Tuesday, July 3, 2007

Vitamin C deficiency

Defi ciency in humans
A defi ciency of vitamin C results in scurvy. Fully
developed scurvy is rarely seen nowadays, but clinical
signs of mild scurvy are found quite frequently in
alcoholics and drug addicts. The symptoms described
below have been observed in patients with scurvy
(Chazan & Mistilis, 1963) and in experimentally induced
scurvy (Hodges et al., 1971).
Early symptoms in adults are weakness, easy fatigue
and listlessness, followed by shortness of breath and
aching bones, joints and muscles. Progressive changes
in the skin then appear after about 4 months of complete
vitamin C deprivation. A horny material piles
up around the openings of hair follicles, and the hair
becomes fragmented and coiled. Red spots of pinpoint
to pinhead size caused by the rupture of small
blood vessels appear fi rst on the feet and ankles, and
then spread upwards. Thereafter, bruises appear over
large areas of skin, particularly on the legs. Bruising
is the manifestation of haemorrhages in subcutaneous
tissue, beneath the periosteum of bones and in
the synovia of joints. The gums become swollen and
bleeding, especially where there is advanced dental
caries. Haemorrhages are caused by rupture of capillaries,
which are fragile because of impaired ascorbic
acid-dependent synthesis of vascular basement
membrane (Priest, 1970). Wounds fail to heal and old
wounds reopen. The sufferer is visibly anaemic due in
part to haemolysis caused by peroxidative damage to
the erythrocyte plasma membrane (Goldberg, 1963).
Vitamin C defi ciency in adults may cause osteoporosis
due to a diminished production of organic matrix
in bones. The corresponding symptoms in infantile
scurvy are impaired ossifi cation and bone growth.
Kinsman & Hood (1971) studied the psychological
aspects of vitamin C defi ciency in healthy volunteers.
They measured four behavioural areas: physical fi tness
(strength, coordination and balance), mental
functions (memory, vigilance and problem solving),
psychomotor performance tasks (reaction time,
manipulative skills and hand–arm steadiness), and
personality. Three areas of change associated with
vitamin C deficiency were found: physical fitness involving
bending or twisting of the legs, psychomotor
tasks, and measures of personality. The changes in the
physical fitness could be accounted for by the pronounced
joint pain in the legs that occurred during
the deficiency period. The decrements in psychomotor
performance were attributed to a reduced motivational
level. The personality changes corresponded
to the classical ‘neurotic triad’ of the Minnesota Multiphasic
Personality Inventory, i.e. hypochondriasis,
depression and hysteria. Elevation of this triad is also
found in prolonged semi-starvation and defi ciencies
of B-complex vitamins.
In fully developed scurvy, as witnessed and recorded
at sea by James Lind in 1752, the body is covered with
spots and bruises, and the skin overlying the joints
becomes discoloured from the haemolysed blood in
and around them. There may be bleeding into the
peritoneal cavity and pericardial sac as well as into
joints. The gums become swollen, spongy and of a
livid blue-red colour. The swelling can develop to such
an extent that the gum tissue completely encases and
hides the teeth. The spongy gums bleed on the slightest
touch and become secondarily infected, leading to
loosening of the teeth and gangrene. Death preceded
by dyspnoea, cyanosis and convulsions is inevitable in
the continuing absence of vitamin C.
19.12.2 Rebound scurvy
Theoretically, the absorption of ascorbic acid could be
impaired on resumption of normal vitamin C inputs
following mega-dosing (>1 g per day), because of insuffi
cient carriers in the enterocyte cell membranes.
Based on experiments with guinea pigs, it is considered
likely that, in humans, renewed synthesis of
carriers will take place well before the onset of scurvy.
During mega-dosing, reduced ascorbate absorption is
accompanied by increased rates of ascorbate catabolism.
In adult guinea pigs, the accelerated catabolism
is not reversible after more than 2 months on subnormal
uptake of ascorbate (Sorensen et al., 1974). Guinea
pigs are thus susceptible to a systemic conditioning
effect known as rebound scurvy, caused by an induction
of ascorbic acid-metabolizing enzymes by high
dietary vitamin C. The body stores of vitamin C are
depleted more rapidly in juvenile guinea pigs than in
adults, increasing the likelihood of rebound scurvy in
juveniles. Solid evidence for the existence of rebound
scurvy in humans is tenuous (Gerster & Moser, 1988),
and reports by Schrauzer & Rhead (1973) and Siegel
et al. (1982) describe only single cases.

Vitamin C and cardiovascular disease

Ascorbic acid through its numerous metabolic and
antioxidant effects may inhibit some of the steps
involved in atherosclerosis and thrombosis, thus
reducing the risk of cardiovascular disease. In a casecontrol
study, Ramirez & Flowers (1980) reported
signifi cantly lower (p < 0.001) leucocyte vitamin C
levels in 101 cases of angiographically documented
cardiovascular disease.
19.11.1 Cholesterol metabolism
Studies of animals that either synthesize (rat, rabbit)
or do not synthesize (guinea pig, monkey) vitamin
C have shown that vitamin C is intimately involved
in cholesterol metabolism. Guinea pigs subjected to
chronic vitamin C defi ciency exhibit increased cholesterol
levels in blood plasma and liver due to slower
conversion of cholesterol to bile acids (Ginter et al.,
1971; Ginter, 1973). The impaired conversion results
from a decreased activity of the rate-limiting liver enzyme
cholesterol 7α-hydroxylase (Horio et al., 1989).
When guinea pigs, rats and rabbits are rendered
hypercholesterolaemic by feeding a high-cholesterol
diet, vitamin C supplementation lowers their blood
cholesterol levels.
19.11.2 Lipoprotein profi le
Diets low in vitamin C lead to a redistribution of
cholesterol among the various plasma lipoproteins.
Vitamin C defi ciency in ODS rats (rats with an hereditary
inability to synthesize ascorbic acid) leads to
an increase in potentially pro-atherogenic LDL cholesterol
and a decrease in HDL cholesterol, resulting in
hypercholesterolaemia (Uchida et al., 1990).
19.11.3 Protection of LDL against
peroxidative modifi cation
Physiological concentrations of ascorbic acid protect
LDL against copper-catalysed peroxidative modifi cation
in vitro, maintaining the ability of LDL to be
recognized by appropriate LDL receptors and not by the scavenger receptor of macrophages (Sakuma
et al., 2001). This protective action preserves LDL’s
indigenous lipid-soluble antioxidants, except for
ubiquinol, the reduced form of coenzyme Q (Retsky
& Frei, 1995). Ascorbic acid spares, rather than
regenerates, LDL-associated α-tocopherol, i.e. prevents
α-tocopherol oxidation in the fi rst place. The
dilemma of whether ascorbate acts as a pro-oxidant or
as an antioxidant when interacting with LDL has been
addressed by Lynch et al. (1996). Ascorbate protects
native or mildly oxidized LDL against further metal
ion-dependent oxidation; only if LDL becomes extensively
oxidized does ascorbate acts as a pro-oxidant.
19.11.4 Effects on nitric oxide-mediated
arterial relaxation
Several studies have shown that an acute application
of ascorbic acid enhanced endothelium-dependent
vasodilation in patients with diabetes, coronary artery
disease, hypertension, hypercholesterolaemia, or
chronic heart failure, and in cigarette smokers (Heitzer
et al., 1996; Levine et al., 1996; Ting et al., 1996; Solzbach
et al., 1997; Ting et al., 1997; Hornig et al., 1998).
Long-term ascorbic acid treatment (500 mg per day)
produced a sustained improvement in endotheliumdependent
vasodilation in patients with coronary artery
disease (Gokce et al., 1999). Kanani et al. (1999)
demonstrated that administration of ascorbic acid
prevents induction of endothelial dysfunction by homocysteine.
These fi ndings may be attributable to the
scavenging of superoxide by ascorbate, thus preventing
the reaction between superoxide and nitric oxide
to form hydroxyl radicals and nitrogen dioxide, both
of which can initiate lipid peroxidation.
Heller et al. (1999) demonstrated that pre-incubation
of cultured endothelial cells with ascorbic acid
led to a three-fold increase of the cellular production
of nitric oxide after stimulation with ionomycin or
thrombin. Ascorbate did not induce the expression of
nitric oxide synthase and appeared to act through an
effect on the availability or affi nity of the enzyme cofactor
tetrahydrobiopterin. The fi ndings suggest that
saturation of the vascular tissue with ascorbate provides
the optimum reaction conditions for adequate
nitric oxide synthesis and that a decrease in intracellular
ascorbate leads to endothelial dysfunction.
19.11.5 Enhancement of prostacyclin
The formation of prostacyclin (PGI2), a member of
the prostaglandin family which protects the arterial
wall against deposition of platelets, is inhibited by hydroperoxides
of unsaturated fatty acids. In vitro studies
have shown that physiological concentrations of
ascorbic acid enhance the formation of prostacyclin
by aortic rings by protecting the cyclooxygenase and
PGI-synthase (Beetens & Herman, 1983).
19.11.6 Effects of vitamin C
Rifi ci & Khachadurian (1993) administered vitamin
C (1 g per day) and vitamin E (800 IU per day), both
separately and in combination, to healthy female and
male subjects and examined oxidation of lipoproteins
in vitro. Vitamin E administration alone produced a
52% inhibition and vitamin C alone a 15% inhibition
of copper-catalysed thiobarbituric acid reactive
substances production; the combination of vitamins
produced a 63% inhibition. Harats et al. (1998) reported
that in young healthy male subjects consuming
a diet high in saturated fats, supplementation with
citrus fruits containing an estimated 500 mg per day
of vitamin C reduced the in vitro susceptibility of
LDL to oxidation. Mosca et al. (1997) reported that
antioxidant supplementation with a combination of
800 IU of vitamin E, 1000 mg of vitamin C and 24 mg
of β-carotene signifi cantly reduced the susceptibility
of LDL to oxidation in patients with coronary artery
disease. The response produced by a similar combination
containing half the amounts of each antioxidant
was non-signifi cant.
19.11.7 Epidemiological studies
Current evidence from epidemiological studies on the
role of vitamin C in the prevention of cardiovascular
disease is inconclusive, with some studies showing a
very strong correlation between vitamin C intake and
incidence of cardiovascular events and other studies
showing no correlation at all (Lynch et al., 1996; Institute
of Medicine, 2000).

