of Czl Bile Acids from Plant Sterols in the Rat*

Transcription

1 THE JOURNAL. OF BIOLOGICAL CHEMISTRY <o 1990 by The American Society for Biochemistry and Molecular Biology, Inc. Vol. 265, No. 14, Issue of May 15, pp Printed in U.S.A. Formation of Czl Bile Acids from Plant Sterols in the Rat* (Received for publication, July 10, 1989) Kirsten Muri BobergS, Erik Lund, Jonas 8lund$, and Ingemar BjGrkhem$ From the$ Institute of Clinical Biochemistry, University of Oslo, Rikshospitalet, Oslo, Norway and the SDepartment of Clinical Chemistry and Research Center, Karolinska Institutet, Huddinge Hospital, Huddinge, Swedens Formation of bile acids from sitosterol in bile-fistulated female Wistar rats was studied with use of 4- Y!- labeled sitosterol and sitosterol labeled with 3H in specific positions. The major part (about 75%) of the 14C radioactivity recovered as bile acids in bile after intravenous administration of [4-14C]sitosterol was found to be considerably more polar than cholic acid, and only trace amounts of radioactivity had chromatographic properties similar to those of cholic acid and chenodeoxycholic acid. It was shown that polar metabolites were formed by intermediate oxidation of the 38-hydroxyl group (loss of 3H from 3a-3H-labeled sitosterol) and that the most polar fraction did not contain a hydroxyl group at C, (retention of 3H in 7a,7/3-3Hz-labeled sitosterol). Furthermore, the polar metabolites had lost at least the terminal 6 or 7 carbon atoms of the side chain (loss of 3H from 22,23-3H2- and 24,2S3H2-labeled sitosterol). Experiments with 3H-labeled 7cy-hydroxysitosterol and 4- *C-labeled 26-hydroxysitosterol showed that none of these compounds was an efficient precursor to the polar metabolites. By analysis of purified most polar products of [4- *Cl sitosterol by radio-gas chromatography and the same products of 7a,7,!?-[2H2]sitosterol by combined gas chromatography-mass spectrometry, two major metabolites could be identified as C& bile acids. One metabolite had three hydroxyl groups (3cu, 15, and unknown), and one had two hydroxyl groups (3~, 15) and one keto group. Considerably less Czl bile acids were formed from [4-14C]sitosterol in male than in female Wistar rats. The Czl bile acids formed in male rats did not contain a 15-hydroxyl group. Conversion of a [4- *C]sitosterol into CZl bile acids did also occur in adrenalectomized and ovariectomized rats, indicating that endocrine tissues are not involved. Experiments with isolated perfused liver gave direct evidence that the overall conversion of sitosterol into Cz, bile acids occurs in this organ. Intravenously injected 7q7&3H-labeled campesterol gave a product pattern identical to that of 4-14C-labeled sitosterol. Possible mechanisms for hepatic conversion of sitosterol and campesterol into Czl bile acids are discussed. The plant sterol sitosterol is structurally similar to cholesterol, differing only by the presence of an ethyl substituent at Cz4 in the side chain (1). The ethyl group has a marked effect * This study was supported by the Norwegian Council on Cardiovascular Diseases and the Swedish Medical Research Council. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 USC. Section 1734 solely to indicate this fact on the absorbability and metabolism of the steroid. In particular, the mechanism of oxidation of the steroid side chain to yield a bile acid must be different between cholesterol and sitosterol. The conversion of cholesterol to cholic and chenodeoxycholic acid involves intermediate formation of 3cr,7a,12~ trihydroxy-5@-cholestanoic acid and 3cu,7cu-dihydroxy-5&cholestanoic acid, respectively (2). The CZ7 steroid side chain is then oxidized by a sequence of reactions similar to that in /3- oxidation of fatty acids (2). These reactions are most efficiently catalyzed by the peroxisomal fraction of the liver (3-5). In the conversion of 3a,7cu,l2Lu-trihydroxy-5P-cholestanoic acid into cholic acid an intermediate formation of 3a,7a,12a,24-tetrahydroxy-5P-cholestanoic acid takes place (6). This reaction proceeds via a AZ4 unsaturated steroid (3a,701,12cu-trihydroxy-24-en-Sfi-cholestanoic acid) and thus involves the participation of a desaturase and a hydratase (6, 7). The presence of an ethyl group at Cn4 should prevent or at least obstruct a side chain oxidation according to the above mechanism. The possibility must be considered that sitosterol is first dealkylated to cholesterol and then converted into bile acids. Attempts to demonstrate such a pathway in mammals have failed hitherto, however (1, 8). Furthermore, sitosterol is a poor substrate for 7cu-hydroxylation, the initial and ratelimiting reaction in bile acid biosynthesis from cholesterol (9). In the rat (10, 11) and in the monkey (12), attempts to demonstrate conversion of sitosterol into natural bile acids have failed. Surprisingly, however, formation of some cholic, chenodeoxycholic, and deoxycholic acid from injected [22,23-3H]sitosterol in man has been reported (13). Several studies have shown that sitosterol is converted into polar compounds in the bile acid fraction of rat bile (8,10,14), but no conclusive identification has been provided. Studies on the metabolism of sitosterol have been hampered by methodological problems. Due to the low solubility and absorbability, only relatively small amounts can be administered, and identification of the products formed is difficult due to the presence of a great excess of products from cholesterol. In the present work we have studied the metabolism of sitosterol in rats by using [4-Y!]sitosterol in combination with sitosterol specifically labeled with H in various posi- tions. Evidence is presented that Gel bile acids are major metabolites of sitosterol. MATERIALS AND METHODS Unlabeled Compounds Sitosterol (24-ethyl-5-cholesten-38-01) (98% pure) was purchased from Alltech Associates Inc. (Deerfield, IL). The purity of this material was ascertained by HPLC (see below). Stigmasterol (24-ethyl- 5,22-cholestadiene-3&ol), fucosterol(24-ethyl-5,24(28)cholestadiene- The abbreviations used are: HPLC, high pressure liquid chromatography; GC, gas chromatography.

