Biotin, a water-soluble vitamin, is an essential cofactor in a variety of enzymes that catalyze reactions that fix carbon dioxide. In mammals, these reactions are component steps in various cellular processes, including fatty acid synthesis, gluconeogenesis, amino acid metabolism and propionate metabolism. Recent studies suggest that biotin status is reduced as a consequence of long-term anticonvulsant therapy (Mock and Dyken 1995). Increases in the urinary excretion of biotin metabolites along with decreases in the urinary excretion of biotin in subjects receiving chronic anticonvulsant treatment suggest that accelerated biotin catabolism leads to biotin wasting and contributes to reduced biotin status (Krause et al. 1988, Mock et al. 1995). To study accelerated biotin catabolism in an animal model, biotin metabolism should resemble that in humans, and anticonvulsant therapy should accelerate biotin catabolism in the species chosen.

In previous studies (Wang et al. 1996), we confirmed the pioneering finding of Lee et al. (1972) that metabolism of radiolabeled biotin injected intraperitoneally in rats resembles biotin metabolism in normal adult subjects. Using the rat model in our laboratory, we studied the effect of anticonvulsant treatment on biotin biotransformation. The studies did not reproduce the accelerated biotin transformation in response to anticonvulsant drugs that we previously observed in both adults and children. Moreover, with current analytic methods, the blood volume of the rat is too small to accurately determine plasma pharmacokinetics of physiologic doses of radiolabeled biotin.

The pig has been used as a model for studying the digestion and absorption of several nutrients and for studying biotransformation of therapeutic drugs (Aranda et al. 1984, Dvorchik 1981, Juchau 1990, Kearns and Hendry 1990, Kearns et al. 1986, Peggins et al. 1986). Thus, the pig is a potential model for studying biotin biotransformation. However, the urinary metabolite profile has not been reported for pigs. In this study, we sought to determine the urinary excretion rates of biotin and its metabolites after intravenous administration of biotin.

MATERIALS AND METHODS

For the 72 h beginning with administration of radiolabeled biotin, pigs were housed in metabolic cages. Urine samples were collected discrete from feces at four time intervals (0-12, 12-24, 24-48 and 48-72 h).

The identities of BNBMK and biotin sulfone were confirmed by thin layer chromatography of the corresponding HPLC fractions as described previously (Zempleni et al. 1996). Briefly, two different solvent systems were used: 1) butanol-acetic acid-water (4:1:1, v/v/v) (Said and Redha 1987) and 2) butanol alone (Lee et al. 1972). The unlabeled BNBMK standard was visualized using p-dimethylamino-cinnamaldehyde as described previously (McCormick and Roth 1970); the labeled metabolite of [¹⁴C]biotin was identified by scintillation counting of areas of the TLC coating.

RESULTS

Table 1. Urinary excretion of biotin and biotin metabolites from pigs administered [3H]biotin intravenously [View Table]

As noted in Table 1, the predominant radioactive form excreted in the urine was the unchanged vitamin. Less than a third is excreted as BSO. About 15% of radioactivity elutes with the wave front of HPLC. We speculate that this peak contains [³H]acetate or ³H₂O (or both), the expected products of the -oxidation of the valeric acid side chain that contains the ³H label.

Because the radioactive label does not reside in the biotin ring, [³H]biotin studies do not allow direct tracing of most of the metabolites of -oxidation, including BNB. Because BNB is a major metabolite (Mock et al. 1993), this is an important limitation. Notwithstanding, the initial use of [³H]biotin was necessary to assess a metabolite profile at tracer doses. To determine whether the administration of physiologic amounts of biotin resulted in substantially different metabolite profiles, tracer doses of [³H]biotin were administered simultaneously with a physiologic dose of [¹⁴C]biotin (pig no. 4, Table 1). The urinary profile of [³H]metabolites for pig no. 4 was similar to those for pigs receiving only tracer doses of [³H]biotin; the proportion of biotin was modestly greater as expected for a greater total dose of biotin. This experiment provides evidence that the administration of masses of biotin in the physiologic range does not saturate metabolic pathways and substantially alter the metabolic profile. The finding that the percentage of incorporation of the tracer dose varies substantially among individual pigs that were treated identically was surprising to us. Whether this could reflect moment to moment regulation of the handling of intravenous infusion of biotin or variation from animal to animal due to genetic factors of nutritional status is not known.

