a broad mixture (Table 4) of menaquinones
is present. The predominant forms appear to be MK-7, MK-8, MK-10, and MK-
11. Kayata et al. (131) have reported that the hepatic menaquinone content of five 24-
month-old infants was approximately sixfold higher than that of three infants less than 2
weeks of age, and another study (132) failed to find menaquinones in neonatal livers.
Although the long-chain menaquinones are potential sources of vitamin K activity in liver,
the extent to which they are utilized is not known. A recent study (124) has demonstrated
that the utilization of MK-9 as a substrate for the vitamin K–dependent carboxylase is
only about 20% as extensive as phylloquinone when the two compounds are present in
the liver in equal concentrations. Recent data have also suggested that MK-4 may play a
unique role in satisfying the vitamin K requirement of some species or tissues. Most
analyses of liver from various species have not detected significant amounts of MK-4.
However, chicken liver has been shown (133,134) to contain more MK-4 than phylloquinone,
and some nonhepatic tissues of the rat have been shown (135) to contain much
more MK-4 than phylloquinone. The significance of these observations and the possibility
that MK-4 may play a unique role in some aspect of vitamin K nutrition is not yet
E. Metabolism
1. Alkylation of Menadione
Animals cannot synthesize the naphthoquinone ring, and this portion of the vitamin must
be furnished in the diet. Bacteria and plants synthesize this aromatic ring system from
shikimic acid, and these pathways have been reviewed (136). It appears that bacterial
synthesis of menaquinones does not usually proceed through free menadione as an intermediate,
but rather 1,4-dihydroxy-2-naphthoic acid is prenylated, decarboxylated, and then
methylated to form the menaquinones. The transformation of menadione to MK-4 in animal
tissues was first observed by Martius and Esser (13). Radioactive menadione was
used (137) to establish that menadione could be converted to a more lipophilic compound
that, on the basis of limited characterization, appeared to be MK-4. It was then demonstrated
(138) that menadione could be converted to MK-4 by an in vitro incubation of rat
or chick liver homogenates with geranylgeranyl pyrophosphate. The activity in chick liver
was much higher than that in the rat. These studies have been extended (139), and it has
been demonstrated that other isoprenoid pyrophosphates could serve as alkyl donors for
menaquinone synthesis. Early studies (140) also demonstrated that administered phylloquinone
or other alkylated forms could be converted to MK-4. It was originally believed
that the dealkylation and subsequent realkylation with a geranylgeranyl side chain occurred
in the liver, but subsequently it was concluded that phylloquinone was not converted
to MK-4 unless it was administered orally. This suggested that intestinal bacterial action
132 Suttie























Vitamin K 133
was required for the dealkylation step. More recent studies (133–135) have demonstrated
that the phylloquinone-to-MK-4 conversion is very extensive in some tissues and it has
been demonstrated that gut bacteria are not needed for this conversion (135a,135b).
2. Metabolic Degradation and Excretion
Menadione metabolism has been studied in both whole animals (142,143) and isolated
perfused livers (135). The phosphate, sulfate, and glucuronide of menadiol have been
identified in urine and bile, and studies with hepatectomized rats (145) have suggested
that extrahepatic metabolism is significant. Early studies of phylloquinone metabolism
(146) demonstrated that the major route of excretion was in the feces and that very little
unmetabolized phylloquinone was present. Wiss and Gloor (147) observed that the side
chains of phylloquinone and MK-4 were shortened by the rat to seven carbon atoms,
yielding a carboxylic acid group at the end that cyclized to form a ?-lactone. This lactone
was excreted in the urine, presumably as a glucuronic acid conjugate. The metabolism of
radioactive phylloquinone has now been studied in humans by Shearer’s group (88). They
found that about 20% of an injected dose of either 1 mg or 45 µg of vitamin K was
excreted in the urine in 3 days, and that 40–50% was excreted in the feces via the bile.
Two different aglycones of phylloquinone were tentatively identified as the 5- and 7-
carbon side-chain carboxylic acid derivatives, respectively. It was concluded that the ?-
lactone previously identified was an artifact formed by the acidic extraction conditions
used in previous studies.
More recent studies of vitamin K metabolism have been conducted following the
discovery of the significance of the 2,3-epoxide of the vitamin (XXV). This metabolite,
commonly called vitamin K oxide, was discovered by Matschiner et al. (148), who were
investigating an observation (122) that warfarin treatment caused a buildup of liver phylloquinone.
This increase in radioactive vitamin was shown to be due to the presence of a
significant amount of a metabolite more polar than phylloquinone that was isolated and
characterized as phylloquinone 2,3-epoxide. Further studies of this compound (149) revealed
that about 10% of the vitamin K in the liver of a normal rat is present as the
epoxide and that this can become the predominant form of the vitamin following treatment
with coumarin anticoagulants.
As might be expected from the effects of coumarin anticoagulants on tissue metabolism
of vitamin K, anticoagulant treatment has a profound effect on vitamin K excretion.
Disappearance of radiolabeled vitamin K from the plasma is not significantly altered by
coumarin anticoagulant administration, but the ratio of plasma vitamin K epoxide to vitamin
K does increase drastically under these conditions. Warfarin administration also
greatly increases urinary excretion and decreases fecal excretion of phylloquinone (150)
The distribution of the various urinary metabolites of phylloquinone is also substantially
altered by warfarin administration. The amounts of the two major urinary glucuronides that
have tentatively been identified as the conjugates of 2-methyl-3-(5?-carboxy-3?-methyl-
2?pentenl)-1,4-naphthoquinone (XXVI) and 2-methyl-3-(3?-carboxy-3?-methylpropyl)-
134 Suttie
1,4-naphthoquinone (XXVII) are substantially decreased, and three new metabolites appear
It is likely that there are a number of excretion products of vitamin K that have not
yet been identified. Matschiner (127) has fed double-labeled phylloquinone—a mixture
of [6,7-3H]phylloquinone and [14C]phylloquinone-(phytyl-U) with a 3H/14C ratio of 66—
to vitamin K–deficient rats and recovered radioactivity in the urine with a 3H/14C ratio of
264. This ratio is higher than would be expected for the excretion of a metabolite with
seven carbons remaining in the side chain and is more nearly that expected from the
excretion of the metabolite with five carbons remaining in the side chain. These data
suggest that some more extensively degraded metabolites of phylloquinone are also
formed. An exchange of the phytyl side chain of phylloquinone for the tetraisoprenoid
chain of MK-4 before degradative metabolism would also be expected to contribute to
an increase in the 3H/14C ratio, and current data suggest that this exchange may occur. It
is doubtful, however, that this exchange is extensive enough to account for the increased
ratio of naphthoquinone ring atoms to side-chain atoms that was observed in this study.
The available data relating to vitamin K metabolism therefore suggest that menadione
is rapidly metabolized and excreted, and that only a relatively minor portion of this
synthetic form of the vitamin gets converted to biologically active MK-4. The degradative
metabolism of phylloquinone and the menaquinones is much slower. The major products
appear to have been identified, but there may be a number of urinary and biliary products
not yet characterized. The ratio of the various metabolites formed is drastically altered
by anticoagulant administration. However, there is no evidence to indicate that the isoprenoid
forms of the vitamin must be subjected to any metabolic transformation before they
serve as a cofactor for the vitamin K–dependent carboxylases.
A. Vitamin K–Dependent Clotting Factors
Soon after Dam’s discovery of a hemorrhagic condition in chicks that could be cured by
certain plant extracts, it was demonstrated that the plasma of these chicks contained a
decreased concentration of prothrombin. At this time, the complex series of reactions
involved in the conversion of circulating fibrinogen to a fibrin clot were poorly understood.
A basis for a clear understanding of the role of the multitude of factors involved in coagulation
came with the realization that many of the proteins involved could be looked at as
zymogens that could be activated by a specific protease and that these modified proteins
could in turn activate still other zymogens. This led to the development of a generalized
‘‘cascade’’ or ‘‘waterfall’’ theory of blood coagulation, which has now been shown (152)
to be an oversimplification. As shown in Fig. 2, the key step is the activation of prothrombin
(factor II) to thrombin by the activated form of factor X (factor Xa). Factor X can be
Vitamin K 135
Fig. 2 Involvement of vitamin K–dependent clotting factors in coagulation. The vitamin K–dependent
proteins (ellipses) circulate as inactive forms of serine proteases until converted to their
active (subscript a) forms. These conversions occur in stages where an active protease, a substrate,
and a protein cofactor (triangles) form a Ca2-mediated associated with a phospholipid surface. The
protein cofactors V and VII are activated by thrombin (IIa) to achieve their full activity. The clotting
system is traditionally divided into two pathways: the extrinsic pathway, which involves a tissue
factor (TF) in addition to blood components, and an intrinsic pathway, which involves components
present in the blood. Protein C is activated by IIa in the presence of an endothelial cell protein called
thrombomodulin (TM). Protein C is not a procoagulant but rather functions in a complex with protein
S to inactivate Va and VIIIa.
activated by two pathways involving the vitamin K–dependent proteins factor VII and
factor IX. These activations are carried out in a Ca2/phospholipid-dependent manner, and
the enzymology of these conversions is now well understood. The four proteins involved in
thrombin generation were collectively called the ‘‘vitamin K–dependent clotting factors’’
for 25 years before protein C and protein S were discovered. These two proteins have an
anticoagulant role as they are able to inactivate the accessory proteins, factor Va and factor
VIIIa (153). A seventh vitamin K–dependent protein, termed protein Z, has been described
but its function is not known.
There was a period of about 40 years between the discovery of vitamin K and the
elucidation of the metabolic function of the vitamin. Theories of the participation of vitamin
K in oxidative phosphorylation somehow being related to the synthesis of specific
proteins could not be substantiated, nor could theories of the vitamin-controlled production
of specific proteins at a transcriptional level be proven. Involvement of an intracellular
precursor in the biosynthesis of prothrombin was first clearly stated by Hemker et al. (154)
who postulated that an abnormal clotting time in anticoagulant-treated patients was due
to a circulating inactive form of plasma prothrombin. Direct evidence of the presence of
a liver precursor protein was obtained when Shah and Suttie (155) demonstrated that
136 Suttie
the prothrombin produced when hypoprothrombinemic rats were given vitamin K and
cycloheximide was not radiolabeled if radioactive amino acids were administered at the
same time as the vitamin. These observations were consistent with the presence of a hepatic
precursor protein pool in the hypoprothrombinemic rat that was rapidly being synthesized
and that could be converted to prothrombin in a step that did not require protein
synthesis. This hypothesis was strengthened by direct observations (156) that the plasma
of patients treated with coumarin anticoagulants contained a protein that was antigenically
similar to prothrombin but lacked biological activity. A similar protein was first demonstrated
in bovine plasma by Stenflo (157), but it appears (158) to be present in low concentrations
or altogether absent in other species.
Studies of this ‘‘abnormal’’ prothrombin (159) demonstrated that it contained normal
thrombin, had the same molecular weight and amino acid composition, but did not
absorb to insoluble barium salts as did normal prothrombin. This difference, and the altered
calcium-dependent electrophoretic and immunochemical properties, suggested a difference
in calcium binding properties of these two proteins that was subsequently demonstrated
by direct calcium binding measurements. The critical difference in the two proteins
was the inability of the abnormal protein to bind to calcium ions, which are needed for
the phospholipid-stimulated activation of prothrombin by factor X (160). These studies
of the abnormal prothrombin clearly implicated the calcium binding region of prothrombin
as the vitamin K–dependent region. Acidic, Ca2-binding peptides could be isolated from
a tryptic digest of the fragment 1 region of normal bovine prothrombin but could not
be obtained when similar isolation procedures were applied to preparations of abnormal
prothrombin. The nature of the vitamin K–dependent modification was elucidated when
Stenflo et al. (161) succeeded in isolating an acidic tetrapeptide (residues 6–9 of prothrombin)
and demonstrating that the glutamic acid residues of this peptide were modified so that
they were present as ?-carboxyglutamic acid (3-amino-1,1,3-propanetricarboxylic acid)
residues (Fig. 3). Nelsestuen et al. (162) independently characterized ?-carboxyglutamic
acid (Gla) from a dipeptide (residues 33 and 34 of prothrombin), and these characterizations
of the modified glutamic acid residues in prothrombin were confirmed by Magnusson
Fig. 3 Structure of ?-carboxyglutamic acid (Gla) and a diagramatic representation of the prothrombin
molecule. Specific proteolysis of prothrombin by factor Xa and thrombin will cleave prothrombin
into the specific large peptides shown: fragment 1 (F-1), fragment 2 (F-2), prethrombin
1 (P-1), prethrombin 2 (P-2), and thrombin (thr). For details of the activation of prothrombin to
thrombin and for sequences of vitamin K–dependent proteins, see Ref. (146). The Gla residues in
bovine prothrombin are located at residues 7, 8, 15, 17, 20, 21, 26, 27, 30, and 33, and they occupy
homologous positions in the other vitamin K–dependent plasma proteins.
Vitamin K 137
et al. (163), who demonstrated that all 10 Glu residues in the first 33 residues of prothrombin
are modified in this fashion.
B. Vitamin K–Dependent Carboxylase
After the vitamin K–dependent step in prothrombin synthesis was shown to be the formation
of ?-carboxyglutamic acid residues, Esmon et al. (164) demonstrated that the addition
of vitamin K and H14CO3 to vitamin K–deficient rat liver microsomal preparations resulted
in the fixation of CO2 into microsomal proteins. It was possible to isolate radioactive
prothrombin from this incubation mixture and show that essentially all of the incorporated
radioactivity was present as ?-carboxyglutamic acid residues in the fragment 1 region of
prothrombin. These observations would appear to offer final proof of the biochemical role
of vitamin K as a cofactor for this microsomal glutamyl carboxylase (Fig. 4).
This enzyme activity was soon shown to be active when solubilized in a number
of detergents, and it was demonstrated (165) that the pentapeptide Phe-Leu-Glu-Glu-Val
would serve as a substrate for this enzyme. Most subsequent studies have utilized this or
a similar peptide substrate, rather than the endogenous microsomal substrates, to study
enzyme activity. The rough microsomal fraction of liver is highly enriched in carboxylase
activity, and lower but significant levels are found in smooth microsomes. Mitochondria,
nuclei, and cytosol have negligible activities. The data obtained from protease sensitivity
studies are consistent with the hypothesis that the carboxylation event occurs on the lumen
side of the rough endoplasmic reticulum (166). Details of the properties of the enzyme
can be found in a number of reviews (167–173). This carboxylation reaction does not
require ATP, and the available data are consistent with the view that the energy to drive
this carboxylation reaction is derived from the reoxidation of the reduced form of vitamin
K. This unique carboxylase requires O2, and studies ruled out the involvement of biotin
in the system. These findings and a direct study of the CO2/HCO3
 requirement indicate
that carbon dioxide rather than HCO3
 is the active species in the reaction. The vitamin
K antagonist 2-chloro-3-phytyl-1,4-naphthoquinone is an effective inhibitor of the carboxylase,
and the reduced form of this analog has been shown to be competitive with the
reduced vitamin site. Polychlorinated phenols are strong inhibitors, and substitution of a
trifluoromethyl group, a hydroxymethyl group, or a methoxymethyl group at the 2-position
also results in inhibitory compounds (174).
A review of studies of the substrate specificity at the vitamin site of the carboxylase
suggests that the only important structural features of this substrate in a detergent-solubilized
system are a 2-methyl-1,4-naphthoquinone substituted at the 3-position with a somewhat
hydrophobic group (174). Methyl substitution of the benzenoid ring has little effect
or increases the apparent Km for the reduced vitamin. A large number of low molecular
Fig. 4 The vitamin K–dependent carboxylation reaction.
138 Suttie
weight peptide substrates of the enzyme have been synthesized, and their assay has failed
to reveal any unique sequence needed as a signal for carboxylation. In general, peptides
with Glu-Glu sequences are better substrates than those with single Glu residues. Gln, DGlu,
or Homo-Glu residues have been demonstrated to be noncarboxylated residues, while
Asp residues are poorly carboxylated. Why only the first of the two adjacent Glu residues
in these synthetic, nonphysiological substrates is carboxylated by the enzyme is not yet
apparent. A consensus sequence within the Gla region that may be important for efficient
carboxylation has been identified (175), but its significance is not known. The manner by
which the enzyme carboxylates its normal physiological protein substrates is not known.
Data obtained by characterizing the partially ?-carboxylated forms of prothrombin secreted
by a dicoumarol-treated cow would suggest that carboxylation may begin at the most
amino-terminal of the 10 potential Gla sites in prothrombin and move toward the more
carboxy terminal sites (176). However, an in vitro study utilizing des-?-carboxyosteocalcin
has shown carboxylation preferentially occurring at the most carboxy terminal of the three
potential Gla sites (177), and this aspect of the carboxylation reaction is not understood.
The stereochemistry of the reactions at the glutamyl residue has also been determined
(178). The incorporation of threo- or erythro-?-fluoroglutamate into a peptide substrate
results in a stereospecific hydrogen abstraction that corresponds to the elimination of the
4-pro-S-hydrogen of the glutamyl residues. Studies with tritium-labeled substrates have
also established that the hydrogen exchange catalyzed by the enzyme in the absence of
CO2 proceeds with a stereospecific abstraction of the same 4-pro-S-hydrogen of the glutamyl
residue that is eliminated in the carboxylation reaction. The stereochemistry associated
with the addition of CO2 to the active intermediate to form Gla has been determined by
the demonstration (179) that carboxylation of 4-S-fluoroglutamate proceeds with inversion
of configuration.
The substrate Phe-Leu-Glu-Glu-Leu, tritiated at the ? carbon of each Glu residue,
was used (180) to demonstrate that the enzyme catalyzed a vitamin KH2-dependent and
O2-dependent, but CO2-independent release of tritium from this substrate, thus establishing
the role of the vitamin in removing the ? hydrogen of the Glu substrate. The 2,3-epoxide
of vitamin K is a coproduct of Gla formation, and at saturating concentrations of CO2
there is an apparent equivalent stoichiometry between epoxide formation and Gla formation
(181). At lower CO2 concentrations a large excess of vitamin K epoxide is produced.
How epoxide formation is coupled to ?-hydrogen abstraction has not yet been firmly established,
but presumably involves an oxygenated intermediate that is on the pathway to
epoxide formation. This enzyme will catalyze a vitamin KH2- and oxygen-dependent exchange
of 3H from 3H2O into the ? position of a Glu residue of a peptide substrate
(182,183). Exchange of 3H from water with the ?-carbon hydrogen is decreased as the
concentration of HCO3
 in the media is increased. Studies utilizing ?-3H-labeled Glu substrates
have also demonstrated a close association between epoxide formation, Gla formation,
and ?-CEH bond cleavage. The efficiency of the carboxylation reaction, Gla/?-CEH
bond cleaved, is independent of Glu substrate concentration, and the data suggest that this
ratio approaches unity at high CO2 concentrations (184). The available data are consistent
with the model shown in Fig. 5, which indicates that the role of vitamin K is to abstract
the ?-methyl hydrogen to leave a carbanion.
A major gap in an understanding of the mechanism of action of this enzyme has
been the assumption that abstraction of the ?-methyl hydrogen requires a strong base
(presumably formed from the vitamin) and the lack of evidence for such an intermediate.
A series of reports by Dowd and co-workers (185,186) has suggested that an initial attack
Vitamin K 139
Fig. 5 Proposed mechanism of action for the vitamin K–dependent carboxylase enzyme. The
available data support the interaction of O2 with the reduced form of vitamin K to form an oxygenated
intermediate that is sufficiently basic to abstract the ?-hydrogen of the glutamyl residue. The products
of this reaction would be vitamin K epoxide and a glutamyl carbanion. The bracketed peroxy, dioxetane,
and alkoxide intermediates have not been identified in the enzyme-catalyzed reaction but are
postulated based on model organic reactions.
of O2 at the carbonyl carbon adjacent to the methyl group results in the formation of a
dioxetane ring, which generates an alkoxide intermediate. This intermediate is hypothesized
to be the strong base that abstracts the ?-methyl hydrogen. This pathway leads to
the possibility that a second atom of molecular oxygen can be incorporated into the carbonyl
group of the epoxide product, and this activity can be followed by utilization of
either 18O2 or vitamin K with 18O incorporated into the carbonyl oxygens in the reaction.
This partial dioxygenase activity of the carboxylase has been verified by a second group
(187), and although the general scheme (188) shown in Fig. 5 is consistent with all of
the available data, the mechanism remains a hypothesis at this time.
140 Suttie
Normal functioning of the vitamin K–dependent carboxylase poses an interesting
question in terms of enzyme–substrate recognition. This microsomal enzyme recognizes
a small fraction of the total hepatic secretory protein pool and then carboxylates 9–12
Glu sites in the first 45 residues of these proteins. Cloning of the vitamin K–dependent
proteins has revealed that the primary gene products contain a very homologous ‘‘propeptide’’
between the amino terminus of the mature protein and the signal peptide (189). This
region appears from both early in vitro and in vivo lines of evidence (190) to be a ‘‘docking’’
or ‘‘recognition site’’ for the enzyme. This domain of the carboxylase substrates has
also been shown (191) to be a modulator of the activity of the enzyme by decreasing the
apparent Km of the Glu site substrate. The structural features of the propeptide domain
that are important in stimulating the activity of the enzyme have been demonstrated (192)
to be the same as those that have been shown to be involved in targeting these proteins
for carboxylation. This propeptide domain is undoubtedly of major importance in directing
the efficient carboxylation of the multiple Glu sites in these substrates.
Significant progress toward a detailed understanding of the properties of the vitamin
K–dependent carboxylase were limited for years by the lack of a purified enzyme. A
preparation that has 500–1000 times the specific activity of microsomes and can be routinely
prepared in 2 days has been available for a number of years (193). The first report
of purification of the enzyme to homogeneity (194) was of a 77-kd protein that was subsequently
acknowledged (195) to represent the purification of the endoplasmic reticulum
heat shock protein BiP or GRP78. Two microsomal proteins of 94 kd (196,197) and 98
kd (198) have been purified and claimed to be the carboxylase. The subsequent expression
of the 94-kd protein in an insect cell line that lacks any endogenous carboxylase activity
utilizing a baculovirus expression system (199) has established that this protein carries
the activity previously associated with crude preparations of the carboxylase. The role of
the 98-kd protein in the overall carboxylase mechanism, if any, is not known.
Although recombinant carboxylase is available, little information on the structure
of this enzyme is available. The affinity of the propeptide region of various vitamin K–
dependent proteins for the carboxylase has been studied (199a), as has the influence of
the Glu substrate on the epoxidase activity (199b) of the enzyme. The fate of proteins that
are not carboxylated has also been investigated, and, at least in the case of prothrombin, it
has been shown (199c) that in the rat, but not in the human or bovine, lack of carboxylation
leads to degradation in the endoplasmic reticulum rather than secretion.
The substrates for the vitamin K–dependent carboxylase contain multiple potential
Gla residues, and the available data (199d, 199e) suggest that carboxylation is an ordered
process leading to complete carboxylation once initiated. The enzyme has been shown
(199f) to be a substrate for itself, and this modification may be involved in control of the
activity of the enzyme.
C. Vitamin K–Dependent Proteins in Skeletal Tissues
Hauschka et al. (200) first reported the presence of Gla in the EDTA-soluble proteins of
chick bone and demonstrated that it was located in an abundant low molecular weight
protein that would bind tightly to BaSO4. Price et al. (201) independently discovered a
low molecular weight Gla-containing protein in bovine bone that bound tightly to hydroxyapatite
and that inhibited hydroxyapatite formation from saturated solutions of calcium
phosphate. This protein was soon sequenced and shown to contain three Gla residues in
a 49-residue sequence (MW 5700) that shows no apparent homology to the vitamin K–
Vitamin K 141
dependent plasma proteins. This protein has been called either ‘‘osteocalcin’’ or bone Gla
protein (BGP), and the terms are used interchangeably by most investigators. A second
protein, matrix Gla protein (MGP), that has been found in bone in cartilage is structurally
related to bone Gla protein (BGP) (202) and is also expressed in numerous other tissues
(203). Protein S, a vitamin K–dependent protein with an antithrombotic role in hemostasis,
has also been shown (204) to be present in bone matrix. A number of reviews of the
properties and metabolic role of these skeletal tissue components are available (205–208).
Clear evidence for the physiological role of these vitamin K–dependent bone proteins
has been difficult to obtain. Early reports tended to assume some role of osteocalcin
in bone mineralization, but no evidence to support this role has been obtained. Price and
Williamson (209) developed a vitamin K–deficient rat model in which animals were maintained
by administration of a ratio of warfarin to vitamin K that prevented hemorrhage
but resulted in bone osteocalcin levels only 2% of normal. No defect in bone size, morphology,
or mineralization was observed in these animals, and calcium homeostasis was not
impaired. Subsequent studies (210) demonstrated that when rats were maintained on this
protocol for 8 months rather than 2 months, a mineralization disorder characterized by
complete fusion of proximal tibia growth plate and cessation of longitudinal growth was
observed. Whether this response is due to a defect in osteocalcin, MGP, or protein S
synthesis is not yet known. These data do, however, suggest that a skeletal vitamin K–
dependent protein is involved in regulating the deposition of bone mineral.
Osteocalcin is synthesized in bone rather than accumulating there after synthesis in
a different organ. However, small amounts do circulate and can be measured by radioimmunoassay.
Circulating osteocalcin is four- to fivefold higher in young children than
in adults and reaches the adult level of 5–7 ng/mL at puberty. Levels are probably slightly
higher in males, and a slight increase with advanced age has been found. In males, it has
been shown that plasma osteocalcin increases in Paget’s disease, bone metastasis, renal
osteodystrophy, hyperparathyroidism, and osteopenia. In general, plasma osteocalcin increased
markedly in Paget’s disease and is also elevated in other diseases characterized
by increased resorption and increased bone formation (211,212).
A number of reports have suggested that vitamin K status might influence skeletal
health. Low concentrations of circulating vitamin K have been reported in patients with
bone fractures, and under-?-carboxylated circulating osteocalcin has been associated with
low bone density and fracture rate (208,212,214). A large multicenter study (214a) has
identified increased under-?-carboxylated osteocalcin as a predictor of hip fracture, and
low phylloquinone intake has been correlated with hip fractures (214b). If vitamin K status
is related to osteoporosis incidence, the anticoagulant-treated population should be at risk.
However, only marginal associations between warfarin treatment and lower bone mineral
density were found in a recent meta-analysis of nine studies (214c). If there is a relationship
between bone health and vitamin K status, it is unlikely to be mediated through the
rather small alterations in the amount of biologically active osteocalcin that have been
observed in these reports. Both osteocalcin and MGP have been cloned, and the phenotype
of mice lacking the osteocalcin gene was increased bone mass and stronger bones (214d)
and of MGP an increase in aortic calcification (214e).
D. Other Vitamin K–Dependent Proteins
Proteins containing Gla residues have been found in mineralized tissues other than bone
and in pathologically calcified tissues (215–217). These proteins have not been extensively
142 Suttie
characterized, and their relationship to osteocalcin is not clear. A Gla-containing protein
has been purified from kidney (218) and shown to be distinct from osteocalcin, but its
function has not been established. The reported presence of a Gla-containing protein in
liver mitochondria (219) has been questioned (220) but later confirmed by a third investigator
(221). Spermatozoa (222) and urine (223) have also been reported as sources of
specific Gla-containing proteins. Specific proteins have not been characterized following
these reports; however, two novel proline-rich proteins have been cloned (223a). The role
of these proteins has not been elucidated, but a vitamin K–dependent protein encoded by
a growth arrest–specific gene (gas-b) (224) with considerable homology to protein S appears
to be important in cellular development. These proteins are not restricted to higher
vertebrates, and proteins containing Gla residues have been found in elasmobranch species
(225), snake venoms (226), and murine gastropods (227).
These reports of purified or nearly purified proteins containing Gla residues make
it abundantly clear that the action of vitamin K–dependent carboxylase is not limited to
the hepatic modification of a few mammalian plasma proteins. With the exception of
muscle, most mammalian tissues and organs contain significant levels of carboxylase activity.
In considering the physiological role of vitamin K–dependent proteins without a
known function, the widespread use of vitamin K antagonists as clinical anticoagulants
should be considered. Patients are routinely given sufficient amounts of coumarin anticoagulants
to depress vitamin K–dependent clotting factor levels of 15–35% of normal. It
might, therefore, have been expected that a number of problems unrelated to clotting
factor synthesis would have been observed in these patients. However, widespread clinical
problems associated with an effect on synthesis of other vitamin K–dependent proteins
during routine anticoagulant therapy has not been observed. Vitamin K metabolism in
these other tissues may not be as sensitive to coumarin anticoagulants as that in liver, or
the normal level of these proteins may be in excess of that needed for their physiological
function. It may also be that these proteins are part of a secondary backup system to a
metabolic pathway that is warfarin-insensitive.
E. Metabolic Effects of 4-Hydroxycoumarins
The oral anticoagulants are not only effective antithrombotic clinical drugs (228), but
they have also been widely used as rodenticides. Vitamin K and warfarin or other 4-
hydroxycoumarins exhibit a general in vivo agonist/antagonist relationship, and early
investigators recognized that the relationship was not what would be expected for the
competition of two compounds for a single active site. When the vitamin K–dependent
carboxylase was discovered, it was shown that the enzyme was not particularly sensitive
to inhibition by these drugs and that in contrast to the situation in animals in vitro inhibition
of the carboxylase enzyme by warfarin could not be reversed by high concentrations of
vitamin. The microsomal associated activities that have been identified as being involved
in the metabolic interconversions of the liver vitamin K pool are shown in Fig. 6. In
addition to the carboxylase/epoxidase, they include a vitamin K epoxide reductase and
two or more vitamin K quinone reductases. The microsomal epoxide reductase requires
a dithiol rather than a reduced pyridine nucleotide for activity and is commonly studied
in vitro with dithiothreitol as a reductant. This membrane-associated activity has been
difficult to purify but has now been reported (228a) to be composed of a microsomal
epoxide hydrase and a second protein. The thioredoxin/thioredoxin reductase systems can
Vitamin K 143
Fig. 6 Vitamin K–related activities in rat liver microsomes. Vitamin K epoxide formed in the
carboxylation reaction is reduced by a warfarin-sensitive pathway ‘‘vitamin K epoxide reductase’’
that is driven by dithiothreitol (DTT) as a reducing agent in in vitro studies. The quinone form of
the vitamin can be reduced to the hydroquinone either by a warfarin-sensitive DTT-driven quinone
reductase or by one or more nicotinamide nucleotide–linked quinone reductases which are less
sensitive to warfarin. Warfarin and other coumarin anticoagulants do not have a significant effect
on the carboxylase/epoxidase activity.
drive the reactions in vitro, and it is possible that this is the physiologically relevant reductase
A real understanding of the mechanism of action of warfarin began when Matschiner
and colleagues (231,232) demonstrated that the 2,3-epoxide of vitamin K was a normal
metabolite in rat liver and that the ratio of vitamin K epoxide to vitamin K was increased
by warfarin administration. Although it was originally thought that these high levels of
vitamin K epoxide inhibited the carboxylase, this was shown to be unlikely (233). The
current general theory is that inhibition of vitamin K epoxide reduction by warfarin prevents
efficient recycling of the vitamin to its enzymatically active form and therefore limits
the action of the carboxylase.
A strain of wild rats that were resistant to the action of the common 4-hydroxycoumarin
anticoagulants were used to establish this theory. The vitamin K epoxide reductase
preparation obtained from livers of the warfarin-resistant rats were relatively insensitive
to inhibition by warfarin (234,235). These preparations were, however, strongly inhibited
by a second 4-hydroxycoumarin, difenacoum, which had been developed as an effective
rodenticide for control of the warfarin-resistant rat population. These data appeared to
provide the final proof that the inhibition of the epoxide reductase by warfarin was related
to its anticoagulant action. The dithiol-dependent vitamin K quinone reductase has also
been reported to be warfarin-sensitive, and it has been shown that this activity, like that of
the epoxide reductase, is less sensitive to warfarin when assayed in tissues of the warfarinresistant
rat (236,237). It is, therefore, likely that the effects of the 4-hydroxycoumarin
144 Suttie
anticoagulants involve not only the reduction of vitamin K epoxide to the quinone but also
the reduction of the quinone to the hydroquinone (238). The NADH-dependent quinone
reductases are less sensitive to warfarin inhibition and constitute a pathway for vitamin
K quinone reduction in the anticoagulant-treated animal. The presence of this pathway
explains the ability of administered vitamin K to counteract the hemorrhagic condition
resulting from a massive dose of warfarin (239).
The alteration of two enzyme activities by what has been assumed to be a single
mutation responsible for the development of warfarin resistance in the wild rat population
raises interesting questions in terms of the structural relationship of the proteins involved.
Whether or not the two microsomal dithiol-dependent activities, vitamin K epoxide reductase
and vitamin K quinone reductase, are catalyzed by the same active site, two active
sites on the same protein, or on two subunits of the same protein is unclear at present. A
significant number of data do, however, support the theory that the same enzyme catalyzes
both activities (240).
The classical method used to define an inadequate intake of vitamin K was to measure
the plasma concentration of one of the vitamin K–dependent clotting factors, prothrombin
(factor II), factor VII, factor IX, or factor X. The various tests of clotting function used
in clinical practice, which are based on the activity of these factors, have been summarized
by Denson and Biggs (241).
Whole-blood clotting times, which were used in early work, are notoriously inaccurate,
variable, and insensitive and should not be used. Tests used at present measure the
time it takes recalcified citrated or oxalated plasma to form a fibrin clot. The standard
‘‘one-stage prothrombin time’’ assay measures clotting time after the addition of calcium
and a lung or brain extract (thromboplastin) preparation to furnish phospholipids and tissue
factor. Variations of this assay have been developed, and commercial reagent kits are
available. Because of the presence of tissue extract, factor IX is bypassed, and the assay
responds to the level of prothrombin and factors VII and X. Of these, factor VII has the
shortest half-life, and its concentration decreases earliest when vitamin K action is impaired.
It is likely, therefore, that these one-stage prothrombin assays often measure the
level of factor VII rather than prothrombin. Specific assays for factors VII and X are also
available but are seldom used in studies of vitamin K sufficiency. The classic assay for
prothrombin is a ‘‘two-stage’’ assay in which thrombin is generated from prothrombin in
one tube and a sample of this added to fibrinogen in a second tube to measure thrombin
concentration. This is a much more tedious procedure; although capable of giving an
accurate measurement of the amount of prothrombin in a plasma sample, it has seldom
been used for routine assay of vitamin K status. A number of snake venom preparations
liberate thrombin from prothromin and have been used (158,242,243) to develop onestage
clotting assays for prothrombin. The enzymes in these preparations do not require
that prothrombin be present in a calcium-dependent phospholipid complex for activation,
and they will therefore activate the descarboxyprothrombin formed in vitamin K–deficient
animals. For this reason they cannot be used to monitor a vitamin K deficiency.
All of these methods depend on the observation of the formation of a fibrin clot as
an end point to assay and are not amenable to automation. The vitamin K–dependent
clotting factors are serine proteases, and in recent years there has been considerable interest
Vitamin K 145
in the development of chromogenic substrates for the assay of these proteins, particularly
factor X and prothrombin. These assays, when utilized to assay prothrombin activity,
actually measure the concentration of thrombin that has been generated from prothrombin
by various methods (242). Because they can be readily adapted to an automated analysis,
these substrates are receiving a great deal of attention as clinical tools to follow anticoagulant
therapy but have not yet been utilized in any attempt to monitor vitamin K adequacy.
Human vitamin K deficiency results in the secretion into the plasma of partially
carboxylated species of vitamin K–dependent proteins. Because they lack the full complement
of ?-carboxyglutamic acid residues, their calcium binding affinity is altered and they
can be separated from normal prothrombin by alterations in their ability to bind to barium
salts or by electrophoresis. Antibodies that are specific for these ‘‘abnormal’’ prothrombins
have been developed and can also be used to detect a vitamin K deficiency. These assays
or similar methods used to detect the concentration of under-?-carboxylated osteocalcin
have greatly increased the sensitivity with which a vitamin K deficiency can be detected
(244). It is also likely that vitamin K status is reflected in alterations of circulating levels
of the vitamin. The extremely low concentration of vitamin K in plasma made these measurements
very difficult at one time, but satisfactory HPLC methods for the determination
of plasma or serum phylloquinone have now been developed (see Sec. VI.C). The amount
of vitamin K found in ‘‘normal’’ plasma appears to be about 0.5 ng/mL, and limited
information on the response of circulating vitamin K to changes in dietary vitamin K is
currently available.
A. Animals
The establishment of a dietary vitamin K requirement for various species has been difficult.
The challenge in demonstrating dietary requirement in many species presumably comes
from the varying degrees to which they utilize the large amount of vitamin K synthesized
by intestinal bacteria and the degree to which different species practice coprophagy. A
spontaneous deficiency of vitamin K was first noted in chicks, and poultry are much more
likely to develop symptoms of a dietary deficiency than any other species. This has usually
been assumed to be due to the rapid transit rate of material through the relatively short
intestinal tract of the chick or to limited synthesis of menaquinones in this species. A
more recent study (133) suggests that limited recycling of vitamin K because of low epoxide
reductase activity may be the cause of the increased requirement.
Ruminal microorganisms synthesize large amounts of vitamin K, and ruminants do
not appear to need a source of vitamin in the diet. Deficiencies have, however, been produced
in most monogastric species. Estimations of vitamin K requirements by different
workers are difficult to compare because of the different forms of the vitamin that were
used and different methods that were employed to establish the requirement. Some studies
have utilized a curative assay in which animals were first made hypoprothrombinemic by
feeding a vitamin K–deficient diet. They were then either injected with the vitamin and
prothrombin levels assayed after a few hours or a day, or fed diets containing various
amounts of the vitamin for a number of days and the response in clotting factor synthesis
noted. Preventive or prophylactic assays have also been used and an attempt made to
determine the minimum concentration of the vitamin that must be present in a diet to
maintain normal clotting factor levels.
146 Suttie
Phylloquinone has often been used in experimental nutrition studies, whereas other
forms of vitamin K are usually used in practical rations. Menadione is usually considered
to be from 20% to 40% as effective as phylloquinone on a molar basis, but this depends
a great deal on the type of assay which is used. It is rather ineffective in a curative assay,
where the rate of its alkylation to a menaquinone is probably the rate-limiting factor, but
often shows activity nearly equal to phylloquinone in a long-term preventive assay. Practical
nutritionists have often preferred to utilize a water-soluble form of menadione, such
as menadione sodium bisulfite complex (MSBC). This compound appears to be about as
active on a molar basis as phylloquinone in poultry rations, and at least in this species,
the activity of menadione, MSBC, and phylloquinone are roughly equal on a weight basis.
Detailed discussions of the vitamin K requirements of various species are available
in articles by Scott (245), Doisy and Matschiner (246), and Griminger (67). The data
indicate that the requirement for most species falls in a range of 2–200 µg vitamin K per
kilogram body weight per day. The data in Table 5, which have been adopted from a table
presented by Griminger (67), give an indication of the magnitude of the requirement for
various species. It should be remembered that this requirement can be altered by age, sex,
or strain, and that any condition influencing lipid absorption or conditions altering intestinal
flora will have an influence of these values (see Sec. X). A considerably higher level
of dietary vitamin K has been recommended for most laboratory animals by the National
Research Council (247). Recommendations for most species are in the range of 3000 µg/
kg of diet, but the rat requirement has been set at 50 µg/kg.
Although this level is sufficient in most cases, it does not prevent all signs of defi-
ciency (111), and the American Institute of Nutrition (248) has now recommended that
purified diets for laboratory rodents should have 750 µg of phylloquinone added to each
kilogram of diet.
B. Humans
The requirement of the adult human for vitamin K is extremely low, and there seems little
possibility of a simple dietary deficiency developing in the absence of complicating factors.
Until recently, the low requirement and the relatively high levels of vitamin K found
in most diets had prevented an accurate assessment of the requirement. Frick et al. (249)
Table 5 Vitamin K Requirements of Various Speciesa
Daily intake concentration
Species (µg/kg per day) (µg/kg diet)
Dog 1.25 60
Pig 5 50
Rhesus monkey 2 60
Rat, male 11–16 100–150
Chicken 80–120 530
Turkey poult 180–270 1200
a Data have been summarized from a more extensive table (Ref. 67) and
are presented as the amount of vitamin needed to prevent the development
of a deficiency. No correction for differences in potency of equal weights
of different forms of the vitamin has been made.
Vitamin K 147
studied the vitamin K requirement of starved intravenously fed debilitated patients given
antibiotics to decrease intestinal vitamin K synthesis. They determined that 0.1 µg/kg per
day was not sufficient to maintain normal prothrombin levels and that 1.5 µg/kg per day
was sufficient to prevent any decreases in clotting factor synthesis. Their data indicate
that the requirement was of the order of 1 µg/kg per day. Two other studies (75,250) with
very limited numbers of subjects successfully decreased vitamin K intake to the extent
that clotting factor activities were lowered and also suggested that the vitamin requirement
of the human is in the range of 0.5–1.0 µg vitamin K per kilogram per day.
A major problem in determining a dietary requirement for vitamin K has been the
relative insensitivity of the commonly used prothrombin time measurement (244). A more
recent study modifying the vitamin K intake of young adults by restriction of foods with
a high phylloquinone content (251) has resulted in two symptoms of a mild deficiency:
increased circulating under-?-carboxylated prothrombin and decreased Gla excretion.
These responses were reversed by additional dietary vitamin, and the data obtained are
consistent with a requirement in the range previously suggested. A more carefully controlled
metabolic ward study (113) found alterations in the same two sensitive measures
of deficiency in subjects consuming about 10 µg phylloquinone per day, and increased
circulating under-?-carboxylated prothrombin has been observed in a second metabolic
ward controlled study (252). Based largely on these studies, the RDA for vitamin K has
been set at 1 µg phylloquinone per kilogram body weight for adults, and 5 µg phylloquinone
per day for infants from birth to 6 months and 10 µg phylloquinone per day from
6 to 12 months. The 1 µg/kg body weight was also applied to children.
This requirement may be subject to change as more data become available. Because
of the close relationship of plasma phylloquinone concentration to recent intakes, this
measurement lacks utility for assessing vitamin K status. Assessment of adequacy by use
of the somewhat insensitive one-stage prothrombin time has meant that a large decrease
in vitamin K–dependent clotting factor synthesis was needed to produce an apparent defi-
ciency. More sensitive clotting assays and the ability to immunochemically detect circulating
des-?-carboxyprothrombin (253) provide an opportunity to monitor much milder forms
of vitamin K deficiency. There are a number of reports that vitamin K status may be
important in maintaining skeletal health (214), and in recent studies (254,255,255a) the
extent of under-?-carboxylation of circulating osteocalcin has been shown to be a very
sensitive criterion of vitamin K sufficiency. It is likely that this method will become increasingly
important in future studies directed at defining the human requirement for vitamin
A. Adult Human Considerations
The human population normally consumes a diet containing a great excess of vitamin K,
but a vitamin K–responsible hypoprothrombinemia can sometimes be a clinically signifi-
cant human problem. O’Reilly (35) has reviewed the potential problem areas and has
pointed out the basic factors needed to prevent a vitamin K deficiency: (a) a normal diet
containing the vitamin (b) the presence of bile in the intestine, (c) a normal absorptive
surface in the small intestine, and (d) a normal liver. Cases of an acquired vitamin K
deficiency do, therefore, occur in the adult population and, though relatively rare, present
148 Suttie
a significant problem for some individuals. It has usually been assumed that a general
deficiency is not possible, but Hazell and Bloch (256) have observed that a relatively high
percentage of an older adult hospital-admitted population has a hypoprothrombinemia that
responds to administration of oral vitamin K. The basis for this apparent increase in vitamin
K requirement was not determined and was probably multicausal. It has, however,
been shown (257) that it is much easier to develop a vitamin K deficiency in older than
in young rats. Whether this is related to anything other than an increased intake of
nutrients/unit of body weight in the younger animals has not been determined.
B. Hemorrhagic Disease of the Newborn
A vitamin K–responsive hemorrhagic disease of the human newborn has been a long
recognized syndrome. The classical disease is seen in infants at the age of 2–5 days, but
‘‘late hemorrhagic disease’’ is a condition also seen in infants older than 1 month (258).
The newborn infant has low levels of the vitamin K–dependent clotting factors (259), and
these normally increase with time. Infants are at risk because the gut of the newborn is
relatively sterile and provides little vitamin K, and human breast milk is low in vitamin
K. Serious problems appear to be associated almost exclusively with breast-fed infants.
Because of the possibility of hemorrhagic episodes, the American Academy of Pediatrics
recommends that phylloquinone be administered parenterally to all newborn infants at a
dose of 0.5–1.0 mg. Commercial infant formulas are routinely fortified with vitamin K.
Improvement in methods for detecting mild vitamin K deficiencies (260) has led to
a renewed interest in vitamin K nutrition in the infant, and the importance of the low
vitamin K intake of breast-fed infants to the development of signs of deficiency is now
more clearly understood. A critical factor in determining the vitamin K status of the breastfed
infant has been shown to be milk intake (261,262). There is currently increased interest
in oral rather than parenteral administration of vitamin K to infants, and efforts are being
made to develop satisfactory products.
C. Effects of Drugs
Early studies (see Ref. 257) of vitamin K action and requirement established the need to
prevent coprophagy to produce a vitamin K deficiency in the rat. It is easily demonstrated
that the vitamin K requirement of the rat is greatly increased under germ-free conditions
(263) and that feeding sulfa drugs increases the vitamin K requirement of the chick (264).
The importance of menaquinones in satisfying the human requirement for vitamin K is
unclear (108). Numerous cases of vitamin K–responsive hemorrhagic events in patients
receiving antibiotics have been reviewed by Savage and Lindenbaum (265). These have
been assumed to be due to decreased menaquinone utilization by these patients, but it is
possible that many cases may represent low dietary intake alone and that the presumed
adverse drug interaction is not always related to the hypoprothrombinemia. The secondand
third-generation cephalosporins have been implicated in a large number of hypothrombinemic
episodes; although it has been suggested that these drugs have a direct effect on
the vitamin K–dependent carboxylase (266), it is more likely that they are exerting a weak
coumarin-like response (267,268).
A series of reports have demonstrated (269,270) that dietary butylated hydroxytoluene
causes a hemorrhagic condition in rats that can be cured by vitamin K supplementation.
The mechanism by which the effect is mediated has not yet been clarified. Phenobarbital
and diphenylhydantoin administration to mothers has been reported (271) to produce a
Vitamin K 149
vitamin K–responsive hemorrhage in the newborn, and clofibrate appears to alter coumarin
responsiveness (272) through an effect on vitamin K utilization of availability. The nature
of these effects has not been extensively investigated.
The widely used coumarin anticoagulants (see Secs. II.C and VII.E) effectively antagonize
the action of vitamin K and also influence metabolism of the vitamin. Clinically,
the effect of this vitamin K antagonism has been thought to be limited to their effect on
clotting factor synthesis. Oral anticoagulant therapy during the first trimester is known to
be associated with development of the fetal ‘‘warfarin syndrome’’ (273). Whether this
outcome is related to the influence of warfarin on the action of osteocalcin, matrix Gla
protein, protein S, or an unknown vitamin K–dependent protein is not known. It is also
possible that this adverse drug effect is related not to the action of warfarin as a vitamin
K antagonist but to an unrelated action of the drug.
D. Influence of Hormones
Early studies of vitamin K requirements indicated that female rats had higher plasma
prothrombin concentrations and a lower vitamin K requirement. Plasma prothrombin levels
are also higher in pregnant rats, and this increased concentration of prothrombin in
female and pregnant rats results from an effect on rate of synthesis rather than rate of
prothrombin degradation (274). Castration of both sexes unifies that vitamin K response,
and in the castrated rat, prothrombin concentrations can be increased with estrogens and
decreased with androgens (275). There are some indications (126) from studies utilizing
[3H]vitamin K1 that the estrogen effects are related to the amount of vitamin needed in
the liver to maintain normal levels of prothrombin. The available evidence suggests that
the influence of estrogens on rate of synthesis is reflected in a higher rate of synthesis
and accumulation of prothrombin precursors in the microsomes (276,277).
The effect of other hormones on vitamin K metabolism or action is less clear. Nishino
(278) has shown that hypophysectomy prevents the estrogen stimulation of prothrombin
synthesis in castrated females, and that prolactin injections protect intact males from
the development of hypoprothrombinemia. It has also been shown that hypothyroidism
in humans results in a decrease in both the rate of synthesis and destruction of the vitamin
K–dependent clotting factors (279). It is likely that the effects of these hormones are
related to rates of synthesis of the proteins involved rather than to any effect on vitamin
K metabolism. But it is possible that there are also effects on the vitamin itself.
E. Other Dietary and Disease State Factors
Early studies of vitamin K functions (280) established that the inclusion of mineral oil in
diets prevented its absorption, and mineral oil has often been used in vitamin K–deficient
diets. High dietary vitamin A has also been recognized for some time to adversely influence
vitamin K action (246). Whether this is a general effect on nonpolar lipid absorption
or a specific vitamin K antagonism is not clear, but it can be observed at relatively low
dietary levels of retinol acetate and retinoic acid (257). Dietary oxidized squalene also
has a potent hemorrhagic effect (257), and a d-?-tocopherol hydroquinone administration
has been shown to produce a vitamin K–responsive hemorrhagic syndrome in the pregnant
rat (281). High vitamin E consumption may be of some clinical significance, as the addition
of vitamin K to the diet of a patient on coumarin anticoagulant therapy has been
shown to result in a hemorrhagic episode (282). It is possible that high vitamin E intakes
may exacerbate a borderline vitamin K deficiency. There are indications that ?-tocopherol
150 Suttie
quinone, rather than ?-tocopherol (283), may be the causative agent, and this may explain
some of the differences in the reported results.
Insufficient assimilation of vitamin K can occur in adults on protracted antibiotic
treatment or those receiving long-term parenteral hyperaliminentation without vitamin K
supplementation. Malabsorption of vitamin K has also occurred as a result of obstructive
jaundice, biliary fistula, pancreatic insufficiency, steatorrhea, or chronic diarrhea. Specific
references to observations of vitamin K–responsive hemorrhagic conditions in these various
diseases states can be found in Refs. 35 and 265.
No hazards attribute to the long-term ingestion of elevated amounts of the natural forms
of vitamin K have been reported (284,285). For treatment of prolonged clotting times
when hemorrhage is not a problem, vitamin K can be given orally or parenterally. If given
orally to patients with impaired biliary function, bile salts should also be administered.
Vitamin K1 is available as the pure compound or as an aqueous colloidal solution that
can be given intramuscularly or intravenously. Some adverse reactions have been noted
following intervenous administration, and unless a severe hemorrhagic episode is present,
intramuscular infection is the recommended route of therapy. Effective therapy requires
synthesis of normal clotting factors, and a couple of hours may be necessary before a
substantial decrease in clotting times is apparent.
The relative safety of phylloquinone and, presumably, menaquinones does not hold
for menadione or its water-soluble derivatives. These compounds can be safely used at
low levels to prevent the development of a deficiency but should not be used as a pharmacological
treatment for a hemorrhagic condition. Although once prescribed for treatment
of the hemorrhagic disease of the newborn, these compounds are known to react with free
sulfhydryl groups of various tissues and to cause hemolytic anemia, hyperbilirubinemia,
and kernicterus. This marked increase in conjugated bilirubin is extremely toxic to the
neonatal brain and has caused death in some instances (284).
1. H. Dam, Cholesterinstoffwechsel in Hu?hnereiern and Hu?nchen. Biochem. Z. 215:475–492
2. W. D. McFarlane, W. R. Graham, and F. Richardson, The fat-soluble vitamin requirements
of the chick. I. The vitamin A and vitamin D content of fish meal and meat meal. Biochem.
J. 25:358–366 (1931).
3. W. F. Holst and E. R. Halbrook, A ‘‘scurvy-like’’ disease in chicks. Science 77:354 (1933).
4. H. Dam. The antihaemorrhagic vitamin of the chick. Biochem. J. 29:1273–1285 (1935).
5. H. Dam. The antihaemorrhagic vitamin of the chick. Occurrence and chemical nature. Nature
135:652–653 (1935).
6. H. J. Almquist and E. L. R. Stokstad, Dietary haemorrhagic disease in chicks. Nature 136:
31 (1935).
7. H. J. Almquist and E. L. R. Stokstad, Hemorrhagic chick disease of dietary origin. J. Biol.
Chem. 111:105–113 (1935).
8. H. Dam, F. Schonheyder, and E. Tage-Hansen, Studies on the mode of action of vitamin K.
Biochem. J. 30:1075–1079 (1936).
9. D. W. MacCorquodale, L. C. Cheney, S. B. Binkley, W. F. Holcomb, R. W. McKee, S. A.
Vitamin K 151
Thayer, and E. A. Doisy, The constitution and synthesis of vitamin K. J. Biol. Chem. 131:
357–370 (1939).
10. P. Karrer, A. Geiger, R. Legler, R. Ru?egger, and H. Salomon, U? ber die isolierung des ?-
Phyllochinones (vitamin K aus Alfalfa) sowie u?ber dessen Entdeckungsgeschechter. Helv.
Chim. Acta 22:1464–1470 (1939).
11. H. J. Almquist and A. A. Klose, Synthetic and natural antihemorrhagic compounds. J. Am.
Chem. Soc. 61:2557–2558 (1939).
12. L. F. Fieser, Synthesis of 2-methyl-3-phytyl-1,4-naphthoquinone. J. Am. Chem. Soc. 61:
2559–2561 (1939).
13. C. Martius and H. O. Esser. U? ber die Konstitution des in Tierkorper aus Methyl Naphthochinongelbildenten
K-Vitamines. Biochem Z. 331:1–9 (1958).
14. H. J. Almquist, The early history of vitamin K. Am. J. Clin. Nutr. 28:656–659 (1975).
15. H. J. Almquist, Vitamin K. Physiol. Rev. 21:194–216 (1941).
16. H. Dam, Vitamin K, its chemistry and physiology. Adv. Enzymol. 2:285–324 (1942).
17. E. A. Doisy, S. B. Binkley, and S. A. Thayer, Vitamin K. Chem. Rev. 28:477–517 (1941).
18. S. L. Booth, K. W. Davidson, and J. A. Sadowski, Evaluation of an HPLC method for the
determination of phylloquinone (vitamin K1) in various food matrices. J. Agric. Food Chem.
42:295–300 (1994).
19. J. T. Matschiner, W. V. Taggart, and J. M. Amelotti. The vitamin K content of beef liver,
detection of a new form of vitamin K. Biochemistry 6:1243–1248 (1967).
20. J. T. Matschiner and W. V. Taggart, Separation of vitamin K and associated lipids by reversed-
phase partition column chromatography. Anal. Biochem. 18:88–93 (1967).
21. IUPAC-IUB, Commission on Biochemical Nomenclature. Nomenclature of quinones with
isoprenoid side chains. Eur. J. Biochem. 53:15–18 (1975).
22. L. F. Fieser, M. Tishler, and W. L. Sampson, Vitamin K activity and structure. J. Biol. Chem.
137:659–692 (1941).
23. P. Griminger, Biological activity of the various vitamin K forms. Vitam Hormones 24:605–
618 (1966).
24. F. Weber and O. Wiss, Vitamin K group: active compounds and antagonists, in The Vitamins,
Vol. 3 (W. H. Sebrell and R. S. Harris, eds.). Academic Press, New York, 1971, pp. 457–
25. J. T. Matschiner and R. G. Bell, Metabolism and vitamin K activity of cis phylloquinone in
rats. J. Nutr. 102:625–630 (1972).
26. J. T. Matschiner and W. V. Taggart, Bioassay of vitamin K by intracardial injection in defi-
cient adult male rats. J. Nutr. 94:57–59 (1968).
27. P. Griminger, Relative vitamin K potency of two water-soluble menadione analogues. Poultry
Sci. 44:211–213 (1965).
28. P. N. Dua and E. J. Day, Vitamin K activity of menadione dimethylpyrimidinol bisulfite in
chicks. Poultry Sci. 45:94–96 (1966).
29. O. W. Charles, B. C. Dilworth, and E. J. Day, Chick bioassay of vitamin K compounds using
dicumarol and pivalyl as anticoagulants. Poultry Sci. 47:754–760 (1968).
30. R. W. Seerley, O. W. Charles, H. C. McCampbell, and S. P. Bertsch, Efficacy of menadione
dimethylpyrimidinol bisulfite as a source of vitamin K in swine diets. J. Anim. Sci. 42:599–
607 (1976).
31. K. P. Link, The discovery of dicumarol and its sequels. Circulation 19:97–107 (1959).
32. H. A. Campbell and K. P. Link. Studies on the hemorrhagic sweet clover disease. IV. The
isolation and crystallization of the hemorrhagic agent. J. Biol. Chem. 138:21–33 (1941).
33. M. A. Stahmann, C. F. Huebner, and K. P. Link, Studies on the hemorrhagic sweet clover
disease. V. Identification and synthesis of the hemorrhagic agent. J. Biol. Chem. 138:513–
527 (1941).
34. E. Renk and W. G. Stoll, Orale Antikoagulantien. Prog. Drug. Res. 11:226–355 (1968).
35. R. A. O’Reilly, Vitamin K and the oral anticoagulant drugs. Annu. Rev. Med. 27:245–261
152 Suttie
36. C. M. Boyle, Case of apparent resistance of Rattus norvegicus Berkenhout to anticoagulant
poisons. Nature 188:517 (1960).
37. W. B. Jackson and D. Kaukeinen, Resistance of wild norway rats in North Carolina to warfarin
rodenticide. Science 176:1343–1344 (1972).
38. W. B. Jackson, A. D. Ashton, and K. Delventhal, Overview of anticoagulant rodenticide
usage and resistance, in Current Advances in Vitamin K Research (J. W. Suttie, ed.). Elsevier,
Science Publisher, New York, 1988, pp. 381–388.
39. M. Lund, Detection and monitoring of resistance to anticoagulant rodenticies in populations
of brown rats (Rattus norvegicus) in Denmark, in Current Advances in Vitamin K Research
(J. W. Suttie, ed.). Elsevier, New York, 1988, pp. 399–405.
40. J. H. Greaves and P. B. Cullen-Ayres, Genetics of Difenacoum resistance in the rat, in Current
Advances in Vitamin K Research (J. W. Suttie, ed.). Elsevier, New York, 1988, pp. 389–
41. H. Kobat, E. F. Stohlman, and M. I. Smith, Hypoprothrombinemia induced by administration
of indandione derivatives. J. Pharmacol. Exp. Ther. 80:160–170 (1944).
42. P. Ren, R. E. Laliberte, and R. G. Bell, Effects of warfarin, phenylindanedione, tetrachloropyridinol,
and chloro-vitamin K1 on prothrombin synthesis and vitamin K metabolism in
normal and warfarin-resistant rats. Mol. Pharmacol. 10:373–380 (1974).
43. J. Lowenthal, J. A. MacFarlene, and K. M. McDonald, The inhibition of the antidotal activity
of vitamin K1 against coumarin anticoagulant drugs by its chloro analogue. Experientia 16:
428–429 (1960).
44. J. Lowenthal, Vitamin K analogs and mechanisms of action of vitamin K, in The Fat-Soluble
Vitamins (H. F. DeLuca and J. W. Suttie, eds.). University of Wisconsin Press, Madison,
1970, pp. 431–446.
45. J. W. Suttie, Anticoagulant-resistant rats: possible control by the use of the chloro analog
of vitamin K. Science 180:741–743 (1973).
46. J. Lowenthal and J. A. MacFarlene, Vitamin K-like and antivitamin K activity of substituted
para-benzoquinones. J. Pharmacol. Exp. Ther. 147:130–138 (1965).
47. M. N. Marshall, Potency and coagulation factor effects of 2,3,5,6-tetrachloropyridinol compared
to warfarin and its antagonism by vitamin K. Proc. Soc. Exp. Biol. Med. 139:806–
810 (1972).
48. N. U. Bang, G. O. P. O’Doherty, and R. D. Barton, Selective suppression of vitamin K–
dependent procoagulant synthesis by compounds structurally unrelated to vitamin K. Clin.
Res. 23:521A (1975).
49. P. A. Friedman and A. E. Griep, In vitro inhibition of vitamin K-dependent carboxylation
by tetrachloropyridinol and the imidazopyridines. Biochemistry 19:3381–3386 (1980).
50. S. B. Binkley, L. C. Cheney, W. F. Holcomb, R. W. McKee, S. A. Thayer, D. W. Mac-
Corquodale, and E. A. Doisy, The constitution and synthesis of vitamin K1. J. Am. Chem.
Soc. 61:2558–2559 (1939).
51. H. J. Almquist, Vitamin K: discovery, identification, synthesis, functions. Fed. Proc. 28:
2687–2689 (1979).
52. L. F. Fieser, Identity of synthetic 2-methyl-3-phytyl-1,4-napththoquinone and vitamin K1. J.
Am. Chem. Soc. 61:2561 (1939).
53. H. Mayer and O. Isler, Vitamin K group—chemistry, in The Vitamins, 2nd ed., Vol. 3 (W.
H. Sebrell and ???
54. H. Mayer and O. Isler, Synthesis of vitamins K, in Methods in Enzymology, Vol. 18C (D. B.
McCormick and L. D. Wright, eds.). Academic Press, New York, 1971, pp. 491–547.
55. H. Mayer and O. Isler, Vitamin K group—industrial preparation, in The Vitamins, 2nd ed.,
Vol. 3 (W. H. Sebrell and R. S. Harris, eds.). Academic Press, New York, 1971, pp. 444–
56. Y. Naruta and K. Maruyama, Regio- and stereocontrolled polyprenylation of quinones. A
new synthetic method of vitamin K series. Chem. Lett.:991–884 (1979).
Vitamin K 153
57. S. Patai, The Chemistry of the Quinoid Compounds, Parts 1 and 2. John Wiley and Sons,
New York, 1974.
58. P. Sommer and M. Kofler, Physiochemical properties and methods of analysis of phylloquinones,
menaquinones, ubiquinones, phostoquinones, menadione, and related compounds.
Vitamins and Hormones 24:349–399 (1966).
59. P. J. Dunphy and A. F. Brodie, The structure and function of quinones in respiratory metabolism,
in Methods in Enzymology, Vol. 18C (D. B. McCormick and L. D Wright, eds.). Academic
Press, New York, 1971, pp. 407–461.
60. H. Mayer and O. Isler, Isolation of vitamin K, in Methods in Enzymology, Vol. 18C (D. B.
McCormick and L. D. Wright, eds.). Academic Press, New York, 1971, pp. 469–491.
61. J. T. Matschiner and J. M. Amelotti, Characterization of vitamin K from bovine liver J. Liver
Res. 9:176–179 (1968).
62. R. C. Williams, J. A. Schmidt, and R. A. Henry, Quantitative analysis of fat-soluble vitamins
by high-speed liquid chromatography. J. Chromatogr. Sci. 10:494–501 (1972).
63. M. J. Shearer, High-performance liquid chromatography of K vitamins and their antagonists,
in Advances in Chromatography, Vol. 21 (J. C. Giddings, E. Grushka, J. Cazes, and P. R.
Brown, eds.). Marcel Dekker, New York, 1983, pp. 243–301.
64. H. Dam and E. Sondergaard, The determination of vitamin K, in The Vitamins, 2nd ed., Vol.
6 (P. Gyorgy and W. N. Pearson, eds.). Academic Press, New York, 1967, pp. 245–260.
65. P. Griminger and O. Donis, Potency of vitamin K1 and two analogs in counteracting the
effects of dicumarol and sulfaquinoxaline in the chick. J. Nutr. 70:361–368 (1960).
66. J. T. Matschiner and E. A. Doisy, Jr., Bioassay of vitamin K in chicks. J. Nutr. 90:97–100
67. P. Griminger, Nutritional requirements for vitamin K-animal studies, in Symposium Proceedings
on the Biochemistry, Assay, and Nutritional Value of Vitamin K and Related Compounds.
Assn. Vitamin Chemists, Chicago, 1971, pp. 39–59.
68. H. Weiser and A. W. Kormann, Biopotency of vitamin K. Int. J. Vitamin Nutr. Res. 53:143–
155 (1983).
69. H. Dam and F. Schonheyder, The occurrence and chemical nature of vitamin K. Biochem.
J. 30:897–901 (1936).
70. H. Dam and J. Glavind. Vitamin K in the plant. Biochem. J. 32:485–487 (1938).
71. R. E. Olson, Vitamin K, in Modern Nutrition in Health and Disease (M. E. Shils and V. R.
Young, eds.). Lea & Febiger, Philadelphia, 1988, pp. 328–339.
72. S. W. Souci, W. Fachmann, and H. Kraut, Food Composition and Nutrition Tables, 1986/
1987, 3rd ed., Wisenschaftliche Verlagsgesellschaft, Stuttgart, 1986.
73. J. A. T. Pennington, Bowes and Church’s Food Values of Portions Commonly Used. J. B.
Lippincott, Philadelphia, 1989.
74. L. R. Richardson, S. Wilkes, and S. J. Ritchey, Comparative vitamin K activity of frozen,
irradiated and heat-processed foods. J. Nutr. 73:369–373 (1961).
75. E. A. Doisy, Vitamin K in human nutrition, in Symposium Proceedings on the Biochemistry,
Assay and Nutritional Value of Vitamin K and Related Compounds. Assn. Vitamin Chemists,
Chicago, 1971, pp. 79–92.
76. M. J. Shearer, V. Allan, Y. Haroon, and P. Barkhan, Nutritional aspects of vitamin K in the
human, in Vitamin K Metabolism and Vitamin K–Dependent Proteins (J. W. Suttie, ed.).
University Park Press, Baltimore, 1980, pp. 317–327.
77. K. Kodaka, T. Ujiie, T. Ueno, and M. Saito, Contents of vitamin K1 and chlorophyll in green
vegetables. J. Jap. Soc. Nutr. Food Sci. 39:124–126 (1986).
78. R. M. Seifert, Analysis of vitamin K1 in some green leafy vegetables by gas chromatography.
J. Agric. Food Chem. 27:1301–1304 (1979).
79. G. Ferland and J. A. Sadowski, Vitamin K1 (phylloquinone) content of edible oils: effects
of heating and light exposure. J. Agric. Food Chem. 40:1869–1873 (1992).
80. Y. Haroon, M. J. Shearer, S. Rahim, W. G. Gunn, G. McEnery, and P. Barkhan, The content
154 Suttie
of phylloquinone (vitamin K1) in human milk, cows’ milk and infant formula foods determined
by high-performance liquid chromatography. J. Nutr. 112:1105–1117 (1982).
81. R. v. Kries, M. Shearer, P. T. McCarthy, M. Haug, G. Harzer, and U. Gobel, Vitamin K1
content of maternal milk: influence of the stage of lactation, lipid composition, and vitamin
K1 supplements given to the mother. Pediatr. Res. 22:513–517 (1987).
82. L. M. Canfield, J. M. Hopkinson, A. F. Lima, G. S. Martin, K. Sugimoto, J. Burri, L. Clark,
and D. L. McGee, Quantitation of vitamin K in human milk. Lipids 25:06–411 (1990).
83. S. L. Booth, J. A. Sadowski, J. L. Weihrauch, and G. Ferland, Vitamin K1 (phylloquinone)
content of foods: a provisional table. J. Food Comp. Anal. 6:109–120 (1993).
83a. S. L. Booth, J. A. Sadowski, and J. A. T. Pennington, Phylloquinone (vitamin K1) content
of foods in the U.S. Food and Drug Administration’s total diet study. J. Agric. Food Chem.
43:1574–1579 (1995).
83b. S. L. Booth, R. Webb, and J. C. Peters, Assessment of phylloquinone and dihydrophylloquinone
dietary intakes among a nationally representative sample of U.S. consumers using 14-
day food diaries. J. Am. Diet. Assoc. 99:1072–1076 (1999).
84. G. Ferland, D. L. MacDonald, and J. A. Sadowski, Development of a diet low in vitamin
K-1 (phylloquinone). J. Am. Dietet. Assoc. 92:593–597 (1992).
85. G. Ferland and J. A. Sadowski, Vitamin K1 (phylloquinone) content of green vegetables:
effects of plant maturtion and geographical growth location. J. Agric. Food Chem. 40:1874–
1877 (1992).
86. D. Hollander, Intestinal absorption of vitamins A, E, D, and K. J. Lab. Clin. Med. 97:449–
462 (1981).
87. R. Blomstrand and L. Forsgren, Vitamin K1-3H in man. Its intestinal absorption and transport
in the thoracic duct lymph. Int. Z. Vitamin Forschung 38:46–64 (1968).
88. M. J. Shearer, A. McBurney, and P. Barkhan, Studies on the absorption and metabolism of
phylloquinone (vitamin K1) in man. Vitamins and Hormones 32:513–542 (1974).
89. K-S. G. Jie, B. L. M. G. Gijsbers, and C. Vermeer, in The Role of Vitamin K-Dependent
Proteins in Tissue Calcification, Ph.D. thesis, University of Limburg, Maastricht, The Netherlands,
1995, pp. 77–83.
89a. A. K. Garber, N. C. Binkley, D. C. Krueger, and J. W. Suttie, Comparison of phylloquinone
bioavailability from food sources or a supplement in human subjects. J. Nutr. 129:1201–
1203 (1999).
90. J. J. Lipsky, Subject review: nutritional sources of vitamin K. Mayo Clin. Proc. 69:462–466
91. M. J. Shearer, Vitamin K metabolism and nutriture. Blood Rev. 6:92–104 (1992).
92. D. Hollander, E. Rim, and P. E. Ruble, Jr., Vitamin K2 colonic and ileal in vivo absorption:
Bile, fatty acids, and pH effects on transport. Am. J. Physiol. 233:E124–E129 (1977).
93. T. Ichihashi, Y. Takagishi, K. Uchida, and H. Yamada, Colonic absorption of menaquinone-
4 and menaquinone-9 in rats. J. Nutr. 122:506–512 (1992).
94. J. M. Conly and K. E. Stein, The absorption and bioactivity of bacterially synthesized menaquinones.
Clin. Invest. Med. 16:45–57 (1993).
95. J. M. Conly and K. Stein, Quantitative and qualitative measurements of K vitamins in human
intestinal contents. Am. J. Gastroenterol. 87:311–316 (1992).
96. D. Hollander and T. C. Truscott, Colonic absorption of vitamin K-3. J. Lab. Clin. Med. 83:
648–656 (1974).
97. D. Hollander and T. C. Truscott, Mechanism and site of vitamin K-3 small intestinal transport.
Am. J. Physiol. 226:1526–1522 (1974).
98. J. Saupe, M. J. Shearer, and M. Kohlmeier, Phylloquinone transport and its influence on ?-
carboxyglutamate residues of osteocalcin in patients on maintenance hemodialysis. Am. J.
Clin. Nutr. 58:204–208 (1993).
99. T. D. Bjornsson, P. J. Meffin, S. E. Swezey, and T. F. Blaschke, Disposition and turnover
Vitamin K 155
of vitamin K1 in man, in Vitamin K Metabolism and Vitamin K–Dependent Proteins (J. W.
Suttie, ed.). University Park Press, Baltimore, 1980, pp. 328–332.
100. M. J. Shearer, C. N. Mallinson, G. R. Webster, and P. Barkhan, Clearance from plasma and
excretion in urine, faeces and bile of an intravenous dose of tritiated vitamin K1 in man. Br.
J. Haematol. 22:579–588 (1972).
101. J. P. Hart, The exploitation of the electrochemical properties of K vitamins for their sensitive
measurement in tissues, in Vitamin K and Vitamin K–Dependent Proteins: Analytical, Physiological,
and Clinical Aspects (M. J. Shearer and M. J. Seghatchian, eds.). CRC Press, Boca
Raton, 1993, pp. 27–54.
102. Y. Haroon, Reaction detection methods for K vitamins and their 2?3?-epoxy metabolite in
liquid chromatography, in Vitamin K and Vitamin K–Dependent Proteins: Analytical, Physiological,
and Clinical Aspects (M. J. Shearer and M. J. Seghatchian, eds.). CRC Press, Boca
Raton, 1993, pp. 55–90.
103. J. A. Sadowski, S. J. Hood, G. E. Dallal, and P. J. Garry, Phylloquinone in plasma from
elderly and young adults: factors influencing its concentration. Am. J. Clin. Nutr. 50:100–
108 (1989).
104. M. J. Shearer, P. T. McCarthy, O. E. Crampton, and M. B. Mattock, The assessment of
human vitamin K status for tissue measurements, in Current Advances in Vitamin K Research
(J. W. Suttie, ed.). Elsevier, New York, 1988, pp. 437–452.
105. S. J. Hodges, M. J. Pilkington, M. J. Shearer, L. Bitensky, and J. Chayen, Age-related changes
in the circulating levels of congeners of vitamin K2, menaquinone-7 and menaquinone-8.
Clin. Sci. 78:63–66 (1990).
106. S. J. Hodges, M. J. Pilkington, T. C. B. Stamp, A. Catterall, M. J. Shearer, L. Bitensky, and
J. Chayen, Depressed levels of circulating menaquinones in patients with osteoporotic fractures
of the spine and femoral neck. Bone 12:387–389 (1991).
107. S. J. Hodges, K. Akesson, P. Vergnaud, K. Obrant, and P. D. Delmas, Circulating levels of
vitamin K1 and K2 decreased in elderly women with hip fracture. J. Bone Miner. Res. 8:
1241–1245 (1993).
108. J. W. Suttie, The importance of menaquinones in human nutrition. Annu. Rev. Nutr. 15:399–
417 (1995).
109. M. Shino, Determination of endogenous vitamin K (phylloquinone and menaquinone-n) in
plasma by high-performance liquid chromatography using platinum oxide catalyst reduction
and fluorescence detection. Analyst 113:393–397 (1988).
110. K. Hirauchi, T. Sakano, T. Nagaoka, A. Morimoto, and S. Masuda, Measurement of K vitamins
in biological materials by high-performance liquid chromatography with fluorometric
detection, in Proceedings of the 7th Symposium on Analytical Chemistry of Biological Substances,
Pharmaceutical Soc. Japan, 1985, pp. 43–46.
111. C. G. Kindberg, and J. W. Suttie, Effect of various intakes of phylloquinone on signs of
vitamin K deficiency and serum and liver phylloquinone concentrations in the rat. J. Nutr.
119:175–180 (1989).
112. B. H. Will and J. W. Suttie, Comparative metabolism of phylloquinone and menaquinone-
9 in rat liver. J. Nutr. 122:953–958 (1992).
