Epigenetic synergies between biotin and folate in the regulation of pro-inflammatory cytokines and repeats.

The protein biotin ligase, holocarboxylase synthetase (HLCS), is a chromatin protein that interacts physically with the DNA methyltransferase DNMT1, the methylated cytosine-binding protein MeCP2 and the histone H3 K9-methyltransferase EHMT1, all of which participate in folate-dependent gene repression. Here we tested the hypothesis that biotin and folate synergize in the repression of pro-inflammatory cytokines and long-terminal repeats (LTRs), mediated by interactions between HLCS and other chromatin proteins. Biotin and folate supplementation could compensate for each other's deficiency in the repression of LTRs in Jurkat and U937 cells. For example, when biotin-deficient Jurkat cells were supplemented with folate, the expression of LTRs decreased by >70%. Epigenetic synergies were more complex in the regulation of cytokines compared with LTRs. For example, the abundance of TNF-α was 100% greater in folate- and biotin-supplemented U937 cells compared with biotin-deficient and folate-supplemented cells. The NF-κB inhibitor curcumin abrogated the effects of folate and biotin in cytokine regulation, suggesting that transcription factor signalling adds an extra layer of complexity to the regulation of cytokine genes by epigenetic phenomena. We conclude that biotin and folate synergize in the repression of LTRs and that these interactions are probably mediated by HLCS-dependent epigenetic mechanisms. In contrast, synergies between biotin and folate in the regulation of cytokines need to be interpreted in the context of transcription factor signalling.

Intrauterine elimination of pyridoxal 5'-phosphate in full-term and preterm infants.

This study addressed the intrauterine elimination of pyridoxal 5'-phosphate (PLP) in 15 preterm and 31 full-term infants, thereby providing estimates of fetal vitamin consumption as well as maternal vitamin requirements during pregnancy. Elimination was calculated as the difference in the plasma PLP concentration between umbilical vein and umbilical artery times the umbilical plasma flow. Plasma flow in the umbilical vein was calculated from pulsed Doppler ultrasonographic determination of blood flow and from the hematocrit value. Plasma PLP concentrations were assayed in maternal and umbilical veins and the umbilical artery; PLP concentrations were similar in preterm and full-term infants (P > 0.05). In both groups of infants the PLP concentration in the umbilical vein (preterm: 100.3 nmol/L; full-term: 63.9 nmol/L) was ninefold higher than in maternal circulation (P < 0.001). In full-term infants, PLP concentrations in maternal and umbilical veins correlated weakly (r = 0.358, P < 0.05), but no significant correlation was found in the preterm group (P > 0.05). The arteriovenous concentration gradient of PLP in cord vessels was higher in preterm infants (15.0 nmol/L) than in full-term infants (2.1 nmol/L), but the difference between groups was not significant (P > 0.05). Preterm infants eliminated 1.7 nmol PLP.kg-1.min-1 in utero, whereas full-term infants eliminated 0.2 nmol PLP.kg-1.min-1 (P < 0.05). The significantly higher plasma flow in preterm infants (116 mL.min-1.kg-1) compared with full-term infants (78 mL.min-1.kg-1) contributed to the higher PLP elimination in preterm infants.

Bioavailability of biotin given orally to humans in pharmacologic doses.

BACKGROUND: Patients with carboxylase deficiency are treated with pharmacologic doses of biotin. OBJECTIVE: We sought to determine the bioavailability of biotin at pharmacologic doses. DESIGN: Biotin was administered orally (2.1, 8.2, or 81.9 micromol) or intravenously (18.4 micromol) to 6 healthy adults in a crossover design with > or =2 wk between each biotin administration. Before and after each administration, timed 24-h urine samples were collected. Urinary biotin and biotin metabolites were analyzed by an HPLC avidin-binding assay. RESULTS: Urinary recoveries of biotin plus metabolites were similar (approximately 50%) after the 2 largest oral doses and the 1 intravenous dose, suggesting 100% bioavailability of the 2 largest oral doses. For unexplained reasons, the apparent recovery of the smallest oral dose was about twice that of the other doses. For all 4 doses, biotin accounted for >50% of the total of biotin and biotin metabolites in urine. Bisnorbiotin (13-23%), biotin-d,l-sulfoxide (5-13%), bisnorbiotin methyl ketone (3-9%), and biotin sulfone (1-3%) accounted for the remainder. The percentage excretion of biotin was greater when biotin was administered intravenously and for the largest oral dose than for the 2 smallest oral doses. CONCLUSION: Our data provide evidence that oral biotin is completely absorbed even when pharmacologic doses are administered. Biotin metabolites account for a substantial portion of total urinary excretion and must be considered in bioavailability studies. We speculate that renal losses of biotin (as a percentage of the dose administered) are moderately elevated when pharmacologic doses of biotin are administered.