2 7968 Conversion of Plant Sterols to Bile Acids 3&01), and a plant sterol mixture (S 5753, approximately 60% sitosterol and 40% campesterol) were products of Sigma. Cholic, deoxycholic, and lithocholic acid were obtained from Supelco Inc. (Bellefonte, PA). and chenodeoxvcholic acid from Larodan Fine Chemicals (Mali& Sweden). All other chemicals and solvents were standard commercial high purity materials. A trihydroxy CT:! bile acid standard was synthesized from 22- cholanoic acid-5-en (obtained from Steraloids, Inc., Wilton, NH). The latter material was treated with an excess of m-chloroperbenzoic acid in chloroform for 1 h at room temperature. The reaction product (mainly 5a,6cu-epoxy-22-cholanoic acid-3/3-01) was extracted from the reaction mixture and treated with 3 M HCl in ethanol/water (l:l, v/v) for 3 h at 70 C. The product isolated had GC properties and a mass spectrum (as trimethylsilyl ether-methyl ester) as expected for 22-cholanoic acid-3&5a,6&triol. A trihydroxy Cnl bile acid standard, 21cholanoic acid-3p,501,6atriol, was synthesized by oxidation of 3P-hydroxypregn-5-en-21-oic acid with osmium tetroxide. 3P-Hydroxypregn-5-en-21-oic acid was synthesized from 3P-hydroxyandrost-5-en-17-one as described by Pyrek et al. (15). The final product had GC properties and a mass spectrum (as trimethylsilyl ether) as expected. A dihydroxy CZ2 bile acid standard with one double bond and two hydroxyl groups (22cholanoic acid-5-ene-3p,7ry-diol) was synthesized from the above 22-cholanoic acid-5-en using the same procedure as described below for synthesis of 7cu-[4- Clhydroxysitosterol from [4- Clsitosterol. The product isolated had GC properties and a mass spectrum (as trimethylsilyl ether-methyl ester) as expected. Labeled Compounds [4-i4C]Sitosterol, 56 mci/mmol, was obtained from Amersham Corp. The material was purified by HPLC immediately before preparation of the injection mixture. [3a- H]Sitosterol, 75 mci/mmol, was obtained by oxidation of sitosterol to the corresponding 3-0x0-A4 compound (16) and subsequent isomerization to the corresponding 3-0x0-A5 compound (17). The latter compound was reduced with H-labeled sodium borohydride (17) to yield [3a-3H]sitosterol. The material was purified by thin-layer chromatography (toluene/ethyl acetate (3:2, v/v) as moving phase) and HPLC before use. [7a,7fl-3Hz]Sitosterol, 757 &i/mmol, was prepared by oxidation of sitosterol acetate with CrOB to yield 7-oxo-sitosterol acetate (18). The material was purified by aluminum oxide chromatography (Woelm III) and was pure as judged by thin-layer chromatography and gasliquid chromatography as trimethylsilyl ether. The material was then converted into 7-oxo-sitosterol-ethylene-mercaptol by the following procedure. 7-oxoSitostero1, 50 mg, was dissolved in 5 ml of ethanol and mixed thoroughly with 0.5 ml of 1,2-dimercaptoethane (Aldrich) and 2.5 ml of 6 N HCl. The mixture was allowed to stand overnight at a temperature of 45 C. The reaction mixture was then diluted with water and extracted with ethyl ether. The organic phase was washed twice with saturated copper(b) acetate, dried over anhydrous MgSO,, and evaporated. The yield of the mercaptol was 32 mg. The material was pure as judged by thin-layer chromatography and gas chromatography of the trimethylsilyl ether. In order to reduce the mercaptol, tritiated Raney Nickel had to be nrenared. A solution of 1.8 e of NaOMe in a mixture of 5 ml of H20. iod@l of tritiated water (0.5Curie, Amersham), and 5 ml of methanol was prepared. In order to dissolve all NaOMe, water was added dropwise until the solution was clear. Nickel-aluminum alloy, 2 g, was added in small portions over a period of 30 min. The reaction mixture was allowed to stand overnight. The liquid was decanted and the catalyst washed several times with small portions of water. The mercaptol, 15 mg, dissolved in a small volume of ether, was added to approximately 1 g of the above tritiated Raney Nickel covered by ml of ether. After the addition, the slurry was stirred for 2 h in order to complete the reaction. The catalyst was filtered off, the filtrate was dried (MgSOJ, and evaporated. The yield was about 4 mg. The material was pure as judged by thin-layer chromatography and gas chromatography as trimethylsilyl ether. 17a,7& H21Sitostero1 was prepared as above with the exception that deuterized Raney Nickel was used. The latter reagent was obtained bv dissolving 1.8 g of NaOMe in a mixture of 5 ml of *Hz0 (99% pure; Sigma). In order to dissolve all NaOMe, the deuterated water was added dropwise until the solution was clear. The nickelaluminum alloy was added in small portions over a period of 30 min. The reaction mixture was allowed to stand overnight. The liquid was decanted, and the catalyst was washed several times with small portions of deuterated water. The material obtained after reduction of the above 7-oxo-sitosterolethylene-mercaptol with the deuterized Raney Nickel was pure as judged by thin-layer chromatography and gas chromatography of the trimethylsilyl ether. The isotope composition when calculated according to Biemann (19) was the following: 10% unlabeled, 34% monodeuterated, and 56% dideuterated molecules. [22,23-3H2]Sitosterol, 222 &i/mmol, was prepared from stigmasterol and tritium gas. PtOL, 3 mg, was activated in diethyl ether, 1 ml, by bubbling HP gas through the suspension for some minutes. A solution of stigmasterol, 3 mg, and tritiated water, 3 ~1 in 1 ml of diethyl ether, was added. Tritiated hydrogen was bubbled through this solution for 4 min. The catalyst was filtered off, and the solvent was evaporated. The yield was about 1.7 mg. The material was pure as judged by thin-layer chromatography and gas chromatography as trimethylsilyl ether. That this procedure gives incorporation of tritium only in the steroid side chain and not in the steroid nucleus was shown by using deuterium gas instead of tritium gas. Under such conditions the material isolated contained deuterium, all of which was located in the side chain according to mass spectrometry of the trimethylsilyl ether. [24,28-3H2]Sitostero1, 263 &i/mmol, was prepared essentially as above with the exception that fucosterol was used as starting material. A similar yield as above was obtained, and the material was pure as judged by thin-layer chromatography and gas chromatography as trimethylsilyl ether. 7a-[7P-3H]Hydroxysitosterol, 20 mci/mmol was prepared from the above 7-oxo-sitosterol by reduction with tritium-labeled sodium borohydride. A mixture of the 7a- and 7p-hydroxy isomers was obtained. The 7a-hydroxy isomer was isolated by preparative thin-layer chromatography using toluene/ethyl acetate (3:7, v/v) as moving phase. The material obtained was pure as judged by thin-layer chromatography and gas chromatography of the trimethylsilyl ether derivative. 