113. P. M. Allison, L. L. Mummah-Schendel, C. G. Kindberg, C. S. Harms, N. U. Bang, and
J. W. Suttie, Effects of a vitamin K-deficient diet and antibiotics in normal human volunteers.
J. Lab. Clin. Med. 110:180–188 (1987).
114. J. W. Suttie, L. L. Mummah-Schendel, D. V. Shah, B. J. Lyle, and J. L. Greger, Vitamin
K deficiency from dietary vitamin K restriction in humans. Am. J. Clin. Nutr. 47:475–480
115. G. Ferland, J. A. Sadowski, and M. E. O’Brien, Dietary induced subclinical vitamin K defi-
ciency in normal human subjects. J. Clin. Invest. 91:1761–1768 (1993).
116. Y. Usui, H. Tanimura, N. Nishimura, N. Kobayashi, T. Okanoue, and K. Ozawa, Vitamin
156 Suttie
K concentrations in the plasma and liver of surgical patients. Am. J. Clin. Nutr. 51:846–852
117. M. J. Thierry and J. W. Suttie, Distribution and metabolism of menadiol diphosphate in the
rat. J. Nutr. 97:512–516 (1969).
118. M. J. Thierry and J. W. Suttie, Effect of warfarin and the chloro analog of vitamin K on
phylloquinone metabolism. Arch. Biochem. Biophys. 147:430–435 (1971).
119. M. J. Thierry, M. A. Hermodson, and J. W. Suttie, Vitamin K and warfarin distribution and
metabolism in the warfarin-resistant rat. Am. J. Physiol. 219:854–859 (1970).
120. T. Konishi, S. Baba, and H. Sone, Whole-body autoradiographic study of vitamin K distribution
in rat. Chem. Pharm. Bull. 21:220–224 (1973).
121. J. P. Green, E. Sondergaard, and H. Dam, Intracellular distribution of vitamin K in beef
liver. Biochim. Biophys. Acta 19:182–183 (1956).
122. R. G. Bell and J. T. Matschiner, Intracellular distribution of vitamin K in the rat. Biochim.
Acta 184:597–603 (1969).
123. S. E. Nyquist, J. T. Matschiner, and D. J. James Morre, Distribution of vitamin K among
rat liver cell fractions. Biochim. Biophys. Acta 244:645–649 (1971).
124. C. K. Reedstrom and J. W. Suttie, Comparative distribution, metabolism, and utilization
of phylloquinone and menaquinone-9 in rat liver. Proc. Soc. Exp. Biol. Med. 209:403–409
125. C. E. Kight, C. K. Reedstrom, and J. W. Suttie, Identification, isolation, and partial purification
of a cytosolic binding protein for vitamin K from rat liver. FASEB J 9:A725 (1995).
126. T. E. Knauer, C. M. Siegfried, and J. T. Matschiner, Vitamin K requirement and the concentration
of vitamin K in rat liver. J. Nutr. 106:1747–1756 (1976).
127. J. T. Matschiner, Occurrence and biopotency of various forms of vitamin K, in The Fat-
Soluble Vitamins (H. F. DeLuca and J. W. Suttie, eds.). University of Wisconsin Press, Madison,
1970, pp. 377–397.
128. J. T. Matschiner, Isolation and identification of vitamin K from animal tissue, in Symposium
Proceedings on the Biochemistry, Assay, and Nutritional Value of Vitamin K and Related
Compounds. Assn. Vitam. Chem., Chicago, 1971, pp. 21–37.
129. K. Uchida and T. Komeno, Relationships between dietary and intestinal vitamin K, clotting
factor levels, plasma vitamin K and urinary Gla, in Current Advances in Vitamin K Research
(J. W. Suttie, ed.). Elsevier, New York, 1988, pp. 477–492.
130. Y. Usui, N. Nishimura, N. Kobayashi, T. Okanoue, M. Kimoto, and K. Ozawa, Measurement
of vitamin K in human liver by gradient elution high-performance liquid chromatography
using platinum-black catalyst reduction and fluorimetric detection. J. Chromatogr. 489:291–
301 (1989).
131. S. Kayata, C. Kindberg, F. R. Greer, and J. W. Suttie, Vitamin K1 and K2 in infant human
liver. J. Pediatr. Gastroenterol. Nutr. 8:304–307 (1989).
132. M. J. Shearer, P. T. McCarthy, O. E. Crampton, and M. B. Mattock, The assessment of
human vitamin K status from tissue measurements, in Current Advances in Vitamin K Research
(J. W. Suttie, ed.). Elsevier, New York, 1988, pp. 437–452.
133. B. H. Will, Y. Usui, and J. W. Suttie, Comparative metabolism and requirement of vitamin
K in chicks and rats. J. Nutr. 122:2354–2360 (1992).
134. M. Guillaumont, H. Weiser, L. Sann, B. Vignal, M. Ledercq, and A. Frederich, Hepatic
concentration of vitamin K active compounds after application of phylloquinone to chickens
on a vitamin K deficient or adequate diet. Int. J. Vitam. Nutr. Res. 62:15–20 (1992).
135. H. H. W. Thijssen and M. J. Drittij-Reijnders, Vitamin K distribution in rat tissues: dietary
phylloquinone is a source of tissue menaquinone-4. Br. J. Nutr. 72:415–425 (1994).
135a. R. T. Davidson, A. L. Foley, J. A. Engelke, and J. W. Suttie, Conversion of dietary phylloquinone
to tissue menaquinone-4 in rats is not dependent on gut bacteria. J. Nutr. 128:220–223
135b. J. E. Ronden, M. J. Drittij-Reijnders, C. Vermeer, and H. H. W. Thijssen, Intestinal flora is
Vitamin K 157
not an intermediate in the phylloquinone-menaquinone-4 conversion in the rat. Biochem.
Biophys. Acta 1379:69–75 (1998).
136. R. Bentley and R. Meganathan, Biosynthesis of vitamin K (menaquinone) in bacteria. Microbiol.
Rev. 46:241–280 (1982).
137. C. Martius, Chemistry and function of vitamin K, in Blood Clotting Enzymology (W. H.
Seegers, ed.). Academic Press, New York, 1967, pp. 551–575.
138. C. Martius, The metabolic relationships between the different K vitamins and the synthesis
of the ubiquinones. Am. J. Clin. Nutr. 9:97–103 (1961).
139. G. H. Dialameh, K. G. Yekundi, and R. E. Olson, Enzymatic alkylation of menaquinone-
O to menaquinones by microsomes from chick liver. Biochim. Biophys. Acta 223:332–338
140. C. Martius, Recent investigations on the chemistry and function of vitamin K, in Ciba Fdn.
Symposium on Quinones in Electron Transport (G. E. W. Wolstenholme and C. M. O’Connor,
eds.). Little, Brown, Boston, 1961, pp. 312–326.
141. M. Billeter, W. Bollinger, and C. Martius, Untersuchungen u?ber die Umwandlung von verfutterten
K-Vitaminen durch Austausch der Seitenkette und die Rolle der Darmbakterien
hierbei. Biochem. Z. 340:290–303 (1964).
142. F. C. G. Hoskin, J. W. T. Spinks, and L. B. Jaques, Urinary excretion products of menadione
(vitamin K3). Can. J. Biochem. Physiol. 32:240–250 (1954).
143. K. T. Hart, Study of hydrolysis of urinary metabolites of 2-methyl-1,4-naphthoquinone. Proc.
Soc. Exp. Biol. Med. 97:848–851 (1958).
144. R. Losito, C. A. Owen, Jr., and E. V. Flock, Metabolism of [C14] menadione. Biochemistry
6:62–68 (1967).
145. R. Losito, C. A. Owen, Jr., and E. V. Flock, Metabolic studies of vitamin K1-14C and menadione-
14C in the normal and hepatectomized rats. Thromb. Diath. Haemorrhag. 19:383–388
146. J. D. Taylor, G. J. Miller, L. B. Jaques, and J. W. T. Spinks, The distribution of administered
vitamin K1-14C in rats. Can. J. Biochem. Physiol. 34:1143–1152 (1956).
147. O. Wiss and H. Gloor, Absorption, distribution, storage and metabolites of vitamin K and
related quinones. Vitam Hormones 24:575–586 (1966).
148. J. T. Matschiner, R. G. Bell, J. M. Amelotti, and T. E. Knauer, Isolation and characterization
of a new metabolite of phylloquinone in the rat. Biochim. Biophys. Acta 201:309–315
149. R. G. Bell, J. A. Sadowski, and J. T. Matschiner, Mechanism of action of warfarin. Warfarin
and metabolism of vitamin K1. Biochemistry 11:1959–1961 (1972).
150. M. J. Shearer, A. McBurney, and P. Barkhan, Effect of warfarin anticoagulation on vitamin-
K1 metabolism in man. Br. J. Haematol. 24:471–479 (1973).
151. M. J. Shearer, A. McBurney, A. M. Breckenridge, and P. Barkhan, Effect of warfarin on
the metabolism of phylloquinone (vitamin K1): dose–response relationships in man. Clin.
Sci. Mol. Med. 52:621–630 (1977).
152. E. W. Davie, K. Fujikawa, and W. Kisiel, The coagulation cascade: initiation, maintenance,
and regulation. Biochemistry 30:10363–10370 (1991).
153. L. H. Clouse and P. C. Comp, The regulation of hemostasis: the protein C system. N. Engl.
J. Med. 314:1298–1304 (1986).
154. H. C. Hemker, J. J. Veltkamp, A. Hensen, and E. A. Loeliger, Nature of prothrombin biosynthesis:
preprothrombinaemia in vitamin K-deficiency. Nature 200:589–590 (1963).
155. D. V. Shah and J. W. Suttie, Mechanism of action of vitamin K: evidence for the conversion
of a precursor protein to prothrombin in the rat. Proc. Natl. Acad. Sci. USA 68:1653–1657
156. P. O. Ganrot and J. E. Nilehn, Plasma prothrombin during treatment with dicumarol. II.
Demonstration of an abnormal prothrombin fraction. Scand. J. Clin Lab. Invest. 22:23–28
158 Suttie
157. J. Stenflo, Dicumarol-induced prothrombin in bovine plasma. Acta Chem. Scand. 24:3762–
3763 (1970).
158. T. L. Carlisle, D. V. Shah, R. Schlegel, and J. W. Suttie, Plasma abnormal prothrombin and
microsomal prothrombin precursor in various species. Proc. Soc. Exp. Biol. Med. 148:140–
144 (1975).
159. J. Stenflo and J. W. Suttie, Vitamin K-dependent formation of ?-carboxyglutamic acid. Annu.
Rev. Biochem. 46:157–172 (1977).
160. C. T. Esmon, J. W. Suttie, and C. M. Jackson, The functional significance of vitamin K
action. Difference in phospholipid binding between normal and abnormal prothrombin. J.
Biol. Chem. 250:4095–4099 (1975).
161. J. Stenflo, P. Fernlund, W. Egan, and P. Roepstorff, Vitamin K dependent modifications of
glutamic acid residues in prothrombin. Proc. Natl. Acad. Sci. USA 71:2730–2733 (1974).
162. G. L. Nelsestuen, T. H. Zytkovicz, and J. B. Howard, The mode of action of vitamin K.
Identification of ?-carboxyglutamic acid as a component of prothrombin. J. Biol. Chem. 249:
6347–6350 (1974).
163. S. Magnusson, L. Sottrup-Jensen, T. E. Petersen, H. R. Morris, and A. Dell, Primary structure
of the vitamin K–dependent part of prothrombin. FEBS Lett. 44:189–193 (1974).
164. C. T. Esmon, J. A. Sadowski, and J. W. Suttie, A new carboxylation reaction. The vitamin Kdependent
incorporation of H14CO3
 into prothrombin. J. Biol. Chem. 250:4744–4748 (1975).
165. J. W. Suttie, J. M. Hageman, S. R. Lehrman, and D. H. Rich, Vitamin K–dependent carboxylase:
development of a peptide substrate. J. Biol. Chem. 251:5827–5830 (1976).
166. T. L. Carlisle and J. W. Suttie, Vitamin K dependent carboxylase: subcellular location of
the carboxylase and enzymes involved in vitamin K metabolism in rat liver. Biochemistry
19:1161–1167 (1980).
167. B. C. Johnson, Post-translational carboxylation of preprothrombin. Mol. Cell. Biochem. 38:
77–121 (1981).
168. R. E. Olson, The function and metabolism of vitamin K. Annu. Rev. Nutr. 4:281–337 (1984).
169. J. W. Suttie, Vitamin K–dependent carboxylase. Annu. Rev. Biochem. 54:459–477 (1985).
170. J. W. Suttie. Vitamin K–dependent carboxylation of glutamyl residues in proteins. BioFactors
1:55–60 (1988).
171. J. W. Suttie, Synthesis of vitamin K–dependent proteins. FASEB J. 7:445–452 (1993).
172. C. Vermeer and M. A. G. de Boer-van den Berg, Vitamin K–dependent carboxylase. Haematologia
18:71–97 (1985).
173. C. Vermeer. ?-Carboxyglutamate-containing proteins and the vitamin K–dependent carboxylase.
Biochem. J. 266:625–636 (1990).
174. A. Y. Cheung, G. M. Wood, S. Funakawa, C. P. Grossman, and J. W. Suttie, Vitamin K–
dependent carboxylase: substrates, products, and inhibitors, in Current Advances in Vitamin
K Research (J. W. Suttie, ed.). Elsevier, New York, 1988, pp. 3–16.
175. P. A. Price, J. D. Fraser, and G. Metz-Virca, Molecular cloning of matrix Gla protein: implications
for substrate recognition by the vitamin K–dependent ?-carboxylase. Proc. Natl.
Acad. Sci. USA 84:8335–8339 (1987).
176. D. J. Liska and J. W. Suttie, Location of ?-carboxyglutamyl residues in partially carboxylated
prothrombin preparations. Biochemistry 27:8636–8641 (1988).
177. M. E. Benton, P. A. Price, and J. W. Suttie, Multi-site-specificity of the vitamin K–dependent
carboxylase: in vitro carboxylation of des-?-carboxylated bone Gla protein and des-?-carboxylated
pro bone Gla protein. Biochemistry 34:9541–9551 (1995).
178. R. Azerad, P. Decottignies-Le Marechal, C. Ducrocq, A. Righini-Tapie, A. Vidal-Cros, S.
Bory, J. Dubois, M. Gaudry, and A. Marquet, The vitamin K–dependent carboxylation of
peptide substrates: stereochemical features and mechanistic studies with substrate analogues,
in Current Advances in Vitamin K Research (J. W. Suttie, ed.). Elsevier, New York, 1988,
pp. 17–23.
179. J. Dubois, C. Dugave, C. Foures, M. Kaminsky, J-C. Tabet, S. Bory, M. Gaudry, and A.
Vitamin K 159
Marquet, Vitamin K dependent carboxylation: determination of the stereochemical course
using 4-fluoroglutamyl-containing substrate. Biochemistry 30:10506–10512 (1991).
180. P. A. Friedman, M. A. Shia, P. M. Gallop, and A. E. Griep, Vitamin K–dependent ?-carbon–
hydrogen bond cleavage and the non-mandatory concurrent carboxylation of peptide bound
glutamic acid residues. Proc. Natl. Acad. Sci. USA 76:3126–3129 (1979).
181. A. E. Larson, P. A. Friedman, and J. W. Suttie, Vitamin K–dependent carboxylase: stoichiometry
of carboxylation and vitamin K 2,3-epoxide formation. J. Biol. Chem. 256:11032–
11035 (1981).
182. J. J. McTigue and J. W. Suttie, Vitamin K-dependent carboxylase: demonstration of a vitamin
K- and O2-dependent exchange of 3H from 3H2O into glutamic acid residues. J. Biol. Chem.
258:12129–12131 (1983).
183. D. L. Anton and P. A. Friedman, Fate of the activated ?-carbon–hydrogen bond in the uncoupled
vitamin K–dependent ?-glutamyl carboxylation reaction. J. Biol. Chem. 258:14084–
14087 (1983).
184. G. M. Wood and J. W. Suttie, Vitamin K-dependent carboxylase. Stoichiometry of vitamin
K epoxide formation, ?-carboxyglutamyl formation, and ?-glutamyl-3H cleavage. J. Biol.
Chem. 263:3234–3239 (1988).
185. P. Dowd, S. W. Ham, and S. J. Geib, Mechanism of action of vitamin K. J. Am. Chem. Soc.
113:7734–7743 (1991).
186. P. Dowd, S-W. Ham, and R. Hershline, Role of oxygen in the vitamin K–dependent carboxylation
reaction: incorporation of a second atom of 18O from molecular oxygen-18O2 into vitamin
K oxide during carboxylase activity. J. Am. Chem. Soc. 114:7613–7617 (1992).
187. A. Kuliopulos, B. R. Hubbard, Z. Lam, I. J. Koski, B. Furie, B. C. Furie, and C. T. Walsh,
Dioxygen transfer during vitamin K–dependent carboxylase catalysis. Biochemistry 31:
7722–7728 (1992).
188. P. Dowd, S. W. Ham, S. Naganathan, and R. Hershline, The mechanism of action of vitamin
K. Annu. Rev. Nutr. 15:419–440 (1995).
189. A. Ichinose and E. W. Davie, The blood coagulation factors: their cDNAs, genes, and expression,
in Hemostasis and Thrombosis: Basic Principles and Clinical Practice, 3rd ed. (R. W.
Colman, J. Hirsh, V. J. Marder, and E. W. Salzman, eds.). J. B. Lippincott, Philadelphia,
1994, pp. 19–54.
190. B. Furie and B. C. Furie, Molecular and cellular biology of blood coagulation. N. Engl. J.
Med. 326:800–806 (1992).
191. J. E. Knobloch and J. W. Suttie, Vitamin K–dependent carboxylase. Control of enzyme
activity by the ‘‘propeptide’’ region of factor X. J. Biol. Chem. 262:15334–15337 (1987).
192. A. Cheung, J. W. Suttie, and M. Bernatowicz, Vitamin K–dependent carboxylase: structural
requirements for peptide activation. Biochim. Biophys. Acta 1039:90–93 (1990).
193. M. C. Harbeck, A. Y. Cheung, and J. W. Suttie, Vitamin K–dependent carboxylase: partial
purification of the enzyme by antibody affinity techniques. Thromb. Res. 56:317–323 (1989).
194. B. R. Hubbard, M. M. W. Ulrich, M. Jacobs, C. Vermeer, C. Walsh, B. Furie, and B. C.
Furie, Vitamin K–dependent carboxylase: affinity purification from bovine liver by using a
synthetic propeptide containing the ?-carboxylation recognition site. Proc. Natl. Acad. Sci.
USA 86:6893–6897 (1989).
195. A. Kuliopulos, C. E. Cieurzo, B. Furie, B. C. Furie, and C. T. Walsh, N-Bromoacetyl peptide
substrate affinity labeling of vitamin K–dependent carboxylase. Biochemistry 31:9436–9444
196. S-M. Wu, D. P. Morris, and D. W. Stafford, Identification and purification to near homogeneity
of the vitamin K–dependent carboxylase. Proc. Natl. Acad. Sci. USA 88:2236–2240
197. S-M. Wu, W-F. Cheung, D. Frazier, and D. Stafford, Cloning and expression of the cDNA
for human ?-glutamyl carboxylase. Science 254:1634–1636 (1991).
198. K. L. Berkner, M. Harbeck, S. Lingenfelter, C. Bailey, C. M. Sanders-Hinck, and J. W.
160 Suttie
Suttie, Purification and identification of bovine liver ?-carboxylase. Proc. Natl. Acad. Sci.
USA 89:6242–6246 (1992).
199. D. A. Roth, A. Rehemtulla, R. J. Kaufman, C. T. Walsh, B. Furie, and B. C. Furie, Expression
of bovine vitamin K–dependent carboxylase activity in baculovirus infected cells. Proc. Natl.
Acad. Sci. USA 90:8372–8376 (1993).
199a. B. Furie, B. A. Bouchard, and B. C. Furie, Vitamin K–dependent biosynthesis of ?-carboxyglutamic
acid. Blood 93:1798–1808 (1999).
199b. T. B. Stanley, S. M. Wu, R. J. Houben, V. P. Mutucumarana, and D. W. Stafford, Role of
the propeptide and ?-glutamic acid domain of factor IX for in vitro carboxylation by the
vitamin K–dependent carboxylase. Biochemistry 37:13262–13268 (1998).
199c. W. Wu, J. D. Bancroft, and J. W. Suttie, Structural features of the kringle domain determine
the intracellular degradation of under-?-carboxylated prothrombin: studies of chimeric rat/
human prothrombin. Proc. Natl. Acad. Sci. USA 94:13654–13660 (1997).
199d. M. E. Benton, P. A. Price, and J. W. Suttie, Multi-site-specificity of the vitamin K–dependent
carboxylase: in vitro carboxylation of des-?-carboxylated bone Gla protein and Des-?-carboxylated
pro bone Gla protein. Biochemistry 34:9541–9551 (1995).
199e. D. P. Morris, R. D. Stevens, D. J. Wright, and D. W. Stafford, Processive post-translational
modification. Vitamin K–dependent carboxylation of a peptide substrate. J. Biol. Chem. 270:
30491–30498 (1995).
199f. K. L. Berkner and B. N. Pudota, Vitamin K–dependent carboxylation of the carboxylase.
Proc. Natl. Acad. Sci. USA 95:466–472 (1998).
200. P. V. Hauschka, J. B. Lian, and P. M. Gallop, Direct identification of the calcium-binding
amino acid ?-carboxyglutamate, in mineralized tissue. Proc. Natl. Acad. Sci. USA 72:3925–
3929 (1975).
201. P. A. Price, A. S. Otsuka, J. W. Poser, J. Kristaponis, and N. Raman, Characterization of a
?-carboxyglutamic acid-containing protein from bone. Proc. Natl. Acad. Sci. USA 73:1447–
1451 (1976).
202. P. A. Price and M. K. Williamson, Primary structure of bovine matrix Gla protein, a new
vitamin K–dependent bone protein. J. Biol. Chem. 260:14971–14975 (1985).
203. J. D. Fraser and P. A. Price, Lung, heart, and kidney express high levels of mRNA for the
vitamin K–dependent matrix Gla protein. J. Biol. Chem. 263:11033–11036 (1988).
204. C. Maillard, M. Berruyer, C. M. Serre, M. Dechavanne, and P. D. Delmas, Protein S, a
vitamin K–dependent protein, is a bone matrix component synthesized and secreted by osteoblasts.
Endocrinology 130:1599–1604 (1992).
205. J. B. Lian, P. V. Hauschka, and P. M. Gallop, Properties and biosynthesis of a vitamin K–
dependent calcium binding protein in bone. Fed. Proc. 37:2615–2620 (1978).
206. P. A. Price, Vitamin K–dependent formation of bone Gla protein (osteocalcin) and its function.
Vitam Hormones 42:65–108 (1985).
207. P. A. Price, Role of vitamin K–dependent proteins in bone metabolism. Annu. Rev. Nutr.
8:565–583 (1988).
208. C. Vermeer, K-S. G. Jie, and M. H. J. Knapen, Role of vitamin K in bone metabolism. Annu.
Rev. Nutr. 15:1–22 (1995).
209. P. A. Price and M. K. Williamson, Effects of warfarin on bone. Studies on the vitamin K–
dependent protein of rat bone. J. Biol. Chem. 256:12754–12759 (1981).
210. P. A. Price, M. K. Williamson, T. Haba, R. B. Dell, and W. S. S. Jee, Excessive mineralization
with growth plate closure in rats on chronic warfarin treatment. Proc. Natl. Acad. Sci. USA
79:7734–7738 (1982).
211. P. A. Price, J. G. Parthemore, and L. J. Deftos, New biochemical marker for bone metabolism.
Measurement by radioimmunoassay of bone Gla protein in the plasma of normal subjects
and patients with bone disease. J. Clin. Invest. 66:878–883 (1980).
212. P. D. Delmas, L. Malaval, M. E. Arlot, and P. J. Meunier, Serum bone Gla-protein compared
to bone histomorphometry in endocrine diseases. Bone 6:339–341 (1985).
Vitamin K 161
213. P. A. Price, Vitamin K nutrition and postmenopausal osteoporosis (editorial). J. Clin. Invest.
91:1268 (1993).
214. N. C. Binkley and J. W. Suttie, Vitamin K nutrition and osteoporosis J. Nutr. 125:1812–
1821 (1995).
214a. P. Vergnaud, P. Garnero, P. J. Meunier, G. Breart, K. Kamihagi, and P. D. Delmas, Undercarboxylated
osteocalcin measured with a specific immunoassay predicts hip fractures in elderly
women: the EPIDOS study. J. Clin. Endocrinol. Metab. 82:719–724 (1997).
214b. D. Feskanich, P. Weber, W. C. Willett, H. Rockett, S. L. Booth, and G. A. Colditz, Vitamin
K intake and hip fractures in women: a prospective study. Am. J. Clin. Nutr. 69:74–79 (1999).
214c. P. J. Caraballo, S. E. Gabriel, M. R. Castro, E. J. Atkinson, and L. J. Melton, Changes in
bone density after exposure to oral anticoagulants: a meta-analysis. Osteoporosis Int. 9:441–
448 (1999).
214d. P. Ducy, C. Desbois, B. Boyce, G. Pinero, B. Story, C. Dunstan, E. Smith, J. Bonadio, S.
Goldstein, C. Gundberg, A. Bradley, and G. Karsenty, Increased bone formation in osteocalcin-
deficient mice. Nature 382:448–452 (1996).
214e. G. Luo, P. Ducy, M. D. McKee, G. J. Pinero, E. Loyer, R. R. Behringer, G. Karsenty, Spontaneous
calcification of arteries and cartilage in mice lacking matrix GLA protein. Nature 386:
78–81 (1997).
215. J. B. Lian, R. J. Levy, J. T. Levy, and P. A. Friedman, Other vitamin K dependent proteins,
in Calcium-Binding Proteins: Structure and Function (F. L. Siegel, E. Carafoli, R. H. Kretsinger,
D. H. MacLennan, and R. H. Wasserman, eds.). Elsevier, New York, 1980, pp. 449–
216. J. A. Helpern, S. J. McGee, and J. M. Riddle, Observations suggesting a possible link between
gamma-carboxyglutamic acid and porcine bioprosthetic valve calcification. Henry Ford
Hosp. Med. J. 30:152–155 (1982).
217. L. J. M. van Haarlem, B. A. M. Soute, H. C. Hemker, and C. Vermeer, Characterization of
Gla-containing proteins from calcified human atherosclerotic plaques, in Current Advances
in Vitamin K Research (J. W. Suttie, ed.). Elsevier, New York, 1988, pp. 287–292.
218. A. E. Griep and P. A. Friedman, Purification of a protein containing ?-carboxyglutamic acid
from bovine kidney, in Vitamin K Metabolism and Vitamin K-Dependent Proteins (J. W.
Suttie, ed.). University Park Press, Baltimore, 1980, pp. 307–310.
219. A. Gardemann and G. F. Domagk, The occurrence of ?-carboxyglutamate in a protein isolated
from ox liver mitochondria. Arch. Biochem. Biophys. 220: 347–353 (1983).
220. D. M. Smalley and P. C. Preusch, Analysis of ?-carboxyglutamic acid by reverse phase HPLC
of its phenylthiocarbamyl derivative. Anal. Biochem. 172:241–247 (1988).
221. X-J. Wen and M. Thierry-Palmer, Identification of a vitamin K–dependent protein in rat
liver mitochondria. FASEB J. 9:A725, 1995.
222. B. A. M. Soute, W. Muller-Ester, M. A. G. de Boer-van den Berg, M. Ulrich, and C. Vermeer,
Discovery of a ?-carboxyglutamic acid-containing protein in human spermatozoa. FEBS Lett.
190:137–141 (1985).
223. Nakagawa, Y., V. Abram, F. J. Kezdy, E. T. Kaiser, and F. L. Coe, Purification and characterization
of the principal inhibitor of calcium oxalate monohydrate crystal growth in human
urine. J. Biol. Chem. 258:12594–12600 (1983).
223a. J. D. Kulman, J. E. Harris, B. A. Haldeman, and E. W. Davie, Primary structure and tissue
distribution of two novel proline-rich ?-carboxyglutamic acid proteins. Proc. Natl. Acad. Sci.
USA 94:9058–9062 (1997).
224. G. Manfioletti, C. Bancolini, G. Avanzi, and C. Schneider, The protein encoded by a growth
arrest–specific gene (gas6) is a new member of the vitamin K-dependent proteins related to
protein S, a negative coregulator in the blood coagulation cascade. Mol. Cell. Biol. 13:4976–
4985 (1993).
225. P. V. Hauschka, E. A. Mullen, G. Hintsch, and S. Jazwinski, Abundant occurrence of ?-
carboxyglutamic acid (Gla)-containing peptides in the marine gastropod family Conidae, in
162 Suttie
Current Advances in Vitamin K Research (J. W. Suttie, ed.). Elsevier, New York, 1988,
pp. 237–243.
226. G. Tans, J. W. P. Govers-Riemslag, J. L. M. L. van Rijn, and J. Rosing, Purification and
properties of a prothrombin activator from the venom of Notechis scutatus scutatus. J. Biol.
Chem. 260:9366–9370 (1985).
227. B. M. Olivera, W. R. Gray, R. Zeikus, J. M. McIntosh, J. Varga, J. R. Victoria de Santos,
and L. J. Cruz, Peptide neurotoxins from fish-hunting cone snails. Science 230:1338–1343
228. J. Hirsh, J. S. Ginsberg, and V. J. Marder, Anticoagulant therapy with coumarin agents, in
Hemostasis and Thrombosis: Basic Principles and Clinical Practice, 3rd ed. (R. W. Colman,
J. Hirsh, V. J. Marder, and E. W. Salzman, eds.). J. B. Lippincott, Philadelphia, 1994, pp.
228a. D. Cain, S. M. Hutson, and R. Wallin, Assembly of the warfarin-sensitive vitamin K 2,3-
epoxide reductase enzyme complex in the endoplasmic reticulum membrane. J. Biol. Chem.
272:29068–29075 (1997).
229. L. J. M. van Haarlem, B. A. M. Soute, and C. Vermeer, Vitamin K–dependent carboxylase.
Possible role for thioredoxin in the reduction of vitamin K metabolites in liver. FEBS Lett.
222:353 (1987).
230. R. B. Silverman and D. L. Nandi, Reduced thioredoxin: a possible physiological cofactor
for vitamin K epoxide reductase. Further support for an active site disulfide. Biochem. Biophys.
Res. Commun. 155:1248 (1988).
231. J. T. Matschiner, R. G. Bell, J. M. Amelotti, and T. E. Knauer, Isolation and characterization
of a new metabolite of phylloquinone in the rat. Biochim. Biophys. Acta 201:309 (1970).
232. R. G. Bell and J. T. Matschiner, Warfarin and the inhibition of vitamin K activity by an
oxide metabolite. Nature 237:32 (1972).
233. J. A. Sadowski and J. W. Suttie, Mechanism of action of coumarins. Significance of vitamin
K epoxide. Biochemistry 13:3696–3699 (1974).
234. A. Zimmermann and J. T. Matschiner, Biochemical basis of hereditary resistance to warfarin
in the rat. Biochem. Pharmacol. 23:1033 (1974).
235. D. S. Whitlon, J. A. Sadowski, and J. W. Suttie, Mechanism of coumarin action: significance
of vitamin K epoxide reductase inhibition. Biochemistry 17:1371 (1978).
236. M. J. Fasco and L. M. Principe, R-and S-warfarin inhibition of vitamin K and vitamin K
2,3-epoxide reductase activities in the rat. J. Biol. Chem. 257:4894 (1982).
237. M. J. Fasco, E. F. Hildebrandt, and J. W. Suttie, Evidence that warfarin anticoagulant action
involves two distinct reductase activities. J. Biol. Chem. 257:11210 (1982).
238. E. F. Hildebrandt and J. W. Suttie, Mechanism of coumarin action: sensitivity of vitamin
K metabolizing enzymes of normal and warfarin resistant rat liver. Biochemistry 21:2406
239. R. Wallin, S. D. Patrick, and L. F. Martin, Vitamin K1 reduction in human liver. Biochem.
J. 260:879 (1989).
240. S. L. Gardill and J. W. Suttie, Vitamin K epoxide and quinone reductase activities: evidence
for reduction by a common enzyme. Biochem. Pharmacol. 40:1055 (1990).
241. K. W. E. Denson and R. Biggs, Laboratory diagnosis, tests of clotting function and their
standardization, in Human Blood Coagulation, Haemostasis and Thrombosis (R. Biggs, ed.).
Blackwell Scientific Publishers, 1972, pp. 278–332.
242. B. R. J. Kirchhof, C. Vermeer, and H. C. Hemker, The determination of prothrombin using
synthetic chromogenic substrates; choice of a suitable activator. Thromb. Res. 13:219–232
243. K. W. E. Denson, R. Borrett, and R. Biggs, The specific assay of prothrombin using the
Taipan snake venom. Br. J. Haematol. 21:219–226 (1971).
244. J. W. Suttie, Vitamin K and human nutrition. J. Am. Dietet. Assoc. 92:585–590 (1992).
245. M. L. Scott, Vitamin K in animal nutrition. Vitam Hormones 24:633–647 (1966).
Vitamin K 163
246. E. A. Doisy and J. T. Matschiner, Biochemistry of vitamin K, in Fat-Soluble Vitamins
(R. A. Morton, ed.). Pergamon Press, Oxford, 1970, pp. 293–331.
247. National Academy of Sciences. Nutritional Requirements of Laboratory Animals, 3rd ed.,
Washington, D.C. 1978.
248. P. G. Reeves, F. H. Nielsen, and G. C. Fahey, Jr., AIN-93 purified diets for laboratory rodents:
final report of the American Institute of Nutrition ad Hoc Writing Committee on the Reformulation
of the AIN-76A Rodent Diet. J. Nutr. 123:1939–1951 (1993).
249. P. G. Frick, G. Riedler, and H. Brogli, Dose response and minimal daily requirement for
vitamin K in man. J. Appl. Physiol. 23:387–389 (1967).
250. R. A. O’Reilly, Vitamin K in hereditary resistance to oral anticoagulant drugs. Am. J. Physiol.
221:1327–1330 (1971).
251. J. W. Suttie, L. L. Mummah-Schendel, D. V. Shah, B. J. Lyle, and J. L. Greger, Development
of human vitamin K deficiency by dietary vitamin K restriction. Am. J. Clin. Nutr. 47:475–
480 (1988).
252. G. Ferland, J. A. Sadowski, and M. E. O’Brien, Dietary induced subclinical vitamin K defi-
ciency in normal human subjects. J. Clin. Invest. 91:1761–1768 (1993).
253. R. A. Blanchard, B. C. Furie, M. Jorgensen, S. F. Kruger, and B. Furie, Acquired vitamin
K-dependent carboxylation deficiency in liver disease. N. Engl. J. Med. 305:242–248
254. K-S. G. Jie, K. Hamulyak, B. L. M. G. Gijsbers, F. J. M. E. Roumen, and C. Vermeer, Serum
osteocalcin as a marker for vitamin K-status in pregnant women and their newborn babies.
Thromb. Haemost. 68:388–391 (1992).
255. M. H. J. Knapen, K. Hamulyak, and C. Vermeer, The effect of vitamin K supplementation
on circulating osteocalcin (bone Gla protein) and urinary calcium excretion. Ann. Intern.
Med. 111:1001–1005 (1989).
255a. S. L. Booth, M. E. O’Brien-Morse, G. E. Dallal, K. W. Davidson, and C. M. Gundberg,
Response of vitamin K status to different intakes and sources of phylloquinone-rich foods:
comparison of younger and older adults. Am. J. Clin Nutr. 70:368–377 (1999).
256. K. Hazell and K. H. Baloch, Vitamin K deficiency in the elderly. Gerontol. Clin. 12:10–17
257. E. A. Doisy, Jr., Nutritional hypoprothrombinemia and metabolism of vitamin K. Fed. Proc.
20:989–994 (1961).
258. P. A. Lane and W. E. Hathaway, Vitamin K in infancy. J. Pediatr. 106:351–359 (1985).
259. M. Andrew, B. Paes, R. Milner, M. Johnston, L. Mitchell, D. M. Tollefsen, and P. Powers,
Development of the human coagulation system in the full-term infant. Blood 70:165–172
260. R. v. Kries, F. R. Greer, and J. W. Suttie, Assessment of vitamin K status of the newborn
infant. J. Pediatr. Gastroenterol. Nutr. 16:231–238 (1993).
261. Kries, R. v., A. Becker, and U. Gobel, Vitamin K in the newborn: influence of nutritional
factors on acarboxy-prothrombin detectability and factor II and VII clotting activity. Eur. J.
Pediatr. 146:123–127 (1987).
262. K. Motohara, I. Matsukane, F. Endo, Y. Kiyota, and I. Matsuda, Relationship of milk intake
and vitamin K supplementation to vitamin K status in newborns. Pediatrics 84:90–93 (1989).
263. B. E. Gustafsson, F. S. Daft, E. G. McDaniel, and J. C. Smith, Effects of vitamin K-active
compounds and intestinal microorganisms in vitamin K–deficient germfree rats. J. Nutr. 78:
461–468 (1962).
264. T. S. Nelson and L. C. Norris, Studies on the vitamin K requirement of the chick. II. Effect
of sulfaquinoxaline on the quantitative requirements of the chick for vitamin K1, menadione
and menadione sodium bisulfite. J. Nutr. 73:135–142 (1961).
265. D. Savage and J. Lindenbaum, Clinical and experimental human vitamin K deficiency, in
Nutrition in Hematology (J. Lindenbaum,ed.). Churchill Livingstone, New York, 1983,
pp. 271–320.
164 Suttie
266. J. J. Lipsky, Mechanism of the inhibition of the ?-carboxylation of glutamic acid by Nmethylthiotetrazole-
containing antibiotics. Proc. Natl. Acad. Sci. USA 81:2893–2897 (1984).
267. H. Bechtold, K. Andrassy, E. Jahnchen, J. Koderisch, H. Koderisch, L. S. Weilemann, H-G.
Sonntag, and E. Ritz, Evidence for impaired hepatic vitamin K1 metabolism in patients treated
with N-methyl-thiotetrazole cephalosporins. Thromb. Haemostas. (Stuttg). 51:358–361
268. K. A. Creedon and J. W. Suttie, Effect of N-methyl-thiotetrazole on vitamin K epoxide reductase.
Thromb. Res. 44:147–153 (1986).
269. H. Suzuki, T. Nakao, and K, Hiraga, Vitamin K deficiency in male rats fed diets containing
butylated hydroxytoluene (BHT). Toxicol. Appl. Pharmacol. 50:261–266 (1979).
270. O. Takahashi and K. Hiraga, Preventive effects of phylloquinone on hemorrhagic death induced
by butylated hydroxytoluene in male rats. J. Nutr. 109:453–457 (1979).
271. K. R. Mountain, A. S. Gallus, and J. Hirsch, Neonatal coagulation defection due to anticonvulsant
drug treatment in pregnancy. Lancet: 1:265 (1970).
272. A. S. Rogen and J. C. Ferguson, Clinical observation on patients treated with atronid and
anticoagulants. J. Atheroscler. Res. 3:671–676 (1963).
273. J. G. Hall, R. M. Pauli, and K. M. Wilson, Maternal and fetal sequelae on anticoagulation
during pregnancy. Am. J. Med. 68:122–140 (1980).
274. J. T. Matschiner and R. G. Bell, Effect of sex and sex hormones on plasma prothrombin and
vitamin K deficiency. Proc. Soc. Exp. Biol. Med. 144:316–320 (1973).
275. J. T. Matschiner and A. K. Willingham, Influence of sex hormones on vitamin K deficiency
and epoxidation of vitamin K in the rat. J. Nutr. 104:660–665 (1974).
276. D. W. Jolly, B. M. Kadis, and T. E. Nelson, Jr., Estrogen and prothrombin synthesis. The
prothrombinogenic action of estrogen. Biochem. Biophys. Res. Commun. 74:41–49 (1977).
277. C. M. Siegfried, G. R. Knauer, and J. T. Matschiner, Evidence for increased formation of
preprothrombin and the noninvolvement of vitamin K–dependent reactions in sex-linked
hyperprothrombinemia in the rat. Arch. Biochem. Biophys. 194:486–495 (1979).
278. Y. Nishino, Hormonal control of prothrombin synthesis in rat liver microsomes, with special
reference to the role of estradiol, testosterone and prolactin. Arch. Toxicol. Suppl. 2:397–
402 (1979).
279. A. T. van Oosterom, P. Kerkhoven, and J. J. Veltkamp, Metabolism of the coagulation factors
of the prothrombin complex in hypothyroidism in man. Thrombos. Haemostas. (Stuttg.) 41:
273–285 (1979).
280. M. C. Elliott, B. Isaacs, and A. C. Ivy, Production of ‘‘prothrombin deficiency’’ and response
to vitamins A, D and K. Proc. Soc. Exp. Biol. Med. 43:240–245 (1940).
281. G. H. Rao and K. E. Mason, Antisterility and antivitamin K activity of d-?-tocopheryl hydroquionone
in the vitamin E–deficient female rat. J. Nutr. 105:495–498 (1975).
282. J. J. Corrigan, Jr., and F. I. Marcus, Coagulopathy associated with vitamin E ingestion. JAMA
230:1300–1301 (1974).
283. L. Uotila, Inhibition of vitamin K-dependent carboxylase by vitamin E and its derivatives,
in Current Advances in Vitamin K Research (J. W. Suttie, ed.). Elsevier, New York, 1988,
pp. 59–64.
284. C. A. Owen, Vitamin K group. XI. Pharmacology and toxicology, in The Vitamins, Vol. 3,
2nd ed. (W. H. Sebrell and R. S. Harris, eds.). Academic Press, New York, 1971, pp. 492–
285. National Research Council, Vitamin Tolerance of Animals, National Academy Press, Washington,
D.C., 1987.
Vitamin E
University of Kentucky, Lexington, Kentucky
Vitamin E was discovered by Evans and Bishop more than 75 years ago as a lipid-soluble
substance in lettuce and wheat, necessary for the prevention of fetal death and resorption
in rats fed a rancid lard diet (1). This substance was designated as vitamin E following
the recognition of vitamin D (2,3). The term ‘‘tocopherol’’ is used after the Greek words
‘‘tokos’’ (childbirth), ‘‘phero’’ (to bring forth), and ‘‘ol’’ (alcohol). ?-Tocopherol was
isolated from wheat germ oil in 1936 (4). It was first synthesized by Karrer (5), and its
structure was determined by Fernholz (6) in 1938. The antioxidant properties of tocopherols
were first reported by Olcott and Emerson in 1937 (7).
In addition to fetal resorption in rats, a number of species-dependent deficiency
symptoms of vitamin E, such as liver necrosis in rats and pigs, erythrocyte hemolysis in
rats and chicken, and white muscle disease in calves, sheep, mice, and mink, were reported
during the 1940s and 1950s (8–10). However, due to the lack of a definite clinical syndrome
attributable to vitamin E deficiency and the difficulty of inducing vitamin E defi-
ciency in human adults, the need for or use of this vitamin had been questioned. In the
late 1960s, the essentiality of vitamin E for humans was recognized in connection with
studies on premature infants in which hemolytic anemia was associated with vitamin E
deficiency (9,11). Subsequent studies have shown that neurologic abnormalities do occur
in association with malabsorption syndromes of various etiologies (9,11,12). In recent
years, the involvement of free radicals in the pathogenesis of degenerative diseases and
the possible prevention or slow-down of the disease process by antioxidants have promoted
a renewed and expanded interest in vitamin E.
166 Chow
Vitamin E is the term suggested for all tocol and tocotrienol derivatives qualitatively
exhibiting the biological activity of ?-tocopherol. The term ‘‘tocopherols’’ is the generic
description for all mono-, di-, and trimethyltocols and tocotrienols, and is not synonymous
with the term ‘‘vitamin E.’’ All eight naturally occurring tocopherol compounds isolated
from plant sources have a 6-chromanol ring (head) and a phytyl side chain (tail) (8–
10,13,14). There are four tocopherols and four tocotrienols that occur naturally, differing
in the number and position of methyl groups on the phenolic ring (Fig. 1). Tocotrienols
have a structure similar to that of tocopherols, except that the side chain contains three
isolated double bonds at the 3?, 7?, and 11? positions.
The tocopherol molecule has three chiral centers (2, 4?, and 8?) in its phytyl tail.
Since a total of eight stereoisomeric forms may exist, tocopherols of unspecified configuration
are more accurately called methyl-substituted tocols (Table 1). The term ‘‘tocol’’
is the trivial designation for 2-methyl-2-(4?, 8?, 12?-trimethyltridecylchroman-6-ol. All naturally
occurring tocopherols (?-, ?-, ?-, and ?-) have the same molecular configuration
Fig. 1 Structural formula of tocopherols.
Vitamin E 167
Table 1 Vitamin E Compounds
Vitamin Trivial name Chemical name
?-Tocopherol 5,7,8-Trimethyltocol 2,5,7,8-Tetramethyl-2-(4?,8?,12?-trimethyltridecyl)-
?-Tocopherol 5,8-Dimethyltocol 2,5,8-Trimethyl-2-(4?,8?,12?-trimethyltridecyl)-6-
?-Tocopherol 7,8-Dimethyltocol 2,7,8-Trimethyl-2-(4?,8?,12?-trimethyltridecyl)-6-
?-Tocopherol 8-Monomethyltocol 2,8-Dimethyl-2-(4?,8?,12?-trimethyltridecyl)-6-
?-Tocotrienol 5,7,8-Trimethyl tocotrienol 2,5,7,8-Tetramethyl-2-(4?,8?,12?-trimethyltridecyl)-
?-Tocotrienol 5,8-Dimethyl tocotrienol 2,5,8-Trimethyl-2-(4?,8?,12?-trimethyltridecyl)-6-
?-Tocotrienol 7,8-Dimethyl tocotrienol 2,7,8-Trimethyl-2-(4?,8?,12?-trimethyltridecyl)-6-
?-Tocotrienol 8-Monomethyl tocotrienol 2,8-Dimthyl-2-(4?,8?,12?-trimethyltridecyl)-6-
[RRR, 2d, 4?d, 8?d, d-tocopherol or ()-tocopherols] in their phytyl groups. Since tocotrienols
have only one chiral center at position 2, they can only have 2d and 2l stereoisomers.
However, the double bonds at positions 3? and 7? of the phytyl tail allow for the existence
of four cis/trans geometrical isomers (a total of eight isomers, theoretically) per tocotrienol
In addition to naturally occurring isomers, several types of synthetic vitamin E,
mainly in ester forms (e.g., ?-tocopheryl acetate and tocopheryl succinate), are available
commercially. The ester form is less susceptible to oxidation and is therefore more suitable
for food and pharmaceutical applications than the free form. To distinguish it from the
synthetic one, the naturally occurring stereoisomer of ?-tocopherol, formerly known as
d-?-tocopherol, has been designated as RRR-?-tocopherol. The totally synthetic ?-tocopherol,
previously known as dl-?-tocopherol or 2dl, 4?dl, 8?dl-tocopherol, which consists
of eight stereoisomers [2d, 4?d, 8?d (RRR), 2l, 4?d, 8?d (SRR), 2d, 4?d, 8?l (RRS), 2l,
4?d,?8?l (SRS), 2d, 4?l, 8?d (RSR), 2l, 4?l, 8?d (SSR), 2d, 4?l, 8?l (RSS), and 2l, 4?l, 8?l
(SSS)] (13,14), has been designated as all-rac ?-tocopherol. All-rac ?-tocopherol is usually
synthesized by the condensation of trimethylhydroquinone with isophytol. The majority
of tocopherol present in plants is not the ? form. Therefore, RRR-?-tocopherol, which
is referred as natural ?-tocopherol, is often obtained following methylation of tocopherol
mixtures isolated from vegetable oils.
Tocopherols are pale yellow or yellow-brown viscous liquids at room temperature.
At high purity they are almost odorless and colorless. Free tocopherols are practically
insoluble in water but are soluble in oils, fats, acetone, ethanol, chloroform, and other
organic solvents. The boiling point of tocopherols at 0.1 mm Hg pressure is 200–220°C,
their molecular weight ranges from 396.6 to 430.69, and the molar absorbance (E1%; 1
cm) at 292–298 nm ranges from 75.8 to 91.4 (Table 2) (14a). Tocopherols can be oxidized
by atmospheric oxygen, and the oxidation is accelerated by heat, light, alkali, and metal
ions. On the other hand, tocopherols are rather stable to heat and alkali when oxygen is
168 Chow
Table 2 Molecular Weight and UV Absorption Maxima and Molar
Absorbance of Vitamin E Compounds
Mol w ?max (nm) E%
1 cm
?-Tocopherol 430.7 292.0 75.8
?-Tocopherol 416.7 296.0 89.4
?-Tocopherol 416.7 298.0 91.4
?-Tocopherol 402.6 298.0 87.3
?-Tocotrienol 424.6 292.5 91.0
?-Tocotrienol 410.6 294.0 87.3
?-Tocotrienol 410.6 296.0 90.5
?-Tocotrienol 396.6 297.0 88.1
Source: Adapted from Ref. 14a.
Generally, the first step of tocopherol oxidation is the formation of resonance-stabilized
chromanoxyl (chroman-6-oxyl) radical, due to the donation of the phenolic hydrogen
to a lipid peroxyl radical (15,16). The chromanoxyl radicals are very reactive toward alkyl
and alkylperoxy radicals. Depending on the severity of the oxidation conditions and the
presence of other chemicals, different tocopherol oxidation products are obtained. For
example, oxidation of ?-tocopherol in polar solvents (water or alcohol) leads to the formation
of 5-formyl and hemiacetal derivatives of 5-hydroxymethylene chromanol (13,17).
In lipophilic solvents, the chromanoxy radicals tend to react mainly via radical–radical
coupling reactions and form dimers or trimers (17,18). In the presence of peroxy radicals,
?-tocopherol is primarily oxidized to 8?-peroxy-substituted tocopherone, which is then
degraded to form ?-tocopherylquinone, plus various tocopherone and quinone epoxides
and spiro-dimer and trimer (19–21). Relatively little is known about the products of and
mechanisms for the antioxidant reactions of tocotrienols. Tocotrienols, however, are expected
to yield the same type of radicals and dimeric products as their corresponding
tocopherols. Two dimers of ?-tocotrienol [5-(?-tocotrienyloxy)-?-tocotrienol and 5-(?-tocotrienyl)-
?-tocotrienol] are the principal oxidation products of ?-tocotrienol-rich Hevea
latex (22).
It has long been recognized that tocopherols possess antioxidant activity (7), and
the biological activity of vitamin E is mainly attributed to their ability to donate their
phenolic hydrogen to lipid free radicals. Tocopherols, as well as tocotrienols, can react
with peroxyl radicals more rapidly than can polyunsaturated fatty acids and, therefore,
are very effective free-radical chain-breaking antioxidants. Antioxidant activity of tocopherols
is determined by their chemical reactivity with molecular oxygen, superoxide radicals,
peroxyl radicals, or other radicals, or by their ability to inhibit autoxidation of fats
and oils. The chemical structures of the tocopherols and tocotrienols support a hydrogendonating
power in the order ?  ?  ?  ? (23). The presence of electron-releasing
substituents ortho and/or para to the hydroxy function increased the lectern density of the
active centers, facilitating the homolytic fission of the O-H bond, increasing the stability of
the phenoxy radical, and improving the reactivity with peroxy radicals (23). Thus, ?-
tocopherol is structurally expected to be a more potent hydrogen donor than ?-, ?-, and
?-tocopherols because these lack one or two ortho-methyl group. The order of ?  ? 
Vitamin E 169
?  ? is generally found in the relative antioxidant activity of the tocopherols in vivo
(see Sec. V). The relative antioxidant activity of tocopherols in vitro, however, varies
considerably depending on the experimental conditions and the assessment method employed.
?-tocopherol and some other isomers, for example, have sometimes been shown
to exhibit a higher antioxidant activity than the ? form in a number of in vitro test systems
(24). Also, tocotrienols are more effective than tocopherols in preventing oxidation under
certain conditions (25). In vitro antioxidant activities of the tocopherols depend not only
on their absolute chemical reactivities toward hydroperoxy and other free radicals, but
also on such factors as tocopherol concentrations, temperature, light, type of substrate and
solvent, and other chemical species, which may act as pro-oxidants and synergists, in the
test system (13,23).
Unlike free tocopherols, tocopheryl esters are much more stable to oxidation, and
do not function as antioxidants in vitro.
A. Intestinal Absorption
Similar to lipid components, intestinal absorption of vitamin E depends on pancreatic
function, biliary secretion, micelle formation, and transport across intestinal membrane
(26,27). Tocopherol is first emulsified and solubilized within bile salt micelles and transported
across the water layer to come in contact with the absorption brush-border membrane
of the enterocyte. Tocopheryl esters, such as tocopheryl palmitate, tocopheryl acetate,
and tocopheryl succinate, are hydrolyzed to free tocopherol during the absorptive
process. Under physiological conditions the hydrolysis occurs in gut lumen, although a
mucosal esterase has also been found in the endoplasmic reticulum of the enterocytes
(28). Free tocopherol is absorbed by a passive diffusion process from small intestine to
the enterocyte (29). The absorption process is nonsaturable, non-carrier-mediated, and
does not require energy. The area at the junction between the upper and middle thirds of
the small intestine appears to be the region for tocopherol uptake by enterocytes (29,30).
Within the enterocyte, tocopherol is incorporated into chylomicrons and secreted into the
intracellular spaces and lymphatics, and thus into the bloodstream. This transfer process
does not appear to require a specific protein.
The efficacy of tocopherol absorption varies considerably depending on the conditions
and methods for its estimation. Generally, the rate of absorption decreases as the
amount taken increases. Medium-chain triglycerides have been shown to enhance the absorption
process, whereas long-chain polyunsaturated fatty acids reduce the absorption of
?-tocopherol (31,32). The adverse effect of dietary long-chain polyunsaturated fatty acids
on vitamin E absorption may partly be due to an increase in oxidation of tocopherol during
In humans, the majority of ?-tocopherol is absorbed via the lymph ducts, and a
small portion is absorbed by the portal vein (26,27). The absorption of tocopherols other
than the ? form is not as well understood. However, studies have shown that there are
no major differences in the rate of intestinal absorption between ?- and non-?-tocopherols,
such as ?-tocopherol, and that excess of the ? form does not reduce the absorption of ?-
tocopherols (33,34). Thus, a higher level of ?-tocopherols than other tocopherols in human
plasma and tissues is the result of factors other than absorption.
170 Chow
B. Plasma Transport
Due to its hydrophobicity vitamin E requires special transport mechanisms in the aqueous
milieu of the plasma, body fluid, and cells. Similar to dietary lipids, tocopherols absorbed
in the small intestine are incorporated into triglyceride-rich chylomicrons, secreted into
intestinal lymph, and then delivered to the liver. There is no discrimination between various
forms of vitamin E during chylomicron secretion (35,36).
Following delivery to the liver from small intestine, tocopherol is repackaged with
very low-density lipoproteins (VLDLs) and secreted into plasma. At this stage, RRR-?-
tocopherol is preferentially secreted with VLDL over SRR-?-tocopherol and RRR-?-
tocopherol (35–37). During the conversion of VLDLs to low-density lipoproteins (LDLs)
in the circulation, a portion of tocopherol is transferred to LDLs. Exchange of tocopherol
also occurs between LDLs and high-density lipoproteins (HDLs) (38). Therefore, tocopherol
is distributed in all lipoproteins (39,40). In humans, the highest concentration of vitamin
E is found in LDL and HDL. Vitamin E can also be exchanged between lipoproteins
and liposomes, and between plasma and red blood cells. While no specific plasma/serum
carrier protein for vitamin E has been reported, a phospholipid transfer protein that accelerates
exchange/transfer of ?-tocopherol between lipoproteins and cells has been identified
in human plasma (41). This protein may play a role in determining the concentration of
tocopherols present in the circulation and in tissues.
Normal plasma vitamin E concentration in humans ranged from 11 to 37 µmol/L
(5 to 16 mg/L), 1.6 to 5 µmol ?-tocopherol/mmol lipid (0.8 to 1.7 mg/g), or 2.5 to 8
µmol ?-tocopherol/mmol cholesterol (2.8 to 8 mg/g). Approximately 10–20% of total
tocopherols in human plasma is the ? form. In malabsorption states, such as cystic fibrosis,
plasma lipids, lipoproteins, and vitamin E concentration are frequently reduced concurrently
C. Tissue Uptake
The mechanisms involved in the uptake of tocopherols by tissue remain unclear. Lipoprotein
lipase bound to the surface of the endothelial lining of capillary walls catabolizes the
triglycerides in the core of chylomicrons and forms chylomicron remnants (26,27). The
latter is then taken up by the parenchymal cells in the liver via the remnant receptors on
the surface of hepatocytes (42,43). Along with the free fatty acids, some vitamin E is also
taken up by peripheral tissues during catabolism of lipoproteins by lipoprotein lipase.
Tocopherol is taken up by tissues via several mechanisms. This includes lipoprotein receptor
(apolipoprotein B/E)–dependent and receptor-independent pathways, independent
transport and cotransport of ?-tocopherol and LDL, and uptake from a number of lipoproteins
(44). The action of lipoprotein lipase is an important process prior to tissue uptake
of tocopherols.
Most tissues, including liver, skeletal muscle, and adipose tissue, have the capacity
to accumulate ?-tocopherol (45–47). In adipose tissue, tocopherol is located mainly in
the bulk lipid droplet and its turnover is slow. The adrenal gland has the highest concentration
of ?-tocopherol, although lung and spleen also contain relatively high concentrations
(47,48). ?-Tocopherol is primarily taken up and located in parenchymal cells (49). The
transfer of tocopherol between parenchymal and nonparenchymal cells takes place after
uptake in parenchymal cells. Parenchymal, but not nonparenchymal, cells have a large
storage capacity for surplus ?-tocopherol and are more resistant to tocopherol depletion
than nonparenchymal cells. Light mitochondria has the highest concentration of ?-tocoVitamin
E 171
pherol, whereas the concentration is low in cytosol (49,50). The majority of tocopherol
is localized in the membranes. For example, approximately three-fourths of mitochondrial
?-tocopherol is found in the outer membrane, and one-fourth is associated with the inner
membrane (49). Also, essentially all tocopherol in the red cells is found in the membranes
D. Hepatic Secretion
Vitamin E taken up by the liver is stored in the parenchymal cells or secreted into the
bloodstream within nascent VLDLs. Following the action of lipoprotein lipase, some vitamin
E in the VLDLs may end up in LDLs as these lipoproteins are transformed in plasma.
Vitamin E in LDLs may again be taken up by liver via the LDLs (apolipoprotein B/E)
receptor or by non-receptor-mediated uptake (26,42,43). Some vitamin E in association
with chylomicrons and VLDL may also be transferred to peripheral cells and HDLs during
lipolysis by lipoprotein lipase. Secreted tocopherols are either rapidly returned from
plasma to the liver during the course of lipid metabolism or excreted via bile. The secretory
pathway via nascent VLDLs from liver seems to be critical in maintaining tocopherol
concentrations in plasma. Studies have shown that discrimination between the stereoisomers
of tocopherol occurs during hepatic secretion of nascent VLDLs (36). RRR-?-
Tocopherol is preferentially incorporated over other tocopherols into VLDLs, which are
then secreted into plasma. This is evidenced by the much higher levels of RRR-?-tocopherol
than either SRR-?- or RRR-?-tocopherol found in the nascent VLDLs and plasma of
cynomolgus monkeys after the same amount of each was given (37). The preference of
a specific hepatic tocopherol-binding protein to RRR-?-tocopherol appears to be responsible
for this discrepancy.
E. Tocopherol-Binding Proteins
A cystosolic tocopherol-binding protein has been shown to selectively facilitate the incorporation
of RRR-?-tocopherols into nascent VLDLs. By preferentially binding the ? form
over the other forms of tocopherols, the binding protein determines which tocopherols
are returned to the liver from plasma via HDLs and LDLs (52–55). The discrimination
between the tocopherol isomers and homologues appears to be due to the preference of
the tocopherol-binding protein to recognize specific structural features: a fully methylated
aromatic ring, a saturated phytyl side chain, and a stereochemical RRR configuration of
the methyl group’s branching of the side chain (14,36). There is a chiral discrimination
between different isomers of ?-tocopherol, with preference retention of the RRR stereoisomer
over the SRR form (35,36). By regulating the binding and/or transfer of tocopherol,
the transfer protein plays a key role in determining plasma concentration and biological
activity of tocopherols (55). The process results in the preferential enrichment of LDL
and HDL with ?-tocopherol in plasma. The critical role of tocopherol-binding protein in
regulating plasma tocopherol concentration has been indicated in patients with familial
isolated vitamin E deficiency (56,57). Those patients have clear signs of vitamin E defi-
ciency (extremely low plasma vitamin E and neurological abnormalities) but have no fat
malabsorption or lipoprotein abnormalities. Absence of the transfer protein in those patients,
which impairs secretion of tocopherol into hepatic lipoproteins (VLDLs), has been
postulated to be responsible for their low plasma vitamin E status (26,55). The human ?-
tocopherol transfer protein gene was shown to be located at chromosome 8q13 (57a).
Studies of the gene structure and mutations of ?-tocopherol transfer protein in patients
172 Chow
with familial vitamin E deficiency have confirmed that mutations of the gene for ?-tocopherol
transfer protein are responsible for isolated vitamin E deficiency (58,58a). The degree
of functionality of the mutant ?-tocopherol transfer protein seems to be associated with
the degree of severity of the neurological damage and age of onset (59).
A tocopherol/binding protein with a molecular weight of 30–36 kd has been purified
from rat liver cytosol (54,55,60,61). The protein binds ?-tocopherol and enhances its transfer
between membranes. The purified protein has two isoforms with isoelectric points at
5.0 and 5.1, respectively (61). The binding protein exhibits a structural homology with
the cellular retinaldehyde-binding protein present only in visual tissues (62). A tocopherolbinding
protein with a molecular weight of 36.6 kd has also been identified and purified
from human liver cytosol (63). An apparently different ?-tocopherol binding protein with
a molecular mass of 14.2 kd has been isolated and purified from rabbit heart (64) and rat
liver and heart (65). The binding of this smaller binding protein to ?-tocopherol is rapid,
reversible, and saturable, and neither ?- nor ?-tocopherol can replace the binding of ?-
tocopherol. It also markedly stimulates the transfer of tocopherol from liposomes to mitochondria.
The binding protein is different from the cytosolic fatty acid–binding protein
and may be involved in intracellular transport and metabolism of ?-tocopherol (66). The
relationship between those two binding proteins, if any, is not known.
F. Metabolism
The major route of excretion of absorbed tocopherol is fecal elimination. Excess ?-tocopherol
and other forms of tocopherols are excreted first into the bile and then into the
feces. After exerting its antioxidant activity tocopherol is first converted to tocopheryl
chromanoxy radical. The chromanoxy radical is readily reverted to tocopherol, and the
process can be facilitated by such reducing agents as glutathione and ascorbate and/or a
yet to be identified enzymic system. Tocopheryl chromanoxy radicals can also form dimer
or trimer, or be further oxidized to form tocopheryl quinone (Fig. 2). A small amount of
?-tocopheryl quinone is found in liver (67). Studies on the metabolic fate of C14-labeled
?-tocopheryl quinone and ?-tocopheryl hydroquinone have showed that there is no conversion
to ?-tocopherol in vivo (68). ?-Tocopheryl quinone, however, can be reduced
likely by an enzymic process to ?-tocopheryl hydroquinone, and then further metabolized
to ?-tocopheronic acid. Higher activity of NADPH-dependent tocopheryl quinone reductase
is found in the mitochondria and microsomes than in the cytosol of rat hepatocytes
(69). Also, NADPH-cytochrome P450 reductase is capable of catalyzing tocopheryl hydroquinone
formation from tocopheryl quinone (69). Tocopheronic acid is subsequently conjugated
with glucuronic acid or other compounds, and excreted in urine (68). The urinary
metabolite, ?-tocopheronic acid, was first isolated by Simon and co-workers (70) after
administration of large doses of ?-tocopherol to rabbits and humans. A portion of bound
or conjugated ?-tocopheryl hydroquinone is secreted in the bile and eliminated in the
feces (68). A small portion of dimer and trimer of ?-tocopherol has also been found in
rat liver (71,72).
Results obtained form recent studies suggest that 2,5,7,8-tetramethyl-1(2?-carboxyethyl)-
6-hydroxy chroman, instead of ?-tocopheronic acid, may be the major urinary metabolite
of ?-tocopherol (72a). The urinary metabolite appears to be formed in liver directly
from a side chain degradation of ?-tocopherol without oxidative splitting of the
chroman ring (Fig. 2). The compound can be oxidized readily to form ?-tocopheronic
Vitamin E 173
Fig. 2 Metabolic fate of ?-tocopherol. GSH represents reduced glutathione; GSSG, oxidized glutathione;
NADPH and NADP, reduced and oxidized nicotinamide dinucleotide phosphate, respectively.
acid (or tocopheronolactone). Similarly, 2,7,8-trimethyl-2-(2?-carboxyethyl)-6-hydroxy
chroman and 2,8-dimethyl-2(2?-carboxyethyl)-6-hydroxy chroman have been identified as
the principal urinary metabolites for ?- and ?-tocopherols, respectively (72b,72c).
A. Deficiency Symptoms
A number of species-dependent and tissue-specific symptoms of vitamin E deficiency have
been reported (Table 3). The development and severity of certain vitamin E deficiency
symptoms are associated with the status of other nutrients, including polyunsaturated fatty
acids, selenium, and sulfur amino acids (8,9). For example, the most common sign of
deficiency is necrotizing myopathy, which occurs in almost all species in the skeletal
muscle, as well as in some heart and smooth muscles. However, in lambs and calves,
myopathy primarily results from selenium deficiency, and in rabbits and guinea pigs, severe
debilitating myopathy develops when fed a low vitamin E diet alone. On the other
174 Chow
Table 3 Pathology of Vitamin E Deficiency
Condition Animal Tissues affected
Reproductive failure:
Embronic degeneration Female: rat, hen, turkey, Vascular system of embryo
Male: rat, guinea pig, ham- Male gonads
ster, dog, cock
Liver, blood, brain, capillaries:
Liver necrosis Rat, pig Liver
Erythrocyte destruction Rat, chick Blood (hemolysis)
Anemia Monkey Bone marrow
Blood protein loss Chick, turkey Serum albumin
Encephalomalacia Chick Cerebellum (Purkinje cells)
Exudative diathesis Chick, turkey Vascular system
Depigmentation Rat Incisors
Kidney degeneration Rat, monkey, mink Kidney tubular epithelium
Steatitis Mink, pig, chick Depot fat
Eosinophilia Rat Blood, small intestine, liver,
stomach, and skeletal
Nutritional myopathies:
Nutritional muscular dystrophy Rabbit, guinea pig, rat, mon- Skeletal muscle
key, duck, turkey
Stiff lamb Lamb, kid Skeletal muscle
White-muscle disease Calf, sheep, mouse, mink Skeletal and heart muscle
Myopathy of gizzard and heart Turkey, poult Gizzard, heart, and skeletal
Neuromuscular disorder Rat, monkey, humans Skeletal muscle, spinal cord
hand, rats manifest a relatively benign myopathy when fed a vitamin E–deficient diet,
and chickens develop no myopathy unless the diet is depleted of both vitamin E and
sulfur amino acids. In the chicken, encephalomalacia, a disorder characterized by localized
hemorrhage and necrosis in the cerebellum, is prevented by vitamin E but not by selenium.
However, in chicken, exudative diathesis (a disorder characterized by increased capillary
permeability) can be prevented by either vitamin E or selenium (8,9). Similarly, eosinophilic
enteritis and eosinophilia in rats is preventable by vitamin E or selenium (73). The
cause of the species-dependent and tissue-specific vitamin E deficiency symptoms remains
to be delineated.
In humans, lower plasma/serum levels of vitamin E (0.5 mg/dL) are associated
with a shorter life span of the red cells and increased susceptibility to hemolytic stress.
Most of the vitamin E deficiency states in humans are associated with fat malabsorption
disorders (9,11,12,74). Low serum vitamin E levels have been observed in patients with
a variety of fat malabsorption conditions (Table 4). Recent studies of children and adults
with diseases that cause fat/vitamin E malabsorption, such as abetalipoproteinemia,
chronic cholestatic hepatobiliary disorder, and cystic fibrosis, as well as patients with the
primary form of vitamin E deficiency, familial isolated vitamin E deficiency syndrome,
have conclusively demonstrated that neurological dysfunction is associated with vitamin
E deficiency (12,26,56).
Vitamin E 175
Table 4 Tocopherol Levels in Human Vitamin E Deficiency States
Mean plasma/
serum ?-tocopherol
Condition (mg/100 mL) Range
Premature birth 0.25 0–0.56
Cystic fibrosis 0.24 0–0.97
Malabsorption syndromes other than cystic fibrosis:
Biliary atresia 0.11 —
Biliary cirrhosis 0 —
Biliary obstruction 0.08 0–0.14
Celiac disease 0.20 0–0.35
Chronic pancreatitis 0.40 0.17–0.79
Gastrectomy 0.36 0.10–0.80
Intestinal lymphangiectasia 0.28 —
Intestinal resection 0.31 —
Nontropical sprue 0.25 0.12–0.32
Regional enteritis plus intestinal resection 0.27 0.15–0.39
Tropical sprue 0.28 —
Ulcerative colitis 0.24 0.17–0.36
Whipple’s disease 0.30 0.14–0.45
Protein-calorie malnutrition 0.48 0–0.91
Normal 1.05 7.7–15.5
Source: After Ref. (11.)
B. Toxicity
Both acute and chronic studies with several species of animals have shown that high doses
of vitamin E are relatively nontoxic. However, extremely high levels of vitamin E
(16,000–64,000 IU/kg diet) have been shown to result in decreased pigmentation and a
waxy, feather like appearance (75), and abnormal mineralization and clotting abnormalities
(76) in the chick. Also, feeding pelicans with a high dose of vitamin E (5000 IU/kg
diet) results in vitamin K deficiency (77). These findings suggest that interference with
absorption or metabolic interaction with vitamins D and K is associated with extreme
dietary levels of vitamin E. In humans, supplementation of 800 IU vitamin E per day or
more for several years has consistently shown no adverse effects (78). Although increased
post-surgery bleeding and one individual case of breeding in patient receiving anticoagulant
therapy has been reported, a recent study (78a) has shown that moderate to large
doses of vitamin E (up to 1200 IU daily) can be safely used in patients receiving warfarin.
A. Biopotency
The utilization or function of tocopherols in biological tissues is governed not solely by
their chemical reactivities but by the biokinetics of their distribution, transport, retention,
and localization as well. Tocopherols differ widely in their antioxidant and biological
activities. A vitamin E activity of RRR-?-tocopherol higher than that of the other forms
176 Chow
suggests that the RRR configuration of the phytyl tail must be optimal for maximum
biopotency. The biological activity of tocopherol is assessed in terms of its relative ability
to prevent such deficiency-related symptoms as fetal resorption-gestation and erythrocyte
hemolysis (Table 5), and, more recently, by a curative myopathy test in rats (8,9,79,79a).
The curative myopathy test is based on the ability of tocopherol to suppress the increase
of pyruvate kinase activity in the plasma of vitamin E–deficient rats. Increased plasma
pyruvate kinase activity resulting from myodegeneration is a sensitive and specific indicator
for assessing the extent of vitamin E deficiency in rats (80). The increase of plasma
pyruvate kinase activity in rats with vitamin E deficiency is related to the nutritional status
of selenium (81). Markedly increases in plasma pyruvate kinase activity is also found in
dystrophic chickens and hamsters (82). Similarly, the pyruvate kinase activity in the serum
of patients with neuromuscular disorders is approximately 20- to 40-fold that of the control
values (83–85).
The biological activity of various forms of tocopherol is expressed as units of activity
in relation to that of all-rec-?-tocopheryl acetate. The relative values in international units
(IU/mg) are: all-rec-?-tocopheryl acetate, 1.00; all-rec-?-tocopherol, 1.10; RRR-?-tocopheryl
acetate, 1.36; RRR-?-tocopherol, 1.49; all-rec-?-tocopheryl succinate, 0.89; and
RRR-?-tocopheryl succinate, 1.21.
B. Biological Functions
Although many biochemical abnormalities are associated with vitamin E deficiency, the
mechanism by which vitamin E prevents various metabolic and pathological lesions has
not yet been elucidated. Vitamin E is the major lipid-soluble chain-breaking antioxidant
found in plasma, red cells, and tissues (14,86,87), and it plays an essential role in maintaining
the integrity of biological membranes. Among the biological functions proposed
for vitamin E, prevention against free-radical–initiated lipid peroxidation tissue damage
is the most important one accepted by most investigators. The antioxidant role of vitamin
E in vivo is supported by the findings that synthetic antioxidants can prevent or lessen
certain vitamin E–deficient symptoms, and that increased production of peroxidation prod-
Table 5 Vitamin E Activity of Tocopherolsa
Structure gestation (%) Hemolysis (%)
RRR-?-tocopherol 100 100
RRR-?-tocopherol 25–50 15–27
RRR-?-tocopherol 8–19 3–20
RRR-?-tocopherol 0.1–3 0.3–2
RRR-?-tocotrienol 21–50 17–25
RRR-?-tocotrienol 4–5 1–5
RRR-?-tocotrienol — —
RRR-?-tocotrienol — —
RRR-?-tocopheryl acetate 91 —
All-rec-?-tocopherolb 74 —
All-rec-?-tocopheryl acetateb 67 —
a Bioassay methods.
b A mixture of eight steroisomers.
Vitamin E 177
ucts, such as malondialdehyde, ethane, and pentane, are found in vitamin E–depleted
animals (9,87). The suggestion that vitamin E may play a structural role in the control of
membrane permeability and stability (88) does not conflict with this view.
Several non-antioxidant functions of vitamin E have also been suggested. Vitamin
E, for example, has been reported to regulate the de novo synthesis of xanthine oxidase
(89), modulate the activities of microsomal enzymes (90,91), and down-regulate protein
kinase activity and cell proliferation (92,92a). It is interesting to note that RRR-?-tocopherol
prevents the inhibition of protein kinase activity and cell growth of smooth-muscle
cells caused by RRR-?-tocopherol (93). Also, vitamin E may regulate immune response
or cell-mediated immunity by modulating the generation of prostaglandins and lipid peroxidation
products, and the metabolism of arachidonic acid (94–96). Additionally, vitamin
E has been shown to be essential for optimally developing and maintaining the integrity
and function of nervous mechanism and skeletal muscle (12,97). Recently, vitamin E has
been shown to be capable of down-regulate mitochondrial generation of superoxide and
hydrogen peroxide (97a,97b). By reducing mitochondrial generation of superoxide and
related ROS, dietary vitamin E not only attenuates oxidative damage but also modulates
the expression and activation of signal transduction pathways and other redox-sensitive
biological modifiers, and thereby may prevent or delay the onset of degenerative tissue
A. Free-Radical–Induced Peroxidative Damage
The harmful effects resulting from inhalation of high concentration of oxygen have been
attributed to the formation of reactive oxygen species rather than molecular oxygen per se.
