|
|
||||||||
Review Article |
1 From the Department of Molecular and Cellular Biology, University of California, Berkeley (BNA, IE-S, and EAS), and the Children's Hospital Oakland Research Institute, Oakland, CA (BNA and IE-S).
2 Supported by grants to BNA from the Ellison Foundation (SS0422-99), the National Foundation for Cancer Research (M2661), the Wheeler Fund of the Dean of Biology, and the National Institute of Environmental Health Sciences Center (ES01896).
3 Address reprint requests to BN Ames, Children's Hospital Oakland Research Institute, 5700 Martin Luther King Jr Way, Oakland, CA 94609. E-mail: bames{at}chori.org.
| ABSTRACT |
|---|
|
|
|---|
Val) [DNA: C-to-T substitution at nucleo-tide 677 (677C
T)] in methylenetetrahydrofolate reductase (NADPH) and the cofactor FAD (in relation to cardiovascular disease, migraines, and rages), the Pro187
Ser (DNA: 609C
T) mutation in NAD(P):quinone oxidoreductase 1 [NAD(P)H dehy-drogenase (quinone)] and FAD (in relation to cancer), the Ala44
Gly (DNA: 131C
G) mutation in glucose-6-phosphate 1-dehydrogenase and NADP (in relation to favism and hemolytic anemia), and the Glu487
Lys mutation (present in one-half of Asians) in aldehyde dehydrogenase (NAD + ) and NAD (in relation to alcohol intolerance, Alzheimer disease, and cancer).
Key Words: Genetic disease therapeutic vitamin use binding defect favism alcohol intolerance autism migraine headaches single nucleotide polymorphisms enzyme mutations review
| INTRODUCTION |
|---|
|
|
|---|
50 human genetic diseases involving defective enzymes can be remedied by high concentrations of the vitamin component of the coenzyme, and that this therapeutic technique can be applied in several other cases, including polymorphisms associated with disease risks, for which molecular evidence suggests that a mutation affects a coenzyme binding site. The nutrients discussed in this review are pyridoxine (618); thiamine (625); riboflavin (627); niacin (632); biotin (637); cobalamin (638); folic acid (641); vitamin K (643); calciferol (645); tocopherol (645); tetrahydrobiopterin (646); S-adenosylmethionine (646); pantothenic acid (646); lipoic acid (647); carnitine (647); hormones, amino acids, and metals (648); and maxi B vitamins (649).
The proportion of mutations in a disease gene that is responsive to high concentrations of a vitamin or substrate may be one-third or greater (13). Determining the true percentage from the literature is difficult because exact response rates in patients are not always reported and much of the literature deals only with individual case reports. The true percentages depend on several factors, such as the nature of the enzyme, the degree of enzyme loss that results in a particular phenotype, how much a small conformational change disrupts the binding site of the particular enzyme, whether the binding site is a hot spot for mutations, and whether dietary administration of the biochemical raises its concentration in the cell. From what is known of enzyme structure, it seems plausible that, in addition to direct changes in the amino acids at the coenzyme binding site, some mutations affect the conformation of the protein, thus causing an indirect change in the binding site.
An alternate form of a gene present in >1% of the population is called a polymorphism. Some polymorphisms that are associated with a phenotype have been shown to alter cofactor binding and affect a large percentage of the population (seeTable 1
for a list of the allelic frequencies of the polymorphisms discussed in this review). Our analysis of metabolic disease that affects cofactor binding, particularly as a result of polymorphic mutations, may present a novel rationale for high-dose vitamin therapy, perhaps hundreds of times the normal dietary reference intake (DRI) in some cases. This area should interest the entire health community because of the considerable percentage of the population affected by polymorphisms, many of which may have outlived their genetic usefulness. The setting of a DRI may become more complicated if a sizable percentage of the population in fact has a higher B-vitamin requirement because of a polymorphism. It seems likely that the examples listed in Table 1
will represent the beginning of a much longer list as genomics advances and awareness of remediable Km mutants increases.
|
There are
40000 human genes. Of the 3870 enzymes catalogued in the ENZYME database (6), 860 (22%) use a cofactor. Any cofactor used by many enzymes is of particular interest, such as the 8 vitamin-derived coenzymes discussed in this review. Although high-dose vitamin remediation seems to be routinely tried for diseases involving enzymes dependent on pyridoxal-P (PLP) and thiamine pyrophosphate (TPP), some of the vitamins, such as riboflavin, pantothenate, folate, and niacin, deserve more attention. Thus, www.KmMutants.org is also intended for the input of physicians, who may examine the benefits of high-dose multivitamin treatment (see the section on maxi B vitamins) for mental or metabolic disorders of unknown cause or report side effects of vitamin treatment. Provided safe dosages are used (Table 2
), there is potentially much benefit and possibly little harm in trying high-dose nutrient therapy because of the nominal cost, ease of application, and low level of risk. Most of the vitamins discussed here appear safe in relatively high doses because the body can discard excess.
|
10 mg/d and symptoms of toxicity can appear at 100 mg/d (10). Ascorbate concentrations are tightly regulated in young men and women and there is an upper limit on steady state plasma ascorbate concentrations of
80 µmol/L (11, 12). At the end of each vitamin section, we discuss the toxicity of the vitamin as well as data on raising tissue concentrations. Note that some reports discussed below did not ascertain the minimal necessary therapeutic level of treatment, but instead used high doses that were thought would produce the desired effect. This treatment strategy is obviously not feasible with all nutrients (because of possible toxicity). Additionally, it is likely that not all administered vitamin is absorbed at very high doses (see the discussion of tissue concentrations and toxicity in the section on riboflavin). In an ideal situation, the lowest adequate therapeutic dosage would be elucidated and used. Nutritional interventions to improve health are likely to be a major benefit of the genomics era. Many coenzyme binding motifs have been characterized, and essential residues for binding have been elucidated. Structural data can be found at the beginning of many sections. It will soon be possible to identify the complete set of genes having cofactor binding sites and the polymorphisms that fall into these regions, with an end goal of using vitamins, and possibly amino acids, hormones, and minerals, to effect a metabolic "tune-up."
Support for some of the views discussed here can be found in the literature. It is clear that many individual researchers have recognized that high-dose vitamin treatment is effective in particular diseases because a mutation affects the affinity of an enzyme for its coenzyme. In particular, Linus Pauling (13) hypothesized in his review entitled Orthomolecular Psychiatry that much mental disease may be due to insufficient concentrations of particular biochemicals in the brain as the result of an inadequate intake of particular micronutrients and that some brain dysfunction may be due to mutations that affect the Km of enzymes: "The still greater disadvantage of low reaction rate for a mutated enzyme with K[m] only 0.01 could be overcome by a 200-fold increase in substrate concentration to [S] = 400. This mechanism of action of gene mutation is only one of several that lead to disadvantageous manifestations that could be overcome by an increase, perhaps a great increase, in the concentration of a vital substance in the body. These considerations obviously suggest a rationale for megavitamin therapy." More recently, high-dose pyridoxine therapy has been suggested as a treatment for improving dysphoric psychological states (eg, loneliness, anxiety, hostility, and depression) by stimulating the production of 2 pyridoxine-dependent neurotransmitters, serotonin and
-aminobutyric acid (14).