26-[4-4C]Hydroxysitosterol (56 mci/mmol, assuming no endogenous dilution) was prepared biosynthetically from [4-i4C]sitosterol (see below). The radiochemical purity was ascertained by radioscanning of thin-layer chromatography. Only one peak was obtained in the radiochromatography with the properties expected for that of 26- hydroxysitosterol. The material contained, however, contaminating unlabeled compounds as judged by gas chromatography as trimethylsilyl ether. The material may also contain 29-hydroxysitosterol which should have about the same chromatographic properties as 26- hydroxysitosterol and which has a mass spectrum almost identical to that compound (20). All labeled sterols were nurified bv HPLC immediatelv before preparation of the injection mixtures. The sterols were then dissolved in soybean oil by sonication and mixed with Intralipid, 200 mg/ml (KabiVitrum AB, Stockholm, Sweden). Animals Wistar rats ( g) were obtained from Miillegaards Breeding Center, Denmark, and with a few specified exceptions, only female rats were used. During anesthesia with an intramuscular injection of Hypnorm-dormicum (2.5 mg of fluanisonum, 0.05 mg of fentanyl, and 1.25 mg of dormicum/ml; 0.15 ml/100 g of body weight), the bile duct was cannulated with a polyethylene tube. The sterol preparation (l- 20 &i, mg) was administered intravenously in a tail vein in most experiments. The animals were kept in restraining cages with free access to water and ordinary rat diet (R 34, Ewos Sverige AB, Sodertalie, Sweden). Subcutaneous injections of 0.9% (w/v) NaCl were given in addition to compensate for water losses during the operation and biliary drainage. Bile was collected 24 and 48 h after the sterol injection. The studies were approved by the ethical committee for biological experiments on animals in Norway. Liver Perfusion Isolated livers from female Wistar rats were perfused essentially as described by Seglen (21). The perfusion medium was ml of a bicarbonate buffer, continuously gassed with 95% 02, 5% CO, by use of an oxygenator chamber with a nylon net on stainless steel net SUDDO~~. The temperature was kept constant at 37 C, and ph was [4-4C]Sitosterol, 14 x i06 dpm, was dissolved in 300 ~1 of ethanol and 1 ml of Intralinid 20% and added to the perfusate. The bile duct was cannulated before the perfusion, and bile was collected during the experimental period of 1 h. Hydrolysis of the bile, extrac-

3 Conversion of Plant Sterols to Bile Acids 7969 tion of bile acids, and HPLC analysis of the ether extract were carried out as described above. Hydrolysis and Extraction Procedures Aliquots of the bile were hydrolyzed in 4 M KOH at 120 C for 18 h. Neutral sterols were extracted with n-hexane. After acidification of the water phase to ph 1.0 with hydrochloric acid, bile acids were extracted with ethyl ether. The ether extract was washed to neutrality with water, blown to dryness under a stream of Nz, and redissolved in chloroform/methanol (2:1, v/v). Suitable aliquots of the hexane and ether extracts were assayed for radioactivity (see below). Experiments were carried out to exclude the possibility that the main radioactivity peaks originated from artificial decomposition of the?-labeled sitosterol molecule. C-Labeled sitosterol was dissolved in ethanol or bile from an untreated rat and hydrolyzed under the same conditions as described above. After treatment in ethanol, the hexane extraction of sitosterol was essentially complete. After ether extraction of the acidified mixture, the material was subjected to HPLC analysis. No radioactivity was detected in the fractions of HPLC containing the major metabolites from injected sitosterol. Chromatography HPLC-Bile acids were separated by HPLC as described previously (22). A Supelcosil LC-18 column (250 X 4.6 mm, 5-pm particle size, from Supelco Inc.) was used with methanol/water (70:30, v/v) as mobile phase and a flow rate of 1 ml/min. The water was adjusted to ph 3.0 with phosphoric acid. The chromatograph was equipped with a constant flow pump (Consta Metric III, Laboratory Data Control, Milton Roy Co., St. Petersburg, FL) and a differential refractometer (R-401, Waters Assoc., Milford, MA). Typical retention times (min) were: 22-cholanoic acid-3&5a,68-triol, 5; 22-cholanoic acid-5-ene- 3&7a-diol, 8; cholic acid, 24; chenodeoxycholic acid, 50; deoxycholic acid, 54; and lithocholic acid, 120. Purification of the sterol precursors and analysis of the hexane extracts of the bile were performed with methanol/water (95:5, v/v) as the mobile phase. The retention times (min) were: 7ol-hydroxysitosterol, 14; 26-hydroxycholesterol, 8; 26- hydroxysitosterol, 10; cholesterol, 41; and sitosterol, 52. Fractions of 1 ml were collected from the outlet of the chromatograph, added to 10 ml of Packard Insta Gel II scintillation liquid (Packard Instrument Co., Downers Grove, IL), and assayed for radioactivity in a Packard Tri-Carb liquid scintillation analyzer, mode Simultaneous estimation of H and C was performed with automatic quenching correction. Zon-exchange Chromatography-For the purpose of structure identification of metabolites from injected sitosterol, fractions containing radioactivity peaks were collected from repeated HPLC injections. These were purified further on a column of g of Lipidexdiethylaminohydroxypropyl; (a generous gift from Prof. Jan Sjovall, Karolinska Institutet) in 72% (v/v) ethanol, according to Almi et al. (23). After application of the sample dissolved in 1 ml of 72% ethanol, the column was eluted at a flow rate of ml/h. An initial washing step with 20 ml of 72% ethanol was followed by elution of unconjugated bile acids with 7.5 ml of 0.1 M acetic acid in 72% ethanol (ph 3.85). Glycine and taurine conjugates and sulfated bile acids could be eluted consecutively. This system thus ensured that possible conjugates of sitosterol-derived bile acids had been adequately hydrolyzed. All radioactive products of labeled sitosterol appeared in the free bile acid fraction. The elution was performed with [3H]deoxycho1ic acid as an internal standard. The radioactivity-containing eluates were taken to dryness and subjected to an additional purification by the same HPLC as above. Gas Chromatography-Mass Spectrometry-Samples of bile purified according to the above procedures were converted to the methyl estertrimethylsilyl ether derivatives (24) and analyzed by combined GCmass spectrometry using an LKB 2091 instrument equipped with 30- m X 0.25-mm DB-1301 column (0.25-pm phase) and operating at 70 ev. The GC oven was held at an initial temperature of 180 C for 2 min and then linearly programmed at a rate of 8 C/min to 290 C. Radio-gas Chromatography-Radio-gas chromatography was performed as described in Ref. 25 using a packed column 1% OV 101 as a stationary phase. The chromatograph was run isothermally at 220 C with a carrier gas flow of 20 ml/min. Thin-layer Chromatography-TLC for isolation of some of the synthesized labeled sterols was carried out on plates precoated with Silica Gel H (Merck). Preparation and Incubation of Liver Mitochondria-The liver was chilled on ice immediately after removal, and a 10% (w/v) homogenate was prepared in 0.25 M sucrose. The nuclear fraction was removed by centrifugation at 800 x g for 10 min. Centrifugation of the supernatant at 65OOxg for 20 min yielded a mitochondrial pellet that was washed twice and resuspended in 50 mm Tris acetate buffer (ph 7.4), reconstituting the original volume. [4- CJSitosterol (5 X lo6 dpm) or [4-?]cholesterol(5 x lo6 dpm) was dissolved in 50 ~1 of acetone and added to an incubation mixture consisting of 1 ml of the above mitochondrial suspension, 10 mm MgCl,, 1.5 mm isocitrate, and 50 mm Tris acetate buffer (ph 7.4) in a total volume of 5 ml. The reaction was carried out for 30 min, terminated, and extracted by the addition of 10 ml of chloroform/methanol (2:1, v/v). The lipid extracts were subjected to HPLC using methanol/water (95:5, v/v) as eluent. 