Superoxide radicals, nitric oxide, lipid alkoxyl and peroxyl radicals are the most significant
reactive oxygen species generated in systems living in aerobic environments. Among
them, peroxy radical derived from polyunsaturated fatty acids has special significance due
to its involvement in lipid peroxidation, the most common indicator of free-radical processes
in living systems (14).
The process of lipid peroxidation (or autoxidation) can be divided into three phases:
initiation, propagation, and termination. In the initiation phase (reaction 1), carbon-centered
lipid radicals (R•) can be produced by proton abstraction from, or addition to, a
polyunsaturated fatty acid (RH) when a free-radical initiator (I*) is present. The initiation
reaction is generally very slow and is dependent on the type of initiator employed. However,
the reaction can be catalyzed by heat, light, trace metals, and/or certain enzymes
(e.g., lipoxygenases). During the propagation phrase, the lipid radical reacts readily with
available molecular oxygen to form a peroxy radical (ROO•) at a very high rate (reaction
2). The peroxyl radical formed can react with another polyunsaturated fatty acid (R?H),
at a slow rate, to form a hydroperoxide (ROOH) and a new carbon-centered radical (reaction
RH ?>
R• (1)
R•  O2 > ROO• (2)
ROO•  R?H > ROOH  R?• (3)
178 Chow
The propagative process can continue until all polyunsaturated fat is consumed or
the chain reaction is broken (termination phase). Free radicals can also be scavenged and
the chain reaction terminated by self-quenching to form dimers (reaction 4) or by the
action of antioxidant (AH) (reaction 5).
R• (or ROO•)  R?• > R-R? (or ROOR?) (4)
R• (or ROO•)  AH > RH (or ROOH)  A• (5)
Of the free-radical process inhibitors, tocopherols are among the most effective
chain breakers available. This is because tocopherols can react more rapidly with peroxy
radicals than do polyunsaturated fatty acid (14). Since lipid peroxy radicals can react with
tocopherols, several orders of magnitude faster (Ka  104–109 M1 s1) than their reaction
with acyl lipids (Ka  10–60 M1 s1) (14,23), one tocopherol molecule can protect up
to 103–108 polyunsaturated fatty acid molecules at low peroxide levels. Similarly, a small
ratio of ?-tocopherol to polyunsaturated fatty acid molecules (1:1500) in the red cell membrane
is sufficient to interrupt the free-radical chain reactions (98).
Tocopherols can also inhibit the oxidations induced by the electronically excited
singlet oxygen. The relative effectiveness of ?-, ?-, ?-, and ?-tocopherols as singlet oxygen
quenchers is 100:55:26:10, respectively (99,100). In addition, tocopherols can react with
hydroxyl, perhydroxy, and superoxide radicals (101,102), as well as nitric oxide (103,104),
and form tocopheryl quinones, epoxytocopherones, tocored, and/or other oxidation products.
The antioxidant property of various tocopherols is dependent on their stoichiometric
factors, the rate of their reaction with peroxy radicals, and their concentrations. Under
certain conditions, tocopherol may exert pro-oxidant property. The pro-oxidant effect of
?-tocopherol reported in in vitro studies appears to be related to its tocopheroxy radicals
(23). When the concentration of tocopheroxy radical is high, the radicals may react reversibly
with unperoxidized lipids and lipid hydroperoxides by chain transfer and generate
alkyl and peroxy radicals, respectively.
Reactive oxygen species may react with cellular components, with resultant degradation
and/or inactivation of essential cellular constituents (105–109). The products derived
from the reaction with reactive oxygen species may be more or less reactive or harmful.
Cell membranes, which contain a relative high proportion of polyunsaturated fatty acids,
are more susceptible to free-radical–induced lipid peroxidation. The process of free-radical–
induced lipid peroxidation has been implicated as a critical initiating event leading
to cell injury or organ degeneration. There are a large variety of conditions that are capable
of initiating or enhancing oxidative stress within the cellular environment. These conditions
include inadequate intake of antioxidants, excess intake of prooxidants, exposure to
noxious chemical or physical agents, strenuous physical activities, injuries and wounds,
and certain hereditary disorders (87).
B. Antioxidant Systems
In view of the potential adverse effects of oxygen and its reactive intermediates, it is
important that a number of enzymatic and nonenzymatic antioxidant systems are present
in the cell. Under normal conditions, the overall antioxidant activity of the cell is able to
control or prevent most of the adverse effects of oxygen and its reactive intermediates.
However, when the antioxidant potential is weakened or oxidative stress is greatly increased,
irreversible damage to the cell may occur. The susceptibility of a given organ or
Vitamin E 179
Fig. 3 Relationship between vitamin E and other antioxidant systems. LH represents membrane
or polyunsaturated lipids; LOOH, lipid hydroperoxides, LOH, hydroxy acid; LOO•, peroxy radical;
L•, alkyl radical; LO•, alkoxy radical; OH•, hydroxyl radical; •O2, superoxide radical; X•, other radicals;
1O2, singlet oxygen; H2O2, hydrogen peroxide; RCHO, aldehydes; , epoxides; Vit.E•,
vitamin E radical; Vit.C•, ascorbate radical or dehydroascorbate; 6-PG, 6-phosphogluconate; GSH,
reduced glutathione; GSSG, oxidized glutathione; NADH and NAD, reduced and oxidized nicotinamide
adenine dinucleotide, respectively; NADPH and NADP, reduced and oxidized nicotinamide
dinucleotide phosphate, respectively; Se selenium; SOD, superoxide dismutase; GP, GSH peroxidase
or phospholipid hydroperoxide GSH peroxidase; GR, GSSG reductase; G-6-PD, glucose-6-
phosphate dehydrogenase; ADH, aldehyde dehydrogenase; EH, epoxide hydrase; AR, ascorbate or
dehydroascorbate reductase.
180 Chow
Fig. 4 Possible regeneration pathways of ?-tocopherol. LOOH represents lipid hydroperoxides;
LOO•, peroxy radical; GSH, reduced glutathione; GSSG, oxidized glutathione; NADH and NAD,
reduced and oxidized nicotinamide adenine dinucleotide, respectively; NADPH and NADP, reduced
and oxidized nicotinamide dinucleotide phosphate, respectively.
organ system to oxidative damage is determined by the overall balance between the extent
of oxidative stress and antioxidant capability.
Major cellular antioxidant mechanisms include (a) direct interaction with oxidants
or oxidizing agents by ascorbic acid, glutathione (GSH), and other reducing agents; (b)
scavenging of free radicals and singlet oxygen by vitamin E, ascorbic acid, carotenoids,
superoxide dismutase, and other scavengers; (c) reduction of hydroperoxides by glutathione
peroxidases and catalase; (d) binding or removal of transition metals by ferritin, transferrin,
ceruloplasmin, albumin, and other chelators; (e) separation or prevention of reactive
oxygen species and other factors from reaching the specific site of action or reacting with
essential cellular components by membrane barriers; and (f) repair of resulting damage
by dietary nutrients and metabolic activities (Fig. 3). While various antioxidant systems
act at different stages and cellular localization, vitamin E occupies a unique position in
the overall antioxidant defense (87,110,111). The essential role of vitamin E in maintaining
the red cell integrity, for example, is supported by the findings that ascorbic acid and GSH
promote oxidative damage to the vitamin E–deficient red cells, while protecting vitamin
E–supplemented red cells (111). The contradictory role of these reducing agents appears
to be the result of (a) their participation in vitamin E regeneration when certain vitamin
E levels are maintained and (b) their involvement in free-radical generation by virtue of
their maintaining transition metal ions in a reduced state (112,113) when vitamin E is low
or depleted. Approximately 1 part of ?-tocopherol is capable of protecting 1000 parts of
Vitamin E 181
lipid molecule in biological membranes (98,114). The efficacy of tocopherol is augmented
by the presence of other antioxidant systems and the factors involved in its regeneration
(Fig. 4).
It has long been recognized that the function of vitamin E is closely interrelated with other
dietary nutrients. In addition to dietary lipids, vitamin E is functionally related to the status
of polyunsaturated fatty acids, selenium, vitamin C, iron, ?-carotene, and sulfur-containing
amino acids.
A. Lipids
Cell membranes contain a high proportion of polyunsaturated fatty acids, and vitamin E
is found predominantly in the membranes. The dietary requirement for vitamin E is related
to the degree of unsaturation of the fatty acids in tissue lipids, which can be altered by
dietary lipids (115,116). The intake of tocopherols generally parallels the intake of polyunsaturated
fatty acids. Foods that contained high polyunsaturated fatty acids generally are
rich sources of tocopherols. A typical U.S. mixed diet has a ratio of approximately 0.5
mg ?-tocopherol equivalent per gram of polyunsaturated fatty acids (114).
B. Selenium
The functional interrelationships between vitamin E and selenium have long been recognized.
Selenium has been shown to prevent or reduce the severity of several symptoms
of vitamin E deficiency, including necrotic liver degeneration and eosinophilic enteritis
in rats and exudative diathesis in chicks (8,9,73). Selenium deficiency in domestic animals
is frequently associated with an increased vitamin E requirement, and the situation is often
aggravated by the oxidation of vitamin E in feed during storage.
The finding that selenium is an integral part of the enzymes GSH peroxidase (117)
and phospholipid hydroperoxide GSH peroxidase (118) provides a reasonable explanation
of the metabolic interrelationship between vitamin E and selenium. Selenium may complement
the antioxidant function of vitamin E by reducing lipid hydroperoxide (ROOH) to
corresponding hydroxy acid (ROH) via the action of GSH peroxidases (reaction 6). This
reduction prevents the decomposition of hydroperoxides to form free radicals that may
initiate further peroxidation and may thus reduce the requirement of vitamin E.
ROOH  2 GSH ???????>
GSH peroxidase
ROH  GSSG  H2O (6)
C. Vitamin C (Ascorbic Acid)
In addition to being an important water-soluble free-radical scavenger and reducing agent,
vitamin C may complement the function of vitamin E and reduce vitamin E requirement.
This effect is partly attributable to the involvement of vitamin C in the regeneration or
restoration of vitamin E (Fig. 4) after exerting its antioxidant function (119–122). However,
unambiguous experimental evidence for the sparing effect of vitamin C on the requirement
of vitamin E has yet to be provided (123). In addition to ascorbic acid, GSH,
in association with an enzyme or enzyme system(s), seem to be involved in the regeneration
of vitamin E (122,124). Also, a GSH-dependent dehydroascorbate reductase and an
182 Chow
NADH-semidehydroascorbate reductase appear to be involved in the regeneration or restoration
of vitamin C (125). Furthermore, dihydrolipoic acid, NADH-cytochrome b5, and
ubiquinol have been shown to play a role in vitamin E regeneration (127–129), and dihydrolipoic
acid is also involved in the regeneration of ubiquinol, ascorbate, GSH, and thioredoxin
(129). While the nature of the vitamin E regenerative process in vivo is not entirely
clear, the process does provide a rational explanation for the fact that it is very difficult
to deplete adult animals or human subjects of the vitamin (115).
D. -Carotene
Of the more than 500 carotenoids that have been identified, about 50 possess some vitamin
A activity (129). In the leafy green and yellow-orange vegetables, the predominant provitamin
A activity is due to a hydrocarbon carotene, ?-carotene, which makes up over 50%
of the total carotenoids. Other carotenoids, such as lycopene and lutein, are found in the
circulation but cannot be converted to retinol (130). In addition to being a precursor of
vitamin A, ?-carotene is an effective quencher of singlet oxygen and free-radical scavenger
(131–133). As a more effective scavenger of singlet oxygen, ?-carotene may complement
vitamin E in protecting cellular components against oxidative damage. ?-Carotene may
also modulate membrane properties important to cell–cell signaling or possibly serve as
an intracellular precursor to retinoic acid (134). Vitamin E has been shown to enhance
the lymphatic transport of ?-carotene and its conversion to vitamin A in the ferret (135).
On the other hand, excessive intake of ?-carotene has been reported to decrease the plasma
level of vitamin E, and excess vitamin E lowers the levels of ?-carotene (136–138).
E. Iron
Separation and removal of iron and other transition metal ions is of critical importance
in preventing iron-catalyzed generation of hydroxyl radicals from hydrogen peroxide (reaction
8). Hydrogen peroxide is readily formed from superoxide following the action of
superoxide dismutase (SOD) (reaction 7).
  2H ??>
H2O2  O2 (7)
H2O2  Fe2 ??> OH  OH•  Fe3 (8)
Transition metal ions, particularly Fe2 and Cu, can exacerbate membrane damage
by catalyzing the decomposition of lipid hydroperoxides to yield more free radicals (reaction
9) and aggravate lipid peroxidation membrane damage (112,113):
ROOH ??> R•, RO•, ROO•, etc. (9)
Dietary iron exists mainly in the ferric state but is absorbed in the ferrous form.
Iron absorbed from the diet or released from ferritin can be sequestered by transferrin,
lactoferrin, citrate, ATP, and other phosphate esters. However, excess iron absorption
occurs in certain pathological conditions. The genetic disease idiopathic hemochromatosis,
for example, causes a condition similar to iron overload, and the effects of iron overload
can be lessened by increased vitamin E intake (139).
Vitamin E 183
F. Sulfur-Containing Amino Acids
It has long been recognized that muscular dystrophy, the condition that occurs in most
animal species fed a vitamin E–deficient diet, appears in the chick only when fed a diet
that contains low-sulfur amino acids (8,9). This interrelationship of vitamin E with sulfur
amino acids appears to be due to a requirement for sulfur amino acids in the synthesis
of GSH (140). GSH is needed for the activity of GSH peroxidase and for the
restoration/regeneration of vitamin E (Fig. 4). Oxidized glutathione (GSSG), in turn, is
reduced by GSSG reductase utilizing NADPH generated in the pentose shunt pathway
(reaction 10).
NADPH  GSSG ??????>
GSSG reductase
2GSH  NADP (10)
Tocopherols exist mainly in the free-alcohol form and are widely distributed in a variety
of plant life. A notable exception is latex lipids, in which the vast majority of tocopherols
(mainly ?-tocotrienol) are present as esters of fatty acids (10,141). Vegetable oils are the
primary dietary sources of tocopherols in the United States.
While ?-tocopherol predominates in most animal species and is significantly more
active biologically than any other form of tocopherol, ?-tocopherol is the major form
consumed in the United States. ?-Tocopherol in soybean oil, corn oil, and other vegetable
oils, accounts for over half of the estimated total tocopherol intake (Table 6). The signifi-
cant sources of vitamin E in U.S. diets include regular margarines, regular mayonnaise
and salad dressings, vitamin E–fortified breakfast cereals, vegetable shortenings, peanut
butter, eggs, cooking oils, potato chips, whole milk, tomato products, and apples (142).
Sheppard et al. (143) have summarized the tocopherol content and vitamin E activity in
various food products based on recent information.
A. Requirements
In 1968, vitamin E was officially recognized as an essential nutrient, and 30 IU was recommended
for men by the Food and Nutrition Board. Reassessment of dietary fat and vitamin
E content led in 1974 to a revised recommendation of 15 IU. In the 1980 recommended
dietary allowance (RDA), efforts were made to take into account the dietary contribution
of non-?-tocopherols, and the requirement was expressed as tocopherol equivalent (TE).
One TE is equal to 1 mg RRR-?-tocopherol, and the other isomers were converted to TE
by multiplying the 1 mg of each isomer by the relative activity factor (e.g., ?-tocopherol,
0.5; ?-tocopherol, 0.1; ?-tocotrienol, 0.3). The tenth RDA expressed in both TE and IU
is shown in Table 7, and the values ranged from 4.5 IU (three ?-tocopherol equivalents
or 3 mg d-tocopherol) in infants, 6.0–10.4 in children (10 years or under), 14.9 in males
(11 years), 11.9 in females (11 years), 14.9 in pregnant women, 16.4 to 17.9 in lactating
women (144). The normal intake of vitamin E in U.S. diets ranges between 4 and 33
TE daily in adults not taking vitamin E supplements, with average values between 11 and
13 TE. This level of intake results in average plasma/serum levels in adults of approximately
9.5 ug ?-tocopherol/mL (9).
184 Chow
?-T ?-T-3 ?-T ?-T-3
Vitamin E 185
Table 7 U.S. Recommended Daily Dietary Allowance for
Vitamin E
International Tocopherol
Age (years) units equivalents
Infants 0–0.5 4.5 3.0
Children 0.5–1.0 6.0 4.0
1–3 8.9 6.0
4–10 10.4 7.0
Males 11 14.9 10.0
Females 11 11.9 8.0
Pregnant 14.9 10.0
Lactating, first
6 mo. 17.9 12.0
Lactating, second
6 mo. 16.4 11.0
Vitamin E deficiency is rarely seen in the population of the United States. When it
occurs, it is usually a result of lipoprotein deficiencies or lipid malabsorption syndromes
(11,12,26). A high intake of polyunsaturated fat may increase the vitamin E requirement.
Increased levels have also been recommended to provide protection against oxidants and
oxidizing agents in the environment. In addition, recent information suggests that a higher
intake of vitamin E is associated with enhanced immune response and reduced risk of
certain chronic diseases, particularly coronary heart disease (see Sec. X).
B. Assessments
The nutritional status of vitamin E is assessed based on the concentration of tocopherols
in body store and its physiological functions. Currently, several indices, such as tocopherol
concentrations in plasma/serum erythrocytes, platelets, or tissues; degree of erythrocyte
hemolysis, and amounts of lipid peroxidation products (e.g., ethane, pentane, and malondialdehyde)
generated, have been used to assess the nutritional status of vitamin E in humans.
Measurement of tocopherol concentrations in plasma/serum is the most commonly
used approach. The tocopherol concentration in erythrocytes is an indicator of longer term
status for vitamin E than that in plasma/serum. Platelets are more sensitive for measurement
of dose response to dietary vitamin E when compared with plasma, erythrocytes, or
lymphocytes (145). Also, platelet tocopherol concentrations are independent of serum lipid
levels (146), an important advantage relative to serum or plasma tocopherol concentrations.
However, none of the above accurately reflects both dietary intake and body stores
of vitamin E. Measurement of ?-tocopherol tissue store, such as adipose tissue, is a reliable
way to assess long-term vitamin E status in humans. However, tissue biopsy is highly
invasive and therefore impractical for application to the general public. Measurement of
tocopherol concentrations in the relatively noninvasive and easily collectible skin and
buccal mucosal cells has shown that the nutritional status of vitamin E, and the concentrations
of ?- and ?-tocopherol are significantly correlated with that of the plasma (147).
The significant plasma–tissue relationships suggest that plasma tocopherol concentrations,
in lieu of their target tissue concentrations, can be used as a reliable indicator of vitamin
E status.
186 Chow
Long-term intake of vitamin E is difficult to assess accurately because diets vary
substantially over time. Also, since tocopherols are widely distributed in foods and products
made from them, their consumption is difficult to quantify based on a dietary history.
Potential sources of error in the use of interviews for obtaining food/nutrient intake information
are as follows: (a) only recent intake is estimated, whereas earlier dietary intake
may be very different; (b) a limited number of foods were analyzed; and (c) errors may
have occurred in subject estimation of food intake frequency and size of portion eaten.
The accuracy of vitamin E intake information obtained through the use of frequency
questionnaires also depends on the reliability of information in the food composition data
used. The tocopherol content of foods is highly variable, depending genetic, seasonal,
processing, storage, and other factors. Also, the tocopherol content varies greatly among
foods, and the tocopherol contents of various foods were obtained using colorimetry procedure
rather than HPLC. In addition, food processing, storage, and culinary practices can
significantly influence the destruction of tocopherols present in the food. Furthermore, the
possible interaction of vitamin E and other nutrients, such as polyunsaturated lipids, ?-
carotene, selenium, sulfur amino acids, and ascorbic acid, further complicates evaluation
of vitamin E status.
A. Immune Response
The immune response generally declines with age, and suboptimal immune responses in
the elderly may be responsible for the increased risk of infections and immune-mediated
diseases (148). Therefore, maintaining proper immune function is important for the elderly.
Several indices of immune response, including responses on delayed-type hypersensitivity
skin tests, antibody production, lymphocyte proliferation, cytokine production, and
counts of the specific subgroups of white blood cells, are influenced by the status of essential
nutrients, including vitamin E. Vitamin E supplementation is associated with enhanced
production of the cytokine interleukin-2 (IL-2), enhanced lymphocyte proliferation, and
decreased production of the immunosuppressive prostaglandin E2 (149). The antioxidant
property is likely to play a role in the immunostimulatory effects of vitamin E for the
immune response. For example, vitamin E supplementation (60, 200, or 800 IU/day) for
a month or longer resulted in significant increases in delayed-type hypersensitivity skin test
response, and T-cell subpopulations and proliferative responses, as well as IL-2 activity in
the healthy, free-living elderly (150,151,151a). These findings suggest that current dietary
vitamin E intake levels of 10–15 IU/day are insufficient to attain optimal immune responses
in the general population. However, the 800 IU/day supplement group showed
lower immune responses on some tests than the 200 IU/day group, and was comparable
with the 60 IU/day or the placebo group, suggesting that more is not necessarily better
B. Cardiovascular Disease
Cardiovascular disease is the leading cause of morbidity and mortality in industrialized
countries. Several dietary factors, including decreased vitamin E status, have been implicated
in the incidence of coronary heart disease. In recent years, several large-scale obserVitamin
E 187
vational and experimental studies have shown that higher intake or serum/plasma levels
of ?-tocopherol are associated with a decreased risk of cardiovascular diseases. For example,
in a study that included 87,245 female nurses and 39,910 male health professionals, the
age-adjusted relative risk of coronary heart disease was found to be significantly lowered in
the highest quintile of vitamin E intake compared with the lowest quintile (132,153). The
apparent benefit was attributable mainly to subjects consuming 100 IU vitamin E or more
daily for 2 years or longer. In another study, which included 21,809 postmenopausal
women, higher food-derived vitamin E intake was associated with decreased death rate
due to coronary heart disease (154). Results obtained from intervention studies generally
corroborated with the epidemiological data. In the Cambridge Heart Antioxidant Study
(155), for example, patients with angiographically proven coronary atherosclerosis receiving
400 or 800 IU of vitamin E for 510 days had a significant reduction of the rate of
major cardiovascular events (47%) and of nonfatal myocardial infarction (77%), in comparison
with those receiving placebo. However, cardiovascular deaths were not signifi-
cantly different between groups.
While the information available suggests that increased intake of vitamin E is associated
with reduced risk of cardiovascular disease, the mechanism by which vitamin E reduces
the risk of cardiovascular disease is not clear. It is possible that the ability of vitamin
E to prevent LDL oxidation (156,157) and/or to reduce platelet adhesion (158,159) may
be responsible. Evidence indicates that enhanced uptake of oxidatively modified LDL by
macrophages through a specific receptor gives rise to lipid-laden foam cells, one of the
earliest stages in the development of atherosclerosis (160–162). Oxidized LDL is toxic
to cells and may be responsible for damage to the endothelial layer and destruction of
smooth-muscle cells (160,162).
C. Cancer
Carcinogenesis is a multiple process, and a large number of chemical and physical agents
can affect the development of cancer in numerous ways. Vitamin E may protect against
cancer development by reacting directly with mutagens/carcinogens, altering metabolic
activation, enhancing the immune system, inhibiting cell proliferation, or other mechanisms.
A large number of observational and experimental studies have been performed
to determine the relationship between vitamin E status and cancer risk (163,164). Also, a
number of well-designed large-scale chemoprevention trials for antioxidant micronutrients
have been conducted in recent years (165–167). However, only a few of those intervention
studies are designed specifically to examine the efficacy of vitamin E. Most of the studies
dealing with head/neck, lung, and colorectal cancer suggest a protective effect of vitamin
E against cancer risk, though reports dealing with the association between vitamin E status
and risk of breast, bladder, and cervical cancer are less consistent (168). Nevertheless,
the majority of studies suggest that low vitamin E status is associated with increased risk
of certain cancer.
D. Neurodegenerative Diseases
Neurodegenerative diseases, such as dementia of Alzheimer type and Parkinson’s disease,
are characterized neurochemically by a transmitter-specific loss of neurons, which progresses
and extends to several neuronal systems over the course of the disease. While the
188 Chow
pathophysiology of the heterogenous disorder is largely unknown, free radical-induced
oxidative damage may play a role in the cellular events leading to the degeneration of
neurons in the brain. In patients with Alzheimer disease, for example, amyloid beta-peptide
accumulates in plaques of the brain. The compound is neurotoxic by a mechanism involving
induction of reactive oxygen species. The concentration of ?-tocopherol is lower in
cerebrospinal fluid from patients with Alzheimer type dementia that the control (169).
Also, the concentrations of basal and iron-ascorbate stimulated lipid peroxidation products
are higher in the inferior temporal cortex of Alzheimer subjects than the controls (170).
These reports suggest that the brain of patients with Alzheimer disease is associated with
inadequate antioxidant status and/or increased oxidative stress. In a double-blind,
multicenter trial, Sano et al. (171) evaluated the effect of ?-tocopherol (2,000 IU daily)
on primary outcome in 341 patients with Alzheimer’s disease of moderate severity over
a 2 year period and found that treatment with ?-tocopherol was beneficial in delaying the
primary outcome of disease progression (death, severe dementia, loss of the ability to
perform basic activities of daily living or the need for institutionalization). The degree of
functionality of the mutant ?-tocopherol transfer protein in patients with familial vitamin
E deficiency are associated with the degree of severity of the neurological damage and
age of onset as well as with plasma vitamin E concentration (59). High doses of vitamin
E can prevent or mitigate the neurological course of this disease. Plasma vitamin E concentrations
in patients increase when they are treated with large doses of vitamin E, presumably
because of the direct transfer of tocopherol from chylomicrons to other circulating
The essentiality of vitamin E for humans had been questioned after the vitamin was identified
as the fat-soluble substance necessary for reproduction in rats more than 75 years
ago. This is because a number of tissue-specific deficiency symptoms can be easily produced
in experimental animals but not in humans. The essentiality of vitamin E for humans
was only recognized in the late 1960s. Subsequent studies of patients with certain
neurological disorders have clearly demonstrated the need of the vitamin in the development
and maintenance of the integrity and function of the nervous mechanism and skeletal
Recent studies also provide a much better understanding of the mechanisms of tocopherol
transport and selective retention of certain forms or isomers of tocopherols. The
finding that tocopherol can be regenerated or restored in association with GSH, ascorbic
acid, and other systems provides a feasible explanation as to why it is difficult to deplete
vitamin E in human adults. Information is available that supports the view that the function
of vitamin E is interrelated with other antioxidant systems and that ascorbic acid, ?-
carotene, sulfur amino acids, and selenium complement the antioxidant function of vitamin
In recent years, public interest in vitamin E has increased drastically. This is mainly
promoted by experimental findings which suggest that vitamin E may protect against oxidative
stress resulting from exposure to environmental agents, and that increased intake
of the vitamin is associated with a reduced risk of cardiovascular disease, cancer, and
other degenerative diseases, as well as with enhanced immune response. However, more
research is needed to establish more definitively the relationship between vitamin E intake
and the risk of degenerative diseases, and to elucidate further the mechanisms by which
Vitamin E 189
vitamin E is involved in altering disease processes. In addition, the optimal requirement
of various population groups for vitamin E against oxidative stress has yet to be established.
1. H. M. Evans and K. S. Bishop, On the existence of a hitherto unrecognized dietary factor
essential for reproduction. Science 56:650 (1922).
2. B. Sure, Dietary requirements for reproduction. II. The existence of a specific vitamin for
reproduction. J. Biol. Chem. 58:693 (1924).
3. H. M. Evans and G. O. Burr, The anti-sterility vitamin fat soluble E. Proc. Natl. Acad. Sci.
USA 11:334 (1925).
4. H. M. Evans, O. H. Emerson, and G. A. Emerson, The isolation from wheat germ oil of an
alcohol, alpha tocopherol, having the properties of vitamin E. J. Biol. Chem. 113:319
5. P. von Karrer, H. Fritzsch, B. H. Ringier, and H. Salmon, Alpha-tocopherol. Helv. Chim.
Acta 21:520 (1938).
6. E. Fernholz, On the constitution of alpha-tocopherol. J. Am. Chem. Soc. 60:700 (1938).
7. H. S. Olcott and O. H. Emerson, Antioxidant properties of the tocopherols. J. Am. Chem.
Soc. 59:1008 (1937).
8. M. L. Scott, Studies on vitamin E and related factors in nutrition and metabolism, in The
Fat-Soluble Vitamins (H. F. DeLuca and J. W. Suttie, eds.), University of Wisconsin Press,
Madison, 1969, pp. 355–368.
9. L. J. Machlin, Vitamin E, in Handbook of Vitamins, 2nd ed. (L. J. Machlin, ed.), Second
ed., Marcel Dekker, New York, 1991, pp. 99–144.
10. C. K. Chow, Vitamin E and blood. World Rev. Nutr. Dietet. 45:133 (1985).
11. J. G. Bieri and P. M. Farrell, Vitamin E. Vitam. Hormones 34:31 (1976).
12. R. J. Sokol, Vitamin E deficiency and neurological disease. Annu. Rev. Nutr. 8:351 (1988).
13. A. Kamal-Eldin and L.-A. Appelqvist, The chemistry and antioxidant properties of tocopherols
and tocotrienols. Lipids 31:671 (1996).
14. G. W. Burton and M. G. Traber. Vitamin E: antioxidant activity, biokinetics, and bioavailability.
Annu. Rev. Nutr. 10:357 (1990).
14a. T. S. Shin and J. S. Godber, Isolation of four tocopherols and four tocotrienols from a variety
of natural sources by semi-preparative high performance liquid chromatography. J. Chromatogr.
678:48 (1994).
15. T. Doba, G. W. Burton, and K. U. Ingold, EPR spectra of some ?-tocopherol model compounds.
Polar and conformational effects and their relation to antioxidant activities. J. Am.
Chem. Soc. 105:6505 (1983).
16. M. Matsuo, S. Matsumoto, and T. Ozawa, Electron spin resonance spectra and hyperfine
coupling constants of the tocopheroxy and 2,2,5,7,8-pentamethylchroman-6-oxyl radicals derived
from vitamin E and its model and deuterated model compounds. Org. Magnet. Reson.
21:261 (1983).
17. C. Suarna, M. Baca, and P. T. Southwell-Keely, Oxidation of the ?-tocopherol model compound
2,2,5,7,8-pentamethyl-6-chromanol in the presence of alcohol. Lipids 27:447
18. J. L. G. Nillson, D. G. Doyle, and K. Folkers, New tocopherol dimers. Acta Chem. Scand.
22:200 (1968).
18a. D. R. Nelan and C. D. Robeson, The oxidation product from ?-tocopherol and potassium
ferricyanide and its reaction with ascorbic and hydrochloric acids. J. Am. Chem. Soc. 84:
2963 (1962).
19. R. Yamaguchi, T. Matsui, K. Kato, and K. Ueno, Reaction of ?-tocopherol with 2,2?-azobis
(2,4-dimethyl-valeronitrile) in benzene. Agric. Biol. Chem. 53:3257 (1989).
190 Chow
20. O. Igarashi, M. Hagino, and C. Inagaki, Decomposition of ?-tocopherol spirodimer by alkaline
saponification. J. Nutr. Sci. Vitaminol. 19:469 (1973).
21. D. C. Liebler, K. L. Kaysen, and T. A. Kennedy, Redox cycles of vitamin E: hydrolysis and
ascorbic acid dependent reduction of 8a-(alkyldioxy) tocopherones. Biochemistry 28:9772
22. C. K. Chow and H. H. Draper, Isolation of ?-tocotrienol dimers from Hevea latex. Biochemistry
9:445 (1970).
23. J. Pokprny, Major factors affecting the autoxidation of lipids, in Autoxidation of Unsaturated
Lipids (H. W. S., Chan, ed.), Academic Press, London, 1987, pp. 141–206.
24. E. A. Serbinova, V. E. Kagan, D. Han, and L. Packer, Free radical recycling and intramembrane
mobility in the antioxidant properties of alpha-tocopherol and alpha-tocotrienol. Free
Rad. Biol. Med. 10:263 (1991).
25. H. S. Olcott and J. Van der Veen, Comparative antioxidant activities of toco and its methyl
derivatives. Lipids 3:331 (1968).
26. H. J. Kayden and M. G. Traber, Absorption, lipoprotein transport, and regulation of plasma
concentrations of vitamin E in humans. J. Lipid Res. 34:343 (1993).
27. C. A. Drevon, Absorption, transport and metabolism of vitamin E. Free Rad. Res. Commun.
14:229 (1991).
28. P. M. Mathias, J. T. Harries, T. J. Peters, and D. P. R. Muller, Studies on the in vivo absorption
of micellar solution of tocopherol and tocopheryl acetate in the rat. J. Lipid Res. 22:
829 (1981).
29. D. Hollander, E. Rim, and K. S. Muralidhara, Mechanism and site of small intestinal absorption
of ?-tocopherol in the rat. Gastroenterology 68:1492 (1975).
30. D. Hollander, Intestinal absorption of vitamins A, E, D, and K. J. Lab. Clin. Med. 97:449
31. M. K. Horwitt, Interrelationship between vitamin E and polyunsaturated fatty acids in adult
men. Vitam. Hormones 20:541 (1962).
32. A. Bjorneboe, G.-E. Soyland, A. Bjorneboe, G. Rajka, and C. A. Drevon, effect of dietary
supplementation with eicosapentaenoic acid in the treatment of atopic dermatitis. Br. J. Dermatol.
117:463 (1987).
33. I. R. Peake, H. G. Windmueller, and J. G. Bieri, A comparison of the intestinal absorption,
lymph and plasma transport, and tissue uptake of ?- and ?-tocopherols in the rat. Biochim.
Biophys. Acta 889:310 (1972).
34. M. G. Traber, H. J. Kayden, J. B. Green, and M. H. Green, Absorption of water-miscible
forms of vitamin E in a patient with cholestasis and in thoracic duct-cannulated rats. Am. J.
Clin. Nutr. 44:914 (1986).
35. M. G. Traber and H. J. Kayden, Preferential incorporation of ?-tocopherol vs ?-tocopherol
in human lipoproteins. Am. J. Clin. Nutr. 49:517 (1989).
36. M. G. Traber, G. W. Burton, K. U. Ingold, and H. J. Kayden, RRR- and SRR-?-tocopherols
are secreted without discrimination in human chylomicrons, but RRR-?-tocopherol is preferentially
secreted in very low density lipoproteins. J. Lipid Res. 31:675 (1990).
37. M. Traber, L. Rudel, G. Burton, L. Hughes, K. Ingold, and H. Kayden, Nascent VLDL
from liver perfusions of cynomolgus monkeys are preferentially enriched in RRR- compared
with SRR-?-tocopherol. Studies using deuterated tocopherols. J. Lipid Res. 31:687
38. M. G. Traber, J. C. Lane, N. Lagmay, and H. J. Kayden, Studies on the transfer of tocopherol
between lipoproteins. Biochim. Biophys. Acta 793:387 (1984).
39. W. A. Behrens, J. N. Thompson, and R. Madere, Distribution of ?-tocopherol in human
plasma lipoproteins. Am. J. Clin. Nutr. 35:691 (1982).
40. E. C. McMormick, D. G. Cornwell, and J. B. Brown, Studies on the distribution of tocopherol
in human serum lipoproteins. J. Lipid Res. 1:221 (1960).
41. G. M. Kostner, K. Oettl, M. Jauhiainen, C. Ehnholm, and H. Esterbauer, Human plasma
Vitamin E 191
phospholipid transfer protein accelerates exchange/transfer of alpha-tocopherol between
lipoproteins and cells. Biochem. J. 305:659 (1995).
42. C. H. Floren and A. Nilson, Binding, interiorization and degradation of cholesteryl ester–
labeled chylomicron-remnant particles by rat hepatocyte monolayers. Biochem. J. 168:483
43. B. C. Sherrill, T. L. Innerarity, and R. W. Mahley, Rapid hepatic clearance of the canine
lipoproteins containing only the E apoprotein by a high affinity receptor. J. Biol. Chem. 255:
1804 (1980).
44. W. Cohn, M. A. Goss-Sampson, H. Grun, and D. P. Muller, Plasma clearance and net uptake
of alpha-tocopherol and low-density lipoprotein by tissue in WHHL and control rabbits.
Biochem. J. 287:247 (1992).
45. L. J. Machlin and E. Gabriel, Kinetics of tissue ?-tocopherol uptake and depletion following
administration of high levels of vitamin E. Ann. NY Acad. Sci. 393:48 (1982).
46. M. G. Traber and H. J. Kayden, Tocopherol distribution and intracellular localization in
human adipose tissue. Am. J. Clin. Nutr. 46:488 (1987).
47. A. Bjorneboe, G.-E. Aa. Bjorneboe, E. Bodd, B. F. Hagen, N. Kveseth, and C. A. Drevon,
Transport and distribution of ?-tocopherol in lymph, serum and liver cells in rats. Biochim.
Biophys. Acta 889:310 (1986).
48. P. J. Hornsby and J. F. Crivello, The role of lipid peroxidation and biological antioxidants
in the function of adrenal cortex. Mol. Cell Endocrinol. 30:123 (1983).
49. A. Bjorneboe, M. S. Nenseter, B. F. Hagen, G.-E. A. Bjorneboe, K. Prydz, and C. A. Drevon,
Effect of dietary deficiency and supplementation with all-rac-?-tocopherol in hepatic content
in rat. J. Nutr. 121:1208 (1991).
50. B. F. Hagen, A. Bjorneboe, G.-E., Aa. Bjorneboe, and C. A. Drevon, Effect of chronic ethanol
consumption on the content of ?-tocopherol in subcellular fractions of rat liver. Alcoholism
(NY) 13:246 (1989).
51. C. K. Chow, Distribution of tocopherols in human plasma and red blood cells. Am. J. Clin.
Nutr. 28:756 (1975).
52. W. A. Behrens and R. Madere, Transfer of ?-tocopherol to microsomes mediated by a partially
purified liver ?-tocopherol binding protein. Nutr. Res. 2:611 (1982).
53. N. Kaplowitz, H. Yoshida, J. Kuhlenkamp, B. Slitsky, I. Ren, and A. Stolz, Tocopherol
binding proteins of hepatic cytosol. Ann. NY Acad. Sci. 570:85 (1989).
54. G. L. Catignani and J. G. Bieri, Rat liver ?-tocopherol binding protein. Biochim. Biophys.
Acta 497:349 (1977).
55. M. G. Traber, Determinations of plasma vitamin E concentrations. Free Rad. Biol. Med. 16:
229 (1994).
56. M. G. Traber, R. J. Sokol, A. Kohlschutter, and H. J. Kayden, Impaired discrimination between
stereoisomers of ?-tocopherol in patients with familial isolated vitamin E deficiency.
J. Lipid Res. 34:201 (1993).
57. A. Kohlschutter, C. Hubner, W. Jansen, and S. G. Lindner, A treatable familial neuromyopathy
with vitamin E deficiency, normal absorption, and evidence of increased consumption
of vitamin E. J. Inher. Metab. Dis. 11:149 (1988).
57a. M. Arita, Y. Sato, A. Miyata, T. Tanabe, E. Takahashi, H.J. Kayden, H. Arai, and K. Inoue.
Human ?-tocopherol transfer protein: cDNA cloning, expression and chromosomal localization.
Biochem. J. 306: 437 (1995).
58. A. Hentati, H. X. Deng, W. Y. Hung, M. Nayer, M. G. Ahmed, X. He, R. Tim, D. A. Stumpf,
and T. Siddique, Human alpha-tocopherol transfer protein: gene structure and mutations in
familial vitamin E deficiency. Ann. Neurol. 39:295 (1996).
58a. K. Quahchi, M. Arita, H.J. Kayden, F. Hentati, M.B. Hamida, R. Sokol, H. Arai, K. Inoue,
J.L. Mandel, and M. Koenig. Ataxia with isolated vitamin E deficiency is caused by mutations
in the ?-tocopherol transfer protein. Nature Genet. 9: 141 (1995).
59. T. Gotoda, M. Arita, H. Arai, K. Inoue, T. Yokota, Y. Fukuo, Y. Yazaki, and N. Yamada,
192 Chow
Adult-onset spinocerebellar dysfunction caused by a mutation in the gene for the alphatocopherol-
transfer protein. N. Engl. J. Med. 333:1313 (1995).
60. H. Yoshida, M. Yusin, I. Ren, J. Kuhlenkamp, T. Hirano, A. Soltz, and N. Kaplowitz, Identi-
fication, purification and immunochemical characterization of a tocopherol-binding protein
in rat liver cytosol. J. Lipid Res. 33:343 (1992).
61. Y. Sato, K. Hagiwara, H. Arai, and K. Inoue, Purification and characterization of the ?-
tocopherol transfer protein from rat liver. FEBS Lett. 288:41 (1991).
62. Y. Sato, A. Arai, A. Miyata, S. tokita, K. Yamamoto, and T. Tanabe, Primary structure of
alpha-tocopherol transfer protein from rat liver. Homology with cellular retinaldehyde-binding
protein. J. Biol. Chem. 268:17705 (1993).
63. Kuhlenkamp, J., M. Ronk, M. Yusin, A. Stolz, and N. Kaplowitz, Identification and purification
of a human liver cytosolic tocopherol binding protein. Protein Expression Purif. 4:382
64. A. K. Dutta-Roy, M. J. Gordon, D. J. Leishman, B. J. Paterson, G. G. Duthie, and W. P.
James, Purification and partial characterization of an alpha-tocopherol-binding protein from
rabbit heart cytosol. Mol. Cell. Biochem. 123:139 (1993).
65. A. K. Dutta-Roy, D. J. Leishman, M. J. Gordon, F. M. Campbell, and G. G. Duthie, Identifi-
cation of a low molecular weight mass (14.2 kDa) ?-tocopherol-binding protein in the cytosol
of rat liver and heart. Biochem. Biophys. Res. Commun. 196:1108 (1993).
66. M. J. Gordon, F. M. Campbell, G. G. Duthie, and A. K. Dutta-Roy, Characterization of a
novel alpha-tocopherol-binding protein from bovine heart cystol. Arch. Biochem. Biophys.
318:140 (1995).
67. A. S. Csallany, H. H. Draper, and S. N. Shah, Conversion of d-?-tocopherol-C14 to tocopheryl-
p-quinone in vivo. Arch. Biochem. Biophys. 98:142 (1962).
68. C. K. Chow, H. H. Draper, A. S. Csallany, and M. Chiu, The metabolism of C14-?-tocopheryl
quinone and C14-?-tocopheryl hydroquinone. Lipids 2:390 (1967).
69. T. Hayashi, A. Kanetoshi, M. Nakamura, M. Tamura, and H. Shirahama, Reduction of alphatocopherolquinone
to alpha-tocopherolhydroquinone in rat hepatocytes. Biochem. Pharmacol.
44:489 (1992).
70. E. J. Simon, A. Eisengart, L. Sundheim, and A. T. Milhorat, The metabolism of vitamin E.
II. Purification and characterization of urinary metabolites of ?-tocopherol. J. Biol. Chem.
221:807 (1956).
71. A. S. Csallany and H. H. Draper, Dimerization of ?-tocopherol in vivo. Arch. Biochem.
Biophys. 100:335 (1963).
72. H. H. Draper, A. H. Csallany, and M. Chiu, Isolation of a trimer of ?-tocopherol from mammalian
liver. Lipids 2:47 (1967).
72a. N. Schltz, M. Leist, M. Petrzika, B. Gassmann, and R. Breigelius-Flohe. Novel urinary metabolite
of ?-tocopherol, 2,5,7,8-tetramethyl-2(2?-carboxyethyl)-6-hydroxyxhroman, as an
indicator of an adequate vitamin E supply? Am. J. Clin. Nutr. 62: 1527S (1995).
72b. W.J. Wechter, D. Kantoci, E.D. Murry, Jr., D.C. D’Amico, M.E. Jung, and W.-H. Wang.
A new endogenous natriuretic factor: LLU-?. Proc. Natl. Acad. Sci. USA. 93: 6002
72c. S. Chiku, K. Hamamura, and T. Nakamura. Novel urinary metabolite of ?-tocopherol in rats.
J. Lipid Res. 25: 40 (1984).
73. C. B. Hong and C. K. Chow, Induction of eosinophilic enteritis and eosinophils in rats by
vitamin E and selenium deficiency. Exp. Mol. Pathol. 48:182 (1988).
74. J. Harries and D. Muller, Absorption of vitamin E in children with biliary obstruction. Gut
12:579 (1971).
75. C. F. Nockels, D. L. Menge, and E. W. Kienholz, Effect of excessive dietary vitamin E in
the chick. Poult. Sci. 55:649 (1976).
76. B. E. March, E. Wong, L. Seier, L. Sim, and L, Biely, Hypervitaminosis E in the chick. J.
Nutr. 103:371 (1973).
Vitamin E 193
77. D. K. Nichols, M. J. Wolff, L. G. Philips, Jr., and R. J. Montali, Coagulopathy in pinkbacked
pelicans (Pelecanus rufescens) associated with hypervitaminosis E. J. Zoo Wildl.
Med. 20:57 (1989).
78. A. Bendich and L. J. Machlin, The safety of oral intake of vitamin E: data from clinical
studies from 1986 to 1991, in Vitamin E in Health and Disease (L. Packer and J. Fuchs),
Marcel Dekker, New York, 1993, pp. 411–416.
78a. J.M. Kim and R.H. White. Effect of vitamin E on the anticoagulant response to warfarin.
Am. J. Cardiol. 77: 545–546 (1996).
79. L. J. Machlin, E. Gabriel, and M. Brin, Biopotency of ?-tocopherol by curative myopathy
bioassay in the rat. J. Nutr. 112:1437 (1982).
79a. H. Weisser and M. Vecchi, Stereoisomers of ?-tocopheryl acetate. II. Biopotencies of eight
stereoisomers, individually or in mixtures, as determined by rat resportion-gestation test. Int.
J. Nutr. Res. 52:351 (1982).
80. C. K. Chow, Increased activity of pyruvate kinase in plasma of vitamin E–deficient rats. J.
Nutr. 105:1221 (1975).
81. C. K. Chow, Effect of dietary vitamin E and selenium on rats: pyruvate kinase, glutathione
peroxidase and oxidative damage. Nutr. Res. 10:183 (1990).
82. P. K. S. Liu, E. A. Barnard, and P. J. Barnard, Blood plasma pyruvate kinase as a marker
of muscular dystrophy. Properties in dystrophic chickens and hamsters, Exp. Neurol. 67:581
83. M. C. Alberts and F. J. Samaha, Serum pyruvate kinase in muscle disease and carrier states.
Neurology 24:462 (1974).
84. I. M. Weinstock, J. R. Behrendt, H. E. Wiltshire, Jr., J. Keleman, and S. Louis, Pyruvate
kinase diagnostic value in neuromuscular disease. Clin. Chim. Acta 80:415 91977).
85. M. Zata, L. J. Shapiro, D. S. Campion, E. Oda, and M. M. Kaback. Serum pyruvate kinase
(PK) and creatinine phosphokinase (CPK) in progressive muscular dystrophies. J. Neurol.
Soc. 36:349 (1978).
86. A. L. Tappel, Vitamin E and free radical peroxidation of lipids. Annu. NY Acad. Sci. 203:
12 (1972).
87. C. K. Chow, Vitamin E and oxidative stress. Free Rad. Biol. Med. 11:215 (1991).
88. J. A. Lucy, Functional and structural aspects of biological membranes: a suggested structural
role of vitamin E in the control of membrane permeability and stability. Ann. New York
Acad. Sci. 203:4 (1972).
89. G. L. Catignani, F. Chytil, and W. J. Darby, Vitamin E deficiency: immunochemical evidence
for increased accumulation of liver xanthine oxidase. Proc. Natl. Acad. Sci. USA 71:1966
90. C. K. Chow and C. Gairola, Influence of dietary vitamin E and selenium on metabolic activation
of chemicals to mutagens. J. Agric. Food Chem. 32:443 (1984).
91. J. Chen, M. P. Goetchius, T. C. Campbell, and G. F. Combs, Jr. Effects of dietary selenium
and vitamin E on hepatic mixed function oxidase activities and in vivo covalent binding of
aflatoxin B1 in rats. J. Nutr. 112:324 (1982).
92. C. W. Mahoney and A. Azzi, Vitamin E inhibits protein kinase C activity. Biochem. Biophys.
Res. Commun. 154:694 (1988).
92a. D. Boscoboinik, A. Szewczyk, C. Hensey, and A. Azzi, Inhibition of cell proliferation by
alpha-tocopherol. Role of protein kinase C. J. Biol. Chem. 226:6188 (1991).
93. A. Tasinato, D. Boscoboinik, G. M., Bartoli, P. Maroni, and A. Azzi, d-Alpha-tocopherol
inhibition of vascular smooth muscle cell proliferation occurs at physiological concentrations,
correlates with protein kinase c inhibition, and is independent of its antioxidant properties.
Proc. Nat. Acad. Sci. USA 92:12190 (1995).
94. S. N. Meydani, Micronutrients and immune function in the elderly. Ann. NY Acad. Sci. 587:
197 (1990).
95. C. E. Douglas, A. C. Chan, and P. C. Choy, Vitamin E inhibits platelet phospholipase A2—
194 Chow
protection of membranes from lipid peroxidative damage. Biochim. Biophys. Acta 876:639
96. E. J. Goetzel, Vitamin E modulates the lipoxygenation of arachidonic acid in leukocytes.
Nature 288:193 (1981).
97. C. J. MacEvilly and D. P. R. Muller, Lipid peroxidation in neural tissues and fractions from
vitamin E–deficient rats. Free Rad. Biol. Med. 20:639 (1996).
97a. C.K. Chow, W. Ibrahim, W. Wei, and A.C. Chan. Vitamin E regulates mitochondrial hydrogen
peroxide generation. Free Rad. Biol. Med. 27: 580 (1999).
97b. A. Lass and R.S. Sohal. Effect of coenzyme Q(10) and alpha-tocopherol content of mitochondria
on the production of superoxide anion radicals. FASEB J. 14: 87 (2000).
98. A. T. Diplock and J. Lucy, The biochemical modes of action of vitamin E and selenium: a
hypothesis. FEBS Lett. 29:205 (1973).
99. G. W. Grams and K. Eskins, Dye-sensitized photoxidation of tocopherols: correlation between
singlet oxygen reactivity and vitamin E activity. Biochemistry 11:606 (1972).
100. C. S. Foote, Quenching of singlet oxygen, in Singlet oxygen (H. H. Wasserman and R. W.
Muray, eds.), Academic Press, New York, 1979, pp. 139–171.
101. K. Fukuzawa and J. M. Gebicki, Oxidation of ?-tocopherol in micelles and liposomes by
hydroxyl, perhydroxyl and superoxide free radicals. Arch. Biochem. Biophys. 226:242
102. M. Nishikimi, H. Yamada, and K. Yagi. Oxidation by superoxide of tocopherols dispersed
in aqueous media with deoxycholate. Biochim. Biophys. Acta 627:101 (1980).
103. D. Lathia and A. Blum, Role of vitamin E as nitrite scavenger and N-nitrosamine inhibitor:
a review. Int. J. Vitam. Nutr. Res. 59:430 (1989).
104. R. V. Cooney, A. A. Franke, P. J. Harwood, V. Hatch-Pigott, L. J. Custer, and L. J. Mordan,
?-Tocopherol detoxification of nitrogen dioxide: superiority to ?-tocopherol. Proc. Natl.
Acad. Sci. USA 90:1771 (1993).
105. B. Saltman, Oxidative stress: a radical view. Semin. Hematol. 26:249 (1989).
106. B. Halliwell, Oxidants and human disease: some new concepts. FASEB J. 1:358 (1987).
107. B. A. Freeman and J. D. Crapo, Biology of disease: free radicals and tissue injury. Lab.
Invest. 47:412 (1982).
108. B. P. Yu, Cellular defense against damage from reactive oxygen species. Physiol. Rev. 74:
139 (1994).
109. H. Sies, Biochemistry of oxidative stress. Angew. Chem. Int. Ed. Engl. 25:1058 (1986).
110. C. K. Chow, Nutritional influence in cellular antioxidant defense systems. Am. J. Clin. Nutr.
32:1066 (1979).
111. J. Wang, C.-J. Huang, and C. K. Chow, Red cell vitamin E and oxidative damage: a dual
role of reducing agents. Free Rad. Res. 24:291 (1996).
112. S. D. Aust and L. A. Morehouse, Role of metals in oxygen radical reactions. Free Rad. Biol.
Med. 1:3 (1986).
113. B. Halliwell and J. M. C. Gutterige, Oxygen radicals and iron in relation to biology and
medicine: some problems and concepts. Arch. Biochem. Biophys. 246:501 (1986).
114. J. G. Bieri and R. P. Evarts, Tocopherols and fatty acids in American diets. The recommended
allowance for vitamin E. J. Am. Dietet. Assoc. 62:147 (1993).
115. M. K. Horwitt, Interrelationships between vitamin E and polyunsaturated fatty acids in adult
men. Vitam Hormones 20:541 (1962).
116. M. K. Horwitt, Data supporting supplementation of human with vitamin E. J. Nutr. 121:424
117. J. T. Rotruck, A. T. Pope, H. E. Ganther, A. B. Swanson, D. G. Hofeman, and W. G. Hoekstra,
Selenium: biochemical role as a component of glutathione peroxidase. Science 179:588
118. F. Ursini, M. Maiorino, and C. Gregolin, The selenoenzyme phospholipid hydroperoxide
glutathione peroxidase. Biochim. Biophys. Acta 839:62 (1985).
Vitamin E 195
119. E. Niki, J. Tsuchiya, R. Tanimura, and Y. Kamiya, Regeneration of vitamin E from alphachromanoxy
radical by glutathione and vitamin C. Chem. Lett. 789 (1982).
120. J. E. Packer, T. F. Slater, and R. L. Wilson. Direct observation of a free radical interaction
between vitamin E and vitamin C. Nature 278:737 (1979).
121. A. C. Chan, K. Tran, T. Raynor, P. R. Ganz, and C. K. Chow, Regeneration of vitamin E
in human platelets. J. Biol. Chem. 266:17290 (1991).
122. H. Wefers, and H. Sies, The protection by ascorbate and glutathione against microsomal
lipid peroxidation is dependent on vitamin E. Eur. J. Biochem. 174:353 (1988).
123. G. W. Burton, V. Wronska, L. Stone, D. O. Foster, and K. U. Ingold. Biokinetics of dietary
RRR-?-tocopherol in the male guinea pig at the three dietary levels of vitamin C and two
levels of vitamin E. Evidence that vitamin C does not ‘‘spare’’ vitamin E in vivo. Lipids
25:199 (1990).
124. C. C. Reddy, R. W. Scholz, C. E. Thomas, and E. J. Massaro, Vitamin E dependent reduced
glutathione inhibition of rat liver microsomal lipid peroxidation. Life Sci. 31:571
125. E. Diliberto, Jr., G. Dean, C. Carter, and P. L. Allen, Tissue, subcellular and submitochondrial
distributions of semidehydroascorbate reductase: possible role of semidehydroascorbate reductase
in cofactor regeneration. J. Neurochem. 39:563 (1982).
126. V. E. Kagan, E. A. Serbinova, T. Forte, G. Scita, and L. Packer, Recycling of vitamin E in
human low density lipoproteins. J. Lipid Res. 33:385 (1992).
127. A. Constantunescu, D. Han, and L. Packer, Vitamin E recycling in human erythrocyte membranes.
J. Biol. Chem. 268:10906 (1993).
128. L. Packer, H. Tritschler, and K. Wessel, Neuroprotection by the metabolic antioxidant ?-
lipoic acid, Free Rad. Biol. Med. 22:359 (1997).
129. A. Bendich and J. A. Olson, Biological actions of carotenoids. FASEB J. 3:1927 (1989).
130. M. S. Micozzi, Foods, micronutrients, and reduction of human cancer, in Nutrition and Cancer
Prevention: Investigating the Role of Micronutrients (T. E. Moon and M. S. Micozzi,
eds.), Marcel Dekker, New York, 1989, pp. 213–241.
131. G. W. Burton, Antioxidant action of carotenoids. J. Nutr. 119:109 (1989).
132. G. W. Burton and K. U. Ingold, ?-Carotene: an unusal type of lipid antioxidant. Science
224:569 (1984).
133. T. A. Kennedy and D. C. Liebler, Peroxyl radical scavenging by beta-carotene in lipid bilayers.
Effect of oxygen partial pressure. J. Biol. Chem. 267:4658 (1992).
134. J. L. Napoli and K. R. Race, The biosynthesis of retinoic acid from retinol by rat tissues in
vitro. Arch. Biochem. Biophys. 255:95 (1987).
135. X. D. Wang, R. P. Marini, X. Rebuterne, J. G. Fox, N. I. Krinsky, and R. M. Russell, Vitamin
E enhances the lymphatic transport of beta-carotene and its conversion to vitamin A in the
ferret. Gastroenterology 108:719 (1995).
136. S. R. Blakely, E. Grundel, M. Y. Jenkins, and G. V. Mitchell, Alterations in ?-carotene and
vitamin E stress in rats fed ?-carotene and excess vitamin A. Nutr. Rev. 10:1035 (1990).
137. L. Arnrich and V. A. Virtamo, Interactions of fat-soluble vitamins in hypervitaminosis. Ann.
NY Acad. Sci. 355:109 (1980).
138. W. C. Willet, M. J. Stampfer, B. A. Underwood, J. O. Taylor, and C. H. Hennekins, Vitamins
A, E and carotene: effects of supplementation on their plasma levels. Am. J. Clin. Nutr. 38:
559 (1983).
139. B. R. Bacon and R. B. Britton, The pathology of hepatic iron overload: a free radical–mediated
process? Hepatology 11:127 (1990).
140. R. A. Leedle and S. D. Aust, the effect of glutathione on the vitamin E requirement for
inhibition of liver lipid peroxidation. Lipids 25:241 (1990).
141. C. K. Chow, H. H. Draper, and C. S. Csallany, Method for the assay of free and esterified
tocopherols. Anal. Biochem. 32:81 (1969).
142. U.S. Department of Agriculture, Human Nutrition Information Service. Nationwide Food
196 Chow
Consumption Survey, 1987–88. Individual intakes, 3 days. Computer tape PB 90-504044.
Springfield, VA. National Technical Information Service, 1990.
143. A. J. Sheppard and J. A. T. Pennington, Analysis and distribution of vitamin E in vegetable
oils and foods. In ‘‘Vitamin E in Health and Disease’’ (L. Packer and J. Fuchs, eds.), Marcer
Dekker, New York, 1993, pp. 9–31.
144. Food and Nutrition Board. Recommended dietary allowances, 10th ed. Washington, D.C.:
National Academy of Sciences, 1989.
145. J. Lehmann, Comparative sensitivities of tocopherol levels of platelets, red blood cells, and
plasma for estimating vitamin E nutritional ststus in the rat. Am. J. Clin. Nutr. 34:2104
146. G. K., Vatassery, A. M. Krezowski and J. H. Eckfeldt. Vitamin E concentrations in human
blood plasma and platelets. Am. J. Clin. Nutr. 37:1020 (1983).
147. Y.-M. Peng, Y.-S. Peng, Y. Lin, T. Moon, D. J. Roe and C. Ritenbaugh. Concentrations and
plasma-tissue-diet relationships of carotenoids, retinoids and tocopherols in humans. Nutr.
Cancer 23:233 (1995).
148. R. K. Chandra, Nutrition and immunity in the elderly: clinical significance. Nutr. Rev. 53:
S80 (1995).
149. S. N., Meydani, M. Meydani, M., L. C. Rall, F. Morrow and J. B. Blumberg. Assessment
of the safety of high-dose, short-term supplementation with vitamin E in healthy orer adults.
Am. J. Clin. Nutr. 60:704 (1994).
150. S. N., Meydani, M. P. Barklund, S. Liu, R. A. Miller, J. G. Cannon, F. D. Morrow, R. Rocklin
and J. B. Blumberg. Vitamin E supplementation enhances cell-mediated immunity in healthy
elderly subjects. Am. J. Clin. Nutr. 52:557 (1990).
151. S. N. Meydani, L. Leka and R. Loszewski. Long-term vitamin E supplementation enhances
immune response in healthy elderly. FASEB J. 8:A272 (1994).
151a. S.N. Meydani, M. Meydani, J.B. Blumberg, L.S. Lynette, G. Siber, R. Loszewski, C. Thompson,
M.C. Pedrosa, R.D. Diamond, and B.D. Stollar. Vitamin E supplementation and in vivo
immune response in healthy elderly subjects. J. Am. Med. Assoc. 277: 1380 (1997).
152. M. J. Stampfer, C. H. Henneckens, J. E. Mason, G. A. Colditz, B. Rosner and W. C. Willett.
Vitamin E consumption and the risk of coronary disease in women. New England J. Med.
328:1444 (1993).
152a. R.K. Chandra. Graying of the immune system. Can nutrient supplements improve immunity
on the elderly? J. Am. Med. Assoc. 277: 1398 (1997).
153. M. J. Stampfer, C. H. Henneckens, A. Ascherio, E. Giovannucci, G. A. Colditz, B. Rosner
and W. C. Willett. Vitamin E consumption and the risk of coronary disease in men. New
England J. Med. 328:1450 (1993).
154. L. H. Kushi, A. R. Folson, R. J. prineas, P. J. Mink, Y. Wu, R. M. and Bostick. Dietary
antioxidant vitamin intake and coronary disease in postmenopausal women. New England
J. Med. 334:1156 (1996).
155. N. G. Stephens, A. Parsons, P. M. Schofield, F. kelly, K. Cheeseman and M. J. Mitchison.
Randonmized controlled trial of vitamin E in patients with coronary disease: Cambridge
Heart Antioxidant Study (CHAOS). Lancet 347:781 (1996).
156. H. Esterbauer, M. Dieber-Rotheneder, M. Striegel and G. Waeg. Role of vitamin E in preventing
the oxidation of low density lipoprotein. Am. J. Clin. Nutr. 53:314S (1991).
157. I. Jialal, C. J. Fuller and B. A. Huet. the effect if alpha-tocopherol supplementation on LDL
oxidation: a dose-response study. Arterioscler. Thromb. Vascular Biol. 15:190 (1995).
158. J. Jandak, M. Steiner and P. D. Richardson. Tocopherol, an effective inhibitor of platelet
adhesion. Blood 73:141 (1989).
159. M. Steiner, Vitamin E: more than an antioxidant. Clin. Cardiol. 16 16S (1993).
160. H. Esterbauer, M. Dieber-Rotheneder, G. Waeg, G. Striegel and G. Jurgens. Biochemical,
structural and functional properties of oxidized low-density lipoprotein. Chem. Res. Toxicol.
3:77 (1990).
Vitamin E 197
161. H. Esterbauer, G. Jurgen, O. Quehenberger and E. Koller. Autoixdation of human low density
lipoprotein: loss of polyunsaturated fatty acids and vitamin E and generation of aldehydes.
J. Lipid Res. 28:495 (1987).
162. D. Steinberg, S. Parthasasathy, T. E. Carew, L. C. Khoo and J.-L. Witztum. Beyond cholesterol-
modifications of low-density lipoprotein that increase its atherogenicity. New England
J. Med. 320:915 (1989).
163. H. Sies, W. Stahl and A. R. Sundquist. Antioxidant functions of vitamins: vitamins E and
C, beta-carotene, and other carotenoids. Ann. New York Acad. Sci. 669:7 (1992).
164. L. Packer, Protective role of vitamin E in biological systems. Am. J. Clin. Nutr. 53:1050S
165. E. R. Greenberg, J. A. Baron, T. D. Tostesom, D. H. Freeman, G. J. Beck, T. A. Colacchino,
J. A. Coller, H. D. Frank, R. W. Haile, J. S. Mandel, D. W. Nierenberg, R. Rothstein,
D. C. Snover, M. M. Stevens, and R. U. van Stolk, A clinical trial of antioxidant vitamins
to prevent corolectoral adenoma. N Engl. J. Med. 331:141 (1994).
166. The Alpha-Tocopherol, Beta-Carotene Cancer prevention Group. The effect of vitamin E
and beta-carotene on the incidence of lung cancer and other cancer in male smokers. N. Engl.
J. Med. 330:1029 (1994).
167. W. J. Blot, J.-Y. Li, R. R. Taylor, W. Guo, S. Dawsey, G.-O. Wang, C. S. Yang, S.-F. Zheng,
M. Gail, G.-Y. Li, Y. Yu, B.-Q. Liu, J. Tangrea, Y.-H. Sun, F. Lie, J. F. Fraumeni, Y.-H.
Zhang, and B. Li, Nutrition intervention trials in Linxian, China: supplementation with specific
vitamin/mineral combinations, cancer incidence, and disease-specific mortality in the
general population. J. Natl. Cancer Inst. 85:1483 (1993).
168. C. K. Chow, Vitamin E and cancer, in Nutrition and Disease (K. K. Carroll and D. Kritchevsky,
eds.), AOCS Press, Champaign, 1994, pp. 173–233.
169. S.W. Barger, D. Horster, K. Furukawa, Y. Goodman, J. Krieglstein, and M.P. Mattson. Tumor
necrosis factors alpha and beta protect neurons against amyloid beta-peptide toxicity:
evidence for involvement of a kappa B-binding factor and attenuation of peroxide and Ca2
accumulation. Proc. Natl. Acad. Sci. USA 92: 9328 (1995).
170. H. Tohgi, T. Abe, M. Nakanishi, F. Hamato, K. Sasaki, and S. Takahashi. Concentrations of
alpha-tocopherol and its quinone derivative in cerebrospinal fluid from patients with vascular
dementia of the Binswanger type and Alzheimer type dementia. Neurosci. Lett. 174: 73
171. M. Sano, C. Ernesto, R.G. Thomas, M.R. Klauber, K. Schafer, M. Grundman, P. Woodbury,
J. Growdon, C.W. Cotman, E. Pfeiffer, L.S. Schneider, and L.J. Thal. A controlled trial of
selegiline, alpha-tocopherol of both as treatment for Alzheimer’s disease. N. Engl. J. Med.
336: 1216 (1997).
This Page Intentionally Left Blank
Bioorganic Mechanisms Important to
Coenzyme Functions
Emory University, Atlanta, Georgia
A. Definition
With the use (and misuse) of terms over significant periods of time, there is sufficient
uncertainty regarding a specific word that it is best to avoid confusion by defining it. This
is the case for the word ‘‘coenzyme,’’ which together with ‘‘cofactor’’ and ‘‘prosthetic
group’’ have to some become ambiguous biochemical jargon (1). As defined succinctly,
a coenzyme is a natural, organic molecule that functions in a catalytic enzyme system
(2). Coenzymes bind specifically within protein apoenzymes to constitute catalytically
competent holoenzymes. Hence, coenzymes are organic cofactors that augment the diversity
of reactions that otherwise would be limited to chemical properties, principally simple
acid–base catalysis, of side-chain substituents from amino acid residues within enzymes.
With tight binding, coenzymes may be referred to as prosthetic groups; with loose binding,
they may be called cosubstrates.
The focus of this chapter will be on the molecular means or mechanisms by which
coenzymes participate as the loci of bond making and breaking steps in holoenzymecatalyzed
reactions. Most coenzymes are more complex metabolic derivatives of watersoluble
vitamins. Since the subject of this volume is vitamins, coverage of coenzyme
mechanisms will bear emphasis on those coenzymes derived from such vitamins covered
in other regards in subsequent chapters. References are sparsely used for typical or recent
material that otherwise has largely become the domain of mechanistically inclined textbooks
of biochemistry (3,4), especially when set for the more advanced student or professional
200 McCormick
B. Groupings
Attempts to place each coenzyme in a singular mechanistic group are sometimes fraught
with difficulty because some can arguably fit into more than one category. For example,
the lipoyl residues of transacylases undergo oxidation–reduction as well as participate in
acyl transfers, but it is the latter that distinguishes the biological function. In at least a
few instances, the tetrahydro forms of pterin coenzymes are oxidized to the dihydro level
during operations of the enzymic systems, e.g., 5,10-methylene tetrahydrofolate in thymidylate
synthase and tetrahydrobiopterin in phenylalanine hydroxylase. More conventionally,
folyl coenzymes are pulled together in a broad mechanistic view under one-carbon
The six groups that are subdivisions of this chapter reasonably imply the mechanistic
and functional connections of principal coenzymes. These are oxidation–reduction reactions,
generation of leaving group potential, acyl activation and transfer, carboxylations,
one-carbon transfers, and rearrangements on vicinal carbons. These groupings follow a
conventional pattern of biochemical recognition of what major purposes are served by
vitamin-derived coenzymes.
A. Nicotinamide Coenzymes
Nicotinamide adenine dinucleotide (NAD) and its phosphate (NADP) are the two natural
pyridine nucleotide coenzymes derived from niacin, as discussed in Chapter 6. Both contain
an N(1)-substituted pyridine-3-carboxamide (nicotinamide) that is essential to function
in redox reactions with a potential near 0.32 V. The nicotinamide coenzymes function
in numerous oxidoreductase systems, usually of the dehydrogenase/reductase type,
which include such diverse reactions as the conversion of alcohols (often sugars and polyols)
to aldehydes or ketones, hemiacetals to lactones, aldehydes to acids, and certain amino
acids to keto acids (6). The general reaction stoichiometry is written as:
Substrate  NAD(P)
Bproduct  NAD(P)H  H
The common stereochemical mechanism of operation of nicotinamide coenzymes
when associated with enzymes involves the stereospecific abstraction of a hydride ion
(H) from substrate, with para addition to one or the other side of C-4 in the pyridine
ring of the coenzyme, as shown in Fig. 1. The second hydrogen of the substrate group
oxidized is concomitantly removed, typically from an electronegative atom (e.g., N, O,
or S), as a proton (H) that ultimately exchanges as a hydronium ion. The sidedness of
the pyridine ring is such that the para hydrogen up from the plane, when the carboxyamide
function is near (or to the right) and the pyridine N is on the left (or at the bottom), is
Fig. 1 The stereospecific hydride ion transfer to and from nicotinamide coenzymes generating
prochiral R and S forms.
Mechanisms for Coenzyme Function 201
prochiral R because it is oriented toward the re face (A side). The hydrogen down from
the plane of the ring is then prochiral S and oriented toward the si face (B side). There
are numerous examples of nicotinamide coenzyme–dependent enzymes that are stereospecific
as regards addition/removal of hydride ion from the prochiral R (A) or S (B) position
relative to the pyridine ring. These include, for A sidedness, NAD-dependent alcohol
dehydrogenase and NADP-dependent cytoplasmic isocitrate dehydrogenase; for B sidedness,
NAD-dependent d-glyceraldehyde-3-phosphate dehydrogenase and NADP-dependent
d-glucose-6-phosphate dehydrogenase (7). In general, A-side dehydrogenases bind
a conformation of the coenzyme in which the nicotinamide ring has an anti orientation
with respect to the ribosyl ring, i.e., the carboxamide groups points away; B-side dehydrogenases
usually have syn conformation. This is probably due to stabilization of the dihydronicotinamide
in the boat conformation attributable to orbital overlap between the lone
n pair of electrons on the pyridine nitrogen and the antiboding ? orbital of the ribosyl CO
moiety, as depicted in Fig. 2. The pseudoaxial hydrogens, either pro-R with anti or
pro-S with syn orientation, are then more easily transferred because of orbital overlap in
the transition state.
Nicotinamide coenzymes also participate in other (nonredox) biological reactions
that involve ADP ribosylations; however, these are properly considered as substrate rather
than coenzymic functions. The larger nucleotide-like structure of nicotinamide coenzymes
is important in regard to coenzymic roles primarily to facilitate recognition and specific
binding by the protein apoenzymes with which they interact.
B. Flavocoenzymes
Flavin mononucleotide (FMN), flavin adenine dinucleotide (FAD), and covalently linked
flavocoenzymes (usually 8?-substituted FAD) are the natural coenzymes derived from
riboflavin (vitamin B2), which is discussed in Chapter 7. All contain an isoalloxazine ring
system with a redox potential near0.2 V when free in solution, but the potential is subject
to considerable variation when bound within functional flavoproteins. This propensity for
being able to shift potential gives flavoenzymes a wider operating range for oxidation–
reduction reactions than is the case for enzymes dependent on nicotinamide coenzymes.
Flavoproteins are variably able to catalyze both one- and two-electron redox reactions.
The ring portion involving nitrogens 1 and 5 and carbon 4a reflects sequential addition
or loss of electrons and hydrogen ions in one-electron processes and addition or
loss of other transitory adducts with two-electron processes. There are nine chemically
discernible redox forms, i.e., three levels of oxidation–reduction and three species for
acid, neutral, and base conditions. Of these only five have biological relevance because
Fig. 2 The syn and anti conformations of bound nicotinamide coenzymes that putatively transfer
pro-(S) and pro-(R) hydrogens, respectively.
202 McCormick
Fig. 3 Biologically important redox states of flavocoenzymes with pKa values for interconversions
of the free species.
of pH considerations (2). These biological forms are summarized in Fig. 3. It should also
be noted that free flavin semiquinones (radicals) are quite unstable, and the free flavin
hydroquinones react very rapidly with molecular oxygen to become reoxidized. This latter
is kinetically dissimilar to reduced nicotinamide coenzymes, which reoxidize more slowly
with O2. Binding to specific enzymes markedly affects the kinetic stability of the halfreduced
and reduced forms of flavocoenzymes and again allows for diverse reactions under
biological situations. An example of a flavoprotein undergoing a one-electron transfer is
with the microsomal NADPH–cytochrome P450 reductase. This enzyme contains both
FAD and FMN, which latter cycles between a neutral semiquinone and fully reduced form
(8). The FAD in the electron-transferring flavoprotein, which mediates electron flow from
fatty acyl coenzyme A (CoA) to the mitochondrial electron transport chain, cycles between
oxidized quinone and anionic semiquinone. In addition, a single-step, two-electron transfer
from substrate can occur in the nucleophilic reactions shown in Fig. 4. Such cases as
hydride ion transfer from reduced nicotinamide coenzymes or the carbanion generated by
Fig. 4 Reaction types encountered with flavoquinone coenzymes and natural nucleophiles to generate
N-5 and C-4a adducts as intermediates.