Although he does not discuss binding defects, Roger Williams (15), another pioneer in the field of biochemical nutrition, also recognized that higher doses of vitamins may be necessary to accommodate for what he calls biochemical individuality: "Individuality in nutritional needs is the basis for the genetotrophic approach and for the belief that nutrition applied with due concern for individual genetic variations, which may be large, offers the solution to many baffling health problems. This certainly is close to the heart of applied biochemistry." [Human genetic variation appears greater than previously thought (16).] Williams' conclusions suggest that genetic and thus biochemical individuality necessitates much nutritional individuality. This is especially relevant in the dawning age of genomics, in which it will someday become routine to screen individuals for polymorphisms and thus treat persons more efficaciously by genotype, rather than just by phenotype.
It also appears that, during aging, oxidation deforms many proteins, thereby decreasing their affinity for their substrates or coenzymes (17). Mechanisms of protein deformation include direct protein oxidation, adduction of aldehydes from lipid peroxidation, and, in the case of membrane proteins, decreases in fluidity of oxidized membranes. This oxidative decay is particularly acute in mitochondria (1820). Thus, feeding high amounts of several mitochondrial biochemicals may reverse some of the decay of aging (17, 2126). Fourteen genetic diseases due to defective mitochondrial proteins are discussed in this review.
The impetus for this review arose while teaching an undergraduate laboratory course in which the students isolated bacterial mutants that grew on a complex medium but not on a minimal medium and characterized the defective gene and pathway. An appreciable percentage of mutant phenotypes could be explained by an increased Km (decreased affinity) of an enzyme, which could then be remedied by higher concentrations of the coenzyme or substrate (27).
| PYRIDOXINE (VITAMIN B-6) |
|---|
|
|
|---|
|
-pyrroline-5-carboxylic acid, which is then converted into proline. Defects in OAT lead to gyrate atrophy of the choroid and retina, an autosomal recessive disease that affects persons of all ages (see OMIM 258870). The disease is characterized by slowly progressive chorioretinal degeneration leading to blindness. Ornithine accumulates 10- to 15-fold when the enzyme is defective and appears to be responsible for much of the pathology of gyrate atrophy (29). In the pyridoxine-responsive forms of the disease, for which doses ranged from 10 to 750 mg/d, it appears that the defective enzyme has a Km defect for PLP; ornithine accumulation is decreased when patients are given high doses of pyridoxine.
The OAT activity in fibroblast extracts of a pyridoxine-responsive patient with the alanine-to-valine substitution at codon 222 (Ala226
Val) increased from 9 to 44 nmol productmg-1h-1 when the concentration of PLP in the assay was increased to 600 µmol/L. The Km from the cell line of a second patient with the same Ala226
Val mutation was 122 µmol/L (control Km: 6 µmol/L). Similar to cells from the second patient, Chinese hamster cells expressing an OAT complementary DNA (cDNA) producing the Ala226
Val protein also exhibited increased OAT activity with the addition of pyridoxine (30).
Vitamin B-6responsive and -nonresponsive patients with gyrate atrophy were shown to have different point mutations resulting in single amino acid changes in the mature enzyme (31). After incubation with 40 µmol PLP/L, OAT activity increased substantially more in fibroblasts from carriers of the pyridoxine-responsive variant than in fibroblasts from control subjects and nonresponsive patients (32). These investigators concluded that "the greater increases in activity seen in pyridoxine-responsive cells when PLP was added to the assay suggest both that the holoenzyme content in these cells is decreased owing to low affinity and that PLP binding to the apoenzyme occurs at a higher concentration."
In another study, 3 patients responded to oral vitamin B-6 (600750 mg/d) with a decrease in serum ornithine and a return to normal of reduced concentrations of serum lysine. Lower doses of vitamin B-6 (1830 mg/d) appeared to work just as well as the high doses (33).
In another study of 9 patients with gyrate atrophy (34), 4 patients responded to pyridoxine, which lowered serum ornithine by
50% in 3 cases. The Km (of OAT for PLP) was 23 µmol/L in the control subjects, 23 µmol/L in the pyridoxine-nonresponsive patients, and 168 µmol/L in the pyridoxine-responsive patients. This higher Km for pyridoxine-responsive patients could be explained by mutations in the binding site that severely reduce coenzyme affinity, whereas nonresponsive patients may harbor more severe mutations that affect a different area of the enzyme.
In a study of Japanese patients in which 1 of 7 patients (all with different mutations) responded to vitamin B-6, the pyridoxine-responsive mutation was found to be Thr181
Met, but the affinity for PLP was not measured (35). In another Japanese study, one patient (of 3) responded to vitamin B-6 (300600 mg/d) with a 60% reduction in serum ornithine concentrations. OAT activity in the fibroblasts from this patient increased up to 25% of normal levels in the presence of 2000 µmol PLP/L, although no significant improvement was observed in acuity or visual field. Thus, vitamin B-6 responsiveness may be due to a mutation in OAT that results in a high Km for PLP (36, 37). A Glu318
Lys mutation of the OAT gene was found in 3 heterozygous patients and 1 homozygous patient, all of whom were vitamin B-6-responsive according to previous in vivo and in vitro studies. Dose-dependent effects of the Glu318
Lys allele were observed in the homozygotes and heterozygotes in 1) OAT activity, 2) increase of OAT activity in the presence of PLP, and 3) apparent Km for PLP with these values approximately doubled in the homozygous individual compared with the heterozygotes. Thus, the highest residual level of OAT activity and mildness of clinical disease correlated directly with the higher number of the mutant Glu318
Lys allele found in the homozygous patient (38).
Many case reports of gyrate atrophy exist; as of 1995, pyridoxine-responsiveness had been observed in 7 of the
150 total documented cases (7/150 = 5%) (39). Whether pyridoxine treatment was actually attempted in each of these cases is unclear. Thus the true response rate may be higher or lower than 5%.
Cystathionine ß-synthase: homocystinuria
Cystathionine ß-synthase (CBS) of the transsulfuration pathway catalyzes the PLP-dependent condensation of homocysteine and serine to form cystathionine. Individuals carrying a defective form of this enzyme (see OMIM 236200) accumulate homocysteine in the blood and urine and display a wide range of symptoms that appear to be due to homocysteine toxicity, including mental retardation, vascular and skeletal problems, and optic lens dislocation. Barber and Spaeth (40) were the first to report pyridoxine-responsiveness with a complete return to normal of the patient's methionine and homocysteine concentrations in plasma and urine. They speculated that "if the deficient enzymatic activity were due to decreased affinity of a defective apoenzyme for its cofactor, activity might be restored by increasing the intracellular concentration of pyridoxal phosphate" (40).
