INTRODUCTION

The formation of bile acids from cholesterol¹ serves two important physiological functions. First, bile acid biosynthesis provides a pathway of cholesterol catabolism by which excess sterol can be disposed. Second, the end products of this pathway serve as natural detergents in the gut that facilitate the solubilization and absorption of dietary sterols, triglycerides, and fat-soluble vitamins. Individual bile acids differ in the positions of the hydroxyl groups on the ring structure of cholesterol and in the length and degree of oxidation of the side chain. To date, the chemical structures of several hundred bile acids in dozens of different species have been elucidated (1). Included in this roster are bile acids that range in size from 19 to 27 carbons and that contain hydroxyl groups with either  or  stereochemistries at one or more of these carbon atoms.

Although the chemical structures of bile acids from different species are diverse, almost all vertebrates contain 7-hydroxylated bile acids (1, 2), which suggests that compounds with this substituent play essential roles in cholesterol and lipid metabolism. Two pathways that lead to the synthesis of 7-hydroxylated bile acids have been described. One is the neutral or microsomal pathway (3, 4) and involves an initial 7-hydroxylation of cholesterol by a microsomal cytochrome P450 enzyme termed cholesterol 7-hydroxylase (EC 1.14.13.17). This enzyme has been extensively studied at the biochemical, genetic, and molecular levels (reviewed in Refs. 5 and 6). The output from the microsomal pathway is tightly regulated by controlling the transcription of the cholesterol 7-hydroxylase gene in the liver (7-10).

The second pathway is the mitochondrial or acidic pathway and utilizes a slightly different sequence of initial steps in the formation of 7-hydroxylated bile acids (3, 4). Cholesterol is first converted in mitochondria to an oxysterol by hydroxylation of the side chain (11). This intermediate is then a substrate for a microsomal oxysterol 7-hydroxylase that is distinct from cholesterol 7-hydroxylase (12-16). Although oxysterol 7-hydroxylase activity has been reported in several tissues of the pig, rat, hamster, and mouse, a cDNA encoding this enzyme has not yet been reported, nor has the regulation of the enzyme, and hence the output of the mitochondrial pathway, been extensively studied.

A genetic demonstration of the physiological importance of 7-hydroxylated bile acids came from mice in which the contribution of the microsomal pathway was eliminated by targeted disruption of the cholesterol 7-hydroxylase gene (17). Approximately 90% of animals homozygous for a null allele die in the first 3 weeks of life from fat and vitamin malabsorption. Death can be prevented by the addition of a 7-hydroxylated bile acid and vitamins to the diet, whereas survival in the absence of these supplements is correlated with the induction of the oxysterol 7-hydroxylase/mitochondrial pathway (16, 17). The activation of this pathway in mice normally occurs at about 3 weeks of age in response to yet undefined regulatory signals (16).

A candidate cDNA for the oxysterol 7-hydroxylase was recently isolated by differential hybridization screening of a hippocampal cDNA library (18). This cDNA, termed hct-1 for hippocampal transcript 1, encoded a cytochrome P450 enzyme that shared 39% sequence identity with cholesterol 7-hydroxylase. The high degree of sequence conservation placed hct-1 in the cytochrome P450 gene family 7 with the official designation of Cyp7b1 (Ref. 19). Besides the hippocampus, the hct-1 mRNA was detected in several nonneuronal tissues, including the liver (18). Subsequent expression studies showed that the hct-1 enzyme is a steroid 7-hydroxylase that acts on dehydroepiandrosterone and pregnenolone (20). These results, together with the expression of hct-1 in the brain, suggested that the enzyme was involved in the synthesis of neurosteroids (20).

In the current study, we provide evidence that hct-1 is also an oxysterol 7-hydroxylase that participates in hepatic bile acid biosynthesis. The biochemical, pharmacological, and ontological properties of the hct-1 cDNA-encoded enzyme match those of an oxysterol 7-hydroxylase activity present in liver microsomes. It is further shown that the oxysterol 7-hydroxylase and the cholesterol 7-hydroxylase enzymes, and hence their respective pathways, are differentially regulated in response to diet.

EXPERIMENTAL PROCEDURES

³²P-Labeled nucleotides and [³H]cholest-5-ene-3,25-diol (25-[³H]hydroxycholesterol) were purchased from NEN Life Science Products. Thin layer chromatography plates (LK5DF silica gel 150 Å) were from Whatman (Clifton, NJ). Reagents used in cDNA cloning and sequencing were from New England Biolabs (Beverly, MA), Boehringer Mannheim, or Life Technologies, Inc. Common laboratory chemicals and stigmastanol were from Sigma. Nafimidone was obtained from D. C. Swinney at Roche Biosciences (Palo Alto, CA).