Mechanisms for Coenzyme Function 203
base abstraction of a substrate proton may lead to attack at the flavin N-5 positions; some
nucleophiles, such as the hydrogen peroxide anion, are added on the C-4a position with
its frontier orbital.
As regards stereochemistry, the transfer of hydrogen on and off N-5 can take place
to or from one or the other face of the isoalloxazine ring system. When visualized with
the benzenoid ring to the left and side chain at top (as in the top of Fig. 3), orientation
is to the si face with re on the opposite side. A fair number of flavoproteins have now
been categorized on this steric basis (7). Examples include the FAD-dependent glutathione
reductase from human erythrocytes, which uses NADPH as substrate and is re side interacting,
and the FMN-dependent spinach glycolate oxidase, which is si side interacting. It
should be noted that reduced flavins free in solution have bent or ‘‘butterfly’’ conformations,
since the dihydroisoalloxazine has a 144° angle between benzenoid and pyrimidinoid
protions. The orientations taken when bound to enzymes, however, may vary and
can influence the redox potential.
There are flavoprotein-catalyzed dehydrogenations that are both nicotinamide coenzyme–
dependent and independent, reactions with sulfur-containing compounds, hydroxylations,
oxidative decarboxylations, dioxygenations, and reduction of O2 to hydrogen peroxide.
The diversity of these systems is covered in periodic symposia on flavins and
flavoproteins (9). The intrinsic ability of flavins to be variably potentiated as redox carriers
upon differential binding to proteins, to participate in both one- and two-electron transfers,
and in reduced (1,5-dihydro) form to react rapidly with oxygen permits wide scope in
their operation.
A. Thiamine Pyrophosphate
Thiamine pyrophosphate (TPP) is the coenzyme derived from thiamine (vitamin B1),
which is discussed in Chapter 8. Though the substituted pyrimidyl portion shares a role
in apoenzymic recognition and binding, the thiazole moiety not only is important in that
regard but is involved with substrates to provide a transitory matrix that provides good
leaving group potential (2,10). Specifically, TPP as a Mg2 ternary complex within enzymes
can react as an ylid, resonance-stabilized carbanion that attacks carbonyl functions
as illustrated in Fig. 5.
In all cases, the bond to be labilized (from R? to the C of the original carbonyl carbon)
must be oriented for maximal ?-? orbital overlap with the periplanar system extending to
the electron-deficient quaternary nitrogen of the thiazole ring. There are two general types
of biological reactions in which TPP functions in so-called active aldehyde transfers. First,
in decarboxylation of ?-keto acids, the condensation of the thiazole moiety of TPP with
the ?-carbonyl carbon on the acid leads to loss of CO2 and production of a resonancestabilized
carbanion. Protonation and release of aldehyde occur in fermentative organisms
such as yeast, which have only the TPP-dependent decarboxylase, but reaction of the ?-
hydroxalkyl-TPP with lipoyl residues and ultimate conversion to acyl-CoA occurs in
higher eukaryotes, including humans with multienzymic dehydrogenase complexes (see
Sec. IV.8). The other general reaction involving TPP is the transformation of ?-ketols
(ketose phosphates). Specialized phosphoketolases in certain bacteria and higher plants
can split ketose phosphates, e.g., d-xylulose-5-phosphate, to simpler, released products,
e.g., d-glyceraldehyde-3-phosphate and acetyl phosphate. However, the reaction of impor204
Fig. 5 Function of the thiazole moiety of thiamine pyrophosphate. In ?-keto acid decarboxylations,
R? is a carboxylate lost as CO2 and R? is a proton generating an aldehyde. In ketolations
(both trans- and phospho-), R and R? are part of a ketose phosphate generating a bound glycoaldehyde
intermediate and R? is an aldose phosphate or inorganic phosphate generating a different ketose
phosphate or acetyl phosphate, respectively.
tance to humans and most animals is a transketolation. Transketolase is a TPP-dependent
enzyme found in the cytosol of many tissues, especially liver and blood cells, where principal
carbohydrate pathways exist. This enzyme catalyzes the reversible transfer of a glycoaldehyde
moiety (?,?-dihydroxyethyl-TPP) from the first two carbons of a donor ketose
phosphate, i.e., d-xylulose-5-phosphate, to the aldehyde carbon of an aldose phosphate,
i.e., d-ribose-5-phosphate, of the pentose phosphate pathway wherein d-sedoheptulose-7-
phosphate and d-glyceraldehyde-3-phosphate participate as the other substrate–product
B. Pyridoxal-5?-Phosphate
Two of the three natural forms of vitamin B6, which is discussed in Chapter 10, can be
phosphorylated to yield directly functional coenzymes, i.e., pyridoxal-5?-phosphate (PLP)
and pyridoxamine-5?-phosphate (PMP). PLP is the predominant and more diversely functional
coenzymic form, although PMP interconverts as coenzyme during transaminations
(2–5,11). At physiological pH, the dianionic phosphates of these coenzymes exist as zwitterionic
meta-phenolate pyridinium compounds. Both natural and synthetic carbonyl reagents
(e.g., hydrazines and hydroxylamines) form Schiff bases with the 4-formyl function
of PLP, thereby removing the coenzyme and inhibiting PLP-dependent reactions.
PLP functions in numerous reactions that embrace the metabolism of proteins, carbohydrates,
and lipids. Especially diverse are PLP-dependent enzymes that are involved in
amino acid metabolism. By virtue of the ability of PLP to condense its 4-formyl substituent
Mechanisms for Coenzyme Function 205
with an amine, usually the ?-amino group of an amino acid, to form an azomethine (Schiff
base) linkage, a conjugated double-bond system extending from the ? carbon of the amine
(amino acid) to the pyridinium nitrogen in PLP results in reduced electron density around
the ? carbon. This potentially weakens each of the bonds from the amine (amino acid)
carbon to the adjoined functions (hydrogen, carboxyl, or side chain). A given apoenzyme
then locks in a particular configuration of the coenzyme–substrate compound such that
maximal overlap of the bond to be broken occurs with the resonant, coplanar, electronwithdrawing
system of the coenzyme complex. These events are depicted in Fig. 6.
Selection of the ? bond to be cleaved (R?-C) is achieved by stabilizing a conformation
of the external aldimine in which the leaving group is orthogonal to the plane of the
ring system, thus ensuring maximal orbital overlap with the ? system in the transition
state. The type of stereochemistry involved is characteristic of a given enzyme system.
Both re and si faces of the PLP complex can orient toward a reactive function, such
as a base, contributed by enzyme protein. This is the case, for instance, with aspartate
aminotransferase whereby a base function abstracts a proton from the ? position and
transfers it to the same side of the ? system (syn transfer), adding it to the si face of the
azomethine group in the ketimine. In this case, the hydrogen is incorporated in a pro-S
position at carbon 4? (the methylene of PMP).
Aminotransferases effect rupture of the ?-hydrogen bond of an amino acid with
ultimate formation of an ?-keto acid and PMP; this reversible reaction proceeds by a
double-displacement mechanism and provides an interface between amino acid metabolism
and that for ketogenic and glucogenic reactions. Amino acid decarboxylases catalyze
breakage of the ?-carboxyl bond and lead to irreversible formation of amines, including
several that are functional in nervous tissue (e.g., epinephrine, norepinephrine, serotonin,
and ?-aminobutyrate). The biosynthesis of heme depends on the early formation of ?-
aminolevulinate from PLP-dependent condensation of glycine and succinyl-CoA followed
by decarboxylation. There are many examples of enzymes, such as cysteine desulfhydrase
and serine hydroxymethyltransferase, that affect the loss or transfer of amino acid side
chains. PLP is the essential coenzyme for phosphorylase that catalyzes phosphorolysis of
the ?-1,4 linkages of glycogen. An important role in lipid metabolism is the PLP-dependent
condensation of l-serine with palmitoyl-CoA to form 3-dehydrosphinganine, a precursor
of sphingolipids. The diversity of PLP functions is covered in periodic symposia
Fig. 6 Operation of pyridoxal-5?-phosphate with a biological primary amine. Loss of R? shown
from the aldimine can be generalized for side-chain cleavages and eliminations, loss of CO2 in
decarboxylations, or loss of a proton in aminotransferations and racemizations.
206 McCormick
that now include other carbonyl compounds as cofactors (pyruvyl enzymes, quinoproteins,
etc.) (12).
A. Phosphopantetheine Coenzymes
4?-Phosphopantetheine is a phosphorylated amide of ?-mercaptoethylamine and the vitamin
pantothenate, which is discussed in Chapter 9. The phosphopantetheinyl moiety serves
as a functional component within the structure of CoA and as a prosthetic group covalently
attached to a seryl residue of acyl carrier protein (ACP). Because of a thiol (sulfhydryl)
terminus with a pKa near 9, phosphopantetheine and its coenzymic forms are readily oxidized
to the catalytically inactive disulfides.
Esterification of the thiol function to many carboxylic acids is the prelude to numerous
acyl group transfers and enolizations (2–5). In classic terminology, the thiol ester
enhances both ‘‘head’’ and ‘‘tail’’ activations of acyl compounds, as illustrated in Fig.
7. In head activation, the carbonyl function is attacked by a nucleophile and releases the
original thiol. The carbonyl carbon of a thiol ester is more positively polarized than would
be an oxy ester counterpart. Hence, the acyl moiety is relatively activated for transfer.
The phosphopantetheine terminus of CoA and ACP both function in acyl transfers, the
latter only within the fatty acid synthase complex. In tail activation, there is greater tendency
for thio than oxy esters to undergo enolization by permitting removal of an ? hydrogen
as a proton. The enolate is stabilized by delocalization of its negative charge between
the ? carbon and the acyl oxygen, which makes it thermodynamically accessible as an
intermediate. Since the developing charge is also stabilized in the transition state that
precedes the enolate, it is also kinetically accessible. This process leads to facile condensation
whereby nucleophilic addition of the carbanion-like carbon of the enolate to a neutral
activated acyl group (another thiol ester) is a favored process. CoA is a good leaving
group from the tetrahedral intermediate. From the above, it follows that thio esters are
more like a ketone than an oxy ester. The degree of resonance electron delocalization
from the overlap of sulfur p orbitals with the acyl ? bond is less than with oxygen in oxy
The myriad acyl thio esters of CoA are central to the metabolism of numerous compounds,
especially lipids and the penultimate catabolites of carbohydrates and ketogenic
Fig. 7 Activations of acyl moieties enhanced by formation of thio esters with phosphopantetheine
Mechanisms for Coenzyme Function 207
amino acids. For example, acetyl-CoA, which is formed during metabolism of carbohydrates,
fats, and some amino acids, can acetylate compounds such as choline and hexosamines
to produce essential biochemicals. It can also condense with other metabolites, such
as oxalacetate, to supply citrate, and it can lead to formation of cholesterol. The reactive
sulfhydryl termini of ACP provide exchange points for acetyl-CoA and malonyl-CoA. The
ACP-S-malonyl thio ester can chain-elongate during fatty acid biosynthesis in a synthase
B. Lipoic Acid
?-Lipoic (thioctic) acid is not a vitamin for the human and other animals because it can
be biosynthesized from longer chain, essential fatty acids. However, it is indispensable
in its coenzymic role, which interfaces with some of the functions of TPP and CoA (2–
4). The natural d isomer of 6,8-dithiooctanoic (1,2-dithiolane-3-pentanoic) acid occurs in
amide linkage to the -amino group of lysyl residues within transacylases that are core
protein subunits of ?-keto acid dehydrogenase complexes of some prokaryotes and all
eukaryotes. In such transacylases, the functional dithiolane ring is on an extended flexible
arm. The lipoyl group mediates the transfer of the acyl group from an ?-hydroxyalkyl-
TPP to CoA in a cyclic system that transiently generates the dihydrolipoyl residue, as
shown in Fig. 8. Hence, the lipoyl/dihydrolipoyl pair in this cycle serves a dual role of
electron transfer and acyl group vector by coupling the two processes.
The three ?-keto acid dehydrogenase complexes of mammalian mitochondria are
for pyruvate and ?-ketoglutarate of the Krebs citric acid cycle and for the branched-chain
?-keto acids from some amino acids. All involve participation of lipoyl moieties within
core transacylase subunits surrounded by subunits of TPP-dependent ?-keto acid decarboxylase,
which generate ?-hydroxyalkyl-TPP, and are further associated with subunits
of FAD-dependent dihydrolipoyl (lipoamide) dehydrogenase. The number of subunits and
their packing varies among cases. For instance, a single particle of E. coli pyruvate dehy-
Fig. 8 Function of the lipoyl moiety within enzymes involved in transacylations following ?-
keto acid decarboxylations (see Fig. 5). In multienzyme dehydrogenases, there is transfer of an acyl
moiety from an ?-hydroxyalkyl thiamine pyrophosphate to the lipoyl group of a transacylase core
and thence to CoA. This results in formation of the dihydrolipoyl group, which is cyclically reoxidized
by the FAD-dependent dihydrolipoyl dehydrogenase.
208 McCormick
Fig. 9 Function of the biotinyl moiety within enzymes involved in carboxylations with the intermediacy
of the putative carbonyl phosphate.
drogenase consists of at least 24 chains of each decarboxylase and transacetylase plus 12
chains of the flavoprotein.
A. Biotin
The characteristics of the vitamin biotin are discussed in Chapter 11. The coenzymic form
of this vitamin occurs only as the vitamin with its valeric acid side chain amide-linked
to the ?-amino group of specific lysyl residues in carboxylase and transcarboxylase (2–
5). The length of this flexible arm (14 A° ) is similar to that encountered with the lipoyl
attachments in transacylases. Biotin-dependent carboxylases operate by a common mechanism
illustrated in Fig. 9. This involves tautomerization of the ureido 1?-N to enhance
its nucleophilicity. Though the two ureido nitrogens are essentially isoelectronic in the
imidazoline portion of biotin, the steric crowding of the thiolane side chain near the 3?-
N essentially prevents chemical additions to position 3?. Phosphorylation of bicarbonate
by ATP to form carbonyl phosphate provides an electrophilic mixed-acid anhydride. This
latter can then react at the nucleophilically enhanced 1?-N to generate reactive N(1?)-
carboxybiotinyl enzyme. This in turn can exchange the carboxylate function with a reactive
center in a substrate, typically at a carbon with incipient carbanion character (13).
There are nine known biotin-dependent enzymes: six carboxylases, two decarboxylases,
and a transcarboxylase. Of these, only four carboxylases have been found in tissues of
humans and other mammals. Acetyl-CoA carboxylase is a cytosolic enzyme that catalyzes
formation of malonyl-CoA for fatty acid biosynthesis. Pyruvate carboxylase is a mitochondrial
enzyme that forms oxalacetate for citrate formation. Propionyl-CoA carboxylase
forms the d isomer of methylmalonyl-CoA on a pathway toward succinyl-CoA in the
Krebs cycle. Finally, ?-methylcrotonyl-CoA carboxylase forms ?-methylglutaconyl-CoA
in the catabolic pathway from l-leucine.
Among natural compounds with a pteridine nucleus, those most commonly encountered
are derivatives of 2-amino-4-hydroxypteridines, which are trivially named pterins. HuMechanisms
for Coenzyme Function 209
mans and other mammals normally require only folic acid as a vitamin of the pterion
type. This is discussed in Chapter 12. Although a number of pterins when reduced to the
5,6,7,8-tetrahydro level function as coenzymes, the most generally utilized are poly-?-
glutamates of tetrahydrofolate (THF) (2,14). Tetrahydropteroylglutamates and forms of
its natural derivatives responsible for vectoring one-carbon units in different enzymic reactions
are shown in Fig. 10. All of these bear the substituent for transfer at nitrogen 5 or
10 or are bridged between these basic centers. The number of glutamate residues varies,
usually from one to seven, but a few to several glutamyls optimize binding of tetrahydrofolyl
coenzymes to most enzymes requiring their function. Less broadly functional, but
essential for some coenzymic roles of pterins, is tetrahydrobiopterin, which cycles with
its quinoid 7,8-dihydro form during O2-dependent hydroxylation of such aromatic amino
acids as in the conversion of phenylalanine to tyrosine. Also, the recently elucidated
molybdopterin functions in some Mo/Fe flavoproteins, such as xanthine dehydrogenase.
Pterion coenzymes, most of them at the tetrahydro level, are sensitive to oxidation.
The interconversions of folate with the initial coenzymic relatives, the tetrahydrofolyl
polyglutamates, involve dihydrofolate reductase, necessary for reducing the vitamin
level compound through 7,8-dihydro to 5,6,7,8-tetrahydro levels. This enzyme is the target
of such inhibitory drugs as aminopterin and amethopterin (methotrexate). A similar dihydropterin
reductase catalyzes reduction of dihydro- to tetrahydrobiopterin. Tetrahydrofolate
is trapped intracellularly and extended to polyglutamate forms that operate with THFdependent
systems. In some cases (e.g., thymidylate synthetase), there is a redox change
in tetrahydro to dihydro coenzyme, which is recycled by the NADPH-dependent reductase.
There are different redox levels for the one-carbon fragment carried by THF systems.
With formate, phosphorylation by ATP leads to formyl phosphate, a reactive mixed-acid
anhydride. This can react with nitrogens at either position 5 or 10 in THF to form the
formyl compounds; upon cyclization, the 5,10-methenyl-THF results. With formaldehyde,
which becomes an electrophilic cation when protonated (CH2OH), the reaction shown
in Fig. 11 ensues. This reaction is initiated at the most basic N-5 (pKa  4.8) of THF to
generate an N-hydroxymethyl intermediate. The high electron charge (high basicity) and
large free valence (high polarizability) confer a low activation energy for electrophilic
substitution at N-5. With loss of a hydroxyl and quaternization of this nitrogen, the onecarbon
unit becomes canonically equivalent to a carbocation, which can react either interor
intramolecularly. In the former instance, reaction with a nucleophile can lead to another
Fig. 10 Structures with numbering for tetrahydropteroyl-l-glutamates, including the biological
derivatives for one-carbon transfers.
210 McCormick
Fig. 11 Reaction of formaldehyde with the basic N-5 of tetrahydrofolyl coenzyme to generate
an N-hydroxymethyl intermediate that leads to a mesomeric carbocation. This latter can react intermolecularly
in a Mannich-like reaction or intramolecularly to form 5,10-methylene tetrahydrofolate.
hydroxymethyl compound; in the latter, 5,10-methylene-THF is formed. For the methyl
level, reduction of the methylene-THF occurs.
An overview of some of the major interconnections among the one-carbon-bearing
THF coenzymes and their metabolic origins and roles include a fair range of reaction
types. As mentioned above, there is generation and utilization of formate. The important
de novo biosynthesis of purine includes two steps wherein glycinamide ribonucleotide
and 5-amino-4-imidazole carboxamide ribonucleotide are transformylated by 5,10-methenyl-
THF and 10-formyl-THF, respectively. In pyrimidine nucleotide biosynthesis, deoxyuridylate
and 5,10-methylene THF form thymidylate and dihydrofolyl coenzyme in a
mechanism whereby the tetrahydro coenzyme is oxidized to the dihydro level, as shown
in Fig. 12. There are conversions of some amino acids, namely, N-formimino-l-glutamate
(from histidine catabolism) with THF to l-glutamate and 5,10-methenyl-THF (via 5-
Fig. 12 The 1,3 hydrogen shift and redox nature of 5,10-methylene–tetrahydrofolyl coenzyme
within thymidylate synthase.
Mechanisms for Coenzyme Function 211
formimino-THF), l-serine with THF to glycine and 5,10-methylene-THF, and l-homocysteine
with 5-methyl-THF to l-methionine and regenerated THF.
Though there are several biologically active forms derived from vitamin B12, discussed
in Chapter 13, only a couple warrant attention as coenzymes in higher eukaryotes (2,15).
The coenzyme B12 (COB12) known to function in most organisms, including humans, is
5?-deoxyadenosylcobalamin. A second important coenzyme form is methylcobalamin
(methyl-B12), in which the methyl group replaces the deoxyadenosyl moiety of CoB12.
Some prokaryotes utilize other bases (e.g., adenine) in this position originally occupied
by the cobalt-coordinated cyanide anion in cyanocobalamin, the initially isolated form of
vitamin B12.
The metabolic interconversion of vitamin B12 as the naturally occurring hydroxocobalamin
(B12a) with other vitamin and coenzyme level forms involves sequential reduction
of B12a to the paramagnetic or radical B12r and further to the very reactive B12s. The latter
reacts in enzyme-catalyzed nucleophilic displacements of tripolyphosphate from ATP to
generate CoB12 or of THF from 5-methyl-THF to generate methyl-B12.
Seemingly all CoB12-dependent reactions react through a radical mechanism, and
all but one (Lactobacillus leichmanii ribonucleotide reductase) involve a rearrangement
of a vicinal group (X) and a hydrogen atom. This general mechanism is illustrated in Fig.
13. For the CoB12-dependent mammalian enzyme, l-methylmalonyl-CoA mutase, X is the
CoA-S-CO- group, which moves with retention of configuration from the carboxyl-bearing
carbon of l(R)-methylmalonyl-CoA to the carbon ? to the carboxyl group in succinyl-
CoA. This reaction is essential for funneling propionate to the tricarboxylic acid cycle.
Without CoB12 (from vitamin B12), more methylmalonate is excreted, but also the CoA
Fig. 13 Radical intermediates in vicinal rearrangements catalyzed by enzymes utilizing 5?-deoxyadenosyl-
212 McCormick
ester competes with malonyl-CoA in normal fatty acid elongation to form instead abnormal,
branched-chain fatty acids. Methyl-B12 is necessary in the transmethylase-catalyzed
formation of l-methionine and regeneration of THF. Without this role, there would not
only be no biosynthesis of the essential amino acid, but increased exogenous supply of
folate would be necessary to replenish THF, which would not otherwise be recovered
from its 5-methyl derivative.
1. O. H. Hasim and N. Azila Adnan, Coenzyme, cofactor and prosthetic group—ambiguous
biochemical jargon, Biochem. Ed. 22:93 (1994).
2. D. B. McCormick, Coenzymes, biochemistry, in Encyclopedia of Human Biology, Vol. 2 (R.
Dulbecco, ed.-in-chief), Academic Press, San Diego, 1991, pp. 527–545.
3. D. E. Metzler, Coenzymes—nature’s special reagents, in Biochemistry: The Chemical Reactions
of Living Cells, Academic Press, New York, 1977, pp. 428–516.
4. P. A. Frey, Vitamins, coenzymes, and metal cofactors, in Biochemistry (G. Zubay, ed.), Wm.
C. Brown, Dubuque, 1993, pp. 278–304.
5. C. T. Walsh, Enzymatic Reaction Mechanisms, Freeman, San Francisco, 1977.
6. D. Dolphin, R. Poulson, and O. Avamovic (eds.), Pyridine Nucleotide Coenzymes, Parts A
and B, Wiley-Interscience, New York, 1987.
7. D. J. Creighton and N. S. R. K. Murthy, Stereochemistry of enzyme-catalyzed reactions at
carbon, The Enzymes, Vol. 19 (D. S. Sigmon and P. D. Boyer, eds.), Academic Press, San
Diego, 1990, pp. 323–421.
8. A. H. Merrill, Jr., J. D. Lambeth, D. E. Edmondson, and D. B. McCormick, Formation and
mode of action of flavoproteins, Annu. Rev. Nutr. 1:281–317 (1981).
9. B. Curti, S. Ronchi, and G. Zanetti (eds.), Flavins and Flavoproteins, Walter de Gruyter,
Berlin, 1991.
10. H. Z. Sable and C. J. Gubler (eds.), Thiamin. Twenty years of progress, Ann. N.Y. Acad. Sci.,
378, (1982).
11. D. Dolphin, R. Poulson, and O. Avamovic (eds.), Pyridoxal Phosphate, Parts A and B, Wiley-
Interscience, New York, 1987.
12. T. Fukui, H. Kagamiyama, K. Soda, and H. Wada (eds.), Enzymes Dependent on Pyridoxal
Phosphate and Other Carbonyl Compounds as Cofactors, Pergamon Press, Oxford, 1991.
13. J. R. Knowles, The mechanism of biotin-dependent enzymes, Annu. Rev. Biochem. 58:195,
14. R. L. Blakley and T. J. Benkovic (Vol. 1) and R. L. Blakley and V. M. Whitehead (Vol. 3),
Folate and Pterins, John Wiley and Sons, New York, 1986.
15. D. Dolphin, (ed.), B12, Vols. 1 and 2, Wiley-Interscience, New York, 1982.
University of Guelph, Guelph, Ontario, Canada
University of Calgary, Calgary, Alberta, Canada
The identification of niacin as a vitamin resulted from an urgent need to cure pellagra,
which ravaged low socioeconomic groups of the southeastern United States in the early
twentieth century and various European populations for the previous two centuries. The
history of pellagra and the progression of early research findings in niacin nutrition are
extensively documented by many writers, including Harris (1) and Carpenter (2). The first
reports of the condition were recorded in the early 1700s in Spain. Corn had been introduced
to Europe from the Americas and quickly became a staple food, as it could produce
more calories per acre than wheat or rye.
In 1735, the Spanish physician Casal became the first to describe the strange new
disease, which he termed mal de la rosa (disease of the rose). He documented the skin
lesions of the disease quite carefully, and the characteristic rash around the neck of pellagrins
is still referred to as ‘‘Casal’s necklace.’’ The disease spread geographically with
the cultivation of corn and became known as pelle agra (rough skin) by the Italian peasantry.
Due to the widespread incidence of pellagra in eighth-century Europe and the dedication
of entire hospitals for victims of the disease, there was a painstaking documentation
of its symptoms (1,2). These include diarrhea and neurological disturbances (dementia)
(3), as well as the sun-sensitive dermatitis, which, along with the eventual death of the
patient, are often referred to as the ‘‘4 D’s’’ of pellagra. The disease was recognized in
populations in Egypt, South Africa, and India in the late 1800s and early 1900s. It reached
epidemic proportions in the United States in the first half of the 1900s, producing at least
214 Kirkland and Rawling
250,000 cases and 7,000 deaths per year for several decades in the southern states alone
(4). Serious outbreaks continue to occur in some developing countries (5). An improved
standard of living and supplementation of some breakfast cereals with niacin have limited
the disease in North America and Europe, but cases of pellagra still occur, although they
are likely underreported due to a lack of familiarity of modern physicians with this disease.
It is also likely that subclinical deficiencies of niacin exist in developed countries; 15%
of women surveyed in Malmo, Sweden, had blood nucleotide pools that indicated a suboptimal
niacin intake (6).
Until the first half of the twentieth century, the etiology of pellagra was unknown.
Early theories suggested that it is a type of leprosy, that it results from a toxin in moldy
corn, or that it is an infectious disease communicated through an insect vector. Gradually,
an association between pellagra and corn consumption became apparent, even in the absence
of mold contamination. This was confirmed by Joseph Goldberger (7), who determined
that a pellagra-preventative (‘‘P-P’’) factor, missing from corn, was necessary to
prevent and cure pellagra in humans. In 1937, Elvehjem and colleagues discovered that
nicotinic acid could cure black tongue in dogs, an early animal model for pellagra (8).
Paradoxically, chemical analysis of corn revealed that it is not especially low in nicotinic
acid content. However, Krehl et al. (9) induced niacin deficiency in rats by feeding the
animals a corn-based diet and then alleviated the symptoms with the addition of casein,
a protein source rich in tryptophan. There had been suggestions that tryptophan deficiency
caused pellagra, and in 1921, a pellagrous patient was treated successfully with tryptophan
supplementation (10). It was eventually realized that tryptophan can be used, with low
efficiency, as a substrate for the synthesis of nicotinamide adenine dinucleotide (NAD)
(11). In 1951, Carpenter’s group found that niacin in corn is biologically unavailable and
can be released only following prolonged exposure to extremes in pH (12). These findings
eventually led to a better understanding of the contribution of a corn-based diet to the
development of pellagra; corn-based diets are low in tryptophan, and the preformed nicotinic
acid is tightly bound to a protein that prevents its absorption. The release of niacin
from this protein at extremes of pH explained the good health of native Americans, who
used corn as a dietary staple: in all cases these societies processed their corn with alkali
prior to consumption. Corn was in use throughout North and South America for thousands
of years as a domesticated crop, and methods of preparation ranged from treatment of
corn with ashes from the fireplace, to the well-known use of lime water (water and calcium
hydroxide) in the making of tortillas. Had the explorers brought native American cooking
techniques along with corn to sixteenth-century Europe, pellagra epidemics might never
have occurred.
Pellagra may be difficult to identify, as the symptoms of dermatitis, diarrhea, and
dementia occur in an unpredictable order, and it is uncommon to find all three aspects
until the disease is very advanced. The earliest sign of deficiency is often inflammation
in the oral cavity, which progresses to include the esophagus and eventually the whole
digestive tract, associated with a severe diarrhea (1). Burning sensations often discourage
food consumption, and the patient progresses to a state of marasmus; the cause of death
in most cases is probably a result of the effects of general malnutrition (poor disease
resistance) and diarrhea (1). Because of the sensitivity to sunlight, and possibly due to
dietary variation, the incidence of pellagra is seasonal. The dermatitis can become severe
rapidly, with exposed skin showing hyperpigmentation, bullous lesions, and desquamation
(Fig. 1). The skin may heal during fall and winter, leaving pink scars, only to revert to
open sores the following summer. The other dramatic symptom of pellagra is dementia,
Niacin 215
(b) (c)
Fig. 1 (a) An Austrian child with pellagra, showing dermatitis on the exposed skin of the face
and hands. Note the unaffected skin on the wrists where the cuffs of the coat are turned up. (b)
Severe pellagrous lesions on the arms of a 32-year-old woman. They did not respond to riboflavin
supplementation but cleared up with the use of Valentine’s whole liver extract, a good source of
niacin. (c) Dermatitis on the legs and feet of a typical pellagra patient, showing the pattern of protection
from the sun afforded by sandals.
[Fig. 1(a) is reproduced with permission from Pellagra: History, Distribution, Diagnosis, Prognosis,
Treatment, Etiology, by Stewart R. Roberts, C.V. Mosby Company, 1914. Fig. 1 (b) and 1 (c) are
reproduced with permission from Clinical Pellagra, by Seale Harris, published by C.V. Mosby, St.
Louis, 1941.]
216 Kirkland and Rawling
which usually goes beyond the type of depression associated with general malnutrition.
Neurological changes in pellagra patients begin peripherally, with signs such as muscle
weakness, twitching and burning feelings in the extremities, and altered gait (13). Early
psychological changes include depression and apprehension, but these progress to more
severe changes, such as vertigo, loss of memory, deep depression, paranoia and delirium,
hallucinations, and violent behavior (14). This type of dementia is very similar to schizophrenia
(15). There are numerous examples of the pellagrous insane committing murder
(16), although it was more common for these patients to turn to suicide. Interestingly,
many insane pellagrins are drawn to water, and their most common form of suicide is
reported to be drowning (16). While there are pathological changes in the spinal cord in
advanced pellagra (17), and some of the motor disturbances are permanent (18), there is
a striking recovery of psychological function when insane pellagra patients are treated
with nicotinic acid, with a disappearance of symptoms in 1–2 days (14). These observations
suggest that a compound derived from niacin is involved in neural signaling pathways.
The first biochemical role established for niacin was its involvement in redox reactions
and energy metabolism. For many years, however, the symptoms of pellagra remained
mysterious when viewed from this perspective. Other nutrients involved in energy
metabolism do cause similar deficiency symptoms. Riboflavin-containing coenzymes are
often coupled with nicotinamide-dependent reactions in the transfer of electrons, but ribo-
flavin deficiency is not very similar to pellagra. Iron is intimately involved in electron
transport and ATP production, and it functions closely with nicotinamide cofactors in
energy metabolism. Iron deficiency and pellagra both cause general weakness, which could
be associated with disrupted energy metabolism, but the pellagra patient displays several
unique and dramatic clinical characteristics. Whereas other nutrient deficiencies cause
dermatitis, only in niacin deficiency is this condition induced by exposure to sunlight.
Although thiamine deficiency also causes changes in energy metabolism and neural function,
it thought to achieve these effects through separate biochemical functions, and this
is likely the case with niacin as well.
Recently, insights into the function of nicotinamide coenzymes in metabolism have
improved our appreciation of the biochemical changes that may underlie the 4 D’s of
pellagra. In 1967, poly(ADP-ribosyl)ation was identified and recognized as a posttranslational
modification of nuclear proteins. Poly(ADP-ribose) synthesis makes use of NAD
as substrate, rather than as an electron-transporting intermediate. Poly(ADP-ribose) formation
has been shown to be important in DNA repair processes, and new knowledge in
this area may soon provide an explanation for the sun-sensitive dermatitis of pellagra.
Mono(ADP-ribosyl)ation, which also uses NAD as a substrate, was characterized, beginning
in the mid-1960s, as a mechanism of action for many bacterial toxins. Mono(ADPribosyl)
ation is now thought to be important in the endogenous regulation of many aspects
of signal transduction and membrane trafficking in eukaryotic cells, but these studies are
in their infancy. In 1989, cyclic ADP-ribose was identified as another product of NAD
metabolism and was shown to have the ability to regulate cellular calcium homeostasis,
a central process in neural transmission. Interestingly, the enzymes that make and degrade
cyclic ADP-ribose and the proteins that bind this second messenger are present in the
brain in relatively large quantities compared with other tissues. Even more recently, the
same enzymes that make cyclic ADP-ribose have been found to produce nicotinic acid
adenine dinucleotide phosphate (NAADP), using NADP as a substrate. NAADP appears
to also have distinct functions in the regulation of intracellular calcium stores. The studies
Niacin 217
of mono(ADP-ribosyl)ation reactions and cyclic ADP-ribose and NAADP function are in
their early stages, but they may soon provide explanations for the dementia of the pellagra
patient and the changes in intestinal cell function that lead to diarrhea.
A. Nicotinic Acid and Nicotinamide
The term niacin is accepted as a broad descriptor of vitamers having the biological activity
associated with nicotinamide, including nicotinamide, nicotinic acid, and a variety of pyridine
nucleotide structures. In the past, niacin has been used to specifically refer to nicotinic
acid (pyridine-3-carboxylic acid; Fig. 2A), but for the purposes of this discussion, ‘‘niacin’’
is used in reference to all forms with vitamin activity, while ‘‘nicotinic acid’’ refers
to pyridine-3-carboxylic acid. Nicotinic acid is a white crystalline solid, stable in air at
normal room temperature. It is moderately soluble in water and alcohol, but insoluble in
ether. An aqueous solution has a maximum ultraviolet (UV) absorbance at 263 nm.
Like nicotinic acid, nicotinamide (niacinamide; pyridine-3-carboxamide) (see structure,
Fig. 2B) is a white crystalline substance with a maximal UV absorbance at 263 nm.
In contrast to nicotinic acid, nicotinamide is highly soluble in water and is soluble in
ether—characteristics that allow separation of the two vitamers.
Fig. 2 Chemical structures of niacin compounds. (A) Nicotinic acid; (B) nicotinamide; (C) NAD;
(D) NADP; (E) site of reduction.
218 Kirkland and Rawling
B. Niacin Coenzymes
The biologically active forms of niacin compounds are the NAD and NADP coenzymes
(Fig. 2C, D). It is the C-4 position on the pyridine ring of the nicotinamide moiety that
participates in oxidation and reduction reactions. Due to the electronegativity of the amide
group and the nitrogen at position 1 on this ring, hydride ions can readily reduce the
oxidized C-4 position. This is the basis for the enzymatic hydrogen transfer reactions that
are ubiquitous among organisms. With respect to the non-redox functions of NAD, the
glycosidic linkage between nicotinamide and ribose is a high-energy bond, and cleavage
of this bond drives all types of ADP-ribose transfer reactions in the forward direction.
The oxidized and reduced forms of the coenzymes are designated NAD or NADP
and NADH or NADPH, respectively. The designations NAD and NADP are used to describe
the total pools. This is often necessary if the method of quantification does not
distinguish between oxidized and reduced forms, or if a general statement about the nucleotide
pool is made. The total pool of all four forms may be referred to as NAD(P). Both
NAD and NADP are white powders that are freely soluble in water and poorly soluble
in ether. Both compounds have strong UV absorption at 340 nm in their reduced forms,
with a weaker absorption at 260 nm when oxidized or reduced. The absorbance at 340
nm is often used to monitor the oxidation or reduction of these cofactors in enzyme assays.
A. Quantification
The traditional analysis for nicotinic acid entails cleavage of the pyridine ring with cyanogen
bromide (the Koenig reaction), followed by reaction with an aromatic amine to yield
a colored product that can be assayed spectrophotometrically (19). Because hydrolysis of
the nicotinamide pyridine ring with cyanogen bromide is not quantitative, it is advisable
to first deamidate nicotinamide enzymatically to form nicotinic acid before preparing samples
for the Koenig reaction. Microbiological assay of nicotinic acid and nicotinamide is
possible using Tetrahymena pyriformis or Lactobacillus plantarum (20). Nicotinic acid
can also be specifically determined microbiologically with Leuconostoc mesenteroides;
this microorganism is an obligate user of nicotinic acid rather than of nicotinamide (21).
Fluorometric measurement of nicotinamide is very sensitive. In this assay, nicotinamide
is converted to N?-methylnicotinamide, which is then reacted with a carbonyl compound,
commonly acetone or acetophenone, thus producing a fluorescent adduct on ring position
4 (22).
These traditional methods for analysis of nicotinic acid and nicotinamide are gradually
being replaced by a variety of new techniques, including gas chromatography and
mass spectroscopy (23) or high-performance liquid chromatography (HPLC) (24). Because
of the importance of these vitamers as pharmacological agents, more research on
pharmacokinetics following supraphysiological doses is warranted. Specific measurement
of nicotinic acid, nicotinamide, and their metabolites will require the use of these more
advanced analytical techniques.
Measurement of pyridine nucleotides is easier than measurement of the vitamin
precursors. Oxidized forms (NAD, NADP) are extracted by acid, usually 1 N perchlorate.
This causes destruction of the reduced forms (NADH, NADPH). The reduced nucleotides
are extracted by base, which causes destruction of the oxidized forms (25). The
Niacin 219
extracted nucleotides are generally quantified by enzyme cycling techniques that recognize
oxidized or reduced nucleotides but are specific to either NAD or NADP (25). The oxidized
forms can also be measured by HPLC techniques that provide additional data on
ATP, ADP, and AMP levels (26).
B. Food Content
The niacin content of human foods is usually expressed as niacin equivalents (NE), which
are equivalent to niacin content (mg)  one-sixtieth tryptophan content (mg). This relationship
may not be accurate if the intake of tryptophan is low (see Sec. IV.A). Furthermore,
the conversion of tryptophan appears to be improved when the intake of niacin is
low (2). Niacin in plant products is mainly in the form of nicotinic acid. Animal products
will initially contain mainly NAD and NADP coenzymes, but little is known about breakdown
of these nucleotides during the aging of meats. Under anerobic conditions, NAD
will be converted to NADH, which may protect it from the wide variety of NAD-catabolizing
enzymes in the cell. Analysis of food products for performed niacin is probably best
accomplished by microbiological methods that respond to nicotinic acid, nicotinamide,
and coenzyme forms of the vitamin. Because plant products contain less tryptophan than
animal products and because the nicotinic acid may be largely bound in unavailable forms,
some grain products such as breakfast cereals are supplemented with nicotinic acid. Such
supplementation, together with the widespread occurrence of niacin and tryptophan in a
mixed diet, has greatly diminished the number of clinically obvious cases of pellagra
in developed countries. Subclinical niacin deficiency may still be common in developed
countries (6,23), and clinical pellagra still occurs in association with alcoholism (27).
Outbreaks of pellagra continue in some areas of the world where populations rely on corn
(or maize) as a staple (5). The degree of nutrient deficiency in a population is always
based on the current perception of optimal intake or function. As our understanding of
niacin function evolves, we should be open to revising these end points and reevaluating
niacin status.
C. Dietary Requirements and Assessment of Status
Because of the important roles of NAD and NADP in energy metabolism, the niacin
requirement is related to energy intake. The recommended daily allowance (RDA) is set
at 6.6 NE/1000 kcal/day in the United States. The new Canadian RDAs for niacin range
from 2 mg niacin/day for infants to 18 NE/d during pregnancy, following a similar relationship
with energy requirement. There is also a minimum intake recommended for those
on low-energy diets, and this may reflect the roles of NAD beyond the typical redox
reactions of energy metabolism. Niacin status has traditionally been tested by measuring
the urinary excretion of various niacin metabolites or the urinary ratio of N-methyl-2-
pyridone-5-carboxamide to N-methylnicotinamide (28). The 2-pyridone form decreases
to a greater extent in response to a low dietary intake, and a ratio of less than 1.0 is
indicative of niacin deficiency. More recently, it has been found that the NAD pool in
red blood cells decreases rapidly during niacin deficiency in men, whereas the NADP
pool is quite stable (29). This has led to the suggested use of NAD/NADP, referred to
as the niacin number, as an easily obtained index of niacin deficiency in humans. Studies
using animal models have also shown that blood NAD pools deplete more rapidly and to
a greater extent than those of tissues such as liver, heart, or kidney (30). This suggests
220 Kirkland and Rawling
that a portion of blood NAD may represent a labile storage pool used to support other
tissues in the early stages of deficiency.
A. Pathways of Synthesis
Although many microorganisms and plants can synthesize the pyridine ring of NAD de
novo from aspartic acid and dihydroxyacetone phosphate (31), animals do not have this
ability. Nicotinic acid, nicotinamide, pyridine nucleotides and tryptophan represent the
dietary sources from which the pyridine ring structure can be derived by most mammalian
species. Animals may also practice caprophagy to take advantage of colonic synthesis of
niacin by microflora.
In 1958, Preiss and Handler (32) proposed a pathway for the conversion of nicotinic
acid to NAD in yeast and erythrocytes (shown in reactions 9, 5, and 6 of Fig. 3). Initially, it
was believed that nicotinamide was also metabolized through the Preiss-Handler pathway,
following the conversion of nicotinamide to nicotinic acid by nicotinamide deamidase
(reaction 8, Fig. 3). However, it was soon demonstrated by Dietrich that nicotinamide
reacts first with phosphoribosyl pyrophosphate and then ATP to produce NAD directly
(reactions 10 and 11, Fig. 3) (33). Subsequent studies supported this route by showing
that inhibition of NAD synthetase, the final enzyme in the Preiss-Handler scheme (reaction
6, Fig. 3), does not affect the incorporation of nicotinamide into NAD (34). It was also
found that deamidation of nicotinamide to nicotinic acid only occurs at supraphysiological
concentrations of nicotinamide (35), which is not surprising, as the Km of nicotinamide
deamidase for nicotinamide is above physiological nicotinamide levels. It has been concluded,
therefore, that the Dietrich pathway is the major route by which physiological
concentrations of nicotinamide supply NAD in mammalian cells (36). It should be noted,
however, that the predominant pathways of nicotinic acid and nicotinamide utilization
change when the vitamers are consumed in pharmacological quantities, such as those used
in the treatment of cardiovascular disease or insulin-dependent diabetes mellitus (IDDM).
This will be discussed later in this chapter (Sec. VII). Interestingly, there is a common
enzyme in the two routes of NAD synthesis. Nicotinamide mononucleotide (NMN) and
nicotinic acid mononucleotide (NAMN) adenylyltransferase activities reside within a common
protein in the nucleus; this enzyme catalyzes the last step in the conversion of nicotinamide
and the second to last step in the conversion of nicotinic acid to NAD (37).
Quinolinic acid, a tryptophan metabolite, can react with phosphoribosyl pyrophosphate
to produce nicotinic acid mononucleotide. This reaction occurs only in the kidney
and in the liver in mammals, due to the localization of quinolinate phosphoribosyltransferase
in these tissues (38). NAD is then synthesized from nicotinic acid mononucleotide
via the Preiss-Handler pathway. A number of reactions are required to convert tryptophan
to quinolinic acid. Most of the ?-amino-?-carboxymuconic-?-semialdehyde, an intermediate
in this pathway, is catabolized to acetyl CoA and CO2 (39). There is, however, a ratelimiting
step in this degradation sequence, which, when overwhelmed, allows some of
the semialdehyde to transform spontaneously into quinolinic acid (reaction 3, Fig. 3) and
ultimately to produce NAD (40). Therefore, the production of NAD from tryptophan is
favored by a low activity of one enzyme (picolinic carboxylase) and a high activity of
another (quinolinate phosphoribosyltransferase), leading to a wide range in the efficiency
of tryptophan utilization among species.
Niacin 221
Fig. 3 Pathways of NAD synthesis in mammals. Reactions 9, 5, and 6 comprise the Preiss-
Handler pathway, while reactions 10 and 11 form the Dietrich pathway. The following enzymes
correspond to the numbered reactions: (1) 5 step conversion; (2) picolinic carboxylase; (3) spontaneous
chemical reaction; (4) quinolinic acid phosphoribosyltransferase; (5) NAMN adenylyltransferase
(enzymes 5 and 11 may be identical proteins); (6) NAD synthetase; (7) NAD glycohydrolases,
various ADP-ribosylation reactions; (8) nicotinamide deamidase; (9) nicotinic acid phosphoribosyltransferase;
(10) nicotinamide phosphoribosyltransferase; (11) NMN adenylyltransferase.
In nutritionally replete humans, there is thought to be a 60 mg/1 mg ratio between
tryptophan supply and niacin formation, although individual variation is significant (41).
Because of this relationship, dietary niacin content is described in niacin equivalents (1
NE  mg niacin  1/60 mg tryptophan). More recent work has suggested that humans
may not utilize tryptophan for niacin synthesis when tryptophan levels are limiting in
the diet (29). In these experiments, young men were placed on a diet containing 6 NE/
day, and their blood NAD levels decreased by 70% over 5 weeks. Addition of 240 mg/
day of tryptophan to this diet had no effect on blood NAD, although it prevented the
decrease in plasma tryptophan that resulted from consumption of the unsupplemented
diet. This tryptophan represented, in theory, an additional 4 NE/day, but it appears that
222 Kirkland and Rawling
protein turnover takes precedence over niacin synthesis when tryptophan levels are low.
NE calculations for diets with marginal niacin status may be inaccurate because of this
relationship. At the same time, tryptophan supplements have been reported to cure pellagra
and a genetic defect in tryptophan absorption, known as Hartnup’s disease, also
causes pellagra-like signs and symptoms. While the efficiency of conversion of tryptophan
to NAD is thought to be increased by niacin deficiency (2), it is also apparent that
at low tryptophan intakes, the need for protein metabolism dominates over conversion
to NAD (29).
Species vary in their ability to synthesize niacin from tryptophan. Cats, adapted to
a diet high in amino acids, have approximately 50 times greater picolinic carboxylase
activity than humans (40), and they demonstrate extremely poor utilization of tryptophan
as a precursor for NAD (42). Rats, on the other hand, are more efficient than humans in
their use of tryptophan for NAD synthesis (ratio of about 33 mg tryptophan: 1 mg niacin)
(43). Mice are resistant to niacin deficiency, even on diets containing as little as 6% casein,
with 7% gelatin, a protein source lacking tryptophan (unpublished data). Guinea pigs are
very susceptible and will eventually die on a niacin-free diet containing 20% casein, demonstrating
a very poor conversion of tryptophan to niacin (44). The capacity for NAD
generation from tryptophan also depends on adequate status of vitamin B6 and minerals,
including copper, iron, and magnesium, all of which are necessary for reactions in this
pathway. When using experimental animals to model niacin deficiency in humans, it is
important to remember the tremendous variability between species in the utilization of
B. Absorption
Both nicotinamide and nicotinic acid can be absorbed through the stomach lining, but
absorption in the small intestine is more rapid. For intact nucleotides, pyrophosphatase
activity in the upper small intestine metabolizes NAD to yield nicotinamide mononucleotide,
which is then quickly hydrolyzed to form nicotinamide riboside and eventually free
Absorption of both nicotinic acid and nicotinamide at low concentrations appears
to be via sodium-dependent facilitated diffusion (39) or by carrier-mediated transport making
use of proton cotransporters and anion antiporters (45). Higher concentrations of both
forms appear to be absorbed by passive diffusion. Once absorbed from the lumen into the
enterocyte, nicotinamide may be converted to NAD or released into the portal circulation.
Although some nicotinic acid moves into the blood in its native form, the bulk of the
nicotinic acid taken up by the enterocyte is converted to NAD via the Preiss-Handler
pathway (39). As required, NAD glycohydrolases in the enterocytes release nicotinamide
from NAD into the plasma as the principal circulating form of niacin. The kinetics of
nicotinamide and nicotinic acid transport and metabolism are influenced significantly by
pharmacological intakes (see Sec. VII).
C. Distribution and Metabolism
Niacin compounds entering the portal circulation are either internalized by erythrocytes
or transported to the liver. Erythrocytes take up nicotinic acid by facilitated diffusion,
using an anionic transport protein that accounts for 30% of the membrane protein content
(39). In spite of this, erythrocytes appear to favor nicotinamide as a precursor for NAD
synthesis (46).
Niacin 223
The liver is the central processing organ for niacin. Aside from its role in the conversion
of tryptophan to NAD, it receives nicotinamide and some nicotinic acid via the portal
circulation, as well as nicotinamide released from other extrahepatic tissues. In the liver,
nicotinic acid and nicotinamide are metabolized to NAD or to yield compounds for urinary
excretion, depending on the niacin status of the organism.
The liver has some capacity for NAD storage. Because hepatic regulation of nicotinamide
phosphoribosyltransferse by ATP and NAD (positive and negative, respectively)
is less effective than in many other tissues, liver NAD concentrations increase significantly
following dietary nicotinamide administration (39). In rats, nicotinic acid and nicotinamide,
at 1000 mg/kg diet, increase blood and liver NAD to a similar and modest extent
(about 50%) (47). NAD glycohydrolases are thought to use the storage pool of hepatic
NAD to produce nicotinamide (48), which is released for replenishment of extrahepatic
tissues. The function of NAD glycohydrolases is controversial, however, and difficult to
distinguish from cyclic ADP-ribose formation and catabolism (see later sections).
The liver plays an important role in the preparation of niacin for urinary excretion,
producing a variety of methylation and hydroxylation products of both nicotinic acid and
nicotinamide. In humans, nicotinamide is primarily methylated to produce N?-methylnicotinamide,
whereas nicotinic acid is conjugated with glycine to form nicotinuric acid. Increasing
levels of the untransformed vitamers can be found in the urine as the level of
niacin ingestion increases (39).
A. NAD Cofactors in Redox Reactions
Oxidation and reduction of the C-4 position on the pyridine ring of NAD coenzymes
is the basis for hydrogen transfer reactions important in oxidative phosphorylation and
biosynthetic reactions. NAD is reduced to NADH in glycolytic reactions, oxidative decarboxylation
of pyruvate, oxidation of acetate in the tricarboxylic acid (TCA) cycle, oxidation
of alcohol, ?-oxidation of fatty acids, and a large number of other cellular oxidation
reactions. The electrons derived from these oxidation reactions are transferred to the electron
transport chain through the oxidation of NADH. The energy resulting from these
transfers is used to generate ATP. In contrast to the central role that NAD(H) plays in
energy expenditure, NADP(H) is essential for the biosynthetic reactions involved in energy
storage. NADPH is produced from reduction of NADP in reactions of the pentose phosphate
pathway and during the malate/pyruvate shuttle across the mitochondrial membrane.
NADPH then acts as a reducing agent for fatty acid production, cholesterol synthesis,
and manufacture of deoxyribonucleotides. NADPH is also notable for its involvement in
glutathione regeneration.
The extra phosphate group carried on NADP permits the cell to separate the oxidation
and reduction activities of the nicotinamide cofactors by allowing most cellular enzymes
to be specific for one of these coenzyme species. Because of this specificity, the
majority of cellular NADP is maintained in the reduced state by the pentose phosphate
pathway and reduction reactions are favored by mass action. Conversely, NAD is predominantly
oxidized, improving the oxidant capabilities of this cofactor under cellular conditions.
The role of NAD coenzymes in redox reactions has been the function classically
associated with niacin since Warburg and Christian showed that nicotinamide was a component
of NADP in 1935 and, in the following year, demonstrated the presence of nicotin224
Kirkland and Rawling
amide in NAD (49). It is probable that early researchers attributed the pathology associated
with niacin deficiency to disruptions in redox cycling, since this was the only function of
niacin known at that time. However, the distinctive clinical signs of pellagra may be better
explained in relation to the functions of NAD described in the following sections.
B. NAD as a Substrate
Although the hydride transfer chemistry of the pyridine nucleotides was initially described
in the 1930s (49), the high-energy glycosidic linkage between nicotinamide and ADPribose
received little attention before the discovery of mono- and poly(ADP-ribosyl)ation
reactions in the 1960s. The energy provided by breaking this bond allows the addition of
ADP-ribose to a variety of nucleophilic acceptors. These include glutamate side chains
and hydroxyl groups of ribose for poly(ADP-ribose) synthesis, a variety of amino acid
side chains in mono(ADP-ribosyl)ation reactions, and an internal ribose linkage for cyclic
ADP-ribose synthesis (Fig. 4). The catabolic activity of NAD glycohydrolase (if this reaction
does exist in isolation) also represents an ADP-ribose transfer, in which water acts
as the nucleophilic acceptor.
C. ADP-Ribose Cyclization and NAADP Synthesis
In 1987, a metabolite of NAD was found to cause intracellular calcium mobilization in
sea urchin eggs (50). The molecule was similar to the potent second messenger inositol-
1,4,5-triphosphate (IP3) in its capacity to promote this response. It acted, however, on
different cellular pools of calcium. Eventually, the compound was identified as cyclic
(ADP-ribose) (51). Cells have two types of channels, the IP3 and the ryanodine receptors,
which regulate the release of calcium from internal stores such as the sarcoplasmic or
endoplasmic reticulum and mitochondria. While the IP3 receptors have been characterized,
the ryanodine receptors have been identified by their nonphysiological stimulation by the
plant alkaloid, ryanodine. In the last few years, evidence has accumulated that cyclic ADPribose
is an endogenous second messenger that controls the ryanodine receptors (52) and
that the release of calcium is due to the binding of cyclic ADP-ribose to a small protean,
which then dissociates from the ryanodine receptor, allowing calcium transport (53).
Calcium concentrations are about 10,000-fold higher outside cells than within the
cytoplasm. Transient increases in cytoplasmic calcium are critical to most types of cell
signaling, including nerve transmission and muscle contraction. When a stimulus causes
a small amount of calcium to be released, there is a positive feedback leading to a much
larger flow of calcium into the cytoplasm. Cyclic ADP-ribose and ryanodine receptors
appear to play an important role in this calcium-induced calcium release (CICR) (52).
ADP-ribosyl cyclase and cyclic ADP-ribose hydrolase are the enzyme activities responsible
for synthesis and degradation of cyclic ADP-ribose, respectively. NAD glycohydrolase,
isolated from canine spleen, actually has both cyclase and hydrolase activities,
leading to an overall reaction that appears to be NAD glycohydrolysis (54). The leukocyte
cell surface antigen CD38 also has these capacities (55). It is possible that other enzymes
previously identified as NAD glycohydrolases are similar in their abilities to synthesize
and catabolize cyclic ADP-ribose.
At the moment, it is unclear whether cyclic ADP-ribose works as a cofactor to
enhance the sensitivity of ryanodine receptors to calcium (as does caffeine) or if it is a
true second messenger that responds directly to an extracellular stimulus (56). In rat pancreatic
? cells, glucose stimulates the formation of cyclic ADP-ribose, which is thought
Niacin 225
Fig. 4 Structures and origin of cyclic ADP-ribose and NAADP.
to induce insulin secretion via CICR (57). In some systems, there is redundancy between
the function of cyclic ADP-ribose and IP3; in sea urchin eggs, for example, the actions
of both must be inhibited to prevent fertilization (52).
In 1995, a metabolite of NADP was also found to cause release of intracellular
calcium stores; it was identified as nicotinic acid adenine dinucleotide phosphate
(NAADP). Surprisingly, this metabolite is formed by the same enzymes that make cyclic
ADP-ribose, although the reaction mechanism seems quite different. NAADP is synthesized
by exchanging the nicotinamide on NADP with nicotinic acid, forming one of the
most potent calcium-releasing agents known (58). This metabolite is effective at nanomolar
concentrations and appears to be physiologically important in various calcium release
systems, including sea urchin eggs, brain microsomes, and pancreatic cell insulin release
(58). Interestingly, NAADP appears to regulate intracellular calcium pools that are distinct
from those controlled by IP3 or cyclic ADP-ribose.
226 Kirkland and Rawling
There are many questions to be answered concerning the roles of cyclic ADP-ribose
and NAADP in cell signaling and how they might be affected by niacin deficiency. ADPribosyl
cyclase activities are widespread; CD38 has been identified in the plasma membrane
of a wide variety of mammalian tissues, and a soluble cyclase has been characterized
from the testis of Aplysia and of dog (59). Another cyclase enzyme from canine spleen
is localized to the endoplasmic reticulum (54). CD38 presents a unusual picture in that
the catalytic domain is extracellular. This ectoenzyme may obtain NAD or NADP from
extracellular fluids and release cyclic ADP-ribose or NAADP to interact with extracellular
receptors. Cyclic ADP-ribose or NAADP may also enter the cell to interact with intracellular
binding proteins. Conversely, CD38 could be internalized under appropriate conditions
and function as an intracellular source of cyclic ADP-ribose, drawing on cytosolic NAD
or NADP pools. To understand the effect of niacin deficiency on these processes, we
need to determine the subcellular localization of the active forms of these enzymes and
their affinity for nucleotide substrates. Later in this chapter we will try to integrate these
functions with other NAD-utilizing reactions during niacin deficiency.
The tissue distribution of cyclic ADP-ribose– and NAADP-metabolizing enzymes
is also interesting. Levels of CD38 and cyclic ADP-ribose hydrolase activities are particularly
high in brain tissue (59,60), and preliminary data show that the small proteins that
bind cyclic ADP-ribose are also high in brain tissue (61). Researchers may find over the
next few years that neural functions mediated by cyclic ADP-ribose and NAADP are
impaired during niacin deficiency and lead to the unusual motor disruptions and dementia
of pellagra.
D. NAD Glycohydrolysis
NAD glycohydrolase enzymes catalyze the donation of the ADP-ribose moiety of NAD
to water, resulting in free ADP-ribose and nicotinamide. There have been many reports
of this activity in mammalian plasma membranes and in the cytosol (62). However, diffi-
culty in isolating the enzymes has prevented clear identification of their form and metabolic
function. The function of these enzymes has never been clearly understood but may
become clearer with the discovery that cyclic ADP-ribose formation and hydrolysis are
often catalyzed by the same enzyme, leading to the misleading appearance of simple NAD
hydrolysis. It is possible that CD38 and other cyclase/hydrolase complexes are identical
to the elusive membrane-bound NAD glycohydrolase (54). To further complicate this
issue, poly(ADP-ribose) polymerase (63), bacterial ADP-ribosylating toxins (64), and endogenous
mono(ADP-ribosyl) transferases (65) all have the ability to use water as an
acceptor of the ADP-ribose moiety, in lieu of a protein, thus leading to glycohydrolaselike
activity. If true NAD glycohydrolase activities exist, their function may be related to
the control of intracellular NAD levels and to the availability of nicotinamide for export
from tissues such as small intestine and liver (62).
E. Mono(ADP-ribosyl)ation
Mono(ADP-ribosyl)ation is the transfer of a single ADP-ribose moiety, derived from
NAD, to an amino acid residue on an acceptor protein (66,67). Of these enzyme systems,
described in bacteriophages, bacteria, and eukaryotes, bacterial toxins are the best characterized.
Various GTP-binding proteins (G proteins) are the targets for Cholera, Pertussis,
Diphtheria and Pseudomonas toxins. Mono(ADP-ribosyl)ation of G?s by Cholera toxin
results in the stimulation of adenylate cyclase activity, leading to disruption of ion transNiacin
port in intestinal epithelial cells and the diarrhea characteristic of this disease. Pertussis
toxin catalyzes the cysteine-specific ADP-ribosylation of several other forms of G proteins,
also leading to an uncoupling of their activities from the associated receptors. Diphtheria
toxin and Pseudomonas exotoxin A ADP-ribosylate a posttranslationally modified histidine
residue, called diphthamide, on polypeptide elongation factor 2 (EF-2), also a G
protein. This single alteration in EF-2 structure renders the enzyme nonfunctional and
halts protein synthesis. Clostridium toxins ADP-ribosylate various actin monomers to prevent
actin polymerization.
The profound metabolic changes elicited by the bacterial toxins indicate that mono
(ADP-ribosyl)ation is a powerful modulator of protein function. Mammalian cells are
believed to contain a variety of endogenous mono-ADP-ribosyltransferases, but most of
these data are based on the labeling of proteins in the presence of radioactive NAD, and the
experiments are usually conducted under nonphysiological conditions. Several transferase
activities have been examined in more detail, including those that modify EF-2, grp78,
rho, and a more general family of arginine-specific transferases. It appears that the list of
endogenous transferases will continue to grow as methodology allows researchers to identify
their protein substrates. Several of the endogenous monotransferase enzymes regulate
the function of G proteins by changing their interaction with GTP, but this is not a universal
mechanism. There is also a broad spectrum of enzyme chemistry, with five or more amino
acid side chains acting as ADP-ribose acceptors. Some of the better characterized cellular
transferases are discussed in separate sections, below.
The best characterized of these activities, with respect to enzyme purification, are
the NAD:arginine ADP-ribosyltransferases and the opposing ADP-ribosylarginine hydrolases.
Several transferases and hydrolases have been purified and characterized from
different species and tissues (67,68). Different transferase proteins are found in nuclear,
cytosolic, and plasma membrane fractions. The enzymes have a fairly broad substrate
specificity in vitro, making it difficult to identify the physiologically important protein
acceptors. Some of the possible acceptor proteins include histones, cytosolic actin, and
Ca2 ATPase (69), but the metabolic roles of these changes in intact cells or tissues are
uncertain. All of the acceptors are modified on arginine side chains, and most of the
enzymes will ADP-ribosylate free arginine in vitro.
The plasma membrane form of NAD:arginine ADP-ribosyltransferase is generating
the most interest. This enzyme is anchored, via glycosylphosphatidylinositol (GPI), to the
outer surface of the plasma membrane (67). The GPI-anchored form of the enzyme has
been identified in several cell types. A recent report has shown that extracellular NAD
causes a depression in cytotoxic T-cell proliferation and that this response is dependent
on the extracellular GPI-anchored transferase. The ecto activity appears to lead to the
mono(ADP-ribosyl)ation of an intracellular protein (70). This curious chain of events is
similar to the ecto activity of the CD38 family of ADP-ribosyl cyclase enzymes, and it
raises further questions about the role of extracellular NAD pools and how they might
respond to niacin deficiency.
EF-2 is a G protein that is essential to the activity of the peptide elongation cycle. It
has a specific histidine residue that is modified posttranslationally to a unique diphthamide
structure. No other proteins are known to carry this amino acid residue. Furthermore, the
formation of diphthamide on EF-2 occurs in all eukaryotic cells, although it is not necessary
for the elongating activity of EF-2. Diphtheria toxin ADP-ribosylates this diphthamide,
rendering EF-2 inactive and leading to the clinical manifestations of diphtheria infection.
With the idea that humans did not evolve diphthamide synthesis on EF-2 as a site
228 Kirkland and Rawling
for attack by diphtheria toxin, investigators have searched for an endogenous regulatory
system that uses a similar mechanism. Because of an inability to accurately measure the
ADP- ribosylation of EF-2, progress has been slow. It appears that EF-2 is modified by
an endogenous transferase when cultured cells are growth-arrested through serum deprivation
(71). Interestingly, this may be an automodification, as the transferase activity copuri-
fies with EF-2.
Rho is a small molecular weight G protein that is involved in cell division and
differentiation through interactions with the cytoskeletal system. The Clostridium botulinum
C3 toxin ADP-ribosylates rho on a specific asparagine side chain, leading to disruption
of the actin cytoskeleton. An endogenous transferase activity has been identified in
brain cytosol that appears to be specific to rho as an acceptor and modifies the same
asparagine residue (72). No physiological conditions have been identified that cause regulation
of rho ADP-ribosylation.
One of the most interesting proposed mechanisms of cellular regulation using mono-
ADP-ribosylation involves the 78-kDa glucose-regulated-protein (GRP78). GRP78 is a
molecular chaperone that aids in the correct folding of secreted proteins in the lumen of
the endoplasmic reticulum. GRP78 may also bind incorrectly glycosylated proteins and
prevent their secretion. When cultured hepatoma cells are deprived of glucose or an essential
amino acid, GRP78 is mono(ADP-ribosyl)ated by a noncharacterized endogenous
transferase. This modification is freely reversible, suggesting the presence of a hydrolase
enzyme. The ADP-ribosylated form of GRP78 appears to be inactive in its chaperone
functions, and the authors suggest that this is a rapid mechanism to decrease the rate of
protein secretion during times of nutritional stress (73). Of interest, GRP78 expression is
increased in cultured cells grown in niacin-deficient medium (74).
Cysteine-bound ADP-ribose residues have been found in the membrane fraction of
liver tissue (75). An enzyme purified from red blood cell membranes has the ability to
ADP-ribosylate a cysteine residue of Gi?. These observations suggest that an endogenous
transferase similar to Pertussis toxin participates in the regulation of adenylate cyclase
There are potential artifacts in the study of mono(ADP-ribosyl)ation reactions. For
some time it was believed that nitric oxide induced an endogenous mono(ADP-ribosyl)
ation of glyceraldehyde-3-phosphate dehydrogenase (76). Research eventually showed
that this was actually a nonenzymatic reaction combining an S-nitrosylation with a subsequent
addition of the whole NAD molecule to the active site (77). Free ADP-ribose also
reacts nonenzymatically with a variety of amino acid side chains, creating the illusion of
mono(ADP-ribosyl)ation (78). Careful determination of the enzymatic nature of proposed
mono(ADP-ribosyl)ation reactions will be necessary to avoid artifacts in the future. At
the same time, it may turn out that nonenzymatic reactions of NAD or ADP-ribose with
proteins are physiologically relevant and that these types of reactions could be affected
by dietary niacin status.
F. Poly(ADP-ribosyl)ation
The presence of a novel adenylate-containing compound in liver nuclear extracts was
described by Mandel’s group in 1966 (79). They identified it as a polymer of ADP-ribose
and later confirmed that NAD was the substrate and that nicotinamide was released in the
reaction. The structure, known as poly(ADP-ribose), was found to be covalently attached
Niacin 229
Fig. 5 Synthesis and degradation of poly(ADP-ribose).
to nuclear proteins. [See Figure 5 for an illustration of poly(ADP-ribose) structure and
1. Poly(ADP-ribose) Synthesis
Poly(ADP-ribose) polymerase (PARP; EC was the first enzyme identified that
had the ability to synthesize poly(ADP-ribose). PARP contains three functional domains
(80). A 42-kDa DNA-binding domain, containing two zinc fingers, is located at the amino
terminus. Together, these zinc fingers allow the enzyme to bind specifically to strand
breaks in DNA and signal the catalytic portion of the protein to initiate poly(ADP-ribose)
synthesis (81). The 55-kDa carboxy terminus region contains the NAD binding and catalytic
sites. Although more than 30 nuclear proteins may act as acceptors, most of the
230 Kirkland and Rawling
poly(ADP-ribose) is synthesized on PARP itself (referred to as automodification). The
middle section of the amino acid sequence of PARP includes a 16-kDa region that has
been identified as an automodification domain. Here there are 15 sites on which PARP
synthesizes poly(ADP-ribose), although more sites for automodification have been found
outside of this area of the protein (82). The synthesis of poly(ADP-ribose) on PARP itself
is critical in the regulation of its interactions with DNA.
The presence of a DNA strand break is detected by the zinc finger close to the amino
terminus, while the second finger is required for catalytic activation (81). Activated PARP
initiates poly(ADP-ribose) synthesis by covalently linking the ADP-ribose portion of an
NAD molecule to a glutamate or aspartate residue on an acceptor protein, with the release
of nicotinamide. A linear sequence of ADP-ribose units is synthesized on the initial protein-
bound monomer. At intervals of 40–50 residues, branch points are created on the
parent polymer chain and subsequently serve as sites for elongation (83). These initiation,
elongation, and branching reactions are all carried out by the catalytic domain at the carboxy
terminus of PARP. Automodification of PARP is now thought to occur by two
PARP molecules working together as a dimer (84). As PARP becomes more poly(ADPribosyl)
ated, it takes on an increasingly negative charge because of the accumulation of
phosphate groups. This creates electrostatic repulsion between automodified PARP and
DNA, causing PARP to dissociate from the DNA nick and stop catalytic activity (85).
It is important to note that inhibition of PARP activity by treatment with competitive
inhibitors or removal of NAD from in vitro systems is very different from removing PARP
from the system. Inactive PARP binds to strand breaks, preventing access by repair enzymes
and possibly impeding the signals that the cell uses to regulate DNA replication
and the cell cycle in response to DNA damage. It is not known whether niacin deficiency
in vivo causes a similar situation to occur. This depends upon the degree of depletion of
NAD and the response of other aspects of poly(ADP-ribose) metabolism, including the
rate of degradation of poly(ADP-ribose).
2. Poly(ADP-ribose) Degradation
It has been suggested that two different enzymes are required to degrade poly(ADPribose).
Poly(ADP-ribose) glycohydrolase cleaves the poly(ADP-ribose) in a combined
endo-/exoglycosidic fashion to release free ADP-ribose units and some free oligomers
(86). ADP-ribosyl protein lyase appears to release the final ADP-ribose residue from the
acceptor proteins, but this activity is poorly characterized. Free ADP-ribose is rapidly
degraded to AMP and ribose-phosphate by pyrophosphatase activity; pyrophosphatase
could be considered a third enzyme involved in poly(ADP-ribose) catabolism (87). It is
not certain why free ADP-ribose is catabolized so rapidly, although it may be to encourage
the forward activity of glycohydrolase or to limit the nonenzymatic glycation of proteins,
as discussed previously (78). The activity of the glycohydrolase is modified by the nature
of the poly(ADP-ribose), with a lower Km and a higher Vmax for longer poly(ADP-ribose)
chains (88). On short to medium-length chains, PARP and glycohydrolase will compete
for the free end of the poly(ADP-ribose) chain. As the ADP-ribose polymer chains elongate,
PARP activity decreases (resulting from automodification) and glycohydrolase affinity
increases, favoring catabolism.
3. Metabolic Roles of Poly(ADP-ribosyl)ation
The single-strand breaks required to activate PARP are produced in vivo through the
processes of DNA replication, transcription, and repair (89). The role played by poly(ADPNiacin
ribose) metabolism in these processes has been best characterized in DNA repair. Durkacz
et al. (90) demonstrated that poly(ADP-ribosyl)ation was required for cells in culture to
repair DNA damage caused by the carcinogen dimethyl sulfate. Exposure of the cells to
3-aminobenzamide (3-AB), a PARP inhibitor, and depletion of cellular NAD (through
the use of nicotinamide-free culture medium) resulted in the same inability to recover from
DNA damage. This landmark study stimulated a wide variety of research to determine
the roles of poly(ADP-ribose) metabolism in the DNA repair process and provided new
perspectives on the possible effects of niacin deficiency in the whole animal.
4. Mechanisms of Action
Poly(ADP-ribose) contributes negative charge to acceptor proteins, and the number of
ADP-ribose units determines the magnitude of this negative charge. The cell controls
poly(ADP-ribose) chain length through the actions of PARP and poly(ADP-ribose) glycohydrolase.
It would appear that the function of poly(ADP-ribose) is dependent on this
anionic nature, which encourages electrostatic repulsion from other polyanions, such as
DNA, and attraction to cations, such as basic DNA-binding proteins.
Quantitatively, much of the poly(ADP-ribose) in whole cells and tissues is associated
with PARP and histones. The extranucleosomal histone, H1, and histone H2B are
modified, leading to localized disruption of the DNA–protein interactions within and
among nucleosomes (89). In a process referred to as ‘‘histone shuttling,’’ the automodifi-
cation of PARP may also contribute to chromatin relaxation by drawing nearby histones
away from the DNA (85). As a result of these modifications, poly(ADP-ribosylation) leads
to a localized relaxation of nucleosomal structure. The exposed DNA is presumably available
for interactions with other DNA-binding proteins such as helicases, topoisomerases,
polymerases, and ligases involved in replication or repair (89).
A number of studies have also demonstrated poly(ADP-ribose) synthesis on enzymes
involved in DNA repair (89). Poly(ADP-ribosyl)ation of most enzymes, including
DNA topoisomerases I and II and DNA polymerases ? and ?, results in inhibition of their
activities (89). In addition, the poly(ADP-ribosyl)ation of Ca2/Mg2-dependent endonuclease
has been shown to suppress the rapid, nonspecific DNA-degrading activity of this
enzyme (91), and this may be important in the regulation of apoptosis. In contrast, DNA
ligase, the enzyme required to anneal the strand ends once excision repair is complete,
appears to be stimulated by poly(ADP-ribosyl)ation (92). Noncovalent association of
PARP with DNA polymerase-? has been found to stimulate the activity of DNA polymerase-
? (93).
Many approaches have been used to determine the role of poly(ADP-ribose) metabolism
in DNA repair, replication, and transcription. In some cases, investigators have created
defined in vitro models, in which PARP can be added or removed, and its activity
can be controlled by the removal of NAD or the addition of inhibitors. However, these
systems lack the complexity of chromatin structure and interactions with the nuclear matrix
that exist in the whole cell. Lindhal and co-workers have used a simplified in vitro repair
system to show that the combined addition of PARP and NAD does not increase the rate
of excision repair. However, in the absence of NAD, PARP addition effectively blocks
repair because of its inability to automodify and leave the strand to allow for repair (94).
These authors have suggested that PARP is not involved in excision repair directly but
may play a role in preventing nonhomologous recombination events between two sites of
damage. In support of this, chemical inhibition of PARP in cells has been shown by others
to increase homologous recombination and sister chromatid exchanges (95). At the same
232 Kirkland and Rawling
time, it is apparent that the in vitro models do not contain the complexity of chromatin
structure, including nucleosomal organization, DNA supercoiling, and nuclear matrix interactions,
seen in intact cells. In attempts to extend these findings to more physiological
models, cells expressing PARP antisense mRNA to deplete PARP polypeptide (96) and
cells selected for very low PARP expression (97) have been studied. While both of these
approaches cause an increased susceptibility to DNA-damaging agents, they lead to con-
flicting conclusions regarding whether there is a direct role for PARP in excision repair
vs. a secondary role in the regulation of recombination events. An excellent way to determine
the role of a protein in the whole animal is to delete the gene using recombinant
techniques and, if possible, create a fully homozygous animal lacking any functional expression
of the gene. This is usually done in mice, and the modified strains are referred
to as ‘‘knockouts.’’ In the past 3 years a series of publications characterizing two new
PARP knockout mouse models has appeared (98). Up to 1 year of life, PARP-null mice
appear to be healthy and fertile. In one model, growth of pups is slower and litter size is
smaller, but PARP does not appear to be essential for normal development. However,
x-ray treated thymocytes from the animals are delayed in their recovery, compared with
thymocytes from controls, and there are increases in sister chromatid exchange frequency.