Pharmacokinetics of orally and intravenously administered riboflavin in healthy humans.

The pharmacokinetics and utilization (flavocoenzyme synthesis) of orally and intravenously administered riboflavin in healthy humans were assessed. After the determination of circadian rhythms of riboflavin concentrations in blood plasma and urine of four males and five females (control period), each of these subjects received three different oral riboflavin doses (20, 40, and 60 mg) and one intravenous bolus injection of riboflavin (11.6 mg). Vitamins were administered in a randomized, cross-over design with 2 wk between each administration. Blood plasma and urine specimens were collected repeatedly over a period of 48 h after each administration. Concentrations of flavocoenzymes and riboflavin were analyzed in blood plasma; riboflavin was assayed in urine. During the control period, a small circadian variation was observed: plasma concentrations and urinary excretion of riboflavin were low during the afternoon (P < 0.05). Pharmacokinetics were calculated using a two-compartment open model. The maximal amount of riboflavin that can be absorbed from a single dose was 27 mg per adult. Half-life of absorption was 1.1 h. First-order rate constants describing distribution and elimination of riboflavin were significantly higher after intravenous than after oral administration (P < 0.01). Release of flavocoenzymes into plasma was low compared with the increase of riboflavin concentrations. 7 alpha-Hydroxyriboflavin was identified in plasma. Clearance data indicated that urinary excretion of riboflavin contributes to one-half of the overall removal of riboflavin from plasma. No sex differences were observed for any of the pharmacokinetic variables (P > 0.05).

Identification of biotin sulfone, bisnorbiotin methyl ketone, and tetranorbiotin-l-sulfoxide in human urine.

In previous studies using the HPLC and avidin-binding assay, five unidentified avidin-binding substances were observed in human urine. The present study investigated the identity of these substances. Urine was collected before and after intravenous administration of 18.5 mumol biotin to healthy adults. Unknown substances 1 and 3 were initially identified as biotin sulfone and bisnorbiotin methyl ketone, respectively, by coelution with authentic standards on HPLC. Identities were confirmed by thin-layer chromatography and by derivatization with p-dimethyl-aminocinnamaldehyde. As expected for biotin metabolites, the urinary excretion of biotin sulfone and bisnorbiotin methyl ketone increased with biotin administration. The urinary excretion of biotin sulfone increased 21-fold from 0.2 nmol/h before to 4.2 nmol/h after administration; the excretion of bisnorbiotin methyl ketone increased 130-fold from 0.4 to 51.8 nmol/h. At presumed steady state in free-living subjects (n = 6), biotin sulfone and bisnorbiotin methyl ketone accounted for 3.6% and 7.9% of total biotin excretion, respectively. Traces of tetranorbiotin-l-sulfoxide were also identified by using thin-layer chromatography and derivatization with p-dimethylaminocinnamaldehyde. However, tetranorbiotin-l-sulfoxide was not detectable in urine by the HPLC and avidin-binding assay because this metabolite has weak avidin-binding affinity. We conclude that biotin sulfone and bisnorbiotin methyl ketone are present in measurable quantities in human urine; their quantitation should allow more accurate studies on human biotin metabolism and turnover.

The efflux of biotin from human peripheral blood mononuclear cells.

