Oxysterol 7α-Hydroxylase Activity by Cholesterol 7α-Hydroxylase (CYP7A)*

A 7α-hydroxylation is necessary for conversion of both cholesterol and 27-hydroxycholesterol into bile acids. According to current theories, cholesterol 7α-hydroxylase (CYP7A) is responsible for the former and oxysterol 7α-hydroxylase (CYP7B) for the latter reaction. CYP7A is believed to have a very high substrate specificity whereas CYP7B is active toward oxysterols, dehydroepiandrosterone, and pregnenolone. In the present study, 7α-hydroxylation of various oxysterols in liver and kidney was investigated. Surprisingly, human cholesterol 7α-hydroxylase, CYP7A, expressed as a recombinant in Escherichia coli and COS cells, was active toward 20(S)-hydroxycholesterol, 25-hydroxycholesterol, and 27-hydroxycholesterol. This enzyme has previously been thought to be specific for cholesterol and cholestanol. A partially purified and reconstituted cholesterol 7α-hydroxylase enzyme fraction from pig liver showed 7α-hydroxylase activity toward the same oxysterols as metabolized by expressed recombinant human and rat CYP7A. The 7α-hydroxylase activity toward 20(S)-hydroxycholesterol, 25-hydroxycholesterol, and 27-hydroxycholesterol in rat liver was significantly increased by treatment with cholestyramine, an inducer of CYP7A. From the present results it may be concluded that CYP7A is able to function as an oxysterol 7α-hydroxylase, in addition to the previously known human oxysterol 7α-hydroxylase, CYP7B. These findings may have implications for oxysterol-mediated regulation of gene expression and for pathways of bile acid biosynthesis. A possible use of 20(S)-hydroxycholesterol as a marker substrate for CYP7A is proposed.

It has recently been reported that several oxysterols including 20(S)-, 22(R)-, 25-, and 27-hydroxycholesterol are ligands of the liver X receptor (LXR), which functions as a ligand-dependent transcription factor in a heterodimeric complex with the retinoid X receptor (5)(6)(7). LXR has been suggested to be an important regulator of cholesterol homeostasis and bile acid biosynthesis, at least in rodents (5)(6)(7)(8). Binding of ligand to the LXR induces transcription of CYP7A, which is the rate-limiting enzyme in the neutral pathway of bile acid biosynthesis. Oxysterols may also affect cholesterol homeostasis by other mechanisms. 20(S)-, 25-, and 27-hydroxycholesterol are considered to suppress the transcription of hydroxymethylglutaryl-coenzyme A (HMG-CoA) reductase, the rate-limiting enzyme in cholesterol biosynthesis (22). According to some reports 7␣hydroxylation of 25-and 27-hydroxycholesterol may alter their regulatory effects (5,23).
Chemical Synthesis of 7␣,20(S)-Dihydroxycholesterol-7␣,20(S)-Dihydroxycholesterol was prepared by allylic oxidation of 20(S)-hydroxycholesterol with the use of linoleic acid and soybean lipoxygenase (31). A mixture of the three allylic oxidation products was obtained (7-oxo-, 7␣-hydroxy-, and 7␤-hydroxy derivatives, with a predominance of the 7-oxo derivative). The steroids were extracted from the incubation mixture, dissolved in ethanol, and reduced with NaBH 4 to convert the 7-oxo derivative into the 7␣-hydroxyand 7␤-hydroxy derivatives. The 7␣hydroxy derivative was converted into the trimethylsilyl ether derivative and analyzed by combined gas chromatography-mass spectrometry. The mass spectrum showed the expected peaks at m/z 619 (M Ϫ 15), m/z 549 (M Ϫ 85, loss of the C 22 -C 27 part of the steroid side chain), as well as m/z 201 (steroid side chain including the OTMSi group at C 20 ).
Purification of Cholesterol 7␣-Hydroxylase from Pig Liver-Cytochrome P450 catalyzing the 7␣-hydroxylation of cholesterol was prepared from pig liver microsomes (12). This cytochrome P450 fraction showed several bands upon SDS-polyacrylamide gel electrophoresis and contained 1-2 nmol of cytochrome P450/mg of protein.