More in Vitamin C

Natural killer cell activity
An in vivo effect of ascorbic acid on enhancement of
human natural killer cell activity has been reported
at a dosage of 60 mg per kg body weight (Vojdani &
Ghoneum, 1993).
19.10.8 Regulation of the complement
component C1q
When guinea pigs were fed tissue-saturating amounts
of vitamin C, plasma C1q concentrations were signifi -
cantly higher than in those animals fed only enough
ascorbate for adequate growth and for the prevention
of scurvy (Haskell & Johnston, 1991). When healthy
men and women were given 500 mg ascorbate three
times daily with meals for 4 weeks, their plasma C1q
levels were not signifi cantly altered (Johnston, 1991).
Hence, signifi cantly enhanced C1q production may
occur only during activation of the immune system,
and not in healthy, non-infected individuals.
19.10.9 Enhancement of lymphocyte blastogenesis
Using cultured spleen cells from an inbred strain of
rat that does not synthesize vitamin C, Oh & Nakano
(1988) observed that ascorbic acid enhanced
lymphocyte blastogenesis through inhibition of the
biosynthesis of immunosuppressive histamine.
19.10.10 Enhancement of interferon
The participation of vitamin C in protection against
some viral infections may be in the enhancement of
interferon biosynthesis as demonstrated in vivo and
in vitro. The level of circulating interferon induced
in mice by inoculation with leukaemia virus was enhanced
by the addition of ascorbate to the drinking
water (Siegel, 1974). Ascorbate also enhanced the interferon
levels produced by cultured human embryo
fi broblasts in response to Newcastle Disease virus
(Dahl & Degré, 1976; Karpin´ ska et al., 1982).
19.10.11 Regulation of cytokines
Vitamin C has an indirect effect on lymphocyte proliferation
through its action on cytokines, as shown in
vitro by Cunningham-Rundles et al. (1993). Ascorbic
acid suppressed proliferation response to interleukin-
2, suggesting a basis for the vitamin’s inhibitory effect
on mitogen-induced lymphocyte proliferation. In
contrast, ascorbic acid enhanced the proliferative
response to interferon-γ, without inhibiting the
production of interferon-γ that accompanied the response
to infl uenza A (Table 19.4).
19.10.12 Clinical application to immunodeficiency diseases
Anderson (1981) administered a single oral daily dose
of 1 g sodium ascorbate to three children suffering
from chronic granulomatous disease as a supplement
to prophylactic trimethoprim–sulphamethoxazole
therapy for 2 years. In all three patients, introduction
of ascorbate to the therapeutic regimen resulted in the
correction of defective neutrophil motility and increased
activity against staphylococci. These responses
were accompanied by a decrease in the frequency of
infection and increased weight and growth rate.

Immune function of Vitamin C

There is a large body of evidence that vitamin C plays
an important role in the biochemistry of the human
immune system, particularly in the stimulation of
phagocytosis. Leucocytes accumulate ascorbic acid
after uptake from the plasma by active transport
(Moser, 1987), suggesting an involvement of the vitamin
in the normal function of these cells. The concentration
of ascorbate in monocytes, for example, is over
80 times higher than that in plasma (Evans et al., 1982)
and macrophages contain about twice as much ascorbate
as neutrophils and monocytes (Schmidt & Moser,
1985). Vitamin C accumulation in activated human
neutrophils is increased as much as ten-fold above the
concentrations present in resting neutrophils as a result
of a novel vitamin recycling mechanism. Extracellular
ascorbate is oxidized to dehydroascorbic acid by
oxidants generated by the activated neutrophil. The
dehydroascorbic acid is preferentially taken up by the
neutrophil and reduced intracellularly to ascorbate
within minutes (Washko et al., 1993). Ascorbate, as an
antioxidant, protects phagocytes from self-destruction
by reactive oxidants (Muggli, 1993). It also neutralizes
reactive oxidants released extracellularly by
activated phagocytes, thereby preventing damage to
surrounding host tissue (Anderson & Lukey, 1987).

Antioxidant role of Vitamin C

Ascorbate is an effective scavenger of all aggressive
reactive oxygen species within the aqueous environment
of the cytosol and extracellular fl uids. These
species include hydroxyl, superoxide anion and nonlipid
peroxyl radicals together with the non-radicals
singlet oxygen and hydrogen peroxide (Sies & Stahl,
1995). Ascorbate reacts with free radicals to produce
the ascorbyl radical and detoxifi ed product through
a single-electron transfer. Fig. 19.8 shows a possible
scheme in which ascorbate can be recycled during the
scavenging process.
Ascorbate is not the only antioxidant in aqueous
systems: other water-soluble antioxidants such as protein thiols and urate are also present. However,
ascorbate is the only endogenous antioxidant that effectively
protects the lipids in blood plasma (and also
low-density lipoprotein) against oxidative damage
initiated by non-lipid peroxyl radicals generated in
the aqueous phase. This is observed as a complete cessation
of lipid peroxidation when ascorbate is added
to plasma; other endogenous antioxidants, including
α-tocopherol, do not have this effect (Frei, 1991). Apparently,
ascorbate traps virtually all peroxyl radicals
generated in the aqueous phase before they can diffuse
into the lipid phase. Thus, ascorbate acts as the fi rst
and major line of antioxidant defence in the protection
of lipoidal plasma constituents and low-density
lipoprotein. In this action, ascorbate spares vitamin
E, the chain-breaking antioxidant in the lipid phase
(Doba et al., 1985).
In its role as a lipid-soluble, chain-breaking antioxidant
in biomembranes and lipoproteins (see Section
9.5), vitamin E (tocopherol, T-OH) scavenges
lipid peroxyl free radicals and itself is converted to
the tocopheroxyl radical (T-O•). Lipid peroxyl radicals,
because of their location in lipid environments,
cannot be scavenged by ascorbate anion. However, in
vitro studies using phospholipid liposomes as model
biomembranes have shown that ascorbate (AH–)
restores the antioxidant activity of vitamin E by converting
the tocopheroxyl radical back to the phenolic
tocopherol (reaction 19.7). Ascorbate works at the
lipid–water interface of membranes, very close to the
polar head groups of tocopherol.
T-O• + AH– → T-OH + A–• (19.7)
Whether vitamin C regenerates vitamin E in vivo is
debatable. Burton et al. (1990) found no evidence for
an interaction between the two vitamins in guinea
pigs and concluded that any such interaction must be
negligible in comparison with the normal turnover
of vitamin E.
As discussed above, ascorbate is an excellent antioxidant
but, paradoxically, it can also behave as a pro-oxidant
at lower concentrations (Buettner & Jurkiewicz,
1996). This crossover effect from pro-oxidant to
antioxidant is dependent on the ability of transition
metals in their reduced forms (e.g. Fe2+ and Cu+) to
catalyse the generation of free radicals. Ascorbate,
being a powerful reducing agent, maintains transition
metals in their catalytic reduced forms. At a high concentration
of ascorbate, the length of free radical chain
reactions will be small owing to ascorbate’s free radical
scavenging action. As the concentration of ascorbate is
lowered, there will come a point where its antioxidant
action is negligible but its capacity to reduce catalytic
metals is still suffi cient. At this crossover point ascorbate
switches from being an antioxidant to a prooxidant.
The antioxidant/pro-oxidant behaviour of
ascorbate has implications in the protection of plasma
LDL from oxidative modifi cation (Section 19.11.3).
The antioxidant action of vitamin C has a wide variety
of protective roles in the body. For example:
• the DNA in human sperm is protected from free
radical damage (Fraga et al., 1991);
• lung tissue is protected from free radical damage
resulting from inhalation of tobacco smoke, pollutants
and ozone;
• ocular tissue is protected from photo-oxidative
damage that can ultimately result in cataract formation;
• the high concentrations of ascorbate in neutrophils
and macrophages and its release on stimulation
protect these phagocytes and host tissue during the
respiratory burst in which reactive oxygen species
are produced to kill phagocytosed pathogens.

Biosynthesis of collagen

Collagen is the major macromolecule of most connective
tissues. It is composed of three α chain
sub units that are wound together to form a triple
helix. Cross-linking gives the molecule a rigid and
inextensible structure. There are over 25 different α
chains that associate to yield 15 different types of collagen.
Type I collagen, which is found in large quantities
in skin and bone, comprises two α1(I)-chains and
one α2(I)-chain. The amino acid composition of collagen
is unusual among animal proteins in that it has
an abundance of proline and 4-hydroxyproline and a
few residues of 3-hydroxyproline and hydroxylysine.
The hydroxyproline residues are necessary for proper
structural conformation and stability; hydroxylysine
residues take part in cross-linking and facilitate subsequent
glycosylation and phosphorylation.
Collagen α chains are synthesized in a precursor
form known as proα chains, which have additional
non-collagenous amino acid sequences (propeptides)
at both amino and carboxyl termini. The presence of
hydroxyproline and hydroxylysine arises through the
post-translational hydroxylation of particular proline
and lysine residues in the polypeptide chain. Within
the cisternae of the rough endoplasmic reticulum,
the newly synthesized proα chains encounter three
hydroxylating enzymes. Two of these enzymes, prolyl-
4-hydroxylase and prolyl-3-hydroxylase, convert
proline residues to 4-hydroxyproline or 3-hydroxyproline
respectively, and the third, lysyl hydroxylase,
converts lysine residues to hydroxylysine. Following
amino acid modifi cation, the propeptides at the carboxyl
termini of two proα1 and one proα2 chains associate
and bond through disulphide bridges. Triple
helix formation then takes place as the protein passes
through the endoplasmic reticulum. Following attachment
of carbohydrate moieties to the carboxy
terminal propeptides, the procollagen molecules are
transported to the cell surface within secretory granules.
Enzymatic removal of the propeptides during
the process of extrusion allows the collagen molecules
to spontaneously assemble into fi brils. These are then
cross-linked by a series of covalent bonds and deposited
into the extracellular matrix.
Ascorbic acid stimulates collagen synthesis through
increased transcription of procollagen genes (Hitomi
& Tsukagoshi, 1996). Also, ascorbic acid is an essential
cofactor for the post-translational hydroxylation of
proline and lysine residues in the polypeptide chain.
Each of the enzymes concerned contains an iron ion
(maintained in the ferrous state by ascorbate) and
requires molecular oxygen and α-ketoglutarate as
co-substrates (Prockop et al., 1979) (Fig. 19.4). The
absence of wound healing is one of the features of
scurvy that can be attributed to impaired collagen
synthesis arising from lack of vitamin C.
The pathway of collagen synthesis is tightly coupled
through feedback regulation (Schwarz et al., 1987).
Proline hydroxylation stabilizes the triple helical conformation of the procollagen. This conformation
increases the secretion rates by six-fold and this in
turn leads to an increase in translational effi ciency.
Therefore ascorbate levels, solely by controlling the
activity of the proline hydroxylation step, can control
the chain of events through the whole pathway.