C57BL/6J mice (male, 3 months old) were obtained from the Jackson Laboratory (Bar Harbor, ME) and were housed individually in a humidity- and temperature-controlled room (22 °C) with alternating 12-h light/12-h dark cycles. Mice were fed a cereal-based rodent diet (7001, Harlan Teklad, Madison, WI), which contained 4% (w/w) fat, 24% (w/w) protein, and 5% (w/w) fiber. Where indicated, this diet was supplemented with 2% (w/w) cholesterol (ICN, Irvine, CA), 2% (w/w) colestid (ICN), 0.5% (w/w) cholic acid (ICN), or a combination of 2% cholesterol and 0.5% cholic acid. Mice (n = 4 or 5 per group) were fed individual diets for 10 days. Body weights for each animal were recorded before and after the feeding period. On the last day of the experiment, mice were fasted for 6 h, anesthetized with nembutal (3 mg/animal, delivered intraperitoneally), and bled from the inferior vena cava. Total plasma cholesterol was enzymatically determined (Cholesterol/HP kit, Boehringer Mannheim). Plasma cholesterol levels (mean ± S.E.) after 10 days on normal diet or diet mixed with 2% cholesterol, 2% colestid, 0.5% cholic acid, or 2% cholesterol plus 0.5% cholic acid were 75.8 ± 4.9, 69.8 ± 3.8, 65.7 ± 2.1, 58.0 ± 7.6, and 131.2 ± 21.5 mg dl¹, respectively. None of the values measured in animals fed supplemented diets were significantly different from those of normal diet-fed mice (i.e. p > 0.01, Student's two-tailed test assuming equal variance).

The wet weights of the livers were determined, and the tissue from each mouse was divided into three aliquots. Two aliquots of 500 mg each were frozen in liquid nitrogen and stored at 70 °C until used for the preparation of mRNA and microsomal membranes, respectively. A third aliquot (~200 mg) was immediately saponified in 5 ml of ethanolic KOH (2 h, 65 °C). Free sterols were extracted with petroleum ether, an internal standard was added (1 mg stigmastanol), and the organic phase was analyzed by gas chromatography to determine total hepatic cholesterol content (21). Hepatic cholesterol levels (mean ± S.E.) on the normal, cholesterol, colestipol, cholic acid, and cholesterol plus cholic acid diets were 2.3 ± 0.05, 3.24 ± 0.19 (p  0.001), 2.44 ± 0.09, 3.41 ± 0.18 (p  0.0003), and 23.5 ± 3.0 (p  0.0001) mg/g of tissue, respectively. Values measured in cholesterol-, cholic acid-, and cholic acid plus cholesterol-fed mice were significantly higher (i.e. p  0.01, Student's two-tailed test assuming equal variance) than those determined in control mice.

Assays for oxysterol 7-hydroxylase activity were carried out as described previously (16). The hepatic oxysterol 7-hydroxylase utilizes both 25-hydroxycholesterol and 27-hydroxycholesterol as substrate (14). In the present study, commercially available 25-[³H]hydroxycholesterol was used as a substrate.

A plasmid vector capable of expressing the murine Cyp7bl protein was constructed via standard methods of genetic engineering. Briefly, an hct-1 cDNA encompassing nucleotides 20 to 1880 (Ref. 18), was ligated into the mammalian expression vector pCMV6 and transformed into E. coli DH5 cells. The desired recombinants were characterized by restriction enzyme mapping and DNA sequencing. The expression plasmid was purified and transfected into human embryonic kidney 293 cells using an MBS transfection kit (Stratagene Corp., La Jolla, CA). Twenty-four h after transfection, the indicated concentrations of 25-[³H]hydroxycholesterol were added to the medium of transfected and mock-transfected cells. At various times after the addition, lipid metabolites were extracted from the medium using Folch reagent (chloroform/methanol, 2:1, v/v) and analyzed by thin layer chromatography (solvent system toluene/ethyl acetate, 2:3, v/v) and autoradiography.