Purification of Oxysterol 7␣-Hydroxylase from Pig Liver-Cytochrome P450 catalyzing the 7␣-hydroxylation of 27-and 25-hydroxycholesterol but not cholesterol was prepared from pig liver microsomes as described previously (16). SDS-polyacrylamide gel electrophoresis of the purified cytochrome P450 fraction showed 1 major and 2-3 minor protein bands. The cytochrome P450 content varied between 2-3 nmol of cytochrome P450/mg of protein.
Purification of Oxysterol 7␣-Hydroxylase from Pig Kidney-Microsomes were prepared from kidneys of castrated although otherwise untreated male pigs. The microsomes were suspended in 50 mM Tris-HCl buffer, pH 7.4, containing 20% glycerol and 0.1 mM EDTA, diluted to a protein concentration of 6 mg/ml and solubilized with 1.8% (w/v) sodium cholate. The solubilized proteins were subjected to polyethylene glycol precipitation (26). The proteins precipitating between 7 and 14% polyethylene glycol were collected by centrifugation, homogenized, and dissolved in 100 mM phosphate buffer, pH 7.4, containing 20% glycerol, 0.1 mM EDTA, and 0.7% sodium cholate. All buffers used in the purification procedure contained 20% glycerol and 0.1 mM EDTA. Phosphate buffer was the potassium salt. Column eluates were monitored by measuring the absorbance at 416 nm. The polyethylene glycol precipitate was subjected to chromatography on octylamine-Sepharose (column, 3 ϫ 40 cm) equilibrated in 100 mM phosphate buffer, pH 7.4, containing 0.7% sodium cholate. Cytochrome P450 was eluted with 100 mM phosphate buffer, pH 7.4, containing 0.37% sodium cholate and 0.08% POEL. Cytochrome P450 fractions were pooled, diluted 1:4 with 20% glycerol and 0.1 mM EDTA and applied to a hydroxylapatite column (3 ϫ 15 cm) equilibrated in 10 mM phosphate, pH 7.4. The hydroxylapatite column was washed with 10 mM phosphate buffer, pH 7.4, containing 0.2% POEL and with 35 mM phosphate buffer, pH 7.4, containing 0.2% POEL. Cytochrome P450 was eluted with 300 mM phosphate buffer, pH 7.4, containing 0.2% POEL. The eluate was concentrated and dialyzed against 10 mM phosphate buffer, pH 8.0 containing 0.1% sodium cholate and 0.4% POEL. The sample was applied to a Q-Sepharose column (2 ϫ 30 cm) equilibrated in dialysis buffer. The column was washed with equilibrating buffer, and cytochrome P450 active for the 7␣-hydroxylation of 27-hydroxycholesterol was eluted with 10 mM phosphate buffer, pH 8.0, containing 500 mM sodium acetate, 0.1% sodium cholate, and 0.4% POEL. The Q-Sepharose chromatography was performed at room temperature, and all other purification steps at 5°C. The cytochrome P450 content in the purified enzyme fraction was about 3 nmol of cytochrome P450/mg of protein.
Expression of Human Cholesterol 7␣-Hydroxylase in E. Coli-The methods used for expression of human CYP7A in E. coli were essentially the same as those described by Karam and Chiang (28) except that the expression plasmid pJL/H7␣1.5 containing cDNA encoding for human CYP7A (28) was transformed into E. coli strain DH5␣ instead of TOPP3 cells.
Purification of Expressed Recombinant Human CYP7A in E. Coli-The purification procedure was essentially the same as that described (28) with the following modifications in the hydroxylapatite chromatography. The sample was dialyzed against 10 mM phosphate buffer, pH 7.4, containing 0.3% sodium cholate and applied to a hydroxylapatite column equilibrated in dialysis buffer. The column was washed with 10 mM phosphate, pH 7.4, containing 0.3% sodium cholate and then with 50 mM phosphate, pH 7.4, containing 0.3% sodium cholate. Protein was eluted with 360 mM phosphate, pH 7.4, containing 0.2% POEL. Eluted enzyme fractions were pooled and dialyzed against 100 mM phosphate buffer, pH 7.4, containing 20% glycerol, 0.1 mM EDTA, and 0.2% POEL prior to assay of catalytic activity.