Both PARP knockout models show extreme sensitivity to radiation injury, and the one
that has been tested is very sensitive to the acute toxicity of a chemical carcinogen that
causes DNA alkylation. It will be important to examine the incidence of spontaneous
carcinogenesis in these mice in long-term experiments, as well as determining their response
to low levels of various carcinogens.
PARP-null mice created an unexpected breakthrough in this field through the finding
that they still made small amounts of poly(ADP-ribose), leading to the discovery of four
new enzymes that have PARP activity. These are distinct gene products that have eluded
discovery due to their small quantities and the small amount of poly(ADP-ribose) that
they synthesize. A larger (142 kDa) enzyme, called tankyrase, associates with telomeres
(99) and may help regulate the activity of telomerase or inhibit recombination events at
chromosome tips. Several PARP homologues are approximately 60 kDa in size, lack both
the traditional DNA-binding domain and the automodification zone, but appear to be activated
by DNA damage (100–102). The 193-kDa subunit of mammalian vaults is also
known to have the ability to synthesize poly(ADP-ribose) (103). Vaults are cytoplasmic
ribonucleoprotein complexes of unknown function, but the identification of a cytoplasmic
PARP activity explains some earlier observations regarding the existence of cytoplasmic
The effects of PARP removal are different from those obtained using inhibitors of
PARP activity. It is probably much better for cells to not express PARP at all than to
express inactive forms or to express PARP under conditions in which substrate is limiting.
Niacin deficiency may fall in this later category, as it has the potential to cause inactive
PARP molecules to accumulate on DNA strand breaks, inhibiting their repair. The probability
of mutations occurring would further increase if this binding also impaired the signaling
events that control cell cycling in the presence of DNA damage. There is evidence
that severe niacin deprivation of cells can prevent the DNA damage–induced accumulation
of p53, and the removal of PARP protein has a similar effect in this model (104). Other
experiments have suggested that p53 is a substrate for poly(ADP-ribosyl)ation and that
PARP and p53 may form a complex that binds to certain DNA sequences (98). Increased
p53 levels help to remove the cell from active cycling and either promote efficient DNA
repair or induce apoptosis. Further DNA replication in the presence of DNA damage favors
Niacin 233
the accumulation of mutations that lead to the neoplastic phenotype. Similarly, if apoptosis
of severely damaged cells is prevented, there is a strong tendency for neoplastic transformation.
The responses of the two PARP knockout models differ with respect to p53 induction.
5. DNA Replication and Transcription, Cellular Differentiation
Various roles have been suggested for poly(ADP-ribose) metabolism in other cellular
functions, but the experimental results are not as conclusive as those for DNA repair. The
amounts of PARP and poly(ADP-ribose) vary during the cell cycle, but inhibition of PARP
activity only changes DNA replication or cell division in some experimental models.
PARP could be functioning in the complex of proteins at the replication fork through
protein–protein interactions, without a requirement for poly(ADP-ribose) synthesis. As
mentioned earlier, PARP is known to have protein–protein interactions with DNA polymerase
?. If these roles do exist, there must be redundancy of function, given that the
PARP knockout mouse grows and develops to adulthood normally (98). Tankyrase, for
example, is an alternate PARP that associates with telomerase and could be important in
regulation of telomerase activity following DNA replication (99).
PARP appears to associate with areas of DNA that are undergoing active transcription.
PARP may modify members of the transcription complex, such as RNA polymerase,
or may stabilize cruciform structures in the DNA, but inhibition or removal of PARP from
cells does not appear to have dramatic effects on their overall capacity for transcription
of RNA (98). PARP itself or poly(ADP-ribose) formation may regulate specific genes;
PARP interacts directly with the AP-2 family of transcription factors (89) and also synthesizes
poly(ADP-ribose) on transcription factors that bind to AP-2 (105).
There are some dramatic effects in the area of cellular differentiation, especially in
cells of myeloid origin. PARP protein is absent from circulating neutrophils; this is the
only known example of a nucleated animal cell that does not contain this enzyme (106).
Acute promyelocytic leukemia cells provide a model for in vitro differentiation to either
neutrophils or monocytes. In this model, neutrophilic differentiation is also associated
with a disappearance of PARP protein (106), and when PARP is artificially expressed,
this pathway of differentiation is blocked (107). PARP also appears to play a causative
role in differentiation in other models using transformed cells in culture (82), but the
normal development of the PARP knockout mouse suggests that redundancy exists in the
whole animal. For all of these functions, it is important to remember the difference between
a loss of catalytic activity that could be caused by niacin deficiency and the removal
of PARP protein in the knockout model.
6. Relationship of Poly(ADP-Ribose) Metabolism to Carcinogenesis In
Given that poly(ADP-ribose) metabolism plays an important role in DNA repair in cultured
cells, one would expect PARP activity to be important in the in vivo process of
carcinogenesis. There are studies using competitive inhibitors that support this conclusion.
Concurrent treatment with 3-aminobenzamide (3-AB), a competitive inhibitor of PARP,
caused a 10-fold increase in the formation of altered hepatic foci (precancerous lesions)
in rats treated with diethylnitrosamine (108).
It is important to note, for much of the following discussion on PARP inhibition
and carcinogenesis, that there are two very different mechanisms in which PARP may be
234 Kirkland and Rawling
involved in cell death following DNA damage. In the first, low to moderate levels of DNA
damage cause a gradual accumulation of genetic changes that initially prevent the cell
from dividing and may eventually kill a portion of the cells through the loss of function
of essential genes or induction of apoptosis. If one or more mutations occur in genes
involved in cell division, this may also lead to the initiation of carcinogenesis. In a second
mechanism, large quantities of DNA damage activate PARP to such an extent that cellular
NAD is decreased to levels that do not support cellular functions, such as ATP synthesis,
and the cell dies quite rapidly. This type of cell death also tends to follow the pattern of
apoptosis and has been portrayed by some as the central mechanism for induction of
apoptosis via DNA damage. This is not an accurate generalization. Berger proposed the
‘‘suicidal NAD depletion’’ model as a mechanism to encourage cell death when DNA
damage was extreme (109). Under these conditions, the risk of carcinogenesis would be
high if cells survive. Inhibition of PARP increases cell injury in the first model because
of an inhibition of DNA repair processes. In the second model, however, PARP inhibition
can save cells from death by NAD depletion, at least in the short term. This can also lead
to an increase in carcinogenesis, due to the survival of cells with significant levels of
DNA damage.
An in vivo example of this second model is the chemical induction of insulin-dependent
diabetes mellitus (IDDM) in animals. Alloxan and streptozotocin are DNA-damaging
agents, working via hydroxyl radical formation and base alkylation, respectively, and are
fairly selective pancreatic ?-cell toxins. They are commonly used to induce IDDM in
animal models. In these models, alloxan and streptozotocin induce poly(ADP-ribose) synthesis
to the point of NAD depletion and cell death. Treatment with PARP inhibitors,
including nicotinamide and 3-AB, results in conservation of pancreatic ? cells and maintenance
of insulin secretion in rats treated with alloxan and streptozotocin (110). However,
the incidence of pancreatic ?-cell tumors in these animals is essentially 100% later in life
(111). Nicotinamide is a relatively weak PARP inhibitor (Ki  10 mM) (112), and it is
rapidly converted to NAD as it enters the cell, so it is unlikely that it accumulates to a
level at which it acts effectively as a PARP inhibitor in whole-animal models. Nicotinamide
probably functions under these conditions as an NAD precursor (47) and by preventing
extreme NAD depletion in damaged ? cells allows cells to survive and eventually
express a neoplastic genotype.
It is interesting to consider whether suicidal NAD depletion (as a direct response to
PARP activation) is relevant to human carcinogenesis. The extreme level of DNA damage
required to initiate this response suggests that it is not a major factor in chronic disease
states, but there are some interesting exceptions. The concentrated autoimmune attack that
initiates IDDM in humans appears to act by this mechanism (see Sec. VII for further
discussion). In addition, experiments with PARP knockout mice have shown that PARP
activation plays a major role in ischemia-reperfusion injury in the brain (98). The sudden
oxidant stress that occurs when blood flow returns to a hypoxic region of the brain causes
a PARP-dependent depletion of NAD, which leads to neuronal cell death. This model
may provide important clues to the effective treatment of stroke victims. However, most
human cancers result from a gradual accumulation of DNA damage under conditions in
which poly(ADP-ribose) synthesis rates are low and NAD depletion is modest. Even very
aggressive models of chemically induced carcinogenesis cause a high incidence of initiation
without severe depletion of NAD (113). If models of cell injury that cause suicidal
depletion of NAD are used to examine the effect of niacin deficiency, dramatic results
can be obtained (114). This is different from the situation in which a modest amount of
Niacin 235
DNA damage causes initiation of apoptosis and NAD depletion occurs as apoptosis progresses.
Cell culture models have a great contribution to make in understanding the relationship
between niacin status and cellular defense; many parameters can be controlled and/
or measured accurately, including NAD levels, degree of DNA damage, progression of
apoptosis, and transformation to a neoplastic phenotype. However, researchers do need
to relate their in vitro models to the degree of NAD depletion and levels of DNA damage
that are possible in the whole animal.
7. Dietary Niacin Deficiency, Poly(ADP-ribose) Metabolism, and
The native population of the Transkei region in South Africa has a high risk for esophageal
cancer (115). A maize-based, low-protein diet is staple for these people, and pellagra is
common. Esophageal ulcerations and esophagitis, common in pellagrins, have been associated
with development of carcinoma of the esophagus. Van Rensburg et al. found a greater
than fivefold increase in the risk of esophageal cancer in Zulu men who ate maize daily
(116). Frequent consumption of maize by natives of the Henan province of China is also
associated with increased esophageal cancer risk (117). Consumption of maize in northeastern
Italy was associated with increased risk of oral, pharyngeal, and esophageal cancers,
especially with heavy consumption of alcohol (118). The frequency of esophageal
cancer appears to increase when maize replaces sorghum as a basic dietary component
(115,119). Low niacin intake in an American population, in western New York state, was
also associated with increased oral cancer risk, when controlled for smoking and alcohol
consumption (120), which appeared to be the initiating factors in the disease. This is
interesting in that the patients were not deficient in niacin to the degree of showing clinical
symptoms of pellagra, and suggests that subclinical niacin deficiency may increase cancer
The Linxian province of northern China also has a very high rate of esophageal and
gastric cancers. Recently, a series of nutritional intervention trials was conducted in this
area (121,122). During 5 years of intervention, a combined supplement of ?-carotene,
vitamin E, and selenium decreased total mortality, total cancer mortality, and stomach
cancer mortality. A parallel group receiving riboflavin and niacin did not show any bene-
fits. It is important to note that this population has traditionally been dependent on rice
as a staple rather than maize, and niacin deficiency is less common than thiamine or
selenium deficiency. There is an important need to design an intervention trial in an area
dependent on maize as a staple, with a significant incidence of existing niacin deficiency
in the population.
A limited number of animal experiments have been conducted to test whether niacin
deficiency plays a causal role in the process of carcinogenesis. Miller and Burns studied
the interaction of niacin deficiency, protein-energy malnutrition, and renal carcinogenesis
in rats (123). The diets were low in tryptophan and total protein (25% of recommended
intake), with nicotinamide at zero, requirement, or 10 times requirement levels. The diets
did not affect tumor size or number, although death due to renal tumor burden may have
been accelerated in the deficient animals. Unfortunately, pyridine nucleotides in the liver
and kidney were not decreased by the deficient diet, and it is difficult to determine if niacin
deficiency had any impact in this experiment. This experiment highlights the difficulties in
working with a nutrient that can be synthesized from an amino acid, requiring, in most
species, an imbalanced amino acid diet to create the vitamin deficiency. Parameters must
be controlled carefully to get the desired nutritional status.
236 Kirkland and Rawling
Van Rensburg et al. (124) developed an animal model of malnutrition and esophageal
cancer, induced by N-nitrosomethylbenzylamine (a DNA-alkylating agent) in cornfed
rats. Addition of 20 mg of nicotinic acid per kilogram of basal diet resulted in reduction
of tumor incidence, size, and progression compared with that seen in rats fed the basal
diet alone. NAD and poly(ADP-ribose) levels were not measured in this study. Another
study has shown that lymphocytes from niacin-deficient rats were more susceptible to
oxygen radical–induced DNA damage (125).
We have developed a model of niacin deficiency in rats that maintains a positive
growth rate. The diet is based on components with minimal niacin content and a mixture
of casein (7%) and gelatin (6%) as sources of protein, designed to limit tryptophan content.
Rats fed this diet develop clinical signs of deficiency, including dermatitis, diarrhea, and
ataxia, and also have decreased hepatic NAD and poly(ADP-ribose) levels (30). However,
when treated with DEN, hepatic poly(ADP-ribose) accumulation was not affected
by niacin deficiency, and there was no long-term effect of diet on the development of
preneoplastic altered hepatic foci (126). It is apparent that hepatic NAD levels, which
decreased from about 900 to about 600 µM in deficient rats, were still adequate to support
the activity of PARP, which has a Km in the range of 20–80 µM (127).
It is likely that, during niacin deficiency, NAD depletion is not uniform in different
organs and tissues. The symptoms of pellagra demonstrate tissue specificity, like most
nutrient deficiencies. One factor that can cause more rapid nutrient depletion is cell turnover.
The bone marrow has the most rapid rate of cell turnover in the body, with a doubling
time of about 12 hours, and these cells leave to other sites in the body, exporting nutrient
resources with them. With this in mind, we recently examined the effect of niacin status
on NAD and poly(ADP-ribose) metabolism in the bone marrow of rats. We found that
NAD depletion in the bone marrow is more dramatic than in any other tissue that we have
measured (20% of control) and that basal poly(ADP-ribose) content is decreased in the
niacin-deficient marrow to almost undetectable levels. The most common problem with
DNA damage to the bone marrow occurs during chemotherapy, so we started to treat
niacin-deficient and control rats with nitrosourea drugs that are known to cause bone marrow
suppression and induce leukemias in the long term. In this model of the side effects
of chemotherapy drugs, niacin deficiency increased the severity of acute bone marrow
suppression (anemia, leukopenia) (128) and increased the rate of development of nitrosourea-
induced leukemias (129). These results may be important, because a large percentage
of cancer patients appear to be niacin deficient (130), and bone marrow suppression
and secondary leukemias are probably the two biggest problems in cancer therapy, beyond
curing the original disease.
Many more experiments have been conducted on the effect of pharmacological supplementation
of nicotinic acid or nicotinamide on various types of carcinogenesis. A wide
range of doses of nicotinamide has been used, in the presence and absence of exogenous
carcinogens, as reviewed by Bryan (131). Nicotinamide alone does not appear to present
risk as a carcinogen, although there may be some risk associated with its use in the prevention
of IDDM (further discussion in Sec. VII). When used in conjunction with carcinogenic
foods or compounds in animal experiments, nicotinamide has a confusing array of effects,
suggesting varied mechanisms of action. When given with diethylnitrosamine, nicotinamide
did not affect liver carcinogenesis, but it did increase kidney neoplasms. Following
streptozotocin treatment, nicotinamide decreased adenoma formation but increased pancreatic
islet cell tumors (111,131). Conversely, nicotinamide has shown significant protecNiacin
tive effects against bladder and intestinal cancers when provided in a diet containing
bracken fern.
There is much less information on the effects of large doses of nicotinic acid on
carcinogenesis. Surprisingly, the best information is probably from human studies, derived
from the long-term use of this compound in the treatment of hypercholesterolemia. Nine
years after a 6-year period of nicotinic acid use to treat hypercholesterolemic patients,
there was a significant decrease in mortality in this group, but this did not appear to be
due to a decrease in cancer incidence (132).
We have recently studied animals supplemented with nicotinamide or nicotinic acid
and treated with DEN (47). While both supplements increased liver NAD to a modest
extent in the absence of carcinogen treatment, only nicotinamide significantly increased
basal poly(ADP-ribose) levels (before DEN), whereas only nicotinic acid increased poly
(ADP-ribose) levels following DEN. Neither supplement affected the development of
DEN-induced preneoplastic altered hepatic foci. In contrast, pharmacological supplementation
of nicotinamide in mouse diets caused a dramatic increase in skin NAD and provided
significant protection against skin cancer induced by UV radiation (see Sec. V.F.9,
below). We have also shown that pharmacological supplementation of both nicotinic acid
and nicotinamide cause large increases in bone marrow NAD and poly(ADP-ribose) in
rats, and decrease the long term development of leukemia (unpublished data). As with
deficiency, it is not surprising that niacin supplementation affects cancer susceptibility in
some tissues, like the marrow and skin, and not others, like the liver. We need to continue
to build our knowledge of whole animal models to appreciate the complexities of niacin
metabolism and allow accurate recommendations for human populations.
Although some of the foregoing responses may have been due to changes in the
functions associated with NAD utilization, there are also a variety of pharmacological
actions that are not related to NAD synthesis. These may be responsible for the inconsistency
of the responses, and some of the potential mechanisms will be discussed in Sec. VII.
8. Niacin Status and Oxidant Lung Injury
Niacin supplementation can decrease the degree of lung injury and fibrosis from a variety
of causes, including exposure to lipopolysaccharide, cyclophosphamide, and bleomycin.
The best defined model uses bleomycin, an antibiotic chemotherapy drug that intercalates
with DNA and induces damage through the local production of oxygen radicals (133).
This is a very severe stress, which causes NAD and ATP depletion, perhaps leading to
suicidal NAD depletion, as hypothesized by Berger (109). Protection by niacin may be
functioning through the maintenance of NAD levels, but nicotinamide is more effective
in increasing NAD, while nicotinic acid is more potent in the reduction of lung pathology
(134). The protection of lung tissue with these supplements may allow much safer use of
bleomycin as a chemotherapy agent, although it is not known whether they will also protect
tumor tissue.
Hyperoxia is another popular model of oxidant stress in the lung and has clinical
relevance to the care of premature infants and patients with adult respiratory distress syndrome.
Hyperoxia has been shown to induce poly(ADP-ribose) synthesis in the lung, and,
although poly(ADP-ribose) synthesis is decreased by niacin deficiency, the deficient state
does not increase the severity of lung damage (135). Consistent with this finding, pharmacological
niacin supplementation is also ineffective in decreasing the severity of hyperoxic
lung damage (136). Why is this response different from that observed following bleomycin
238 Kirkland and Rawling
treatment? Bleomycin induces a more sudden and severe stress to pulmonary cells,
and it targets DNA specifically as an intracellular target, leading to depletion of NAD and
ATP. Hyperoxic damage occurs gradually over a period of 5 or more days (137) and
does not cause NAD depletion (135). The majority of oxygen radicals emanate from the
mitochondria and endoplasmic reticulum, and oxidative damage is distributed among the
intracellular compartments. Perhaps of greatest interest in this model, hyperoxia actually
increases lung NAD content in the niacin-deficient animal to almost that of niacin-replete
controls. This result strongly suggests that enzymes or transport systems involved in NAD
turnover are regulated in response to certain aspects of cellular damage, perhaps via the
oxidant stress response described for the induction of fos and jun (138). The degree of
oxidant stress and the time course of the stress, which differ between hyperoxia and bleomycin
toxicity, are likely to be important in determining the ability of cells or tissues to
9. Niacin Status and Skin Injury
Because of the sun sensitivity displayed by pellagrins, there has been a long-standing
interest in the potential of niacin to improve skin health (139). Unfortunately, very little
is known about the effect of niacin deficiency on the susceptibility to UV light or chemical
carcinogens. Rainbow trout are more prone to UV light–induced skin damage when niacin-
deficient (140), but little is known about other nonhuman species, or whether this
finding correlates with a change in ADP-ribose metabolism.
On the other hand, many studies have been conducted using pharmacological doses
of various forms of the vitamin. Nicotinic acid acts as a vasodilator in the skin, leading
to an increase in blood flow through the microvasculature. Interestingly, this occurs with
both oral and topical use and appears, in both cases, to be caused by changes in prostaglandin
production (141,142). Over the years, a wide variety of treatment regimens and different
forms of nicotinic acid have been used to treat various skin disorders (139). The effects
on blood flow occur only at supraphysiological levels of the vitamin and are probably not
caused by modulation of NAD pools. However, nicotinic acid supplementation at 1–10
g/kg of diet was shown to decrease UV-induced skin cancers in mice (143). The decrease
was linear through the supplementation range, as was the increase in skin NAD content.
This work shows that very large dietary doses of niacin may continue to influence cancer
susceptibility through modulation of NAD pools, although the authors also showed that
niacin supplementation improved immune surveillance of tumor cells.
Nicotinamide supplementation has also been shown to protect the skin from DNAdamaging
agents in some animal models. Large doses of nicotinamide, given intraperitoneally,
decreased the skin damage caused by sulfur mustard (144). Used as a chemical warfare
agent in the First World War, this is a DNA-alkylating agent: that causes edema,
necrosis, and microvesical formation in the exposed skin, associated with NAD depletion.
Interestingly, nicotinamide prevents the depletion of NAD and subsequent pathological
changes, without protecting against the earlier pathological changes that precede NAD
It seems obvious that the most critical cellular functions of the niacin-containing nucleotides
are those of electron transport and energy metabolism. A loss in the capacity to
deliver reducing equivalents to the electron transport chain would be similar to poisoning
Niacin 239
the cell with cyanide or suffocating from a lack of oxygen. It makes sense, then, that
these functions will be strongly protected when NAD levels start to deplete during niacin
deprivation. Cultured cells, in the absence of DNA damage, can grow and divide with
less than 5% of control NAD levels (90,104,145), leaving us with a variety of questions
to be answered. How do the other pathways of NAD utilization, including poly-, mono-,
and cyclic ADP-ribose formation, compete for these limiting substrate pools? What is the
nature of this competition at the cellular level with respect to compartmentalization between
the nucleus, cytoplasm, and mitochondria? What is the role of extracellular NAD
in the function of mono(ADP-ribosyl)transferase and ADP-ribosyl cyclase enzymes on
the outer surface of the cell? How are nicotinic acid and nicotinamide distributed among
tissues during deficiency, and does this contribute to the distinctive signs and symptoms
of pellagra? How do these interactions lead to the specific metabolic lesions that cause
the sun-sensitive dermatitis, diarrhea, and dementia?
Possible mechanisms for unequal utilization of NAD at the subcellular level include(
a) variation in the affinity of enzymes for NAD (Km) and (b) compartmentalization.
Km values are used to describe the affinity of an enzyme for its substrate and are defined
as the concentration of substrate required to support 50% of the maximal activity. A lower
Km indicates a higher affinity and suggests that an enzyme will compete effectively with
enzymes having a higher Km as NAD concentrations fall during deficiency. Some caution
in interpretation is required; enzyme kinetics may change during purification, especially
for membrane-bound proteins.
The Km of PARP for NAD is thought to be between 20 and 80 µM (127). In certain
cultured cells, the ability to synthesize poly(ADP-ribose) decreases when cellular NAD
content drops to less than half of control levels (145), showing that the synthesis of poly
(ADP-ribose) is one of the most sensitive pathways of NAD utilization. This is similar
to the proportionate decrease in NAD during niacin deficiency in many tissues in vivo,
but tissues vary in their absolute concentrations of NAD, and direct extrapolation to
intact tissues could be inaccurate. For example, in rats, liver NAD decreases by close
to 50% during deficiency but is still at about 500 µM absolute concentration (30).
It has been stated that poly(ADP-ribosyl)ation is the aspect of NAD utilization that
is most sensitive to niacin deficiency because of a much higher Km of PARP for NAD.
Do the Km values of other NAD-utilizing enzymes suggest that the sensitivity of poly
(ADP-ribose) metabolism may be unique? There are scores of dehydrogenase enzymes
that use NAD as an electron acceptor, producing NADH for utilization in the electron
transport chain. Glyceraldehyde phosphate dehydrogenase is a cytosolic enzyme that is
critical to the flow of substrates through glycolysis. It uses NAD as an oxidant and has
a Km for this cofactor of 13 µM (146), smaller than that of PARP, indicating a higher
affinity. Other cytosolic enzymes may have slightly higher affinities for NAD than PARP,
including alcohol and aldehyde dehydrogenases (17–110 µM and 16 µM, respectively)
(146). In the mitochondria, isocitrate is oxidatively decarboxylated in the TCA cycle by
a dehydrogenase with a Km for NAD of 78 µM(146), which is similar to PARP. However,
the mitochondrial form of malate dehydrogenase is also critical to the flow of substrate
through the TCA cycle, and its Km for NAD has been reported as 540 µM (146). If these
data are correct, the TCA cycle appears to require compartmentalization of NAD during
niacin deficiency, and the role of the mitochondria in this regard will be discussed below.
With respect to cyclic ADP-ribose synthesis, the purified microsomal cyclase from
canine spleen has a Km of 10 µM for NAD (54). With access to cytosolic NAD pools,
this enzyme should maintain its catalytic activity during niacin deficiency. The CD38
240 Kirkland and Rawling
cyclase is reported to have a Km of 15 µM for NAD (147). While this is a relatively high
affinity for substrate, the curious aspect of this enzyme is that it faces the exterior of the
cell. Does it have a requirement for extracellular NAD or access to intracellular pools?
Cultured kidney epithelial cells synthesize cyclic ADP-ribose, but they require permeabilization
to use NAD in the medium and require over 500 µM for a half-maximal response
(148). There are many questions to be answered in this area of research, but the potential
for cyclic ADP-ribose metabolism to be affected by niacin deficiency is worth considering.
Mono(ADP-ribosyl)transferases are a very diverse group. The only published data
on affinity for NAD refers to the arginine specific transferases. In a family of transferases
from turkey erythrocytes, two cytosolic enzymes have Km values of 7 and 36 µM, while
a transferase from the membrane fraction has a Km of 15 µM (67). These enzymes would
appear to compete with PARP under conditions of limiting NAD pools, but a transferase
from chicken liver nuclei has a Km of between 200 and 500 µM, and a transferase from
mammalian skeletal and heart muscle displays a Km of 560 µM (67). It appears that some
mono(ADP-ribosyl)transferases may be quite sensitive to niacin deficiency. The resulting
changes in cell signaling might not appear as problems in cell culture models but could
nevertheless present significant problems in the whole organism. As discussed earlier,
there is a mono(ADP-ribosyl)transferase anchored to the outer surface of the plasma membrane,
which, like CD38, appears to require extracellular NAD. This enzyme appears to
cause the ADP-ribosylation of an intracellular protein, leading to a depression in T-cell
proliferation. Although no attempt has been made to determine the Km of this enzyme for
NAD, levels of NAD as low 1 µM in the culture medium are effective in decreasing
cell proliferation (70).
It becomes apparent that the physical partitioning of NAD within the cell is a key
factor in the availability of the molecule for various metabolic functions. The cytoplasmic
pool provides substrate for soluble enzymes, as well as for those on the endoplasmic
reticulum and on the inside of the plasma membrane. These would support a host of redox
reactions and the activity of a variety of poorly defined mono(ADP-ribosyl) transferases
and ADP-ribosylcyclases. The mitochondria isolate a pool of NAD that is predominantly
involved in electron transport, although mono(ADP-ribosyl)ation reactions have been reported
in this organelle. Nuclear NAD is probably used mainly for poly(ADP-ribosyl)
ation reactions, but mono(ADP-ribosyl)transferases also are located here. The least studied
pool is extracellular NAD, which appears to play a role in some ADP-ribosylcyclase and
mono(ADP-ribosyl)transferase activities. How distinct are these pools, and how do they
respond to the progression of niacin deficiency?
Because of the presence of nuclear pores, it is unlikely that nuclear and cytosolic
NAD concentrations would differ to any great extent. However, the final step in the
synthesis of NAD from nicotinamide is catalyzed by a nuclear enzyme (Fig. 3) (36). The
same enzyme catalyzes the second to last step in the conversion of nicotinic acid to NAD,
but the last enzyme in this pathway is in the cytosol. This means that all of the NAD
synthesized in the cell from newly arrived nicotinamide, or from nicotinamide released
by any of the ADP-ribosylation reactions in the cell, will be available first to nuclear
reactions, of which poly(ADP-ribosyl)ation is likely to predominate. Nicotinic acid will
lead to the production of cytosolic NAD, which may favor different patterns of utilization.
The mitochondria are well equipped to regulate NAD levels. The inner mitochondrial
membrane is essentially impermeable to all forms of NAD(P). Reducing equivalents
in the form of NADH must be transformed via shuttle mechanisms to enter the mitochondria
for ATP production. How does the mitochondrion produce or obtain NAD, and what
Niacin 241
levels does it maintain? Some researchers believe that mitochondria synthesize NAD (149)
whereas others suggest that slow, high-affinity carriers bring the necessary NAD from the
cytosol (150,151). The net requirement is probably modest, as most of the reactions identi-
fied in this organelle do not degrade the cofactor. The important question concerns the
ability of mitochondria to concentrate NAD, and it appears that they have potent mechanisms
to accomplish this. NAD levels in hepatocyte mitochondria appear to be about
10-fold higher than in the cytoplasm, with absolute concentrations in the neighborhood
of 5 mM (152). These NAD concentrations would support enzymes with relatively low
affinities for NAD, such as malate dehydrogenase (146). With this ability to concentrate
NAD, the mitochondrial pool could be very well protected during niacin deficiency.
The plasma pool of NAD is poorly characterized. Levels of noncellular NAD in
blood samples are extremely low (23), and there is no information on the response of this
pool to dietary niacin intake. Further research will be required to determine if this source
of NAD has any physiological role.
In addition to the competition for NAD at the cellular level, organs and tissues vary
in their ability to conserve NAD pools or compete for precursors during the progression of
niacin deficiency (30). Various tissues also start with different levels of NAD, which
may act as reserves during deficiency. Blood NADP is more stable than NAD during
niacin deficiency (29), but it may change in other tissues, and the impact of niacin defi-
ciency on NAADP metabolism is not known. These are some of the concepts that must
be appreciated as we progress toward a better understanding of the biochemical basis of
the pathologies of pellagra, a disease whose clinical symptoms remain unexplained at the
molecular level.
Levels of niacin in excess of the RNI have been used in attempts to treat Hartnup’s disease,
carcinoid syndrome, poor glucose tolerance, atherosclerosis, schizophrenia, hyperlipidemia,
IDDM, and a variety of skin disorders. In some countries, during the shortages of
proper medical supplies caused by World War II, nicotinic acid became a popular drug
because the dramatic flushing reaction that it caused in the skin was interpreted as a sign
of the potency of the treatment (153). Currently, nicotinic acid and nicotinamide are used
mainly in the prevention of cardiovascular disease and IDDM, respectively. Very few
nutrients are prescribed medicinally in North America for pharmacological purposes that
are mechanistically distinct from their known nutrient functions. Both nicotinic acid and
nicotinamide fall into this category, and the pharmacological effects of these two vitamers
appear to be surprisingly unrelated.
Large oral doses of the two vitamers may be absorbed and distributed quite differently
than when provided at levels found in a normal diet. However, there is very little
information, especially from humans, on the pharmacokinetics of these compounds. In
normal diets, nicotinic acid is obtained mainly from plant products, while nicotinamide
and preformed nucleotides are derived mainly from animal-based food products. The intestine
and liver are active in the conversion of nicotinic acid to NAD and subsequently to
nicotinamide for release into the bloodstream, causing normal plasma levels of nicotinamide
to be greater than those of nicotinic acid (23). Large oral doses of nicotinic acid will
overcome this regulation, but the concentrations that reach peripheral tissues in patients are
poorly defined (153). There are two reasons for this: first, quantification of specific vitamin
forms is laborious, and, second, nicotinic acid was approved as a drug under a ‘‘grand242
Kirkland and Rawling
father’’ clause, which excused it from the rigorous testing required by potential new drugs
on the market.
In the case of nicotinamide, which is the main circulating form of the vitamin under
normal conditions, large doses can have a depressing effect on relative availability. This
is caused by the induction of deamidase enzymes in gut microflora and also by attaining
concentrations in the Km range of similar enzymes in the liver, leading to formation of
nicotinic acid (154). However, pharmacokinetic studies of nicotinamide in humans show
that plasma levels are significantly elevated by large oral doses (155). More work is needed
in defining the pharmacokinetics of both forms of niacin, including an assessment of their
effect on NAD pools and their turnover in various tissues.
A. Nicotinic Acid
1. Nicotinic Acid and Hyperlipidemia
Historically, nicotinic acid has been administered to patients with a variety of disorders,
often more for the dramatic skin reaction than proven curative powers. Its most successful
use is for the treatment of hyperlipidemia (156). At high doses, nicotinic acid causes
several changes in lipid and lipoprotein metabolism, including inhibition of lipolysis in
adipose tissue (157), inhibition of the synthesis and secretion of very low-density lipoprotein
(VLDL) by the liver (158), lowering of serum lipoprotein(a) levels (156), and increasing
serum levels of high-density lipoprotein (HDL) (159).
These mechanisms of action of nicotinic acid appear to be unrelated to the formation
of pyridine nucleotides or to the actions of nicotinamide. Decreased lipolysis in adipose
tissue is due to an inhibition of adenylate cyclase activity (157). The resulting drop in
cAMP levels leads to the decreased mobilization of fatty acids. It is not known whether
nicotinic acid binds to a cellular receptor or interacts directly with adenylate cyclase. The
decrease in fatty acid release from adipose tissue is at least partially responsible for the
drop in VLDL formation by the liver as well as for the subsequent drop in LDL levels,
although there may also be direct effects on liver lipid metabolism. Studies with radioactive
acetate suggest that nicotinic acid inhibits cholesterol synthesis at the level of
3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase (156), and this may also
play a role in the lowering of VLDL and LDL levels. Unlike many treatments for hyperlipidemias,
nicotinic acid also increases circulating levels of HDL (159), the beneficial lipoprotein
that removes cholesterol from vascular tissue, but the mechanisms involved are
uncertain. The skin flush of the face and upper trunk is a very rapid response to nicotinic
acid, and it appears to be caused by the local formation of prostaglandins (142), although
the underlying mechanism is unclear.
2. Nicotinic Acid Toxicity
There are a few drawbacks to using high levels of oral nicotinic acid. As mentioned above,
the short-term side effects may include vasodilation, burning or stinging sensations in the
face and hands, nausea, vomiting, and diarrhea. In the longer term, there may be varying
degrees of hyperpigmentation of the skin, abnormal glucose tolerance, hyperuricemia,
peptic ulcers, hepatomegaly, and jaundice (156). The chronic doses of time-release nicotinic
acid have been reported to cause more hepatotoxicity (160), but this is controversial.
It should be noted that all drugs used in the treatment of hyperlipidemia have some side
effects, many of which can be managed through changes in dose. Interestingly, nicotinic
acid use for 6 years by patients with cardiovascular disease led to a decrease in all-cause
Niacin 243
mortality measured 8 years after the drug use was discontinued (132). Frequently, the
choice of treatment is financially based, with the cost of nicotinic acid treatment being a
fraction of that of the newer medications.
B. Nicotinamide
In the past, nicotinamide has been used in the treatment of schizophrenia (161), but more
effective drugs have replaced it in this field. It is now being tested as a chemotherapy
agent; in this application, nicotinamide potentiates the cytotoxic effects of chemotherapy
and radiation treatment against tumor cells (162), an action that appears to be due to
increased blood flow and oxygenation of tumor tissue (163). However, most of the current
interest in nicotinamide involves its potential use in the prevention or delay of onset of
IDDM (156,164).
1. Nicotinamide and IDDM
Interest in this area started with the finding that nicotinamide could prevent diabetes induced
by the ?-cell toxins alloxan and streptozotocin (110). It was soon shown that the
? cells were killed by an interesting mechanism; severe DNA damage led to excessive
activation of PARP, which depleted cellular NAD levels to the point that ATP synthesis
could not be maintained. The researchers concluded that nicotinamide was preventing
diabetes by inhibiting PARP activity and preventing the depletion of NAD pools. Since
nicotinamide is not a very high-affinity inhibitor of PARP (112), and cellular levels tend
to stay low due to active conversion to NAD, it seems likely that protection in this model
was due to the use of nicotinamide as a precursor for NAD synthesis. We have shown
that large oral doses of nicotinamide increase NAD and poly(ADP-ribose) levels in the
liver (47) and more recently, in the bone marrow (unpublished results). It is unlikely that
oral nicotinamide could act as a PARP inhibitor in the pancreas, under similar conditions.
Interestingly, all of the animals protected from chemically induced diabetes by nicotinamide
developed insulin producing ?-cell tumors (111), a form of cancer that is particularly
lethal in humans (165). It is not surprising that cells rescued from NAD depletion
due to extreme DNA damage, either through inhibition of PARP or through pharmacological
support of NAD pools, would be at risk for neoplastic growth due to the survival of
cells with significant levels of DNA damage.
One concern in promoting the clinical use of nicotinamide in the prevention of
IDDM in humans is that there will also be a long-term risk of pancreatic cancer if survival
of ? cells can be maintained in the face of autoimmune attack. In spite of the theoretical
difficulties and the potential dangers of inhibiting PARP in the face of genotoxic damage
to the pancreas, many investigators in the diabetes field are comfortable with the theory
of PARP inhibition (166). Fortunately, there are other aspects of the response to nicotinamide
that paint a more optimistic picture, although not a very clear one.
In humans, the onset of IDDM occurs spontaneously by immune recognition of ?-
cell antigens. This is associated with leukocyte infiltration and the presence of anti–islet
cell antibodies in the serum. The spontaneous occurrence of IDDM is similar in the nonobese
diabetic (NOD) mouse. When nicotinamide is given to weanling NOD mice, the
onset of diabetic symptoms is prevented or delayed (167). The important distinction is
whether nicotinamide prevents ?-cell death by protecting against autoimmune attack or
by maintaining the ? cell following immune attack. There appear to be several ways in
which nicotinamide can prevent the initial damage to the cells of the pancreas. In some
244 Kirkland and Rawling
experimental models, nicotinamide treatment diminishes some aspects of the immune response
itself, including the ability of monocytes to attack cells labeled with antibodies
(168), the rate of movement of lymphocytes into tissues (169), and the production of
nitric oxide in response to inflammatory cytokines (170). Other studies have shown that
nicotinamide acts as a free-radical scavenger to protect the ?-cell targets even after the
immune response is directed against them (171,172). These mechanisms may provide an
explanation for the prevention of IDDM by nicotinamide and lead to more effective protocols
in the future.
In a very different experimental model, nicotinamide was found to decrease the
severity of diabetes in response to partial pancreatectomy. This appeared to be due to a
stimulation of ?-cell proliferation, leading to an increase in the size of islets (173). The
? cells normally have a low capacity for cell regeneration, and it has been suggested that
this predisposes to the development of human diabetes (110). When fetal pancreas is
pretreated with nicotinamide, there is an acceleration in the reversal of diabetes after transplantation
of the islet cells into diabetic nude mice (174). This is an important model for
future directions in the treatment of fully developed IDDM, and these experiments show
that nicotinamide has a positive influence on ?-cell regeneration and differentiation.
Nicotinamide may have an effect on mono(ADP-ribosyl)ation reactions, either by
acting as an inhibitor or by enhancing substrate (NAD) levels. In addition, there is a new
feeling that cyclic ADP-ribose could have an important role in immune regulation. Cyclase
enzymes like CD38 are important proteins in leukocytes, and drugs that interfere with
cyclic ADP-ribose regulation of calcium release are immune suppresors (166).
2. Clinical Trials
Encouraged by the type of data summarized above, a number of experiments have been
done with human subjects. In the majority of these studies, patients were recruited in an
early stage of clinically apparent diabetes. Since these subjects retain a varying degree of
?-cell function, it is not surprising that the results have been inconsistent. However, a
number of treatment protocols have been successful in inducing remission in some patients
(175) and increasing the residual level of plasma insulin for up to 2 years after diagnosis
The animal models show that nicotinamide treatment should start before the disease
process is in an advanced stage. To do this in human populations, researchers must identify
the susceptible population. Blood levels of islet cell antibodies, human leukocyte antigen,
and family history are used as predictors of IDDM for subject recruitment. Several intervention
trials starting in the early stages of the disease have produced interesting results.
In a 2-year study, 13 of 14 high-risk children treated with oral nicotinamide remained
disease-free, while all of the children in the control group became diabetic (177). In another
study, nicotinamide treatment for 8 months protected plasma insulin in high-risk
children, in spite of the fact that anti–islet cell antibody levels were not decreased (178).
This suggests that the autoimmune response is affected downstream of autoantibody production,
or, alternatively, that ?-cell defenses are improved by nicotinamide treatment.
These results have encouraged the organization of several larger studies, including
the European-Canadian Nicotinamide Diabetes Intervention Trial (ENDIT), which will
involve the screening of about 30,000 people for subjects that are anti–islet cell antibody–
positive with normal glucose tolerance (164). These subjects will be treated with nicotinamide
or placebo for 5 years, and the results should provide significant information on
Niacin 245
the prevention of diabetes by nicotinamide, its mechanism of action, and even the longterm
risk of carcinogenesis.
3. Nicotinamide Toxicity
The levels of nicotinamide that are used in the treatment of IDDM (about 3 g/day) have
not been reported to cause any adverse side effects on an acute basis. Larger doses (about
10 g/day) have been known to cause liver injury (parenchymal cell injury, portal fibrosis,
cholestasis) (179). Chronic intake of nicotinamide can also induce a methyl-group defi-
ciency state due to the methylation reactions involved in excretion (180,181), and physicians
recommending nicotinamide therapy should ensure that the subjects have an adequate
intake of methyl donors such as choline and methionine. Methyl donor deficiency
also appears to increase the risk of carcinogenesis, and more work from this perspective
is needed to define the safe use of nicotinamide. The most serious potential side effect
of nicotinamide use would be the induction of pancreatic tumors in patients at risk for
IDDM, and it will take many years to assess this. There is comfort from the fact that studies
of the NOD mouse have not reported pancreatic tumors during long-term nicotinamide
Niacin deficiency has the potential to alter redox reactions, poly and mono(ADP-ribose)
synthesis, and the formation of cyclic ADP-ribose and NAADP. During niacin deficiency,
the metabolic changes that lead to the dramatic signs and symptoms of pellagra will likely
be tissue-specific and reflect subcellular competition for NAD pools. The effect of chronic
niacin undernutrition on human health, especially the process of carcinogenesis, appears
to be an exciting area that deserves more attention. With a rapidly broadening perspective
on the biochemical roles for niacin in metabolism, identification of optimal niacin nutriture
should be possible in the coming decade.
Supplementation of nicotinic acid and nicotinamide above the dietary requirement
may affect some of the same processes, but these compounds have distinctive pharmacological
properties, some of which may be unrelated to their currently defined nutrient
functions. Future research in models of niacin deficiency and supplementation may lead
us to reevaluate the accepted metabolic roles of niacin and create new guidelines for niacin
The authors thank Dr. William Bettger, Dr. William Woodward, Dr. Elaine Jacobson, and
Dr. Myron Jacobson for critical reading of this manuscript. Dr. Doug Lanska helped us to
find historical references and early photographs. The Cancer Research Society (Montreal,
Canada) has supported research on niacin status and ADP-ribose metabolism, which made
the writing of this chapter possible.
1. S Harris. Clinical Pellagra. St. Louis: C.V. Mosby, 1941.
2. KJ Carpenter. Pellagra. Stroudsburg: Hutchinson Ross, 1981.
246 Kirkland and Rawling
3. SR Roberts. Pellagra, History, Distribution, Diagnosis, Prognosis, Treatment, Etiology. St.
Louis: C.V. Mosby, 1914.
4. AJ Bollet. Politics and pellagra: the epidemic of pellagra in the U.S. in the early twentieth
century. Yale J Biol Med 65:211–221, 1992.
5. P Malfait, A Moren, JC Dillon, A Brodel, G Begkoyian, MG Etchegorry, G Malenga, P
Hakewill. An outbreak of pellagra related to changes in dietary niacin among Mozambican
refugees in Malawi. Int J Epidemiol 22:504–511, 1993.
6. EL Jacobson. Niacin deficiency and cancer in women. J Am Coll Nutr 12:412–416, 1993.
7. J Goldberger. The relation of diet to pellagra. JAMA 78:1676–1680, 1922.
8. CA Elvehjem, RJ Madden, FM Strong, DM Woolley. Relation of nicotinic acid and nicotinic
acid amide to canine black tongue. J Am Chem Soc 59:1767–1768, 1937.
9. WA Krehl, LJ Teply, CA Elvehjem. Corn as an etiological factor in the production of a
nicotinic acid deficiency in the rat. Science 101:283, 1945.
10. WH Sebrell. History of pellagra. Fed Proc 40:1520–1522, 1981.
11. C Heidelberger, EP Abraham, S Lepkovsky. Concerning the mechanism of the mammalian
conversion of tryptophan into nicotinic acid. J Biol Chem 176:1461–1462, 1948.
12. J Laguna, KJ Carpenter. Raw versus processed corn in niacin-deficient diets. J Nutr 45:21–
28, 1951.
13. Buniva. Observations in pellagra: it would not appear to be contagious. In: KJ Carpenter,
ed. Pellagra. Stroudsburg: Hutchinson Ross, 1981, pp 11–12.
14. TD Spies, WB Bean, WF Ashe. Recent advances in the treatment of pellagra and associated
deficiencies. In: KJ Carpenter, ed. Pellagra. Stroudsburg: Hutchinson Ross, 1981, pp 213–
15. A Hoffer. Pellagra and schizophrenia. Psychosomatics 11:522–525, 1970.
16. S Harris. Clinical Pellagra. St. Louis: C.V. Mosby, 1941, p 366.
17. S Harris. Clinical Pellagra. St. Louis: C.V. Mosby, 1941, p 192.
18. S Harris. Clinical Pellagra. St. Louis: C.V. Mosby, 1941, p 295.
19. M Freed. Methods of Vitamin Assay. New York: The Association of Vitamin Chemists Inc.,
1966, p 169.
20. W Friedrich. Vitamins. New York: Walter de Gruyter, 1988, pp 473–542.
21. TR Guilarte, K Pravlik. Radiometric-microbiologic assay of niacin using Kloeckera brevis:
analysis of human blood and food. J Nutr 113:2587–2594, 1983.
22. HW Huff, WA Perlzweig. The fluorescent condensation product of N1-methylnicotinamide
and acetone: II. A sensitive method for the determination of N1-methylnicotinamide. J Biol
Chem 167:157–167, 1947.
23. EL Jacobson, AJ Dame, JS Pyrek, MK Jacobson. Evaluating the role of niacin in human
carcinogenesis. Biochimie 77:394–398, 1995.
24. J Stein, A Hahn, G Rehner. High-performance liquid chromatographic determination of nicotinic
acid and nicotinamide in biological samples applying post-column derivatization resulting
in bathmochrome absorption shifts. J Chromatogr B Biomed Appl 665:71–78,
25. R Roskoski. Determination of pyridine nucleotides by fluorescence and other optical techniques.
In: D Dolphin, R Poulson, O Avramovic, eds. Pyridine Nucleotide Coenzymes:
Chemical, Biochemical and Medical Aspects. New York: John Wiley and Sons, 1987, pp
26. DP Jones. Determination of pyridine dinucleotides in cell extracts by high-performance liquid
chromatography. J Chromatogr 225:446–449, 1981.
27. JL Spivak, DL Jackson. Pellagra: an analysis of 18 patients and a review of the literature.
Johns Hopkins Med J 140:295–309, 1977.
28. SE Sauberlich. Nutritional aspects of pyridine nucleotides. In: D Dolphin, R Poulson, O
Avramovic, eds. Pyridine Nucleotide Coenzymes: Chemical, Biochemical and Medical Aspects.
New York: John Wiley and Sons, 1987, pp 599–626.
Niacin 247
29. CS Fu, ME Swendseid, RA Jacob, RWMcKee. Biochemical markers for assessment of niacin
status in young men: levels of erythrocyte niacin coenzymes and plasma tryptophan. J Nutr
119:1949–1955, 1989.
30. JM Rawling, TM Jackson, ER Driscoll, JB Kirkland. Dietary niacin deficiency lowers tissue
poly(ADP-ribose) and NAD concentrations in Fischer-344 rats. J Nutr 124:1597–1603,
31. JW Foster, AG Moat. Nicotinamide adenine dinucleotide biosynthesis and pyridine nucleotide
cycle metabolism in microbial systems. Microbiol Rev 44:83–105, 1980.
32. J Preiss, P Handler. Biosynthesis of diphosphopyridine nucleotide I. Identification of Intermediates.
J Biol Chem 233:488–500, 1958.
33. LS Dietrich, L Fuller, IL Yero, L Martinez. Nicotinamide mononucleotide pyrophosphorylase
activity in animal tissues. J Biol Chem 241:188–191, 1966.
34. H Grunicke, HJ Keller, M Liersch, A Benaguid. New aspects of the mechanism and regulation
of pyridine nucleotide metabolism. Adv Enzyme Regul 12:397–418, 1974.
35. U Delabar, M Siess. Synthesis and degradation of NAD in guinea pig cardiac muscle: II.
Studies about the different biosynthetic pathways and the corresponding intermediates. Basic
Res Cardiol 74:571–593, 1979.
36. BM Olivera, AM Ferro. Pyridine nucleotide metabolism and ADP-ribosylation. In: Hayashi
O, Ueda K., eds., ADP-ribosylation Reactions: Biology and Medicine. New York: Academic
Press, 1982, pp 19–40.
37. S Sestini, C Ricci, V Micheli, G Pompucci. Nicotinamide mononucleotide adenylyltransferase
activity in human erythrocytes. Arch Biochem Biophys 302:206–211, 1993.
38. AG Moat, JW Foster. Biosynthesis and salvage pathways of pyridine nucleotides. In: D Dolphin,
M Powanda, R Poulson, eds. Pyridine Nucleotide Coenzymes: Chemical, Biochemical
and Medical Aspects. New York: John Wiley and Sons, 1982, Part B, pp 1–24.
39. LM Henderson. Niacin. Annu Rev Nutr 3:289–307, 1983.
40. M Ikeda, H Tsuji, S Nakamura, A Ichiyama, Y Nishizuka, O Hayaishi. Studies on the biosynthesis
of nicotinamide adenine dinucleotide. II. A role of picolinic carboxylase in the biosynthesis
of nicotinamide adenine dinucleotide from tryptophan in mammals. J Biol Chem 240:
1395–1401, 1965.
41. MK Horwitt, AE Harper, LM Henderson. Niacin-tryptophan relationships for evaluating niacin
equivalents. Am J Clin Nutr 34:423–427, 1981.
42. AC Da Silva, R Fried, RC De Angelis. Domestic cat as laboratory animal for experimental
nutrition studies; niacin requirements and tryptophan metabolism. J Nutr 46:399–409, 1952.
43. LV Hankes, LM Henderson, WL Brickson, CA Elvehjem. Effect of amino acids on the
growth of rats on niacin-tryptophan deficient rations. J Biol Chem 174:873–881, 1948.
44. ME Reid. Nutritional studies with the guinea pig. VII. Niacin. J Nutr 75:279–286, 1961.
45. H Takanaga, H Maeda, H Yabuuchi, I Tamai, H Higashida, A Tsuji. Nicotinic acid transport
mediated by pH-dependent anion antiporter and proton cotransporter in rabbit intestinal
brush-border membrane. J Pharm Pharmacol 48:1073–1077, 1996.
46. V Micheli, HA Simmonds, S Sestini, C Ricci. Importance of nicotinamide as an NAD precursor
in the human erythrocyte. Arch Biochem Biophys 283:40–45, 1990.
47. TM Jackson, JM Rawling, BD Roebuck, JB Kirkland. Large supplements of nicotinic acid
and nicotinamide increase tissue NAD and poly(ADP-ribose) levels but do not affect diethylnitrosamine-
induced altered hepatic foci in Fischer-344 rats. J Nutr 125:1455–1461,
48. C Bernofsky. Physiology aspects of pyridine nucleotide regulation in mammals. Mol Cell
Biochem 33:135–143, 1980.
49. NO Kaplan. History of the pyridine nucleotides. In: D Dolphin, R Poulson, O Avramovic,
eds. Pyridine Nucleotide Coenzymes: Chemical, Biochemical and Medical Aspects. New
York: John Wiley and Sons, 1987, Part A, pp 1–20.
50. DL Clapper, TF Walseth, PJ Dargie, HC Lee. Pyridine nucleotide metabolites stimulate cal248
Kirkland and Rawling
cium release from sea urchin egg microsomes desensitized to inositol trisphosphate. J Biol
Chem 262:9561–9568, 1987.
51. H Kim, EL Jacobson, MK Jacobson. Position of cyclization in cyclic ADP-ribose. Biochem
Biophys Res Commun 194:1143–1147, 1993.
52. HC Lee. Cyclic ADP-ribose: a calcium mobilizing metabolite of NAD. Mol Cell Biochem
138:229–235, 1994.
53. N Noguchi, S Takasawa, K Nata, A Tohgo, I Kato, F Ikehata, H Yonekura, H Okamoto.
Cyclic ADP-ribose binds to FK506-binding protein 12.6 to release Ca2 from islet microsomes.
J Biol Chem 272:3133–3136, 1997.
54. H Kim, EL Jacobson, MK Jacobson. Synthesis and degradation of cyclic ADP-ribose by
NAD glycohydrolases. Science 261:1330–1333, 1993.
55. E Zocchi, L Franco, L Guida, U Benatti, A Bargellesi, F Malavasi, HC Lee, A De Flora. A
single protein immunologically identified as CD38 displays NAD glycohydrolase, ADPribosyl
cyclase and cyclic ADP-ribose hydrolase activities at the outer surface of human
erythrocytes. Biochem Biophys Res Commun 196:1459–1465, 1993.
56. A Galione. Cyclic ADP-ribose, the ADP-ribosyl cyclase pathway and calcium signalling.
Mol Cell Endocrinol 98:125–131, 1994.
57. S Takasawa, K Nata, H Yonekura, H Okamoto. Cyclic ADP-ribose in insulin secretion from
pancreatic beta cells [see comments]. Science 259:370–373, 1993.
58. HC Lee. NAADP: An emerging calcium signaling molecule. J Membr Biol 173:1–8, 2000.
59. HC Lee, R Graeff, TF Walseth. Cyclic ADP-ribose and its metabolic enzymes. Biochimie
77:345–355, 1995.
60. M Mizuguchi, N Otsuka, M Sato, Y Ishii, S Kon, M Yamada, H Nishina, T Katada, K Ikeda.
Neuronal localization of CD38 antigen in the human brain. Brain Res 697:235–240, 1995.
61. SH Snyder, DM Sabatini, MM Lai, JP Steiner, GS Hamilton, PD Suzdak. Neural actions of
immunophilin ligands. Trends Pharmacol Sci 19:21–26, 1998.
62. J Yamauchi, S Tanuma. Occurrence of an NAD glycohydrolase in bovine brain cytosol.
Arch Biochem Biophys 308:327–329, 1994.
63. Y Desmarais, L Menard, J Lagueux, GG Poirier. Enzymological properties of poly(ADPribose)
polymerase: characterization of automodification sites and NADase activity. Biochem
Biophys Acta 1078:179–186, 1991.
64. R Antoine, C Locht. The NAD-glycohydrolase activity of the pertussis toxin S1 subunit.
Involvement of the catalytic HIS-35 residue. J Biol Chem 269:6450–6457, 1994.
65. RE West, Jr., J Moss. Amino acid specific ADP-ribosylation: specific NAD: arginine mono-
ADP-ribosyltransferases associated with turkey erythrocyte nuclei and plasma membranes.
Biochemistry 25:8057–8062, 1986.
66. RV Considine, LL Simpson Cellular and molecular actions of binary toxins possessing ADPribosyltransferase
activity. Toxicon 29:913–936, 1991.
67. A Zolkiewska, IJ Okazaki, J Moss. Vertebrate mono-ADP-ribosyltransferases. Mol Cell Biochem
138:107–112, 1994.
68. T Takada, IJ Okazaki, J Moss. ADP-ribosylarginine hydrolases. Mol Cell Biochem 138:119–
122, 1994.
69. M Tsuchiya, M Shimoyama. Target protein for eucaryotic arginine-specific ADP-ribosyltransferase.
Mol Cell Biochem 138:113–118, 1994.
70. J Wang, E Nemoto, AY Kots, HR Kaslow, G Dennert. Regulation of cytotoxic T cells by
ecto-nicotinamide adenine dinucleotide (NAD) correlates with cell surface GPI-anchored/
arginine ADP-ribosyltransferase. J Immunol 153:4048–4058, 1994.
71. WJ Iglewski. Cellular ADP-ribosylation of elongation factor 2. Mol Cell Biochem 138:131–
133, 1994.
72. T Maehama, N Sekine, H Nishina, K Takahashi, T Katada. Characterization of botulinum C3-
catalyzed ADP-ribosylation of rho proteins and identification of mammalian C3-like ADPribosyltransferase.
Mol Cell Biochem 138:135–140, 1994.
Niacin 249
73. BE Ledford, GH Leno. ADP-ribosylation of the molecular chaperone GRP78/BiP. Mol Cell
Biochem 138:141–148, 1994.
74. S Chatterjee, MF Cheng, SJ Berger, NA Berger. Induction of M(r) 78,000 glucose-regulated
stress protein in poly(adenosine diphosphate-ribose) polyme. Cancer Res 54:440–4411,
75. LJ McDonald, J Moss. Enzymatic and nonenzymatic ADP-ribosylation of cysteine. Mol Cell
Biochem 138:221–226, 1994.
76. J Zhang, SH Snyder. Nitric oxide stimulates auto-ADP-ribosylation of glyceraldehyde-3-
phosphate dehydrogenase. Proc Natl Acad Sci USA 89:9382–9385, 1992.
77. LJ McDonald, J Moss. Nitric oxide and NAD-dependent protein modification. Mol Cell Biochem
138:201–206, 1994.
78. EL Jacobson, D Cervantes-Laurean, MK Jacobson. Glycation of proteins by ADP-ribose.
Mol Cell Biochem 138:207–212, 1994.
79. P Chambon, JD Weill, J Doly, MT Strosser, P Mandel. On the formation of a novel adenylic
compound by enzymatic extracts of liver nuclei. Biochem Biophys Res Commun 25:638–
643, 1966.
80. I Kameshita, Z Matsuda, T Taniguchi, Y Shizuta. Poly(ADP-ribose) synthetase. Separation
and identification of three proteolytic fragments as the substrate-binding domain, the DNAbinding
domain, and the automodification domain. J Biol Chem 259:4770–4776, 1984.
81. M Ikejima, S Noguchi, R Yamashita, T Ogura, T Sugimura, DM Gill, M Miwa. The zinc
fingers of human poly(ADP-ribose) polymerase are differentially required for the recognition
of DNA breaks and nicks and the consequent enzyme activation. Other structures recognize
intact DNA. J Biol Chem 265:21907–21913, 1990.
82. D Lautier, J Lagueux, J Thibodeau, L Menard, GG Poirier. Molecular and biochemical features
of poly(ADP-ribose) metabolism. Mol Cell Biochem 122:171–193, 1993.
83. CC Kiehlbauch, N Aboul-Ela, EL Jacobson, DP Ringer, MK Jacobson. High resolution fractionation
and characterization of ADP-ribose polymers. Anal Biochem 208:26–34, 1993.
84. H Mendoza-Alvarez, R Alvarez-Gonzalez Poly(ADP-ribose) polymerase is a catalytic dimer
and the automodification reaction is intermolecular. J Biol Chem 268:22575–22580,
85. FR Althaus, L Hofferer, HE Kleczkowska, M Malanga, H Naegeli, P Panzeter, C Realini.
Histone shuttle driven by the automodification cycle of poly(ADP-ribose)polymerase. Environ
Mol Mutagen 22:278–282, 1993.
86. S Desnoyers,GM Shah, G Brochu, JC Hoflack, A Verreault, GG Poirier. Biochemical properties
and function of poly(ADP-ribose) glycohydrolase. Biochimie 77:433–438, 1995.
87. A Miro, MJ Costas, M Garcia-Diaz, MT Hernandez, JC Cameselle. A specific, low Km
ADP-ribose pyrophosphatase from rat liver. FEBS Lett 244:123–126, 1989.
88. K Hatakeyama, Y Nemoto, K Ueda, O Hyashi. Poly(ADP-ribose) glycohydrolase and ADPribosyl
group turnover. In: ADP-Ribose Transfer Reactions. Berlin: Springer-Verlag, 1989,
p 47.
89. D D’Amours, S Desnoyers, I D’Silva, GG Poirier. Poly(ADP-ribosyl)ation reactions in the
regulation of nuclear functions. Biochem J 342 (Pt 2):249–268, 1999.
90. BW Durkacz, O Omidiji, DA Gray, S Shall. (ADP-ribose)n participates in DNA excision
repair. Nature 283:593–596, 1980.
91. MF Denisenko, VA Soldatenkov, LN Belovskaya, IV Filippovich. Is the NAD-poly (ADPribose)
polymerase system the trigger in radiation-induced death of mouse thymocytes? Int
J Radiat Biol 56:277–285, 1989.
92. Y Ohashi, K Ueda, MKawaichi, O Hayaishi. Activation of DNA ligase by poly(ADP-ribose)
in chromatin. Proc Natl Acad Sci USA 80:3604–3607, 1983.
93. CM Simbulan,M Suzuki, S Izuta, T Sakurai, E Savoysky, K Kojima, K Miyahara, Y Shizuta,
S Yoshida. Poly(ADP-ribose) polymerase stimulates DNA polymerase alpha by physical
association. J Biol Chem 268:93–99, 1993.
250 Kirkland and Rawling
94. MS Satoh, T Lindahl. Role of poly(ADP-ribose) formation in DNA repair. Nature 356:356–
358, 1992.
95. TJ Jorgensen, JC Leonard, PJ Thraves, A Dritschilo. Baseline sister chromatid exchange in
human cell lines with different levels of poly(ADP-ribose) polymerase. Radiat Res 127:107–
110, 1991.
96. T Stevnsner, R Ding, M Smulson, VA Bohr. Inhibition of gene-specific repair of alkylation
damage in cells depleted of poly(ADP-ribose) polymerase. Nucleic Acids Res 22:4620–4624,
97. S Chatterjee, NA Berger. Growth-phase-dependent response to DNA damage in poly(ADPribose)
polymerase deficient cell lines: basis for a new hypothesis describing the role of
poly(ADP-ribose) polymerase in DNA replication and repair. Mol Cell Biochem 138:61–
69, 1994.
98. Y Le Rhun, JB Kirkland, GM Shah. Cellular responses to DNA damage in the absence of
poly(ADP-ribose) polymerase. Biochem Biophys Res Commun 245:1–10, 1998.
99. S Smith, I Giriat, A Schmitt, T de Lange. Tankyrase, a poly(ADP-ribose) polymerase at
human telomeres [see comments]. Science 282:1484–1487, 1998.
100. M Johansson. A human poly(ADP-ribose) polymerase gene family (ADPRTL): cDNA cloning
of two novel poly(ADP-ribose) polymerase homologues. Genomics 57:442–445,
101. H Berghammer, M Ebner, R Marksteiner, B Auer. pADPRT-2: a novel mammalian polymerizing(
ADP-ribosyl)transferase gene related to truncated pADPRT homologues in plants and
Caenorhabditis elegans. FEBS Lett 449:259–263, 1999.
102. JC Ame, V Rolli, V Schreiber, C Niedergang, F Apiou, P Decker, S Muller, T Hoger, J
Menissier-de Murcia, G de Murcia. PARP-2, A novel mammalian DNA damage-dependent
poly(ADP-ribose) polymerase. J Biol Chem 274:17860–17868, 1999.
103. VA Kickhoefer, AC Siva, NL Kedersha, EM Inman, C Ruland, M Streuli, LH Rome. The
193-kD vault protein, VPARP, is a novel poly(ADP-ribose) polymerase. J Cell Biol 146:
917–928, 1999.
104. CM Whitacre, H Hashimoto, ML Tsai, S Chatterjee, SJ Berger, NA Berger. Involvement of
NAD-poly(ADP-ribose) metabolism in p53 regulation and its consequences. Cancer Res 55:
3697–3701, 1995.
105. JM Rawling, R Alvarez-Gonzalez. TFIIF, a basal eukaryotic transcription factor, is a substrate
for poly(ADP-ribosyl)ation. Biochem J 324 (Pt 1):249–253, 1997.
106. M Bhatia, JB Kirkland, KA Meckling-Gill. Modulation of poly(ADP-ribose) polymerase
during neutrophilic and monocytic differentiation of promyelocytic (NB4) and myelocytic
(HL-60) leukaemia cells. Biochem J 308 (Pt 1):131–137, 1995.
107. M Bhatia, JB Kirkland, KA Meckling-Gill. Overexpression of poly(ADP-ribose) polymerase
promotes cell cycle arrest and inhibits neutrophilic differentiation of NB4 acute promyelocytic
leukemia cells. Cell Growth Differ 7:91–100, 1996.
108. S Takahashi, D Nakae, Y Yokose, Y Emi, A Denda, S Mikami, T Ohnishi, Y Konishi.
Enhancement of DEN initiation of liver carcinogenesis by inhibitors of NADADP ribosyl
transferase in rats. Carcinogenesis 5:901–906, 1984.
109. NA Berger. Poly(ADP-ribose) in the cellular response to DNA damage. Radiat Res 101:4–
15, 1985.
110. H Okamoto. The role of poly(ADP-ribose) synthetase in the development of insulin-dependent
diabetes and islet B-cell regeneration. Biomed Biochim Acta 44:15–20, 1985.
111. T Yamagami, A Miwa, S Takasawa, H Yamamoto, H Okamoto. Induction of rat pancreatic Bcell
tumors by the combined administration of streptozotocin or alloxan and poly(adenosine
diphosphate ribose) synthetase inhibitors. Cancer Res 45:1845–1849, 1985.
112. PW Rankin, EL Jacobson, RC Benjamin, J Moss, MK Jacobson. Quantitative studies of
inhibitors of ADP-ribosylation in vitro and in vivo. J Biol Chem 264:4312–4317, 1989.
113. JM Rawling, ER Driscoll, GG Poirier, JB Kirkland. Diethylnitrosamine administration in
Niacin 251
vivo increases hepatic poly(ADP-ribose) levels in rats: results of a modified technique for
poly(ADP-ribose) measurement. Carcinogenesis 14:2513–2516, 1993.
114. JR Wright, Jr., J Mendola, PE Lacy. Effect of niacin/nicotinamide deficiency on the diabetogenic
effect of streptozotocin. Experientia 44:38–40, 1988.
115. GP Warwick, JS Harington. Some aspects of the epidemiology and etiology of esophageal
cancer with particular emphasis on the Transkei, South Africa. Adv Cancer Res 17:121–
131, 1973.
116. SJ Van Rensburg, ES Bradshaw, D Bradshaw, EF Rose. Oesophageal cancer in Zulu men,
South Africa: a case-control study. Br J Cancer 51:399–405, 1985.
117. J Wahrendorf, J Chang-Claude, QS Liang, YG Rei, N Munoz, M Crespi, R Raedsch, D
Thurnham, P Correa. Precursor lesions of oesophageal cancer in young people in a highrisk
population in China [see comments]. Lancet 2:1239–1241, 1989.
118. S Franceschi, E Bidoli, AE Baron, C La Vecchia. Maize and risk of cancers of the oral
cavity, pharynx, and esophagus in northeastern Italy [see comments]. J Natl Cancer Inst 82:
1407–1411, 1990.
119. SJ Van Rensburg. Epidemiologic and dietary evidence for a specific nutritional predisposition
to esophageal cancer. J Natl Cancer Inst 67:243–251, 1981.
120. JR Marshall, S Graham, BP Haughey, D Shedd, R O’Shea, J Brasure, GS Wilkinson, D
West. Smoking, alcohol, dentition and diet in the epidemiology of oral cancer. Eur J Cancer
B Oral Oncol 28B:9–15, 1992.
121. WJ Blot, JY Li, PR Taylor, W Guo, S Dawsey, GQ Wang, CS Yang, SF Zheng, M Gail,
GY Li. Nutrition intervention trials in Linxian, China: supplementation with specific vitamin/
mineral combinations, cancer incidence, and disease-specific mortality in the general population
[see comments]. J Natl Cancer Inst 85:1483–1492, 1993.
122. GQ Wang, SM Dawsey, JY Li, PR Taylor, B Li, WJ Blot,WMWeinstein, FS Liu, KJ Lewin,
H Wang. Effects of vitamin/mineral supplementation on the prevalence of histological dysplasia
and early cancer of the esophagus and stomach: results from the General Population
Trial in Linxian, China. Cancer Epidemiol Biomarkers Prev 3:161–166, 1994.
123. EG Miller, H Burns, Jr. N-nitrosodimethylamine carcinogenesis in nicotinamide-deficient
rats. Cancer Res 44:1478–1482, 1984.
124. SJ Van Rensburg, JM Hall, PS Gathercole. Inhibition of esophageal carcinogenesis in cornfed
rats by riboflavin, nicotinic acid, selenium, molybdenum, zinc, and magnesium. Nutr
Cancer 8:163–170, 1986.
125. JZ Zhang, SM Henning, ME Swendseid. Poly(ADP-ribose) polymerase activity and DNA
strand breaks are affected in tissues of niacin-deficient rats. J Nutr 123:1349–1355,
126. JM Rawling, TM Jackson, BD Roebuck, GG Poirier, JB Kirkland. The effect of niacin defi-
ciency on diethylnitrosamine-induced hepatic poly(ADP-ribose) levels and altered hepatic
foci in the Fischer-344 rat. Nutr Cancer 24:111–119, 1995.
127. K Ueda, M Kawaichi, O Hayshi. Poly(ADP-ribose) Synthetase. In: O Hayashi, K Ueda, eds.
ADP-ribosylation Reactions, Biology and Medicine. New York: Academic Press, 1982, pp
128. AC Boyonoski, LM Gallacher, MM ApSimon, RM Jacobs, GM Shah, GG Poirier, JB Kirkland.
Niacin deficiency in rats increases the severity of ethylnitrosourea-induced anemia and
leukopenia. J Nutr 130:1102–1107, 2000.
129. AC Boyonoski, LM Gallacher, MM ApSimon, RM Jacobs, GM Shah, GG Poirier, JB Kirkland.
Niacin deficiency increases the sensitivity of rats to the short and long term effects of
ethylnitrosourea treatment. Mol Cell Biochem 193:83–87, 1999.
130. RI Inculet, JA Norton, GE Nichoalds, MM Maher, DE White, MF Brennan. Water-soluble
vitamins in cancer patients on parenteral nutrition: a prospective study. JPEN J Parenter
Enteral Nutr 11:243–249, 1987.
131. GT Bryan. The influence of niacin and nicotinamide on in vivo carcinogenesis. In: LA Poirier,
252 Kirkland and Rawling
PM Newberne, MW Pariza, eds. Advances in Experimental Medicine and Biology. Vol.
206—Essential Nutrients in Carcinogenesis. New York: Plenum Press, 1986, pp 331–338.
132. PL Canner, KG Berge, NK Wenger, J Stamler, L Friedman, RJ Prineas, W Friedewald.
Fifteen year mortality in Coronary Drug Project patients: long-term benefit with niacin. J
Am Coll Cardiol 8:1245–1255, 1986.
133. SN Giri, R Blaisdell, RB Rucker, Q Wang, DM Hyde. Amelioration of bleomycin-induced
lung fibrosis in hamsters by dietary supplementation with taurine and niacin: biochemical
mechanisms. Environ Health Perspect 102 (suppl 10):137–147, 1994.
134. A Nagai, H Matsumiya, M Hayashi, S Yasui, H Okamoto, K Konno. Effects of nicotinamide
and niacin on bleomycin-induced acute injury and subsequent fibrosis in hamster lungs. Exp
Lung Res 20:263–281, 1994.
135. JM Rawling,MMApSimon, JB Kirkland. Lung poly(ADP-ribose) and NAD concentrations
during hyperoxia and niacin deficiency in the Fischer-344 rat. Free Radic Biol Med 20:865–
871, 1996.
136. SG Jenkinson, RA Lawrence, DL Butler. Inability of niacin to protect from in vivo hyperoxia
or in vitro microsomal lipid peroxidation. J Toxicol Clin Toxicol 19:975–985, 1982.
137. JD Crapo. Morphologic changes in pulmonary oxygen toxicity. Annu Rev Physiol 48:721–
731, 1986.
138. S Bergelson, R Pinkus, V Daniel. Intracellular glutathione levels regulate Fos/Jun induction
and activation of glutathione S-transferase gene expression. Cancer Res 54:36–40, 1994.
139. JK Wilkin, TB Bentley, DL Latour, EW Rosenberg. Nicotinic acid treatment of skin disorders.
Br J Dermatol 100:471–472, 1979.
140. HA Poston, MJ Wolfe. Niacin requirement for optimum growth, feed conversion and protection
of rainbow trout, Salmo gairdneri, from ultraviolet-B radiation. J Fish Dis 8:451–460,
141. JD Morrow, JA Awad, JA Oates, LJ Roberts. Identification of skin as a major site of prostaglandin
D2 release following oral administration of niacin in humans. J Invest Dermatol 98:
812–815, 1992.
142. JK Wilkin, G Fortner, LA Reinhardt, OV Flowers, SJ Kilpatrick, WC Streeter. Prostaglandins
and nicotinate-provoked increase in cutaneous blood flow. Clin Pharmacol Ther 38:273–
277, 1985.
143. HL Gensler, T Williams, AC Huang, EL Jacobson. Oral niacin prevents photocarcinogenesis
and photoimmunosuppression in mice. Nutr Cancer 34:36–41, 1999.
144. JJ Yourick, JS Dawson, CD Benton, ME Craig, LW Mitcheltree. Pathogenesis of 2,2?-dichlorodiethyl
sulfide in hairless guinea pigs. Toxicology 84:185–197, 1993.
145. EL Jacobson, V Nunbhakdi-Craig, DG Smith, HY Chen, BL Wasson, MK Jacobson. ADPribose
polymer metabolism: implications for human nutrition. In: GG Poirier, P Moreau, eds.
ADP-Ribosylation Reactions. New York: Springer-Verlag, 1992, pp 153–162.
146. TE Barnum ed. Enzyme Handbook. New York: Springer-Verlag, 1969.
147. RM Graeff, TF Walseth, K Fryxell, WD Branton, HC Lee. Enzymatic synthesis and characterizations
of cyclic GDP-ribose. A procedure for distinguishing enzymes with ADP-ribosyl
cyclase activity. J Biol Chem 269:30260–30267, 1994.