Renal reabsorption of Vitamin C

General principles
The kidney actively reabsorbs ascorbate present in the
glomerular fi ltrate, thereby maximizing vitamin C
conservation in the body and helping the intestine to
maintain the circulating vitamin in its useful, reduced
state. The kidneys of all mammals handle vitamin C in
a similar manner. Renal reabsorption of vitamin C is
an essential process for humans as, without it, urinary
loss would far exceed the average daily intake of the
vitamin. Although species that have the ability to synthesize
ascorbic acid might be able to replace that lost
in the urine, the metabolic costs would be high.
Transport mechanisms
Ascorbic acid
Renal uptake of the L-ascorbate anion at the brushborder
membrane of the absorptive cell of the proximal
convoluted tubule is, like intestinal uptake in the
human or guinea pig, a sodium-coupled, secondary
active transport system (Rose, 1986; Bowers-Komro
& McCormick, 1991). Unlike the corresponding intestinal
transport system, however, the renal system
is electrogenic, indicating a Na+/ascorbate– coupling
ratio of 2:1 (Toggenburger et al., 1981). As the loaded
carrier bears a net positive charge, its transport is accelerated
by the negative membrane potential. Rapid
renal reabsorption of ascorbate is essential considering
that the transit time in the proximal tubule is only
about 10 s. Ascorbate is transported across the basolateral
membrane by sodium-independent facilitated
diffusion (Bianchi & Rose, 1985a).
Dehydroascorbic acid
The mechanism of dehydroascorbic acid transport in
renal brush-border (Bianchi & Rose, 1985b) and basolateral
(Bianchi & Rose, 1985a) membrane vesicles
appears to be facilitated diffusion. A favourable gradient
for continued renal reabsorption is maintained
dehydroascorbic acid (Rose, 1989). Dehydroascorbic
acid is taken up also from the blood across the basolateral
cell membrane and subsequently reduced to
ascorbate, which is then returned to the circulation
(Rose, 1989). The kidney participates with the intestine
and blood components in promoting reduction
of dehydroascorbic acid derived from the blood.

Intestinal absorption of Vitamin C

General principles
Approximately 80–90% of the vitamin C content of
a given foodstuff exists in the reduced form, ascorbic
acid; the remainder is in the oxidized form, dehydroascorbic
acid. Ascorbic acid and dehydroascorbic
acid are absorbed by separate transport mechanisms
in animal species that depend upon dietary vitamin
C (Fig. 19.3). Inside the absorptive cell (enterocyte)
of the intestinal epithelium, dehydroascorbic acid is
enzymatically reduced and the accumulated ascorbic
acid is transported across the basolateral membrane
to the bloodstream. In addition to uptake at the
brush-border membrane, dehydroascorbic acid from
the bloodstream can be taken up at the basolateral
membrane, reduced within the cell, and returned to
the circulation in the form of the useful and non-toxic
ascorbic acid. The serosal uptake of dehydroascorbic
acid from the bloodstream and intracellular reduction
to ascorbic acid take place in animal species which do
not have a dietary vitamin C requirement as well as
those species that do. The ability of the enterocyte to
absorb dehydroascorbic acid effi ciently is important
because, apart from the indigenous dehydroascorbic
acid content of the diet, additional oxidation of
ascorbic acid occurs in the gastrointestinal tract as
the vitamin functions to maintain other nutrients
such as iron in the reduced state. The intestinal uptake
and reduction of dehydroascorbic acid explains
why this compound, orally administered, maintains
plasma concentrations of ascorbic acid and prevents
scurvy. The overall system of intestinal transport and
metabolism is designed to maximize the conservation
of vitamin C and also to maintain the vitamin in its
non-toxic reduced state, whether it is derived from the
diet .

Effi ciency of ascorbate absorption in humans
The usual dietary intake of vitamin C ranges from 30–
180 mg per day and over this range the effi ciency of
absorption is 70–90% (Institute of Medicine, 2000).
Brush-border uptake by the sodium-coupled, secondary
active transport mechanism reaches its maximum
rate at a relatively low luminal concentration. Beyond
physiological intakes, absorption becomes progressively
less effi cient, falling from 75% of a single 1-g
dose to 16% of a single 12-g dose (Table 19.2). This
fall-off in effi ciency occurs because absorption of
high luminal concentrations of vitamin C takes place
mainly by simple diffusion, and this passive movement
proceeds at a very low rate.
The ingestion of eight 0.125-g doses of ascorbate
spaced throughout the day produced a 72% increase
in absorption compared to a single 1-g dose (Yung et
al., 1981). The absorption effi ciency of a single dose
can be improved if the ascorbate is ingested in the
form of a sustained-release capsule (Sacharin et al.,
1976). The ingestion of 1 g of ascorbate immediately
after a fatty meal produced a 69% increase in absorption
compared to the same dose given on an empty
stomach (Yung et al., 1981). The divided dose effect
is consistent with a saturable absorption mechanism,
while the after-meal effect indicates a slowing of gastric
Adaptive regulation of ascorbate absorption in
guinea pigs
In the guinea pig, intestinal absorption of ascorbate
is adaptively regulated in a transient and reversible
manner by the level of dietary ascorbate (Karasov et
al., 1991). The mechanism of regulation is an increase
or decrease in the number of carriers at both brushborder
and basolateral membranes of enterocytes in
response to low or high concentrations of ascorbate
in the blood. The rationale for adaptive regulation
is that carriers are most needed at low dietary ascorbate
levels; at excessive levels the required amount of
ascorbate can be absorbed by fewer carriers, aided by
passive diffusion. As ascorbate does not provide metabolizable
energy, there is nothing to gain from the
cost of synthesizing and maintaining carriers when
the vitamin supply is in excess. The issue of adaptive
regulation has not been examined in humans.

Vitamin C

The structure of vitamin C, designated as a hex uronic
acid, was established in 1933 at the University of
Birmingham in England by Walter Haworth and
his associates, who also accomplished its synthesis.
Szent-Györgyi and Haworth renamed hexuronic acid
‘L-ascorbic acid’ to convey its antiscorbutic properties;
the new name was offi cially accepted in 1965.
Both Szent-Györgyi and Haworth were to be awarded
the Nobel Prize in 1937, the former for Physiology
and Medicine and the latter for Chemistry. Synthetic
ascorbic acid proved to have identical physicochemical
and biological properties to the vitamin C isolated
from plant or animal tissues, and there was no difference
in biological potency between the synthetic and
natural products. In 1934, Reichstein and Grüssner
in Switzerland worked out a chemical route for synthesizing
ascorbic acid commercially, starting from

The term ‘vitamin C’ refers to both ascorbic acid
and dehydroascorbic acid, since the latter oxidation product is reduced back to ascorbic acid in the body.
The principal natural compound with vitamin C activity
is L-ascorbic acid. There are two enantiomeric
pairs (mirror images) of the 2-hexenono-1,4-lactone
structure; namely, L- and D-ascorbic acid and L- and
D-isoascorbic acid (Fig. 19.1). D-Ascorbic acid and
L-isoascorbic acid are devoid of vitamin C activ- ity and do not occur in nature. D-Isoascorbic acid
(commonly known as erythorbic acid) is an epimer
of L-ascorbic acid, the structural difference being the
orientation of the hydrogen and hydroxyl group at
the fi fth carbon atom. D-Isoascorbic acid is also not
found in natural products, apart from its occurrence
in certain microorganisms. It possesses similar reductive
properties to L-ascorbic acid, but exhibits only
5% of the antiscorbutic activity of L-ascorbic acid in
guinea pigs (Pelletier & Godin, 1969).
At around neutral pH, ascorbic acid exists as the
ascorbate anion due to facile ionization of the hydroxyl
group on C-3. Ascorbate is easily and reversibly
oxidized to dehydro-L-ascorbic acid, forming the
ascorbyl radical (also known as semidehydroascorbate)
as an intermediate (Fig. 19.2). The delocalized
nature of the unpaired electron in the ascorbyl radical
makes it a relatively unreactive free radical and two
ascorbyl radicals can react together non-enzymatically
to produce ascorbate and dehydroascorbic acid. In
the body, enzymes are available to reduce the ascorbyl
radical and dehydroascorbic acid back to ascorbate.
Dehydroascorbic acid is not a true organic acid as
it contains no readily ionizable protons. In aqueous
solution, dehydroascorbic acid exists not as the
2,3-diketo compound, but as the bicyclic hemiketal
hydrate. In buffered solution at neutral or alkaline
pH, dehydroascorbic acid undergoes a non-reversible
oxidation in which the two rings open to give 2,3-
diketogulonic acid in a straight-chain structure.

Dietary sources of Vitamin C
Fresh fruits (especially citrus fruits and blackcurrants)
and green vegetables constitute rich sources
of vitamin C. Potatoes contain moderate amounts
but, because of their high consumption, represent the
most important source of the vitamin in the British
diet. Liver (containing 10–40 mg per 100 g), kidney
and heart are good sources, but muscle meats and cereal
grains do not contain the vitamin in measurable
amounts. Human milk provides enough ascorbic acid
to prevent scurvy in breast-fed infants, but preparations
of cow’s milk are a poor source owing to oxidative
losses incurred during processing.

Causes of vitamin B12 malabsorption

Vitamin B12 deficiency results from a failure of intestinal
absorption or subsequent transport to the tissues;
it is rarely, if ever, caused by a lack of B12 in the diet.
Disorders of vitamin B12 absorption and transport
have been discussed by Kapadia & Donaldson (1985)
and just a few examples of malabsorption are mentioned
• Elderly people are prone to atrophic gastritis, a
condition in which the gastric oxyntic mucosa
atrophies to such an extent that virtually no hydrochloric
acid or intrinsic factor is secreted.
• In patients with diverticula, strictures and fi stulas
of the small intestine, stagnant regions of the lumen
may become contaminated with colonic bacteria
which can take up much of the dietary vitamin B12
passing by. Bacteria can take up vitamin B12 bound
to intrinsic factor, although not as avidly as they can
take up the free vitamin. Intrinsic factor and bacteria
have a similar affi nity for B12, so it is possible
that bacterial uptake could take place following the
vitamin’s release from haptocorrin but before its
transfer to intrinsic factor.
• The fish tapeworm Diphyllobothrium latum competes
with the host for B12, making it less available
for absorption.
• A common inherited disorder is an auto-immune reaction
with formation of antibodies against intrinsic
factor. Such cases involve two types of antibody: type
I prevents intrinsic factor from binding to cobalamin
and type II blocks the binding of the intrinsic factor–
cobalamin complex to the ileal receptor.

Vitamin B12

A type of anaemia attributed to a digestive disorder
was reported by Combe in 1822 and later recognized
as pernicious anaemia by Addison in 1849. It was not
until 1926 that Minot and Murphy started to cure patients
suffering from pernicious anaemia by feeding
them with large amounts of raw liver. The idea for this treatment originated from the discovery by Whipple
that dietary liver improved haemoglobin production
in iron-defi cient dogs. In 1929, Castle showed that the
intestinal absorption of the ‘antipernicious anaemia
principal’ required prior binding to a specifi c protein
(intrinsic factor) secreted by the stomach.
Research into isolating the active principal from
liver was hampered by the inability to induce pernicious anaemia in animals. For many years, the
only known bioassay was the haemopoietic response
of patients with the disease. Eventually, in 1948, a red
crystalline substance having the clinical activity of
liver and designated as vitamin B12 was isolated by
several independent scientifi c groups. The success of
one group, headed by Folkers (Merck and Co., USA),
was largely attributable to a microbiological assay developed
by Shorb in 1947. The complicated structure
of vitamin B12 was established by Hodgkin using X-ray
crystallography in 1955. Its complete chemical synthesis
was achieved in 1973, but because of the large
number of stages required (over 70) the procedure is
of no commercial interest.