Kim and Rosenberg (41) showed that CBS activity was 5% of that of control subjects in pyridoxine-responsive homocystinuric patients, who had markedly elevated plasma and urinary concentrations of methionine and homocystine. The mutant synthases had a 20-fold lower affinity for PLP. A 2- to 3-fold increase in the Km for homocysteine and serine was found in one vitamin B-6-responsive patient, although the Km for PLP was not measured. The maximum reaction rate (Vmax) was also reduced. It was suggested that pharmacologic doses of pyridoxine led to increased cellular concentrations of PLP and increased enzymatic activity (41).
One group showed cell lines from pyridoxine-responsive patients to have higher Km values for PLP (155, 145, 195, and 200 µmol/L) than control values (52, 52, and 85 µmol/L), whereas nonresponsive patients had the highest values (990 and 4000 µmol/L). It was noted that, in general, about one-half of CBS-deficient patients respond to pyridoxine with a lowering of homocysteine and serine concentrations to normal (42). A 21-y-old pyridoxine-responsive individual had a 3- to 4-fold elevated apparent Km of CBS for PLP as measured in fibroblast extracts (43). An Ala114
Val substitution was present in this individual, which is only 5 residues away from the lysine residue, Lys119, that binds PLP. These investigators concluded that in vivo responsiveness in individuals with some residual CBS activity is related both to the affinity of the mutant aposynthase for PLP and to the capacity of cells to accumulate PLP (43).
In one report (44), a G-to-A substitution at nucleotide 797 (797G
A; amino acid substitution: Arg266
Lys) was found in most pyridoxine-responsive patients. Seven of 12 patients were responsive to pyridoxine (40900 mg/d), which greatly decreased total plasma homocysteine.
Pyridoxine (501000 mg/d) markedly reduced homocysteine excretion in a group of pyridoxine-responsive patients: patient 1, 867 to 10 µmol/d; patient 2, 1021 to 79 µmol/d; patient 3, 15 µmol/L (blood concentration) to undetectable concentrations; and patient 4, plasma amino acids reverted to normal (45). It appears that a missense mutation (Ile278
Thr) is common (41%) in pyridoxine-responsive patients and that patients who are responsive to pyridoxine usually have a milder clinical phenotype than do nonresponsive patients.
The idea that pyridoxine-responsive patients have an increased Km was supported in a review of the mechanism of pyridoxine-responsive disorders (46). A review of the CBS deficiencies (2) found 629 patients in the literature: 231 (37%) were vitamin B-6responsive, 231 (37%) were vitamin B-6nonresponsive, 67 (11%) were intermediate in response, and 100 (16%) had not been classified. A decade later, the field was reviewed again and it was suggested that dosages of pyridoxine of 500 mg/d for 2 y appear to be safe, but that 1000 mg/d should not be exceeded (47).
A database of mutations in CBS (48) lists and maps >100 pathogenic mutations (including >70 missense mutations) that span all 7 exons of the CBS gene. Although the Schiff-base forming lysine has been assigned to nucleotide 119 in exon 3, it is difficult to say which domains are responsible for PLP binding. Serine should be tested clinically in addition to pyridoxine for treating patients, with the use of homocysteine concentrations as a measure of efficacy, because the Km of CBS for serine was shown to be increased in several cases (41). Oral serine administration (500 mgkg-1d-1) raises serine concentrations in plasma and cerebrospinal fluid (49), although very high doses (1400 mgkg-1d-1) can result in adverse effects (50).
Vitamin B-6-therapy may be valuable in more than just severe homozygous CBS-deficient cases: heterozygous parents of CBS-deficient patients also have significantly increased homocysteine concentrations (51). Increased homocysteine is a risk factor for cardiovascular disease (52). Heterozygosity for CBS deficiency may be present in 1% or 2% of the population.
Erythroid specific
-aminolevulinic acid synthase: X-linked sideroblastic anemia
Erythroid specific
-aminolevulinic acid synthase (ALAS2; 5-aminolevulinate synthase), with its PLP cofactor, is located in the mitochondria of animal cells and catalyzes the condensation of glycine and succinyl-CoA to form
-aminolevulinic acid, the first and rate-limiting step in the series of reactions that makes heme for incorporation into hemoglobin. Defects in ALAS2 are responsible for the most common inherited form of sideroblastic anemia, which is X-linked (see OMIM 301300). Because iron is transported to the mitochondria whether or not it is combined with heme, deficiencies in heme lead to iron deposits in erythroblast mitochondria and increased ringed sideroblasts in the marrow (53). About one-third of patients with sideroblastic anemia respond to pyridoxine (1), with doses ranging from 50 to 600 mg/d.
Three generations in a family originally described by Cooley in 1945 were found to have an ALAS2 gene with an A-to-C point mutation that results in an ALAS2 variant with reduced activation by PLP. The specific activity of the mutant enzyme was
26% of normal in the presence of 5 µmol PLP/L. The PLP cofactor activated or stabilized the purified mutant enzyme in vitro, consistent with the pyridoxine-responsive anemia in affected patients. It was hypothesized that the mutation alters the local secondary structure and possibly perturbs the overall conformation, thus decreasing stability, reducing the affinity for PLP, or both (54).
A point mutation (663G
A) in the ALAS2 gene of an 8-mo-old Japanese male led to pyridoxine-responsive sideroblastic anemia. The activity of the mutant enzyme (Arg204
Gln) expressed in vitro was 15% of that of the control; with the addition of PLP, the activity of the mutant enzyme increased to 35% (55).
Among 6 other cases (1 patient and 5 kindreds), 4 had an amino acid substitution at a PLP binding site of ALAS2 that reduced the affinity of ALAS2 for its coenzyme (1, 54, 5658). One of the ALAS2 mutations, Gly291
Ser, reduced enzyme activity to 10% of normal; enzymatic activity was increased with the addition of PLP in vitro (57). In another mutation, Thr388
Ser, activity was decreased to
50% of wild-type but was raised by pyridoxine supplementation (1). A mutation found in a highly conserved region of exon 9, Ile471
Asn, was found in a 30-y-old Chinese man with the pyridoxine-responsive form of XLSA (56). Prokaryotic expression of the normal and mutant cDNAs showed that the mutant construct had lower enzymatic activity than did the normal enzyme and required higher concentrations of PLP to achieve maximal activation. The amino acid substitution occurred in the exon containing the putative PLP binding site, which may account for the reduced ability of the enzyme to catalyze the formation of
-aminolevulinic acid. Another study showed that a large number of probands have mutations in exon 9, the exon containing the PLP binding site (59). Furuyama et al (60) reported an ALAS2 mutation that results in 12% of normal ALAS activity, which increases to 25% in the presence of PLP.