148. KW Beers, EN Chini, HC Lee, TP Dousa. Metabolism of cyclic ADP-ribose in opossum
kidney renal epithelial cells. Am J Physiol 268:C741–C746, 1995.
149. H Grunicke, HJ Keller, B Puschendorf, A Benaguid. Biosynthesis of nicotinamide adenine
dinucleotide in mitochondria. Eur J Biochem 53:41–45, 1975.
150. JL Purvis, JM Lowenstein. The relationship between intra- and extramitochondrial pyridine
nucleotides. J Biol Chem 236:2794–2803, 1961.
151. A Behr, H Taguchi, RK Gholson. Apparent pyridine nucleotide synthesis in mitochondria:
an artifact of NMN and NAD glycohydrolase activity? Biochem Biophys Res Commun 101:
767–774, 1981.
152. ME Tischler, D Friedrichs, K Coll, JR Williamson. Pyridine nucleotide distributions and
Niacin 253
enzyme mass action ratios in hepatocytes from fed and starved rats. Arch Biochem Biophys
184:222–236, 1977.
153. M Weiner, J van Eys. Nicotinic Acid: Nutrient-Cofactor-Drug. New York: Marcel Dekker,
154. DA Bender, BI Magboul, D Wynick. Probable mechanisms of regulation of the utilization
of dietary tryptophan, nicotinamide and nicotinic acid as precursors of nicotinamide nucleotides
in the rat. Br J Nutr 48:119–127, 1982.
155. A Petley, B Macklin, AG Renwick, TJ Wilkin. The pharmacokinetics of nicotinamide in
humans and rodents. Diabetes 44:152–155, 1995.
156. JR DiPalma, WS Thayer. Use of niacin as a drug. Annu Rev Nutr 11:169–187, 1991.
157. C Marcus, T Sonnenfeld, B Karpe, P Bolme, P Arner. Inhibition of lipolysis by agents acting
via adenylate cyclase in fat cells from infants and adults. Pediatr Res 26:255–259, 1989.
158. SM Grundy, HY Mok, L Zech, M Berman. Influence of nicotinic acid on metabolism of
cholesterol and triglycerides in man. J Lipid Res 22:24–36, 1981.
159. JD Alderman, RC Pasternak, FM Sacks, HS Smith, ES Monrad, W Grossman. Effect of a
modified, well-tolerated niacin regimen on serum total cholesterol, high density lipoprotein
cholesterol and the cholesterol to high density lipoprotein ratio. Am J Cardiol 64:725–729,
160. JI Rader, RJ Calvert, JN Hathcock. Hepatic toxicity of unmodified and time-release preparations
of niacin. Am J Med 92:77–81, 1992.
161. A Hoffer. Megavitamin B-3 therapy for schizophrenia. Can Psychiatr Assoc J 16:499–504,
162. MR Horsman. Nicotinamide and other benzamide analogs as agents for overcoming hypoxic
cell radiation resistance in tumours. A review. Acta Oncol 34:571–587, 1995.
163. DJ Chaplin, MR Horsman, MJ Trotter. Effect of nicotinamide on the microregional heterogeneity
of oxygen delivery within a murine tumor. J Natl Cancer Inst 82:672–676, 1990.
164. MT Behme. Nicotinamide and diabetes prevention. Nutr Rev 53:137–139, 1995.
165. JA Norton, JL Doppman, RT Jensen. Cancer of the endocrine system. In: VT DeVita, S
Hellman, SA Rosenberg, eds. Cancer: Principles and Practice of Oncology. Philadelphia:
J.B. Lippincott Co., 1989, pp 1324–1331.
166. H Kolb, V Burkart. Nicotinamide in type 1 diabetes. Mechanism of action revisited. Diabetes
Care 22 (suppl 2):B16–B20, 1999.
167. S Reddy, NJ Bibby, RB Elliott. Early nicotinamide treatment in the NOD mouse: effects on
diabetes and insulitis suppression and autoantibody levels. Diabetes Res 15:95–102, 1990.
168. H Nakajima, K Yamada, T Hanafusa, H Fujino-Kurihara, J Miyagawa, A Miyazaki, R Saitoh,
Y Minami, N Kono, K Nonaka. Elevated antibody-dependent cell-mediated cytotoxicity and
its inhibition by nicotinamide in the diabetic NOD mouse. Immunol Lett 12:91–94, 1986.
169. K Yamada, T Hanafusa, H Fujino-Kurihara, A Miyazaki, H Nakajima, J Miyagawa, N
Kono, K Nonaka, S Tarui. Nicotinamide prevents lymphocytic infiltration in submandibular
glands but not the appearance of anti-salivary duct antibodies in non-obese diabetic (NOD)
mice. Res Commun Chem Pathol Pharmacol 50:83–91, 1985.
170. DL Eizirik, S Sandler, N Welsh, K Bendtzen, C Hellerstrom. Nicotinamide decreases nitric
oxide production and partially protects human pancreatic islets against the suppressive effects
of combinations of cytokines. Autoimmunity 19:193–198, 1994.
171. W Sumoski, H Baquerizo, A Rabinovitch. Oxygen free radical scavengers protect rat islet
cells from damage by cytokines. Diabetologia 32:792–796, 1989.
172. B Kallmann, V Burkart, KD Kroncke, V Kolb-Bachofen, H Kolb. Toxicity of chemically
generated nitric oxide towards pancreatic islet cells can be prevented by nicotinamide. Life
Sci 51:671–678, 1992.
173. Y Yonemura, T Takashima, K Miwa, I Miyazaki, H Yamamoto, H Okamoto. Amelioration of
diabetes mellitus in partially depancreatized rats by poly(ADP-ribose) synthetase inhibitors.
Evidence of islet B-cell regeneration. Diabetes 33:401–404, 1984.
254 Kirkland and Rawling
174. O Korsgren, A Andersson, S Sandler. Pretreatment of fetal porcine pancreas in culture with
nicotinamide accelerates reversal of diabetes after transplantation to nude mice. Surgery 113:
205–214, 1993.
175. P Vague, R Picq, M Bernal, V Lassmann-Vague, B Vialettes. Effect of nicotinamide treatment
on the residual insulin secretion in type 1 (insulin-dependent) diabetic patients. Diabetologia
32:316–321, 1989.
176. P Pozzilli, N Visalli, R Buzzetti, MG Baroni, ML Boccuni, E Fioriti, A Signore, C Mesturino,
L Valente, MG Cavallo. Adjuvant therapy in recent onset type 1 diabetes at diagnosis and
insulin requirement after 2 years. Diabete Metab 21:47–49, 1995.
177. RB Elliott, HP Chase. Prevention or delay of type 1 (insulin-dependent) diabetes mellitus
in children using nicotinamide. Diabetologia 34:362–365, 1991.
178. R Manna, A Migliore, LS Martin, E Ferrara, E Ponte, G Marietti, F Scuderi, G Cristiano,
G Ghirlanda, G Gambassi. Nicotinamide treatment in subjects at high risk of developing
IDDM improves insulin secretion. Br J Clin Pract 46:177–179, 1992.
179. SL Winter, JL Boyer. Hepatic toxicity from large doses of vitamin B3 (nicotinamide). N
Engl J Med 289:1180–1182, 1973.
180. YA Kang-Lee, RW McKee, SM Wright, ME Swendseid, DJ Jenden, RS Jope. Metabolic
effects of nicotinamide administration in rats. J Nutr 113:215–221, 1983.
181. MM ApSimon, JM Rawling, JB Kirkland. Nicotinamide megadosing increases hepatic poly
(ADP-ribose) levels in choline-deficient rats. J Nutr 125:1826–1832, 1995.
Riboflavin (Vitamin B2)
Weill Medical College of Cornell University and Memorial Sloan-Kettering Cancer
Center, New York, New York
In the pioneering studies by McCollum and Kennedy (1) in the early part of the twentieth
century, water-soluble tissue extracts were found to be effective in the prevention of the
deficiency state of pellagra in experimental animals. As studies progressed, it became
evident that there were at least two distinct fractions of these extracts, one of which was
heat-labile and the other heat-stable.
Further studies of this heat-stable fraction showed that it was a complex containing
a yellow growth factor. This factor had fluorescent properties and was later purified and
named riboflavin (B2) (2). Other components of this fraction were later identified as niacin,
which was the true antipellagra compound, and vitamin B6, which was particularly effective
in preventing dermatitis in animals.
The physiological role of the yellow growth factor remained obscure until the landmark
discovery by Warburg and Christian (3) in 1932 of ‘‘yellow enzyme’’ or ‘‘old yellow
enzyme.’’ This protein was found to be composed of an apoenzyme and a yellow cofactor
serving as a coenzyme. This coenzyme was shown subsequently to contain an isoalloxazine
ring (4) and a phosphate group (5).
Synthesis of riboflavin was accomplished by Kuhn et al. (6) and Karrer et al. (7).
The flavin coenzyme riboflavin-5?-phosphate (flavin mononucleotide, FMN), was identi-
fied in 1937 by Theorell (8). In 1938, Warburg and Christian (9) clarified the structure
of flavin adenine dinucleotide (FAD), formed from FMN. While many enzymes utilize
FMN and FAD as cofactors, flavins bound covalently to specific tissue flavoproteins
also have been found to have major biological significance, as reviewed by McCormick
256 Rivlin and Pinto
Riboflavin is defined chemically as 7,8-dimethyl-10-(1?-d-ribityl)isoalloxazine. The planar
isoalloxazine ring provides the basic structure not only for riboflavin (vitamin B2) but for
the naturally occurring phosphorylated coenzymes that are derived from riboflavin (Fig.
1). These coenzymes include FMN, FAD, and flavin coenzymes linked covalently to specific
tissue proteins, generally at the 8-? methyl position of the isoalloxazine ring (10).
Among the key mammalian enzymes with covalently bound FAD are included sarcosine
and succinic dehydrogenases, which are located in the matrix and inner mitochondrial
membrane, respectively; monoamine oxidase in the outer mitochondrial membrane;
and l-gulonolactone oxidase, found particularly in liver and kidney microsomes of those
mammalian species capable of synthesizing ascorbic acid from its precursors (11).
The sequence of events in the synthesis of the flavin coenzymes from riboflavin and
its control by thyroid hormones are shown in Fig. 2. Thyroid hormones regulate the activities
of the flavin biosynthetic enzymes (12), the synthesis of the flavoprotein apoenzymes,
Fig. 1 Structural formulas of riboflavin, flavin mononucleotide (riboflavin-5?-phosphate, FMN)
and flavin adenine dinucleotide (FAD).
Riboflavin (Vitamin B2) 257
Fig. 2 Metabolic pathway of conversion of riboflavin into FMN, FAD, and covalently bound
flavin, together with its control by thyroid hormones (12).
and the formation of covalently bound flavins (13). The first biosynthetic enzyme, flavokinase,
catalyzes the initial phosphorylation from ATP of riboflavin to FMN. A fraction
of FMN is directly utilized in this form as a coenzyme. The largest fraction of FMN,
however, combines with a second molecule of ATP to form FAD, the predominant tissue
flavin, in a reaction catalyzed by FAD synthetase, also called FAD pyrophosphorylase.
The covalent attachment of flavins to specific tissue proteins occurs after FAD has been
synthesized. A sequence of phosphatases returns FAD to FMN and FMN, in turn, to
riboflavin (12). Most flavoproteins utilize FAD rather than FMN as coenzyme for a wide
variety of metabolic reactions. Microsomal NADPH–cytochrome P450 reductase is highly
unusual in containing FMN and FAD in equimolar ratios.
Riboflavin is yellow and has a high degree of natural fluorescence when excited by
UV light, a property that can be utilized conveniently in its assay. There are a number of
variations in structure in the naturally occurring flavins. Riboflavin and its coenzymes are
sensitive to alkali and to acid, particularly in the presence of UV light. Under alkaline
conditions, riboflavin is photodegraded to yield lumiflavin (7,8,10-trimethylisoalloxazine),
which is inactive biologically. Riboflavin is photodegraded to lumichrome (7,8-dimethylalloxazine)
under acidic conditions, a product that is also biologically inactive (10). Thus,
an important physical property of riboflavin and its derivatives is their sensitivity to UV
light, resulting in rapid inactivation. Therefore, phototherapy of neonatal jaundice and of
certain skin disorders may promote systemic riboflavin deficiency. The structure–function
relationships of the various biologically active flavins have been comprehensively reviewed
A. Riboflavin Deficiency
Clinical riboflavin deficiency is not recognizable at the bedside by any unique or characteristic
physical features. The typical glossitis, angular stomatitis, and dermatitis observed
is not specific for riboflavin deficiency and may be due to other vitamin deficiencies as
258 Rivlin and Pinto
well. In fact, when dietary deficiency of riboflavin occurs, it is almost invariably in association
with multiple nutrient deficits (15).
With the onset of riboflavin deficiency, one of the adaptations that occurs is a fall
in the hepatic free riboflavin pool to nearly undetectable levels, with a relative sparing of
the pools of FMN and FAD that are needed to fulfill critical metabolic functions (16).
Another adaptation to riboflavin deficiency in its early stages is an increase in the de
novo synthesis of reduced glutathione from its amino acid precursors, in response to the
diminished conversion of oxidized glutathione back to reduced glutathione. This may represent
a compensatory reaction resulting from depressed activity of glutathione reductase,
a key FAD-requiring enzyme (Fig. 3).
An emerging concept is that dietary inadequacy is not the only cause of riboflavin
deficiency and that certain endocrine abnormalities, such as adrenal and thyroid hormone
insufficiency, drugs, and diseases, may interfere significantly with vitamin utilization
(17,18). Psychotropic drugs, such as chlorpromazine, antidepressants [including imipramine
and amitriptyline (19)], cancer chemotherapeutic drugs (e.g., adriamycin), and some
antimalarial agents [e.g., quinacrine (20)], impair riboflavin utilization by inhibiting the
conversion of this vitamin into its active coenzyme derivatives. Figure 4 shows the structural
similarities among riboflavin, imipramine, chlorpromazine, and amitriptyline. Alco-
Fig. 3 Diagrammatic representation of metabolic adaptations to riboflavin deficiency. Diminished
conversion of oxidized glutathione to reduced glutathione occurs as a result of decreased activity
of glutathione reductase, an enzyme that utilizes FAD as a coenzyme. Levels of reduced glutathione
are maintained by enhanced de novo synthesis from precursors (31,32).
Riboflavin (Vitamin B2) 259
Fig. 4 Structural similarities among riboflavin, imipramine, chlorpromazine, and amitriptyline.
hol appears to cause riboflavin deficiency by inhibiting both its digestion from dietary
sources and its intestinal absorption (21).
In approaching riboflavin deficiency, as well as other nutrient deficiencies, it may
be useful to think in terms of risk factors. Thus, the consequences of a poor diet may be
intensified if the patient is also abusing alcohol, is using certain drugs for prolonged periods,
is extremely elderly, or has malabsorption or an underlying illness affecting vitamin
In experimental animals, hepatic architecture is markedly disrupted in riboflavin
deficiency. Mitochondria in riboflavin-deficient mice increase greatly in size, and cristae
increase in both number and size (22). These structural abnormalities may disturb energy
metabolism by interfering with the electron transport chain and metabolism of fatty acids.
Villi decrease in number in the rat small intestine; villus length increases, as does the rate
of transit of developing enterocytes along the villus (23). These findings of structural
abnormalities together with accelerated rate of intestine cell turnover (24) may help to
explain why dietary riboflavin deficiency leads to both decreased iron absorption and increased
iron loss from the intestine.
There are many other effects of riboflavin deficiency on intermediary metabolism,
particularly on lipid, protein, and vitamin metabolism. Of particular relevance to vitamin
metabolism is the fact that the conversion of vitamin B6 to its coenzyme derivative, pyridoxal
5?-phosphate, may be impaired (25). Riboflavin deficiency has been studied in many
animal species and has several vital effects, foremost of which is failure to grow. Additional
effects include loss of hair, skin disturbances, degenerative changes in the nervous
system, and impaired reproduction. Congenital malformations occur in the offspring of
female rats that are riboflavin-deficient. The conjunctiva becomes inflamed, the cornea is
vascularized and eventually opaque, and cataract may result (26).
Changes in the skin consist of scaliness and incrustation of red-brown material consistent
with changes in lipid metabolism. Alopecia may develop, lips become red and
260 Rivlin and Pinto
Riboflavin (Vitamin B2) 261
swollen, and filiform papillae on the tongue deteriorate. During late deficiency anemia
develops. Fatty degeneration of the liver occurs. Important metabolic changes occur, so
that deficient rats require 15–20% more energy than control animals to maintain the same
body weight. Thus, in all species studied, riboflavin deficiency causes profound structural
and functional changes in an ordered sequence. Early changes are very readily reversible.
Later anatomical changes, such as formation of cataract, are largely irreversible despite
treatment with riboflavin.
In humans, as noted above, the clinical features of human riboflavin deficiency do
not have absolute specificity. Early symptoms may include weakness, fatigue, mouth pain
and tenderness, burning and itching of the eyes, and possibly personality changes. More
advanced deficiency may give rise to cheilosis, angular stomatitis, dermatitis, corneal vascularization,
anemia, and brain dysfunction. Thus, the syndrome of dietary riboflavin defi-
ciency in humans has many similarities to that in animals, with one notable exception.
The spectrum of congenital malformations observed in rodents with maternal riboflavin
deficiency has not been clearly identified in humans (26).
B. Food-Related Issues
In the United States today, the most significant dietary sources of riboflavin are meat and
meat products, including poultry and fish, and milk and dairy products, such as eggs and
cheese. In developing countries, plant sources contribute most of the dietary riboflavin
intake. Green vegetables, such as broccoli, collard greens, and turnip greens, are reasonably
good sources of riboflavin. Natural grain products tend to be relatively low in riboflavin,
but fortification and enrichment of grains and cereals has led to a great increase in
riboflavin intake from these food items.
The food sources of riboflavin are similar to those of other B vitamins. Therefore,
it is not surprising that if a given individual’s diet has inadequate amounts of riboflavin,
it will very likely be inadequate in other vitamins as well. A primary deficiency of dietary
riboflavin has wide implications for other vitamins, as flavin coenzymes are involved in
the metabolism of folic acid, pyridoxine, vitamin K, niacin, and vitamin D (27).
Several factors in food preparation and processing may influence the amount of
riboflavin that is actually bioavailable from dietary sources. Appreciable amounts of ribo-
flavin may be lost with exposure to UV light, particularly during cooking and processing.
Prolonged storage of milk in clear bottles or containers may result in flavin degradation
(28). Fortunately, most milk is no longer sold in clear bottles. There has been some controversy
as to whether opaque plastic containers provide greater protection than do cartons,
particularly when milk is stored on a grocery shelf exposed to continuous fluorescent
It is highly likely that large amounts of riboflavin are lost during the sun-drying of
fruits and vegetables. The precise magnitude of the loss is not known but varies with the
duration of exposure. The practice of adding sodium bicarbonate as baking soda to green
vegetables to make them appear fresh can result in accelerated photodegradation of ribo-
flavin. The riboflavin content of common food items is shown in Table 1 (29).
A. Absorption, Transport, Storage, Turnover, and Excretion
Since dietary sources of riboflavin are largely in the form of coenzyme derivatives, these
molecules must be hydrolyzed prior to absorption. Very little dietary riboflavin is found
262 Rivlin and Pinto
as free riboflavin from sources in nature. Under ordinary circumstances the main sources
of free riboflavin are commercial multivitamin preparations.
The absorptive process for flavins occurs in the upper gastrointestinal tract by specialized
transport involving a dephosphorylation-rephosphorylation mechanism, rather
than by passive diffusion. This process is sodium-dependent and involves an ATPase
active transport system that can be saturated (30). It has been estimated that under normal
conditions the upper limit of intestinal absorption of riboflavin at any one time is approximately
25 mg (31). This amount represents approximately 15 times the RDA. Therefore,
the common practice of some megavitamin enthusiasts to consume massive doses of multivitamins
has little benefit with respect to riboflavin, as the additional amounts would be
passed in the stool. Dietary covalent-bound flavins are largely inaccessible as nutritional
sources. In experimental animals, the uptake of riboflavin from the intestine is increased
in dietary riboflavin deficiency (32).
A number of physiological factors influence the rate of intestinal absorption of ribo-
flavin (17). Diets high in psyllium gum appear to decrease the rate of riboflavin absorption,
whereas wheat bran has no detectable effect. The time from oral administration to peak
urinary excretion of riboflavin is prolonged by the antacids aluminum hydroxide and magnesium
hydroxide. Total urinary excretion is unchanged by these drugs, however, and
their major effect appears to be delaying the rate of intestinal absorption rather than inhibiting
net absorption. Alcohol intake interferes with both the digestion of food flavins into
riboflavin and the direct intestinal absorption of the vitamin (21). This observation suggests
that the initial rehabilitation of malnourished alcoholic patients may be accomplished more
efficiently with vitamin supplements containing riboflavin rather than with food sources
comprising predominantly phosphorylated flavin derivatives.
There is evidence that the magnitude of intestinal absorption of riboflavin is increased
by the presence of food. This effect of food may be due to decreasing the rates
of gastric emptying and intestinal transit, thereby permitting more prolonged contact of
dietary riboflavin with the absorptive surface of the intestinal mucosal cells. In general,
delaying the rate of gastric emptying tends to increase the intestinal absorption of riboflavin.
Bile salts also increase the absorption of riboflavin (17).
A number of metals and drugs form chelates or complexes with riboflavin and ribo-
flavin-5?-phosphate that may affect their bioavailability (30). Among the agents in this
category are the metals copper, zinc and iron; the drugs caffeine, theophylline, and saccharin;
and the vitamins nicotinamide and ascorbic acid; as well as tryptophan and urea. The
clinical significance of this binding is not known with certainty in most instances and
deserves further study.
In human blood, the transport of flavins involves loose binding to albumin and tight
binding to a number of globulins. The major binding of riboflavin and its phosphorylated
derivatives in serum is to several classes of immunoglobulins, i.e., IgA, IgG, and IgM
Pregnancy induces the formation of flavin-specific binding proteins. It has been
known for many years that there is an avian riboflavin carrier protein, genetically controlled,
that determines the amount of riboflavin in chicken eggs (33). Absence of the
protein in autosomal recessive hens results in massive riboflavinuria because there is no
mechanism for retaining and binding the vitamin in serum. Eggs become riboflavin-defi-
cient, and embryonic death occurs between the tenth and fourteenth day of incubation.
Administration of antiserum to the chicken riboflavin-carrier protein leads to termination
of the pregnancy.
Riboflavin (Vitamin B2) 263
A new dimension to concepts of plasma protein binding of riboflavin in mammals
was provided by the demonstration that riboflavin-binding proteins can also be found
in serum from pregnant cows, monkeys, and humans. A comprehensive review of ribo-
flavin-binding proteins covers the nature of the binding proteins in various species and
provides evidence that, as in birds, these proteins are crucial for successful mammalian
reproduction (10). The pregnancy-specific binding proteins may help transport riboflavin
to the fetus.
Serum riboflavin-binding proteins also appear to influence placental transfer and
fetal/maternal distribution of riboflavin. There are differential rates of uptake of riboflavin
at the maternal and fetal surfaces of the human placenta (34). Riboflavin-binding proteins
regulate the activity of flavokinase, the first biosynthetic enzyme in the riboflavin-to-FAD
pathway (16).
The urinary excretion of flavins occurs predominantly in the form of riboflavin;
FMN and FAD are not found in urine. McCormick et al. have identified a large number
of flavins and their derivatives in human urine. Besides the 60–70% of urinary flavins
contributed by riboflavin itself, other derivatives include 7-hydroxymethylriboflavin (10–
15%), 8?-sulfonylriboflavin (5–10%), 8-hydroxymethylriboflavin (4–7%), riboflavinyl
peptide ester (5%), and 10-hydroxyethylflavin (1–3%), representing largely metabolites
from covalently bound flavoproteins and intestinal riboflavin degradation by microorganisms.
Traces of lumiflavin and other derivatives have also been found. These findings
were described fully by McCormick (10).
Ingestion of boric acid greatly increases the urinary excretion of riboflavin (17).
This agent when consumed forms a complex with the side chain of riboflavin and other
molecules having polyhydroxyl groups, such as glucose and ascorbic acid. In rodents,
riboflavin treatment greatly ameliorates the toxicity of administered boric acid. This treatment
should also be effective in humans with accidental exposure of boric acid, although
in practice it may be difficult to provide adequate amounts of riboflavin due to its low
solubility and limited absorptive capacity from the intestinal tract.
Urinary excretion of riboflavin in rats is also greatly increased by chlorpromazine
(19). Levels are twice those of age- and sex-matched pair-fed control rats. In addition,
chlorpromazine accelerates the urinary excretion of riboflavin during dietary deficiency.
Urinary concentrations of riboflavin are increased within 6 h of treatment with this
The major function of riboflavin, as noted above, is to serve as the precursor of the flavin
coenzymes, FMN and FAD, and of covalently bound flavins. These coenzymes are widely
distributed in intermediary metabolism and catalyze numerous oxidation–reduction reactions.
Because FAD is part of the respiratory chain, riboflavin is central to energy production.
Other major functions of riboflavin include drug and steroid metabolism, in
conjunction with the cytochrome P450 enzymes, and lipid metabolism. The redox functions
of flavin coenzymes include both one-electron transfers and two-electron transfers
from substrate to the flavin coenzyme (10).
Flavoproteins catalyze dehydrogenation reactions as well as hydroxylations, oxidative
decarboxylations, dioxygenations, and reductions of oxygen to hydrogen peroxide.
Thus, many different kinds of oxidative and reductive reactions are catalyzed by flavoproteins.
264 Rivlin and Pinto
A. Antioxidant Activity
In the wake of contemporary interest in the dietary antioxidants, one vitamin that is often
neglected as a member of this category is riboflavin. Riboflavin does not have significant
inherent antioxidant action, but powerful antioxidant activity is derived from its role as
a precursor to FMN and FAD. A major protective role against lipid peroxides is provided
by the glutathione redox cycle (35). Glutathione peroxidase breaks down reactive lipid
peroxides. This enzyme requires reduced glutathione, which in turn is regenerated from
its oxidized form (GSSG) by the FAD-containing enzyme glutathione reductase. Thus,
riboflavin nutrition should be critical for regulating the rate of inactivation of lipid peroxides.
Diminished glutathione reductase activity should be expected to lead to diminished
concentrations of reduced glutathione that serve as substrate for glutathione peroxidase
and glutathione S-transferase, and therefore would limit the rate of degradation of lipid
peroxides and xenobiotic substances (36).
Furthermore, the reducing equivalents provided by NADPH, the other substrate required
by glutathione reductase, are primarily generated by an enzyme of the pentose
monophosphate shunt, glucose-6-phosphate dehydrogenase. Taniguchi and Hara (37), as
well as our laboratory (38), have found that the activity of glucose-6-phosphate dehydrogenase
is significantly diminished during riboflavin deficiency. This observation provides
an additional mechanism to explain the diminished glutathione reductase activity in vivo
during riboflavin deficiency and the eventual decrease in antioxidant activity.
There have been a number of reports (39,40) in the literature indicating that riboflavin
deficiency is associated with compromised oxidant defense and furthermore that supplementation
of riboflavin and its active analogs improves oxidant status. Investigators
have shown that riboflavin deficiency is associated with increased hepatic lipid peroxidation
and that riboflavin supplementation limits this process (37–40). In our laboratory,
we have shown that feeding a riboflavin-deficient diet to rats increases basal as well as
stimulated lipid peroxidation (36).
B. Riboflavin and Malaria
Recent reports provide increasing evidence that riboflavin deficiency is protective against
malaria both in experimental animals and in humans (41,42). With dietary riboflavin defi-
ciency, parasitemia is decreased dramatically, and symptomatology of infection may be
diminished. In a study with human infants suffering from malaria, normal riboflavin nutritional
status was associated with high levels of parasitemia (43,44). In a similar fashion,
supplementation of iron and vitamins, which included riboflavin, to children resulted in
increased malarial parasitemia (45).
Further evidence for a beneficial role of riboflavin deficiency in malaria is provided
by studies utilizing specific antagonists of riboflavin, e.g., galactoflavin and 10-(4?-chlorophenyl)-
3-methylflavin (46). These flavin analogs, as well as newer isoalloxazines derivatives
(47), are glutathione reductase inhibitors and possess clear antimalarial efficacy (46).
The exact mechanism by means of which riboflavin deficiency appears to inhibit malarial
parasitemia is not yet established. One possibility relates to effects on the redox status of
erythrocytes, which is an important determinant of growth of malaria parasites. Protection
from malaria is afforded by several oxidant drugs, vitamin E deficiency, and certain genetic
abnormalities compromising oxidative defense (35).
It is well known that malaria parasites (Plasmodium berghei) are highly susceptible
Riboflavin (Vitamin B2) 265
to activated oxygen species. Parasites are relatively more susceptible than erythrocytes to
the damaging effects of lipid peroxidation (35). We have hypothesized that the requirement
of the parasites for riboflavin should be higher than that of the host cells and therefore
that marginal riboflavin deficiency should be selectively detrimental to parasites. Support
for this hypothesis is that the uptake of riboflavin and its conversion to FMN and FAD
are significantly higher in parasitized than in unparasitized erythrocytes and furthermore
that the rate of uptake of riboflavin is proportional (48) to the degree of parasitemia. These
results strongly suggest that parasites have a higher requirement for riboflavin than do
host erythrocytes.
C. Riboflavin and Homocysteine
A subject of great contemporary interest is the possible role of homocysteine in the pathogenesis
of vascular disease, including cardiovascular, cerebrovascular, and peripheral vascular
disorders (49). Blood levels of folic acid sensitively determine serum homocysteine
concentrations (50). As shown in Fig. 5, N-5-methyltetrahydrofolate is a cosubstrate with
homocysteine in its inactivation by conversion to methionine. Methylcobalamin is also a
coenzyme in this enzymatic reaction. Vitamin B6 is widely recognized for its importance
in the inactivation of homocysteine by serving as coenzyme of two degradative enzymes,
cystathionine ?-synthase and cystathioninase.
However, it is not commonly appreciated that riboflavin also has a role in homocysteine
metabolism, as the flavin coenzyme, FAD, is required by methyltetrahydrofolate
reductase, the enzyme responsible for converting N-5,10-methylenetetrahydrofolate into
N-5-methyltetrahydrofolate. Thus, the efficient utilization of dietary folic acid requires
adequate riboflavin nutrition. Furthermore, a mutation leading to a heat-sensitive form of
methylenetetrahydrofolate reductase has recently been identified (51). Its role in vascular
disease has not been defined. Further research is required to determine whether the serum
levels of homocysteine and the prevalence of vascular disease can be correlated directly
Fig. 5 Diagrammatic representation of homocysteine metabolism, showing the roles of folic acid,
B12, B6, and riboflavin. FAD is the coenzyme for methylenetetrahydrofolate reductase.
266 Rivlin and Pinto
with indices of riboflavin nutrition, and whether effects of marginal as well as overt defi-
ciency of riboflavin are clinically significant with respect to vascular disease.
D. Fat Metabolism
The important role of riboflavin in fat metabolism has been highlighted by recent demonstrations
that in certain rare inborn errors administration of riboflavin may be therapeutic.
In acyl-CoA dehydrogenase deficiency, infants present with recurrent hypoglycemia and
lipid storage myopathy and increased urinary excretion of organic acids. Clinical improvement
has occurred rapidly after riboflavin supplementation (52,53). Three varieties of the
disorder occur, all of which involve flavoproteins of various types. Five patients with a
mitochondrial disorder associated with NADH dehydrogenase deficiency were improved
by riboflavin treatment (54).
There is general agreement that dietary riboflavin intake at many times the RDA is without
demonstrable toxicity (10,55–57). Because riboflavin absorption is limited to a maximum
of about 25 mg at any one time (10), the consumption of megadoses of this vitamin
would not be expected to increase the total amount absorbed. Furthermore, classical animal
investigations showed an apparent upper limit to tissue storage of flavins that cannot be
exceeded under ordinary circumstances (58). The tissue storage capacity for flavins is
probably limited by the availability of proteins capable of providing binding sites. Thus,
protective mechanisms prevent tissue accumulation of excessive amounts of the vitamin.
Because riboflavin has very low solubility, even intravenous administration of the vitamin
would not introduce large amounts into the body. FMN is more water-soluble than ribo-
flavin but is not ordinarily available for clinical use.
Nevertheless, the photosensitizing properties of riboflavin raise the possibility of
some potential risks. Phototherapy in vitro leads to degradation of DNA and increases in
lipid peroxidation, which may have implications for carcinogenesis and other disorders.
Irradiation of rat erythrocytes in the presence of FMN increases potassium loss (59). Topical
administration of riboflavin to the skin may increase melanin synthesis by stimulation
of free-radical formation. Riboflavin forms an adduct with tryptophan and accelerates the
photo-oxidation of this amino acid (60). Further research is needed to explore the full
implications of the photosensitizing capabilities of riboflavin and its phosphorylated derivatives.
The photosensitization of vinca alkaloids by riboflavin may distort the results of
testing of the efficacy of cytotoxic drugs if the studies are carried out in the presence of
visible light, as is usually done (61). This property of riboflavin needs to be considered
in drug evaluations, inasmuch as cell death will occur even without the addition of the
drug being analyzed.
Riboflavin is capable of reacting with chromate (VI), forming a complex and then
increasing the DNA breaks due to a chromium-induced free-radical mechanism (62).
Treatment of mouse FM3A cells greatly increases mutation frequency and the extent of
cellular DNA damage in the presence of light (63). There appears to be increasing evidence
that in the presence of visible light, riboflavin and its degradative product, lumiflavin,
may enhance mutagenicity (64).
Riboflavin (Vitamin B2) 267
On the other hand, recent studies (65) confirm earlier reports (66) that riboflavin
deficiency may enhance carcinogenesis by increasing activation of carcinogens, particularly
nitrosamines. Riboflavin may possibly provide protection against damage to DNA
caused by certain carcinogens through its action as a coenzyme with a variety of cytochrome
P450 enzymes.
It is important to establish the role of riboflavin as a dietary factor capable of preventing
carcinogenesis while at the same time determining the full implication of the
photosensitizing actions of riboflavin on mutagenesis and carcinogenesis. There are reports
raising the possibility that deficient riboflavin nutritional status, together with shortages
of other vitamins, may possibly enhance development of precancerous lesions of the
esophagus in China (67,68) and in Russia (69).
There are a variety of methods available for analysis of riboflavin and its derivatives.
Bioassays measure the growth effect of vitamins but lack the precision of more sensitive
analytical procedures (70). Fluorometric procedures take advantage of the inherent fluorescent
properties of flavins (71). Some degree of purification of the urine or tissues may
be required before analysis as there is often significant interference by other natural substances
that leads to quenching of fluorescence and methodological artefacts. A procedure
has been developed for measuring riboflavin by competitive protein binding that is applicable
to studies in human urine (72). Riboflavin binds specifically to the avian egg white
riboflavin-binding protein (33) and thereby provides the basis for quantitative analysis
(73). Other procedures are also in use based on binding to specific apoenzymes, such as
d-amino acid oxidase. Currently, procedures using high-pressure liquid chromatography
(HPLC) have been widely applied as they have precision and can be utilized for analysis
of riboflavin in pure form as well as in biological fluids and tissues (74). HPLC is the
method most widely employed for determination of flavins in the blood and other tissues.
In clinical studies that involve individual patients as well as population groups, the
status of riboflavin nutrition is generally evaluated by determining the urinary excretion
of riboflavin (75) and the erythrocyte glutathione reductase activity coefficient (EGRAC)
(75). Urinary riboflavin determinations are made in the basal state, in random samples,
in 24-h collections, or after a riboflavin load test. Normal urinary excretion of riboflavin
is approximately 120 µg/g creatinine per 24 h or higher. It is useful to express urinary
excretion in terms of creatinine to verify the completeness of the collection and to relate
excretion to this biological parameter. Expressed in terms of the total amount, riboflavin
excretion in the normal adult is about 1.5–2.5 mg/day, which is very close to the
recommended dietary allowance of the National Academy of Sciences.
In deficient adult individuals, urinary riboflavin excretion is reduced to about 40
µg/g creatinine per 24 hr. Thus, deficient individuals have reduced urinary excretion,
reflecting diminished dietary intake and depleted body stores. Normal urinary excretion
is reduced with age, may be reduced by physical activity (as discussed below), and is
stimulated by elevated body temperature, treatment with certain drugs, and various stressful
conditions associated with negative nitrogen balance (76). Interpretation of urinary
riboflavin excretion must be made with these factors in mind.
Another potential drawback to utilizing urinary riboflavin excretion as an assessment
of nutritional status of this vitamin is that the amount excreted reflects recent intake very
sensitively. Thus, if an individual has been depleted for a long time but consumes a food
268 Rivlin and Pinto
item high in riboflavin, the urinary excretion determined a few hours later may not be in
the deficient range, but is likely to be normal or even elevated.
It is for this reason that attention has been directed to the development of assessment
techniques that more accurately reflect long-term riboflavin status. The method most
widely employed that largely meets these needs is assay of EGRAC as noted above. The
principle of the method is that the degree of saturation of the apoenzyme with its coenzyme,
FAD, should reflect the body stores of FAD. In deficient individuals, relative unsaturation
of the apoenzyme with FAD leads to decreased basal activity of the enzyme. Therefore,
the addition of FAD to the enzyme contained in a fresh erythrocyte hemolysate from
deficient individuals will increase activity in vitro to a greater extent than that observed
in a preparation from well-nourished individuals in whom the apoenzyme is more saturated
with FAD.
The EGRAC is the ratio of enzyme activity with to that without addition of FAD
in vitro. In general, most studies indicate that an activity coefficient of 1.20 or less indicates
adequate riboflavin status, 1.2–1.4 borderline-to-low status, and greater than 1.4 a clear
riboflavin deficiency (75,76).
It must be kept in mind that a number of physiological variables influence the results
of this determination. In the inherited disorder of glucose-6-phosphate dehydrogenase de-
ficiency, associated with hemolytic anemia, the apoenzyme has a higher affinity for FAD
than that of the normal erythrocyte that will affect the measured EGRAC. Thyroid function
affects glutathion reductase activity, the coefficient being elevated in hypothyroidism and
decreased in hyperthyroidism (77), reflecting the fact that hypothyroidism has many biochemical
features in common with those of riboflavin deficiency (17).
The latest RDAs issued by the Food and Nutrition Board (55) call for adult males
aged 19–50 years to consume 1.7 mg/day and 51 years of age or older, 1.4 mg. Adult
females from 19 to 50 years of age should consume 1.3 mg/day and from age 51 years
or older 1.2 mg. It is recommended that intake be increased to 1.6 mg/day during pregnancy,
to 1.8 mg/day early in lactation, and to 1.7 mg/day later in lactation. Infants should
consume 0.4–0.5 mg/day and children 0.8–1.2 mg/day, depending on age.
There has been some concern as to whether these figures are applicable to other
population groups around the world. Chinese tend to excrete very little riboflavin, and
their requirement may be lower than that of Americans (78). Adults in Guatemala appear
to have similar requirements in individuals older than 60 compared to those 51 years or
younger (79). This finding may not necessarily be relevant to populations of other countries.
The requirements of various national groups require further study. Environmental
factors, protein-calorie intake, physical activity, and other factors may have an impact on
riboflavin status. More research is needed on the requirements of the extremely old, who
form an increasingly large proportion of the population. They are also the population
group that consumes the largest number of prescribed and over-the-counter medications.
A point of interest is whether riboflavin requirements are increased in individuals
who exercise compared to those who are sedentary. In women aged 50–67 who exercised
vigorously for 20–25 min/day, 6 days a week, both a decrease in riboflavin excretion and
a rise in the EGRAC was noted, findings consistent with a marginal riboflavin-deficient
state, as shown in Table 2 (80). Supplementation with riboflavin did not, however, improve
exercise performance. These investigators observed compromised riboflavin status as well
in young women exercising vigorously (81). Similar observations of reduced urinary ribo-
flavin excretion and elevated EGRAC were made in young Indian males who exercised
actively (82).
Riboflavin (Vitamin B2) 269
Table 2 Group Means (n  7) for EGRAC and Urinary Riboflavin
Excretion During Both Exercise and Nonexercise Periodsa
Group Nonexercise Exercise
EGRAC 1.224  0.079b 1.283  0.067c
Urinary riboflavin (mg/day) 0.17  0.11 0.14  0.10d
EGRAC 1.070  0.031 1.109  0.045c
Urinary riboflavin (mg/day) 0.66  0.49 0.46  0.21d
a x  SD. EGRAC, erythrocyte glutathione reductase activity coefficient.
b Significantly greater than high-riboflavin group, p  0.0005.
c Significantly different from nonexercise: p  0.0001.
d Significantly different from nonexercise: p  0.01.
Source: Ref. 62.
To determine whether the status of riboflavin nutrition influences metabolic responses
to exercise, blood lactate levels were determined in a group of physically active
college students from Finland before and after the exercise period. A number of the students
were initially in a state of marginal riboflavin deficiency. Following supplementation
with vitamins, including riboflavin, that produced improvement in the elevated EGRAC,
the blood lactate levels were unaffected and were related only to the degree of exercise
Thus, to date, while exercise clearly produces biochemical abnormalities in riboflavin
metabolism, it has not been shown that these abnormalities lead to impaired performance,
nor has it been shown that riboflavin supplementation leads to improved exercise
Research was supported in part by the Clinical Nutrition Research Unit Grant 1-PO1-CA-
29502 from the National Institutes of Health, and by grants from the Stella and Charles
Guttman Foundation, the Sunny and Abe Rosenberg Foundation, the Rosenstiel Foundation,
the Isadore Rosenfeld Heart Foundation, and the Frank J. Scallon Foundation and
an industrial agreement with Wakunaga of America Co., Ltd.
1. E. V. McCollum and C. Kennedy, The dietary factors operating in the production of polyneuritis.
J. Biol. Chem. 24:491, 1916.
2. A. D. Emmett and G. O. Luros, Water soluble vitamins. I. Are the antineuritic and the growthpromoting
water soluble B vitamines the same? J. Biol. Chem. 43:265, 1920.
3. O. Warburg and W. Christian, Uber ein neues Oxydationsferment und sein Absorptionsspektrum.
Biochem. Z. 254:438, 1932.
4. K. G. Stern and E. R. Holiday, Zur Konstitution des Photo-flavins; Versuche in der Alloxazine-
Reihe. Ber. Dtsch. Chem. Gessellsch. 67:1104, 1934.
5. H. Theorell, Reindarstellung (kristallisation) des gelben Atmungsfermentes und die reversible
Spaltung desselben. Biochem. Z. 272:155, 1934.
270 Rivlin and Pinto
6. R. Kuhn, K. D. Reinemund, H. Kaltschmitt, R. Strobele, and H. Trischmann, Synthetisches
6,7-Dimethyl-9-d-riboflavin. Naturwissenschaften 23:260, 1935.
7. P. Karrer, K. Scopp, and F. Benz, Synthesis of flavins IV. Helv. Chim. Acta 18:426, 1935.
8. H. Theorell, Die Freie Eiweisskomponente des gelben Ferments und ihre Kupplung mit Lactoflavinphosphorsaure.
Biochem. Z. 290:293, 1937.
9. O. Warburg and W. Christian, Co-Ferment der d-Aminosaure-Deaminase. Biochem. Z. 295:
261, 1938.
10. D. B. McCormick, Riboflavin, in Modern Nutrition in Health and Disease, 8th ed. (M. E.
Shils, J. A. Olson, and M. Shike, eds.), Lea and Febiger, Philadelphia, 1994, p. 366.
11. K. Yagi, Y. Nakagawa, O. Suzuki, and N. Ohishi, Incorporation of riboflavin into covalentlybound
flavins in rat liver. J. Biochem. 79:841, 1976.
12. R. Rivlin, Medical progress: riboflavin metabolism. N. Engl. J. Med. 283:463, 1970.
13. J. T. Pinto and R. S. Rivlin, Regulation of formation of covalently bound flavins in liver and
cerebrum by thyroid hormones. Arch. Biochem. Biophys. 194:313, 1979.
14. A. H. Merrill, Jr., J. D. Lambeth, D. E. Edmondson, and D. B. McCormick, Formation and
mode of action of flavoproteins. Annu. Rev. Nutr. 1:281, 1981.
15. R. S. Rivlin, Vitamin deficiency, in Conns’ Current Therapy (R. E. Rakel, ed.), W. B. Saunders,
Philadelphia, 1994, p. 551.
16. S. Fass and R. S. Rivlin, Regulation of riboflavin-metabolizing enzymes in riboflavin defi-
ciency. Am. J. Physiol. 217:988, 1969.
17. R. S. Rivlin, Medical aspects of vitamin B2, in Chemistry and Biochemistry of Flavins (F.
Muller, ed.), CRC Press, Boca Raton, FL, 1990, p. 201.
18. J. A. Cimino, S. Jhangiani, E. Schwartz, and J. M. Cooperman, Riboflavin metabolism in the
hypothyroid human adult. Proc. Soc. Exp. Biol. Med. 184:121, 1987.
19. J. T. Pinto, Y. P. Huang, and R. S. Rivlin, Inhibition of riboflavin metabolism in rat tissues
by chlorpromazine, imipramine and amitriptyline. J. Clin. Invest. 67:1500, 1981.
20. P. Dutta, J. T. Pinto, and R. S. Rivlin, Antimalarial effects of riboflavin deficiency. Lancet
2:1040, 1985.
21. J. T. Pinto, J. P. Huang, and R. S. Rivlin, Mechanisms underlying the differential effects of
ethanol upon the bioavailability of riboflavin and flavin adenine dinucleotide. J. Clin. Invest.
79:1343, 1987.
22. B. Tandler, R. A. Erlandson, and E. L. Wynder, Riboflavin and mouse hepatic cell structure
and function. I. Ultrastructural alterations in simple deficiency. Am. J. Pathol. 52:69,
23. E. A. Williams, R. D. E. Rumsey, and H. J. Powers, Cytokinetic and structural responses of
the rat small intestine to riboflavin depletion. Br. J. Nutr. 75:315, 1996.
24. H. J. Powers, L. T. Weaver, S. Austin, and J. K. Beresford, A proposed intestinal mechanism
for the effect of riboflavin deficiency on iron loss in the rat. Br. J. Nutr. 69:553, 1993.
25. D. B. McCormick, Two interconnected B vitamins: riboflavin and pyridoxine. Physiol. Rev.
69:1170, 1989.
26. G. A. Goldsmith, Riboflavin deficiency, in Riboflavin (R. Rivlin, ed.), Plenum Press, New
York, 1975, p. 221.
27. R. S. Rivlin, Disorders of vitamin metabolism: deficiencies, metabolic abnormalities and excesses,
in Cecil Textbook of Medicine, 19th ed. (J. H. Wyngaarden, L. H. Smith, Jr., J. C.
Bennett, and F. Plum, eds.), W. B. Saunders, Philadelphia, 1991, p. 1170.
28. R. L. Wanner, Effects of commercial processing of milk and milk products on their nutrient
content, in R. S. Harris and H. V. Loesecke (eds.), The Nutritional Evaluation of Food Processing,
John Wiley and Sons, New York, 1960, p. 173.
29. A. H. Ensminger, M. E. Ensminger, J. E. Konlande, and J. R. K. Robson, Food and Nutrition
Encyclopedia, CRC Press, Boca Raton, FL, 1994, p. 1927.
30. D. B. McCormick, Riboflavin, in Present Knowledge in Nutrition, 6th ed. (M. L. Brown, ed.),
International Life Sciences Institute, Washington, DC, 1990, p. 146.
Riboflavin (Vitamin B2) 271
31. J. Zempleni, J. R. Galloway, and D. B. McCormick, Pharmacokinetics of orally and intravenously
administered riboflavin in healthy humans. Am. J. Clin. Nutr. 63:54, 1996.
32. H. M. Said and R. Mohammadkhani, Uptake of riboflavin across the brush border membrane
of rat intestine: regulation by dietary vitamin levels. Gastroenterology 105:1294, 1993.
33. C. O. Clagett, Genetic control of the riboflavin carrier protein. Fed. Proc. 30:127, 1971.
34. J. Dancis, J. Lehanka, and M. Levitz, Placental transport of riboflavin: differential rates of
uptake at the maternal and fetal surfaces of the perfused human placenta. J. Obstet. Gynecol.
158:204, 1988.
35. P. Dutta, Disturbances in glutathione metabolism and resistance to malaria: current understanding
and new concepts. J. Soc. Pharm. Chem. 2:11, 1993.
36. R. S. Rivlin and P. Dutta, Vitamin B2 (riboflavin). Relevance to malaria and antioxidant activity.
Nutr. Today 30:62, 1995.
37. M. Taniguchi and T. Hara, Effects of riboflavin and selenium deficiencies on glutathione and
related enzyme activities with respect to lipid peroxide content of rat livers. J. Nutr. Vitamin
29:283, 1983.
38. P. Dutta, J. Seirafi, D. Halpin, J. T. Pinto, and R. S. Rivlin, Acute ethanol exposure alters
hepatic glutathione metabolism in riboflavin deficiency. Alcohol 12:43, 1995.
39. T. Miyazawa, K. Tsuchiya, and T. Kaneda, Riboflavin tetrabutyrate: an antioxidative synergist
of alpha-tocopherol as estimated by hepatic chemiluminescence. Nutr. Rep. Int. 29:157,
40. T. Miyazawa, C. Sato, and T. Kaneda, Antioxidative effects of ?-tocopherol and riboflavinbutyrate
in rats dosed with methyl linoleate hydroperoxide. Agric. Biol. Chem. 47:1577, 1983.
41. P. Kaikai and D. I. Thurnham, The influence of riboflavin deficiency on Plasmodium berghei
infections in rats. Trans. R. Soc. Trop. Med. Hyg. 77:680, 1983.
42. B. S. Das, D. B. Das, R. N. Satpathy, J. K. Patnaik, and T. K. Bose, Riboflavin deficiency
and severity of malaria. Eur. J. Clin. Nutr. 42:277, 1988.
43. D. I. Thurnham, S. J. Oppenheimer, and R. Bull, Riboflavin status and malaria in infants in
Papua New Guinea. Trans. R. Soc. Trop. Med. Hyg. 77:423, 1983.
44. S. J. Oppenheimer, R. Bull, and D. I. Thurnham, Riboflavin deficiency in Madang infants.
Papua N. Guinea Med. J. 26:17, 1983.
45. C. J. Bates, H. J. Powers, and W. H. Lamb, Antimalarial effects of riboflavin deficiency in
Madang infants. Papua N. Guinea Med. J. 26:17, 1983.
46. K. Becker, R. I. Christopherson, W. B. Cowden, N. H. Hunt, and R. H. Schirmer, Flavin
analogs with antimalarial activity as glutathione reductase inhibitors. Biochem. Pharmacol.
39:59, 1990.
47. A. Schonleben-Janas, P. Kirsch, P. R. E. Mittl, R. H. Schirmer, and R. L. Krauth-Siegel,
Inhibition of human glutathione reductase by 10-arylisoalloxazines: crystallographic, kinetic
and electrochemical studies. J. Med. Chem. 39:1549, 1996.
48. P. Dutta, Enhanced uptake and metabolism of riboflavin in erythrocytes infected with Plasmodium
falciparium. J. Protozool. 38:479, 1991.
49. I. A. Graham, L. E. Daly, H. M. Refsum, et al., Plasma homocysteine as a risk factor for
vascular disease. The European concerted project. JAMA 277:1775, 1997.
50. C. J. Boushey, S. A. A. Beresford, G. S. Omenn, and A. G. Motulsky, A quantitative assessment
of plasma homocysteine as a risk factor for vascular disease. JAMA 274:1049, 1995.
51. P. Frosst, H. J. Blom, R. Milos, et al., A candidate genetic risk factor for vascular disease: a
common mutation in methylenetetrahydrofolate reductase. Nature Genet. 10:111, 1995.
52. G. Uziel, B. Garavaglia, E. Ciceri, I. Moroni, and M. Rimoldi, Riboflavin-responsive glutaric
aciduria type II presenting as a leukodystrophy. Pediatr. Neurol. 13:333, 1995.
53. P. L. J. A. Bernsen, F. J. M. Gabreels, W. Ruitenbeek, and H. L. Hamburger, Treatment of
complex I deficiency with riboflavin. J. Neurol. Sci. 118:181, 1993.
54. U. A. Walker and E. Byrne, The therapy of respiratory chain encephalomyopathy: a critical
review of the past and present perspective. Acta Neurol. Scand. 92:273, 1995.
272 Rivlin and Pinto
55. Food and Nutrition Board, National Research Council. Recommended Dietary Allowances,
10th ed., National Academy Press, Washington, DC, 1989, p. 132.
56. R. S. Rivlin, Effect of nutrient toxicities (excess) in animals and man: riboflavin, in Handbook
of Nutrition and Foods (M. Recheigl, ed.), CRC Press, Boca Raton, FL, 1979, p. 25.
57. J. M. Cooperman and R. Lopez, Riboflavin, in Handbook of Vitamins: Nutritional, Biochemical,
and Clinical Aspects (L. J. Machlin, ed.), Marcel Dekker, New York, 1984, p. 299.
58. H. B. Burch, O. H. Lowry, A. M. Padilla, and A. M. Combs, Effects of riboflavin deficiency
and realimentation on flavin enzymes of tissues. J. Biol. Chem. 223:29, 1956.
59. F. S. Ghazy, T. Kimura, S. Muranishi, and H. Sezaki, The photodynamic action of riboflavin
on erythrocytes. Life Sci. 21:1703, 1977.
60. M. Salim-Hanna, A. M. Edwards, and E. Silva, Obtention of a photo-induced adduct between
a vitamin and an essential amino acid. Binding of riboflavin to tryptophan. Int. J. Vitam. Nutr.
Res. 57:155, 1987.
61. C. Granzow, M. Kopun, and T. Krober, Riboflavin-mediated photosensitization of vinca alkaloids
distorts drug sensitivity assays. Cancer Res. 55:4837, 1995.
62. M. Sugiyama, K. Tsuzuki, X. Lin, and M. Costa, Potentiation of sodium chromate (VI)–
induced chromosomal aberrations and mutation by vitamin B2 in Chinese hamster V79 cells.
Mutat. Res. 283:211, 1992.
63. T. Bessho, K. Tano, S. Nishimura, and H. Kasai, Induction of mutations in mouse FM3A cells
by treatment with riboflavin plus visible light and its possible relation with formation of 8-
hydroxyguanine (7,8-dihydro-8-oxoguanine) in DNA. Carcinogenesis 14:1069, 1993.
64. H. Kale, P. Harikumar, S. B. Kulkarni, P. M. Nair, and M. S. Netrawali, Assessment of the
genotoxic potential of riboflavin and lumiflavin. B. Effect of light. Mutat. Res. 298:17, 1992.
65. R. P. Webster, M. D. Gawde, and R. K. Bhattacharya, Modulation of carcinogen-induced
DNA damage and repair enzyme activity by dietary riboflavin. Cancer Lett. 98:129, 1996.
66. R. S. Rivlin, Riboflavin and cancer: a review. Cancer Res. 33:1997, 1973.
67. N. Munoz, J. Wahrendorf, L. J. Bang, M. Crespi, and A. Grassi. Vitamin intervention on
precancerous lesions of the esophagus in a high-risk population in China. Ann. N.Y. Acad.
Sci. 534:618, 1988.
68. J. Wahrendorf, N. Munoz, J. B. Lu, D. I. Thurnham, M. Crespi, and F. X. Bosch. Blood retinol
and zinc riboflavin status in relation to precancerous lesions of the esophagus: findings from
a vitamin intervention trial in the People’s Republic of China. Cancer Res. 48:2280, 1988.
69. D. G. Zaridze, J. U. Bukin, Y. N. Orlov, et al., Relationship between esophageal mucosa
pathology and vitamin deficit in population with high frequency of esophageal cancer. Vop.
Onkol. 35:939, 1989.
70. H. Baker and O. Frank, Analysis of riboflavin and its derivatives in biologic fluids and tissues,
in Riboflavin (R. S. Rivlin, ed.), Plenum Press, New York, 1975, p. 49.
71. O. A. Bessey, O. H. Lowry, and R. H. Love, Fluorometric measure of the nucleotides of
riboflavin and their concentration in tissues, J. Biol. Chem. 180:755, 1949.
72. A. G. Fazekas, C. E. Menendez, and R. S. Rivlin, A competitive protein binding assay for
urinary riboflavin, Biochem. Med. 9:167, 1974.
73. M. J. Kim, H. J. Kim, J. M. Kim, B. Kim, S. H. Han, and G. S. Cha, Homogeneous assays for
riboflavin mediated by the interaction between enzyme-biotin and avidin-riboflavin conjugates.
Anal. Biochem. 231:400, 1995.
74. J. L. Chastain and D. B. McCormick, Flavin catabolites: identification and quantitation in
human urine. Am. J. Clin. Nutr. 46:830, 1987.
75. H. E. Sauberlich, J. H. Judd, G. E. Nichoalds, H. P. Broquist, and W. J. Darby, Application
of the erythrocyte glutathione reductase assay in evaluating riboflavin nutritional status in a
high school student population. Am. J. Clin. Nutr. 25:756, 1972.
76. R. S. Rivlin, Riboflavin in Present Knowledge in Nutrition (E. E. Ziegler and L. J. Filer, Jr.,
eds.), ILSI Press, Washington, DC, 1996, p. 167.
77. C. E. Menendez, P. Hacker, M. Sonnenfeld, R. McConnell, and R. S. Rivlin, Thyroid hormone
Riboflavin (Vitamin B2) 273
regulation of glutathione reductase activity in rat erythrocytes and liver. Am. J. Physiol. 226:
1480, 1974.
78. T. A. Brun, J. Chen, T. C. Campbell, et al., Urinary riboflavin excretion after a load test in
rural China as a measure of possible riboflavin deficiency. Eur. J. Clin. Nutr. 44:195, 1990.
79. W. A. Boisvert, I. Mendoza, C. Castenada, et al., Riboflavin requirement of healthy elderly
humans and its relationship to macronutrient composition of the diet. J. Nutr. 123:915, 1993.
80. L. R. Trebler Winters, J.-S. Yoon, H. J. Kalkwarf, J. C. Davies, M. G. Berkowitz, J. Haas,
and D. A. Roe, Riboflavin requirements and exercise adaptation in older women. Am. J. Clin.
Nutr. 56:526, 1992.
81. A. Z. Belko, E. Obarzanek, R. Roach, M. Rotter, G. Urban, S. Weinberg, and D. A. Roe,
Effects of aerobic exercise and weight loss on riboflavin requirements of moderately obese,
marginally deficient young women. Am. J. Clin. Nutr. 40:553, 1984.
82. M. J. Soares, K. Satyanarayana, M. S. Bamji, C. M. Jacobs, Y. Venkata Ramana, and S.
Sudhakar Rao, The effect of exercise on the riboflavin status of adult men. Br. J. Nutr. 69:
541, 1993.
83. M. Fogelholm, I. Ruokonen, J. T. Laakso, T. Vuorimaa, and J.-J. Himberg, Lack of association
between indices of vitamin B1, B2 and B6 status and exercise-induced blood lactate in young
adults. Int. J. Sport. Nutr. 3:165, 1993.
This Page Intentionally Left Blank
Mahidol University, Bangkok, Thailand
Although Neiching, the Chinese medical book, mentioned beriberi in 2697 bc, it was not
known for centuries that this illness was due to thiamine deficiency. In 1884, Takaki, a
surgeon general of the Japanese navy, concluded from his observations that beriberi was
caused by a lack of nitrogenous food components in association with excessive intake of
nonnitrogenous food. Though he did not reach the correct conclusion, the dietary changes
introduced into the Japanese navy resulted in a drastic reduction in the incidence of beriberi.
In 1890, the Dutch physician Eijkman, working in Java, discovered that fowls fed
boiled polished rice developed a polyneuritis that resembled beriberi in humans, and this
polyneuritis could be prevented or cured by rice bran or polishings. He suggested that the
toxic principle was present in polished rice but could be neutralized by some protective
factor in rice polishings. His associate, Grijns, in 1901 extracted a water-soluble protective
factor for polyneuritis from rice bran. In 1911, Funk, a chemist at the Lister Institute in
London, was convinced that he had isolated the antiberiberi principle possessing an amine
function from rice bran extracts and named it ‘‘vitamine.’’ However, his crystalline substance
was shown later to have little antineuritic activity (1–3).
Studies by Frazier and Stanton in 1909–1915 showed that human beriberi was a
deficiency disease that responded to extracts from rice polishings. This led Vedder, a U.S.
army physician, to recommend to Roger Williams, a chemist working in Manila, that the
protective substance be isolated. In 1926, Jansen and Donath, Dutch chemists working in
Java, succeeded in isolating and crystallizing antiberiberi factor from rice bran extracts.
The trivial name ‘‘aneurine’’ was suggested by Jansen. Later, Roger Williams isolated
sufficient quantities of thiamine to elucidate its structure. The chemical synthesis of thiamine
was accomplished in 1936 (1–3).
276 Tanphaichitr
Fig. 1 Structural formulas of thiamine and thiamine pyrophosphate.
The biochemical functions of thiamine were identified in 1935 by Thompson and
Johnson who showed that blood pyruvate levels were elevated in persons with thiamine
deficiency. During 1936–1938, Peters and his colleagues at Oxford University demonstrated
that thiamine was essential for pyruvate metabolism; the term biochemical lesion
was coined to describe the failure of thiamine-deficient pigeons to metabolize pyruvate.
In 1937, Lohman and Schuster discovered the active coenzyme form of thiamine to be
thiamine pyrophosphate (TPP; cocarboxylase) (Fig. 1). In 1953, Horecker and his associates
established the coenzymic role of TPP for transkotase (2–5).
A. Structure and Nomenclature
The chemical name of thiamine, formerly known as vitamin B1, vitamin F, aneurine, or
thiamine, is 3-(4-amino-2-methylpyrimidin-5-ylmethyl)-5-(2-hydroxyethyl)-4-methylthiazolium
(Fig. 1) (6). The free vitamin is a base. It is isolated or synthesized and handled
as a solid thiazolium salt, e.g., thiamine chloride hydrochloride, thiamine mononitrate.
B. Synthesis
1. Chemical Synthesis
The original synthesis of thiamine was accomplished almost simultaneously in 1936/37
by Williams et al., Todd and Bergel, and Andersag and Westphal by slightly different
procedures that are used today with only slight modifications for commercial production.
Either the pyrimidine and thiazole rings can be prepared separately and condensed via
the bromide, or the pyrimidine ring can be synthesized and the thiazole ring added to it.
Steps of chemical synthesis of thiamine have been reviewed (2).
2. Biosynthesis
Thiamine cannot be synthesized by animals to any significant extent. Many microbial
species can synthesize thiamine if pyrimidine and thiazole are available, and still others
have a dependence on thiamine. Higher plants can synthesize thiamine de novo. In these
Thiamine 277
organisms, the pyrimidine precursor of thiamine is synthesized by a different pathway
from the pyrimidine precursor of the pyrimidine nucleotides (2). Thiamine biosynthesis
requires phosphorylation of pyrimidine and thiazole before synthesis of the thiamine monophosphate
(TMP) molecule by thiamine phosphate pyrophosphorylase. TMP is then
hydrolyzed by phosphatase to yield free thiamine. Thiamine pyrophosphokinase, which
is responsible for the synthesis of TPP from the free thiamine, has been found in yeast
and mammalian tissue. E. coli can synthesize TPP from TMP. Thiamine triphosphate
(TTP) is synthesized in brain by TPP-ATP phosphoryltransferase (Fig. 2). However, studies
of 14C-thiamine turnover show very little incorporation of 14C into TTP. TPP must be
bound to an endogenous protein to act as a substrate, whereas free TPP does not affect
the reaction (3).
C. Chemical Properties
Thiamine hydrochloride is a white crystalline substance. It is readily soluble in water,
only partly soluble in alcohol and acetone, and insoluble in other fat solvents. In the dry
form, it is stable at 100°C. The thiamine stability in aqueous solutions depends upon pH.
Below pH 5 it is quite stable to heat and oxidation; above pH 5 it is rapidly destroyed
Fig. 2 Biosynthetic pathway of thiamine and its phosphorylated derivatives. Pyrimidine, 2-
methyl-4-amino-5-hydroxymethylpyrimidine; thiazole, 4-methyl-5-(2-hydroxymethyl)thiazole; MP,
monophosphate; TTP, thiamine triphosphate.
278 Tanphaichitr
Fig. 3 Structural formula of thiochrome.
by autoclaving, and at pH 7.0 or higher by boiling. Thiamine is readily cleaved at the
methylene bridge into 2-methyl-4-amino-5-methylpyrimidylsulfonate and 4-methyl-5-(2-
hydroxyethyl)thiazole by sulfite treatment at pH 6.0 or above. At pH 8.0 or above, thiamine
turns yellow and is destroyed by a complex series of irreversible reactions. In strong
alkaline solution with the presence of oxidizing agents, e.g., potassium ferricyanide, thiamine
is converted to thiochrome (Fig. 3), which is fluorescent and is used to determine
thiamine in foods, feeds, pharmaceutical preparations, and biological fluids. Thiamine is
precipitated by iron and ammonium citrate, tannin, and various alkaloids (2,4,5).
Thiamine forms esters by combining with various acids at the hydroxyethyl side
chain. The most important esters are TMP, TPP, and TTP. The chemical synthesis of
TPP by phosphorylation of thiamine and the separation from TMP and TTP provide its
commercial quantities (2). The two most common commercial preparations of thiamine are
thiamine hydrochloride and thiamine mononitrate, which are widely available in various
countries. Other commercial forms that are produced by Japanese pharmaceutical industries
include thiamine allyl disulfide (TAD), thiamine propyl disulfide (TPD), thiamine
tetrahydrofurfuryl disulfide (TTFD), and O-benzoyl thiamine disulfide (Fig. 4). They are
more lipid-soluble and less water-soluble, better retained in the body than thiamine hydrochloride,
and readily converted to thiamine in vivo (2,7).
D. Molecular Structure and Biological Activity
Studies in animals using various thiamine analogs have shown that both pyrimidine and
thiazole moieties are needed for its vitaminic activity, which is maximal when only one
methylene group bridges the two moieties. In the thiazole portion, the quaternary nitrogen
and a hydroxyethyl group at C-5 are needed, as is the amino group at C-4 in the pyrimidine
portion (Fig. 1) (2,4,5).
E. Thiamine Analogs with Antagonistic Action
Several synthesized thiamine analogs are shown to be thiamine antagonists, which are
employed to produce thiamine deficiency in animals. They are categorized into compounds
having substitutions on the pyrimidine or thiazole rings, compounds having modifications
of the pyrimidine or thiazole rings, and compounds having anticoccidial action. The representatives
of these groups of thiamine antagonists are oxythiamine, pyrithiamine, and amprolium
(Fig. 5) (2,4).
1. Compounds with Substitutions on the Pyrimidine or Thiazole Rings
Extensive studies of thiamine-catalyzed reactions in both nonenzymatic and enzymatic
systems by Breslow and Krampitz have shown that the catalytic function of thiamine
involves C-2 of the thiazole with its ability to form a carbanion. Thus, substitution at C-
2 of thiazole, i.e., 2-methylthiamines completely eliminates both its catalytic and biologic
activities (2).
Thiamine 279
Fig. 4 Structural formulas of thiamine derivatives.
An amino group at the 4-position on the pyrimidine ring plays an essential role in
the release of the aldehyde adduct from the thiazole C-2 after splitting of the ECECE
bond. Oxythiamine is a potent thiamine antagonist because it can form oxythiamine pyrophosphate,
which can bind with thiamine-dependent apoenzymes but cannot complete the
reaction (2).
2. Compounds with Modifications of the Pyrimidine or Thiazole Rings
Among various compounds having modifications of the pyrimidine or thiazole rings, only
pyrithiamine, in which a ECHCCHE has been substituted for S, is a potent thiamine
Among the aforementioned two groups of thiamine antagonists, only oxythiamine
and pyrithiamine have been studied extensively in animals. Oxythiamine produces most
280 Tanphaichitr
Fig. 5 Structural formulas of thiamine antagonists.
of the usual symptoms and biochemical changes associated with thiamine deficiency. The
anorexia and inanition after oxythiamine treatment appear much earlier and more severe
than those after pyrithiamine treatment or thiamine deprivation. Since oxythiamine is incapable
of crossing the blood–brain barrier, it never produces the neurological manifestations.
On the contrary, pyrithiamine produces chiefly neurological manifestations (2).
3. Compounds Having Anticoccidial Action
Chick anticoccidial compounds contain pyrimidine, which is combined through the methylene
bridge to a quaternary nitrogen of a pyridine ring. Since they do not contain a hydroxyethyl
group they cannot be phosphorylated. Thus, they are not good inhibitors of the TPPdependent
enzymes. A large number of anticoccidial compounds have been produced by
various substitutions on the pyrimidine and pyridine rings. In small doses they inhibit
chiefly thiamine transport across bacterial cell walls and hence exert anticoccidial action,
whereas at larger doses they also inhibit thiamine transport across the chick intestinal wall
and hence produce thiamine deficiency in the chicks. Amprolium, 1-(4-amino-2-n-propyl-
5-pyrimidinylmethyl)-2-picolinium bromide hydrobromide, is most extensively used commercially
A. Food Sources
Table 1 shows thiamine content in milligrams per 100 g edible portion of selected foods
compiled from the food composition table for use in East Asia published by the U.S.
Department of Health, Education and Welfare and the Food and Agricultural Organization
Thiamine 281
Table 1 Thiamine Content of Some Foods
Thiamine Thiamine
Food and description (mg/100 ga) Food and description (mg/100 g)
Yeast Mung bean:
Baker’s, dried 2.33 Whole seeds, dried 0.53
Brewer’s, debittered 15.61 Strip, dried 0.05
Pork Peanut:
Carcass, fresh, lean 0.69 Raw 0.97
Ham, smoked, lean 0.70 Boiled 0.56
Beef Soybean:
Carcass, fresh, very lean 0.07 Whole mature seeds, dried 0.66
Chicken, raw Milk, unsweetened 0.05
Very young birds 0.08 Curd, raw, plain 0.04
Mature birds 0.08 Maize, whole-kernel:
Liver, raw Yellow 0.29
Beef 1.68 Popcorn, popped, oil added 0.12
Chicken 1.92 Starch 0
Duck 1.32 Rice:
Hog 3.57 Bran 1.26
Kidney, raw Home-pounded 0.20
Beef 1.58 Milled 0.10
Hog 1.31 Cooked, milled 0.02
Abalone, raw 0.24 Noodles, freshly made 0.04
Bass, sea, raw 0.15 Noodles, cooked Trace
Catfish, freshwater, raw 0.10 Wheat:
Crab, sea, blue 0.05 Whole grain, hard, red winter 0.37
Lobster, raw 0.01 Germ 2.10
Duck eggs, raw: Bran 0.54
Whole 0.16 Asparagus: green, raw 0.15
White Trace Bamboo shoots, raw 0.09
Yolk 0.54 Cabbage, Chinese, raw 0.07
Hen eggs, raw: Carrot, raw 0.06
Whole 0.10 Apple, raw 0.02
White 0.01 Banana, ripe 0.03
Yolk 0.24 Guava, raw 0.05
Cheese, blue or cheddar 0.04 Orange, raw 0.04
Ice cream, regular 0.05 Butter, salted 0.01
Milk, cow: fluid, 3.5% fat 0.04 Lard 0
Milk, human, fluid, whole 0.02 Soybean oil 0
Tallow 0.07
a Edible portion.
Source: Compiled from Food Composition Table for Use in East Asia (8).
of the United Nations (8). Though thiamine is found in a large variety of animal and
vegetable products, it is abundant in only a few foods. Excellent sources of thiamine are
yeast, especially dried brewer’s yeast; liver and kidney; lean pork; and legumes, including
mung bean, soybean, and peanut. In cereal grains, thiamine is low in the endosperm but
high in the germ. Thus, the thiamine contents in rice bran and home-pounded rice is higher
282 Tanphaichitr
than that in milled rice. In eggs, thiamine is also unevenly distributed, being negligible
in egg white but present in egg yolk. Thiamine is absent in fats, oils, and refined sugars.
Milk and milk products, sea foods, fruits, and vegetables are not good sources of thiamine.
Foods lacking in thiamine are manufactured, i.e., milled rice, cereal flours, refined
sugars, separated animal fats, and separated vegetable oils. It should be emphasized that
Table 1 shows food sources rich in thiamine without consideration to the amount being
regularly consumed. For example, yeast and wheat germ are rich sources of thiamine, but
they are normally eaten in small amounts.
In most animal products, 95–98% of thiamine is present as TMP, TPP, and TTP,
with about 80–85% as TPP, whereas in plant products thiamine occurs in the nonphosphorylated
form (2).
B. Bioavailability
Thiamine status depends not only on the thiamine content of foods but on its bioavailability.