Expression of Human CYP7A in COS Cells-COS-M6 cells were cultured in Dulbecco's Modified Eagle Medium supplemented with 10% fetal calf serum and antibiotics. The cells were transfected with the pSVL vector containing cDNA encoding for human CYP7A (29). In control experiments, COS cells were transfected with pSVL vector without the CYP7A cDNA insert. Transfection was carried out by electroporation in 0.4-cm cuvettes (Gene Pulser II, Bio-Rad), using a single pulse of 0.4 kV and 100 microfarads. In each experiment, 20 ϫ 10 6 cells were transfected with 20 g of DNA in a volume of 0.8 ml of phosphate-buffered saline containing calcium chloride and magnesium chloride (Dulbecco's, Life Technologies, Inc.). Following transfection, the cells were cultured for 48 h on 60-or 100-mm plates in medium containing 20(S)-hydroxycholesterol (5 M), 22(R)-hydroxycholesterol (5 M), 25-hydroxycholesterol (5 M), 27-hydroxycholesterol (0.5 or 5 M), or dehydroepiandrosterone (5 M). The substrates were dissolved in dimethyl sulfoxide. Catalytic activity toward cholesterol was assayed with endogenous substrate only. Following incubation, medium and cells were harvested separately. The cells were suspended in 1 ml of 50 mM Tris acetate buffer and homogenized (sonicated for 5 s). Cells and medium were extracted with trichloroethane/methanol (2:1) or ethyl acetate, and the organic phase was analyzed for 7␣-hydroxylated metabolites as described below.
Expression of Rat CYP7A in COS Cells-Expression of rat cholesterol 7␣-hydroxylase cDNA was performed by inserting the 2.2-kilobase pair XhoI-XhoI fragment into the pSVL expression vector as described (30). Transfection of COS-M6 cells was carried out as described above for expression of human CYP7A. A pSVL vector with the rat CYP7A cDNA inserted in reversed position was used as a negative control. The transfected cells were cultured for 48 h in medium containing 20(S)-, 25-, or 27-hydroxycholesterol (5 M). Extraction and analysis of 7␣-hydroxylated metabolites was performed in the same way as in experiments with human CYP7A.
Induction of Rat Liver Cholesterol 7␣-Hydroxylase by Cholestyramine Treatment-Female Harlan Sprague Dawley rats (M&B, Ry, Denmark) weighing about 250 g were fed a chow diet (R36, Lactamin AB, Vadstena, Sweden) supplemented with cholestyramine. The animals were housed together (3-4 in each cage) under controlled conditions (22.4°C, 50% air humidity, 12-h light cycle) and were acclimatized for 1 week before the experiment started. The animals had free access to food and water during the experiment. The rats were maintained either on a regular diet (n ϭ 6 rats) or a diet supplemented with 3% (w/w) cholestyramine (n ϭ 7 rats) for 6 days and were then killed using CO 2 . Liver microsomes were prepared from untreated and cholestyramine-treated rats according to standard methods. The microsomes were suspended in 50 mM Tris acetate buffer, pH 7.4, containing 20% glycerol and 0.1 mM EDTA and stored at Ϫ20°C until analysis of enzymatic activities.
Cultures of Human Embryonic Kidney Cells-Human embryonic kidney cells (293 cells) were seeded at 7.5 ϫ 10 5 cells per 60-mm tissue culture dish in Dulbecco's Modified Eagle Medium supplemented with 10% fetal calf serum and antibiotics. Enzymatic activity toward cholesterol, oxysterols, dehydroepiandrosterone, and pregnenolone in these cells was examined by addition of 5-8 M substrates dissolved in either dimethyl sulfoxide or 45% 2-hydroxypropyl-␤-cyclodextrin to the medium and incubation for 24 or 72 h. Following incubation with substrate, the medium was collected and extracted, and the organic phase was analyzed for 7␣-hydroxylated metabolites as described below.
Incubation Procedures-Incubations with purified enzyme fractions or microsomes were carried out at 37°C for 10, 20, 40, or 60 min. The The K m values for dehydroepiandrosterone and pregnenolone with the oxysterol 7␣-hydroxylase fraction were about 2 times higher than those for 25-and 27-hydroxycholesterol (16). In incubations with purified microsomal enzyme fractions, 2 units of NADPH-cytochrome P450 reductase were added (32). Triton X-100 at a concentration of 0.05% (w/v) and 5 mM dithiothreitol were added in incubations with cholesterol and in all incubations with CYP7A purified from E. coli (26). Triton X-100 was not added in incubations with rat liver microsomes. Incubations with intact microsomes (0.2-1 mg) were performed as for purified microsomal enzyme fractions except that the addition of NADPH-cytochrome P450 reductase was omitted. Incubations with oxysterols and cholesterol were terminated with 5 ml of trichloroethane/ methanol (2:1). The incubations with dehydroepiandrosterone and pregnenolone were terminated with 5 ml of ethyl acetate.
