Purification of a 3beta-hydroxy-delta5-C27-steroid dehydrogenase from pig liver microsomes active in major and alternative pathways of bile acid biosynthesis.

A 3β-hydroxy-Δ5-C27-steroid dehydrogenase active in bile acid biosynthesis was purified from pig liver microsomes by solubilization with sodium cholate and by chromatography on DEAE-Sepharose, aminohexyl-Sepharose, and blue Sepharose. The last step in the purification procedure was preparative isoelectric focusing in a Rotofor cell. The final enzyme preparation showed only one protein band upon SDS-polyacrylamide gel electrophoresis. The isoelectric point was estimated to about 7.0 and the apparent Mr was 36,000. The purified enzyme catalyzed the conversion of 7α-hydroxycholesterol, 7α,25-dihydroxycholesterol, 7α,27-dihydroxycholesterol, and 3β,7α-dihydroxy-5-cholestenoic acid into the corresponding 3-oxo-Δ4 compounds. The enzyme was inactive with C19 and C21 steroids as substrates. The enzyme was also inactive with C27 steroids having the 7-hydroxy group in β- instead of α-position. The Km was found to be 0.30 and 0.32 μM with 7α-hydroxycholesterol and 7α,27-dihydroxycholesterol as substrates, respectively. NAD+ was the preferred cofactor. A monoclonal antibody raised against the 3β-hydroxy-Δ5-C27-steroid dehydrogenase was prepared. After coupling to Sepharose, the antibody was able to bind the dehydrogenase and to decrease the conversion of 7α-hydroxycholesterol into 7α-hydroxy-4-cholest-3-one by more than 90%. The N-terminal amino acid sequence was determined and found to be similar but not identical with those of known 3β-hydroxy-Δ5-steroid dehydrogenases active in steroid hormone biosynthesis. Thus, the purified enzyme active toward C27 steroids in bile acid biosynthesis appears to represent a novel type of 3β-hydroxy-Δ5-steroid dehydrogenase.

In the major pathways of bile acid biosynthesis, the initial 7␣-hydroxylation of cholesterol is followed by oxidation of the 3␤-hydroxy group and isomerization of the ⌬ 5 double bond (1). The oxidation and the isomerization are believed to be catalyzed by a single enzyme (2,3). Other pathways to primary bile acids, not involving 7␣-hydroxylation of cholesterol, have been described (4 -6). It was shown recently that in the pig and in man 7␣-hydroxylation of 27-oxygenated steroids is catalyzed by enzyme(s) different from the cholesterol 7␣-hydroxylase (7)(8)(9). It is not known whether the oxidation at C-3 of the side chain oxygenated steroids and 7␣-hydroxycholesterol is catalyzed by the same enzyme.
In 1981, Wikvall (3) purified a 3␤-hydroxy-⌬ 5 -C 27 -steroid dehydrogenase from rabbit liver microsomes. This enzyme was active toward 7␣-hydroxycholesterol. Side chain oxygenated steroids were not tested as substrates for the enzyme (3). In rabbit, the extent of 7␣-hydroxylation of 27-oxygenated intermediates is very low. Consequently, alternative pathways involving 27-oxygenated intermediates may not be important in rabbit. The aim of the present study was to purify and characterize pig liver microsomal 3␤-hydroxy-⌬ 5 -C 27 -steroid dehydrogenase(s) active in bile acid biosynthesis.
