Epimerization of chenodeoxycholic acid to ursodeoxycholic acid by human intestinal lec

Six strains of lecithinase-lipase-negative Clostridia, isolated from human feces, were capable of oxidizing chenodeoxycholic acid to 3a-hydroxy-7-keto-5P-cholanoic acid and of epimerizing it to ursodeoxycholic acid. The identity of the reaction products was confirmed by coniparing their mass spectra, obtained by combined gas-liquid chromatography-mass spectrometry, with those of authentic reference compounds. 3a-Hydroxy-7-keto-5/3-cholanoic acid was reduced by growing cultures of all clostridial strains to chenodeoxycholic acid and to ursodeoxycholic acid, the latter being the preferred conversion product of most strains. However, ursodeoxycholic acid was not attacked by any of the strains. Growth kinetic experiments with three strains showed that chenodeoxycholate was transformed during the log or lag phase. N o bile acid conversion could be seen during the stationary phase. While the concentration of chenodeoxycholic acid decreased and that of ursodeoxycholic acid increased tending towards plateaus, the concentration of 3a-hydroxy-7-keto-5P-cholanoic acid passed through a maximum.a We propose a reaction sequence with 3a-hydroxy-7-keto-5/3-cholanoic acid as an intermediate for the epimerization of chenodeoxycholic acid to ursodeoxycholic acid. This demonstration is the first using isolated bacterial strains.-Edenharder, R., and T. Knaflic. Epimerization of chenodeoxycholic acid to ursodeoxycholic acid by human intestinal lecithinase-lipase-negative Clostridia. J . Lipid Res. 198 1 . 22: 652-658. Supplementary key words bile acids . gas-liquid chromatography-mass spectronietry . growth kinetics and bile acid transformation (1 , 2, 4, 11 18), oxidation of the hydroxyl groups at C3,C7, and CI2, reduction of carbonyl moieties to either CY or /3 hydroxyl functions (1-3, 8, 12, 14-16, 19-21), and epimerization at C, of bile acids with the 5P configuration (22, 23). Other reported microbial degradations include side chain metabolism (24), degradation to nonsteroidal compounds (25), introduction of carbon double bonds into the steroid nucleus (26, 27), hydrolytic cleavage of sulfate esters (28, 29), and esterification of the 3P-hydroxyl group with fatty acids (27). Microbial epimerization of the Sa-hydroxyl group of bile acids has been demonstrated with fecal dilutions (3, 30) and with isolated strains (2, 15). Preliminary reports of 1 Pa-hydroxyl epimerization have also appeared (5, 15). N o intestinal microorganisms responsible for the epimerization of the 7a-hydroxyl group have previously been identified, although fecal bile acids with 7P-hydroxyl groups have been detected (31, 32). More recently, 7cr-epimerization was demonstrated with fecal dilutions (30, 32-35). In this communication we present evidence for 7a-hydroxyl group epimerization in chenodeoxycholic acid by isolated strains of lecithinase-lipase-negative Clostridia and we propose a mechanism for the epimerization process by these bacteria. The primary human bile acids, cholic and chenodeoxycholic acid, are secreted as conjugates of glycine and taurine into the duodenum and are mostly reabsorbed in the ileum. T h e remaining fraction reaches the large bowel, where the bile acids are extensively metabolized by the intestinal flora. Known transformations include deconjugation of conjugated bile acids to yield free bile acids (1 lo), dehydroxylation, mainly at the C, hydroxyl group of the s t e r i d nucleus Abbreviations: The following are systematic names of bile acids referred to in the text by trivial names: cholic acid, 3a,7a,12atrihydroxy-5p-cholanoic acid; chenodeoxycholic acid, 3a,7a-dihydroxy-5P-cholanoic acid; ursodeoxycholic acid, 3a,7a-dihydroxy-50-cholanoic acid; deoxycholic acid, 3a, 1 Pa-dihydroxy-5pcholanoic acid; lithocholic acid, 3a-hydroxy-5p-cholanoic acid. TFA, trifluoroacetyl; HFIP, hexafluoroisopropyl; DMCS, dimethylchlorosilane; GLC, gas-liquid chromatography; GLC-MS, gasliquid chromatography-mass spectrometry. This work is part of‘ the thesis of T. Knaflic to be submitted as partial requirement for the degree of “Doktor der Medizin” to the faculty of medicine, Johannes Gutenberg Universitat MainL. 652 Journal of Lipid Research Volume 22, 1981 by gest, on O cber 7, 2017 w w w .j.org D ow nladed fom MATERIAL AND METHODS Bacterial strains, media, and culture conditions Lecithinase-lipase-negative Clostridia, isolated from feces of patients with large bowel cancer, were obtained from Dr. W. Doll, Institut fur Hygiene und Mikrobiologie der Universitat Wiirzburg (D-8700 Wurzburg). These strains were originally grown from spores, surviving 10 min of heating at 80°C. Stock cultures were maintained in a cooked meat medium (Oxoid2, CM 81), supplemented with Schaedler broth (Oxoid2, CM 497) and 1 ml/liter of 0.1% resazurine solution, pH 7.6. Unless otherwise stated, this medium containing 0.25 mM sodium chenodeoxycholate, ursodeoxycholate, or 3a-hydroxy-7-keto-5P-cholanoate was used for transformation experiments. T o account for the possible necessity of enzyme induction, two serial bacterial transfers were performed; both tests were incubated anaerobically at 37°C until the late stationary phase. Bile acid transformation was then checked with selected samples. All media used were prereduced and anaerobically sterilized. Bacteria were transferred as described by Holdeman and Moore (36), with the modification that a multiple gas supply head with glass Pasteur pipets was used in a sterile hood (CEAG-Schirp, model DF600)3. In general, screw-cap vials with additional butyl rubber stoppers were used for bacterial cultivation.