Biotin, BNB and BSO were excreted in easily detectable amounts (Fig. 2). Trends of the mean percentages of total urinary radioactivity attributable to biotin, BNB and BSO are depicted in Figure 3. The excretion of biotin decreased significantly after the first day. Excretion of BNB and BSO tended to increase, although the increases were not significant. These changes in the metabolite profile seem to reflect equilibration of [¹⁴C]biotin and its metabolites in the body pools. For the cumulative 72-h excretion, biotin accounted for 29 ± 4%, BNB 15 ± 5%, and BSO 16 ± 5% of total excreted radioactivity (means ± SD, n = 5).

The remaining 40% of total radioactivity was distributed among many smaller peaks (Fig. 2). The peak at 7 min was identified as biotin sulfone based on the HPLC retention time compared with authentic biotin sulfone and on the observation that the metabolites of both [¹⁴C]biotin and [³H]biotin retained their radioactive labels. The metabolite that eluted at 22 min was identified as BNBMK based on HPLC retention time compared with authentic BNBMK and on the observation that part of the ³H label was retained. Bisnorbiotin methyl ketone is produced as a by-product of -oxidation (Im et al. 1970, Kazarinoff et al. 1972, McCormick and Wright 1971). [³H]Biotin loses two ³H during the first dehydrogenation from the valeric acid form R--C³H₂--C³H₂--COOH to the enoic acid form R--C³H==C³H--COOH. Hydration of R--C³H==C³H--COOH yields the -hydroxy acid form R--C(OH)³H--C(³H)H--COOH. Another ³H is lost during the next dehydrogenation to the -keto acid form R--C(O)--C(³H)H--COOH. Thus, in theory, biotin retains one of its four ³H during the formation of BNBMK. The identifications of biotin sulfone and BNBMK were confirmed by thin layer chromatography of the radioactive urinary metabolites isolated by HPLC. Each metabolite co-eluted with authentic biotin sulfone and BNBMK on thin layer chromatography using two different solvent systems.

Biotin sulfone accounted for 6 ± 9% and BNBMK accounted for 10 ± 6% of the total radioactivity excreted in the 3-d period. These contributions are similar to the 4% for biotin sulfone and 8% for BNBMK of total biotin plus metabolites detected in human urine by avidin-binding assay (Zempleni et al. 1996).

DISCUSSION

Two observations from Kopinski et al. (1989a) support the hypothesis that the pig is a good model for biotin nutriture in humans. One is the observation that urinary biotin excretion, but not the serum concentration of biotin, is a good index of biotin status. Analogous conclusions have been reported for experimental biotin deficiency induced in adults (Mock et al. 1995). In addition, these investigators reported reduced ratios of unsaturated to saturated fatty acids in lipids extracted from the liver (Kopinski et al. 1989a). These findings are similar to observations concerning serum fatty acids in human biotin deficiency (Mock 1996, Mock et al. 1988). To our knowledge, the biotransformation of biotin and the subsequent urinary excretion of metabolites has not been previously reported in swine.

To study biotin biotransformation and its regulation at the cellular level and under controlled conditions, an animal model that resembles human biotin metabolism would be useful. As a first step in characterizing the pig model, we determined the urinary metabolite profile for tracer and physiologic doses of radiolabeled biotin. Of the total biotin, BNB and BSO excreted over the 72 h following an intravenous injection of [¹⁴C]biotin, biotin accounted for about 51%, BNB for about 24% and BSO for about 25%. Thus, the metabolite profile of radiolabeled biotin was similar to that of unlabeled biotin at steady state in human subjects (Mock et al. 1993).

The similarity seems to extend to minor metabolites as well. Bisnorbiotin methyl ketone and biotin sulfone were excreted in small but easily detectable amounts that resemble the contribution of these metabolites to the steady-state profile of biotin metabolites in free-living human subjects (Zempleni et al. 1996). Bisnorbiotin methyl ketone and biotin sulfone have only recently been identified in the urine of normal adults; these metabolites account for 8% and 4%, respectively, of the total excretion of biotin plus the other identified metabolites (Zempleni et al. 1996).

Bisnorbiotin and BSO appeared in the 0-6-h urine sample, suggesting that removal of biotin from the blood stream, biotransformation into inactive metabolites, and subsequent urinary excretion must be a fairly rapid and ongoing process. A similar rapid excretion was observed for BNBMK.

Because similar amounts of biotin and its two principal metabolites BNB and BSO were excreted by pigs and by humans, we suggest that biotin catabolism in pigs infused with physiologic amounts of radiolabeled biotin resembles that of meal-fed humans. This finding also suggests that the biotransformation of biotin is similar in pigs and humans. Collectively, the data from this study support the conclusion that the pig is an appropriate model for further investigation of biotin biotransformation and regulation of biotin metabolism.