Peripheral blood mononuclear cells (PMBCs) are readily available for sampling and are a useful model for studying biotin metabolism in human cells. To better understand biotin handling by PMBCs, we investigated the mechanism(s) and kinetics of biotin efflux from PMBCs. Human PMBCs were incubated with [(3)H]biotin at 475 pmol/L to load the cells. The [(3)H]biotin-loaded cells were then harvested and incubated in [(3)H]biotin-free media for up to 20 hours. At various intervals, aliquots of the PMBC suspensions were collected and analyzed for intracellular [(3)H]biotin. [(3)H]Biotin efflux from cells at 37 degrees C was fast and triphasic; the half-lives for the three elimination phases were 0.2 +/- 0.02 hours, 1.2 +/- 0.1 hours, and 21.9 +/- 13.6 hours. Such a triphasic [(3)H]biotin efflux could reflect (1) rapid efflux of free biotin, (2) slower release of biotin bound to intracellular molecules, and (3) even slower release from carboxylases in cellular organelles. Incubation at 4 degrees C rather than 37 degrees C increased the [(3)H]biotin retained at 20 hours from 27% to 85%. This observation is consistent with transporter-mediated efflux. When cellular glucose utilization was reduced by 2-deoxy-d-glucose and sodium fluoride, [(3)H]biotin efflux was similar to controls, suggesting that biotin efflux does not directly require metabolic energy. When [(3)H]biotin-loaded cells were incubated in external medium containing unlabeled biotin analogs, [(3)H]biotin efflux was accelerated approximately two times compared with incubation in a biotin-free medium. This observation suggests that biotin efflux is mediated by the same transporter that mediates biotin uptake from the extracellular medium (i.e., classic countertransport).

Biotin biochemistry and human requirements.

Human biotin turnover and requirements can be estimated on the basis of (1) concentrations of biotin and metabolites in body fluids, (2) activities of biotin-dependent carboxylases, and (3) the urinary excretion of organic acids that are formed at increased rates if carboxylase activities are reduced. Recent studies suggest that the urinary excretions of biotin and its metabolite bisnorbiotin, activities of propionyl-CoA carboxylase and beta-methylcrotonyl-CoA carboxylase in lymphocytes, and urinary excretion of 3-hydroxyisovaleric acid are good indicators of marginal biotin deficiency. On the basis of studies using these indicators of biotin deficiency, an adequate intake of 30 microg (123 nmoles) of biotin per day is currently recommended for adults. The dietary biotin intake in Western populations has been estimated to be 35 to 70 microg/d (143-287 nmol/d). Recent studies suggest that humans absorb biotin nearly completely. Conditions that may increase biotin requirements in humans include pregnancy, lactation, and therapy with anticonvulsants or lipoic acid.

Human peripheral blood mononuclear cells: ; Inhibition of biotin transport by reversible competition with pantothenic acid is quantitatively minor.

A transporter present in intestinal cells and in choriocarcinoma cells has been shown to transport both pantothenic acid and biotin at similar transporter affinities. However, the concentration of pantothenic acid in most foods and biological fluids is approximately 200 times the concentration of biotin; theoretically, pantothenic acid might substantially reduce biotin transport via competition. In the present study, we sought to determine whether pantothenic acid reduces biotin transport by the biotin transporter in peripheral blood mononuclear cells (PBMC). PBMC were isolated from human blood by gradient centrifugation. Incubations with [(3)H]biotin and pantothenic acid were conducted at physiologic concentrations. Intracellular [(3)H]biotin was quantified after washing by liquid scintillation counting. Pantothenic acid at 10 to 1,000 nmol/L reduced biotin (475 pmol/L) uptake by less than 12% (P < 0.05). Based on Lineweaver-Burk plots, the competition was reversible. Several structural analogs of pantothenic acid at 1,000 nmol/L reduced biotin transport by only 7 to 15% (P = 0.13). No pattern of molecular structure required for recognition by the transporter was apparent. Extracellular pantothenic acid did not affect biotin efflux from [(3)H]biotin-loaded PBMC (P > 0.05), suggesting that countertransport of extracellular pantothenic acid and intracellular biotin does not increase biotin efflux from PBMC. We conclude that the physiologic effects of pantothenic acid on the transport of biotin in PBMC are likely to be quantitatively minor.

The utilization of intravenously infused pyridoxine in humans.

The utilization of a usual dose of intravenously infused pyridoxine (100 mg pyridoxine hydrochloride) was investigated in ten healthy males. Blood plasma and erythrocytes were investigated by means of high-performance liquid chromatography. Detectable metabolites in blood plasma were pyridoxine, pyridoxal 5'-phosphate, pyridoxal and 4-pyridoxic acid. In erythrocytes pyridoxine, pyridoxal 5'-phosphate, pyridoxal, and pyridoxamine 5'-phosphate were found. From their concentration-time curves rate constants of elimination and invasion, volume of distribution (pyridoxine) and the areas under the curves were calculated. Values for concmax and tmax are reported. A comparison with earlier results of oral pyridoxine administration revealed a better utilization after intravenous than after oral application, i.e. a greater build-up of coenzyme forms. A regulatory phenomenon in erythrocytes caused by high doses of pyridoxine is described. In view of the potential toxicity of pyridoxine the doses used in parenteral nutrition are called into question.