26-Hydroxysitosterol could be isolated by this method in a yield of about 0.5%. RESULTS Pattern of Metabolites from [4- VISitosterol-After intravenous or intraperitoneal administration of [4-Wlsitosterol to female bile fistula rats, 7.2 f 1.1% (mean f S.E., n = 4) of the injected dose was recovered in bile after 24 h. During the next 24 h, an additional 6.2 & 1.2% was recovered. This degree of excretion is in accordance with previous investigations (8). Of the recovered radioactivity, 85 f 2% had extraction properties as bile acids. The ether extract of a 24-h bile sample from a rat injected with [4-Wlsitosterol was analyzed by HPLC, revealing the chromatogram shown in Fig. 1. The distribution of radioactivity within this chromatogram is described further in Table I, in which data from several experiments are included. The majority (75%) of the 14C-labeled acidic products were very polar compounds being eluted before cholic acid in the reversed-phase HPLC system. Two major radioactivity peaks were detected, the largest appearing at 4 ml and the other at 9 ml. These fractions contained 26% (fractions 4 and 5 combined), and 18% (fractions 8, 9, and 10 combined), respectively, of the total amount of radioactivity. Smaller amounts J Elutlon volume (ml) FIG. 1. Reversed-phase HPLC of ether extract of bile from a rat injected with [4- %]sitosterol. The mobile phase was methanol/water (70:30, v/v; the water adjusted to ph 3.0) at a flow rate of 1 ml/min. 14C activity (M) in l-ml fractions and refractive index detector response (---) are shown. Peak Z, cholic acid; peak ZZ, chenodeoxycholic acid. TABLE Distribution of radioactivity in reversed-phase HPLC of ether extracts of bile from rats injected [4- *Clsitosterol The mobile phase was methanol/water (70:30, v/v; the water adjusted to ph 3.0). The injected dose varied between 3 and 20 &i in the different experiments. Distribution I of radioactivity More polar HPLC HPLC Region of Region of than cholic fractions fractions cholic scid chenodeoxyacid cholic acid Mean (n = 8) SE %

4 I 7970 Conversion of Plant SteroLs to Bile Acids of radioactive compounds co-chromatographed with cholic acid (8%) and chenode-oxycholic/deoxycholic acid (9%). Further analysis of the latter material by radio-gas chromatography showed that the retention time was clearly different from that of cholic acid and chenodeoxycholic acid (as methyl ester-trimethylsilyl derivatives). Only about 5% of the radioactivity represented less polar compounds than deoxycholic acid, with very small peaks eluting at about 70 and 110 ml. Lithocholic acid appeared after 120 ml in this system. Most probably the major part of the metabolites of [4-14C] sitosterol was present as conjugates in bile. Attempts to extract the metabolites with ether from unhydrolyzed acidified bile failed, and most of the radioactivity was retained in the water phase. After hydrolysis, however, most of the radioactivity could be obtained in the ether phase when extracted as above. To obtain more specific information on the structure of the major polar metabolites derived from sitosterol, experiments with various double-isotope mixtures were performed. Experiments with Sitosterol Labeled in Specific Positions with 3H--3H-label was incorporated into specific positions of the sitosterol molecule, which was then presented to bile fistula rats in combination with [4-4C]sitosterol. A summary of this series of experiments is given in Table II. [~w~h] Sitosterol was injected together with [4-4C]sitosterol with the purpose of unveiling whether the 3/3-hydroxyl group had been oxidized or not. In the metabolism of cholesterol to bile acids, the 7a-hydroxylation is followed by a saturation of the double bond and an epimerization of the 3@-hydroxyl group to the 3a-position (2). A 3-oxo,A4 steroid is intermediate in the reaction (2). The 3H to 14C ratio of 12.0 in the injected material was reduced to 0.11 and 0.28 in the metabolites eluted at 4 and 9 ml, respectively. This reduction in the 3H to 14C ratio indicates that 3H has been lost from the 3cu-position of sitosterol, most likely by the mechanism described above. Accordingly, the compounds in fractions 4 and 9 are likely to be 3ahydroxylated saturated steroids in analogy to the normal bile acids. After injection of a mixture of [7-3H2]sitosterol and [4-i4C] sitosterol with a 3H to 14C ratio of 0.98, no loss of 3H relative to i C occurred to fraction 4. Hydroxylation at C7 is therefore not a step in the formation of the metabolites in fraction 4. In fact the 3H to 14C ratio was increased by about 50% in fraction 4 as compared with the ratio in the injected material. This may be due to an isotope effect. The H to 14C ratio in fraction 9 was about half that in the injected materials, indicating that a 7~ or 7@-hydroxyl group could be present. The mechanism of side chain oxidation was elucidated by injecting sitosterol, specifically 3H labeled in the side chain, TABLE H to 92 ratios in injection mixture and HPLC fractions of ether extracts of bile from rats injected [4- C]sitosterol + sitosterol H labeled in specific positions II Combination of H/W in sterols (dpm x 106) ln ection miwtllre Fraction 4 Fraction 9 3a-[ H]Sitosterol (11.52) + [4- T!]sitosterol (0.96) together with the isotope containing 14C at C4 in the steroid nucleus. Two different variants of side chain labeling were used. [22,23-3H2]Sitostero1, synthesized from stigmasterol, was administered together with [4-i4C]sitosterol with a 3H to 14C ratio of As shown in Table II, practically all of the 3H label was removed from the major metabolites. Cleavage of the CZs side chain at a proximal site is the most likely interpretation of this result. Thus, a fragment containing at least 6 carbon atoms had been lost. Support for this contention was provided in the next experiment in which [24,28-3H2]sitosterol, synthesized