Incubation Conditions-It was observed that different incubation conditions strongly affected the 7␣-hydroxylase activities of enzymes from different sources. The most striking finding was that presence of Triton X-100 in the incubation mixture increased the 7␣-hydroxylase activity in incubations with E. coli-expressed CYP7A about 10-fold. Replacement of Triton X-100 with CHAPS gave about the same rate of hydroxylation as for incubations without detergent. 7␣-Hydroxylase activity toward cholesterol in rat liver microsomes was only slightly increased by the addition of Triton X-100. In contrast to the effects on cholesterol 7␣-hydroxylation, the presence of Triton X-100 decreased the 7␣-hydroxylation of dehydroepiandrosterone by rat liver microsomes about 10-fold.
Analysis of Incubations with Cholesterol-Formation of 7␣-hydroxycholesterol in microsomes and purified fractions was analyzed using [4-14 C]cholesterol and thin layer chromatography as described previously (16).
Cholesterol 7␣-hydroxylase activity in COS cells was assayed with the endogenous cholesterol as substrate. Following extraction of cell homogenate and medium, the samples were incubated with cholesterol oxidase and subjected to HPLC in the same way as incubations with 27-hydroxycholesterol except that the mobile phase was hexane/isopropanol (98:2). The retention time for the enzymatically formed 7␣-hy-droxy-4-cholesten-3-one in this system was 16 -17 min, which was identical to the retention time of the authentic reference compound.
Assays of 25-hydroxycholesterol 7␣-hydroxylase activity in COS cells were mostly performed with unlabeled 25-hydroxycholesterol (16). These incubations were analyzed in the same way as the incubations with 27-hydroxycholesterol. Testosterone was added as an internal recovery standard and after extraction, the samples were incubated with cholesterol oxidase and subjected to HPLC using hexane/isopropanol 92:8 as the mobile phase. The retention time for 7␣,25-dihydroxy-4-cholesten-3-one in this system was 10 -11 min.
Analysis by Combined Gas Chromatography-Mass Spectrometry-The 7␣-hydroxylated products from incubations with 20(S)-and 27-hydroxycholesterol were converted into trimethylsilyl ether and analyzed by combined gas chromatography-mass spectrometry. In the latter analysis a HP 5973 Quadropole instrument was used equipped with a HP 1989 1A-102 column. The initial temperature was 180°C followed by a rise of 20°C/ min up to 250°C and then 5°C/min up to 300°. The temperature was kept at 300°C for 11.5 min.
Analysis of Incubations with Dehydroepiandrosterone and Pregnenolone-Analysis of formation of 7␣-hydroxydehydroepiandrosterone and 7␣-hydroxypregnenolone was carried out as described previously (16).
Other Methods-NADPH-cytochrome P450 reductase was prepared from pig liver microsomes as described by Yasukochi and Masters (32). Protein concentrations in microsomal fractions and cell homogenates were determined by the method of Lowry (33). The concentration in purified protein fractions was determined by measuring the absorbance at 280 nm (concentration in mg/ml ϭ absorbance of protein at 280 nm). Cytochrome P450 concentration in purified fractions was estimated by measuring the absorbance at 416 nm as an indication of total heme content (34).
SDS-polyacrylamide gel electrophoresis was performed according to Laemmli (35) with modifications as described previously (16), and the gels were silverstained according to the methods described by Wray et al. (36).