In accordance with the literature on nutrition and
pharmacology, the term vitamin B12 is used in this
text as the generic descriptor for all cobalamins that
exhibit antipernicious anaemia activity. Individual
cobalamins will be referred to by their specifi c names
(e.g. cyanocobalamin).
The cobalamin molecule depicted in Fig. 18.1 contains
a corrin ring system and a cobalt atom, which
may assume an oxidation state of (I), (II) or (III).
There are two vitamin B12 coenzymes with known
metabolic activity in humans, namely methylcobalamin
and 5´-deoxyadenosylcobalamin (frequently
abbreviated to adenosylcobalamin and also known
as coenzyme B12). The methyl or adenosyl ligands of
the coenzymes occupy the X position in the corrin
structure. The coenzymes are bound intracellularly
to their protein apoenzymes through a covalent peptide
link, or in milk and plasma to specifi c transport
proteins. The enzyme-bound cobalamins exist as
Cyanocobalamin is the most stable of the vitamin
B12-active cobalamins and is the one mostly used in
pharmaceutical preparations and food supplementation.
Aqueous solutions of cyanocobalamin are stable
in air at room temperature if protected from light. On
exposure to light, the cyano group dissociates from
cyanocobalamin and hydroxocobalamin is formed.
In neutral and acid solution hydroxocobalamin exists
in the form of aquocobalamin (Gräsbeck & Salonen,
1976). This photolytic reaction does not cause a loss
of activity.

Naturally occurring vitamin B12 originates solely
from synthesis by bacteria and other microorganisms
growing in soil or water, in sewage, and in the
rumen and intestinal tract of animals. Any traces of
the vitamin that may be detected in plants are due to
microbial contamination from the soil or manure or,
in the case of certain legumes, to bacterial synthesis in
the root nodules.
Vitamin B12 is ubiquitous in foods of animal origin
and is derived from the animal’s ingestion of cobalamin-
containing animal tissue or microbiologically
contaminated plant material, in addition to vitamin absorbed from the animal’s own digestive tract. Liver
is the outstanding dietary source of the vitamin, followed
by kidney and heart. Muscle meats, fi sh, eggs,
cheese and milk are other important food sources.
Vitamin B12 activity has been reported in yeast, but
this has since been attributed to the presence of noncobalamin
corrinoids or vitamin B12 originating
from the enriching medium (Herbert, 1988). About
5 to 30% of the reported vitamin B12 in foods may be
microbiologically active non-cobalamin corrinoids
rather than true B12 (National Research Council,
Vitamin B12 in foods exists in several forms as reported
by Farquharson & Adams (1976). Meat and fi sh
contain mostly adenosyl- and hydroxocobalamins;
these compounds, accompanied by methylcobalamin,
also occur in dairy products, with hydroxocobalamin
predominating in milk. Sulphitocobalamin is found
in canned meats and fi sh. Cyancobalamin was only
detected in small amounts in egg white, cheeses and
boiled haddock.

Chromosome damage: implications for cancer

Division of cells with unrepaired or misrepaired DNA
damage leads to mutations. If these relate to critical
genes, such as proto-oncogenes or tumour suppressor
genes, cancer may result. Folate is essential for
DNA synthesis and repair through its role in purine
and pyrimidine synthesis. Its role in the synthesis of
S-adenosylmethionine (SAM) is also relevant to cancer.
SAM donates its methyl group to DNA, among
other acceptors, and a defi ciency of folate can lead to
hypomethylation of DNA. As DNA methylation is a
mechanism for silencing transcription (Ng & Bird,
1999), hypomethylation of DNA has the potential
to alter the normal control of gene expression. Hypomethylation
also alters chromatin conformation,
thereby allowing access of DNA-damaging agents and
endonucleases, which destabilize the DNA and make
it more susceptible to strand breaks (Kim et al., 1997).
Imbalanced DNA methylation is a common occurrence
in carcinogenesis (Laird & Jaenisch, 1994).
Low cytosolic levels of 5,10-methylene-THF associated
with folate defi ciency result in decreased synthesis
of deoxythymidine monophosphate (dTMP) and
the accumulation of deoxyuridine monophosphate
(dUMP). This leads to DNA polymerase-mediated
incorporation of dUMP into the DNA molecule in
place of dTMP. Normal DNA repair processes remove
the misincorporated dUMP, forming transient singlestrand
breaks (nicks) that could result in a doublestrand
break if two opposing nicks are formed. Kim
et al. (1997) showed that, in rats, dietary folate depletion
is capable of producing DNA strand breaks and
hypomethylation within a highly conserved region of
the p53 tumour suppressor gene. The p53 gene was
chosen for study because alterations in it have been
implicated in >50% of human cancers. On a genomewide
basis such alterations either did not occur or
were delayed, indicating some selectivity for the exons
examined within the p53 gene.
In epidemiological studies, dietary folate deficiency
is associated with an increased risk of several specific malignancies, notably cancer of the cervix, lung,
colorectum and brain (Glynn & Albanes, 1994). The
presence of micronucleated erythrocytes in marginal
folate defi ciency is indicative of chromosomal damage
(Everson et al., 1988). Both high DNA dUMP
levels and elevated erythrocyte/reticulocyte micronucleus
frequency are reversed by folate administration
(Blount et al., 1997). Duthie & Hawdon (1998)
showed that a dietary intake of folate adequate for the
prevention of clinical defi ciency may not be suffi cient
to maintain DNA stability.
In a study of American male physicians, Ma et al.
(1997) showed that the C677T polymorphism in the
MTHFR gene reduces the risk of colorectal cancer.
Subjects with the homozygous mutation (15% in controls)
had half the risk of colorectal cancer (odds ratio
0.49; 95% confi dence interval 0.27 to 0.87) compared
with the homozygous normal or heterozygous genotypes.
The protection due to the polymorphism was
absent in subjects with folate defi ciency and reduced
in those with high alcohol consumption. It can be
reasoned that, provided folate status is adequate, the
reduced activity of the thermolabile MTHFR enzyme
variant would lead to an increased level of intracellular
5,10-methylene-THF and this would reduce the
likelihood of dUMP misincorporation into DNA.

Megaloblastic anaemia

A defi ciency of folate gives rise to a type of anaemia
known as megaloblastic anaemia. A clinically indistinguishable
anaemia is also produced by vitamin
B12 deficiency but, because it is accompanied by
neurological damage, it is referred to as pernicious
anaemia. Both types of anaemia are the result of abnormal
nuclear maturation caused by impaired DNA
synthesis. The impaired DNA synthesis is presumed
to be attributable to reduced intracellular levels of
polyglutamyl 5,10-methylene-THF. As shown in
reaction 17.7, this important folate is involved in
the formation of deoxythymidine monophosphate,
one of the two pyrimidine nucleotide constituents of
DNA. The defect in DNA synthesis leads to a variety of
secondary disturbances which result in the premature
death of many haemopoietic cells in the bone marrow,
possibly without ever completing the S phase of cell

Megaloblastic anaemia caused by folate defi ciency
manifests as megaloblastosis of the bone marrow
and macrocytosis of the circulating erythrocytes.
Examination of the bone marrow is of great diagnostic
importance. The erythrocyte precursor cells
(erythroblasts) in the bone marrow fail to proliferate
rapidly and exist as gigantic cells called megaloblasts
at all stages of maturation. It is the existence of such
cells that gives rise to the term megaloblastic anaemia.
The increased size of megaloblasts is apparent both in
the cytoplasm and in the nucleus. The nuclei contain
smaller quantities of condensed chromatin than the
nuclei of normoblasts of similar maturity and thus
have an open, sieve-like appearance.
The circulating erythrocytes which are derived from
mature megaloblasts are also abnormally large and are
referred to as macrocytes. The mean corpuscular volume
of macrocytes ranges from 100 to 160 μm3 compared
with 90 to 95 μm3 for normal erythrocytes. The
macrocytes are generally oval in shape (erythrocytes
are biconcave discs) but some fragmented and irregularly
shaped cells are also present. There is a reduction
in red cell count and sometimes incredibly low values
are found. The haemoglobin content of individual macrocytes is increased owing to their larger size, but
there is little change in the haemoglobin concentration
of whole blood.
White cells and platelets are also produced in the
bone marrow. In the differentiating granulocyte series
of white blood cells (neutrophils, eosinophils and
basophils) giant, abnormally shaped metamyelocytes
are found. Megakaryocytes (precursors of platelets)
may also be larger than usual. In advanced folate/B12
defi ciency, the total white cell count and platelet count
may be low. Circulating neutrophils are characterized
by an increased number of nuclear segments. The
presence of hypersegmented neutrophils is a valuable
clue in diagnosing folate/B12 deficiency when red cell
changes are masked by a coexistent iron defi ciency or
the anaemia of chronic disease.
As in all cases of anaemia, the body adjusts its
cardiopulmonary system to compensate for the diminished
oxygen-carrying capacity of the blood, so
in mild anaemia the subject may not be aware of any
problems. Eventually, however, the progressing anaemia
leads to symptoms of weakness, fatigue, shortness
of breath and palpitations. The sufferer may also
experience headache, irritability and an inability to
concentrate. Visible signs of megaloblastic anaemia
in white-skinned people are a marked pallor and a
slight jaundice, giving the skin a distinctive lemonyellow
Megaloblastosis is not confi ned to developing cells
in the bone marrow – all other rapidly dividing cell
types will be affected, including epithelial cells lining
the gastrointestinal, respiratory, and urinogenital
tracts. A notable feature is glossitis where the tongue
is sore at the edges, bright red in colour and smooth in
texture. Gastrointestinal disturbances caused by defective
gut epithelia have adverse consequences upon
overall nutritional status. Male infertility results from
impaired spermatogenesis.
Folate defi ciency can result, in the absence of disease,
from reduced ingestion or absorption, or from
increased utilization. Dietary folate defi ciency is common
among people who, for various reasons, eat little
fruit or fresh vegetables. Absorption is impaired in
alcoholics. There is an increased utilization of folate
during pregnancy owing to the need to transfer an
extra 100–300 μg of folate per day to the fetus (Beatty
& Wickramasinghe, 1993). Experimental folate
defi ciency is diffi cult to produce under normal circumstances,
but the study of patients suffering from
malabsorption problems such as tropical sprue or the
use of folate antagonists has yielded much clinical