Biotin homeostasis during the cell cycle.

Peripheral blood mononuclear cells (PBMC) accumulate biotin by a Na-dependent energy-requiring transporter. This transporter might be the so-called Na-dependent multivitamin transporter, but kinetic observations suggest the existence of a second, more specific, biotin transporter. PBMC respond to proliferation by increased uptake of biotin; the increase is probably mediated by an increased number of transporters on the cell surface. The inferred increase in the biotin transporter synthesis is relatively specific. The increased uptake of biotin into proliferating PBMC is consistent with the hypothesis that these cells have an increased demand for biotin. Indeed, proliferating PBMC increase expression of genes encoding beta-methylcrotonyl-CoA carboxylase and propionyl-CoA carboxylase, generating a quantitatively significant increased demand for biotin as a coenzyme in newly-synthesized carboxylases. Moreover, expression of the holocarboxylase synthetase gene increases, consistent with the synthesis of new holocarboxylases. In addition, proliferating PBMC increase both the density of biotinylation of histones and the mass of biotinylated histones per cell, suggesting a potential role for biotin in transcription and replication of DNA.

The transport of thiamine, riboflavin and pyridoxal 5'-phosphate by human placenta.

In the present examination the concentrations of thiamine, riboflavin and pyridoxal 5'-phosphate in blood plasma of pregnant women and venous and arterial cord plasma were determined. In maternal plasma the concentration was 4.5 nmol/l (thiamine), 22.2 nmol/l (PLP), 8.7 nmol/l (free riboflavin) and 84.5 nmol/l (FAD + FMN). In venous cord plasma the concentration was 45.9 nmol/l (thiamine), 112.1 nmol/l (PLP), 40.6 nmol/l (free riboflavin) and 49.1 nmol/l (FAD + FMN). Therefore the gradients of concentration between maternal plasma and venous cord plasma were 1:10 for thiamine, 1:4.7 for free riboflavin and 1:5 for PLP. For the coenzyme forms of vitamin B2 the maternal circulation showed the higher concentration (1.7:1). Therefore an active transplacentar transport mechanism was assumed. The vitamin concentrations in cord arteria were significantly lower than that in cord vene, indicating a massive retention by the fetus.

The intrauterine turnover of thiamin in preterm and full-term infants.

In thirteen preterm infants, 45 full-term infants, and their mothers thiamin was analyzed in plasma from maternal veins, umbilical arteries, umbilical veins, and placental tissue. The blood flow in the umbilical veins was determined by pulsed Doppler ultrasonography. Thiamin-dependent transketolase was measured in erythrocytes from full-term infants and their mothers. Plasma thiamin concentrations in umbilical veins from preterm infants (227.0 +/- 85.0 nmol/L) and full-term infants (121.3 +/- 103.3 nmol/L) were seven times greater than maternal concentrations (p < 0.005). Maternal and umbilical thiamin concentrations were lower in the full-term group compared to the preterm group (p < 0.05). Arteriovenous concentration gradients were not feasible. The blood flow in the umbilical veins was higher in full-term compared to preterm infants (p < 0.05). However, intrauterine thiamin supply (plasma thiamin concentration times umbilical plasma flow) and placental thiamin concentrations were not different between preterm and full-term infants. Thiamin saturation of transketolase was greater in fetal than in maternal erythrocytes (p < 0.005); severe thiamin deficiency was not observed. Our findings suggest that thiamin turnover is similar in early and late pregnancy. Fetal tissue uptake of thiamin is not substantial. Transketolase activities suggest that thiamin status is sufficient even in late pregnancy.

The identification and kinetics of 7 alpha-hydroxyriboflavin (7-hydroxymethylriboflavin) in blood plasma from humans following oral administration of riboflavin supplements.