Microsomal membranes from mouse liver were prepared after Dounce homogenization by sequential centrifugation as described below. Microsomal membranes from 293 cells were prepared by harvesting cells 24 h after transfection, washing once with ice cold phosphate-buffered saline, and swelling for 10 min in a hypotonic buffer containing 10 mM Hepes, pH 7.6, 1.5 mM MgCl₂, 10 mM KCl, 1 mM EDTA, 1 mM EGTA, 9000 trypsin inhibitory units/ml aprotinin, 10 µg/ml leupeptin, 5 µg/ml pepstatin, and 5 µM phenylmethylsulfonyl fluoride. Cells were disrupted by passing the suspension through a 22.5-gauge needle 15 times. The nuclei and cell debris were removed by a 5-min centrifugation at 1000 × g at 4 °C. The supernatant was transferred to an ultracentrifuge tube, and microsomal membranes were pelleted by centrifugation at 4 °C for 30 min at 130,000 × g. Microsomal membranes from mouse liver and transfected 293 cells were resuspended in Buffer A (50 mM Tris-HCl, pH 7.4, 20% (v/v) glycerol, 1 mM EDTA, and protease inhibitors as described above) at a final concentration of 20 mg of membrane protein/ml and assayed for oxysterol 7-hydroxylase activity in the presence of the indicated concentrations of nafimidone.

For each incubation, the required amount of an ethanolic solution of inhibitor was dried down in a test tube and preincubated with 100 µg of microsomal protein, 2 mM dithiothreitol, 0.03% (v/v) Triton X-100 in buffer A for 10 min at 37 °C. 25-[³H]Hydroxycholesterol was added to a final concentration of 0.2 µM, and the reaction (total volume 0.5 ml) was started by the addition of NADPH to a final concentration of 1.5 mM. Mouse liver microsomes were incubated for 10 min, and 293 cell membranes were incubated for 30 min, both at 37 °C. Thereafter, 6 ml of methylene chloride were added to each incubation to terminate the reaction and to extract 25-hydroxycholesterol metabolites. The organic phase was dried down under nitrogen, and lipids were redissolved in 40 µl of Folch reagent and analyzed by thin layer chromatography (solvent system toluene/ethyl acetate, 2:3, v/v) and autoradiography. Radiolabeled products were quantitated using a System 200 Imaging Scanner (Bioscan, Inc., Washington, D. C.).

A 15-amino acid peptide (QDLLKRYYRHDDSEIG) derived from residues 266-281 of the cDNA-deduced sequence of the murine Cyp7b1 protein (18), was synthesized by Biosynthesis Inc. (Lewisville, TX), as a multiple-antigen peptide. The multiple-antigen peptide was used to immunize four New Zealand White rabbits (male, 3 months old). For the initial injection, 250 µg of multiple-antigen peptide were emulsified with complete Freund's adjuvant and administered intramuscularly. Over the next 3 months, four boosts were given using the same dose of multiple-antigen peptide but with incomplete Freund's adjuvant. Antibody titers in serum samples were tested in immunoblotting experiments using as antigen extracts from 293 cells transfected with a murine Cyp7bl cDNA expression vector.

Samples were first converted to trimethylsilyl ethers by treatment with 50 µl of pyridine/hexamethyldisilazane/trimethylchlorosilane (3:2:1, v/v/v) at 60 °C for 30 min. After evaporation of the solvent under argon, the residue was dissolved in hexane and transferred to an autosampler vial. Gas chromatography-mass spectrometry was performed on a Hewlett Packard 5890 Series II Plus gas chromatograph equipped with an HP-5MS capillary column (30 m × 0.25 mm, 0.25-µm phase thickness), connected to an HP 5972 mass selective detector and an HP 7673A automatic sample injector. The oven temperature program was 180 °C for 1 min, 20 °C/min to 250°, followed by 5 °C/min to 300 °C, where the temperature was kept for 8.0 min. Helium was the carrier gas. The gas chromatograph was operated in constant flow mode at a flow rate of 0.8 ml of helium/min. The injector was operated in splitless mode at 270 °C. The detector transfer line was maintained at 280 °C.