In plasma, 70–80% of homocysteine is bound to
plasma proteins, chiefl y albumin; only about 1% circulates
as free homocysteine. The remaining 20–30%
circulates as homocysteine disulphide (homocystine)
or as the mixed disulphide, homocysteine–cysteine.
Plasma homocysteine assays measure total homocysteine,
which is the sum of the homocysteine moieties
present in all of the above forms. Because variable
changes in plasma homocysteine concentration have
been observed post-prandially, it is customary to obtain
measurements in the fasting state. Normal levels
of fasting plasma homocysteine are considered to be
between 5 and 15 μmol L–1. Higher fasting values are
classifi ed arbitrarily as moderate (16–30), intermediate
(31–100) and severe (>100 μmol L–1) hyperhomocysteinaemia.
The methionine loading test has been
used to accentuate abnormalities of the homocysteine
metabolic pathways. The test measures fasting plasma
homocysteine before and 2 hours after an oral dose
of methionine (100 mg per kg body weight). An elevated
post-loading homocysteine level indicates an
Hyperhomocysteinaemia can result from inherited
defects in enzymes necessary for either trans-sulphuration
or remethylation and from acquired defi ciencies
in vitamin coenzymes. Renal insuffi ciency can
also lead to hyperhomocysteinaemia. Subclinical
folate defi ciency is commonly associated with hyperhomocysteinaemia,
presumably because of decreased
remethylation of homocysteine. Kang et al. (1987)
found elevated total homocysteine levels in 84% of
subjects with subnormal folate levels. The mean homocysteine
level in the low-folate subjects was about
four-fold greater than the mean level in the control
An association between mild hyperhomocysteinaemia
and increased risk of occlusive vascular disease in
the coronary, cerebral and peripheral arteries has been
demonstrated in case-control (Selhub et al., 1995; European
Concerted Action Project, 1997) and prospective
(Stampfer et al., 1992; Arneson et al., 1995; Perry
et al., 1995) studies. Plasma homocysteine concentration
is a strong predictor of mortality in patients with
angiographically confi rmed coronary artery disease
(Nygard et al., 1997). Whether hyperhomocysteinaemia
is a causal risk factor for the disease or simply
a marker of another prothrombotic risk factor(s) is
debatable (Kuller & Evans, 1998).
Up to 30% of patients with coronary artery disease
had homocysteine elevations that were 10–50%
greater than the level observed among normal subjects
(Clarke et al., 1991). Subjects with hyperhomocysteinaemia
have a two-fold to three-fold increase in
risk of developing cardiovascular disease or venous
thrombosis (den Heijer et al., 1998). In vitro studies
have shown that high concentrations of homocysteine
can promote a prothrombotic state at the luminal surface
of the blood vessel (Lentz, 1998). An association
between impaired endothelium-dependent vasodilation
and hyperhomocysteinaemia was demonstrated
in children with homozygous homocystinuria (Celermajer
et al., 1993), in monkeys fed a methionine-enriched
diet (Lentz et al., 1996), in methionine-loaded
healthy humans (Bellamy et al., 1998; Chambers et
al., 1998) and in non-induced hyperhomocysteinaemic
healthy middle-aged (Woo et al., 1997) and
elderly (Tawakol et al., 1997) humans. In healthy
human subjects, even physiological increments in
plasma homocysteine following oral administration
of methionine or an animal protein meal impaired
endothelium-dependent vasodilatation (Chambers et
al., 1999a). Plasma homocysteine concentration can
be decreased by dietary supplementation with folic
acid, which suggests that hyperhomocysteinaemia
may be a treatable risk factor for vascular disease.

Folate homeostasis

The majority of 5-methyl-THF arriving at the liver
from the intestine and taken up is not demethylated
and converted to polyglutamate; instead it is quickly
released for distribution to extrahepatic tissues. The
initial route for this distribution is the enterohepatic
circulation, whereby the folate is discharged into the
bile and subsequently reabsorbed by the small intestine
before re-entering the systemic circulation.
Accompanying 5-methyl-THF in the bile are larger
amounts of non-methylated tetrahydrofolates which
represent folates salvaged from dying cells such as
senescent erythrocytes and hepatocytes (Shin et al.,
1995). Any folic acid that might have been absorbed
and released into the portal circulation without
modifi cation is exclusively taken up by the liver and
either converted into one-carbon derivatives of THF
prior to rapid release into bile or polyglutamated and
incorporated into the hepatic folate pool (Steinberg,
1984). Hepatic reduction and derivatization of folic
acid provides another source of non-methylated tetrahydrofolates
present in bile (Shin et al., 1995).
The recycling of folate via the enterohepatic pathway
may account for as much as 50% of the folate
that ultimately reaches the extrahepatic tissues. Disruption
of the enterohepatic cycle by bile drainage
results in a fall of the serum folate level to 30–40% of
normal within 6 hours – a much more dramatic drop
than that seen with a folate-defi cient diet. Eventually,
the serum folate level stabilizes, despite continuing
losses in the bile. This suggests a net fl ux of folate into
the plasma compartment from tissue pools. Release
of stored folate from cells of any tissue requires hydrolysis
of the polyglutamates to monoglutamates by
intracellular conjugase.
The maintenance of a normal level of plasma folate
depends on regular increments of exogenous folate
from the diet. The enterohepatic circulation of folate
evens out the intermittent intake of dietary folate. The
liver plays a major role in maintaining folate homeostasis
because of its capacity to store about 50% of the
total body folate, its relatively rapid folate turnover,
and the large folate fl ux through the enterohepatic
circulation (Steinberg, 1984). In situations of dietary
folate defi ciency, the liver does not respond by releasing
its folate stores. Rather, the non-proliferating,
less metabolically active tissues mobilize their folate
stores and return monoglutamyl folate to the liver.
This folate is released by the liver via the enterohepatic
cycle and distributed to the tissues that most require
it – in particular, those with actively proliferating
cells. Preferential uptake of folate by certain tissues
(e.g. placenta and choroid plexus) is made possible
by the presence of the folate receptor on their cellular
surfaces. The kidney plays its part in conserving
body folate by actively reabsorbing folate from the
glomerular fi ltrate. In addition, a pathway exists that
is capable of salvaging folate released from senescent

Uptake of 5-methyl-THF by sinusoidal membrane
vesicles isolated from human liver is an electroneutral
active transport process, which is pH-dependent, sodium-
independent and appears to involve co-transport
with hydrogen ions mediated by the reduced
folate carrier (Horne et al., 1993). This would require
a mechanism for maintaining a gradient of H+ across
the basolateral membrane, but how this is accomplished
is not known for certain. Sinusoidal membrane
vesicles isolated from rat hepatocytes contain a
Na+–H+ exchange system (Arias & Forgac, 1984) and
it can be speculated that the H+ could be conducted
along the membrane and interact with the carrier,
thereby generating a ‘localized’ proton gradient that
could energize active transport of 5-methyl-THF.

Absorption of dietary folate

In the human, the entire small intestine is capable of
absorbing monoglutamyl folate. Absorption is somewhat
greater in the proximal than in the distal jejunum
which, in turn, is much greater than in the ileum.
Folate transport across the brush-border membrane
of the enterocyte proceeds by two parallel processes
(Selhub & Rosenberg, 1981). At physiological concentrations
(<5 μM) of luminal folate, transport occurs
primarily by a saturable process, whereas at higher
concentrations, transport occurs by a non-saturable
process with characteristics of simple diffusion. Zimmerman
et al. (1986) produced data which suggest
that the latter process occurs in part through a conductance
pathway that involves anionic folate and a
cation (perhaps Na+) whose membrane permeation
properties affect the rate of folate transport. The saturable
component is discussed in the following with no
further mention of unsaturable transport.
Transport of folate is mediated by the reduced folate
carrier and is markedly infl uenced by changes in pH
(Schron, 1991). Folate exists primarily as an anion at
the pH of the lumenal contents. In vitro studies using
everted rat jejunal rings showed that absorption was
maximal at pH 6.3 and fell off sharply between pH
6.3 and 7.6 (Russell et al., 1979). In studies using
brush-border membrane vesicles (Said et al., 1987),
folate uptake increased as the pH of the incubation
buffer was decreased from 7.4 to 5.5. This increase in
folate uptake appeared to be partly mediated through
folate–/OH– exchange and/or folate–/H+ co-transport
mechanisms driven by the proton gradient across the
membrane and partly through a direct effect of acidic
pH on the carrier. Inhibition of folate transport by
the anion transport inhibitor DIDS suggested the
involvement of the folate–/OH– exchange mechanism.
Data reported by Mason et al. (1990) suggest that
the effect of pH on the carrier is attributable to an
increased affi nity of the carrier for its folate substrate.
The physiological relevance of the pH dependency
may be related to the existence of the acidic microclimate
at the luminal surface of the jejunum. This socalled
‘unstirred layer’ has a pH that is approximately
2 units lower than the bulk luminal pH and therefore
provides the necessary extracellular acidic conditions
for folate uptake.
Said et al. (1987) also found that folate transport
across the brush-border membrane was saturable,
competitively inhibited by structural analogues of
folic acid, unaffected by transmembrane electrical potential,
and Na+-independent. The human intestinal
reduced folate carrier has been cloned and characterized
at the molecular level (Nguyen et al., 1997).
Said et al. (1997) studied the intracellular regulation
of intestinal folate uptake using monolayers of
cultured mature IEC-6 epithelial cells. These cells are
derived from the proximal small intestine of a normal
rat and possess all of the cellular structures of native
enterocytes. Uptake of folic acid by IEC-6 cells was
similar to that of the native small intestine. Intracellular
cyclic AMP was found to affect the uptake of
folic acid independently of protein kinase A. Protein
tyrosine kinase also affected uptake, but protein kinase
C and Ca2+/calmodulin mediated pathways had
no signifi cant effect.
During intestinal transport some of the folate is
converted within the enterocyte to 5-methyl-THF in
a pH-dependent manner (Strum, 1979). This conversion
is extensive at pH 6.0 and negligible at pH 7.5
presumably because dihydrofolate reductase, the
rate-limiting enzyme in the reduction and methylation
process, has an acidic pH optimum. The percent
conversion is reduced by increasing the concentration
of folate in the mucosal medium, thus indicating saturation
of the process. Since at higher concentrations
most transported folate remains unmodifi ed, intestinal
conversion of absorbed folic acid is not obligatory
for transport into the circulation.
The mechanism of folate exit from the enterocyte
into the lamina propria of the villus is also carriermediated
and sensitive to the effect of anion exchange inhibition. In addition, the exit mechanism is electroneutral
and Na+-independent and has a higher
affi nity for the substrate than has the system at the
brush-border membrane (Said & Redha, 1987).
Absorption of milk folate by the suckling infant
Milk from humans and several species of other mammals
contains a folate-binding protein (FBP) which
may be important for folate absorption by the suckling
infant. In neonates, the uptake of folate bound
to milk FBP occurs preferentially in the ileum as opposed
to the jejunum. The incomplete development
of pancreatic and intestinal absorptive functions
could allow the FBP to reach the ileum without being
digested. This situation was demonstrated by Salter
& Mowlem (1983) who showed that a proportion
of goat’s milk FBP administered orally to neonatal
goats survived along the length of the small intestine.
Protease inhibitors inherent to colostrum may assist
the passage of bound folate along the small intestine
(Laskowski & Laskowski, 1951). In vitro, the addition
of goat’s milk FBP to the medium enhanced the transport
of 5-methyl-THF in brush-border membrane
vesicles isolated from the small intestine of neonatal
goats (Salter & Blakeborough, 1988). Mason & Selhub
(1988) observed that the characteristics of FBP-bound
folate absorption in the suckling rat resemble in some
respects the characteristics of endocytotic absorption
of macromolecules – a well-documented feature of
the suckling mammal’s intestinal physiology.
Adaptive regulation of folate absorption
Said et al. (2000) induced folate defi ciency in rats
by feeding a folate-defi cient diet that contained an
antibiotic to decrease the bacterial synthesis of folate
in the intestine. Using everted intestinal sacs and
brush-border membrane vesicles, they showed that
folate deprivation causes a specifi c up-regulation in
the transport of physiological concentrations of folic
acid across the brush-border membrane of both the
small and large intestines. The effect in the small intestine
took place not only in the jejunum, but also in
the ileum, a region that does not usually absorb folate.
The up-regulation was mediated through an increase
in the number and/or activity of functional reduced
folate carriers (increased Vmax) with no signifi cant effect
on the affi nity of the transport system (unchanged
Km). The up-regulation was associated with a marked
increase in the levels of carrier mRNA and protein,
suggesting a possible involvement of transcriptional
regulatory mechanisms. In addition to the up-regulation
of transepithelial folate transport, folate defi -
ciency was associated with a 10-fold increase in the
activity of brush-border membrane conjugase. The
intestine is therefore able to maximize its ability to
extract the limited amount of folate ingested during
periods of deprivation.