Following the administration of different oral (20, 40, 60 mg) and intravenous (11.6 mg) doses of riboflavin to healthy humans and female patients with liver cirrhosis (oral 40-mg dose), 7 alpha-hydroxyriboflavin (7-hydroxymethylriboflavin) was identified in blood plasma by fluorescence after high-performance liquid and thin-layer chromatographies, and by its absorbance spectrum. The apparent first-order absorption rate constant of 7 alpha-hydroxyriboflavin was 1.2 per hour in healthy subjects. Plasma peak concentrations of 40 nmol/l in males and 20 nmol/l in females (p < 0.01) were achieved within two hours. Peak concentrations and areas under the plasma curves (smaller in females, p < 0.01) of 7 alpha-hydroxyriboflavin were 5 to 16% of those observed for riboflavin. Healthy females showed an approximately 2.5-fold faster disposition of 7 alpha-hydroxyriboflavin from plasma than males (p < 0.01). Correction of peak concentrations and areas under the plasma curves by the rate constants of disposition led to the finding of approximately equal amounts of 7 alpha-hydroxyriboflavin released into plasma by both sexes (p > 0.05). No significant influence of different oral riboflavin doses on 7 alpha-hydroxyriboflavin kinetics was found (p > 0.05). Liver cirrhosis had no significant effect on the amount of 7 alpha-hydroxyriboflavin released into blood plasma (p > 0.05). However, the failure to detect this metabolite following intravenous riboflavin administration indicates a substantial influence of gastrointestinal- or liver-passage.

The metabolism of riboflavin in female patients with liver cirrhosis.

The metabolism of vitamin B2 was studied in five female patients with liver cirrhosis of varying etiology. Following the oral administration of 40 mg (106.3 mumol) riboflavin, plasma concentrations of riboflavin and flavo-coenzymes as well as urinary riboflavin excretion were analyzed over a period of 48 h. Results were compared to data obtained for healthy controls (Zempleni J. et al, Am. J. Clin. Nutr., 1996 [15]). About 18% of the administered vitamin was recovered from patients' urine, indicating an absorption similar to healthy subjects (p > 0.05). The area under the riboflavin plasma concentration vs time curve was 1.2-fold larger among patients than controls, but the difference was not significant (p > 0.05). Riboflavin peak concentrations in plasma (315.6 nmol/l) and times when those concentrations were achieved (3.0 h) were similar to those found for healthy subjects (p > 0.05). Flavocoenzyme peak plasma concentrations were increased 1.4-fold above their baseline levels in cirrhotics which was equal to controls (p > 0.05). 7 alpha-Hydroxyriboflavin was detected in the plasma of patients. Distribution and elimination kinetics of riboflavin were analyzed by using a two-compartment open model; the riboflavin plasma disposition rate constants of the patients (k alpha = 0.7232 h-1; k beta = 0.0627 h-1) were not different from controls (p > 0.05). No differences between both groups were found regarding renal excretion (renal clearance, first-order rate constants for renal excretion; p > 0.05). In conclusion, patients with liver cirrhosis of varying etiology and varying medical treatment did not show alterations of riboflavin turnover.

Pharmacokinetics of vitamin B6 supplements in humans.

The use of vitamin B6 supplements is widespread today. Doses used are often elevated far above the physiological range and reach levels up to 600-fold higher than recommended dietary allowances for healthy people. While the toxic effects caused by chronic high doses of vitamin B6 have been described earlier, pharmacokinetic data on vitamin B6 supplements are rare. This article reviews the pharmacokinetic data of vitamin B6 from human subjects.

Determination of riboflavin and flavocoenzymes in human blood plasma by high-performance liquid chromatography.

A high-performance liquid chromatographic method for determining riboflavin, flavin adenine dinucleotide, and flavin mononucleotide in human blood plasma is presented. Flavocoenzymes are determined as flavin mononucleotide after acid hydrolysis of flavin adenine dinucleotide. Metabolites are separated by reversed-phase column chromatography and quantified by their native fluorescence. Criteria of quality are (riboflavin/flavocoenzymes): coefficients of variation 2.8/4.6% (intra-assay) and 2.8/4.4% (inter-assay); recovery 82.4/94.4%; detection limit < 3.0/9.0 nmol/l. Because sample preparation requires only few steps, and the retention times are short ( < 5 min), this method is recommended for use in routine analysis.

Proliferation of peripheral blood mononuclear cells increases riboflavin influx.