A novel missense mutation in the ALAS2 gene, 1754A
G, in a patient with 53% ALAS activity had 20% activity when expressed in bacteria but 32% in the presence of PLP. Although the mutation, which results in the substitution of glycine for serine, lies outside of exon 9, it is possible that it induces a conformational change that may alter PLP binding to the protein (61). Other mutations located outside exon 9 have also been reported to influence PLP binding (53).
Two cases that appeared late in life have also been analyzed (58). A 77-y-old man and an 81-y-old woman with initial diagnoses of refractory anemia with ringed sideroblasts (which is typically unresponsive to pyridoxine) were found to respond very well to pyridoxine (100 mg/d in the man and 600 mg/d in the woman); hemoglobin concentrations increased in both patients after treatment. The mutations Lys299
Gln and Ala172
Thr were found in the man and woman, respectively. The Ala172
Thr mutation resulted in decreased in vitro stability of bone marrow ALAS2 activity. ALAS2 from both patients showed marked thermolability. Addition of PLP in vitro stabilized the mutant enzymes, which is consistent with the observed in vivo response to pyridoxine. This late-onset form can be distinguished from refractory anemia and ringed sideroblasts by microcytosis, pyridoxine responsiveness, and ALAS2 mutations. These findings emphasize the need to consider all elderly patients with microcytic sideroblastic anemia as candidates for ALAS2 defects, especially if pyridoxine-responsiveness is demonstrated. These investigators concluded, "A decline in PLP availability or metabolism may have precipitated the late onset of XLSA in these patients. An age-related decline in pyridoxine metabolism in combination with a reduced vitamin intake has been described in elderly populations" (58).
It appears that supplementation with glycine, as well as pyridoxine, may be beneficial in new patients and that supplementation with glycine may be beneficial in patients who do not respond to pyridoxine, because the Km for glycine may be affected in some ALAS mutations. For example, the Gly142
Cys constructed mutant has a 4-fold increased Km for glycine (62). If a patient had such a mutation, increased plasma glycine concentrations might increase ALAS activity. One report showed an increase in plasma and cerebrospinal fluid glycine after the administration of 200 mgkg-1d-1 (49).
Kynureninase: xanthurenic aciduria and mental retardation
Kynureninase, a PLP-requiring enzyme involved in tryptophan degradation, catalyzes the conversion of kynurenine and 3-hydroxykynurenine to anthranilic acid and 3-hydroxyanthranilic acid, respectively (see OMIM 236800). Mutations in the kynureninase gene cause mental retardation in children and an excessive urinary output of 3-hydroxykynurenine and kynurenine (and their metabolites, xanthurenic and kynurenic acids). The condition was normalized in 2 children with pyridoxine doses of
30 mg/d. Significantly decreased kynureninase activity in a liver biopsy sample was markedly increased with the addition of PLP, suggesting that a mutation caused a modification in the binding site of the coenzyme (63). A follow-up study confirmed that the defective enzyme was a Km mutant (64). (See the discussion of autism in this section.)
Glutamic acid decarboxylase: seizures in newborns and intelligence quotient deficits
Glutamic acid decarboxylase (GAD; glutamate decarboxylase), a PLP enzyme, converts glutamic acid, an excitatory amino acid, to
-aminobutyric acid, the most important inhibitory neurotransmitter in the central nervous system (up to one-third of synapses in the brain use
-aminobutyric acid as an inhibitory signal). Defects in GAD result in seizures in newborns (see OMIM 266100), but it is not clear whether the seizures are due to too little
-aminobutyric acid or too much glutamic acid (65). Intravenous injection of 100200 mg pyridoxine generally stops the seizures (66). One infant with pyridoxine-dependent seizures was shown to have decreased
-aminobutyric acid production; the seizures stopped within 5 min of the administration of 100 mg pyridoxine. More than 50 cases of pyridoxine-dependent seizures have been reported since 1954 (67).
In one study of 28 infants with seizures, 3 infants had the pyridoxine-responsive phenotype (68). In a study of 120 infants with documented repeated and intractable seizures, only 2 infants responded to pyridoxine administration, suggesting that either only a small percentage of seizures are responsive or that there are many other causes of seizures that are not due to mutations in this vitamin B-6 enzyme (69). There appear to be other enzyme defects that can lower PLP and cause pyridoxine-dependent seizures: one patient had decreased
-aminobutyric acid and increased glutamic acid in the brain, but no significant difference in GAD activity was found between the patient and control subjects, and PLP concentrations were markedly reduced (70).
In a cell line from an infant with pyridoxine-dependent seizures, GAD activity was increased when the enzyme was incubated with high PLP. The investigators speculated that the metabolic abnormality in this disorder may be a binding abnormality between GAD apoenzyme and PLP (71).
A 13-y-old child who died with seizures in progress had elevated glutamic acid and decreased
-aminobutyric acid concentrations in the frontal and occipital cortices but not in the spinal cord; concentrations of all other amino acids, except for cystathionine, were normal. PLP was reduced in the frontal cortex, and GAD activity comparable to that of control subjects was detected when the PLP concentration was >50 µmol/L (70). Response rates from 2 reports give a cumulative pyridoxine-responsiveness of 3%, although seizures may be due to many different defective genes (68, 69).
Intelligence quotients are decreased in pyridoxine-dependent GAD patients, suggesting that the amount of pyridoxine administered should be adjusted to optimally retain intellectual capacity, and not just to stop seizures. A prospective open study found that an increased dose of pyridoxine was associated with an improvement in intelligence quotient. It was suggested that pyridoxine dependency has a wider range of clinical features than classic neonatal seizures and causes specific impairments of higher function, some of which may be reversible by vitamin B-6 therapy (72).
A treatment of asthma, theophylline, depresses PLP concentrations, and may cause seizures by decreasing
-aminobutyric acid production. Pyridoxine treatment reduces theophylline-induced seizures in both mice and rabbits (73).
A linkage analysis study of 2 families argued that pyridoxine-dependent seizures are not due to a Km mutation because the base pair substitutions found in the patients' enzymes were also found in control subjects, and different maternal alleles were passed on to 2 affected children in one family (74). However, some forms of pyridoxine-dependent seizures, which are likely a disease of multiple etiologies, are probably due to Km defects affecting PLP binding by GAD (75).
-Cystathionase: cystathioninuria, mental retardation, and diabetes
After the formation of cystathionine by CBS, another PLP enzyme in the transsulfuration pathway,
-cystathionase (cystathionine
-lyase), converts cystathionine into cysteine and
-ketobutyrate, completing the transfer of sulfur from homocysteine to cysteine. Enzymatic defects (see OMIM 219500) result in cystathionine accumulation in the urine and tissues. The clinical features can include mental retardation, convulsions, thrombocytopenia, nephrogenic diabetes insipidus, and diabetes mellitus. High-dose pyridoxine therapy can markedly reduce concentrations of cystathionine in the urine and blood of deficient patients; it was suggested that vitamin B-6 responsiveness "can best be explained by a structural alteration of the apoenzyme, resulting in failure to combine normally with the coenzyme" (76). This binding theory is supported by others: "The B-6-responsive form results from the synthesis of an aberrant enzyme protein exhibiting altered interaction with the coenzyme, thereby resulting in an inherited increase in the requirement for vitamin B-6" (77). A high percentage of the cases can be ameliorated by supplementation with pyridoxine, which is associated with a reactivation of the defective enzyme and a major decrease in urinary cystathionine excretion; 33 of 37 cases (89%) were found to be pyridoxine-responsive (47). This high percentage is puzzling; one possible explanation is that more severe mutant genes cause lethality and that most of the remaining genes code for a protein with a partial activity and increased Km.