Enzyme Purification-Liver microsomes were prepared from castrated, otherwise untreated, 6-month-old male pigs. The microsomes were suspended in 100 mM Tris-HCl buffer, pH 7.5, containing 20% (v/v) glycerol and 0.1 mM EDTA, solubilized with 1.8% (w/v) sodium cholate (6 mg protein/ml) and precipitated with polyethylene glycol (11). The proteins precipitating between 14 and 22% (w/v) polyethylene glycol 6000 were collected by centrifugation and dissolved in 10 mM phosphate buffer, pH 7.6, containing 0.4% (w/v) of polyoxyethylene(10)lauryl ether (POELE). 1 The buffer also contained 20% (v/v) glycerol and 0.1 mM EDTA as did all other buffers used in the purification procedure. Phosphate buffers were used as the potassium salt. The dissolved precipitate was subjected to chromatography on DEAE-Sepharose (5 ϫ 30 cm) equilibrated in 10 mM phosphate buffer, pH 7.6, containing 0.4% PO-ELE. The column was washed with the equilibrating buffer and then eluted with a linear gradient of 40 -120 mM phosphate buffer, pH 7.6, containing 0.4% POELE. Fractions containing activity toward 7␣-hydroxycholesterol were pooled and concentrated using a Minitan unit (Millipore) with four 10 kDa membranes. The concentrated sample was dialyzed against 10 mM phosphate buffer, pH 7.6, containing 0.2% POELE and 0.5% sodium cholate and applied to a second DEAE-Sepharose column (3 ϫ 30 cm) equilibrated in the same buffer. The column was eluted with 35 mM phosphate buffer, pH 7.6, containing 0.2% POELE and 0.5% sodium cholate. The fractions containing enzyme activity were pooled and non-ionic detergent was removed by treatment with Amberlite XAD-2 (12) prior to dilution with five volumes of 100 mM phosphate buffer, pH 7.4, containing 0.5% sodium * This work was supported by the Swedish Medical Research Council (Project 03X-218) 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 U.S.C. Section 1734 solely to indicate this fact.
cholate. The fraction was applied to an aminohexyl-Sepharose column (3 ϫ 10 cm) equilibrated in the same buffer containing 0.7% sodium cholate. The column was washed with the equilibrating buffer and protein was eluted with 100 mM phosphate buffer, pH 7.4, containing 0.3% sodium cholate and 0.1% POELE. Fractions containing enzyme activity were pooled and concentrated using a Diaflo ultrafiltrating unit (Amicon) with a PM 10 filter and then dialyzed against 100 mM phosphate buffer, pH 7.4, containing 0.4% POELE. The dialysate was subjected to chromatography on a blue Sepharose column (2 ϫ 10 cm) equilibrated in the dialysis buffer. The column was washed with the equilibrating buffer and then with the same buffer containing 60 mM KCl. Enzyme activity was eluted with the equilibrating buffer containing 300 mM KCl. The fractions containing enzyme activity were collected, concentrated with a PM 10 filter and dialyzed against 10 mM phosphate buffer, pH 7.0, containing 0.5% POELE. The dialysate was subjected to preparative isoelectric focusing using a Rotofor unit (Bio-Rad), prefocused for one hour with preblended ampholines ranging from pH 5.0 to 8.0 in 20% glycerol and 0.5% POELE. After 4 h of focusing, the fractions were collected and assayed for enzyme activity. Purity of each active fraction was tested by SDS-polyacrylamide gel electrophoresis and enzymatically active fractions showing a single protein band were pooled and dialyzed against 10 mM phosphate buffer, pH 7.4. The dialyzed preparation was used for enzymatic studies. The purification procedure is summarized in Fig. 1.
Protein was determined by the biuret reaction enhanced with bicinchoninic acid (13). When appropriate, the protein was first precipitated with trichloroacetic acid and sodium deoxycholate (14).
Production and Purification of Monoclonal Antibodies-An 8-weekold female mouse of the Balb/c strain was used. Purified 3␤-hydroxy-⌬ 5 -C 27 -steroid dehydrogenase was emulsified with an equal volume of Freund's complete adjuvant. Fifty g of the enzyme was injected intraperitoneally. Five weeks later, another 100 g of the same antigen in 0.9% NaCl were injected intraperitoneally. Three days later the mouse was killed and the spleen was used as source of antibody-producing cells. The spleen cells were mixed 5:1 with Sp 2/0 myeloma cells and fused using 50% polyethylene glycol. The hybridoma cells were screened with enzyme-linked immunosorbent assay after hypoxanthine/aminopterin/thymidine selection. Cloning of positive cultures, purification of immunoglobulins, coupling to Sepharose, enzyme-linked immunosorbent assay, and immunoblotting were performed as described previously (15). An antibody designated mAb 4F5 was found to be specific for the 3␤-hydroxy-⌬ 5 -C 27 -steroid dehydrogenase.