Catalytic Properties of the Human Cholesterol 7␣-Hydroxylase Recombinant Expressed in E. coli-The results with partially purified pig liver enzyme fractions suggested that 20(S)and 22(R)-hydroxycholesterol were 7␣-hydroxylated in the liver by a 7␣-hydroxylase that was different from the so called oxysterol 7␣-hydroxylase. To examine this further, the catalytic properties of the expressed recombinant CYP7A was studied. Expression of human cholesterol 7␣-hydroxylase in E. coli resulted in a catalytically active protein with a very high 7␣hydroxylase activity toward cholesterol (Table I). Furthermore, this enzyme efficiently 7␣-hydroxylated 20(S)-hydroxycholesterol. Surprisingly, the expressed recombinant human CYP7A showed high 7␣-hydroxylase activity also toward 25-and 27hydroxycholesterol. 7␣-Hydroxylase activity toward 22(R)-hydroxycholesterol, dehydroepiandrosterone, or pregnenolone could not be detected (Table I). The identity of the enzymatically formed 7␣,27-dihydroxycholesterol was confirmed by combined gas chromatography-mass spectrometry. The mass spectrum of the product formed in incubations with 20(S)hydroxycholesterol was identical to that of synthetic 7␣,20(S)dihydroxycholesterol (see "Experimental Procedures").  (Tables II  and III). In contrast, no 7␣-hydroxylase activity was found toward 22(R)-hydroxycholesterol or dehydroepiandrosterone. 7␣-Hydroxycholesterol was found both within the cells and in the cell medium of transfected cells, whereas 7␣-hydroxylated metabolites formed from oxysterols were found almost exclusively in the cell medium (Tables II and III). This is in agreement with reported findings that side chain hydroxylated cholesterol derivatives are more easily secreted from some cells (37). In most of the experiments the cells were incubated with 5 M substrate. 27-Hydroxycholesterol was also added at a concentration of 0.5 M, which corresponds to the reported physiological concentration of this oxysterol (37,38). Also at this low concentration, 27-hydroxycholesterol was efficiently 7␣-hydroxylated by transfected cells. An increase of the concentration of added 27-hydroxycholesterol markedly decreased the 7␣-hydroxylase activity toward cholesterol (Table III).
Control experiments with COS cells transfected with pSVL vector without the CYP7A cDNA showed some endogenous activity in the COS cells toward 25-and 27-hydroxycholesterol. However, the endogenous activity in control cells toward these oxysterols was 3-12 times lower than the activity in cells expressing recombinant CYP7A (Tables II and III). No detectable amounts of 7␣-hydroxycholesterol or 7␣,20(S)-dihydroxycholesterol was found in COS cells transfected with pSVL vector without the CYP7A cDNA (Table II).
The 7␣-hydroxylase activity toward cholesterol was assayed with endogenous substrate only because previous investigations have indicated that addition of exogenous cholesterol to COS cells transfected with CYP7A cDNA does not further increase the 7␣-hydroxylase activity (15). Most of the experiments with oxysterols as substrates were performed with unlabeled substrate, which requires conversion of the products into 3-oxo-⌬ 4 derivatives with cholesterol oxidase prior to HPLC analysis. In a separate experiment, transfected COS cells were incubated with radiolabeled 25-hydroxycholesterol and assayed using thin-layer chromatography. Two polar products formed from 25-hydroxycholesterol were detected in this experiment; one with a mobility as expected for 7␣,25-dihydroxycholesterol and the other with a mobility as expected for 7␣,25-dihydroxy-4-cholesten-3-one. The cell medium of transfected cells contained about 10 times more 7␣,25-dihydroxy-4cholesten-3-one than 7␣,25-dihydroxycholesterol, although the samples had not been incubated with cholesterol oxidase. Furthermore, large amounts of 7␣-hydroxy-4-cholesten-3-one were detected in analyses by gas chromatography-mass spectrometry, where samples not treated with cholesterol oxidase were used. From these results it may be concluded that COS cells contain endogenous 3␤-hydroxy-⌬ 5 -C 27 -steroid dehydrogenase/ isomerase activity.
Effects of 7-Oxocholesterol on the 7␣-Hydroxylation of 27-Hydroxycholesterol by Pig Liver Enzyme Fractions-To further study the hepatic 27-hydroxycholesterol 7␣-hydroxylase activity, we incubated the pig liver cholesterol 7␣-hydroxylase and oxysterol 7␣-hydroxylase fractions (see Table I) with 7-oxocholesterol, an inhibitor of CYP7A (39). There was a marked difference in the effects of 7-oxocholesterol on the activity in the two fractions (Fig. 2). This compound, added in concentrations up to 10 M, had very little effect on the 27-hydroxycholesterol 7␣-hydroxylase activity in the oxysterol 7␣-hydroxylase fraction. In contrast, the 27-hydroxycholesterol 7␣-hydroxylase activity in the cholesterol 7␣-hydroxylase fraction was inhibited by about 80% at a concentration of 3 M 7-oxocholesterol (Fig. 2).