Previously we demonstrated that proliferation of peripheral blood mononuclear cells (PBMC) causes a five-fold increase in cellular uptake of biotin; this increase is mediated by an increased number of biotin transporters on the PBMC surface. In the present study, we investigated the specificity of this phenomenon by determining whether the cellular uptake of riboflavin also increases in proliferating PBMC and whether the increase is also mediated by an increased number of transporters per cell. We characterized [3H]riboflavin uptake in both quiescent and proliferating PBMC. In quiescent PBMC, [3H]riboflavin uptake exhibited saturation kinetics and was reduced by addition of unlabeled riboflavin (P < 0.05) or lumichrome (P < 0.01). These observations are consistent with transporter-mediated uptake. [3H]Riboflavin uptake was reduced at 4 degrees C compared with 37 degrees C (P < 0.01) and by 2, 4-dinitrophenol (P < 0.05) but not by ouabain or incubation in sodium-free medium. These data provide evidence for an energy-dependent but sodium-independent transporter. Proliferating PBMC accumulated approximately four times more [3H]riboflavin than quiescent PBMC (P < 0.05). Because both transporter affinity and transporter number per cell (as judged by maximal transport rate) were similar in quiescent and proliferating PBMC, we hypothesize that the increased riboflavin uptake by proliferating PBMC reflects only increased cellular volume. To test this hypothesis, PBMC volume was reduced using hyperosmolar medium; [3H]riboflavin uptake decreased to about 50% of isotonic controls (P < 0.01). Thus we conclude that proliferating PBMC increase cellular content of riboflavin and biotin by two different mechanisms.

Intrauterine vitamin B2 uptake of preterm and full-term infants.

Intrauterine uptake of vitamin B2 in preterm and full-term infants was examined. Factors of influence on vitamin supply were considered. Forty-four women and their infants were included in the study. Fetal vitamin uptake was calculated as arteriovenous concentration gradient in cord plasma times umbilical plasma flow. Concentration of vitamin B2 (free riboflavin and flavocoenzymes) was determined by high performance liquid chromatography of placental tissue and blood plasma (maternal vein, umbilical artery, umbilical vein). Flavocoenzymes were analyzed as flavin mononucleotide after acid hydrolysis of flavin adenine dinucleotide. Umbilical plasma flow was measured using pulsed Doppler sonography. Both free riboflavin and flavocoenzymes were transferred from the maternal plasma to the umbilical vein, but only free riboflavin was accumulated (approximately 1:4 for preterm and full-term infants, respectively). Flavocoenzyme concentration was higher in the umbilical vein than in the umbilical artery (p < 0.05). This indicated a median uptake of flavocoenzymes of 1.5 nmol/min.kg in preterm infants and 0.4 nmol/min.kg in full-term infants (preterm versus full-term, p < 0.01). Fetal vitamin supply depended on umbilical plasma flow and on maternal vitamin status (the latter was shown only in full-term infants). No dependence on placental vitamin concentration was observed (p > 0.05). Concentration of free riboflavin was higher in umbilical artery than in umbilical vein (p < 0.05). This indicated a release of free riboflavin from fetal tissues independent of gestational age (0.4 nmol/min.kg, preterm; 0.2 nmol/min.kg, full-term; p > 0.05).

Marginal biotin deficiency is teratogenic.

Recent studies of biotin status during pregnancy provide evidence that a marginal degree of biotin develops in a substantial proportion of women during normal pregnancy. Several lines of evidence suggest that, although the degree of biotin deficiency is not severe enough to produce the classic cutaneous and behavioral manifestations of biotin deficiency, the deficiency is severe enough to produce metabolic derangements in women and that characteristic fetal malformations occur at a high rate in some mammals. Moreover, our analysis of data from a published multivitamin supplementation study provide significant albeit indirect evidence that the marginal degree of biotin deficiency that occurs spontaneously in normal human gestation is teratogenic. Investigation of potential mechanisms provides evidence that biotin transport by the human placenta is weak. Further, proliferating cells accumulate biotin at a rate five times faster than quiescent cells; this observation suggests that there is an increased biotin requirement associated with cell proliferation. Perhaps this requirement arises from the need to synthesize additional biotin-dependent holocarboxylases or provide additional biotin as a substrate for biotinylation of cellular histones. Reduced activity of the biotin-dependent enzymes acetyl-CoA carboxylase and propionyl-CoA carboxylase can cause alterations of lipid metabolism and might theoretically lead to alterations of polyunsaturated fatty acid and prostaglandin metabolism that derange normal skeletal development.