To incubations with 7␣-hydroxycholesterol, testosterone was added as an internal recovery standard. The organic phase was evaporated under nitrogen, dissolved in mobile phase, and subjected to high performance liquid chromatography (HPLC) on a silica column (LiChrosorb, 150 mm) with hexane/isopropanol (96:4) as mobile phase. Compounds with 3-oxo-⌬ 4 structure were monitored at 240 nm. Peak areas of product and internal standard were measured and the conversions were calculated from a standard curve. The retention times were 6.2 min for 7␣-hydroxy-4-cholesten-3-one and 12.2 min for testosterone. The standard curve was prepared using known amounts of 7␣-hydroxy-4-cholesten-3-one and was linear in the part used for calculation of conversions. The testosterone used as internal standard was purified on HPLC in the same system as was used for the incubations.
To incubations with 7␣,27-dihydroxycholesterol,11␤-hydroxyprogesterone was added as an internal recovery standard. The organic phase was collected, evaporated under nitrogen and dissolved in mobile phase. The samples were subjected to HPLC on a silica column (LiChrosorb, 150 mm) with hexane/isopropanol (92:8) as mobile phase. Compounds with 3-oxo-⌬ 4 structure were monitored at 240 nm. Peak areas of product and internal standard were measured and the conversions were calculated from a standard curve. The retention times were 9.8 min for 7␣,27-dihydroxy-4-cholesten-3-one and 13.2 min for 11␤-hydroxyprogesterone. The standard curve was prepared using known amounts of 7␣,27-dihydroxy-4-cholesten-3-one and was linear in the part used for calculation of conversions. The 11␤-hydroxyprogesterone used as internal standard was purified on HPLC in the same system as was used for the incubations.
Incubations with 27-hydroxycholesterol were analyzed in a similar system as the two described above for 7␣-hydroxycholesterol and 7␣,27dihydroxycholesterol. The mobile phase was hexane/isopropanol (98:2)

FIG. 4. The effect of protein (A), time (B), cofactor concentration (C), pH (D), and substrate concentration (E), on the conversion of 7␣-hydroxycholesterol to 7␣-hydroxy-4-cholesten-3-one by purified 3␤-hydroxy-⌬ 5 -C 27 -steroid dehydrogenase.
Assay conditions, other than the varied factor, were as described under "Experimental Procedures," except for the substrate curve where 0.025 g of protein was incubated for 5 min. and the retention time was 11.6 for 27-hydroxy-4-cholesten-3-one. Androstenedione was used as internal recovery standard and had a retention time of 15.1 min.
Thirty pmol of the dehydrogenase were incubated with 10, 25, and 50 g of the antibody-coupled Sepharose in 300 l of 100 mM PO 4 , pH 7.4, 20% glycerol, 0.1 mM EDTA, and 0.1% CHAPS. An irrelevant Sepharose-bound monoclonal antibody raised against mitochondrial CYP27 from pig liver and Sepharose without antibody were used as controls.
After incubation in a rotating mixer for 1 h at room temperature followed centrifugation (4000 ϫ g, 10 min, 4°C). The Sepharose was washed twice with 200 l of 100 mM PO 4 , pH 7.4, 20% glycerol, 0.1 mM EDTA, and 0.1% CHAPS and pooled. The supernatants were used for assay of remaining enzyme activity. The reaction mixtures were incubated with 20 g of 7␣-hydroxy-cholesterol for 10 min at 37°C. The incubations were terminated, extracted and analyzed as described under "Incubation Procedure." Amino-terminal Amino Acid Sequence Analysis-Purified enzyme, about 100 pmol, was subjected to SDS-polyacrylamide gel electrophoresis and blotted onto a polyvinylidene difluoride-membrane (Bio-Rad) as described by Towbin et al. (19), Matsudaira (20), and Moos et al. (21). The membrane was stained with 0.025% (w/v) of Coomassie Brilliant Blue R-250 in 40% (v/v) methanol and destained in 50% methanol. The protein bands were cut out and N-terminal amino acid sequence was determined as described previously (22) in a Applied Biosystems 473A sequencer.