Effects of Cholestyramine Treatment on the 7␣-Hydroxylase Activities in Rat Liver Microsomes-To study the physiological importance of the CYP7A-catalyzed 7␣-hydroxylation of oxysterols, we examined the 7␣-hydroxylation of cholesterol, 20(S)hydroxycholesterol, 25-hydroxycholesterol, 27-hydroxycholesterol, and dehydroepiandrosterone in microsomes prepared from rats treated with cholestyramine, an inducer of CYP7A (40,41). Dehydroepiandrosterone was included in this experiment as a negative control and as a marker substrate for CYP7B. Treatment with cholestyramine increased cholesterol 7␣-hydroxylation about 3-fold. 7␣-Hydroxylation of 20(S)-hydroxycholesterol was increased to the same extent. Dehydroepiandrosterone 7␣-hydroxylation, however, was clearly not  25-hydroxycholesterol (B). Transfection, incubation, and analysis were performed as described under "Experimental Procedures." After incubation for 48 h, cells and medium from each plate were harvested separately. The samples were extracted and incubated with cholesterol oxidase. All samples were analyzed for 7␣-hydroxylated products by HPLC. Initially, the formation of 7␣,20(S)-dihydroxy-4-cholesten-3-one or 7␣,25-dihydroxy-4-cholesten-3-one was analyzed using hexane/isopropanol, 94:6 or 92:8, respectively, as the mobile phase. The eluates were collected and rechromatographed with hexane/isopropanol (98:2) to analyze the formation of 7␣-hydroxy-4-cholesten-3-one. The data given are the means of two experiments (results from the separate experiments are shown in brackets). stimulated by cholestyramine treatment (Fig. 3). This result is in accordance with the data reported by Brown and Boyd (41). If anything, treated rats seemed to show a slight decrease in dehydroepiandrosterone 7␣-hydroxylase activity (not statistically significant) in comparison to untreated rats. Cholestyramine treatment stimulated the 7␣-hydroxylation of 25-hydroxycholesterol and 27-hydroxycholesterol, although not to the same extent as the 7␣-hydroxylation of cholesterol and 20(S)-hydroxycholesterol. The increase of 7␣-hydroxylase activity in cholestyramine-treated rats as compared with untreated rats was about 70% for 25-hydroxycholesterol 7␣hydroxylation and 50% for 27-hydroxycholesterol 7␣-hydroxylation. These differences were statistically significant (p Ͻ 0.01) (Fig. 3). Catalytic Properties of Rat CYP7A Recombinant Expressed in COS Cells-To ascertain that rat CYP7A showed the same substrate specificity as human CYP7A, COS cells were transfected with the pSVL vector containing rat CYP7A cDNA (30), and the cells were incubated with 5 M 20(S)-, 25-, or 27hydroxycholesterol for 48 h. The 7␣-hydroxylase activity mediated by rat CYP7A in transfected COS cells toward 20(S)-, 25-, and 27-hydroxycholesterol was 3.1, 0.8 and 2.6 nmol/mg, respectively. The 7␣-hydroxylase activity toward cholesterol varied between 0.8 and 2.5 nmol/mg in different experiments. The activity toward cholesterol tended to decrease when transfected cells were incubated in the presence of 27-hydroxycholesterol in a similar manner as was seen in experiments with human CYP7A (see Table III).