from fucosterol, was the precursor. The 3H label was almost entirely lost during formation of the major metabolites from this compound as well (Table II). At some stage the C&-C& substituent had therefore been removed. Attempts to Identify Sitosterol Metabolites in Fractions 4 and 9 of HPLC-The presence of large amounts of several other polar bile components rendered structure identification of the major metabolites very difficult. As shown in Fig. 2, radio-gas chromatography of the trimethylsilyl ether-methyl ester derivative of HPLC fraction 4 after purification gave at least three peaks containing radioactivity (peaks A-C). Under the specific chromatographic conditions used (packed column OV 101, constant temperature at 220 C), the retention time of the three radioactive peaks was about half that of cholic acid. All of the above chromatographic properties are those expected for Cpl or CZ2 bile acids with at least two hydroxyl groups in the steroid nucleus or side chain. A trihydroxylated CZZ bile acid (see Materials and Methods ) appeared in fraction 5 in the HPLC system used and had a retention time on GC similar to the polar metabolites of [4-4C]sitosterol. In spite of extensive purification of the labeled sitosterol metabolites prior to analysis by combined gas chromatography-mass spectrometry (HPLC + Lipidex-diethylaminohydroxypropyl + HPLC), several peaks occurred in the capillary gas chromatogram. It was difficult to state exactly which of the peaks occurring in the capillary gas chromatography were metabolites of sitosterol. To approach this problem we synthesized deuterium-labeled sitosterol in order to detect the [7- Ha]Sitosterol (0.63) + [4-l%] sitosterol (0.64) [22,23-3H2]Sitostero1 (1.76) + [4- Clsitosterol (1.05) [24,28-3H2]Sitostero1 (1.02) + [4- Clsitosterol (0.57) Time FIG. 2. Radio-gas chromatogram (1% OD 101, isothermally at 220 C) of the trimethylsilyl ether-methyl ester derivative of HPLC fraction 4 of bile from a rat injected with [4-14C] sitosterol. Imlnl

5 Conversion of Plant Sterols to Bile Acids metabolites from the pattern of isotopes in the mass spectrometric analysis. In one case the sitosterol administered had an isotope content of 10,34, and 56% of unlabeled, monodeuterated, and dideuterated steroid, respectively. The sitosterol reisolated from bile had the following composition: 86,4, and 10% for unlabeled, monodeuterated, and dideuterated steroid, respectively. The more than &fold dilution of isotope in the animal experiment could be due to sitosterol present in the soybean oil and in the Intralipid used for administration, as well as in the diet. The sitosterol metabolites could thus be expected to contain about the same low isotope content as the reisolated sitosterol. Fig. 3 shows a typical capillary gas chromatogram contained in connection with analysis of the trimethylsilyl ether-methyl ester derivative of purified fraction 4. Most probably peak I corresponds to the radioactive compound A in Fig. 2. Furthermore, peak IV corresponds to the radioactive compound C in Fig. 2. It is possible that peaks II and III together form peak B in the radio chromatography shown in Fig. 2. When analyzing materials from experiments with [ HJsitosterol, a low isotope incorporation could be found at least in the compounds corresponding to peaks I and III and possibly IV. Thus, these compounds are likely to be metabolites of sitosterol. At least peaks I, III, and IV were found in all purified fractions 4 analyzed. The mass spectrum of the compound corresponding to peak I showed a weak ion at m/z 596, probably the molecular ion (Fig. 4). In addition there were abundant ions at m/z 581 (M- 15), 506 (M-90), 491 (M-90-15), 496 (M-100), 481 (M-lOO- 15), 447 (M-90-59), 416 (M-2 x 90), 406 (M ), 401 (M- 2 X 90-15), 326 (M-3 X 90), 311 (M-3 X 90-15). All these ions are in accordance with a trihydroxy Cpl bile acid. A prominent ion was also present at m/z 202. One fraction 4 was analyzed as trimethylsilyl ether only, which caused the higher ions of the compound inclusive that at m/z 202 to shift upward with 58 mass units. This is in accordance with the contention that there is one carboxylic group in the compound. The fragment at m/z 202 thus contains the two-carbon side chain. In addition, the fragment must contain three carbon atoms from the D-ring of the steroid and a trimethylsilyl group. The trimethylsilyl group may be located at CIS, C6, C17, or C&. If all three hydroxyl groups are present in the steroid nucleus, a peak Time FIG. 3. Capillary gas chromatogram (DB-1301) of trimethysilyl ether-methyl ester derivative of purified (see Materials and Methods ) HPLC fraction 4 of bile from a rat injected with [4- %]sitosterol. (min) FIG. 4. Mass spectrum of the compound corresponding to peak I of the capillary gas chromatogram shown in Fig. 3. Purified HPLC fraction 4 of bile from a rat injected with [4- C] sitosterol was analyzed as the trimethylsilyl ether-methyl ester derivative. should be expected at m/z 253. Due to a high background, however, it was not possible to evaluate by these means how many hydroxyl groups are present in the nucleus. The loss of a fragment with m/z 100 from the molecular ion and from M-90 ion is of interest. A most probable explanation for this loss is presence of a trimethylsilyloxo function at CIS and a CH2-COOCHa group at C,T. The fragment would then be lost by cleavage between CIS and C,G and between C,, and CIT. The fragmentation is in complete accord with that reported for trimethylsilyl ether derivatives of 15-hydroxylated corticosteroid metabolites (26), 15-hydroxylated CZI bile acids (27), and 15-hydroxylated 24-ethyl-SD-cholestan-3a-ol (28). According to the above experiments with 3a-[3H]sitosterol, the metabolites from sitosterol are likely to contain a 3cyhydroxyl group. This was confirmed by oxidation of the material with 3cu-hydroxysteroid dehydrogenase and NAD. This treatment led to a drastic reduction of the height of peak I in the gas chromatography. It may thus be concluded that peak I corresponds to a SLYhydroxylated CZ1 bile acid with two additional hydroxyl groups in the steroid nucleus. One of these two hydroxyl groups is most probably located at Cls. A synthetic trihydroxylated C& bile acid, 21-cholanoic acid- 3/$5o,Ga-trio1 gave a mass spectrum (as methyl ester-trimethylsilyl derivative) very similar to that of peak I, with the exception that the specific fragments due to the trimethylsilyloxo function at CIS were lacking. The compound corresponding to peak III in the chromatogram in Fig. 3 was found to have a molecular ion at m/z 522 (as methyl ester-trimethylsilyl ether) (Fig. 5). Other prominent peaks were at m/z 507 (M-15), 462 (M-60), 432 (M-90), 422 (M-loo), 417 (M-90-15), 372 (M-90-60), 342 (M-2 x 90) and 202. These fragments could be expected for a Csl bile acid similar to that corresponding to peak I, with one of the three hydroxyl groups oxidized to a keto group. Another possibility could be a C,, bile acid with two hydroxyl groups. After treatment of the material with sodium borohydride, peak III disappeared in the gas chromatography, indicating the presence of a keto group. After treatment with a 3ohydroxysteroid dehydrogenase and NAD, the peak was decreased in height, indicating the presence of a hydroxyl group in 3ol-position. The loss of a fragment with m/z 100 and the presence of the ion at m/z 202 in the mass spectrum of the compound corresponding to peak III indicate that this compound also has a CH,-COOH side chain and a hydroxyl group at CIS. The