Calculation of K m Values-The calculation of the K m values with 7␣-hydroxycholesterol and 7␣,27-dihydroxycholesterol were carried out with linear regression analysis.

Purification of 3␤-Hydroxy-⌬ 5 -C 27 -steroid Dehydrogenase-
The purification of 3␤-hydroxy-⌬ 5 -C 27 -steroid dehydrogenase from pig liver microsomes is summarized in Table I. After two DEAE-Sepharose chromatography steps, all heme-containing proteins were completely removed. The preparation at this stage was slightly yellow, which was still the case after aminohexyl-Sepharose chromatography. Most of the activity was bound to the blue Sepharose, as is common for many NAD ϩand NADP ϩ -dependent dehydrogenases.
The large loss of activity in some of the chromatographic steps can only be partly explained by activity in adjacent fractions. Inactivation of enzyme molecules during the purification may be substantial. In support of this it was found that when purified enzyme was preincubated with buffer without glycerol for 10 min at 37°C, only 10% of the activity was left compared with enzyme preincubated with buffer containing 20% glycerol. In, and after, the last step in the purification, the preparative isoelectric focusing (Fig. 2), the loss of activity was considerable. The focusing process and/or components in the buffer, e.g. nonionic detergent in very high concentration that can not be dialyzed, could be possible explanations for this loss.
The final, colorless, preparation was purified at least 1,500fold as compared with the original microsomal fraction, although the above mentioned activity loss indicates an even higher degree of purification. The preparation showed only one protein band upon SDS-polyacrylamide gel electrophoresis, and the apparent M r was 36,000 (Fig. 3). The fact that the preparation is colorless argues against a possible flavin component. This is in line with findings of Wikvall (3). The apparent M r of the 3␤-hydroxy-⌬ 5 -C 27 -steroid dehydrogenase purified from pig liver microsomes in the present communication was much lower than that of the corresponding enzyme (M r ϭ 46,000) purified from rabbit liver microsomes (3). Immunological and structural comparisons are not possible to carry out since attempts to characterize the rabbit liver enzyme further FIG. 5. SDS-PAGE and immunoblotting of the pig liver microsomal 3␤-hydroxy-⌬ 5 -C 27 -steroid-dehydrogenase. Samples containing the indicated amounts of protein were subjected to SDS-PAGE, electrophoretic transfer and immunoblotting with mAb 4F5 as described under "Experimental Procedures." Lane 1, purified 3␤-hydroxy-⌬ 5 -C 27 -steroid dehydrogenase (0.5 g); lane 2, pig liver microsomes (100 g).