Deconjugation of polyglutamyl folate

The folates naturally present in foods exist largely in
protein-bound form, the predominant vitamers being
polyglutamyl forms of THF, 5-methyl-THF and 10-
formyl-THF (Gregory, 1984). Folylpolyglutamates,
being large and strongly electronegative molecules,
are not transportable into cells and, before they can be
absorbed, they must be hydrolysed to monoglutamate
forms. None of the known proteases in saliva, gastric
juice or pancreatic secretions are capable of splitting
the γ-peptide bonds in the polyglutamyl side chain.
Polyglutamyl folate can, however, be hydrolysed by
folate conjugase, which is a trivial name for pteroylpolyglutamate
hydrolase, EC (also known
as folylpoly-γ-glutamyl carboxypeptidase). As much
as 50–75% of dietary polyglutamyl folate can be absorbed
after deconjugation to monoglutamyl folate
(Butterworth et al., 1969). The presence of conjugase
activity in many raw foods of both plant and animal
origin results in a high proportion of the dietary folate
being already monoglutamyl when presented to the
intestinal mucosa (Gregory, 1989).
Two folate conjugases have been found in human
jejunal tissue fractions. One, a brush-border exopeptidase,
has a pH optimum of 6.7–7.0 and is activated by
Zn2+. The other, an intracellular endopeptidase of
mainly lysosomal origin, has a pH optimum of 4.5
and no defi ned metal requirement. The brush-border
conjugase splits off terminal glutamate residues one
at a time and is thought to be the principal enzyme in
the hydrolysis of polyglutamyl folate. Brush-border
conjugases from the jejunum of the human and pig
possess similar enzymatic properties (Gregory et al.,
1987) and thus the porcine enzyme can be used to
study folate bioavailability in humans. Interestingly,
the human and the pig are the only species in which
intestinal brush-border conjugase activity has been
demonstrated. Bhandari & Gregory (1990) showed
that extracts from certain foods (e.g. legumes, tomatoes
and orange juice) can inhibit brush-border conjugase
activity from human and porcine intestine in vitro. Organic acids may be responsible for this inhibition
(Wei & Gregory, 1998). Such inhibition may be
a factor affecting the bioavailability of polyglutamyl
folates in diets containing these foods. The intracellular
conjugase may play no role in the digestion of
dietary folate, being, instead, concerned with folate
metabolism within the enterocyte (Halsted, 1990).
Signifi cant conjugase activity has also been reported
in the pancreatic juice of pigs and humans (Gregory,
1995). Bhandari et al. (1990) found the porcine pancreatic
enzyme to be Zn2+-dependent with maximum
activity at pH 4.0–4.5. Feeding stimulated secretion of
pancreatic juice, including conjugase activity. Chandler
et al. (1991) calculated that conjugase activity
in porcine pancreatic juice was minor relative to the
activity of the jejunal brush-border conjugase.


In 1931, a research group led by Lucy Wills showed
that an autolysed yeast preparation (Marmite),
which was therapeutically ineffective against the
pernicious anaemia caused by vitamin B12 defi ciency,
was effective against nutritional megaloblastic anaemia
in pregnant women. These researchers induced
a similar anaemia in monkeys which then responded
to crude liver extracts. Other substances that cured
specifi c defi ciency anaemias in monkeys and chicks
were isolated from yeast by different research groups
and assigned the names ‘vitamin M’ and ‘vitamin Bc’.
Another substance isolated from liver was shown to
be essential to the growth of Lactobacillus casei and
therefore called the ‘L. casei factor’. In 1941, Mitchell
and co-workers processed four tons of spinach leaves
to obtain a purifi ed substance with acidic properties
which was an active growth factor for rats and L. casei.
They named the factor ‘folic acid’ (from folium, the
Latin word for leaf). Eventually, all of the above substances
proved to be the same when Angier’s group in
1946 accomplished the synthesis and chemical structure
of folic acid.

The term ‘folate’ is used as the generic descriptor
for all derivatives of pteroic acid that demonstrate
vitamin activity in humans. The structure of the
parent folate compound, folic acid, comprises a bicyclic
pterin moiety joined by a methylene bridge to
p-aminobenzoic acid, which in turn is coupled via an
α-peptide bond to a single molecule of L-glutamic
acid (Fig. 17.1, top).
(Note: In the present context, the term ‘folic acid’
refers specifi cally to pteroylmonoglutamic acid
which, with reference to the pteroic acid and glutamate
moieties, can be abbreviated to PteGlu. ‘Folate’ is
a non-specifi c term referring to any folate compound
with vitamin activity. ‘Folacin’ is a non-approved term
synonymous with ‘folate’.)
Folic acid is not a common natural physiological
form of the vitamin. In most natural foods, the pteridine
ring is reduced to give either the 7,8-dihydrofolate
(DHF) or 5,6,7,8-tetrahydrofolate (THF) (see Fig. 17.1).
These reduced forms can be substituted with a covalently
bonded one-carbon adduct attached to nitrogen positions 5 or 10 or bridged across both positions. The
following substituted forms of THF are important
intermediates in folate metabolism: 10-formyl-THF, 5-
methyl-THF, 5-formimino-THF, 5,10-methylene-THF
and 5,10-methenyl-THF (see Fig. 17.1).
An important structural aspect of the 5,6,7,8-tetrahydrofolates
is the stereochemical orientation at
the C-6 asymmetric carbon of the pteridine ring. Of
the two stereoisomers, 6S and 6R (formerly called 6l
and 6d), only the 6S is biologically active and occurs in
nature. Methods of chemical synthesis of tetrahydrofolates,
whether by catalytic hydrogenation or chemical
reduction, yield a racemic product (i.e. a mixture
of both stereoisomers).
All folate compounds exist predominantly as polyglutamates,
containing typically from fi ve to seven
glutamate residues in γ-peptide linkage. The γ-peptide
bond is unique in mammalian biochemistry.
Folate conjugates are abbreviated to PteGlun derivatives,
where n is the number of glutamate residues; for
example, 5-CH3-H4PteGlu3 refers to triglutamyl-5-
methyltetrahydrofolic acid.
Methotrexate (4-amino-10-methylfolic acid; Fig.
17.2) is a folate antagonist which is used as an anticancer

Dietary sources
Polyglutamyl folate is an essential biochemical constituent
of living cells, and most foods contribute some
folate. The folates generally exist in nature bound to
proteins (Baugh & Krumdieck, 1971) and they are
also bound to storage polysaccharides (various types
of starch and glycogen) in foods (Cerná & Káš, 1983).
In the United States, dried beans, eggs, greens, orange
juice, sweet corn, peas and peanut products are good
sources of folate that are inexpensive and available all
the year round.

Humans and other mammals cannot synthesize
folate in their tissues and thus they must obtain the
vitamin from exogenous sources via intestinal absorption.
The intestine is exposed to two sources of
folate: (1) dietary folate and (2) folate synthesized
by bacteria in the large intestine. The latter source is
available to the host tissues through direct absorption
in the colon.
It is fundamental in folate metabolism that folate
monoglutamates are the circulatory and membrane-
transportable forms of the vitamin, whereas
polyglutamates are the intracellular biochemical and
storage forms.
17.4.1 Folate transport proteins
Cellular uptake of folate involves two functionally different
membrane transport proteins: (1) the reduced
folate carrier, which is an organic anion exchange
protein present in the plasma membrane of a wide
variety of cells, and (2) the less ubiquitous folate receptor,
which internalizes folate by a receptor-mediated
process. The affi nities of these proteins for folates
and antifolates differ signifi cantly (Table 17.2). After
internalization, folates are retained in the cytoplasm
by polyglutamylation.
In certain specifi c cell types, such as human placental
trophoblast cells, a functional coordination between
the two transport proteins has been proposed.
Enterocytes and hepatocytes lack the folate receptor
and so folate transport in the human intestine and
liver is mediated solely by the reduced folate carrier.

Dietary deficiency of Biotin

Biotin is so widely distributed in foods that it is doubtful
whether a true dietary defi ciency of the vitamin
has ever occurred in human adults capable of utilizing
it. Artificial biotin defi ciency states have been induced
in healthy volunteers by feeding low-biotin diets containing
a high proportion of raw egg white. An initial
dry scaly dermatitis was followed by non-specifi c
symptoms that included anorexia and extreme lassitude.
All of the symptoms responded to injections
of 150–300 μg of biotin per day.