Factors influencing bioavailability are thiamine losses after handling and processing,
ethanol consumption, presence of antithiamine factors (ATFs), and folate and protein status
1. Thiamine Losses After Handling and Processing of Foods
As noted above, several factors, including pH, temperature, oxidation, solubility, radiation,
and the use of sulfites in handling and processing of foods, affect thiamine losses.
a. pH
Thiamine is stable at acid pH and becomes unstable at pH 7.0 or higher. In cooking
practice, the addition of sodium bicarbonate to green beans and peas to retain green color
or to dried beans to facilitate softening can lead to large losses of thiamine (2,5).
b. Temperature
Thiamine in tissues that is bound to protein is more stable to thermal destruction than is
free thiamine. Thiamine losses on cooking and canning of meats, baking of breads, and
cooking of vegetables are 25–85%, 5–35%, and 0–60%, respectively. The variation depends
on temperature, time, and types of food products. In pasteurization, sterilization,
spray-drying, roller-drying, and condensing of milk, thiamine losses are 9–20%, 30–50%,
10%, 15%, and 40%, respectively. Though freezing has no effect on the thiamine content
of foods, thawing causes thiamine losses into the drip fluid. Thiamine losses on storage
of canned fruits and vegetables at usual room temperatures are 0–20% for a year. Processing
foods at higher temperature and pH in the presence of oxygen or other oxidants
lead to the formation of thiamine sulfides and disulfides, thiochrome, and other oxidation
products. Only thiamine sulfides and disulfides still retain the biological activity of thiamin
Thiamine is cleaved by residual chlorine (0.2–2.0 ppm of chlorine) in proportion
to the rise in temperature, pH, and concentration of residual chlorine. When rice is boiled
in an electric rice cooker, the thiamine in the rice is cleaved by residual chlorine. When
rice is cooked in tap water containing 200 ng/mL of chlorine or distilled water without
chlorine, thiamine losses from rice are 56.5% and 32.2%, respectively. The degradation
of thiamine depends on the amount of ionization of residual chlorine governed by the pH
of water. At pH below 5.0, the residual chlorine is present as HClO and the ratio of ClO
increases in proportion to an increase in pH:
Thiamine 283
HClO s H  ClO
pH 5 < pH 7.5 > pH 10
Thus, ClO plays a key role in the cleavage of thiamine to hydroxymethylpyrimidine
and 4-methyl-5-?-hydroxymethythiazide. Though hydroxymethylpyrimidine can induce
convulsions in mice, the amount converted from thiamine in rice by chlorine in tap water
is so small that its toxic effect should be negligible. However, thiamine loss in polished
rice due to residual chlorine during the cooking process needs attention because rice is
the major source of dietary thiamine in Asian population (11).
c. Solubility
Since thiamine is highly water-soluble, significant thiamine losses occur during cooking
of foods when there is the excessive use of water that is then discarded. About 85% of
thiamine is lost by discarding the water after soaking the rice (5,9,10). Though parboiling
rice causes the thiamine to move from the outer layers to the inner layers of the rice
kernel, thus leading to greater thermal stability and smaller losses due to washing of the
thiamine, parboiled rice is not widely available or widely consumed by the Asian population.
d. Oxidation
Processing foods in the presence of oxygen or other oxidants can lead to the formation
of thiamine sulfides and disulfides, thiochrome, and other oxidation products. This can
lead to irreversible loss of thiamine activity (2).
e. Radiation
Thiamine is also destroyed by x-rays, ?-rays, and UV irradiation. UV irradiation leads to
the production of 2-methyl-4-amino-5-aminoethylpyrimidine (2).
f. Sulfites
Sulfites destroy thiamine. Sulfites that form in the treatment of fruits during dehydration
with SO2 destroy most of the thiamine present (5).
2. Ethanol
Excessive alcohol ingestion is associated with thiamine deficiency. Ethanol given orally
or intravenously inhibits intestinal thiamine uptake (2,5,12,13). Ethanol reduces the rate
of intestinal absorption and the net transmural flux of thiamine. Ethanol inhibits only the
active but not the passive component of thiamine transport by impeding the cellular exit
of thiamine across the basolateral or serosal membrane. The impairment of thiamine exit
from the enterocyte correlates with the Na,K-ATPase activity. This relationship may be
at least partly related to the effect of ethanol in increasing the fluidity of enterocyte brushborder
and basolateral membranes (13). The maximum absorption of 35S-thiamine in alcoholics
may be as low as 1.5 mg, similar to that seen in patients with intestinal resection
in whom the reduction is attributable to a decrease in receptor sites (12).
3. ATFs (Antithiamine Factors)
ATFs occurring in foods can alter thiamine structure and reduce the biological activity
of thiamine. ATFs are divided into two groups, according to their stability to heat, i.e.,
thermolabile and thermostable.
284 Tanphaichitr
a. Thermolabile ATFs
The thermolabile ATFs include thiaminase I (EC and thiaminase II (EC
Thiaminase I is found in viscera of freshwater fish (mainly in viscera), shellfish, fern, and
a limited number of sea fish and plants, and produced in microorganisms, e.g., Bacillus
thiaminolyticus and Clostridium thiaminolyticus. It catalyzes the cleavage of thiamine by
an exchange reaction with an organic base or a sulfhydryl compound involving a nucleophilic
displacement on the methylene group of the pyrimidine moiety of thiamine. Thiaminase
II is found in several microorganisms, e.g., Bacillus aneurinolyticus, Candida
aneurinolytica, Trichosporon, and Oospora. It directly catalyzes the hydrolysis of thiamine
(Fig. 6) (14,15).
Thiaminases are usually inactive in living cells. They are activated when they are
homogenized in a water solution around pH 4–8 or excreted from cells or microorganisms
into the medium. The thiamine cleavage in the diets containing thiaminase can occur
during storage or preparation prior to ingestion or during passing through the gastrointestinal
tract. Continual intake of raw freshwater fish with or without fermentation, raw shell-
fish, and ferns are risk factors for the development of thiamine deficiency (5,9,10,14–16).
b. Thermostable ATFs
The thermostable ATFs have been demonstrated in fern, tea, betel nut, and a large number
of other plants and vegetables and some animal tissues. In animal tissues, it is thought
that myoglobin, hemoglobin, and hemin bind thiamine. The ATFs found in plants and
vegetables are related to ortho- and para-polyphenolic compounds, e.g., caffeic acid (3,4-
dihydroxycinnamic acid), chlorogenic acid [3-(3,4-dihydroxycinnamoyl) quinic acid], and
tannic acid (tannin) (5,9). In vitro studies reveal that the antithiamine activity of polyphenols
requires pH 6.5 or higher and oxygen. The high pH is necessary for the ionization
of the polyphenols and the opening of the thiazole moiety of thiamine at C-2 to yield the
SH form of thiamine. Oxygen helps the oxidation and polymerization of polyphenols to
yield active quinones and relatively less active polymerized products. Oxidation of the
SH form of thiamine by the active quinones leads to the formation of thiamine disulfide.
Further hydrolysis and oxidation yield products that do not exhibit biological activity of
thiamine. Ascorbic acid and other reducing agents prevent the formation of quinone, thiamine
disulfide, or both. However, chemical modification of thiamine by the oxidation of
Fig. 6 Clevage of thiamine by thiaminase I and thiaminase II. Thiaminase I activity requires
organic bases or thiol compounds.
Thiamine 285
quinone may not be the only mechanism to reduce thiamine bioavailability. High concentrations
of divalent cations, including calcium and magnesium ions present in water, augment
the precipitation of thiamine by tannin, which makes thiamine less available for
intestinal absorption. Ascorbic, tartaric, and citric acids present in many fruits and vegetables
can lower such precipitation presumably by sequestering divalent cations.
Human studies have shown that tea drinking, tea leaf chewing, coffee or decaffeinated
coffee drinking, and betel nut chewing lead to biochemical thiamine depletion.
Ascorbic acid intake from either pharmaceutical preparations or from foods improves
thiamine status of the subjects (16,17).
Studies in rats have demonstrated that prolonged tea consumption, in contrast to
consumption of water, lowered (60–80%) blood transketolase, brain total thiamine, brain
transketolase activity, as well as ?-ketoglutarate and pyruvate dehydrogenase activities.
Thiamine deficiency caused by prolonged consumption of tea in rats was reversed by
discontinuing tea consumption; brain thiamine and brain and blood transketolase activities
were restored to normal levels within a week, whereas the time required for complete
restoration of pyruvate dehydrogenase and ?-ketoglutarate activities is about 2 weeks.
Tea-treated rats receiving intraperitoneal injection of 4 mg of thiamine 3 h prior to decapitation
had TPP-dependent enzymic activities exceeding the normal levels, which suggests
that the decrease in thiamine-dependent enzymic activities may be due not only to the lack
of TPP but also to the decreased synthesis of apoenzymes (18). Impairment of synthesis of
whole-brain acetylcholine was also observed in rats consuming tea (19).
4. Folate and Protein Status
Folate-deficient rats absorb thiamine less efficiently than pair-fed controls (20). These data
imply that folate may have a role in maintaining the integrity of the active transport process
of thiamine (13). Subjects with folate or protein deficiency show a significant reduction
in the maximum absorption of 35S-thiamine. The interaction of protein-energy malnutrition
and thiamine absorption is demonstrated by an increase in the maximum absorption of
35S-thiamine in malnourished alcoholics, with a decreased maximum absorption of thiamine
after correction of protein-energy malnutrition (21).
Ingested thiamine is fairly well absorbed, rapidly converted to phosphorylated forms,
stored poorly, and excreted in the urine in a variety of hydrolyzed and oxidized products
A. Absorption
The small intestine absorbs thiamine by two mechanisms. At concentrations exceeding 2
µM, thiamine is absorbed by passive diffusion; at concentrations below 2 µM, thiamine
is absorbed by an active process. The findings that the lack of Na or inhibition of ATPase
with ouabain blocks the uptake of thiamine at low concentrations by the intestinal cells
suggest the presence of a specific carrier. This is supported by the isolation of a thiaminebinding
protein (TBP) associated with thiamine transport into and out of the cell of Escherichia
coli. The 4-amino and the imidazole quaternary nitrogen are necessary for thiamine
uptake by the rat small intestine, whereas the 2-methyl and 5-hydroxyethyl appear to be
286 Tanphaichitr
necessary for thiamine binding to the carrier protein. Active thiamine absorption is greatest
in the jejunum and ileum (2,5,9,13,22,23).
A high percentage of the thiamine in the epithelial cells is phosphorylated, whereas
the thiamine arriving on the serosal side of the mucosa is largely free thiamine. Thus, the
entry of thiamine into the mucosal cells is linked with a carrier-mediated system that is
dependent either on thiamine phosphorylation-dephosphorylation coupling or on some
metabolic energetic mechanisms, possibly activated by Na. Thiamine exit from the mucosal
cell on the serosal side is dependent on Na and on the normal function of ATPase
at the serosal pole of the cell (2,5,9,13,23).
Investigations of the absorption of 35S-thiamine in humans indicate that its intestinal
transport is rate-limited and behaves according to Michaelis-Menten kinetics, which
yielded a Vmax of 8.3  2.4 mg and a Km of 12.0  2.4 mg (12). Earlier studies by Japanese
investigators have also shown that a single oral dose of thiamine greater than 6 mg does
not result in increased urinary thiamine excretion. The results imply that the excessive
amount of oral thiamine over 6 mg cannot be absorbed (22) and are consistent with the
subsequent studies in six healthy Australian volunteers receiving an oral dose of 10 mg
of thiamine. In the Australian study, the mean serum thiamine rose only marginally at 30
min from 5.1 to 5.9 µg/L and peaked at 7.2 µg/L—an increase of 42%. Six hours after
the test dose, the serum thiamine concentration had fallen back to its basal level (24). On
the contrary, healthy subjects taking each of the following physiological doses of thiamine
for a consecutive period of 5 days in the following order: 150, 450, 750, 1050, 1350,
1650, 1950, and 2250 µg had a highly significant correlation between the oral dose and
urinary excretion of thiamine (y  41.0239  0.0465, df  68, r  0.8618, P  0.001)
(25). Thus, all of the results indicate that passive absorption of thiamine is not significant
in humans.
B. Transport
Thiamine is carried by the portal blood to the liver. In normal adults, 20–30% of plasma
thiamine is protein-bound, all of which appears to be TPP (24). The transport of thiamine
into erythrocytes seems to be a facilitated diffusion process, whereas it enters other cells
by an active process (2,26). Erythrocytes contain mainly TPP (24).
Like the intestinal absorption of thiamine, thiamine transport across the blood–brain
barrier also involves two different mechanisms. However, the saturable mechanism at the
blood–brain barrier may be dependent on membrane-bound phosphatases. In this regard,
it differs from the energy-dependent processes described for the gut and cerebral cortex
cells. Studies in the rat reveal that a saturable mechanism with a mean Km of 2.2 nmol/
mL and a mean Vmax of 7.3 nmol/g per hour accounts for 95% of cerebellar and 91% of
cerebral cortex uptake at physiological plasma thiamine concentrations. TMP transport
rates are 5–10 times lower than those of thiamine. Thiamine uptake rates are 10 times
the maximal rate of thiamine loss from the brain (3).
A TBP has been identified in chicken egg (white and yolk) and liver (27) as well
as in the serum of pregnant rats (28). The molecular weight of highly purified TBP isolated
from egg white was 38,000. The TBP binds to 14C-thiamine with a molar ratio of 1 and
an association constant of 0.3 µM (27). Estrogen induces hepatic synthesis of TBP and
modulates its plasma level. The vital role of TBP in the transfer of thiamine across the
placenta is suggested by the observation that the passive immunization of pregnant rats
Thiamine 287
(4–16 days) with antibodies to chicken TBP but not ovalbumin resulted in fetal resorption
C. Tissue Distribution and Storage
The total amount of thiamine in a normal adult is approximately 30 mg. High concentrations
are found in skeletal muscles, heart, liver, kidneys, and brain. About 50% of the
total thiamine is distributed in the muscles. In spinal cord and brain, the thiamine level
is about double that of the peripheral nerves. Leukocytes have a 10-fold higher thiamine
concentration than erythrocytes. Thiamine has a relatively high turnover rate in the body.
The biological half-life of 14C-thiamine is 9–18 days. Besides, thiamine is not stored in
large amounts or for any period of time in any tissue (2,5,9,22).
D. Metabolic Modification
Of the total thiamine in the body, about 80% is TPP, 10% is TTP, and the remainder is
TMP and thiamine. The three tissue enzymes known to participate in formation of the
phosphate esters are thiamine pyrophosphokinase, which catalyzes the formation of TPP
from thiamine and ATP; TPP-ATP phosphoryltransferase, which catalyzes the formation
of TTP from TPP and ATP (Fig. 2); and thiamine pyrophosphatase, which hydrolyzes
TPP to form TMP (2,3,5,29).
At least 25–30 metabolites have been noted to occur in the urine of rats and men
given thiamine labeled in either the pyrimidine or thiazole moiety. Those metabolites that
have been identified are 2-methyl-4-amino-5-pyrimidine carboxylic acid (pyrimidine carboxylic
acid), 2-methyl-4-amino-5-hydroxymethylpyrimidine, 4-methylthiazole-5-acetic
acid (thiazole acetic acid), 3-(2?-methyl-4?-amino-5?-pyrimidylmethyl)-4-methylthiazole-
5-acetic acid (thiamine acetic acid), 2-methyl-4-amino-5-formylaminomethylpyrimidine,
and 5-(2-hydroxyethyl)-4-methylthiazole (thiazole). Of these, pyrimidine carboxylic acid,
thiazole acetic acid, and thiamine acetic acid are the major metabolites excreted in the urine
of rats and men. Rat liver alcohol dehydrogenase is involved in the in vivo metabolism of
thiamine and its thiazole moiety to their corresponding acids. In vitro experiments indicate
that although thiamine itself is a poor substrate for the enzyme, its thiazole moiety is
oxidized at a faster rate than is ethanol (30).
E. Excretion
Thiamine and its metabolites are mainly excreted in the urine. Very little thiamine is
excreted in the bile. Early milk has low thiamine levels. Thiamine administered by oral
or parenteral route is rapidly converted to TPP and TTP in the tissues. Thiamine in excess
of tissue needs, as well as binding and storage capacity, is rapidly excreted in the urine
in the free form (2,4,5,24,25).
Functions of thiamine can be categorized into established and plausible functions. The
established functions of thiamine are biochemical functions in which TPP serves as the
coenzyme of biochemical reactions, whereas the plausible functions are neurophysiological
288 Tanphaichitr
A. Biochemical Functions
In mammalian systems, TPP functions as the Mg2-coordinated coenzyme for the active
aldehyde transfers, which include the oxidative decarboxylation of ?-keto acids and transketolase
reaction (2,4,5,31). The key feature of TPP is that the carbon atom between the
nitrogen and sulfur atoms in the thiazole ring is much more acidic than most CCHE
groups. It ionizes to form a carbanion, which readily adds to the carbonyl group of ?-
keto acids or ketose. The positively charged ring nitrogen of TPP then acts as an electron
sink to stabilize the formation of a negative charge, which is necessary for decarboxylation.
Protonation then gives hydroxyethyl TPP (5).
1. Oxidative Decarboxylation of ?-Keto Acids
Pyruvic, ?-ketoglutaric, and branched-chain ?-keto acids undergo oxidative decarboxylation.
a. Oxidative Decarboxylation of Pyruvic Acid
The net reaction of oxidative decarboxylation of pyruvate catalyzed by the pyruvate dehydrogenase
complex (PDHC) is
Pyruvate  CoA  NAD > acetyl CoA  CO2  NADH  H
In addition to the stoichiometric coenzymes consisting of CoA and NAD, TPP, lipoic
acid, and flavin adenine dinucleotide (FAD) also serve as the coenzymes. The PDHC is
localized in the mitochondrial inner membrane and is an organized assembly of three
enzymes. The conversion of pyruvate to acetyl CoA consists of four steps (2,4,5,9,31).
First, pyruvate is decarboxylated after it combines with TPP is catalyzed by pyruvate
dehydrogenase. Second, the hydroxyethyl group attached to TPP is oxidized by the disul-
fide groups of lipoamide to form an acetyl group and concomitantly transferred to lipoamide
to yield acetyl lipoamide. This step is catalyzed by the lipoic acid-bound dihydrolipoyltransacetylase.
Third, the acetyl group is transferred from acetyl lipoamide to CoA
to form acetyl CoA catalyzed by dihydrolipoyl transacetylase. Fourth, the regeneration of
the oxidized form of lipoamide is catalyzed by the FAD-dependent dihydrolipoyl dehydrogenase;
a hydride ion is transferred to an FAD prosthetic group of the enzyme and then
to NAD (Fig. 7) (31,32).
b. Oxidation Decarboxylation of ?-Ketoglutaric Acid
The net reaction of oxidative decarboxylation of ?-ketoglutarate, taking place in the tricarboxylic
acid (TCA) cycle and catalyzed by the ?-ketoglutarate dehydrogenase complex
(?-KGDHC), is as follows:
?-Ketoglutarate  CoA  NAD > succinyl CoA  CO2  NADH  H
The coenzyme requirements and the steps in formation of succinyl CoA are analogous to
the oxidative decarboxylation of pyruvate (2,4,5,9).
c. Oxidative Decarboxylation of Branched-Chain ?-Keto Acids
The oxidative decarboxylation of the three branched-chain ?-keto acids, i.e., ?-ketoisocaproate,
?-keto-?-methylvalerate, and ?-ketoisovalerate, to yield isovaleryl CoA, ?-
methylbutyryl CoA, and isobutyryl CoA, respectively, is catalyzed by the branched-chain
?-keto acid dehydrogenase complex (BC ?-KADHC), which is analogous to complexes
of pyruvate and ?-ketoglutarate (2,4,5,9).
Thiamine 289
Fig. 7 The pyruvate dehydrogenase complex reactions. TPP, thiamine pyrophosphate; Lip, lipoic
The aforementioned three ?-keto acid multienzyme complexes play the significant
role in energy generating pathways (Fig. 8). The PDHC irreversibly commits pyruvic acid,
a three-carbon intermediate derived from glucose catabolism, to convert to acetyl CoA,
which participates in three important metabolic fates: complete oxidation of the acetyl
group in the TCA cycle for energy generation; conversion of an excess acetyl CoA into
the ketone bodies; and transfer of the acetyl units to the cytosol with subsequent biosynthesis
of sterols and long-chain fatty acids. Because PDHC occurs at a significant branch
point in the metabolic pathways, its activity is strictly regulated by two separate types of
mechanisms. The first mechanism is competitive end-product inhibition of catalytic PDHC
by acetyl CoA and NADH. The second mechanism involves covalent modification of the
PDHC by phosphorylation/dephosphorylation (inactive PDHC/active PDHC) mechanism,
mediated by a specific protein kinase that is tightly bound to the PDHC and by a specific
phosphoprotein phosphatase that is much less tightly associated with the PDHC (31,32).
The pyruvate dehydrogenase kinase/phosphatase system is regulated by a number
of factors. Inactivation of PDHC is accomplished by an Mg2-ATP-dependent pyruvate
dehydrogenase kinase. Pyruvate dehydrogenase kinase is inhibited by ADP, pyruvate,
CoA, NAD, TPP, and calcium, whereas it is stimulated by NADH and acetyl CoA; TPP
inhibits phosphorylation by binding at the catalytic site of pyruvate dehydrogenase to
promote a conformational change, which in turn causes one of the serine hydroxy groups
on the subunit of pyruvate dehydrogenase to become less accessible to pyruvate dehydrogenase
kinase. The pyruvate dehydrogenase phosphatase reaction requires Mg2. It is stimulated
by Ca2 and inhibited by NADH, and the inhibition is reversed by NAD (31).
The primary metabolic fate of acetyl CoA in most cells is its complete oxidation
in the TCA cycle with the generation of 2 CO2, 2 GTP, and 4 reducing equivalents consisting
of 3 NADH and 1 FADH2, which generate 11 ATP by the electron transport system.
Thus, the integrity of the TCA cycle is critical for the provision of cellular energy. ?-
KGDHC is the only enzyme in the TCA cycle that requires TPP as coenzyme. The equilibrium
of the ?-KGDHC reaction lies strongly toward succinyl CoA formation. In this reaction,
the second molecule of CO2, the second reducing equivalent (NADH  H) of the
290 Tanphaichitr
Fig. 8 Pathways dependent on thiamine pyrophosphate. P, phosphate; TPP, thiamine pyrophosphate;
PDHC, pyruvate dehydrogenase complex; ?-KGDHC, ?-ketoglutarate dehydrogenase complex;
BC ?-KADHC, branched-chain ?-keto acid dehydrogenase complex; TK, transketolase.
TCA cycle, and succinyl CoA, an energy-rich thioester compound, are produced. Succinyl
CoA is conserved in a substrate level phosphorylation reaction in the next step of the
TCA cycle. It is at the level of this reaction where ?-ketoglutarate may leave the TCA
cycle to be converted to glutamate by glutamate dehydrogenase, a mitochondrial enzyme,
in the presence of NADH or NADP and ammonia. Unlike PDHC, the ?-KGDHC is not
regulated by a protein kinase. ATP, GTP, NADH, and succinyl CoA inhibits the ?-
KGDHC activity, whereas Ca2 stimulates its activity (32).
Where the BCAAs valine, leucine, and isoleucine are present in excess over what
is needed for protein synthesis, they are degraded through several steps to provide energy.
The initial step in catabolism is reversible transamination with ?-ketoglutarate to form
the corresponding ?-keto acids, i.e., ?-ketoisovalerate, ?-ketoisocaproate, and ?-keto-?-
Thiamine 291
methylvalerate, respectively. The second step is irreversible oxidative decarboxylation of
the three branched-chain ?-keto acids by BC ?-KADHC to yield CO2, NADH that feeds
electrons to the electron transport system, and isobutyryl CoA, isovaleryl CoA, or ?-
methylbutyryl CoA. These three analogs of fatty acyl CoA are oxidized by specific dehydrogenases
to form the corresponding ?,?-unsaturated compounds. The remainder of the
isoleucine degradation pathway is identical to that of fatty acid oxidation, with the provision
of acetyl CoA and propionyl CoA, whereas that of the valine degradation provides
propionyl CoA. The remainder of the leucine degradation is a combination of reactions
used in ? oxidation and ketone body synthesis with the provision of acetyl CoA and
acetoacetate, which may be converted to two acetyl CoA. The acetyl CoA formed from
the degradation of the BCAAs enters the acetyl CoA pool, which can be used for any of the
aforementioned metabolic functions. Propionyl CoA formed from isoleucine and valine for
the most part is not further oxidized as a fatty acid but is utilized for the formation of
succinyl CoA, which is further metabolized via the TCA cycle. Thus, the three carbon
atoms derived from valine and isoleucine may be completely oxidized to CO2, but they
can be incorporated into carbohydrate by oxidation of succinate to oxaloactate and the
reactions of gluconeogenesis. On the contrary, both acetyl CoA and acetoacetate of leucine
degradation are characteristic of fatty acid oxidation (33).
2. Transketolase Reaction
A TPP-dependent transketolase found in the cytosol catalyzes the reversible transfer of a
glycolaldehyde moiety from the first two carbons of a donor ketose phosphate to the
aldehyde carbon of an aldose phosphate in the pentose phosphate pathway. These reactions
are as follows:
Xylulose-5-phosphate  ribose-5-phosphate s glyceraldehyde-3-phosphate
Xylulose-5-phosphate  erythrose-4-phosphate s glyceraldehyde-3-phosphate
The metabolic significance of the pentose phosphate pathway is not to obtain energy
from the oxidation of glucose in animal tissues. Its primary purpose is to generate NADPH,
which serves as a hydrogen and electron donor in reductive biosynthetic reactions, including
the biosynthesis of fatty acids; to convert to pentoses, particularly ribose-5-phosphate,
which are components of RNA, DNA, ATP, CoA, NAD, and FAD; and to catalyze the
interconversion of C3, C4, C6, and C7 sugars, some of which can enter the glycolytic
sequence (2).
B. Neurophysiological Functions
There is increasing evidence of the roles of thiamine in neurotransmitter function and
nerve conduction (3,34).
1. Neurotransmitter Function
Though the findings are inconsistent, abnormal metabolism of four types of neurotransmitters
(acetylcholine, catecholamine, serotonin, and amino acids) has been reported in thiamine
deficiency. Acetylcholine, ?-aminobutyric acid (GABA), glutamate, and aspartate
are produced primarily through the oxidative metabolism of glucose (3,34).
292 Tanphaichitr
a. Acetylcholine
Studies in thiamine-deficient rats have revealed that there is no change in regional brain
acetylcholine levels, but there are reductions in acetylcholine turnover and acetylcholine
utilization in the cortex, midbrain, diencephalon, and brain stem. These results suggest
the depression of central cholinergic mechanisms in thiamine-deficient rats. Since choline
acetyltransferase is unaltered in thiamine deficiency, decreased PDHC activity limiting
acetyl CoA production may underlie these changes (3).
b. Catecholamines
Decreased synthesis of catecholamines was demonstrated in severely thiamine-deficient
rat brain; treatment with thiamine normalized catecholamine synthesis within 2 h, although
some neurological signs persist (34). There is also evidence indicating that behavioral
deficits in rats persisting after reversal of thiamine deficiency are linked to significant
reductions in the norepinephrine content of cortex, hippocampus, and olfactory bulbs (3).
c. Serotonin
Decreased uptake of serotonin by cerebellar synaptosomes and significant increase in 5-
hydroxyindoleacetic acid, a serotonin catabolite, occurs in thiamine-deficient rats. These
findings are supported by autoradiographic evidence of decreased serotonin uptake in
indoleaminergic afferents of cerebellum in thiamine-deficient rats (3,34).
d. Amino Acids
The levels of four amino acids with putative neurotransmitter functions, i.e., glutamate,
aspartate, GABA, and glutamine, are decreased in thiamine-deficient rat brain, with
marked changes in the cerebellum (3).
2. Nerve Conduction
That thiamine plays a special role in neurophysiology that is independent of its coenzyme
function was originally postulated by von Muralt (35). The effect of thiamine antagonists
on conduction in peripheral nerves, the localization of thiamine and thiamine pyrophosphatase
in peripheral nerve membranes as opposed to axoplasm, the release of thiamine from
nerve preparations by electrical stimulation, the association of released thiamine with
membrane fragments of brain, spinal cord, and sciatic nerves all support the aforementioned
view (2–5,9,35). This thiamine release appears to be due to hydrolysis of TPP and
TTP. Tetrodotoxin, local anesthetics, acetylcholine, and serotonin have been shown to
cause release of thiamine from nerve and nerve membrane segments (36–38). The finding
that thiamine-phosphorylated derivatives are associated with the sodium channel protein
has led to the hypothesis that TTP plays a fundamental role in the control of sodium
conductance at axonal membranes (39).
Cardiac failure, muscle weakness, peripheral and central neuropathy, and gastrointestinal
malfunction have been observed in both animals and humans on diets restricted in thiamine
Although the precise biochemical defect responsible for the pathophysiological manifestations
of thiamine deficiency is not established, experimental evidence indicates that
thiamine has three major roles at the cellular level. The first relates to energy metabolism
Thiamine 293
and concerns the oxidative decarboxylation of ?-keto acids, the inhibition of which leads
to a failure of ATP synthesis. The second concerns synthetic mechanisms as reflected by
the transketolase reaction yielding NADPH and pentose. The third deals with the functions
of neurotransmitters and nerve conduction. In Wernicke’s disease, failure of energy metabolism
predominantly affects neurons and their functions in selected areas of the central
nervous system. Glial changes may be caused by biochemical lesions that affect transketolase
and nucleic acid metabolism. Membranous structures are altered in a visible manner
and secondary demyelination ensues (9).
Thiamine deficiency in free-living populations and hospitalized patients can be caused by
inadequate intake, decreased absorption, and defective transport of thiamine, impaired
biosynthesis of TPP, increased requirement, and increased loss of thiamine. These mechanisms
may reinforce each other.
A. Inadequate Intake of Thiamine
1. Inadequate Intake of Thiamine with Concomitant High Intake
of Carbohydrate
All of the published reports indicate that beriberi, the clinical condition of thiamine defi-
ciency, in Asian populations is always associated with high intake of carbohydrate and
insufficient intake of thiamine (5,9,10,30,40–42).
Recent studies in Thailand (10,43–46) and Japan (47–49) have confirmed that inadequate
thiamine intake accompanied by high intake of carbohydrate derived from milled
rice and refined carbohydrate is still the major cause of thiamine deficiency in Asian
populations. For instance, dietary assessment in a village in northeastern Thailand has
repeatedly shown that the villagers’ protein and energy intakes are mainly supplied from
a rice-based diet; 78% of their energy intake is carbohydrate, mainly derived from milled
rice. Milled rice is a poor source of thiamine (Table 1), and about 85% of its thiamine
content is lost when the water in which it has been soaked is discarded (10,44,50,51).
Thus, their thiamine intake is low, e.g., 0.23 mg/1000 kcal (44). These dietary data are
consistent with the high prevalence of biochemical thiamine depletion in that population
(10,43,44). The adverse effects of the ATFs on thiamine status caused by consumption
of raw fermented fish and chewing betel nuts have also been demonstrated in northeastern
Thais, whereas inadequate thiamine status is also detected in northern Thais who regularly
chew fermented tea leaves (16).
The reappearance of beriberi in Japan in the 1970s is related to high carbohydrate
intake derived from milled rice, instant noodles, and sweet carbonated drinks, coupled
with low thiamine intake (47–49).
2. Infants Breast-Fed by Thiamine-Deficient Mothers
The mean ( SEM) concentrations of thiamine in breast milk from mothers residing in
rural areas of central, northern, and northeastern Thailand were 93  9.9, 102  6.5, and
119  6.3 µg/L. These figures are lower than the concentration of thiamine in breast milk
of 142  1.4 µg/L from the North American mothers. The thiamine concentration in
mothers with complaints suggestive of thiamine deficiency has been reported to be 87 
294 Tanphaichitr
4.8 µg/L (52,53). These findings partly explain the prevalences of infantile beriberi in
Thailand (10,54–58).
3. Chronic Alcoholism
The major cause of thiamine deficiency in Western populations is excessive intake of
alcohol (ethanol). Chronic alcoholics may eat nothing for days, and when they eat their
diet is often high in carbohydrate and low in thiamine (59). Excessive ethanol ingestion
creates a metabolic burden in the human body. The liver is primarily responsible for the
first two steps of ethanol metabolism:
Ethanol Acetaldehyde
Acetaldehyde Acetate
The first step, catalyzed by alcohol dehydrogenase, generates acetaldehyde and NADH,
and occurs in the cytosol. The second step, catalyzed by aldehyde dehydrogenase, generates
acetate and NADH but occurs in the mitochondrial matrix space. The NADH generated
by the second step can be used directly by the mitochondrial electron transfer chain,
whereas the NADH generated by the first step cannot be used directly and must be oxidized
back to NAD by the glycerol phosphate and/or malate-aspartate shuttles. Since the intake
of ethanol demands the hepatic disposal of the NADH, intake of even moderate amounts
of ethanol results in inhibition of important processes requiring NAD, including gluconeogenesis
and fatty acid oxidation (60).
Hepatic mitochondria have a limited capacity to oxidize acetate to CO2 because the
activation of acetate to acetyl CoA requires GTP, which is the product of the succinyl
CoA synthetase reaction. The TCA cycle, and therefore GTP synthesis, is inhibited by
NADH levels during ethanol oxidation. Much of the hepatic acetate generated from ethanol
escapes the blood to other cells containing mitochondria, which oxidize it to CO2 and
H2O through the TCA cycle (Fig. 8). The heart is the most active organ for the disposal
of acetate (60). This process cannot be achieved with an inadequate supply of thiamine.
Acetaldehyde can also escape from the liver. Though its concentration in the blood
is lower than that of acetate, acetaldehyde is a reactive compound that readily forms covalent
bonds with functional groups of biologically important compounds (60).
Not only insufficient thiamine intake but also other factors are involved in the development
of thiamine deficiency in chronic alcoholics. These include ethanol inhibition of
thiamine absorption, decrease in thiamine absorption secondary to protein or folate defi-
ciency, impaired conversion of thiamine to TPP in the presence of alcoholic liver injury,
and reduced hepatic storage of thiamine (2,12,13,21,26,30,61).
4. Patients on Parenteral Nutrition
Inadequate intake of nutrients, including thiamine, is common in hospitalized patients.
The cause of their inadequate thiamine intake can be either primary or secondary. Primary
inadequate intake of thiamine is caused by insufficient supply of food quantitatively and
qualitatively. In secondary inadequate intake of thiamine, illness or treatment affects the
thiamine status by altering a person’s eating behavior or impairing his or her ingestion.
Primary and secondary mechanisms frequently reinforce each other (62). In a dietary assessment
in 656 hospitalized patients over a period of 2 years, 57 of the patients had a
Thiamine 295
low daily intake of thiamine (63), and in 19% the intake was less than half the recommended
daily allowances (RDAs) for thiamine (64). This result warrants the supplementation
of thiamine in hospitalized patients to ensure their adequate thiamine status. Special
attention should be paid to patients on parenteral nutrition in which glucose is the major
source of energy supply because inadequate or absent thiamine creates beriberi (65–73).
B. Decreased Absorption of Thiamine
The ethanol inhibition of thiamine absorption and the defective absorption of thiamine in
folate deficiency states are well documented (2,12,13,21,26,30). Patients with intestinal
resection are also prone to develop thiamine deficiency because jejunum and ileum are
the sites of thiamine absorption (12,74).
C. Defective Transport of Thiamine
Thus far, there is no acquired condition that is known to cause defective transport of
thiamine. However, there is a concern about the adequacy of the transport of thiamine
from mothers across the placenta to the fetus. Measurement of erythrocyte or whole-blood
transketolase activity (ETKA or WBTKA) and the enhancement from added TPP to basal
assays (referred to as the TPP effect, TPPE) is used for thiamine assessment. Using this
approach to assessment, thiamine status in cord blood and in mothers within 24 h of
delivery (75) or 6–12 h before delivery (76) indicates that newborns have better thiamine
status than their mothers. These results are consistent with the findings of previous studies
that (a) cord blood has a higher thiamine content than maternal blood (77) and (b) fetal
thiamine status may be normal even when the mother is thiamine-deficient (78). These
findings imply that there is an efficient transport of thiamine across the placental barrier
and that the fetus can sequester thiamine.
D. Impaired Biosynthesis of TPP
The conversion of thiamine to TPP may be impaired in the presence of alcoholic liver
injury, possibly because of decreased availability or use of ATP, and may be associated
with decreased activities of pyruvate dehydrogenase and transketolase due to apoenzyme
or magnesium deficiency (12,13). Besides, the decreases in hepatic and cerebral transketolase
activities are accompanied by a parallel fall in thiamine concentrations in liver and
brain (79). Since the transketolase activity correlates with TPP concentrations (80), these
findings suggest a lowering in the TPP concentrations. However, thiamine deficiency in
patients with fulminant hepatic failure can be corrected by intravenous administration of
100 mg of thiamine hydrochloride twice daily. This indicates that the bioconversion of
thiamine to TPP is still possible in the presence of acute hepatocellular necrosis when the
patient is treated with pharmacological doses of thiamine (81).
E. Increased Requirement of Thiamine
Increased anabolism occurs in various physiological conditions, including pregnancy, lactation,
infancy and childhood, adolescence, and increased physical activity, whereas a
number of pathological conditions create increased catabolism. Increased anabolism and
catabolism raise the requirements of several nutrients, including thiamine.
296 Tanphaichitr
1. Pregnancy
Studies of urinary excretion of thiamine, blood thiamine levels, and ETKA all indicate
that the requirement for thiamine in women increases during pregnancy. This increase
appears to occur early in pregnancy and to remain constant throughout (64,75,76,82).
Thus women with marginal thiamine status who are pregnant and pregnant women with
inadequate intake of thiamine are vulnerable to develop beriberi. This is well illustrated
in the report on women with hyperemesis gravidarum presenting with Wernicke’s encephalopathy
having the symptom of vomiting for a month, and Wernicke’s encephalopathy
usually occurred following thiamine-free intravenous glucose infusion (83,84).
2. Lactation
Thiamine requirement also increases during lactation due to the loss of thiamine in milk
and increased energy consumption during lactation. The healthy lactating woman secretes
approximately 0.2 mg of thiamine in her milk each day (85).
3. Infancy and Childhood
Because the metabolic rate of infants and children is greater and the turnover of nutrients
more rapid than in adults, the unique nutritional needs for growth and development are
superimposed on higher maintenance requirements than those of adults. Whole-blood total
thiamine concentrations are higher in infants aged 0–3 months than those aged 3–12
months, i.e., 258  63 (mean  SD) and 214  44 nmol/L, respectively. After a large
decline in the first year of life, the concentrations become stable and equal to those in
adults: 187 39 nmol/L. The overall decrease is due mainly to a decrease in the phosphorylated
thiamine. Since these changes occur relatively independently of the changes in
hematocrit characteristically seen in infancy, it may be related to changes in metabolic
activity, growth rate, or other variables of biological activity (86). The findings that
changes in whole-blood thiamine concentrations during infancy and childhood are physiological
are supported by nonsignificant differences in WBTKA and TPPE between healthy
infants and children aged 1–6 years (55). However, inadequate thiamine status is commonly
found in breast-fed infants by thiamine-deficient mothers and children on the ricebased
diets (10,54–58).
4. Adolescence
The nutritional requirements of adolescents are influenced primarily by the normal events
of puberty and the simultaneous spurt of growth. Adolescence is the only time in extrauterine
life when growth velocity increases and exerts a major influence on nutritional requirements.
Since thiamine is involved in energy metabolism, adolescents on milled rice as
their staple diets are vulnerable to develop thiamine deficiency (10,45,46,48,87,88).
5. Increased Physical Activity
For most people, the second largest component of total energy expenditure is the energy
expended in physical activity. Thus, high physical activity can precipitate the development
of beriberi in those with marginal thiamine status (10,88).
6. Increased Catabolism
Increased catabolism raises the requirements for energy and thiamine. Thus, catabolic
conditions, such as fever, infection, trauma, or surgery, can precipitate the development
of beriberi in those with marginal thiamine status (10,89–91).
Thiamine 297
F. Increased Loss of Thiamine
Renal patients on chronic dialysis (92,93) and patients with congestive heart failure treated
with furosemide, a diuretic (94,95), may develop beriberi, which in part is due to an
increased renal excretion of thiamine.
The clinical manifestations of beriberi vary with age. Infantile beriberi is commonly found
in infants aged 2–3 months who are being breast-fed by thiamine-deficient mothers. The
clinical manifestations of beriberi in older children and adolescents are similar to those
in adults, and as such are classified as adult beriberi (5,9,10,30,41,96–99).
A. Infantile Beriberi
The predominant symptoms of infantile beriberi include edema, dyspnea, oliguria, aphonia,
cardiovascular disturbances, and gastrointestinal disorders, including loss of appetite,
vomiting, or milk diarrhea. Infantile beriberi may be categorized into four subtypes based
on the dominant features, i.e., cardiac, aphonic, pseudomeningitic, or a combination
1. Cardiac or Acute Fulminating Form
Infants with cardiac beriberi usually present with acute onset of manifestations, including
a loud piercing cry, cyanosis, dyspnea, vomiting, tachycardia, and cardiomegaly; death
may occur within a few hours after onset unless thiamine is administered.
2. Aphonic Form
Aphonic forms of infantile beriberi are generally less acute. The striking feature is the
tone of the child’s cry, which varies from hoarseness to complete aphonia due to paralysis
of the recurrent laryngeal nerve. A decrease in deep tendon reflexes and blepharoptosis
are detected in some infants.
3. Pseudomeningitic Form
Infants exhibit vomiting, nystagmus, purposeless movement of the extremities, and convulsion.
However, no abnormality of cerebrospinal fluid (CSF) is demonstrated.
4. Mixed Form
Infants with beriberi may present with the combined manifestations of cardiac disturbance,
aphonia, and pseudomeningitis.
In a study of 16 infants with beriberi, the prevalence of cardiovascular system disorders
was 87% with dyspnea, 81% with tachycardia (heart rate more than 130 beats/min),
73% with cardiomegaly (confirmed by chest x-ray), 63% with vomiting, 62% with hepatomegaly,
56% with edema, 19% with cyanosis, and 6% with shock, whereas those of nervous
system disorders included 81% with decreased deep tendon reflexes, 50% with aphonia,
12% with convulsion, and 6% with blepharoptosis (10,56).
B. Adult Beriberi
Older children, adolescents, and adults may present with dry (paralytic or nervous), wet
(cardiac), or cerebral beriberi (5,9,10,30,41,96–99).
298 Tanphaichitr
1. Dry Beriberi
The predominant feature in dry beriberi is peripheral neuropathy, characterized by a symmetrical
impairment of sensory, motor, and reflex functions that affects the distal segments
of limbs more severely than the proximal ones, calf muscle tenderness, and difficulty in
rising from a squatting position (5,10,46,99).
Pathological changes in the peripheral nerve in beriberi neuropathy are unusual axonal
degeneration and accumulation of flattened sacs or tubules in the axoplasm of large
myelinated fibers in untreated patients. In beriberi patients treated with thiamine, there
are numerous Bungner’s bands and clusters of thin myelinated fibers, indicating active
regeneration, which correspond well with clinical improvement (47). Examination of CSF
in patients with dry beriberi reveals normal findings (47,99).
2. Wet Beriberi
In addition to peripheral neuropathy, common signs found in wet beriberi include edema,
tachycardia, wide pulse pressure, cardiomegaly, and congestive heart failure (5,10,46,99).
In some patients, there is a sudden onset of a cardiac manifestation known as acute fulminant
or ‘‘shoshin’’ beriberi, with the predominant features of tachycardia, dyspnea, cyanosis,
cardiac enlargement, and circulatory collapse (41).
a. Electrocardiographic Findings
Considerable attention has been given to the electrocardiographic changes that occur in
wet beriberi (99,100). The serial electrocardiographic study in 11 patients with wet beriberi
whose thiamine deficiency was confirmed by ETKA and TPPE revealed the following
findings: 8 patients had a prolonged QT interval, which became normal in 5 patients within
2 h to 28 days after receiving the thiamine treatment. Seven patients showed T-wave
abnormalities in the precordial leads, consisting of flat or inverted T waves that became
normal within 7–40 days following administration of thiamine. Three patients showed
relatively low QRS complex voltage, which increased after the appearance of diuresis
following the thiamine treatment. The increase of QRS complex voltage and the rapid
reduction of heart size were observed concurrently with clinical improvement. Since the
pattern of the electrocardiographic voltage in wet beriberi resembles that in pericardial
effusion, the transient low QRS voltage in patients with wet beriberi is most likely caused
by pericardial effusion (99,101,102).
b. Hemodynamic Findings
Studies in occidental patients with wet beriberi have shown hemodynamic findings of high
cardiac output with low peripheral vascular resistance (103,104); an acute response to
thiamine administration, i.e., decrease in cardiac output with concomitant increase in total
peripheral resistance, was observed within 37 min (104) and 2 h (103).
A hemodynamic study of six Thai nonalcoholic patients with wet beriberi whose
diagnosis was confirmed by ETKA and TPPE revealed two types of hemodynamic findings.
The first type is consistent with occidental reports (103–105); it consisted of four
patients showing high cardiac output and low peripheral and pulmonary vascular resistance.
Within 15 min of intravenous administration of thiamine HCl, their cardiac output
was reduced and vascular resistance was increased, as was blood pressure (99,106,107).
It appears that the major cause of the high cardiac output is vasomotor depression, the
precise mechanism of which is not understood, but which leads to a reduced systemic
vascular resistance (108).
Thiamine 299
The second type of hemodynamic finding involved one patient with normal cardiac
output and another with low cardiac output. The low cardiac output may be due to myocardial
damage sustained by repeated episodes of thiamine deficiency; the increase in cardiac
output after thiamine treatment suggests the improvement of myocardial metabolism by
the pharmacological dose of thiamine (99,106,107). Thus, cardiac output in wet beriberi
may be variable. Normal or low cardiac output does not exclude the diagnosis.
3. Wernicke–Korsakoff Syndrome
The specific factor responsible for most, if not all, of the clinical manifestations of the
Wernicke–Korsakoff syndrome is a deficiency of thiamine. Wernicke–Korsakoff syndrome
constitutes the most common alcoholic-nutritional affliction of the central nervous
system (3,24,108–110). Although alcoholism is the major cause of thiamine deficiency
in Wernicke–Korsakoff syndrome, iatrogenic causes, including parenteral glucose administration
and chronic dialysis, can aggravate the syndrome in patients with marginal thiamine
status (65–68,83,84).
a. Wernicke’s Disease or Encephalopathy
Wernicke originally described an illness of acute onset characterized by mental disturbance,
paralysis of eye movements, and ataxia of gait. Since Wernicke’s time, views regarding
Wernicke’s disease have undergone considerable modification. The diagnosis is
based on the triad of ocular motor signs, ataxia, and derangement of motor functions
The ocular motor signs are the most readily recognized abnormalities. They include
a paresis or paralysis of abduction accompanied by horizontal diplopia, strabismus, and
nystagmus. In advanced cases, there may be complete loss of ocular movement, and the
pupils may become miotic and nonreacting. Blepharoptosis is rare. The ocular motor signs
are attributable to lesions in the brain stem affecting the abducens nuclei and eye movement
centers in the pons and rostral midbrain.
The ataxia affects stance and gait. Persistent ataxia is related to the loss of neurons
in the superior vermis of the cerebellum; extension of the lesion into the anterior parts of
the anterior lobes accounts for the ataxia of individual movements of the legs.
A derangement of mental function is found in about 90% of patients and takes one
of the following three forms. First, a global confusional-apathetic state is most common,
characterized by profound listlessness, inattentiveness, indifference to the surroundings,
and disorientation. Second, a disproportional disorder of retentive memory, i.e., Korsakoff’s
amnesic state, occurs in some patients. Third, the symptoms of alcohol withdrawal
may be found in a relatively small number of patients. The amnesic defect is related to
lesions in the diencephalon, specifically those in the medial dorsal nuclei of the thalamus.
Most patients with Wernicke’s disease display peripheral neuropathy and some cardiovascular
abnormalities, including tachycardia, exertional dyspnea, and postural hypotension
b. Korsakoff’s Psychosis
Victor (109) has stated that Wernicke’s disease and Korsakoff’s psychosis are not separate
diseases; Korsakoff’s psychosis is the psychic component of Wernicke’s disease. Thus,
the clinical manifestations should be called Wernicke’s disease when the amnesic state
is not evident and the Wernicke–Korsakoff syndrome when both the ocular-ataxic and
amnesic manifestations are present.
300 Tanphaichitr
Though the specific role of thiamine in Wernicke–Korsakoff syndrome is established
and most patients with Wernicke–Korsakoff syndrome are alcoholics, only a few
alcoholics are affected (3,109–111). Genetic abnormalities in ETK may underlie a predisposition
to Wernicke–Korsakoff syndrome (112). However, the investigation on the hysteretic
properties of human transketolase with emphasis on its dependency on TPP concentration
has revealed the substantial lag in formation of active holotransketolase and
interindividual differences and cell type variation from the same individual in the lag
period; these findings suggest the existence of mechanisms for the loss of transketolase
activity during thiamine deficiency and may explain, at least in part, the differential sensitivity
to deficiency demonstrated by tissues and individuals (113).
A. Treatment
Thiamine should be promptly administered to beriberi patients. The dosage usually ranges
from 50 to 100 mg of thiamine HCl given intravenously or intramuscularly for 7–14 days,
after which 10 mg/day can be administered orally until the patient recovers fully. To
prevent the recurrence of beriberi, patients should be advised to change their dietary habits
and to stop their alcohol drinking (5,10,30,46,56).
The rationale to administer a large dose of thiamine is to replenish the thiamine
store, which is consistent with the positive correlation between thiamine concentration in
serum and CSF (114), to stimulate the TPP-dependent reactions optimally, which is consistent
with the aforementioned thiamine-dependent hysteretic behavior of human transketolase
(113), and to improve cardiovascular disorders evidenced by the hemodynamic
studies (99,103–107). The parenteral route is used initially to ensure the bioavailability
of thiamine.
Several thiamine derivatives, especially thiamine propyl disulfide (TPD) and tetrahydrofurfuryldisulfide
(TTFD) (Fig. 4), are useful for oral administration to beriberi patients
because even by the oral route they produce a significantly higher thiamine level in the
blood, erythrocytes, and CSF than thiamine HCl and TPP at the same dose of 50 mg, and
their effectiveness is equivalent to that produced by parenteral administration of thiamine
HCl or TPP in healthy subjects (115). The effective intestinal absorption of TTFD is also
demonstrated at 100- and 180-mg dose levels in healthy subjects (116). The rapid absorption
and transfer to the tissue of TPD is also evident by the increased levels of thiamine
in portal vein, hepatic vein, and femoral artery (117). Oral administration of 50 mg of
TPD to alcoholics with thiamine deficiency also restored their ETKA to normal and eliminated
lateral rectus palsy in patients with Wernicke’s disease (117).
The bioavailability of TTFD in multivitamin tablets has also been shown in healthy
subjects taking the vitamin preparation containing 10 mg of TTFD daily for 7 days and
10 or 20 mg of TTFD daily for 44 weeks. Their baseline blood thiamine concentrations
increased gradually during the long-term administration, with a dose-dependent response
(118). The dramatic response of patients with cardiac beriberi to the daily intravenous
administration of 50 mg of TTFD has also been reported (119).
B. Response to Thiamine Administration
In wet beriberi, within 6–24 h of thiamine administration improvement can be observed
in terms of less restlessness; disappearance of cyanosis; reductions in heart rate, respiratory
Thiamine 301
rate, and cardiac size; and clearing of pulmonary congestion. Dramatic improvement after
thiamine treatment is also observed in infants with cardiac beriberi (10,46,55,56,99).
It is difficult to use the response to thiamine administration as a criterion for immediate
diagnosis in dry beriberi or infants with aphonia because more time elapses before
improvement is observed. In a study of 21 patients with beriberi, impaired sensation disappeared
after 7–120 days of thiamine treatment; motor weakness found in 12 patients was
recovered within 60 days; but absent or hypoactive ankle and knee jerks remained for
several months after treatment (10,46,99).
The response of patients with Wernicke’s disease to thiamine treatment shows a
characteristic pattern. Occular palsies may begin to improve within hours to several days
after the administration of thiamine. Sixth-nerve palsies, ptosis, and vertical gaze palsies
recover completely within 1–2 weeks in most cases, but vertical gaze–evoked nystagmus
may persist for months. Ataxia improves slowly and approximately half of patients recover
incompletely. Apathy, drowsiness, and confusion also recede gradually (109).
The outcome of the Korsakoff amnesic state varies. It ranges from complete or
almost complete recovery in less than 20% of patients to slow and incomplete recovery
in the remainder. The residual state is characterized by large gaps in memory, usually
without confabulation, and an inability of the patient to sort out events in the proper
temporal sequence (109).
The sequential development of impaired thiamine status can be divided into four stages.
It begins with an inadequate thiamine intake. The resulting depletion of thiamine in tissues
leads to metabolic derangement of various organs, and clinical manifestations of beriberi
appear (Table 2). Thus, the subject’s history, laboratory tests, and physical examination
are the bases for evaluation of thiamine status in humans (62).
A. Subject’s History
In addition to dietary assessment, demographic data, medical history, family history, and
psychosocial history in each subject should be recorded because all of these factors can
affect the thiamine intake of the subject. Several methods including dietary scan, dietary
record, and/or 24-h dietary recall have been employed to assess the thiamine intake in
various populations, and comparisons are made between the mean thiamine intakes and
RDAs for thiamine. For instance, dietary assessment by dietary record in northeastern
Thai villagers revealed a mean daily energy intake of 2441 kcal, with a mean daily thiamine
intake of 0.56 mg or 0.23 mg per 1000 kcal (44), which were only 51% or 46% of
Table 2 Sequential Development of Inadequate
Thiamine Status
Stages of inadequate thiamine status Methods of assessment
I Inadequate thiamine intake Subject’s history
II Tissue desaturation of thiamine Laboratory assessment
III Metabolic derangement Laboratory assessment
IV Clinical manifestations Physical examination
302 Tanphaichitr
the RDAs for thiamine for the North American population (64). Dietary assessment based
on 24-h recall in the North American male population aged 65–75 years showed a mean
daily energy intake of 1805 kcal and a mean daily thiamine intake of 1.40 mg or 0.68
mg/1000 kcal (120), which are 116% or 136% of the 1989 RDAs for thiamine (64). It
is evident that the northeastern Thai villagers have lower intake of thiamine than the
elderly North American population. These dietary data are consistent with the biochemical
findings on their thiamine status (44,120).
B. Laboratory Tests
Two crucial properties of laboratory tests must be considered when they are employed
for the assessment of thiamine status. The tests must not be overly invasive; they should be
sensitive enough to identify inadequate thiamine status prior to the appearance of clinical
manifestations (Table 1) and to assess the efficacy of subsequent thiamine treatment.
Various biochemical tests have been developed for detecting thiamine deficiency
or assessing the adequacy of thiamine status in humans. These include the measurement
of urinary thiamine excretion; blood thiamine level; thiamine concentration in CFS, blood
pyruvate, lactate, and ?-ketoglutarate levels; and ETKA and TPPE (2,3,5,9,10,26,46,121,
1. Urinary Thiamine Excretion
Prior to the development of the measurement of ETKA and TPPE, the most commonly
used approach to thiamine status assessment was the measurement of urinary levels of
thiamine by the thiochrome method after a one-step purification by ion exchange chromatography.
Microbiological assays were also common and utilized Lactobacillus viridescens
(121). Currently, in addition to ETKA and TPPE measurement, thiamine levels in
urine are estimated by HPLC (122).
There is a reasonably close correlation between the development of a thiamine defi-
ciency and decreased urinary thiamine excretion. Approximately 40–90 and 100 µg or
more of thiamine is excreted in the urine daily in adults on intakes of 0.30–0.36 and 0.50
mg of thiamine per 1000 kcal, respectively, whereas daily urinary thiamine excretion falls
to 5–25 µg with the daily intake of 0.2 mg of thiamine per 1000 kcal. Thus, measurement
of the 24-h urinary thiamine excretion is useful in evaluating human thiamine status. Daily
urinary thiamine excretions of 100, 40–99, and 40 µg are considered to designate
acceptable, low, and deficient thiamine status, respectively. However, it is usually not
feasible to collect 24-h urine samples with large groups under survey conditions. Consequently,
random urine samples are obtained, preferably during fasting state, with the simultaneous
determination of thiamine and creatinine concentrations. A correlation between
the urinary thiamine excretion per gram of creatinine and daily thiamine intake in milligrams
per 1000 kcal has been observed. Urinary thiamine excretions of 66, 27–65, and
27 µg/g of creatinine in adults are considered to designate acceptable, low, or deficient
thiamine status. The corresponding figures in pregnant women during the second trimester
are 55, 23–54, and 23 µg/g creatinine, whereas those during the third trimester are
50, 21–49, and 21 µg/g creatinine (121).
Healthy children have markedly higher levels of urinary thiamine excretion per gram
creatinine than adults. Thus, children aged 1–3, 4–6, 7–12, and 13–15 years are considered
to have acceptable thiamine status when their urinary thiamine excretions are 176,
121, 181, and 151 µg/g creatinine, respectively (121).
Thiamine 303
Though urinary thiamine excretion per gram creatinine may be a useful criterion
for surveying thiamine status in a large population, it is not always a good index for
diagnosing beriberi, e.g., of 15 patients with beriberi prior to thiamine treatment, 8(53%),
3(20%), and 4(27%) had urinary thiamine excretions of 27, 27–65, and 66 µg/g creatinine,
respectively (46).
Although a number of thiamine load tests of a known dose of thiamine administered
orally or intramuscularly have been of value in evaluating the extent of thiamine depletion
in tissue stores, the most commonly used procedure is to administer 5 mg of thiamine
parenterally and to measure the urinary thiamine excretion in the following 4-h period.
Thiamine-deficient subjects usually excrete less than 20 µg of thiamine following the test
dose. However, this procedure is inconvenient in nutrition surveys (121).
2. Blood Thiamine Level
Levels of thiamine in whole blood, plasma or serum, and erythrocytes have been employed
to assess human thiamine status. The thiamine concentrations in these biological samples
can be determined by chemical methods, including the thiochrome procedure and colorimetric
techniques utilizing diazotized p-aminoacetophenone or ethyl-p-aminobenzoate;
microbiological assays using Ochromonas danica or a streptomycin-resistant mutant strain
of Lactobacillus fermenti; and HPLC (2,24,26,86,114,121,122). The specificity, sensitivity,
and other limitations of each method must be carefully reviewed whenever employing
it for assessing blood thiamine levels because blood contains only about 0.8% of the total
body thiamine. Since most of the blood thiamine is in the blood cells, the serum level is
still much lower. The thiamine concentration in leukocytes is about 10 times that in erythrocytes.
In the interpretation of blood thiamine tests, the hematocrit must be taken into
consideration because erythrocyte mass usually contributes 75% or more of the total thiamine
(2,86). Although leukocytes contribute 15% or less of the blood thiamine concentration,
the interpretation of a low-normal blood thiamine level in a subject with leukocytosis
must be done with caution (86).
a. Chemical Methods
The determination of total thiamine concentration in whole blood or serum by the thiochrome
method is based on an oxidative conversion of thiamine (by ferricyanide in an
alkaline solution) to thiochrome (Fig. 3), which is highly fluorescent and is extracted from
the aqueous phase with isobutanol for subsequent fluorometric measurement. Oxidation
of phosphorylated thiamine with ferricyanide yields a thiochrome derivative not extractable
with isobutanol. Pretreatment of whole blood or serum with phosphatase preparations,
e.g., potato acid phosphatase, is necessary for obtaining the total thiamine concentration.
Thus, the total and nonphosphorylated thiamine are measured directly, whereas the phosphorylated
thiamine is calculated from total thiamine minus nonphosphorylated thiamine
The colorimetric method based on the coupling of diazotized p-aminoacetophenone
with thiamine to produce an insoluble purple-red compound that can be measured photometrically.
Uric acid and ascorbic acid interfere with the development of the color. Besides,
the method is not sufficiently sensitive to detect the low concentration of blood
thiamine (2,123).
b. Microbiological Assays
Baker et al. (124) developed a protozoan method using the phytoflagellate Ochomonas
danica to measure thiamine concentrations in whole blood, erythrocytes, leukocytes, CSF,
304 Tanphaichitr
and other tissues. Ochromonas danica has a sensitive and specific intact thiamine requirement;
it can assay 375 pmol/L (100 pg/mL) of thiamine in biological fluids and tissues.
It has been shown that reduced blood thiamine levels determined by this protozoan method
occurred within 2 days in rats on thiamine-deficient diet. At 15 days there was almost no
circulating thiamine, and the thiamine contents in liver and muscle fell to 25% of normal,
whereas the ETKA fell after 10 days, and after only 20 days was reduced to nearly 50%
of normal (125).
Thiamine concentrations in serum and erythrocytes can also be accurately measured
by an automated method using a streptomycin-resistant strain of Lactobacillus fermenti
as the test organism. This method is sensitive to 1.88 nmol/L (0.5 µg/L) of thiamine. The
organism responds to both free and phosphorylated thiamine; its growth response to TPP
is approximately 30% greater than that obtained with thiamine HCl (24,26).
c. High-Performance Liquid Chromatography
HPLC determinations of free and phosphorylated thiamine are now frequently carried out
to assess thiamine status (114,123,126–128); a sensitivity for the detection of 0.05 pmol
of thiamine has been reported (128). Besides, erythrocyte TPP levels in healthy adults
determined by HPLC are in general agreement with those determined by the enzymatic
method utilizing yeast pyruvic carboxylase apoenzyme (123,126). The validity of assessing
thiamine status by erythrocyte TPP levels is supported by the following findings: i.e.,
erythrocyte TPP levels in rats fed a thiamine deficient diet fell more rapidly than ETKA
(123); over an intake range of up to 3.4 mg of thiamine per day in eight hospitalized
patients, their erythrocyte TPP levels increased proportionately with the thiamine intake
(129); and a significantly positive correlation between ETAK and whole-blood TPP levels
in 127 adults (r  0.26, p  0.003) as well as a significantly negative correlation between
?-ETKA (ETKA with added TPP/ETKA without added TPP) (r  0.44, p  0.0001)
Unlike urinary thiamine excretion, there is no accepted guideline for the interpretation
of blood thiamine levels. Tables 3 and 4 show normal thiamine levels in biological
fluids determined by the protozoan (125) and HPLC methods (114), respectively.
3. Thiamine Concentration in CSF
In adults, total thiamine in CSF exists in two forms, i.e., 62% as TMP and 38% as thiamine.
The total thiamine, thiamine, and TMP in CSF are four, two, and eight times higher than
the corresponding forms in serum. Besides, there is a correlation between serum and CSF
concentrations of thiamine. TMP is formed from the dephosphorylation of TPP and is
further dephosphorylated into thiamine (Fig. 2). Since TPP is not found in CSF, the high
concentration gradient of TMP from serum to CSF suggests the existence of an active
Table 3 Normal Thiamine Levels in Biological Fluids Determined by the
Protozoan Method Using Ochromonas danica (125)
Thiamine (nmol/L) Thiamine (µg/L)
Specimen Range Mean Range Mean
Whole blood 94–282 120 25–75 32
Serum 563–957 788 150–255 210
CSF 38–113 60 10–30 16
Thiamine 305
Table 4 Concentrations of Various Forms of Thiamine in Blood and Cerebrospinal Fluid in
Adults Determined by HPLC
Subjecta Thiamine formsd (nmol/L)
Author (y) N Specimen Free TMP TTP Total
Tallaksen et al. 23–58 40 Serum 5  23 2–12 — NA
(114) 21–83 31b Serum 8.6  3.9 4.0  1.5 — 12.7  4.2
21–83 45b,c CSF 16.9  8.3 28.1  7.9 — 44.9  14.0
Fidanza et al. 20–82 127 Whole blood 126  35
(127) (64–202)
a All of the subjects were healthy except those with superscripts b and c.
b Blood and CSF were collected simultaneously from 31 patients with severe back pain but were otherwise
healthy during performance of myelogram.
c Only CSF samples were collected from another 14 similar patients.
d Figures are mean  SD or range. NA, not available from the reports.
transport of TMP, which is supported by a saturation mechanism (114). Though simultaneous
measurements of thiamine in serum and CSF reflect the integrity of blood–brain
barrier and thiamine supply to the brain, it is impractical and unjustifiable to assess thiamine
status by the measurement of CSF thiamine in a free-living population. Determination
of CSF thiamine in hospitalized patients with dry beriberi or Wernicke–Korsakoff
syndrome should be based on the risk and benefit of the patients.
4. Blood Pyruvate, ?-Ketoglutarate, and Lactate Levels
In beriberi patients, pyruvic and ?-ketoglutaric acids accumulate in the tissue (Fig. 8) and
their concentration in the blood rises. Part of accumulated pyruvic acid may also be converted
to lactic acid. However, these changes lack specificity and consistency. The fasting
levels of blood pyruvate have frequently been found to be normal in thiamine deficiency
and only rise above normal following a glucose load (99,121). Thus, a glucose load test,
termed the carbohydrate metabolic index (CMI), has been developed; blood glucose (G),
pyruvate (P), and lactate (L) levels are measured at 1 h after the administration of 1.8 g
of glucose per kilogram of body weight and 5 min after completion of a standardized
exercise. CMI is then calculated as follows;
(L  G/10)  (15P  G/10)
The upper normal limit of CMI is 15, based on the studies of experimental thiamine
deficiency in humans (130). However, this procedure requires too much cooperation and
technical ability to be useful routinely (121).
5. Blood Transketolase Activity
Blood is a feasible specimen to be obtained for the determination of transketolase activity,
which can be measured in erythrocytes, leukocytes, or whole blood.
a. ETKA and TPPE
Brin and his co-workers (131) have demonstrated that ETKA and its percent stimulation
in vitro by added TPP, called TPPE, are sensitive methods to detect thiamine deficiency
306 Tanphaichitr
both in rats and in humans. Healthy adults fed diets containing 190 µg of thiamine daily
for 8 weeks showed a decline in ETKA as early as 8 days, and their ETKA at 17–20
days was markedly lower than that in the control subjects; their TPPE was 15% and 34%
at days 8 and 31, respectively, whereas TPPE in the control subjects was less than 10%
throughout the study; and following thiamine treatment, their ETKA and TPPE returned
to normal values. These findings indicate the specificity of TPP on the depressed ETKA
in thiamine deficiency in humans. In this study, the urinary thiamine excretion in these
thiamine-deficient subjects was sharply reduced to less than 50 µg/day by the second
week and to less than 20 µg/day by the third week, whereas their CMI was below 15
throughout the study. These findings indicate the sensitivity of ETKA and TPPE in detecting
human thiamine deficiency. Subsequent works have established the validity of
ETKA and TPPE for the diagnosis of various forms of beriberi in infants and adults, and
have proved that they are useful indicators of biochemical thiamine depletion prior to the
appearance of beriberi (2,3,5,9,10,30,43,46,56,121,130). The measurement of ETKA and
TPPE represents a functional test of thiamine adequacy and hence may be a more reliable
indicator of thiamine insufficiency than urinary thiamine excretion (121).
ETKA can be measured by a colorimetric procedure involving incubating hemolyzed
erythrocyte samples for 30 or 60 min at 37–43°C in a buffered medium with an excess
of ribose-5-phosphate with and without added TPP and then determining the amount of
ribose-5-phosphate used, sedoheptulose-7-phosphate formed, or the hexoses produced
(26,121). TPPE is then calculated as follows:
TPPE in % 
ETKA with added TPP  ETKA without added TPP
ETKA without added TPP
In some reports, the stimulating effect of TPP on ETKA is expressed as the activity coeffi-
cient of ETKA (?-ETKA), i.e., ETKA with added TPP/ETKA without added TPP (127).
ETKA obtained without the addition of TPP represents the basal level of enzyme
activity and is dependent on the availability of endogenous erythrocyte TPP. The addition
of TPP permits estimation of the amount of apoenzyme uncomplexed as well as the maximum
potential activity. A thiamine deficiency results in a reduction in TPP, which results
in a reduction of the apotransketolase activity (121). This is supported by a correlation
between transketolase activity and TPP concentrations (80).
A modification of the Dreyfus method (132) has been employed for the determination
of ETKA in our laboratory and ETKA is expressed in international units (IU), which
is equivalent to the number of micromoles of sedoheptulose-7-phosphate formed per minute
per liter of erythrocytes.
A colorimetric procedure has been developed that distinguishes transketolase activity
from overall ribose-5-phosphate utilization by measuring ribose disappearance and
sedoheptulose formation in the same color reaction (133). More direct spectrometric methods
have evolved based on the change in absorption at 340 nm due to the decrease in
NADH. For example, glyceraldehyde-3-phosphate oxidation has been used (134). A semiautomated,
continuous-flow procedure for determining unstimulated ETKA and TPP-stimulated
ETKA has also been described (135). The sensitivity, reliability, and precision of
this approach is improved by eliminating hemoglobin interference.
Table 5 shows Brin’s guideline (136) and the guideline of the Interdepartmental
Committee on Nutrition for National Defense (ICNND) (121) for interpreting TPPE levels
in evaluating thiamine status. Both guidelines are agreeable for the TPPE levels of normal
Thiamine 307
Table 5 Guidelines for Interpreting TPPE levels in
Evaluating Thiamine Status in All Ages
Brin’s guideline (136) ICNND’s guideline (121)
Category TPPE (%) Category TPPE (%)
Normal 0–14 Acceptable 0–15
Marginally deficient 15–24 Low 16–20
Severely deficient 25 Deficient 20
or acceptable thiamine status, whereas the TPPE levels for grading severity of thiamine
deficiency are different. The validity of the TPPE levels of 0–15% for acceptable thiamine
status is substantiated by the finding that 94% of beriberi patients prior to thiamine treatment
showed TPPE of 16% or higher (46,99). However, both ETKA and TPPE must be
considered in assessing the adequacy of thiamine status because the patterns of these two
biochemical parameters in acute and chronic thiamine deficiency are different.
Subjects with acute thiamine deficiency have low ETKA with concomitantly high
TPPE levels. Their unstimulated ETKA 1 h after the parenteral administration of thiamine
is equivalent to or higher than the TPP-stimulated ETKA prior to thiamine treatment and
their TPPE is in the acceptable range of 0–15% (10,46,99).
Experimental studies in rats (137) and humans (138) have shown that after a certain
period of thiamine deprivation the TPP added in vitro cannot restore ETKA fully. Thus,
despite low ETKA, high TPPE is not seen. These data are supported by findings that the
TPP added in vitro does not increase ETKA in patients with beriberi prior to thiamine
treatment (46,99) and in free-living populations with chronic inadequate thiamine intake
(43) to the unstimulated ETKA in normal adults. Also, there is no significant difference
in TPPE between subjects with adequate thiamine intake and those with chronically low
intake of thiamine, despite significantly lower ETKA in the latter group (43).
The low unstimulated ETKA in cirrotic patients can be due either to the impaired
conversion of thiamine to TPP or to apotransketolase deficiency. In those with an impairment
of thiamine conversion to TPP, there is no increase in the low ETKA after following
thiamine treatment; only the added TPP in vitro raises their ETKA. In those with apotransketolase
deficiency, there are no increases in ETKA after thiamine treatment and the
addition of TPP in vitro; in some instances, the transketolase level returns to normal when
positive nitrogen balance is restored (125).
Apotransketolase deficiency, evidenced by decreases in unstimulated ETKA and
TPP-stimulated ETKA with no change in TPPE, has also been shown in patients with
insulin-dependent diabetes mellitus, patients with non–insulin-dependent diabetes mellitus
(139), and uremic patients (140).
Elevated ETKA with normal TPPE is documented in patients with pernicious anemia
(139,141,142), whereas patients with megaloblastic anemia caused by folate deficiency
or fish tapeworm do not exhibit elevated ETKA (141). Since TPPE is only rarely increased
in pernicious anemia with or without neuropsychiatric manifestations, thiamine deficiency
should not be a feature of this disease and does not take part in the genesis of hyperpyruvicacidemia
(142). Because transketolase activity is normally greater in young erythrocytes
than in older cells, it has been suggested that patients with pernicious anemia have
a larger population of young erythrocytes than healthy individuals (139,142). This suggestion
is supported by findings of increased ETKA in patients with iron deficiency anemia
308 Tanphaichitr
responding to the iron treatment, in most patients with congenital hemolytic anemia, and
in some patients with acute leukemia; and the normalization of unstimulated ETKA after
3–5 weeks of vitamin B12 treatment as well as the significantly negative correlation between
hemoglobin levels and TPP-stimulated ETKA (142).
b. Leukocyte Transketolase Activity
Erythrocytes have a limited ability to respond to stimuli because they have a long halflife
and lack a nucleus, whereas leukocytes would be expected to respond more rapidly
to changes in nutritional status and perhaps more nearly reflect the status of other cells
in the body. Study in the rats has shown leukocyte transketolase activity to be a reliable
indicator of changes in thiamine status. It is not affected by dietary levels of carbohydrate,
protein, and fat; it responds well to changes in dietary thiamine levels; it reaches a maximum,
which does not increase further when excess thiamine is fed; and there is a parallel
change of transketolase activity in leukocytes and liver (143). However, the use of leukocytes
in assessing thiamine status in humans imposes greater technical difficulties than
the use of erythrocytes.
Dreyfus (132) has demonstrated that WBTK and TPPE can be employed to assess thiamine
status in humans and to confirm the diagnosis of Wernicke’s disease. The usefulness of
WBTK and TPPE is substantiated by subsequent studies in assessing thiamine status in
pregnant women and their cord blood (76), in infants and children (55), in patients with
thyrotoxicosis (144), and patients with wet beriberi (145). Thus, in certain circumstances,
whenever the separation of erythrocytes is not feasible, a whole-blood specimen can be
used to determine transketolase activity and TPPE.
C. Physical Examination
Though physical signs of beriberi appear in the last stage of thiamine deficiency (Table
2), physicians should be able to detect such abnormalities in patients for immediate and
long-term nutritional management (62). The delay in diagnosing beriberi affects the morbidity
and mortality rates of the patients. Beriberi should be suspected in infants being
breast-fed by thiamine-deficient mothers; such infants would have a loud piercing cry,
dyspnea, cyanosis, cardiac failure, and aphonia (10,97).
Common suggestive signs in dry beriberi include glove-and-stocking hypoesthesia
of pain and touch sensations, loss of ankle and/or knee reflexes, tenderness of calf muscle,
difficulty in rising from squatting position, and aphonia. However, other possible known
causes of peripheral neuropathy must be carefully ruled out (10,46,97,99).
Patients with wet beriberi exhibit both peripheral neuropathy and edema. Severe
cases show tachycardia, wide pulse pressure, cardiac enlargement, and pulmonary congestion
Wernicke’s disease must be suspected in chronic alcoholics presenting with the
triad of ocular motor signs, ataxia, and derangement of motor functions. The diagnosis
of Wernicke–Korsakoff syndrome should be made in those having both the ocular-ataxic
and amnestic manifestations (109).
As already mentioned in Sec. IX.B, drastic responses to thiamine treatment are observed
in patients with cardiac beriberi or Wernicke’s disease, whereas more time elapses
before improvement is observed in patients with dry beriberi or Korsakoff’s psychosis.
Laboratory tests are required in the latter conditions to confirm the diagnosis.
Thiamine 309
There is concern about the impact of persisting subclinical or borderline thiamine defi-
ciency on the health status of the population. The major hazard for subjects with subclinical
thiamine deficiency is the increased risk of beriberi when faced with extreme physiological
or pathological conditions, e.g., pregnancy, lactation, high physical activity, infection, or
surgery. This is well illustrated in a study in Japanese university students (49). Routine
physical examination in 766 students revealed irritability with mild cardiovascular signs
in 42 students. Of these 42 students, 16 had whole-blood thiamine levels below 50 ng/
mL. A review of 2754 chest x-ray films indicated that 93 students had cardiothoracic
ratios greater than 50%. Of these 93 students, 44 had whole-blood thiamine levels below
50 ng/mL. The analysis of lifestyle in 59 students with low whole-blood thiamine levels
revealed that 50–60% ate their meals in restaurants and 39–47% did not eat breakfast;
86% of students with cardiac enlargement and low blood thiamine levels undertook strenuous
daily exercise; only 20% were aware of their own abnormalities. The improvement
of cardiothoracic ratio was obtained in 63% of 83 subjects by eliminating strenuous exercise
and administering sound nutritional advice.