RESULTS

To determine if a Cyp7b1 cDNA encoded an oxysterol 7-hydroxylase activity, a murine cDNA was inserted into an expression vector, and the resulting plasmid was transfected into cultured human embryonic kidney 293 cells. A substrate for the murine oxysterol 7-hydroxylase, 25-[³H]hydroxycholesterol (16), was then added to the cell media at a final concentration of 0.12 µM. The conversion of this compound into [³H]cholest-5-ene-3,7,25-triol (7-hydroxylated 25-hydroxycholesterol) was monitored by thin layer chromatography. In initial experiments, the presence of an endogenous human oxysterol 7-hydroxylase activity in 293 cells prevented an unambiguous demonstration of plasmid-encoded enzyme activity (data not shown). To overcome this problem, an inhibitor (nafimidone) of cytochrome P450 enzymes² was used to selectively block the human oxysterol 7-hydroxylase in the 293 cells. The concentration of nafimidone required to inhibit 50% of endogenous oxysterol 7-hydroxylase enzyme activity in intact 293 cells was approximately 3 µM, whereas concentrations above 500 µM were required to inhibit activity expressed from a transfected murine Cyp7b1 cDNA (data not shown).

In the presence of 50 µM nafimidone, mock-transfected cells did not metabolize 25-[³H]hydroxycholesterol to more polar compounds during the 8-h course of the experiment (Fig. 1, lanes 1-4). Cells transfected with the murine Cyp7b1 expression vector began producing a product identified as 7-hydroxylated 25-hydroxycholesterol within 30 min of incubation with substrate (lane 5). With longer incubation times, the amount of this product increased, and another metabolite identified as [³H]7,25-dihydroxy-cholest-4-ene-3-one (Fig. 1, lanes 6-8) appeared; this secondary metabolite arises due to an endogenous 3-hydroxysteroid dehydrogenase enzyme activity present in the 293 cells. The chemical structures of these products were previously determined by thin layer chromatography and gas chromatography-electron impact mass spectrometry using authentic standards (16). We concluded from these experiments that the murine Cyp7b1 cDNA encodes an oxysterol 7-hydroxylase active against 25-hydroxycholesterol.

Cyp7b1 cDNA encodes an oxysterol 7-hydroxylase activity.

When the transfection experiments with the murine Cyp7b1 cDNA were carried out in the absence of nafimidone, additional cDNA-specific, polar products arising from 25-[³H]hydroxycholesterol were detected (Fig. 2). Within 2 h, a major product with a mobility slower than 7-hydroxylated 25-hydroxycholesterol was detected (lane 2). With longer incubation times, the amount of this metabolite rose and several additional minor products appeared (lanes 3 and 4). Mock-transfected cells produced only low levels of 7-hydroxylated 25-hydroxycholesterol (data not shown). To obtain sufficient mass to determine the chemical structure of the initial metabolite produced in transfected cells, the number of dishes in the experiment was increased by a factor of 30. After an 8-h incubation with 25-[³H]hydroxycholesterol, the sterol products were separated by preparative thin layer chromatography, and the compound of interest was isolated from the silica gel plate by extraction with organic solvents.

Gas chromatography of the isolated compound revealed a single peak indicative of a high degree of purity (Fig. 3A). The mass spectrum of this compound is shown in Fig. 3B. A prominent peak at m/z 632 corresponded to the M-90 peak of a tetrahydroxylated cholesterol species. The m/z 131 ion was characteristic of metabolites with hydroxylation at carbon 25, while the ion at m/z 251 indicated the presence of one double bond and three hydroxyl groups on the steroid nucleus. Finally, the m/z 129 ion was characteristic of a fragment containing the ^5,6-unsaturated bond and the 3-hydroxyl group on carbon 3. The positions of three of the four hydroxyl groups in the metabolite could thus be assigned to carbons 3, 7, and 25 based on previous analyses of Cyp7b1-generated products and on the known structure of the starting substrate.

The location of the fourth hydroxyl group was determined by incubating the metabolite with 2,2-dimethoxypropane, which produces an acetonide derivative if vicinal (adjacent) hydroxyl groups are present on the treated compound. An acetonide was produced at high efficiency (data not shown), which was indicative of vicinal hydroxylation at either carbons 2 and 3 or carbons 3 and 4. Position 4 was ruled out by the absence of prominent ions at m/z 147 or 181 in the mass spectrum (Fig. 3B; Ref. 22), which left carbon 2 as the recipient of the additional hydroxyl group. The configuration of the hydroxyl group at this carbon was not determined. Taken together, these data indicated that the identity of the novel polar product arising from cells transfected with the Cyp7b1 cDNA was cholest-5-ene-2,3,7,25-tetrol (Fig. 3C). Cyp7b1 was thus both an oxysterol 7-hydroxylase and a 2-hydroxylase.