6 7972 Conversion of Plant Sterols to Bile Acids, 202 I I I 342 IQJh~i,~I~ I ~,lllul, 372 II Jll, 452 r /,I I, FIG. 5. Mass spectrum of the compound corresponding to peak III of the capillary gas chromatogram shown in Fig. 3. Purified HPLC fraction 4 of bile from a rat injected with [4-l%] sitosterol was analyzed as the trimethysilyl ether-methyl ester derivative. keto group must be located in the steroid nucleus, but it was not possible to locate it from the mass spectrometric data available. A keto group at C, is, however, excluded in view of the retention of 3H in [701,7P-3H2]sitosterol during its conversion into the most polar metabolites. Peak IV was found to give a mass spectrum with prominent ions at m/z 524, 434, 424, 416, and 202. The 18 shift is most probably due to the loss of water, indicating a free hydroxyl group. This hydroxyl group must be sterically hindered for the trimethylsilyl reagent. After derivatization under more vigorous conditions (70 C overnight), peak IV disappeared completely in the gas chromatography, whereas the ratio of peak I to peak III had increased. The most likely explanation is that peaks I and IV correspond to the same compound, which is difficult to derivatize completely with trimethylsilyl reagent unless rather tedious conditions are used. This also means that the hydroxyl group in an unknown position in the compound corresponding to peak IV must be at least to some extent sterically hindered. As could be expected from above, treatment with 3a-hydroxysteroid dehydrogenase and NAD led to disappearance of peak IV, indicating that it is a 30(- hydroxysteroid. In order to confirm further the identity of the major metabolites in fraction 4 as 15-hydroxylated CZ1 bile acids, the fraction was oxidized with chromium oxide. After conversion into the methyl ester-trimethylsilyl derivative, the material gave one predominant peak with chromatographic properties as expected for a CZ1 bile acid with three keto groups. The mass spectrum of the compound showed an intense ion at m/ z 374 (M) and a weak one at 346 (M-28), as expected for a cyclic ketone. In addition, there were ions at m/z 274 (Mloo), 246 (M-128), and m/z 128. The ion at m/z 128 was the base peak. The ion at m/z 128 is likely to correspond to the ion at m/z 202 discussed above, containing a keto group instead of a trimethylsilyloxo group. In addition to the above predominant peak, there was a minor peak containing material with essentially the same mass spectrum. Most probably this represents a positional isomer. The mass spectra of the oxidized material in fraction 4 are those expected for oxidized products of the above trihydroxy- and dihydroxy-mono-keto CZ1 bile acids. The uniformity of the products from oxidation indicates that this fraction must contain almost entirely CZl bile acids with three oxygen functions. At least 80-90% of the material must be 15-hydroxylated steroids. The material in HPLC fractions 7-9 could not be identified with certainty due to presence of contaminating nonsteroid compounds. A predominant compound in this fraction had a mass spectrum as could be expected for a Cpl bile acid with two trimethylsilyloxo groups. This compound did not carry a 15-hydroxyl group however, and alternative explanations for the fragmentation could not be completely excluded. In one analysis of HPLC fractions 7-9, small amounts were seen of a compound with a mass spectrum consistent with a trihydroxylated steroid nucleus and a Czs side chain with two carboxyl groups. This compound was not seen in all analyses and could be a metabolite of campesterol. Experiments with Male Wistar Rats-After intravenous administration of [4-14C]sitosterol to male bile fistula rats, the amount of radioactivity recovered in bile was less than half of that recovered from the female rats. The amount of radioactivity in fraction 4, corresponding to hydroxylated CpI bile acids, was only about 10% of that in the corresponding experiments with female bile fistula rats. In spite of considerable efforts, no trace of 15-hydroxylated CZ1 bile acids would be found in bile from the male rats. Small amounts of a trihydroxy CpI bile acid lacking a 15-hydroxyl group could however be identified in fraction 4 of the HPLC chromatogram. The methyl ester-trimethylsilyl derivative gave a mass spectrum with peaks at m/z 596 (molecular ion), 581 (M-15), 506 (M-90), 491 (M-90-15), 416 (M-2 x 90), 401 (M-2 X 90-15), and 311 (M-3 x 90-15). The radioactive compounds in HPLC fractions 7-9 could not be identified with certainty. Also in these fractions there were no traces of 15-hydroxylated Cpl bile acids. As in the case above with female rats, a compound was found with a mass spectrum as could be expected for a CZ1 bile acid with two trimethylsilyloxo groups. Site of Conversion of Sitosterol into Czl Bile Acids-In view of the fact that endocrine tissue are able to convert both cholesterol and sitosterol into CZ1 steroid hormones (see below), it was considered important to exclude the possibility that the Czl bile acids are formed in the adrenals or ovaries. Female rats ovariectomized or adrenalectomized just prior to cannulation of the bile duct and administration of [4-14C] sitosterol gave, however, the same pattern of products in bile as above. It is thus evident that endocrine tissues are not involved. In a previous work it was shown that intraperitoneally administered sitosterol is recovered to more than 95% in the liver and blood and that only trace amounts are formed in other organs (29). From this finding and from the present finding that the same pattern and yield of products were obtained after intravenous and intraperitoneal administration, the liver seems to be the most likely organ for the conversion. More direct evidence was obtained in two experiments with isolated perfused livers from female Wistar rats. After 1 h of perfusion with [4-14C]sitosterol in the medium, a small amount of bile could be recovered from the bile duct (300 and 350 ~1, respectively). About 80% of the radioactivity obtained in the two bile samples had chromatographic properties as Co1 bile acids. It was calculated that the total conversion into CZl bile acid from the [4-4C]sitosterol in the perfusion medium during the time of the experiment (1 h) was about 15% of that occurring during the same time in a corresponding experiment with intraperitoneally administered [4-YJ]sitosterol to a bile fistula rat. Urinary Excretion of C,, Bile Acids-After intravenous administration of [4- Clsitosterol to an intact female rat in a metabolic cage, urine was collected and analyzed in the same way as the bile samples. The first 24-h portion of urine was found to contain radioactive products with chromatographic properties as CB1 bile acids. The amounts of these products corresponded to about 0.4% of the radioactivity administered,