Alanineglyoxylate aminotransferase: hyperoxaluria and renal failure
Alanineglyoxylate aminotransferase is a liver-specific enzyme that uses a PLP cofactor to transfer the amino group from alanine to glyoxylate, forming serine and pyruvate. A primary hyperoxaluria (see OMIM 259900) caused by a functional deficiency of the peroxisomal alanineglyoxylate aminotransferase results in an accumulation of glyoxylate that is converted to oxalate, resulting in renal deposits of calcium oxalate and renal failure. In one study, large doses of pyridoxine reduced urinary oxalate excretion in 2 of 3 patients with primary hyperoxaluria (78). Posttreatment oxalate concentrations were between pretreatment and control concentrations, and the effect of pyridoxine was maintained for 6 mo.
A review of hyperoxaluria indicates that pharmacologic doses of pyridoxine are of benefit and that a Km mutant may be responsible. Pyridoxine treatment may overcome the effects of mutations in the gene encoding alanineglyoxylate aminotransferase that might interfere with cofactor binding (79). It is suggested that as many as 30% of patients with type I primary hyperoxaluria respond to pyridoxine (80). A review of pyridoxine treatment, which discussed 2 recent reports including 18 patients, stated that 50% of patients are unresponsive to pyridoxine, whereas oxaluria is normalized in 20% of patients and somewhat reduced (but not to normal concentrations) in the remaining 30% (81).
Physicians may consider treating with alanine in addition to pyridoxine to determine the optimum cocktail for minimizing oxalate accumulation. We have not seen any reports in which plasma alanine concentrations were measured after the administration of high doses.
Aromatic-L-amino-acid decarboxylase: developmental delay
Aromatic-L-amino acid decarboxylase (AAD; see OMIM 107930) is a homodimeric PLP-containing enzyme synthesizing 2 important neurotransmitters: dopamine and serotonin (82). After the hydroxylation of tyrosine to form dihydroxyphenylalanine, catalyzed by tyrosine hydroxylase, AAD decarboxylates dihydroxyphenylalanine to form dopamine. Dopamine is sequentially broken down to dihydroxyphenylacetaldehyde by monoamine oxidase B [amine oxidase (flavin-containing)], to dihydroxyphenylacetic acid by aldehyde dehydrogenase, and finally to homovanillic acid by catechol O-methyltransferase. Tryptophan 5-monooxygenase produces 5-hydroxytryptophan, which is also decarboxylated by AAD to give rise to serotonin. Serotonin is broken down to 5-hydroxyindoleacetic acid. AAD deficiency is an autosomal recessive inborn metabolic disorder characterized by combined serotonin and dopamine deficiency.
The first reported cases of AAD deficiency were monozygotic twins with extreme hypotonia and oculogyric crises (83). AAD activity was severely reduced and concentrations of dihydroxyphenylalanine and 5-hydroxytryptophan were elevated in cerebrospinal fluid, plasma, and urine. Pyridoxine (100 mg/d) lowered dihydroxyphenylalanine concentrations in cerebrospinal fluid, but treatment with either bromocriptine or tranylcypromine was required for clinical improvement. Another AAD-deficient patient, with similar presentation, also had greatly reduced activity of AAD in plasma (84). Similar to the first reported cases, combined treatment with pyridoxine, bromocriptine, and tranylcypromine produced some clinical improvement. Several other cases of AAD deficiency have apparently benefited from high-dose pyridoxine treatment (85).
Housekeeping
-aminolevulinic acid synthase: sideroblastic anemia
The mapping of a second
-aminolevulinic acid synthase (5-aminolevulinate synthase) gene, ALAS1, to an autosome, chromosome 3, rules it out as the site of the primary defect in X-linked sideroblastic anemia. It was concluded that this gene is a housekeeping form of ALAS (see OMIM 125290) because it is expressed in all cell types including erythroid cells; thus, the gene is designated ALAS1 to distinguish it from the red cellspecific form, ALAS2 (86).
In a study of 20 patients with sideroblastic anemia, 3 patients showed low ALAS activity that was corrected by PLP in vitro, and 2 other patients were found to be responsive to pyridoxine (20 mg/d) (87). When 1 of these 2 patients was taken off pyridoxine, ALAS activity, as measured in bone marrow, fell markedly unless PLP was added in vitro. Additionally, the Km of the enzyme for PLP was substantially greater (2.5 times) than that of a control sample. However, it is unclear whether these are ALAS1 or ALAS2 defects.
A 70-y-old who exhibited an attack of polymorphic, hypochronic anemia, with increased serum iron and numerous ringed sideroblasts in the bone marrow, was determined to have pyridoxine-responsive primary acquired sideroblastic anemia (88). Administration of pyridoxine (initially 200 mg/d, then 600 mg/d) caused a complete remission of all hematologic abnormalities. ALAS activity was increased to 50% of control with 600 mg pyridoxine/d. The activity could be further increased to 100% of control in vitro with 1000 µmol/L PLP. This defect could also be in ALAS1 or ALAS2.
ß-Alanine
-ketoglutarate transaminase: Cohen syndrome
ß-Alanine
-ketoglutarate transaminase (AKT; 4-aminobutyrate aminotransferase) is involved in the formation of malonic semialdehyde from ß-alanine. Children with AKT deficiency have Cohen syndrome (see OMIM 216550), which involves hypotonia, midchildhood obesity, mental deficiency, and facial, oral, ocular, and limb anomalies. A case report of a girl with features of the syndrome reported a response to 100 mg pyridoxine/d for 1 mo, with a normalization of electroencephalogram and a subsiding of lethargy. The girl was hospitalized once when she missed a week of pyridoxine treatment, but reinstatement of the treatment resulted in more improvement. Cultured skin fibroblasts from the girl showed a toxic response to ß-alanine with a 50% reduction in growth. The addition of 100 µmol pyridoxine/L to the cells abolished the toxic effects and increased AKT activity more than 2-fold (89).