A unique opportunity to study dietary biotin
defi ciency was presented to Baugh et al. (1968) by a
62-year-old female patient who had consumed six raw
eggs and 4 pints of skimmed milk daily for 18 months.
This diet had been recommended by a physician (illadvisedly,
as we now know) as a dietary supplement
to provide a high intake of essential amino acids to
aid liver regeneration following a diagnosis of liver
cirrhosis. During this dietary period, the patient took
vitamin supplements, which included biotin, and
she also received 100 μg of vitamin B12 by injection
monthly. Thus the stage was set, unintentionally, for
the development of biotin defi ciency due to the avidin
content of raw egg whites, uncomplicated by defi ciencies
of other vitamins or common nutrients. Clinical
manifestations included anorexia, nausea, vomiting,
glossitis, pallor, depression, lassitude, substernal pain,
scaly dermatitis and desquamation of the lips. All
symptoms cleared or improved markedly after 2 to
5 days of parenteral (by injection) vitamin therapy
providing 200 μg of biotin daily, while the patient
continued her pre-treatment diet. In contrast to other
case reports, the patient did not exhibit anaemia,
muscle pains, hypercholesterolaemia or electrocardiographic
Seborrhoeic dermatitis of the scalp and a more
generalized dermatitis known as Leiner’s disease have
been reported in breast-fed infants when the mother
is malnourished. These symptoms are relieved when
biotin is administered to the mother.

Inherited defects of biotin metabolism

There are two known congenital disorders of biotin
metabolism: (1) holocarboxylase synthetase (HCS)
defi ciency and (2) biotinidase defi ciency. Both disorders
are inherited as an autosomal recessive trait
and both lead to defi ciency of the four biotin-dependent
carboxylases, a condition known as multiple
carboxylase defi ciency (MCD). Because of the vital
role of these enzymes in protein, carbohydrate and
lipid metabolism, their defi ciency leads to severe lifethreatening
The two forms of MCD usually become symptomatic
in early infancy or childhood. The incidence of
biotinidase defi ciency is about one in 60 000 and that
of HCS defi ciency seems to be even lower. The underlying
cause of HCS defi ciency is decreased affi nity
of HCS for biotin resulting in reduced formation of
holocarboxylases with physiological concentrations
of biotin. In biotinidase defi ciency, MCD results from
progressive development of biotin defi ciency due to
inability to liberate biotin from the biocytin or short
biotinyl peptides that remain after metabolic degradation
of biotin-containing carboxylases. The recycling
of biotin salvaged from degraded enzymes is essential
to maintain an adequate supply of the vitamin. A lack
of biotinidase results in excessive urinary excretion of
biocytin and this raises the requirement for biotin to
above normal intakes.
The two forms of MCD differ biochemically in that
the HCS-defi cient patients have normal plasma biotin
concentrations but decreased carboxylase activities,
whereas patients with biotinidase defi ciency have
subnormal plasma biotin concentrations and normal
carboxylase activities. MCD arising from either inherited
defect causes a block in the biotin-dependent
metabolic pathways with characteristic accumulation
and urinary excretion of organic acids such as lactate,
3-hydroxyisovalerate, 3-methylcrotonylglycine and

The clinical presentation and age of onset of MCD
are extremely variable. HCS defi ciency may present
in the fi rst days of life, while biotinidase defi ciency
usually becomes manifest between the second and
fi fth months of age, depending on the amount of free
biotin in the diet. However, onset of HCS defi ciency
is delayed (2–21 months) in some cases and therefore
classifying the two disorders as neonatal- or earlyonset
and infantile- or late-onset MCD should be discouraged.
Clinical symptoms common to both disorders
include neurological abnormalities (hypotonia,
seizures, ataxia) and cutaneous changes (skin rash,
alopecia). In healthy persons receiving an adequate
diet, biotinidase activity in the brain is relatively very
low (Suchy et al., 1985) and so the brain relies largely
on biotin that is transferred across the blood–brain
barrier. This feature of the brain could explain the
rapid onset of neurological symptoms observed in
biotinidase deficiency.


The discovery and recognition of biotin as a member
of the water-soluble vitamin B complex resulted
from several independent lines of investigation. In
1933, Franklin E. Allison and his colleagues at the US
Department of Agriculture reported that the growth
and respiration of Rhizobium trifolii, a nitrogen-fi xing
bacterium found in the root nodules of legumes,
were stimulated by a factor, ‘coenzyme R’, extractable
from various organic sources. By the early 1920s, several investigators had isolated from various organic
sources crude fractions that contained a novel
growth factor for yeast. Eventually, in 1936, Fritz
Kögl and B. Tönnis, organic chemists at the University
of Utrecht in Germany, isolated from dried egg
yolk a crystalline substance that strongly stimulated
the growth of yeast. This growth factor, which Kögl
and Tönnis named ‘biotin’, was later shown to have
exactly the same stimulatory effect on Rhizobium as
coenzyme R. In this respect, at least, the two factors
were identical.

Further progress came from the fi eld of animal nutrition.
In 1927, Margaret A. Boas at the Lister Institute
of Preventive Medicine in London observed toxicity
in rats when raw egg white was used as a source of
protein in the animals’ diet. After a few weeks the rats
developed dermatitis and haemorrhages of the skin,
their hair fell out, their limbs became paralysed, they
lost considerable weight, and eventually they died.
Only raw or cold-dried egg white produced the toxicity;
cooking made the egg white harmless. This toxicity,
which Boas called egg white injury, was prevented
by a ‘protective factor X’ present in liver and other
sources. Paul György showed that biotin had the same
protective action against egg white injury as did protective
factor X, which he renamed ‘vitamin H’ (German
Haut, skin). György also showed that vitamin H
concentrates supported the growth of biotin-requiring
bacteria. In 1940 György and Vincent du Vigneaud
independently isolated crystalline vitamin H from
liver concentrates. It was soon proven that biotin and
vitamin H were one and the same compound.
The chemical structure of biotin was established by
du Vigneaud’s group in 1942 and in the following year
the vitamin was synthesized at the Merck Company,

Biotin is present in all natural foodstuffs, but the
content of even the richest sources is very low when
compared with the content of most other water-soluble
vitamins. Biotin is not commonly used in fortifi ed
foods, apart from infant formulas. Typical values of
some rich natural sources of biotin include ox liver
(33 μg per 100 g), whole eggs (20 μg per 100 g), dried
soya beans (65 μg per 100 g) and peanuts (72 μg per
100 g) (Holland et al., 1991). Other good sources include
yeast, wheat bran, oatmeal and some vegetables.
Muscle meats, fi sh, dairy products and cereals contain
smaller amounts, but are important contributors to
the dietary intake. Most of the biotin content of animal
products, nuts, cereals and yeast is in a protein-bound
form. A higher percentage of free, water-extractable
biotin occurs in vegetables, green plants, fruit, milk
and rice bran (Lampen et al., 1942).

Vitamin B6 deficiency

Vitamin B6 is widely distributed in foods, and any diet
so poor as to be insuffi cient in this vitamin would
most likely lack adequate amounts of other B-group
vitamins. For this reason, a primary clinical defi ciency
of B6 in the adult human is rarely encountered under
normal circumstances.
In a well-controlled study conducted by Hodges et
al. (1962), six healthy male volunteers were divided
into pairs and given a basic formulated diet administered
by nasogastric tube twice daily. The fi rst pair of
men received a complete formula including pyridoxine;
the second pair received the same diet without
pyridoxine; and the third pair were given the anti-vitamin
deoxypyridoxine in addition to the pyridoxinefree
diet. The men receiving pyridoxine-free diets, but
not those receiving complete diets, developed adverse
symptoms and signs of illness, which were more
severe in the men given the anti-vitamin. The most
obvious symptoms were gastrointestinal disturbances
and epithelial changes. Both men in the anti-vitamin
group had scaling of the skin, foul breath, severe gingivitis,
soreness and discoloration of the tongue and
dry cracked lips. No objective neurological changes
could be demonstrated. After vitamin B6 was restored
to their diet and the anti-vitamin discontinued, one
man recovered promptly and the other recovered
Vitamin B6 deprivation imposed at certain stages of
brain development interferes with the orderly process
of neuronal development (Kirksey et al., 1990). In the
1950s, an occurrence of convulsions in infants was
traced to an unfortifi ed liquid milk-based canned formula
that had undergone autoclaving in manufacture
(Coursin, 1954). There is some circumstantial evidence
that convulsions resulting from vitamin B6 defi
ciency may be caused by an insuffi cient production
of γ-aminobutyric acid, the major neurotransmitter
in the brain (Ebadi, 1978). However, a meaningful
correlation among vitamin B6 defi ciency, concentration
of γ-aminobutyric acid and convulsion has not
been established.

Animal studies for Vitamin B6

Kumar & Axelrod (1968) reported a lowered level of
circulating antibodies and a dramatic reduction in
the number of antibody-forming cells in the spleens
of vitamin B6-defi cient rats immunized with sheep
erythrocytes. This decreased cellular immune response
was independent of the inanition associated
with the defi ciency and was restored to normal by the
administration of PN shortly before immunization.
Robson & Schwarz (1975) reported a dramatic
85–95% reduction in the number of thoracic duct
lymphocytes and a signifi cant reduction in cellular
immunocompetence in vitamin B6-defi cient rats.
These conclusions were based on the results of two
tests: (1) the in vitro mixed lymphocyte reaction
(MLR) and (2) the in vivo normal lymphocyte transfer
reaction (NLT). In the MLR, lymphocytes from test
Lewis strain rats (in this case, vitamin B6-defi cient and
control rats) are cultured with genetically dissimilar
lymphocytes taken from normally nourished F1 hybrid
rats. If the lymphocytes are immunocompetent,
they will become activated and then they will proliferate
and transform into the larger lymphoblasts. The
extent of blastogenesis is quantitated by exposing
the cultures to [3H]thymidine and then measuring
the incorporation of the radioactivity into DNA. In
the NLT, lymphocytes from donor Lewis rats (the test
rats) are injected into the ventral abdominal wall of F1
hybrid rats. Immunologically competent donor cells
produce a graft-versus-host reaction in the skin of the
F1 rat. The impaired proliferation of lymphocytes and
loss of cellular immunocompetence may perhaps be
attributed to a cessation of T-lymphocyte development
within the thymus of the vitamin B6-defi cient
The development of functional T lymphocytes
depends on humoral factors secreted by thymic epithelial
(TE) cells. To investigate the effects of dietary
vitamin B6 defi ciency on TE cell function, Willis-Carr
& St. Pierre (1978) used three groups of Lewis strain
rats as cell donors: (1) normal (control) rats, (2) rats
maintained for 2 weeks on a vitamin B6-defi cient diet
and (3) rats whose thymus glands had been surgically
removed 24 hours after birth (neonatally thymectomized
rats). Spleen, bone marrow and mesenteric
lymph nodes were removed from each donor and
washed cells from these lymphoid tissues were exposed
to monolayers of TE cells. The TE monolayers
were made from (1) normal, (2) vitamin B6-defi cient
and (3) ‘post B6’ rats, i.e. rats placed back on a regular
diet for 3 weeks after the original 2-week B6-defi cient
diet. Exposure of T-lymphocyte precursors from
B6-defi cient or neonatally thymectomized donors to
normal TE monolayers resulted in their conversion to
functional T lymphocytes, as measured by their response
in MLR and to mitogens. However, TE monolayers
from B6-defi cient rats were unable to effect such
a maturation of T lymphocytes. When the defi cient
rats were returned to a normal diet, TE cell function
was restored. The authors suggested that the cause
of defective cellular immunocompetence following
vitamin B6 deprivation is the inability of TE cells to
effect the differentiation of T-lymphocyte precursors
to functional T lymphocytes. Vitamin B6 defi ciency
did not impair T-lymphocyte precursors, which could
be stimulated to differentiate by exposure to normal
TE cell monolayers. Presumably, the observed effect
of vitamin B6 defi ciency is due to a blocking of the
biosynthesis and/or release of a humoral factor that is
produced by TE cells. Chandra & Puri (1985) found
signifi cantly reduced serum thymic factor activity in
rats fed diets restricted in vitamin B6 and given 4-deoxypyridoxine
hydrochloride in their drinking water.
Vitamin B6-defi cient mice exhibited impaired
production and reduced activity of cytotoxic T lymphocytes
(Sergeev et al., 1978; Ha et al., 1984). Antibody-
mediated cytotoxicity, macrophage phagocytosis
and natural killer cell activity were not affected by
the level of vitamin B6 intake (Ha et al., 1984).