More studies are needed to verify subclinical thiamine deficiency in persons having
psychological, cardiovascular, and neurological symptoms. The application of dietary assessment
and laboratory tests should be conducted in populations with chronic low intake
of thiamine and appropriate dietary guidelines be implemented to improve their thiamine
Thiamine is essential in all phases of metabolism. The recommended thiamine intake is
expressed in terms of total caloric intake. The current U.S. recommended thiamine allowances
are 0.5 mg/1000 kcal for children, adolescents, and adults, and 0.4 mg/1000 kcal
for infants. These recommendations are based on the assessment of the effects of varying
levels of dietary thiamine on the occurrence of deficiency signs, on the excretion of thiamine
or its metabolites, and on ETKA. A minimum thiamine intake of 1.0 mg/day is
recommended for adults who consume less than 2000 kcal daily. An additional thiamine
intake of 0.4 mg/day is recommended throughout pregnancy to accommodate maternal
and fetal growth and increased maternal caloric intake. To account for both the thiamine
loss in milk and increased energy consumption during lactation, an increment of 0.5 mg/
day is recommended throughout lactation (64).
Vitamin-responsive or vitamin-dependent diseases are a group of genetically determined
metabolic disorders in which either a vitamin-dependent enzymatic step or a reaction
involving the conversion of a vitamin to its active cofactor form is defective, causing the
abnormal accumulation of metabolites or substrates in the blood. In these diseases, blood
vitamin levels are normal. The basic metabolic defect involves the structure of the apoenzyme,
its coenzyme binding sites, or some aspect of coenzyme synthesis (146).
Thiamine-responsive diseases include maple syrup urine disease, pyruvate decarboxylase
deficiency, subacute necrotizing encephalopathy (Leigh’s disease), and thiamineresponsive
megaloblastic anemia. Such patients require pharmacological doses of thiamine
to alleviate their clinical manifestations (3,26,146,147).
310 Tanphaichitr
Excess thiamine is easily cleared by the kidneys. Although there is some evidence of
toxicity from large doses given parenterally, there is no evidence of thiamine toxicity by
oral administration; oral doses of 500 mg taken daily for a month were found to be nontoxic
Although much has been accomplished in terms of the nutritional, biochemical, physiological,
pharmacological, and molecular aspects of thiamine, humans, especially those on
rice-based diets and chronic alcoholics, are still facing the problem of inadequate thiamine
status, ranging from subclinical thiamine deficiency to beriberi. To combat the problem
of thiamine deficiency, existing nutritional knowledge should be disseminated to motivate
the public to consume appropriate foods in their daily life. To achieve this dietary goal,
both governmental and private agencies must work closely together and recognize the
role of community participation.
1. R. R. Williams, Toward the Conquest of Beriberi, Harvard University Press, Cambridge,
MA, 1961.
2. C. J. Gubler, Thiamin, Handbook of Vitamins: Nutritional, Biochemical, and Clinical Aspects
(L. J. Machlin, ed.), Marcel Dekker, New York, 1984, p. 233.
3. R. H. Haas, Thiamin and the brain, Annu. Rev. Nutr. 8:483 (1988).
4. D. B. McCormick, Thiamin, in Modern Nutrition in Health and Disease, 7th ed. (M. E. Shils
and V. R. Young, eds.), Lea & Febiger, Philadelphia, 1988, p. 355.
5. V. Tanphaichitr, Thiamin, in Modern Nutrition in Health and Disease, Vol. 1, 8th ed. (M. E.
Shils, J. A. Olson, and M. Shike, eds.), Lea & Febiger, Philadelphia, 1994, p. 359.
6. AIN Committee on Nomenclature, Nomenclature policy: generic descriptors and trivial
names for vitamins and related compounds, J. Nutr. 117:7 (1987).
7. C. Kawasaki, From natural thiamine to synthetic thiamine: the brief history of thiamine preparations,
in Review of Japanese Literature on Beriberi and Thiamine (N. Shimazono and E.
Katsura, eds.), Igaku Shoin Ltd., Tokyo, 1965, p. 288.
8. U.S. Department of Health, Education, and Welfare, and Food and Agriculture Organization
of the United Nations, Food Composition Table in East Asia, DHEW Publication No. (NIH)
73-465 (1972), Bethesda, MD, 1972.
9. V. Tanphaichitr and B. Wood, Thiamin, in Nutrition Review’s Present Knowledge in Nutrition,
5th ed. (R. E. Olson, H. P. Broquist, C. O. Chichester, W. J. Darby, A. C. Kolbye, Jr.,
and R. M. Stalvey, eds.), Nutrition Foundation, Washington, DC, 1984, p. 273.
10. V. Tanphaichitr, Epidemiology and clinical assessment of vitamin deficiencies in Thai children,
in Child Health in the Tropics (R. E. Eeckels, O. Ransome-Kuti, and C. C. Kroonenberg,
eds.), Martinus Nijhoff, Dordrecht, 1985, p. 157.
11. N. Yagi and Y. Itokawa, Cleavage of thiamine by chlorine in tap water, J. Nutr. Sci. Vitaminol.
25:281 (1979).
12. C. M. Leevy, Thiamin deficiency and alcoholism, Ann. N.Y. Acad. Sci. 378:316 (1982).
13. A. M. Hoyumpa, Jr., Mechanisms of thiamin deficiency in chronic alcoholism, Am. J. Clin.
Nutr. 33:2750 (1980).
14. K. Murata, Thiaminase, in Review of Japanese Literature on Beriberi and Thiamine (N.
Shimazono and E. Katsura, eds.), Igaku Shoin Ltd., Tokyo, 1965, p. 220.
Thiamine 311
15. K. Murata, Actions of two types of thiaminase on thiamin and its analogues, Ann. N.Y. Acad.
Sci. 378:146 (1982).
16. S. Vimokesant, S. Kunjara, K. Rungruangsak, S. Nakornchai, and B. Panijpan, Beriberi
caused by antithiamin factors in food and its prevention, Ann. N.Y. Acad. Sci. 378:123 (1982).
17. D. M. Hilker and J. C. Somogy, Antithiamins of plant origin: their chemical nature and mode
of action, Ann. N.Y. Acad. Sci. 378:137 (1982).
18. P. Ruenwonsa and S. Pattanavibag, Decrease in the activities of thiamine pyrophosphate
dependent enzymes in rat brain after prolonged tea consumption, Nutr. Rep. Int. 27:713
19. P. Ruenwonsa and S. Pattanavibag, Impairment of acetylcholine synthesis in thiamine defi-
cient rats developed by prolonged tea consumption, Life Sci. 34:365 (1984).
20. L. Howard, C. Wagner, and S. Schenker, Malabsorption of thiamine in folate deficient rats,
J. Nutr. 107:775 (1977).
21. A. D. Thomson, H. Baker, and C. M. Leevy, Patterns of 35S thiamin hydrochloride absorption
in the malnourished alcoholic patient, J. Lab. Clin. Med. 76:35 (1970).
22. K. Inouye and E. Katsura, Diagnosis, prevention and therapy of beriberi, in Review of Japanese
Literature on Beriberi and Thiamine (N. Shimazono and E. Katsura, eds.), Igaku Shoin
Ltd., Tokyo, 1965, p. 64.
23. G. Rindi and U. Ventura, Thiamine intestinal transport, Physiol. Rev. 52:821 (1972).
24. R. E. Davis, G. C. Icke, J. Thom, and W. J. Riley, Intestinal absorption of thiamin in man
compared with folate and pyridoxal and its subsequent urinary excretion, J. Nutr. Sci. Vitaminol.
30:475 (1984).
25. R. M. Bayliss, R. Brooks, J. McCulloch, J. M. Kuyl, and J. Metz, Urinary thiamin excretion
after oral physiological doses of the vitamin, Int. J. Vitam. Nutr. Res. 54:161 (1984).
26. R. E. Davis and G. Icke, Clinical chemistry of thiamin, Adv. Clin. Chem. 23:93 (1983).
27. K. Muniyappa, U. S. Murphy, and P. R. Adiga, Estrogen induction of thiamin carrier protein
in chicken liver, J. Steroid Biochem. 9:888 (1978).
28. P. R. Adiga and K. Muniyappa, Estrogen induction and functional importance of carrier
proteins for riboflavin and thiamin in the rat during gestation, J. Steroid Biochem. 9:829
29. N. Shimazono, Metabolism of thiamine in the animal body, Review of Japanese Literature
on Beriberi and Thiamine (N. Shimazono and E. Katsura, eds.), Igaku Shoin Ltd., Tokyo,
1965, p. 149.
30. V. Tanphaichitr, Thiamin, Nutrition Review’s Present Knowledge in Nutrition, 4th ed. (D. M.
Hegsted, C. O. Chichester, W. J. Darby, K. W. McNutt, R. M. Stalvey, and E. H. Stotz,
eds.), Nutrition Foundation, New York, 1976, p. 141.
31. H. Behal, B. B. Buxton, J. G. Robertson, and M. S. Olson, Regulation of the pyruvate dehydrogenase
multienzyme complex, Annu. Rev. Nutr. 13:497 (1993).
32. M. S. Olson, Bioenergetics and oxidative metabolism, in Textbook of Biochemistry with Clinical
Correlations, 3rd ed. (T. M. Devlin, ed.), Wiley-Liss, New York, 1992, p. 237.
33. A. H. Mehler, Amino acid metabolism of the individual amino acids, in Textbook of Biochemistry
with Clinical Correlations, 3rd ed. (T. M. Devlin, ed.), Wiley-Liss, New York, 1992,
p. 491.
34. J. P. Blass, Thiamin and the Wernicke-Korsakoff syndrome, in Vitamins in Human Biology
and Medicine (M. H. Briggs, ed.), CRC Press, Boca Raton, FL, 1981, p. 107.
35. A. von Muralt, The role of thiamine in neurophysiology, Ann. NY Acad. Sci. 98:499 (1962).
36. J. R. Cooper and J. H. Pincus, The role of thiamine in nervous tissue, Neurochem. Res. 4:
223 (1979).
37. Y. Itokawa and J. R. Cooper, Thiamin release from nerve membranes by tetrodoxin, Science
166:759 (1969).
38. Y. Itokawa and J. R. Coper, Ion movements and thiamine in nervous tissue. I. Intact nerve
preparations, Biochem. Pharmacol. 19:985 (1970).
312 Tanphaichitr
39. E. Schoffeniels, Thiamine phosphorylated derivatives and bioelectrogenesis, Arch. Int. Physiol.
Biochem. 91:233 (1983).
40. T. Oiso, Thiamine intake and its deficiency signs among the Japanese based on the results
of the national nutrition survey, in Review of Japanese Literature on Beriberi and Thiamine
(N. Shimazono and E. Katsura, eds.), Igaku Shoin Ltd., Tokyo, 1965, p. 275.
41. E. Katsura and T. Oiso, Beriberi, in Nutrition in Preventive Medicine (G. H. Beaton and
J. M. Bengoa, eds.), World Health Organization, Geneva, 1976, p. 136.
42. M. S. Bamji, Vitamin deficiencies in rice-eating-populations: effects of B-vitamin supplements,
Nutritional Adequacy, Nutrient Availability and Needs (J. Mauron, ed.), Birkha?user
Verlag, Basel, 1983, p. 245.
43. V. Tanphaichitr, N. Lerdvuthisopon, S. Dhanamitta, and A. Valyasevi, Thiamin status in
northeastern Thais, Intern. Med. 6:43 (1990).
44. M. Kimura, N. Sato, and Y. Itokawa, Thiamin status of inhabitants on North-East Thailand,
Trace Nutrients Res. 4:163 (1988).
45. V. Tanphaichitr, Adolescent beriberi, Thai J. Intern. Med. 4:42 (1984).
46. V. Tanphaichitr, S. L. Vimokesant, S. Dhanamitta, and A. Valyasevi, Clinical and biochemical
studies of adult beriberi, Am. J. Clin. Nutr. 8:1017 (1970).
47. K. Takahashi, T. Kittagawa, and M. Shimao, Acute polyneuritis associated with edema: a
recent revival of beriberi neuropathy in Japan, Jpn. J. Med. 15:214 (1976).
48. C. Kawai, A. Wakabayashi, T. Matsumara, and Y. Yui, Reappearance of beriberi heart disease
in Japan, Am. J. Med. 69:383 (1980).
49. Y. Hatanaka and K. Ueda, Incidence of subclinical hypovitaminosis of B1 among university
students found by a field study in Ehime, Japan, J. Osaka Univ. 31:83 (1981).
50. V. Tanphaichitr, N. Lerdvuthisopon, S. Dhanamitta, and H. P. Broquist, Carnitine status in
Thai adults, Am. J. Clin. Nutr. 33:876 (1980).
51. A. Valyasevi, S. B. Halstead, S. Pantuwatana, and C. Tankayl IV. Dietary habits, nutritional
intake, and infant feeding practices among residents of a hypo-hyperendemic area, Am. J.
Clin. Nutr. 20:1340 (1967).
52. A. Valyasevi, S. Vimokesant, and S. Dhanamitta, Chemical compositions of breast milk in
different locations of Thailand, J. Med. Assoc. Thailand 51:348 (1968).
53. I. G. Macy, Composition of human colostrum and milk, Am. J. Dis. Child. 78:589 (1949).
54. O. Thanangkul and J. A. Whitaker, Childhood thiamine deficiency in northern Thailand, Am.
J. Clin. Nutr. 18:275 (1966).
55. B. Pongpanich, N. Srikrikrich, S. Dhanamitta, and A. Valyasevi, Biochemical detection of
thiamin deficiency in infants and children in Thailand, Am. J. Clin. Nutr. 27:1399 (1974).
56. L. Cheunchit and V. Tanphaichitr, Infantile beriberi, Supartprasong Hosp. Gaz. 1:97 (1975).
57. S. Sintarat and S. Pansuravet, Infantile beriberi in Nakornpanom Hospital, J. Nutr. Assoc.
Thailand 14:184 (1980).
58. V. Nivatvongs and S. Anantkosol, Infantile beriberi: a retrospective study of 52 cases, Bull.
Khon Khaen Univ. Health Sci. 11 (Suppl.):41 (1985).
59. M. Victor, Deficiency diseases of the nervous system secondary to alcoholism, Postgrad.
Med. 50:75 (1971).
60. R. A. Harris and D. W. Crabb, Metabolic interrelationships, in Textbook of Biochemistry
with Clinical Correlations, 3rd ed. (T. M. Devlin, ed.), Wiley-Liss, New York, 1992, p.
61. J. P. Bonjour, Vitamins and alcoholism. IV. Thiamin, Int. J. Vitam. Nutr. Res. 50:321 (1980).
62. V. Tanphaichitr, Evaluation of nutritional status, Medical Practice of Preventive Nutrition
(M. L. Wahlqvist and J. S. Vobecky, eds.), Smith-Gordon and Co. Ltd., London, 1994, p. 333.
63. A. Lemonine, C. L. Devahat, J. L. Codacciont, A. Monges, P. Bermound, and R. M. Salkeld,
Vitamin B1, B2, B6 and C status in hospital patients, Am. J. Clin. Nutr. 33:2595 (1980).
64. Committee on the Tenth Edition of the RDAs, Food and Nutrition Board, Recommended
Dietary Allowances, 10th ed., National Academy Press, Washington, DC, 1989, p. 125.
Thiamine 313
65. G. Blennow, Wernicke encephalopathy following prolonged artificial nutrition, Am. J. Dis.
Child. 129:1456 (1975).
66. A. M. Nadel and P. C. Burger, Wernicke encephalopathy following prolonged intravenous
therapy, JAMA 235:2403 (1976).
67. F. A. Baughman, Jr., and J. P. Papp, Wernicke’s encephalopathy with intravenous hyperalimentation:
remarks on similarities between Wernicke’s encephalopathy and the phosphatedepletion
syndrome, Mt. Sinai J. Med. 43:48 (1976).
68. J. Kramer and J. A. Goodwin, Wernicke’s encephalopathy: complication of intravenous hyperalimentation,
JAMA 238:2176 (1976).
69. R. J. Velez, B. Myers, and M. S. Guber, Severe acute metabolic acidosis (acute beriberi):
an avoidable complication of total parenteral nutrition, J. Parent. Ent. Nutr. 9:216 (1985).
70. P. Laselve, P. Demolin, and L. Hohzapfel, Shoshin beriberi: an unusual complication of
prolonged parenteral nutrition, J. Parent. Ent. Nutr. 10:102 (1986).
71. D. Oriot, C. Wood, R. Gottesman, and G. Huault, Severe lactic acidosis related to acute
thiamine deficiency, J. Parent. Ent. Nutr. 15:105 (1991).
72. J. Zak III, D. Burns, T. Lingenfelser, E. Steyn, and I. N. Marks, Dry beriberi: unusual complication
of prolonged parenteral nutrition, J. Parent. Ent. Nutr. 15:200 (1991).
73. M. Comabella, A. Canton, X. Montalban, and A. Codina, Iatrogenic fulminant beriberi, Lancet
346:182 (1995).
74. K. Hiroi, Y. Goto, J. Ischikawa, K. Kida, and H. Matsuda, A case of beriberi accompanying
short bowel, Acta Paediatr. Jpn. 37:84 (1995).
75. P. Migasena, S. Changbumrung, V. Supawan, and J. Limitrakarn, Erythrocyte transketolase
activity in blood of mothers, umbilical cord and newborn babies in Thailand, J. Nutr. Sci.
Vitaminol. 20:371 (1974).
76. V. Tanphaichitr and N. Lerdvuthisopon, Thiamin status in primigravidas and in cord blood,
in Vitamins and Minerals in Pregnancy and Lactation (H. Berger, ed.), Raven Press, New
York, 1988, p. 71.
77. L. B. Slobody, M. M. Willner, and J. Mestern, Comparison of vitamin B1 levels in mothers
and their newborn infants, Am. J. Dis. Child. 77:736 (1949).
78. K. Tripathy, Erythrocyte transketolase activity and thiamin transfer across human placenta,
Am. J. Clin. Nutr. 21:739 (1968).
79. O. Frank and H. Baker, Vitamin profile in rats fed stock or liquid ethanolic diets, Am. J.
Clin. Nutr. 33:221 (1980).
80. L. G. Warnock, C. R. Prudhomme, and C. Wagner, The determination of thiamin pyrophosphate
in blood and other tissues and its correlation with erythrocyte transketolase activity,
J. Nutr. 108:421 (1978).
81. D. Labadarios, J. E. Rossouw, J. B. McConnell, M. Davis, and R. Williams, Thiamin defi-
ciency in fulminant hepatic failure and effects of supplementation, Int. J. Vitam. Nutr. Res.
47:17 (1977).
82. S. Heller, R. Salkeld, and W. F. Korner, Vitamin B1 status in pregnancy, Am. J. Clin. Nutr.
27:1221 (1974).
83. P. S. Bergin and P. Harvey, Wernicke’s encephalopathy and central pontine myelonolysis
associated with hyperemesis gravidarum, Br. Med. J. 35:517 (1992).
84. P. J. Galloway, Wernicke’s encephalopathy and hyperemesis gravidarum, Br. Med. I. 305:
1096 (1992).
85. P. A. Nail, M. R. Thomas, and B. S. Eakin, The effect of thiamin and riboflavin supplementation
on the level of those vitamins in human breast milk and urine, Am. J. Clin. Nutr. 33:
198 (1980).
86. D. T. Wyatt, D. Nelson, and R. E. Hillman, Age-dependent changes in thiamin concentrations
in whole blood and cerebrospinal fluid in infants and children, Am. J. Clin. Nutr. 53:530
87. D. A. Gans and A. E. Harper, Thiamin status of incarcerated and nonincarcerated adoles314
cent males: dietary intake and thiamin pyrophosphate response, Am. J. Clin. Nutr. 53:1471
88. V. Pattamakom, P. Rattanasiri, C. Roongpisuthipong, and V. Tanphaichitr, Thiamin malnutrition
in institutionalized subjects, Thai J. Intern. Med. 4:1 (1984).
89. R. F. Butterworth, C. Gaudreau, J. Vincelette, A. M. Bourgault, F. Lamothe, and A. M.
Nutine, Thiamine deficiency and Wernicke’s encephalopathy in AIDS, Metab. Brain Dis. 6:
207 (1991).
90. R. Boldorini, L. Vago, A. Lechi, F. Tedeschi, and G. R. Tragattoni, Wernicke’s encephalopathy:
occurrence and pathological aspects in a series of 400 AIDS patients, Acta Biomed.
Ateneo. Parmense. 63:43 (1992).
91. W. W. Coon and L. S. Bizer, Subclinical thiamine deficiency in postoperative patients, Surg.
Gynecol. Obstet. 121:315 (1967).
92. E. Descombes, C. A. Dessibury, and G. Fellay, Acute encephalopathy due to thiamine defi-
ciency (Wernicke’s encephalopathy) in a chronic hemodialyzed patient: a case report, Clin.
Nephrol. 35:171 (1991).
93. V. Jayadha, J. H. Deck, W. C. Halliday, and H. S. Smyth, Wernicke’s encephalopathy in
patients on peritoneal dialysis or hemodialysis, Ann. Neurol. 2:78 (1987).
94. H. Seligmann, H. Halkin, S. Rauchfleisch, N. Kaufmann, R. Tal, M. Motro, Z. Vered, and
D. Ezra, Thiamine deficiency in patients with congestive heart failure receiving long-term
furosemide therapy: a pilot study, Am. J. Med. 92:151 (1991).
95. J. A. Brady, C. L. Rock, and M. R. Horneffer, Thiamin status, diuretic medications, and the
management of congestive heart failure, J. Am. Diet. Assoc. 95:541 (1995).
96. K. Inouye and E. Katsura, Clinical signs and metabolism of beriberi patients, in Review of
Japanese Literature on Beriberi and Thiamine (N. Shimazono and E. Katsura, eds.), Igaku
Shoin Ltd., Tokyo, 1965, p. 29.
97. S. Chaithiraphan, V. Tanphaichitr, and T. O. Cheng, Nutritional heart disease, in The International
Textbook of Cardiology (T. O. Cheng, ed.), Pergamon Press, New York, 1986, p. 864.
98. J. Salcedo, Jr., Experience in the etiology and prevention of thiamin deficiency in the Philippine
Islands, Ann. N.Y. Acad. Sci. 378:568 (1982).
99. V. Tanphaichitr, Clinical and biochemical studies in beriberi, A thesis submitted to the Faculty
of Graduate Studies in partial fulfillment of the requirements for the degree of master
of science, University of Medical Sciences, Bangkok, 1968.
100. R. A. Pallister, Electrocardiogram in Oriental beriberi, Trans. Soc. Trop. Med. Hyg. 48:490
101. Y. Sukumalchantra, V. Tanphaichitr, V. Tongmitr, and B. Jumbala, Electrocardiographic
changes in cardiac beriberi, J. Med. Assoc. Thailand 57:80 (1974).
102. Y. Sukumalchantra, V. Tanphaichitr, V. Tongmitr, and B. Jumbala, Electrocardiographic
changes in adult beriberi: changes in the QRS voltage, Mod. Med. Asia 11:6 (1975).
103. W. J. Lahey, D. B. Arst, M. Silver, C. R. Kleeman, and P. Kunkel, Physiologic observations
on a case of beriberi heart disease, Am. J. Med. 14:248 (1953).
104. P. I. Wagner, Beriberi heart disease. Physiologic data and difficulties in diagnosis, Am. Heart
J. 69:200 (1965).
105. M. Akbarian, N. A. Yankapoules, and W. H. Abelmann, Hemodynamic studies in beriberi
heart disease, Am. J. Med. 41:197 (1966).
106. V. Thongmitr, Y. Sukumalchantra, V. Tanphaichitr, and B. Jumbala, Hemodynamic studies
in adult beriberi, J. Med. Assoc. Thailand 56:703 (1973).
107. Y. Sukumalchantra, V. Thongmitr, V. Tanphaichitr, and B. Jumbala, Variability of cardiac
output in beriberi heart disease, Mod. Med. Asia 12:7 (1976).
108. W. S. Colacci and E. Braunwald, Cardiac tumors, cardiac manifestations of systemic diseases,
and traumatic cardiac injury, in Harrison’s Principles of Internal Medicine, Vol. 1, 13th ed.
(K. J. Isselbacher, E. Braunwald, J. D. Wilson, J. B. Martin, A. S. Fauci, and D. L. Kasper,
eds.), McGraw-Hill, New York, 1994, p. 1101.
Thiamine 315
109. M. Victor and J. B. Martin, Nutritional and metabolic diseases of the nervous system, in
Harrison’s Principles of Internal Medicine, Vol. 2, 13th ed. (K. J. Isselbacher, E. Braunwald,
J. D. Wilson, J. B. Martin, A. S. Fauci, and D. L. Kasper, eds.), McGraw-Hill, New York,
1994, p. 2238.
110. P. M. Dreyfus and M. Seyal, Diet and nutrition in neurologic disorders, in Modern Nutrition
in Health and Disease, Vol. 2, 8th ed. (M. E. Shils, J. A. Olson, and M. Shike, eds.), Lea &
Febiger, Philadelphia, 1994, p. 1349.
111. J. C. M. Brust, Alcoholism, in Meritt’s Textbook of Neurology, 9th ed. (L. W. Rowland,
ed.), Williams & Wilkins, Baltimore, 1995, p. 967.
112. O. E. Pratt, H. K. Rooprii, G. W. Shaw, and A. D. Thomas, The genesis of alcoholic brain
tissue injury, Alcohol Alcoholism 25:217 (1990).
113. C. K. Singleton, S. R. Pekovich, B. A. McCool, and P. R. Martin. The thiamin-dependent
hysteretic behavior of human transketolase: implications for thiamin deficiency, J. Nutr. 125:
189 (1995).
114. C. M. E. Tallaksen, T. Bohmer, and H. Bell, Concentrations of the water-soluble vitamins
thiamin, ascorbic acid, and folic acid in serum and cerebrospinal fluid of healthy individuals,
Am. J. Clin. Nutr. 56:559 (1992).
115. H. Baker, A. D. Thomson, O. Frank, and C. M. Leevy, Absorption and passage of fat- and
water-soluble thiamin derivatives into erythrocytes and cerebrospinal fluid of man, Am. J.
Clin. Nutr. 27:676 (1974).
116. N. Kitamori and Y. Itokawa, Pharmacokinetics of thiamin after oral administration of thiamin
tetrahydrogurfuryl disulfide to humans, J. Nutr. Sci. Vitaminol. 39:465 (1993).
117. H. Baker and O. Farnk, Absorption, utilization and clinical effectiveness of allithiamines
compared to water-soluble thiamines, J. Nutr. Sci. Vitaminol. 22 (Suppl.):63 (1976).
118. Y. Itokawa, M. Kimura, K. Nishino, M. Mino, M. Kitagawa, M. Matsuoka, and H. Otsuka,
Blood levels and urinary excretion of thiamin and riboflavin during oral administration of
multivitamin tablets to healthy adults, J. Nutr. Sci. Vitaminol. Special No.:438 (1992).
119. W. Djoenaidi, H. Notermanus, and G. Dunca, Beriberi cardiomyopathy, Eu. J. Clin. Nutr.
46:227 (1992).
120. F. L. Iber, J. P. Blass, M. Brain, and C. M. Leevy, Thiamin in the elderly-relation to alcoholism
and neurological degenerative disease, Am. J. Clin. Nutr. 6:1067 (1982).
121. H. E. Sauberlich, J. H. Skala, and R. P. Dowdy, Laboratory Tests for the Assessment of
Nutritional Status, CRC Press, Cleveland, 1974, p. 22.
122. H. E. Sauberlich, Newer laboratory methods for assessing nutriture of selected B-complex
vitamins, Annu. Rev. Nutr. 4:377 (1984).
123. L. G. Warnock, C. R. Prudhomme, and C. Wagner, Determination of thiamin pyrophosphate
in blood and other tissues, and its correlation with erythrocyte transketolase activity, J. Nutr.
108:421 (1978).
124. H. Baker, O. Frank, J. J. Fennelly, and C. M. Leevy, A method for assessing thiamine status
in man and animals, Am. J. Clin. Nutr. 14:197 (1964).
125. H. Baker and O. Frank, Clinical Vitaminology, Interscience, New York, 1968, p. 7.
126. L. G. Warnock, The measurement of erythrocyte thiamin pyrophosphate by high-performance
liquid chromatography, Anal. Biochem. 126:394 (1982).
127. F. Fidanza, M. S. Simonetti, A. Floridi, M. Coddini, and R. Fidanza, Comparison of methods
for thiamin and riboflavin nutriture in man, Int. J. Vitam. Nutr. Res. 59:40 (1989).
128. M. Kimura, T. Fujita, S. Nishida, and Y. Itokawa, Differential fluorometric determination
of picogram levels of thiamine, thiamine monophosphate, diphosphate, and triphosphate using
high-performance liquid chromatography, J. Chromatogr. 188:417 (1980).
129. J. S. Powers, J. Zimmer, K. Meurer, E. Manske, J. C. Collins, and H. L. Greene, Direct assay
of vitamins B1, B2 and B6 in hospitalized patients: relationship to level of intake, J. Parent.
Int. Nutr. 7:315 (1993).
130. M. K. Horwitt and O. Kreiser, The determination of early thiamine-deficient state by estima316
tion of blood lactic and pyruvic acids after glucose administration and exercise, J. Nutr. 37:
411 (1949).
131. M. Brin, Thiamine deficiency and erythrocyte metabolism, Am. J. Clin. Nutr. 12:107 (1963).
132. P. M. Dreyfus, Application of blood transketolase determinations, N. Engl. J. Med. 267:596
133. L. G. Warnock, A new approach to erythrocyte transketolase measurement, J. Nutr. 100:
1057 (1978).
134. H. J. Smeets, H. Muller, and J. A. de Weal, NADH dependent transketolase assay in erythrocyte
hemolysates, Clin. Chim. Acta 33:379 (1971).
135. P. P. Waring, D. Fisher, J. McDonnell, E. L. McGown, and H. E. Sauberlich, A continuous-
flow (Auto Analyzer II) procedure for measuring erythrocyte transketolase activity, Clin.
Chem. 28:2206 (1982).
136. M. Brin, Functional evaluation of nutritional status: thiamin, in Newer Methods of Nutritional
Biochemistry, Vol. 3 (A. A. Albanese, ed.), Academic Press, New York, 1967, p. 407.
137. M. Brin, Erythrocyte transketolase in early thiamine deficiency, Ann. N.Y. Acad. Sci. 98:528
138. M. S. Bamji, Transketolase activity and urinary excretion of thiamin in the assessment of
thiamin-nutrition status of Indians, Am. J. Clin. Nutr. 23:52 (1970).
139. B. Kjosen and S. H. Seim, The transketolase assay of thiamine in some diseases, Am. J.
Clin. Nutr. 30:1591 (1977).
140. M. Kuriyama, A. Mizuma, R. Yokomine, A. Igata, and Y. Otuji, Erythrocyte transketolase
activity in uremia, Clin. Chim. Acta 108:169 (1980).
141. T. Markkanen, Transketolase activity of red blood cells in conditions of haematological interest,
Acta Haematol. 39:321 (1968).
142. D. G. Wells and V. Marks, Anaemia and erythrocyte transketolase activity, Acta Haematol.
47:217 (1972).
143. C. H. Cheng, M. Koch, and R. E. Shank, Leukocyte transketolase activity as an indicator
of thiamin nutriture in rats, J. Nutr. 106:1678 (1976).
144. A. Konttinen and M. Vitherkoski, Blood transketolase and erythrocyte glucose-6-phosphate
dehydrogenase activities in thyrotoxicosis, Clin. Chim. Acta 22:145 (1968).
145. V. Tanphaichitr, N. Lerdvuthisopon, and P. Boongird, Clinical and nutritional studies in adult
beriberi, in Abstracts of the 4th Asian and Oceanean Congress of Neurology, Sompong Press,
Bangkok, 1975, p. 151.
146. P. M. Dreyfus and M. Seyal, Diet and nutrition in neurologic disorders, in Modern Nutrition
in Health and Disease, Vol. 2, 8th ed. (M. E. Shils, J. A. Olson, and M. Shike, eds.), Lea &
Febiger, Philadelphia, 1994, p. 1349.
147. K. Bartlett, Vitamin-responsive inborn errors of metabolism, Adv. Clin. Chem. 23:141
Pantothenic Acid
University of Minnesota, St. Paul, Minnesota
In 1933, R. J. Williams gave the name pantothenic acid to a substance that he proved
necessary for the growth of the yeast Saccharomyces cerevisiae (1). Independently, a
factor was isolated from liver that effectively cured dermatitis in chickens caused by a
deficiency in B vitamins (2). Pantothenic acid was biochemically separated from pyridoxine
(vitamin B6), a copurifying component of the vitamin B2 complex, by its lack of adsorption
to Fuller’s earth (3). Once separated, these two vitamins were shown to reverse different
animal B2 complex deficiency disorders, with pantothenic acid curing chicken
dermatitis and vitamin B6 curing dermatitis in rats.
In 1940, pantothenic acid was successfully synthesized by R. J. Williams and R. T.
Major (4). However, its biologically functional form was not discovered until 1947, when
F. Lipmann and his colleagues (5) demonstrated that pantothenic acid was contained in
coenzyme A (CoA), an essential cofactor for such acetylation reactions as that of sulfonamide
in the liver and choline in the brain. The biochemical structure of CoA was published
in 1953 (6); and the central roles of CoA in the mitochondrial tricarboxylic acid
cycle, fatty acid synthesis and degradation and other metabolic processes were elucidated.
Within the past 15 years, there has been increasing evidence for a novel role of CoA in
donating acetyl and acyl groups to many cellular proteins. This chapter will discuss the
newly discovered role of CoA in the modification of proteins, a role that may ultimately
be linked to specific effects of pantothenate deficiency.
Microorganisms synthesize pantothenic acid by joining pantoic acid and ?-alanine in
amide linkage (7). The main route for human biosynthesis of CoA (7) is through phos-
318 Plesofsky
phorylation of pantothenic acid to pantothenic acid 4?-phosphate, whose condensation
with cysteine in an ATP-dependent reaction forms 4?-phosphopantothenoylcysteine.
Decarboxylation yields 4?-phosphopantetheine, whose metabolite pantetheine is an essential
growth factor for the yogurt-producing bacterium Lactobacillus bulgaricus. The anhydride
addition of adenosine 5?-monophosphate to 4?-phosphopantetheine, followed by
phosphorylation of the ribose 3?-hydroxyl, produces CoA (Fig. 1). The CoA sulfhydryl
group, which is derived from cysteine, is the active site of esterification to acetate or acyl
groups. As an alternative to being attached to diphosphoadenosine in CoA, 4?-phosphopantetheine
may be covalently linked to a protein (8). Proteins that participate in fatty acid
metabolism, such as the acyl carrier protein of bacteria and mitochondria and the eukary-
Fig. 1 Coenzyme A and intermediates.
Pantothenic Acid 319
otic fatty acid synthetase, are linked to 4?-phosphopantetheine by a phosphodiester bond.
Citrate lyase of anaerobic bacteria and enzymes that are involved in the nonribosomal
synthesis of peptide antibiotics, such as tyrocidin and gramicidin S, are also linked to
The two types of chemical reactions in which CoA and other pantothenate-containing
molecules participate are acyl group transfer and condensation (9). In acyl group
transfer, there is nucleophilic addition to the carbonyl group thioesterified to CoA, followed
by new ester bond formation and displacement of CoA. Condensation reactions
involve acidification of the ? carbon of the acyl group thioesterified to CoA and attachment
of the ? carbon to an electrophilic center, leading to carbon–carbon bond formation or
cleavage. The first step in the tricarboxylic acid cycle of respiratory metabolism is the
condensation of acetyl-CoA with oxaloacetic acid to yield citric acid. Fatty acid synthesis
involves both types of reactions (10). Malonate is transferred from CoA to the enzymelinked
phosphopantetheine, and acetate is transferred to an enzyme sulfhydryl group. The
introduced acetate or the growing fatty acid chain condenses with the pantetheine-linked
Since pantothenic acid is essential to all forms of life, it is widely distributed in nature
and readily available in food sources. Foods that are particularly rich in pantothenic acid
(11) are liver, kidney, yeast, egg yolk, and broccoli, which contain at least 50 µg of the
vitamin per gram dry weight. Extremely high levels of pantothenate are found in royal
bee jelly (511 µg/g) and ovaries of tuna and cod (2.32 mg/g). The pantothenate content
of human milk increases fivefold within the first 4 days after parturition, from 2.2 to 11.2
µmol/L (48 to 245 µg/dL), a level similar to that found in cow’s milk. Although pantothenic
acid is relatively stable at neutral pH, cooking is reported to destroy 15–50% of
the vitamin in meat, and vegetable processing is associated with pantothenate losses of
37–78% (12). Multivitamin preparations commonly contain the alcohol derivative of pantothenic
acid, panthenol, because it is more stable than pantothenate to which it is converted
by humans (13). Calcium and sodium salts of d-pantothenate are also available as
vitamin supplements.
Formal recommended daily allowances (RDAs) have not been established for pantothenic
acid, but the recommended daily intake is 18–32 µmol (4–7 mg) for adults. Among
younger age groups, 9 µmol (2 mg) pantothenate daily is recommended for infants and
18–23 µmol (4–5 mg) for children 7–10 years of age (14). The pantothenate content of
the average American diet, estimated at 5.8 mg/day (15), is consistent with these recommendations
(12). Even when the intake of pantothenate was less than 4 mg/day, as indicated
by a study of adolescents (16), blood concentrations of the vitamin were in the
normal range (0.91–2.74 µmol/L). The excess appeared to be excreted, since pantothenate
excretion was highly correlated with dietary intake.
CoA that is ingested from dietary sources is hydrolyzed in the intestinal lumen to pantothenic
acid, which is absorbed into the bloodstream by a sodium-dependent transport
mechanism (17). After being circulated in the plasma, pantothenate is taken up into most
cells by cotransport with sodium ions; sodium cotransport is also responsible for placental
320 Plesofsky
absorption of pantothenate from maternal circulation (18). However, the level of CoA
synthesized in tissues does not reflect the amount of pantothenate available (12). Phosphorylation,
the first step in the conversion of pantothenic acid to CoA, which is catalyzed by
pantothenate kinase, is the primary regulatory site of CoA synthesis in bacteria and rat
hearts (19). Feedback inhibition of pantothenate kinase by the products CoA and acyl-
CoA was found to be relieved by carnitine, the carrier of fatty acids into mitochondria (20).
All of the enzymes required for CoA synthesis are located in the cytoplasm. Nevertheless,
mitochondria must also be a final site for CoA synthesis, since 95% of CoA is found in
mitochondria and CoA itself does not cross the mitochondrial membranes (19). CoA is
ultimately hydrolyzed to pantothenate in multiple steps, with the final, unique step being
the hydrolysis of pantetheine to cysteamine and pantothenate, which is excreted in the
urine (21).
A. Metabolism and Synthesis of Biological Molecules
CoA plays a central role (9) in the energy-yielding oxidation of glycolytic products and
other metabolites through the tricarboxylic acid cycle. As acetyl-CoA condenses with
oxaloacetate, in the first step of the cycle, to yield citrate, and as succinyl-CoA it provides
the energy for substrate level phosphorylation of guanosine diphosphate (GDP). Furthermore,
the synthesis of many essential molecules depends on CoA, as does their degradation.
Fatty acids and membrane phospholipids, including the regulatory sphingolipids,
require CoA for synthesis, and the synthesis of amino acids such as leucine, arginine, and
methionine includes a pantothenate-dependent step. CoA participates, as well, in the ?-
oxidation of fatty acids and the oxidative degradation of amino acids, making the catabolic
products available to the respiratory tricarboxylic acid cycle. Pantothenic acid, incorporated
in CoA, is also required for synthesis of isoprenoid-derived compounds, such as
cholesterol, steroid hormones, dolichol, vitamin A, vitamin D, and heme A. Through succinyl-
CoA, pantothenate is essential to the synthesis of ?-aminolevulinic acid, which is a
precursor of the corrin ring in vitamin B12 and the porphyrin rings in hemoglobin and the
cytochromes. CoA donates an essential acetyl group to the neurotransmitter acetylcholine,
as well as to the sugars N-acetylglucosamine, N-acetylgalactosamine, and N-acetylneuraminic
acid, components of glycoproteins and glycolipids.
B. N-Terminal Acetylation of Proteins
The exposed N-terminal amino acid of 50–90% of soluble eukaryotic proteins is modified
with acetate donated by CoA (22). Usually the terminal methionine is cleaved and the
second amino acid, typically alanine or serine, is cotranslationally acetylated on its ?-
amino group. In experiments suggesting that this acetylation may protect proteins from
proteolytic degradation, certain acetylated proteins were resistant to ubiquitin-dependent
proteolysis in vitro, whereas their unacetylated counterparts were degraded (23). More
sophisticated experiments, however, resulted in the proteolysis of acetylated proteins as
well. An additional factor found to be required for degradation of N-terminally acetylated
proteins was the protein synthesis elongation factor EF-1? (24). Despite its uncertain
function, the importance of N-terminal acetylation was demonstrated genetically in Saccharomyces
cerevisiae, where cell cycle progression and sexual development were disPantothenic
Acid 321
rupted in a strain defective in N-terminal acetyltransferase activity (25). The unacetylated
proteins in this mutant strain did not evince destabilization due to lack of acetylation.
Acetylation alters the structure of certain proteins. For example, it increases the N-terminal
?-helical content of calpactin I, a calcium-binding protein, which requires acetylation to
assemble with its regulatory subunit (26).
A unique type of N-terminal acetylation occurs during processing of certain mammalian
peptide hormones from their polyprotein precursors, and this acetylation strongly
affects hormone activity. Both adrenocorticotropic hormone (ACTH), a steroidogenic hormone,
and ?-lipotropin, a lipolytic hormone, are processed from the common precursor
pro-opiomelanocortin. ACTH is processed, in turn, to ?-melanocyte-stimulating hormone
(MSH), and ?-lipotropin is processed to the opioid ?-endorphin (27). Both ?-MSH and
?-endorphin become acetylated, but they differ in terms of the tissues in which they are
acetylated and the effect of acetylation on activity. Both become N-terminally acetylated
in the intermediate pituitary, but only MSH is acetylated in brain, and neither hormone
is acetylated in the anterior pituitary (8). Acetylation stimulates the activity of ?-MSH
(28) but inactivates ?-endorphin by inhibiting its binding to opioid receptors (29). A single
N-acetyltransferase activity, identified in the intermediate but not the anterior lobe of rat
and bovine pituitary, may be responsible for modifying both hormones (30), providing a
mechanism for differentially activating two distinct products of a single precursor.
C. Internal Acetylation of Proteins
1. Histones
The histones and ?-tubulin are two major groups of proteins that undergo selective, reversible
acetylation on the -amino group of internal lysine residues. Other DNA-binding
proteins, including the high-mobility group proteins HMG1 and HMG2 and protamines,
are also subject to acetylation. Within the nucleosome are a tetramer of histones H3 and
H4 and two dimers of histones H2A and H2B, around which is wrapped 146 bp of DNA.
These core histones of the octamer become acetylated within their basic amino terminal
regions, with H3 and H4 each having four possible acetylation sites. A study of preferred
acetylation sites in mammalian histones (31) showed that in H3, lysine 14 was acetylated
first, followed by lysine 23 and lysine 18. Monoacetylated H4 was modified exclusively
at lysine 16, with further additions progressing from lysine 12 to lysine 8 and lysine 5.
In histone H2B, lysine 12 and lysine 15 were the preferred acetylation sites over lysine
5 and lysine 20. Histone H2A has only one acetylation site.
Histones that become highly acetylated are associated chiefly with DNA that is
either newly replicated or contains genes that are actively transcribed (8,32). Different
acetylation sites may be utilized in these two processes. In the ciliated protozoan Tetrahymena
thermophila, where these activities are easily distinguished, histone H4 becomes
diacetylated on lysines 4 and 7 in transcriptionally active chromatin, whereas newly assembled
chromatin contains H4 that is diacetylated on lysines 4 and 11 (33). In Physarum
polycephalum, replicating DNA has acetylation in all four core histones, whereas only
histones H3 and H4 have high levels of acetylation in actively transcribed genes (34).
The temporal patterns of acetylation also differ. Newly synthesized histones are only transiently
acetylated as they assemble with replicating chromatin, whereas histones in transcriptionally
active chromatin are continually and dynamically acetylated.
The presence of highly acetylated H3 and H4 noticeably affects nucleosomal structure,
in part through charge neutralization of the modified lysines. The more open, unfolded
322 Plesofsky
configuration of acetylated chromatin is indicated by its increased sensitivity to nucleases
and its increased salt solubility (35). Agarose gel electrophoresis also showed that histone
acetylation leads to decompaction of the chromatin fiber (36). Hyperacetylation of H3 and
H4 has been demonstrated to lead to a reduction in the linking number change per nucleosome,
indicating a decrease in the negative supercoiling within nucleosomes (37). For the
H2A/H2B dimer, acetylation makes its nucleosomal association particularly labile, which
may facilitate its exchange with newly synthesized histones (38).
Acetylated histones are enriched in genes that are being actively transcribed. The
acetylated chromatin fraction from chick embryo erythrocytes, selected by an antibody
against the acetylated lysine of histones, was strongly enriched in actively transcribed ?-
d-globin gene sequences (39). However, histone acetylation does not correspond exactly
to the transcribed sequences, either spatially or temporally. Acetylation extends into the
interspersed, nontranscribed DNA of the ?-globin locus, as well as being present in the
transcribed region (40). Furthermore, acetylation does not respond to the switch in globin
gene transcription that occurs between 5 and 15 days of embryo development, since acetylated
histones were associated with both globin genes at both stages of development
(39). Similarly, acetylated histones were associated with the gene for the platelet-derived
growth factor B chain before its induction by phorbol esters (41). Histone acetylation,
therefore, may be a precondition for transcription, rather than specifying expression of
particular genes. In vitro experiments indicate that histone acetylation is required for specific
proteins to bind to chromatin. For example, acetylation was required for transcription
factor IIIA (TFIIIA) binding to the 5S RNA gene (42). Acetylation of core histones also
indirectly facilitated the binding of transcription factors USF and GAL4-AH to nucleosomes,
by inhibiting the binding of the internucleosomal histone H1, which represses
transcription factor binding (43).
The chromosomal distribution of acetylated histone H4 was observed in the polytene
chromosomes of Drosophila larvae by immunofluorescence, using antibodies developed
against specific, acetylated lysines (44). The single X chromosome in male larvae,
which is very actively transcribed, was the chief location of H4 acetylated at lysine 16,
an acetylated form that was not observed in male autosomes or in any chromosomes
of female larvae. Immunolabeling of mammalian chromosomes showed that acetylated
H4 was concentrated in chromosomal regions enriched in coding DNA, but acetylated
H4 was largely absent from the inactive X chromosome of female mammalian cells; the
few immunolabeled bands corresponded mainly to regions containing known expressed
genes (45).
That acetylation of histone H4 affects the transcription of certain genes was demonstrated
in S. cerevisiae, where mutations were made in H4 that either delete the N-terminal
tail or substitute its lysine residues. Deletion of the basic tail or an uncharged substitution
of lysine 16 led to derepression of the silent mating locus, suggesting that the chargeneutralizing
acetylation of lysine 16 may be responsible for activating transcription (46).
In contrast, transcription of the GAL1 and PHO5 genes was inhibited by similar mutations
of H4, in which multiple lysines were substituted (47). However, a neutralizing mutation
was still much less inhibitory to transcription than an arginine substitution. Mutations of
the N-terminal sequences of the other three core histones had no or little effect on transcription
of these genes or the mating type locus.
Unlike gene transcription, the replication of DNA does not require histone acetylation
for polymerase enzymes to gain access to the nucleosomal DNA. However, newly
synthesized histones that assemble with the replicated DNA become transiently acetylated
Pantothenic Acid 323
on their N-terminal tails, a modification that is detected in the presence of the deacetylase
inhibitor butyrate. Histones are deposited onto DNA in a two-step process that begins
with the deposition of H3/H4 tetramers, followed by H2A/H2B dimers. Diacetylation of
H4 precedes its deposition onto DNA and its assembly with newly synthesized H3, H2A,
and H2B (48). Histone H2B is also highly acetylated in newly assembled nucleosomes
(49). In the absence of DNA synthesis, newly synthesized H2A and H2B exchange preferentially
into acetylated chromatin regions, assembling with acetylated H3 and H4, but
newly synthesized H3 and H4 remain cytosolic (48). Deposition-related histone acetylation
inhibits the binding of histone H1 to newly assembled chromatin, thereby reducing
higher order nucleosome interactions (50).
Other chromatin-associated molecules that undergo acetylation include HMG1 and
HMG2, which are acetylated by the same N-acetyltransferase as the histones (51). It is
apparently the acetylated form of HMG1 that serves as a histone assembly factor and
stimulates DNA polymerase activity (52). The same acetyltransferase also acetylates polyamines,
which interact with chromatin during DNA synthesis (51).
2. ?-Tubulin
A subunit of microtubules is the other class of proteins that becomes acetylated on the
?-amino group of lysine residues. Microtubules are essential structural components of
eukaryotic cells that are central to chromosome segregation in nuclei and to the cytoplasmic
cytoskeleton, where they affect cell shape, motility, and organelle movement.
They also constitute the major component of flagella and cilia. Microtubles are assembled
from ?-tubulin/?-tubulin dimers that polymerize and depolymerize dynamically. A subset
of microtubules has been found to contain ?-tubulin that is acetylated on lysine 40 (53).
Acetylation occurs on the ?-tubulin after it is incorporated into the microtubule, which
is a better substrate than the tubulin dimer for the isolated acetylase activity (54), and
deacetylation appears to be coupled to microtubule depolymerization. The acetylated microtubules
are more stable to depolymerizing agents, such as colchicine, than the unacetylated
microtubules; in turn, drugs that stabilize microtubules, such as taxol, induce ?-
tubulin acetylation (55). Tubulin turnover in neurons was found to be much slower in
areas where acetylated ?-tubulin was concentrated, such as the neurite shafts, than in the
cell body and growth cones (56). These observations suggest that acetylation may help
to stabilize microtubules.
The distribution of acetylated ?-tubulin in various types of cells has been detected
by an antibody specific for the modified tubulin and has been found to be nonrandom. In
cerebellar neurons, which have abundant microtubules, acetylated ?-tubulin was found to
be concentrated in axons, relative to dendrites, during early neuronal development (57),
although in mature neurons it was present also in thick dendritic trunks. It was excluded
from growth cones at neurite tips in dorsal root ganglion neurons, and acetylated ?-tubulin
was also absent from leading edge microtubules of migrating 3T3 cells. Acetylated microtubules
underlie motor endplates in chick muscle fibers and they contribute to vesicle and
organelle transport, whereas growing or motile regions of the cells appear to be depleted
of acetylated microtubules (58). Individual axonal microtubules actually contain separate
regions of acetylated and unacetylated ?-tubulin along their length, with the older regions
corresponding to the acetylated domains (59). It was suggested that the acetylation of ?-
tubulin may be a step in cell commitment. When fibroblast cells were induced to become
neuron-like, they formed an acetylated microtubule bundle that subsequently extended
into a neurite (60).
324 Plesofsky
Cells undergoing mitosis and meiosis also show nonrandom, varying distribution
of acetylated microtubules. During arrest in meiotic metaphase, acetylated ?-tubulin was
found predominantly at the poles of unfertilized mouse oocytes (61). At meiotic anaphase,
the spindle became labeled, but by telophase only the meiotic midbody microtubules were
acetylated. Changes in the distribution of acetylated microtubules were also described for
the preimplantation development of mouse embryo cells (62). Acetylated microtubules
were generally associated with the cell cortex before differentiation. As cellular asymmetry
developed, the acetylated microtubule subpopulation relocated to the basal part of the cell
cortex and, after asymmetrical cell division, became concentrated in the inside cells rather
than the outside cells.
D. Acylation of Proteins
During the past 15 years it has been discovered that a wide variety of cellular proteins
are modified with long-chain fatty acids, donated by CoA (8). This type of covalent modi-
fication affects the location and activity of many proteins (Table 1), including those that
Table 1 Proteinsa That Are Modified with Fatty Acids
Palmitate Myristate myristate
Rasv, Rasc Transducin Go, Gi protein,
Gs protein ? subunit ARF proteins ? subunit
Receptors: Srcv, Srcc p56lck, p59fyn
Rhodopsin Protein kinase A Insulin receptor
Dopamine (D1, D2) Calcineurin ecNOS
?2A-, ?2-Adrenergic Recoverin
Thyrotropin-releasing hormone MARCKS
Choriogonadotropin Band 4.2
Iron-transferrin IgM heavy chain
Nicotinic acetylcholine Interleukin-1?, 1?
CD4, CD44, CD36, P-selectin TNF-?
HLA-B heavy chain NADH-cytochrome b5 reductase
Band 3 NADH dehydrogenase, B18
Ankyrin cytochrome c oxidase, su 1
Gap junction proteins
Myelin proteolipid subunit
Cysteine string proteins
Glutamic acid decarboxylase65
GAP-43, SNAP-25
Methylmalonate semialdehyde
Aldehyde and glutamate dehydrogenases
a Discussed in text.
Pantothenic Acid 325
have a central role in signal transduction. The two fatty acids that are usually added to
proteins are myristic acid, a rare 14-carbon saturated fatty acid, and the more common
16-carbon palmitic acid. These two fatty acids are attached to proteins by distinct mechanisms,
and they affect proteins differently. The myristoylation that has been best characterized
occurs cotranslationally, by amide linkage to the ?-amino group of an N-terminal
glycine, and it is irreversible. Palmitate, in contrast, is added posttranslationally and forms
a reversible ester bond with a cysteine or serine. The greater hydrophobicity of palmitate
results in strong membrane association of the acylated protein, whereas myristoylation
leads either to weak membrane association or interactions with proteins. The reversibility
of palmitoylation allows this modification to be regulatory.
There are stringent sequence requirements for a protein to become N-terminally
myristoylated (63). An N-terminal glycine, originally in the second position, is absolutely
required, and there is preference for a small neutral subterminal residue and a serine in
the sixth position. The myristoyltransferase that has been isolated shows a specific preference
for myristoyl-CoA over other acyl-CoA donors (63). The N-terminal myristoylation
was shown to be essential for viability in S. cerevisiae, where gene disruption of N-myristoyltransferase
proved to be recessively lethal (64). In contrast, enzymes that transfer palmitate
show little substrate specificity, either for the fatty acyl group transferred from
CoA or the peptide sequence required for protein acylation.
1. GTP-Binding Proteins
The small GTP-binding proteins comprise an extensive group of influential cellular
proteins that become acylated with myristic acid and/or palmitic acid. This group of
nucleotide-binding proteins includes the monomeric Ras superfamily of proteins, heterotrimeric
G proteins, and ADP-ribosylating factors (ARFs). Several of these proteins are also
modified by isoprenoid groups whose synthesis requires CoA, although it is not the lipid
Covalent modification of the Ras proteins, both cellular and viral, has been extensively
studied. Most Ras proteins undergo two types of modifications at their carboxyl
end, isoprenylation and methylation on the processed terminal cysteine and palmitoylation
of a nearby cysteine; in some proteins a basic region functionally replaces palmitoylation
(65). The viral Ras proteins, such as those of Harvey murine and Kirsten sarcoma viruses,
are deficient in GTP hydrolysis and are oncogenic. Both isoprenylation and palmitoylation
are required for viral Ras proteins to bind to the plasma membrane and transform cells,
with palmitoylation strengthening the weak activity conferred by isoprenylation (65). A
15-carbon farnesyl group modifies Ras proteins, and farnesyltransferase inhibitors are effective
antitumor agents in Ras-transformed cells. However, the Ras-related proteins and
most cellular proteins are isoprenylated with the 20-carbon geranylgeranyl (66). The cellular
Ras protein is a critical component of the mitogen-activated protein (MAP) signaling
kinase cascade, serving to direct the protein kinase Raf to the plasma membrane. When
Raf itself was experimentally engineered to contain a farnesylation signal, Ras proved to
be unnecessary for signal transduction (67).
The mammalian Rab proteins are Ras-related proteins that are proposed to regulate
specific stages of vesicular transport to the cell surface (68). These proteins appear to target
vesicles to specific acceptor membranes by binding to the vesicles until their delivery to
membranes. The cycling of Rab proteins between membranes and cytosol is related to
the membrane budding and fusion that drives vesicular movement. Rab proteins are modi-
fied at their C termini with geranylgeranyl, which is required for their association with
326 Plesofsky
organellar membranes, as well as for interaction of a Rab protein with its GDP dissociation
inhibitor (GDI), which releases the Rab protein from membranes (69). A defect in a protein
with sequence homology to GDIs has been implicated in retinal degradation or choroideremia
in humans (66).
The ? subunits of several heterotrimeric G proteins, which mediate transmembrane
signaling, are esterified to palmitate near their N termini on a third-position cysteine. The
Gs? subunit, which mediates hormonal stimulation of cAMP synthesis, requires palmitoylation
both for binding to membranes and for stimulation of adenylyl cyclase (70). Hormonal
activation leads to GTP binding by ?s and to its rapid depalmitoylation, which is
accompanied by relocalization of ?s to the cytosol. In addition to palmitate, myristic acid
is added to the N terminus of those G-protein ? subunits that are members of the Gi and
Go subfamilies. Prior myristoylation increases these proteins’ membrane affinity, which
is required for the addition of palmitate (71). Both types of acylation were found to contribute
to the membrane affinity of the ?o subunit and its binding to the ?? subunits. Mutation
of the palmitoylated cysteine of Go1? disrupted its exclusive localization to membranes,
even in the presence of N-myristoylation, whereas mutation of the myristoylated glycine,
preventing both fatty acylations, caused ?o to become chiefly cytosolic (72). It has been
suggested that dual acylation of these G-protein ? subunits and certain Src-related proteins
(discussed below) directs their localization to plasma membrane caveolae, possibly via
association with glycosylphosphatidylinositol (GPI)-anchored membrane proteins (72). In
addition to myristate, heterogeneous fatty acids, such as laurate and unsaturated 14-carbon
fatty acids, modify the N terminus of the ? subunit of transducin, the photoreceptor G
protein that is stimulated by photolyzed rhodopsin. The nature of the modifying fatty acid
appears to affect the strength of the interaction of transducin ? with ?? and may influence
the speed of visual excitation (73). Since the ? subunits of G proteins are isoprenylated
at their C termini (66), increasing their membrane affinity, these heterotrimeric G proteins,
which play a central role in cell signaling, undergo all three types of modifications with
hydrophobic moieties.
The small GTP-binding proteins that are involved in vesicular transport, known as
ARFs, are N-terminally myristoylated, and this modification is essential for their functions
in transport. In mammalian cells the ARF protein is the principal coat component of non–
clathrin-coated vesicles, and its binding to the Golgi membrane appears to mark the site
of budding of a future vesicle. Whereas recombinant myristoylated ARF5 bound to Golgi
membranes in a GTP-dependent and temperature-dependent manner, the nonmyristoylated
ARF5 did not bind to membranes at all (74). The N-terminal region of ARF1 may be
stabilized by myristoylation in an amphipathic ? helix, conducive to membrane binding.
Myristoylation also apparently stabilizes the interaction of ARF1 with GTP in the presence
of membranes and enables ARF1 to release GDP at high physiological concentrations of
Mg2 (75).
In addition to the myristoylation of ARFs, reconstitution of in vitro transport through
the Golgi stacks requires palmitoyl-CoA (76). The addition of acyl-CoA was necessary
for the budding of vesicles from donor Golgi and for the fusion of these vesicles with
acceptor Golgi cisternae. A nonhydrolyzable analog of palmitoyl-CoA or an inhibitor of
the acyl-CoA synthetase inhibited vesicle budding (76). The source of this acyl group
requirement is not known.
2. Protein Kinases and Phosphatases
Like palmitoylation of Ras proteins, myristoylation of the Src protein provided the first
model of this type of acylation. Src, first identified in Rous sarcoma virus, is a tyrosine
Pantothenic Acid 327
kinase that has both cellular and viral homologues. The viral Src protein is responsible
for cell transformation by the oncogenic viruses, and its N-terminal myristoylation is necessary
both for the protein’s association with the plasma membrane and for its transforming
capability. Expression of a mutant, nonmyristoylated pp60src protein resulted in a
small group of membrane-associated proteins showing decreased tyrosine phosphorylation
(77). pp60src was also prevented from associating with membranes by addition of a less
hydrophobic analog of myristic acid, which is a possible candidate for an antitumor agent
(78). Mutation of the cellular Src protein at its N-terminal glycine to prevent myristoylation
proved inhibitory to its activity in mitosis, presumably because c-Src is activated by a
membrane-bound phosphatase (79).
Although pp60src itself is not additionally acylated with palmitate, several Src-related
tyrosine kinases, such as p56lck and p59fyn, are dually acylated at their N termini (80).
These kinases associate with proteins that are anchored in the outer plasma membrane by
glycosylphosphatidylinositol modification, and dual acylation of the kinases is necessary
for this association. When the palmitoylated cysteines at positions 3 and 6 in p59fyn and
at positions 3 and 5 in p56lck were mutated to serines, these two kinases no longer associated
with the GPI-anchored protein DAF (decay-accelerating factor). On the other hand,
when the reverse mutations were made in pp60src, from serines at positions 3 and 6 to
cysteines, the Src protein uncharacteristically associated with DAF (81). GPI-anchored
proteins, present in the membranes of T cells, B cells, monocytes, and granulocytes, are
thought to transduce external activation signals through their interaction with tyrosine
kinases in the inner membrane layer.
The catalytic subunit of cAMP-dependent protein kinase is also myristoylated at its
N terminus. Myristoylation does not appear to affect its association with membranes, since
the catalytic subunit is anchored in the membrane by the regulatory subunit until dissociation
by cAMP. A comparison of the properties of the myristoylated vs. nonmyristoylated
catalytic subunit indicated that acylation increased the thermal structural stability of the
kinase, possibly by stabilizing an intramolecular interaction between the N-terminal domain
and an internal hydrophobic surface (82). N-terminal myristoylation also occurs in
the B subunit of calcineurin (83), a calmodulin-dependent phosphatase that is the target
of the immunosuppressant cyclosporin A.
3. Membrane Receptors
Many membrane-anchored receptors are acylated with palmitic acid. Heterotrimeric Gprotein–
coupled receptors all have a similar domain structure, consisting of an extracellular
N terminus, seven transmembrane ? helices, and an intracellular C terminus, leading
to three intracellular and three extracellular loops. Near the seventh transmembrane region,
the C-terminal tail has a conserved cysteine whose palmitoylation has been characterized
in rhodopsin, the dopamine (D1 and D2) receptors, the ?2-adrenergic receptor, the ?2Aadrenergic
receptor, thyrotropin-releasing hormone receptor, and the choriogonadotropin
receptor. This acylation appears to lead to the formation of a fourth intracellular loop.
Fluorescence quenching studies with rhodopsin, the retinol-binding photoreceptor of retinal
rod cells, showed that two cysteines at positions 322 and 323, modified by palmitic
acid, are integrated into the cell membrane (84). Furthermore, the light excitation of rhodopsin
is regulated by recoverin, an N-myristoylated protein that responds to changing
levels of calcium. Calcium binding alters the conformation of recoverin to expose the Nmyristoyl
group, which can then bind to membranes (85).
The regulatory functions of receptor palmitoylation have been studied in the dopamine
and adrenergic receptors. The human receptor for dopamine (D1) incorporates palmi328
tate shortly after cells are exposed to hormone, concurrent with cellular desensitization
to dopamine (86). The receptor also becomes phosphorylated and dissociates from the
cell surface. By mutating the palmitoylated cysteine of the ?2-adrenergic receptor, it was
determined that acylation is important both for receptor coupling with Gs (the stimulatory
G protein) and for agonist-promoted desensitization (87). However, the corresponding
mutation in the ?2A-adrenergic receptor had a different effect. Mutation of the cysteine
abolished the downregulation of receptor number that occurs after prolonged exposure to
hormone (88). In contrast, mutational analysis showed that palmitoylation of thyrotropinreleasing
hormone and choriogonadotropin receptors slowed their internalization after prolonged
agonist exposure (89).
Other transmembrane receptors that are palmitoylated include the iron-transferrin
receptor, the insulin receptor, and the nicotinic acetylcholine receptor. In addition to palmitoylation,
the insulin receptor is also reported to be linked to myristic acid (90). Rat adipocytes
contain a membrane glycoprotein, related to CD36, that becomes palmitoylated
on an extracellular domain in response to insulin and energy depletion (91). Several cells
of blood and the immune system, such as T lymphocytes, macrophages, and platelets,
contain receptors that become palmitoylated. CD4, a surface glycoprotein of T lymphocytes
and macrophages that functions in cell adhesion and signaling during antigen recognition,
also serves as receptor for the human immunodeficiency virus (HIV). CD4 becomes
palmitoylated on two cysteines at the junction of its transmembrane and cytoplasmic domains,
but this modification does not apparently affect either membrane localization of
the receptor or its interactions with proteins such as p56lck (92). However, palmitoylation
of a lymphoma glycoprotein, closely related to CD44 of T lymphocytes, was found to be
required for its in vitro binding to ankyrin, the protein that mediates attachment of the
cell membrane to the cytoskeleton (93). The heavy chain of the human histocompatibility
antigen HLA-B is thioesterified to palmitate. An adhesion receptor in human platelets, Pselectin
(CD62), is palmitoylated on a cysteine in its cytoplasmic tail, a domain that is
responsible for the sorting of P-selectin into secretory granules and its endocytosis into
cells (94).
4. Proteins Associated with the Cytoskeleton
An N-myristoylated protein, rich in alanine, that is a major substrate of protein kinase C
(MARCKS) has been identified as a protein that cross-links actin filaments. MARCKS
is thought to be involved in the cytoskeletal rearrangements that occur during neuronal
transmitter release, leukocyte activation, and growth factor–induced mitosis. The initial
binding of MARCKS to the plasma membrane requires myristoylation, which provides
one of its two membrane binding sites. Phosphorylation within the second membrane
binding site occurs upon cell activation and results in dissociation of MARCKS from the
membrane (95).
Membrane proteins in erythrocytes are linked to the cytoskeleton by a series of
acylated proteins. The anion transport protein Band 3, which is palmitoylated (96), is
linked to Band 4.2, which is a major myristoylated protein of the membrane (97). Band
4.2, in turn, is linked to ankyrin, the cytoskeleton attachment protein, which is itself palmitoylated
(8). In addition, the cytoskeletal protein spectrin becomes palmitoylated in its ?
subunit, and this acylated subpopulation is more tightly membrane-associated than the
unmodified spectrin (98). Other cytoskeleton-associated proteins whose palmitoylation has
been reported are vinculin, fibronectin, and a subpopulation of actin. Gap junction proteins
of heart and eye lens also become palmitoylated (99).
Pantothenic Acid 329
5. Neuronal Proteins
One of the first proteins found to be covalently modified by palmitate was the proteolipid
subunit of brain myelin, in which palmitate is esterified to a threonine residue (8). Cysteine
string proteins (csp) are located in synaptic terminals and are required for neurotransmitter
release. Fatty acylation of multiple cysteines, 10 in Torpedo, appears to be a mechanism
for attachment of Tcsp to vesicles, where they interact with presynaptic calcium channels
in the plasma membrane (100). Palmitoylation of acetylcholinesterase, which degrades
the neurotransmitter acetylcholine, results in its anchorage in the cell membrane. Palmitate
attachment to an esterase subpopulation correlated with its attachment to the outer cell
surface, whereas most of the expressed esterase was unmodified and secreted from cultured
cells (101). A major isoform of glutamic acid decarboxylase (GAD65) is located in synaptic
vesicles of GABA-secreting neurons and in microvesicles of pancreatic ? cells. Palmitoylation
of GAD65 in its N-terminal domain is reported to increase its membrane affinity,
possibly regulating its membrane anchorage (102).
Protein acylation appears to play an important role during neuronal development.
The reversible palmitoylation of both GAP-43 and SNAP-25, in developing brains, was
proposed to influence growth cone motility and process outgrowth (103). GAP-43 is a
major component of the growth cone membranes of elongating axons, and SNAP-25 is
a synaptic protein involved in later stages of axon growth. Whereas the nonacylated form
of GAP-43 stimulates the G protein, Go, palmitoylation at two cysteines near its N terminus
inhibits this activity and keeps GAP-43 membrane-bound (104). Nitric oxide inhibits the
acylation of GAP-43 and SNAP-25, along with that of other neuronal proteins (103), and
it inhibits the growth of cultured neurites by causing growth cone collapse. It was suggested
that these effects of nitric oxide reflects its role in regulating process outgrowth
and remodeling in vivo (103). The endothelial form of nitric oxide synthase (ecNOS),
which is involved in smooth-muscle relaxation, is itself regulated by dual acylation (105).
N-terminal myristoylation is required for the membrane localization of ecNOS and for its
reversible modification with palmitate. Depalmitoylation follows addition of the agonist
bradykinin and results in ecNOS relocalizing to the cytosol and becoming phosphorylated.
6. Myristoylated Proteins of the Immune System
Several proteins of the immune system are modified with myristate on internal lysine
residues. These include the heavy chain of µ immunoglobulins, which becomes acylated
during transport to the surface of developing B cells (106). Precursors of the cytokines
interleukin (IL)–1? and 1? are myristoylated on lysines (107), as is the tumor necrosis
factor (TNF) ? precursor (108). After processing, the mature forms of IL-1? and IL-1? are
secreted and bind to receptors. The precursor of IL-1?, however, which is the predominant
myristoylated species, also exists as a plasma membrane–associated protein (107). The
acylated lysines in the propiece of the TNF-? precursor are near a membrane-spanning
segment and are thought to facilitate membrane insertion of the precursor, which, like the
mature TNF-?, is active in mediating inflammation (108).
7. Mitochondrial Proteins
Mitochondrial proteins involved in oxidation–reduction reactions have been reported to
become fatty acylated. The NADH–cytochrome b5 reductase, located at the outer mitochondrial
membranes of animal cells, as well as at microsomes, is N-terminally myristoylated
(109). In the inner membrane, there is N-myristoylation of subunit B18 of complex
330 Plesofsky
I of the electron transport chain, the NADH:ubiquinone oxidoreductase, which was detected
by cDNA sequencing and electrospray mass spectrometry (110). Complex I also
contains the pantothenate-containing mitochondrial acyl carrier protein, identified as an
enzyme subunit by its stoichiometry (111). The core catalytic subunit 1 of cytochrome c
oxidase, the terminal electron carrier of the respiratory chain, is modified by myristic acid
on an internal lysine residue (112). The oxidase subunit 1 is one of only a few proteins
that are encoded by mitochondrial genes and synthesized in mitochondria; its modification
suggests that there is a mitochondrial N-myristoylating activity.
Dehydrogenases involved in oxidative catabolism also become acylated, a modification
that inhibits their enzymatic activity and may be a mechanism for feedback regulation
(113). A decrease in mitochondrial energy level was found to correlate with increased
acylation of an active-site cysteine in methylmalonate semialdehyde dehydrogenase
(MMSDH). It was proposed that long-chain fatty acids accumulate when MMSDH and
other dehydrogenases utilize the NAD required to oxidize fatty acids. The accumulation
of these fatty acids, in turn, promotes acylation of the dehydrogenases, whose inhibition
allows ?-oxidation of fatty acids to proceed. Aldehyde dehydrogenase and glutamate dehydrogenase
are two other mitochondrial NAD-dependent dehydrogenases that become
acylated (113).
The human effects of pantothenate deficiency have been detected chiefly by the administration,
either intentional or unintentional, of pantothenic acid antagonists. Hopantenate is
an analog of pantothenic acid, which contains GABA in place of ?-alanine. Also an agonist
of GABA, hopantenate was used in Japan as a cerebral stimulant for retarded individuals
and to alleviate symptoms of tardive dyskinesia induced by tranquilizers. It was found,
however, that treatment with hopantenate produced severe side effects in patients, including
lactic acidosis, hypoglycemia, and hyperammonemia (114). It ultimately led to acute
encephalopathy with hepatic steatosis resembling Reye’s syndrome. A study in which
dogs were administered hopantenate over several weeks showed that they suffered similar
effects (115). This study also characterized the abnormalities that developed in hepatic
mitochondria, which became enlarged, had increased numbers of cristae, and contained
crystalloid inclusions. These mitochondrial alterations differed from those associated with
Reye’s syndrome. The side effects of hopantenate treatment are apparently caused by
induction of pantothenic acid deficiency. Dogs that were given an equivalent amount of
pantothenic acid at the same time as calcium hopantenate did not develop these disorders
When individuals who suffer seizures are treated with the anticonvulsant drug
valproic acid, liver damage may follow in association with decreased levels of CoA
and acetyl-CoA. These effects appear to be due to sequestration of CoA in esters of valproic
acid and its metabolites (20). Coadministration of pantothenic acid, carnitine, and
acetylcysteine, along with valproate or its derivative, relieved these metabolic defects in
developing mice (20). Skeletal muscles of the murine model of Duchenne’s muscular
dystrophy show reduced energy metabolism, as indicated by reduced heat production,
relative to control muscles. The addition of pantothenate resulted in increased cytoplasmic
synthesis of CoA and increased the thermogenic response to glucose in the diseased muscles
(116). Pantothenic acid also protected rats against liver damage and peroxidation
produced by carbon tetrachloride (117). Furthermore, surgical wound healing was greatly
Pantothenic Acid 331
improved by the administration of pantothenic acid, apparently due to its anti-inflammatory
properties. In vitro studies show that calcium pantothenate dampens the responses
of neutrophils (polymorphonuclear leukocytes) to activation by stimulatory peptides and
cytokines (118).
Analogs of pantothenic acid have been used experimentally to induce pantothenate
deficiency. The antagonist ?-methylpantothenate, in combination with a pantothenatedeficient
diet, produced deficiency symptoms in humans that included headache, fatigue,
insomnia, intestinal disturbances, and paresthesia of hands and feet. A decrease in the
eosinopenic response to ACTH, loss of antibody production, and increased sensitivity to
insulin were also reported (119). Naturally occurring pantothenate deficiency has been
detected only under conditions of severe malnutrition. World War II prisoners in the Philippines,
Japan, and Burma suffered from a disease in which they experienced numbness
in their toes and painful burning sensations in their feet. Pantothenic acid was specifically
required to relieve these symptoms of nutritional melalgia (120).
Many studies have documented the physical effects on animals of a pantothenaterestricted
diet (11). Rats develop hypertrophy of the adrenal cortex, which is followed by