Autism
Autism (a developmental disorder that involves impaired social interactions and deviant behavior) and its associated behaviors are thought to affect 5 in 10000 individuals [and as many as 1 in 300 in some US communities (90)]. Autism may be due to defects in a PLP-requiring enzyme or enzymes involved in the metabolism of serotonin and dopamine, although a genetic link to a vitamin B-6requiring enzyme has not been established. The most replicated clinical sign of autism is an elevation of whole-blood serotonin (5-hydroxytryptamine), which is found in >30% of patients (91). Increased concentrations of homovanillic acid, a breakdown product of dopamine, have also been found in several autistic patients. Pyridoxine therapy has been reported to be successful in autism, raising the possibility that a PLP-requiring enzyme might be defective in those patients responsive to vitamin B-6. (PLP is a coenzyme that forms a Schiff base with an amino group in its catalytic action, so that enzymes with PLP metabolize amino acids or other amines, such as dopamine and serotonin.) The only PLP-requiring enzyme directly involved with the synthesis or degradation of dopamine and serotonin is AAD (ie, dihydroxyphenylalanine decarboxylase). The finding in some vitamin B-6responsive patients, namely elevated homovanillic acid that is at least partially reversible with pyridoxine therapy (92), does not suggest a defect in this enzyme. Additionally, cases of AAD deficiency have been reported in the literature (see the discussion of AAD above) and result only in a very severe inborn metabolic disorder involving deficient concentrations of dopamine and serotonin.
It remains to be seen whether other enzymes in the metabolic pathways of these neurotransmitters may be responsible for the various forms of autism that involve altered neurotransmitter metabolism. Autism is diagnosed by clinical, not biochemical, indexes. Thus, if different autistic patients harbor mutations in different metabolic enzymes, it may be possible to reverse the effects of autism by targeting a treatment to each individual patient. In addition to PLP, the coenzymes FAD, NAD, S-adenosylmethionine, tetrahydrobiopterin, and ascorbate are used by enzymes in serotonin and dopamine metabolism.
Because so little is known about the biochemical basis of this condition, it is difficult to associate a treatment response with a particular biochemical or physiologic pathway; however, there have been many reports of successful treatment of autism with pyridoxine. In a survey involving 4000 questionnaires completed by parents of autistic children, high-dose vitamin B-6 and magnesium treatment (n = 318) elicited the best response; for every parent reporting behavioral worsening with the treatment, 8.5 parents reported behavioral improvement. The next best results were with the acetylcholine precursor, deanol (n = 121); 1.8 parents reported a favorable response for every 1 patient who reported worsening (93).
Sixteen autistic patients previously shown to respond to vitamin B-6 treatment were reassessed and given vitamin B-6 or a placebo in a double-blind study (94). Behavior deteriorated significantly during B-6 withdrawal, and 11 of 15 children behaved better when given
300 mg vitamin B-6/d. The authors speculated that vitamin B-6 therapy may correct, or partially correct, a tryptophan-related metabolic error because of a marked increase in serotonin efflux from platelets of autistic children and because large doses of vitamin B-6 elevate serotonin concentrations (95).
A double-blind trial involving 60 autistic children found that vitamin B-6 (30 mg pyridoxine hydrochloridekg-1d-1 up to 1 g/d) and magnesium (1015 mgkg-1d-1) were more helpful than either supplement alone in ameliorating the various effects of autism. Patients receiving the combined treatment showed a significant (P < 0.02) decrease in homovanillic acid excretion (from 6.6 to 4.4 µmol/mmol creatine) and significant clinical improvement (96).
Tryptophan metabolism was studied in 19 children with various forms of psychosis including autism. Four children (including at least one who was autistic) who had abnormal tryptophan metabolite ratios were treated with 30 mg pyridoxine/d, whereupon biochemical features normalized (97). It was thought that these children had kynureninase defects because the kynureninase reaction required greater than normal amounts of PLP to proceed normally (see the discussion of kynureninase above).
More than a dozen other reports (with up to 190 participants) since 1965 and a review of controlled trials (98) have reported improvements in autistic patients with vitamin B-6 and often magnesium supplementation (99102), although the conclusion that pyridoxine is an effective treatment of autism has been challenged: "interpretation of these positive findings needs to be tempered because of methodological shortcomings inherent in many of the studies" (92). A rebuttal (103) to this critique leaves the matter somewhat unresolved. Evidence supporting the hypothesis that defects in enzymes involved in neurotransmitter biosynthesis may be responsible for some forms of autism comes from a study showing that tetrahydrobiopterin, the cofactor for tyrosine and tryptophan hydroxylases, elicited behavioral improvements in 6 children with autism (104).
Tardive dyskinesia
The long-term use of neuroleptic drugs for the attenuation of psychotic disorders such as schizophrenia can lead to tardive dyskinesia, a neurologic movement disorder characterized by rapid, repetitive, uncontrolled movements. There may be >1 million cases of tardive dyskinesia in the United States today and there is some speculation that deranged metabolism of amino acidderived neurotransmitters is responsible for the disease. The involvement of PLP in dopamine, serotonin, and
-aminobutyric acid metabolism may be the reason for the first clinical applications of pyridoxine in the treatment of tardive dyskinesia; pyridoxine-responsiveness has been reported.
A double-blind, placebo-controlled crossover study found high doses of pyridoxine (
400 mg/d) to be effective in reducing symptoms of tardive dyskinesia in patients with schizophrenia (105). Pyridoxine or placebo was added to the normal neuroleptic treatment of all 15 patients in the study for 4 wk at a time, split by a 1-wk washout period. Pyridoxine treatment invoked improvements in both the dyskinetic movement and Parkinsonian subscales with returns to baseline with removal from pyridoxine. An earlier pilot study by the same group showed significant clinical improvement in 4 of 5 tardive dyskinesia patients given 100 mg pyridoxine/d on top of their normal treatment (106). Three of the responders also showed significant improvement on the brief psychiatric rating scale.
The relation between tardive dyskinesia susceptibility and polymorphisms in dopamine and serotonin receptor genes has been a focus of exploration. Homozygosity for the Ser9
Gly polymorphism in the dopamine D3 receptor was higher in schizophrenics with tardive dyskinesia (22%) than without (4%), suggesting that the glycine allele may be a risk factor for developing tardive dyskinesia (107). A similar study supports the involvement of Ser9
Gly in tardive dyskinesia risk, although the presence of tardive dyskinesia was higher in heterozygotes than in either homozygous group (108). The Thr102
Cys polymorphism in the serotonin type 2A receptor gene has also been investigated, although contradictory results leave the matter unresolved as to which allele may be associated with schizophrenia, tardive dyskinesia, or both. A polymorphism might code for a receptor that has a decreased neurotransmitter binding and the capacity to be stimulated by a pyridoxine-induced increase of neurotransmitter level (95), but before any such hypothesis is taken seriously more detailed biochemical evidence is necessary.