Regulation of steroid hormone action

Steroid hormones (androgens, oestrogens, progestins
and the corticosteroids) are able to enter cells, bind
to receptors and directly regulate gene transcription.
The synthesized proteins carry out the ultimate effect
of the hormone. Evidence accrued in the following
studies suggest a physiological role for vitamin B6 in
modulating steroid hormone action.
The infl uence of B6 vitamers and analogues on the
physical properties of the glucocorticoid receptor
has been studied using in vitro receptor preparations
(Allgood et al., 1990a). The combined data showed
that, among these compounds, only PLP can directly
associate with the glucocorticoid receptor and alter
several of its properties, including molecular conformation,
surface charge, susceptibility to exogenous
proteolysis, DNA binding capacity and subcellular
localization. The last two properties are requisites for
regulation of target gene expression.
In a number of rat studies, Bender’s group produced
in vivo evidence that vitamin B6 may be involved in the
normal physiological action of steroid hormones. In
male rats, Symes et al. (1984) showed that the uptake
and accumulation of tracer [1,2,6,7–3H]testosterone
in the nucleus of the prostate gland were signifi cantly
increased in vitamin B6-defi cient animals compared
with vitamin B6-adequate controls. In a corresponding
study of female rats (Bowden et al., 1986), the
animals were segregated according to the phase of
the oestrous cycle to avoid the inherent variations of
both plasma concentration of oestrogen and the concentration
of oestrogen receptor in the uterus during
the course of the oestrous cycle. As for testosterone in
the male, uptake and accumulation of tracer [2,4,6,7–
3H]17β-oestradiol in uterine nuclei were signifi cantly
increased in vitamin B6-defi cient animals throughout
the oestrous cycle; there were no signifi cant differences
at anoestrus.
Bender’s group also found evidence of enhanced
sensitivity to steroid hormone action in vitamin B6-
defi cient rats of both sexes. In the male, testosterone is
secreted by the interstitial cells of Leydig in the testes,
but only when these cells are stimulated by luteinizing
hormone (LH) released by the anterior pituitary
gland in response to hypothalamic gonadotropin-releasing
hormone (Gn-RH). Circulating testosterone
exerts negative feedback control at the level of the
hypothalamus, switching off the supply of pituitary
LH and thereby stopping testicular secretion of testosterone
(Fig. 14.9). Symes et al. (1984) found that
the plasma concentration of testosterone in vitamin
B6-defi cient male rats was only 25% of that in vitamin B6-adequate controls. This unexplainable reduction
in plasma testosterone was not accompanied by a
reduction in the relative weight of the prostate gland
as might have been expected; neither was it accompanied
by a rise in plasma LH. These two observations
suggest that there may be enhanced sensitivity of the
hypothalamus to negative feedback by testosterone in
vitamin B6 defi ciency, leading to normal (or reduced)
plasma concentrations of LH and normal growth of
the prostate despite considerably reduced circulating
concentrations of testosterone.
In the female rat, ovarian secretion of oestrogen is
stimulated by LH and follicle-stimulating hormone
(FSH) released from the anterior pituitary gland in
response to hypothalamic Gn-RH. During most of
the oestrous cycle, circulating oestrogen exerts negative
feedback control at the level of the hypothalamus,
suppressing the release of LH and FSH. The major
event of ovulation is preceded by a massive outfl ow
of LH from the pituitary (the pre-ovulatory surge)
caused by positive feedback of oestrogen upon the
hypothalamus. Bowden et al. (1986) reported that in
ovariectomized rats, doses of ethynyl-oestradiol that
had no effect on circulating LH in control animals
(i.e. submaximal doses) lowered plasma LH levels in
vitamin B6-defi cient animals. As in the male rat, this
suggests that vitamin B6 defi ciency leads to enhanced
sensitivity of the hypothalamus to negative feedback
by steroid hormone.
Allgood et al. (1990b) investigated the infl uence of
PLP on glucocorticoid receptor-dependent gene expression
by introducing a reporter gene with a defi ned promoter into a cell culture line. The results showed
that, under conditions of moderate vitamin B6 defi -
ciency, the glucocorticoid receptor becomes a more
effi cient activator of gene transcription. Conversely,
high concentrations of vitamin B6 suppress activation
of transcription. The modulatory effects of PLP
concentration occurred through a novel mechanism
that did not involve changes in glucocorticoid receptor
mRNA or protein levels, or the receptor’s ligand
binding capacity. Analogous effects of PLP were
found with the oestrogen, androgen and progesterone
receptors (Allgood & Cidlowski, 1992). Vitamin
B6 appears to modulate steroid hormone-mediated
gene expression through its infl uence on a functional
or co-operative interaction between steroid hormone
receptors and the transcription factor NF1 (Allgood
et al., 1993).

Brain homeostasis for Vitamin B6

Vitamin B6 levels in brain are homeostatically regulated.
Although it is relatively easy to produce symptomatic
vitamin B6 deficiency in animals, levels of the
vitamin in brain (and heart) are somewhat better
maintained in defi ciency states than they are in liver
and kidney. Conversely, massive parenteral doses
(daily intravenous injections of 200 mg kg–1 of PN for
3 days to rabbits) elevated the brain levels of PLP by an
average of only 39% (Spector & Shikuma, 1978).
In brain, vitamin B6 exists predominantly in the enzymatically
active forms PLP and PMP at concentrations
much higher than in plasma and cerebrospinal
fl uid (CSF). In rabbits, the concentrations of vitamin
B6 in plasma, CSF, brain and choroid plexus were, respectively,
0.30, 0.39, 8.90 and 15.10 μmol L–1 or kg–1
(Spector, 1978a).
Spector studied the in vitro uptake and release of
tritium-labelled vitamin B6 in rabbit brain slices and
isolated choroid plexuses (Spector, 1978b; Spector
& Greenwald, 1978). Uptake of [3H]PN by both tissues
was inhibited by (1) low temperature (2°C) and
dinitrophenol, demonstrating energy dependence;
(2) pyridoxal azine, demonstrating dependence on
the activity of intracellular pyridoxal kinase; and (3)
unlabelled non-phosphorylated B6 vitamers and, to
lesser extent, phosphorylated B6 vitamers, demonstrating
saturability of the uptake system. There was
no detectable metabolism of [3H]PN to [3H]pyridoxic
acid in brain slices or choroid plexus. From 70 to 80%
of the labelled vitamin B6 in both tissues was phosphorylated
after a 30-minute incubation in [3H]PN.
Phosphorylated B6 vitamers were taken up much less
readily than non-phosphorylated vitamers. These
studies are not conclusive in separating active transport
from metabolic trapping because both pyridoxal
kinase and active transport require ATP, and therefore
depletion of ATP could affect either process. Furthermore,
dinitrophenol is known to inhibit mammalian
pyridoxal kinase as well as preventing ATP synthesis
by uncoupling oxidative phosphorylation from electron
fl ow through the electron-transport chain.
The activity of pyridoxal kinase in brain is unimpaired
by moderate and severe vitamin B6 defi ciency
(McCormick et al., 1961). Spector & Shikuma (1978)
showed that pyridoxal kinase activity and vitamin B6
accumulation by brain slices and choroid plexus are
not affected by various drugs that alter the concentrations
of PLP or biogenic amines in brain.
Spector (1978a) showed that, during one pass
through the cerebral circulation, [3H]PN was cleared
from the circulation no more rapidly than mannitol.
Mannitol, a molecule of similar size and shape to
PN, is known to be transported by diffusion. Spector
(1978a) also confi rmed in vivo, by injection directly
into the ventricular CSF of rabbits, that non-phosphorylated
B6 vitamers enter brain cells by a saturable
accumulation process. Kinetic studies conducted by
Spector & Greenwald (1978) revealed marked differences
in the uptake of vitamin B6 by choroid plexus
and brain. The half-saturation concentrations and
rate maxima for accumulation were ~0.2 μM and
1.0–2.0 μmol kg–1 per 30 min for brain slices and
7.0 μM and 40 μmol kg–1 per 30 min for isolated
choroid plexus. Assuming the absence of a membrane
carrier for vitamin B6 uptake, the kinetic constants
refer to the binding of substrate to pyridoxal kinase
and are therefore values of Km and Vmax. The differences
in the constants for vitamin B6 uptake by brain
cells and choroid plexus are presumably due to factors
(e.g. intracellular pH) that cause variations in enzyme
This makes the choroid plexus, rather than brain cells,
the likely source of the phosphorylated B6 vitamers
in CSF.
Vitamin B6 is transported in the reverse direction
(i.e. from brain and/or CSF into blood) more rapidly
than mannitol (Spector, 1978a). This suggests that the
transport mechanism for vitamin B6 in this direction
involves a mechanism other than simple diffusion.
In conclusion, Spector’s data show that circulating
vitamin B6 can enter the brain via the blood–CSF
barrier (choroid plexus). The fi nding that PN was
extracted no more rapidly than mannitol during one
pass through the cerebral circulation argues against
signifi cant entry of PN via the blood–brain barrier.
There is a saturable transport system (Km = 0.7 μM)
within the choroid plexus that regulates the entry
of free (unbound) non-phosphorylated B6 vitamers
from plasma into the CSF. The vitamers fi nds their
way into the extracellular space of brain and enter
brain cells by a high-affi nity (Km = ~0.2 μM) saturable
accumulation system. The transport system in both
choroid plexus and brain cells appears to be simple
(or possibly facilitated) diffusion accelerated by the
concentration gradient created by phosphorylation of
the transported B6 vitamers (metabolic trapping). PN,
PL and PM have comparable affi nity for the vitamin
B6 transport systems, as they also do for pyridoxal
kinase. Excessive concentrations of phosphorylated
B6 vitamers within brain cells are dephosphorylated
intracellularly and transported out of the cells.