FIG. 6. Effect of mAb 4F5 on the conversion of 7␣-hydroxycholesterol into 7␣-hydroxy-4-cholesten-3-one by the pig liver microsomal 3␤-hydroxy-⌬ 5 -C 27 -steroid dehydrogenase. Thirty pmol of the dehydrogenase were incubated with the indicated amounts of Sepharose-bound 26C5 (irrelevant antibody) and 4F5 and assayed for enzymatic activity as described under "Experimental Procedures." Inactivated CNBr-Sepharose without antibody was used as control. Results are given as percentage of control values for enzymatic activity in incubations without antibody. The 100% control value was 917 nmol⅐min Ϫ1 ⅐mg Ϫ1 of dehydrogenase. have not been successful. 2 Assay Conditions and Substrate Specificity of Purified 3␤-Hydroxy-⌬ 5 -C 27 -steroid Dehydrogenase-Assay conditions were determined with 7␣-hydroxycholesterol as substrate (Fig.  4A-E). Only one product was detected, 7␣-hydroxy-4-cholesten-3-one. NAD ϩ was the preferred cofactor and the enzyme was saturated with about 0.2 mM NAD ϩ . NADP ϩ was only slightly active in the oxidation. The rate of oxidation was almost linear up to 10 min and 0.5 g of enzyme. The pH optimum was around 7.5. The purified enzyme catalyzed the conversion, in order of efficiency, of 7␣-hydroxycholesterol, 7␣,27-dihydroxycholesterol, 3␤,7␣-dihydroxy-5-cholestenoic acid, and 7␣,25-dihydroxycholesterol into the corresponding 3-oxo-⌬ 4 compounds (Table II). The apparent K m for 7␣-hydroxycholesterol as substrate was 0.30 M and for 7␣,27-dihydroxycholesterol as substrate 0.32 M. Calculated V max values in these experiments were low compared to the activity obtained directly after preparative isoelectric focusing. The K m for 7␣-hydroxycholesterol was considerably lower and the V max considerably higher than those reported previously for the rabbit liver enzyme (3).
Immunological Studies-A monoclonal antibody raised against the pig liver microsomal 3␤-hydroxy-⌬ 5 -steroid dehydrogenase was prepared by immunization of a mouse with the purified enzyme. The antibody designated mAb 4F5 recognized the purified enzyme as well as a protein of the same apparent M r in pig liver microsomes (Fig. 5). The antibody was coupled to Sepharose and incubated with 3␤-hydroxy-⌬ 5 -steroid dehydrogenase. After incubation, the antibody-coupled Sepharose was removed and the supernatant was assayed for dehydrogenase activity. The conversion of 7␣-hydroxycholesterol into 7␣-hydroxy-4-cholesten-3-one was studied. Sepharose without antibody and Sepharose coupled to an irrelevant antibody directed against mitochondrial CYP27 from pig liver were used as controls. Fig. 6 shows that mAb 4F5, but not the irrelevant antibody, was able to bind the dehydrogenase and immunoprecipitate the enzymatic activity by more than 90%. These results confirm that the 36,000 M r protein is the 3␤-hydroxy-⌬ 5 -C 27steroid dehydrogenase.
Structural Characterization-The N-terminal amino acid se-quence was determined to be Ala-Gln-X-Leu-Val-Tyr-Leu-Val-Gln-Gly-Ala-Gly-Ile-Phe-Leu-Gly-Glu-Arg. This sequence shows similarities with known sequences of other 3␤-hydroxy-⌬ 5 -steroid dehydrogenases (23) (Table III). Oxidation at the C-3 position in bile acid biosynthesis is thus catalyzed by an enzyme related to but different from those catalyzing the same reaction in steroid hormone biosynthesis. This conclusion is supported by the substrate specificity of the enzyme.
Biological Significance-In alternative pathways in bile acid biosynthesis bypassing cholesterol 7␣-hydroxylase, oxidation of the 3␤-hydroxy group must be performed on other intermediates than 7␣-hydroxycholesterol. Two such intermediates proposed in previous studies are 7␣,27-dihydroxycholesterol and 3␤,7␣-dihydroxy-5-cholestenoic acid (7,8). Indeed, the purified 3␤-hydroxy-⌬ 5 -C 27 -steroid dehydrogenase was found to catalyze the oxidation of these two steroids as well as 7␣-hydroxycholesterol, the intermediate in the major pathways to primary bile acids. This finding shows that one single enzyme can perform the oxidation/isomerization reaction on all potential intermediates in bile acid biosynthesis. The fact that only bile acids with 3␤-hydroxy-⌬ 5 configuration are found in the rare inborn error of metabolism, 3␤-hydroxy-⌬ 5 -C 27 -steroid dehydrogenase deficiency, supports the contention that only one enzyme is involved in the oxidation of the 3␤-hydroxy group (24,25).