7␣-Hydroxylation of Oxysterols in Cultured Human Embryonic Kidney Cells-The 7␣-hydroxylase activity in human embryonic kidney cells (293 cells) toward oxysterols, cholesterol, dehydroepiandrosterone, and pregnenolone was examined. These cells displayed high 7␣-hydroxylase activity toward 25hydroxycholesterol, 27-hydroxycholesterol, dehydroepiandrosterone, and pregnenolone. In contrast, 7␣-hydroxylase activity toward cholesterol, 20(S)-hydroxycholesterol, or 22(R)-hydroxycholesterol could not be detected (Table V). DISCUSSION The current study identifies three new substrates for the expressed recombinant human CYP7A, an enzyme previously thought to have a high substrate specificity. A role for CYP7A as an oxysterol 7␣-hydroxylase in vivo is supported by the finding that a cholesterol 7␣-hydroxylase fraction from pig liver is active toward all oxysterols metabolized by the expressed recombinant CYP7A. The inhibition experiments with 7-oxocholesterol also support the contention that CYP7A catalyzes the 7␣-hydroxylation of 27-hydroxycholesterol. Furthermore, the experiments with cholestyramine-treated rats support a physiological role for CYP7A in the 7␣-hydroxylation of 20(S)-, 25-, and 27-hydroxycholesterol. As expected, the data on 25and 27-hydroxycholesterol suggest that CYP7A is not the sole enzyme responsible for the 7␣-hydroxylation of these oxysterols, which is consistent with the concept that CYP7B is  The table shows the 7␣-hydroxylated metabolites formed by COS cells transfected with pSVL vector containing full-length CYP7A cDNA cultured in the presence of 0.5 or 5 M 27-hydroxycholesterol. Transfection, incubation, and analysis were performed as described under "Experimental Procedures." After incubation for 48 h, cells and medium from each plate were harvested separately. The samples were extracted and incubated with cholesterol oxidase. All samples were analyzed for 7␣-hydroxylated products by HPLC. Initially, the formation of 7␣,27dihydroxy-4-cholesten-3-one was analyzed using hexane/isopropanol (92:8) as the mobile phase. The eluate from this chromatography was collected and rechromatographed with hexane/isopropanol (98:2) to analyze the formation of 7␣-hydroxy-4-cholesten-3-one. The data given are the means of two or three experiments. ( important in these reactions (18 -21). From the experiments with dehydroepiandrosterone, it may be concluded that CYP7B is not induced by cholestyramine treatment. The finding that 22(R)-hydroxycholesterol is 7␣-hydroxylated in experiments with porcine but not human enzyme may reflect species differences with respect to the 7␣-hydroxylation of oxysterols. The present experiments do not exclude that a yet unknown cytochrome P450 may catalyze 7␣-hydroxylation of 22(R)-hydroxycholesterol in pig liver.
CYP7A is considered to be expressed only in the liver whereas CYP7B mRNA has been detected in extrahepatic tissues, including kidney (21). The present study shows that pig kidney cytochrome P450 and human kidney cells do not 7␣-hydroxylate cholesterol, 20(S)-hydroxycholesterol, or 22(R)-hydroxycholesterol. These results support the conclusion that CYP7B is not responsible for 7␣-hydroxylation of these steroids.
The results of this investigation may have implications for oxysterol-mediated regulation of gene expression as well as for pathways of bile acid biosynthesis. Furthermore the results provide new insight into the substrate specificity of 7␣-hydroxylating cytochrome P450 enzymes in liver and kidney.
LXR and 7␣-Hydroxylation of Oxysterols-The oxysterols 7␣hydroxylated by CYP7A (20(S)-, 25-, and 27-hydroxycholesterol) are reported to be ligands for the LXR receptor (5-7). Thus, some of the ligands of the liver X receptor, a nuclear receptor that induces expression of CYP7A in rodents, are metabolized by the very enzyme they induce. To our knowledge there is no information on what effects the 7␣-hydroxylated derivatives of 20(S)-or 27-hydroxycholesterol may have on the LXR. However, Janowski et al. (5) reported that 25-hydroxycholesterol is a several-fold more potent ligand for LXR␣ than 7␣,25-dihydroxycholesterol. It may be speculated that 7␣-hydroxylation of LXR ligands could reflect a way of controlling the level of expression of CYP7A by means of a feed-back mechanism.