7 Conversion of Plant Sterols to Bile Acids which is about 10% of the corresponding excretion in bile. Due to the relatively small amounts excreted in urine, no further investigations were made on the urinary metabolites and their state of conjugation. Pattern of Metabolites from 7a-[7p-3H]Hydroxysitosterol- In one experiment a female rat was injected with 7a-[7P-3H] hydroxysitosterol combined with [4-14C]sitostero1 to assess whether the metabolites from sitosterol were formed by an initial 7a-hydroxylation. The presence of a 7cu-hydroxyl group resulted in a markedly different distribution of radioactivity within the HPLC chromatogram. Fig. 6 demonstrates that the very polar products, particularly the metabolites at 4 and 9 ml, were absent. Less than 2% of the radioactivity eluted before cholic acid, whereas more co-chromatographed with cholic (16%) and chenodeoxy/deoxycholic acid (16%). A considerable amount of radioactivity from 7a- [7P-3H] hydroxysitosterol was associated with compounds less polar than these bile acids. It may be concluded that it is less likely that synthesis of the metabolites of sitosterol in fractions 4 and 9 involves initial participation of a 7a-hydroxylase. 7a-[7/3- HlHydroxysitosterol was more rapidly excreted in the bile than the simultaneously injected [4-4C]sitosterol. During the first 24-h period 57% of the given dose of the 7cuhydroxylated sitosterol had been eliminated in the bile but only 14% of the injected [ *C]sitosterol. The ratio of 3H to 14C in the ether extract of the 24-h bile sample was 4.7 as compared with 0.88 in the injection mixture. In comparison with sitosterol, 7a-hydroxysitosterol was thus more rapidly converted to acidic products. The clearance of 7oc-hydroxysitosterol from the serum was also more rapid and complete, and the 3H to 14C ratio was only 0.04 when the rat was killed after 48 h. Pattern of Metabolites from 26-[4-4C]Hydroxysitosterol- Mitochondrial 26-hydroxylation of cholesterol is the initial step in a minor pathway in bile acid biosynthesis (2). The mitochondrial w-hydroxylating system has a broad substrate specificity, and Aringer et al. (20) showed that 26- and 29- hydroxy derivatives of sitosterol are formed to a limited extent in rat liver mitochondria. A biosynthetic pathway to bile acids from sitosterol might therefore possibly involve an initial w- hydroxylation. Mitochondria were incubated with [4-14C]sitosterol under conditions optimal for side chain hydroxylation, as described under Materials and Methods. The isolated product, 26-hydroxysitosterol (or a mixture of 26- plus FIG. 7. Reversed-phase HPLC of ether extract of bile from a rat injected with [7-3H&ampesterol and [4-4C]sitosterol. The mobile phase was methanol/water (70:30, v/v; the water adjusted to ph 3.0). H activity (W) and 14C activity (C--Q) in l-ml fractions and refractive index detector response (-) are shown. 29-hydroxysitosterol) was given intravenously to a bile fistula rat. Within 24 h the majority (77%) of the injected dose had been excreted in the bile, predominantly present in the ether extract. Using methanol/water (70:30, v/v) as solvent, a small fraction of the radioactivity (<l%) appeared with the same retention times as those in the corresponding experiments with [ 4C]sitosterol. Reproducible peaks were thus eluted at 4 and 9 ml. A somewhat larger peak co-chromatographed with cholic acid. The major metabolite of 26-[4-4C]hydroxysitosterol eluted at 5 ml in a system with methanol/water (95:5, v/v) as mobile phase. The identity of this metabolite was never established, but the chromatographic properties were those expected for monocarboxylated sitosterol. It may be concluded that at least a minor part of the major metabolites of sitosterol may be formed by an initial w-hydroxylation. Patterns of Metabolites from [7- HJCampesterol-The plant sterol campesterol differs from the sitosterol by having a CZs structure with a methyl group attached to C2.,. If the side chain is shortened by a cleavage at CZ1 or CT., the metabolites from campesterol might possibly be identical to those of sitosterol. Injection of a mixture of the two plant sterols, [7-3Hz]campesterol and [4-4C]sitosterol, was performed. Fig. 7 shows the resulting chromatogram of the ether extract of a 24-h bile sample. There was a striking similarity between the 3H and the 14C activity tracings, with all the major 3H peaks coinciding with those of 14C. FIG. 6. Reversed-phase HPLC of ether extract of bile from a rat injected with 7a-[7&3H]hydroxysitosterol and [4-l%] sitosterol. The mobile phase was methanol/water (70:30, v/v; the water adjusted to ph 3.0) at a flow rate of 1 ml/min. H activity (M) and 14C activity (o---o) in l-ml fractions and refractive index detector response (-) are shown. Peak I, cholic acid, peak II, chenodeoxycholic acid. DISCUSSION Previous attempts to identify polar metabolites of the plant sterol sitosterol in rat bile have been unsuccessful. Several metabolic pathways and final products have been suggested (1, 12, 30), but no conclusive evidence to confirm these has been presented. Since the CZ4 ethyl group of sitosterol is likely to prevent an attack by the enzymes involved in the normal side chain oxidation of cholesterol, the preservation of an intact C& side chain has been considered most likely. The possibility of a more proximal side chain cleavage with loss of seven or more carbon atoms including the ethyl substituent and formation of a CZ1 or CT2 bile acid has never been proposed previously. The crucial finding in the present investigation was the formation of very polar radioactive metabolites from [4- *Cl sitosterol but not from 22,23-3H2- or 24,28- H2-labeled sitosterol. It should be pointed out that most previous work on the metabolism of sitosterol has been performed with use of 22,23- HZ-labeled sitosterol. Formation of C,, or CT2 bile acids is not possible to detect with such an isotope label. The possibility must be considered that the apparent loss