Tissue concentrations and toxicity
Pyridoxine's active role in ameliorating many cases of genetic disease involving enzymes that require a PLP cofactor is clear. Plasma PLP concentrations correlate well with tissue PLP concentrations in rats (109), and thus serve as a good indicator of vitamin B-6 status. There is a linear relation between vitamin B-6 intake and plasma concentrations of PLP (up to an intake of 3 mg/d in humans, which correlates with a plasma concentration of 60 nmol/L) (7). This proportional relation has been shown to hold even at 25 mg/d, resulting in a plasma PLP concentration of 200 nmol/L (110). A double-blind study investigating high-dose vitamin B-6 treatment of tardive dyskinesia showed that baseline plasma PLP (49 nmol/L) could be raised >14 times (690 nmol/L) safely with 400 mg/d pyridoxine (105). A rat study referenced in the DRI publication showed that extremely large doses are well absorbed (7).
The higher concentrations of PLP likely facilitate apoenzyme-coenzyme interaction, and hence higher enzymatic activity, although it should be noted that numerical discrepancies do exist in the literature. Normal serum PLP concentrations appear to be
60 nmol/L, whereas control Km concentrations have been described in the µmol/L range for some enzymes (eg, OAT and CBS).
An upper limit exists as to pyridoxine administration. Although dosages in the hundreds of milligrams have been safely applied, reports exist of neurotoxic effects with very high vitamin B-6 usage. One review advises avoiding doses >1000 mg pyridoxine/d (47). The tolerable upper intake level (UL) of pyridoxine for normal use is 100 mg/d (7); however, the severity of some genetic diseases has reasonably prompted physicians to prescribe higher doses.
| THIAMINE (VITAMIN B-1) |
|---|
|
|
|---|
Branched-chain
-ketoacid dehydrogenase: maple syrup urine disease (ketoacidosis, mental retardation, and ataxia)
The branched-chain
-ketoacid dehydrogenase (BCKAD) multienzyme mitochondrial complex is composed of 3 subunits: an E1 component (TPP-dependent decarboxylase) containing
and ß subunits, an E2 component (lipoate-containing acyltransferase), and an E3 component (FAD- and NAD-containing dihydrolipoyl dehydrogenase), the latter of which is also a component of pyruvate and
-ketoglutarate dehydrogenases. BCKAD is responsible for the oxidative decarboxylation of
-ketoacids of the 3 branched-chain amino acids valine, leucine, and isoleucine. Genetic defects in the complex cause maple syrup urine disease (see OMIM 248600), which involves ketoacidosis, mental retardation, ataxia, and sometimes blindness as a result of the accumulation of
-keto acids.
In 1985, thiamine responsiveness was reported in 12 patients who were fed thiamine in doses ranging from 10 to 1000 mg/d (112). The accumulation of ketoacids was shown to return to normal after thiamine feeding (113). In one thiamine-responsive maple syrup urine disease cell line (WG-34) the Km for TPP (as measured via BCKAD decarboxylation activity) was found to be 16 times higher than normal (114). The sequence of the WG-34 mutant has been determined and unexpectedly, the dihydrolipoamide acyltransferase component of the complex was found to be altered. It is possible that the presence of a normal E2 is essential for the efficient binding of TPP to E1 (115). Other genes from thiamine-responsive patients have been sequenced and the E2 subunit was found to be altered in
2 other patients (116118). It appears that mutations in E2 are responsible for the thiamine-responsive versions of maple syrup urine disease and it has been suggested that this E2 defect impairs the E1-E2 interaction where the TPP molecule must bind, thus increasing the cell's requirement for thiamine and TPP (116).
The crystal structure of the TPP binding portion of the BCKAD complex has been determined as well as the effects of various maple syrup urine disease mutations on the enzyme. One mutation (E1
N222S) increased the Km for TPP in a nonresponsive patient. The other 3 mutations, which are described as affecting cofactor binding, all resulted in nonresponsive maple syrup urine disease. Another residue, E1ß N126, which is altered in some patients with maple syrup urine disease, affects interface interaction in the complex and may be involved with subunit association and K+ binding (111). (See the discussion of potassium in the section on hormones, amino acids, and metals.) It remains to be seen whether some of the other mutations, specifically in the intermediate or thiamine-responsive patients, also result in an increased Km for TPP.
In 9 thiamine-dependent cases, the decarboxylation activity (which is a measurement of overall BCKAD activity) ranged from 3% to 40% of normal (119). In one case, the mutant enzyme was shown to be heat labile and stabilized by increased TPP (120). No adequate reports of the percentage of cases that are remediable by thiamine are available. Although the enzyme complex also uses NAD, FAD, and dihydrolipoic acid as cofactors in addition to the Mg2+ salt of TPP, it appears that the therapeutic application of niacin, riboflavin, and lipoic acid has not been attempted. NAD, CoA, and Mg2+ were tried in cell culture but were ineffective (121).
In one case, supplementation with oral thiamine reversed the blindness that sometimes accompanies the disease (122). Another noteworthy thiamine-responsive case involved a compound heterozygote with a large deletion and a 1002G
A transition at an exon 8 splice site that resulted in exon skipping and the transcription of different length mRNAs (117). The mechanism for this thiamine response remains to be explained.
The data suggest that combination therapy with thiamine, lipoic acid, riboflavin, nicotinamide, and adequate potassium [for which the recommended dietary allowance intake is 2000 mg/d (9)] may be optimal for the initial treatment of patients with maple syrup urine disease. Potassium might be beneficial because K+ is required for the stabilization of E1 by TPP. The stabilizing effect of K+ on BCKAD was shown in the rat liver BCKAD enzyme (123) as well as in the human E1ß protein (124). Both groups observed a dependence of enzyme activity on the concentration of potassium salts.
Pyruvate decarboxylase: Leigh disease (lactic acidosis, ataxia, and mental retardation)
Pyruvate decarboxylase is part of the pyruvate dehydrogenase multienzyme mitochondrial complex (PDHC) that uses TPP, lipoic acid, CoA, FAD, and NADH coenzymes to catalyze the conversion of pyruvate to acetyl-CoA (see OMIM 312170). The gene encoding the E1
peptide of the E1 subunit (pyruvate decarboxylase), which binds TPP, is located on the X chromosome. Genetic defects in the complex can lead to lethal lactic acidosis, psychomotor retardation, central nervous system damage, ataxia, muscle fiber atrophy, and developmental delay (125).
X-linked genetic defects in PDHC cause pyruvate and lactate accumulation and encephalomyelopathy. Twenty-six patients responded to high intakes of thiamine, ranging from 20 to 3000 mg/d. In 2 sisters (126), lipoic acid (100 mg/d) plus thiamine (3000 mg/d) were found to give the best remediation. In several cases, the mutation was shown to increase the Km of the E1 subunit for TPP and reduce the Vmax (127, 128). In several cases in which lactate was measured, thiamine lowered lactate concentrations significantly (127, 129, 130).
In a study of 13 thiamine-responsive PDHC-deficient patients, some had a decreased affinity of PDHC for TPP that was responsive to TPP, whereas the PDHC activity of others increased at high TPP concentrations with no statement about enzyme affinity (131). Another group of patients with lactic acidemia and muscle fiber atrophy had TPP-responsive PDHC enzymes (1.82 and 2.63 nmolmin-1mg protein-1 with 400 µmol/L TPP compared with 0.28 and 0.02 nmolmin-1mg protein-1, respectively, with 0.1 µmol/L TPP) (132).