7␣-Hydroxylating Enzymes in Bile Acid Biosynthesis-The surprising finding that 27-hydroxycholesterol is a substrate for cholesterol 7␣-hydroxylase have yet unclear implications for our understanding of the biosynthesis of bile acids. The data do not contradict the concept of two 7␣-hydroxylases acting in the neutral and acidic pathways of bile acid biosynthesis. However, the results show that CYP7A has the ability to participate in both pathways. The results of the cholestyramine induction experiments indicate that CYP7A may have a regulatory function also in the acidic pathway. Our findings in rats are consistent with the data obtained by Martin et al. (14) in hamsters. These authors reported that cholestyramine treatment of animals maintained on a cholesterol-free diet resulted in a similar increase in hepatic 7␣-hydroxylation of cholesterol and 27hydroxycholesterol as compared with untreated litter mates (14). The quantitative contributions by CYP7A and CYP7B for  27-hydroxycholesterol 7␣-hydroxylation in vivo remain to be investigated. Possible species differences cannot be excluded. The catalytic activity of recombinant CYP7A toward 27-hydroxycholesterol is high and the K m for 27-hydroxycholesterol 7␣-hydroxylation by CYP7A is the same as for cholesterol. A physiologically significant role for CYP7A in 27-hydroxycholesterol 7␣-hydroxylation is supported by the results obtained with COS cells. In these experiments, 27-hydroxycholesterol was efficiently 7␣-hydroxylated by living transfected cells at physiological concentrations of substrate (37,38) in an environment containing large amounts of endogenous cholesterol, competing for the enzyme active site (38). It is notable that presence of 27-hydroxycholesterol substantially decreased the 7␣-hydroxylase activity of these cells toward cholesterol.
The present results showing that CYP7A is able to 7␣-hydroxylate 25-and 27-hydroxycholesterol appear to be in contrast to some previous studies (14,15). Toll et al. (15) reported that recombinant human CYP7A, expressed in COS cells, did not show any significant 7␣-hydroxylase activity toward 25hydroxycholesterol. In those experiments, COS cell homogenate was incubated with 25-hydroxycholesterol for 60 min, 48 h after transfection. This means that the enzyme was exposed to 25-hydroxycholesterol for only 60 min whereas the incubation time for endogenous cholesterol was 49 h. However, in the present experiments the oxysterols were added to the cell medium immediately after transfection. Thus, the incubation time for oxysterols and cholesterol was the same. It is noteworthy that a large part of the formed 7␣-hydroxycholesterol was retained within the cells, whereas almost all of the more polar 7␣-hydroxylated derivatives formed from oxysterols were excreted into the medium. Martin et al. (14) concluded that 7␣hydroxylation of cholesterol and 27-hydroxycholesterol is performed by different enzymes in hamster liver. Although it seems clear from this and other studies (12)(13)(14)(15)(16)(17) that there must be more than one 7␣-hydroxylase involved in these reactions, the data obtained by Martin et al. (14) do not exclude overlapping substrate specificity of these enzymes.
It is possible that CYP7A and CYP7B have different roles in bile acid biosynthesis during different periods of life. Setchell and collaborators (42) described a newborn child with severe neonatal cholestasis with a mutation in the CYP7B gene. The authors concluded that CYP7B is critical for bile acid biosynthesis in infants. Although this patient had a normal CYP7A gene, cholesterol 7␣-hydroxylase activity was not detectable in the liver samples. Interestingly, control samples from the livers of normal infants (Ͻ1 year of age) also lacked detectable cholesterol 7␣-hydroxylase activity.
Substrate Specificity of CYP7A and CYP7B-Whereas the presence of an accessible hydrophilic group on the steroid side chain seems to be obligatory for catalysis by CYP7B, this enzyme apparently does not distinguish between C 27 -and C 19 / C 21 -steroids. CYP7A on the other hand, although effectively 7␣-hydroxylating several C 27 -steroids, shows very low, if any, 7␣-hydroxylase activity toward dehydroepiandrosterone or pregnenolone. In contrast to CYP7B, the presence of a hydroxyl group on the side chain reduces the catalytic activity of CYP7A. Information on the substrate specificity of CYP7A and CYP7B may be of interest for studies of the structure-function relationship of steroid-metabolizing enzymes.
From the present data it may be concluded that 27-hydroxycholesterol 7␣-hydroxylation is not specific for CYP7B. Studies on CYP7B should include assays of 7␣-hydroxylase activity toward dehydroepiandrosterone to distinguish between the contributions of CYP7A and CYP7B toward the reaction.
The finding that 20(S)-hydroxycholesterol is a good substrate for CYP7A opens the possibility that 20(S)-hydroxycholesterol could be used as a specific marker in studies of this enzyme. Because of the presence of endogenous cholesterol in microsomes and cell cultures, cholesterol 7␣-hydroxylation is sometimes difficult to measure. The activity toward 20(S)-hydroxycholesterol is high and can be measured in a simple and inexpensive way.