8 7974 Conversion of Plant Sterols to Bile Acids of H from the side chain of the specifically labeled sitosterol may be due to an isotope effect. Primary isotope effects with H can be expected to be of the magnitude of 5-20 (31). Since the retention of H in the polar metabolites was less than 5%, such an explanation seems highly unlikely. In any case, an isotope effect of this magnitude would not be expected to occur if the side chain cleavage only involves the final three or five carbon atoms. Under the experimental conditions employed, complete loss of H from [22,23- HJsitosterol is consistent with formation of a bile acid containing 23 or less carbon atoms. A bile acid with a carboxylic group at CZ3 and a H label at CZ2 could be expected to lose most of the H label in the alkaline hydrolysis step due to enolization. In the search for C2i-J& metabolites of sitosterol, it was considered important to evaluate the degree of dilution of the labeled sitosterol with unlabeled sitosterol. According to the experiments with [ H.Jsitosterol the intravenously administered sitosterol was diluted more than 5-fold with sitosterol from dietary and endogenous sources. It may be mentioned in this connection that rat liver microsomes contain about 1.2 pg of sitosterol/mg of protein, which corresponds to about 5% of the cholesterol content (32). From the present excretion data and from the degree of dilution of [ HJsitosterol it was calculated that the daily biliary excretion of polar metabolites of sitosterol in a bile fistula rat could be expected to reach at least several hundreds of pg. Among the polar metabolites formed from sitosterol, only two could be identified with certainty as CZl bile acids (major metabolites in HPLC fraction 4). These two metabolites corresponded to 20-30% of all acid metabolites excreted in bile. In total, about 75% of all the metabolites had chromatographic properties as the CZ1 bile acids. Further attempts to identify also the other polar bile acids are in progress. C&i bile acids have never hitherto been found in any biological system other than human meconium. Thus, Pyrek et al. (15) reported presence of small amounts of 3&hydroxypregn-5-en-21-oic acid in such material. The metabolic origin of this bile acid could not be established. It is known that microorganisms are able to catalyze 15- hydroxylation of bile acids (27), but hitherto 15-hydroxylated bile acids have never been detected in mammalian bile. Highly polar metabolites of corticosterone are known to be excreted in 15-hydroxylated form in bile from female rats (26), and female rat liver microsomes contain a 15@-hydroxylating system active on steroid sulfates (33). A specific cytochrome P- 450 species (P-45O15,,) catalyzing 15cu-hydroxylation of testosterone in female mice liver microsomes and male mice kidney microsomes has been described recently (34). The 15-hydroxyl group in the present C,i bile acids may have been introduced before or after degradation of the steroid side chain. In this connection a previous investigation by Aringer (28) is of great interest. He showed that a saturated analogue to sitosterol, 24-ethyl-5P-cholestan-3a-01, was 15-hydroxylated by rat liver microsomes. 24-Ethyl-5p-cholestane-3a-ol may be a metabolite of sitosterol, and a 15-hydroxyl group may thus be introduced already at this stage. The stereochemistry of the 15-hydroxyl group in the CZ1 bile acids formed from sitosterol could not be established in the present work. It is evident, however, that the hydroxylation was highly sex specific since no 15-hydroxylated CZ1 bile acids could be found in bile from male rats. In view of the finding that the overall conversion of sitosterol into CZ1 bile acids was much lower in male rats than in female rats, it is tempting to suggest that a 15-hydroxylation facilitates the degradation of the sitosterol side chain to yield a CZ1 bile acid. The absence of a hydroxyl group at C7 in the most polar Cpl bile acids and the presence of such a hydroxyl group in some of the metabolites are noteworthy. Rat liver contains two 7cu-hydroxylating systems: one highly specific and active on cholesterol, and one less specific, acting on conjugates of deoxycholic acid (2). The specific cholesterol 7a-hydroxylating system has a very low activity toward sitosterol (9). It is possible, however, that the less specific Icu-hydroxylating system active toward deoxycholic acid could have been responsive for the formation of at least one of the metabolites of sitosterol. In view of the fact that the introduction of a 7a-hydroxyl group is a rate-limiting step in the normal biosynthesis of CZ4 bile acids from cholesterol, it was considered to be of interest to study the metabolic fate of 7Lu-hydroxysitosterol. This compound was not converted into the polar metabolites obtained from [4- Clsitosterol. Part of the radioactive metabolites from 7cu-[7P-3H]hydroxysitosterol had chromatographic properties on HPLC similar to those of cholic acid. Due to lack of material no further attempts were made to identify this metabolite. No conclusive evidence was obtained that normal CZ4 bile acids are formed from sitosterol or 7a-hydroxylated sitosterol in rat liver. If the steroid side chain in sitosterol is degraded by a mechanism similar to that occurring in the conversion of cholesterol into bile acids, w-hydroxylation of the side chain would be an important step. 4-Y-Labeled 26-hydroxysitosterol (possibly contaminated with 29-hydroxysitosterol) was, however, converted into CZ, bile acids only to a very small degree. Small amounts of bile acids were formed in this experiment with chromatographic properties as could be expected for steroids with an intact CZ9 steroid side chain. In one experiment with 4-i4C-labeled sitosterol, trace amounts of dicarboxylic CZs bile acid could be identified. The presence of these metabolites was, however, not a reproducible finding, and most probably initial w-hydroxylation of the steroid side chain represents a minor pathway in the metabolism of sitosterol in rats. The situation may be different in other species, however, and it has been reported that monocarboxylated CZ9 bile acids are formed from sitosterol in monkeys (12). All of the above information suggests a major mechanism for side chain cleavage of sitosterol in female Wistar rats which is different from that involved in the synthesis of normal CZ4 bile acids. The possibility must be considered that the mechanism for formation of CZ1 bile acids may be similar to that involved in the formation of Csl steroids from cholesterol in endocrine tissues. In guinea pigs, administration of [3H]sitosterol led to excretion of labeled cortisol in urine (35). It has also been reported that sitosterol could serve as a substrate in reactions forming Csl and Cl9 steroids in mitochondria from rat adrenals, ovaries, testes, and placenta (36). From experiments with adrenalectomized and ovariectomized rats, the possibility was clearly excluded that endocrine tissues are involved in the formation of Cpl bile acids. Furthermore, & - oo& To& FIG. 8. Hypothetical mechanism for the formation of CzI bile acids from sitosterol. I