A female infant with West syndrome (a unique epileptic syndrome with frequently poor prognosis and spasms associated with elevated blood and cerebrospinal fluid lactate concentrations) had thiamine-responsive PDHC deficiency (133). Lactate concentrations were lowered and symptoms disappeared when the infant was administered dichloroacetate and high doses of thiamine (500 mg/d). The patient carried the mutation Gly89
Ser in exon 3 of her PDHC E1
gene, resulting in a decreased affinity for TPP. PDHC activity was activated in vitro with the addition of TPP. Gly89
Ser, along with 4 other mutations in thiamine-responsive PDHC-deficient patients (His44
Arg, Arg88
Ser, Arg263
Gly, and Val389
fs) are in regions outside the TPP binding site in exon 6 (133). However, these mutations may affect overall protein conformation and indirectly decrease cofactor binding affinity. Pyridoxine was shown to be effective in several cases of West syndrome (134).
Three point mutations in E1-deficient patients were recreated in vitro. One mutation, Met181
Val, exhibited a 250-fold increased Km for TPP and, in addition to another studied mutation (Pro188
Leu) is involved with TPP binding. It was mentioned that an aspartate and an asparagine residue form H-bonds with and coordinate the divalent cation (Mg2+ or Ca2+). This cation interacts with the oxygen groups of the pyrophosphate portion of TPP (135). Thus, encouraging adequate intake of magnesium and calcium in PDHC-deficient patients by physicians seems reasonable. The Met181
Val mutation also raised the Km for pyruvate 3-fold (135). Physicians might consider treating PDHC-deficient patients with precursors to all of the cofactors used by the enzyme complex: thiamine, lipoic acid, pantothenate, riboflavin, and niacin.
Thiamine transporter, thiamine pyrophosphokinase, and
-ketoglutarate dehydrogenase: thiamine-responsive megaloblastic anemia
Thiamine-responsive megaloblastic anemia (see OMIM 249270) can be caused by defects in a putative thiamine transporter, thiamine pyrophosphokinase (TPK), and
-ketoglutarate dehydrogenase [KGDH; oxoglutarate dehydrogenase (lipoamide)]. The putative thiamine transporter, encoded by SLC19A2, is homologous to reduced folate carrier proteins and may bring thiamine into cells. TPK is responsible for the phosphorylation of thiamine to TPP cofactor. KGDH is one of the thiamine-dependent dehydrogenases that binds TPP by an E1 carboxylase (see also the discussions of BCKAD and PDHC above).
Mutations in SLC19A2 (136, 137) and defects in TPK (138, 139) and KGDH (140, 141) have all been found in patients with thiamine-responsive megaloblastic anemia. Thiamine-responsive megaloblastic anemia, first described by Rogers et al (142) in 1969, is an autosomal recessive condition with an early onset and is characterized by the triad of megaloblastic anemia, diabetes mellitus, and sensorineural deafness.
Mutations in the gene for the putative thiamine transporter, SLC19A2 (see OMIM 603941), were found in all affected individuals in 6 families with thiamine-responsive megaloblastic anemia (136). Another study found similar results and supports the putative role of SLC19A2 in some forms of thiamine-responsive megaloblastic anemia (137). There is evidence that there is a low-affinity thiamine transporter and that this transporter is responsible for the clinical thiamine-responsiveness, partially correcting for the decreased intracellular thiamine concentrations that result from the defective high-affinity transporter, SLC19A2 (143). The low-affinity version may be the recently identified thiamine transporter SLC19A3 (144). Such a bypass would not involve overcoming a Km defect. Both thiamine transport and TPK were thought to be the enzymes affected in a group of 7 patients with thiamine-responsive megaloblastic anemia (143).
TPK activity was reduced in a patient with thiamine-responsive megaloblastic anemia in whom 60 d of thiamine therapy (50 mg/d) normalized concentrations of free and phosphorylated thiamine (139). Thiamine responsiveness was found in 2 similar cases of thiamine-responsive megaloblastic anemia with deficient TPK activity (138). After thiamine ingestion (75 mg/d), which raised erythrocyte TPP concentrations 1.5- and 2-fold in the patients, hematologic findings returned to normal and insulin requirements decreased by 66%. (See OMIM 606370.)
Deficient KGDH activity (see OMIM 203740) in one patient with thiamine-responsive megaloblastic anemia was stimulated by TPP titration. Near normal activity was reached with 0.75 µmol TPP/L, whereas control subjects were not responsive (140). The KGDH activity in another patient was 2% of that of a control subject, and a defect in binding of TPP to the KGDH complex was suggested (141). Because the KGDH complex uses other coenzymes including lipoic acid, CoA, FAD, and NAD, patients may benefit initially from a high-dose mixture of thiamine, lipoic acid, pantothenate, riboflavin, and niacin, but controlled clinical investigations are needed to validate or reject this hypothesis.
Oxidation of
-ketoglutarate, pyruvate + malate, and malate + palmitate: lactic acidosis and cardiomyopathy
Cardiomyopathy and lactic acid accumulation in a neonate was remedied by feeding thiamine (50 mg/d), carnitine (2 g/d), and riboflavin (50 mg/d), which reversed the high blood lactate concentrations and other symptoms. The patient showed a deficiency in the oxidation of all substrates tested: pyruvate,
-ketoglutarate, and palmitate. After freezing and thawing and addition of essential cofactors (TPP, CoA-SH, NAD), the activities of the ketoacid dehydrogenases became normal. The apparent deficiency may have been caused by a primary deficiency in one of the cofactors or by a defect at the level of thiamine. Although the precise metabolic defect was not assessed, it was concluded that the patient was responsive to thiamine (145). A similar case (146) was also reversed by thiamine (50 mg/d).
Tissue concentrations and toxicity
The effectiveness of thiamine administration in these diseases involving several mutant genes seems clear. It has been shown that a 10-mg dose of thiamine raised serum thiamine concentrations to 24 nmol/L; concentrations returned to baseline (17 nmol/L) 6 h later (147). With higher pharmacologic doses, namely, repetitive 250-mg amounts taken orally and 500 mg/d given intramuscularly, nearly 1 wk was required for steady state plasma concentrations to be reached (148). It seems apparent that thiamine administration raises both TPP and thiamine concentrations in serum, but we have not found documentation of this.
There is no defined UL for thiamine because of its relative safety. Adverse effects of thiamine have been documented, although they appear to be rare. For example, in a study of 989 patients, 100 mg thiamine hydrochloride/d given intravenously resulted in a burning effect at the injection site in 11 patients and pruritus in 1 (149).
| RIBOFLAVIN (VITAMIN B-2) |
|---|
|
|
|---|