The transfection studies were consistent with the notion that the enzyme encoded by the Cyp7b1 cDNA was an oxysterol 7-hydroxylase of bile acid biosynthesis. To gain further support for this hypothesis, we exploited the ability of nafimidone to block oxysterol 7-hydroxylase activity. The concentration of nafimidone required to inhibit 50% of enzyme activity in mouse microsomes (the IC₅₀ value) was approximately ~14 µM (Fig. 4). To determine if the oxysterol 7-hydroxylase enzyme activity encoded by the mouse Cyp7b1 cDNA showed the same inhibition profile, microsomes prepared from transfected 293 cells were assayed for nafimidone sensitivity. The inhibition curve generated with the transfected cell membranes was indistinguishable from that produced with liver microsomes (Fig. 4). The IC₅₀ value for the recombinant enzyme was ~17 µM. The concentrations of nafimidone required to achieve similar inhibition in intact cells were far greater (see above), suggesting that nafimidone crosses the cell membrane poorly. The similarity in nafimidone inhibition profiles obtained with hepatic and transfected cell membranes supported the idea that the Cyp7b1 cDNA encoded a hepatic oxysterol 7-hydroxylase enzyme activity in mice.

Nafimidone inhibition of oxysterol 7-hydroxylase activity in microsomes and transfected cells.

Oxysterol 7-hydroxylase enzyme activity is absent or low in the livers of young mice and is induced around the time of weaning (16). If the murine Cyp7b1 gene encodes the oxysterol 7-hydroxylase of bile acid biosynthesis, then the expression of Cyp7b1 mRNA and protein should closely follow the developmental induction of enzyme activity. To test this idea, livers were isolated from mice ranging in age from 5 to 55 days. Three to 10 animals per time point were used. Poly(A)⁺-enriched mRNA and microsomal membranes were prepared from pooled samples. RNA blotting was performed using a radiolabeled Cyp7b1 cDNA probe. Oxysterol 7-hydroxylase activity in the microsomal membranes was assayed with 25-[³H]hydroxycholesterol as substrate, and immunoblotting was performed using a polyclonal antipeptide antibody.

As shown in Fig. 5, there was a very good correlation between the level of oxysterol enzyme activity (upper panel), and the levels of both the Cyp7b1 mRNA (middle panel) and protein (lower panel). The time course of induction closely paralleled our previous findings (16) that the hepatic oxysterol 7-hydroxylase activity is induced around the third week of life (i.e. day 18, Fig. 5). Hybridization with a control probe (cyclophilin) showed equal amounts of RNA in each lane of the blot, and Ponceau S staining of the immunoblots showed equal amounts of protein in each lane (data not shown).

Induction of hepatic oxysterol 7-hydroxylase in mice.

To compare and contrast the regulation of this enzyme and that of cholesterol 7-hydroxylase, mice were placed on diets having well-characterized effects on the regulation of cholesterol 7-hydroxylase. These included normal laboratory diet or diets supplemented with 2% colestipol, 2% cholesterol, 0.5% cholic acid, or 2% cholesterol plus 0.5% cholic acid. After 10 days, livers were isolated from the different groups, pooled, and assayed for both cholesterol and oxysterol 7-hydroxylase mRNA levels as well as oxysterol 7-hydroxylase enzyme activity and protein levels.

As expected from previous work (23, 24), colestipol-containing diets increased steady state levels of cholesterol 7-hydroxylase mRNA (2.8-fold), as did the addition of cholesterol (1.4-fold) (Fig. 6A). Also, as expected, diets containing cholic acid or cholesterol plus cholic acid led to a marked suppression of the cholesterol 7-hydroxylase mRNA (Fig. 6A). In contrast, dietary supplementation with either colestipol or cholesterol had little or no effect on oxysterol 7-hydroxylase mRNA levels (Fig. 6A) or on enzyme activity or protein levels (Fig. 6B), whereas the addition of cholic acid or cholesterol plus cholic acid decreased the amount of oxysterol 7-hydroxylase mRNA activity and protein levels by 40-70% (Fig. 6, A and B). Under these experimental conditions, oxysterol 7-hydroxylase mRNA, protein, and activity closely paralleled one another.

Effect of diet on hepatic oxysterol 7-hydroxylase expression. Livers were isolated from groups of mice maintained on diets supplemented with the indicated sterols or drug and assayed for the presence of oxysterol 7-hydroxylase activity, mRNA and protein, and cholesterol 7-hydroxylase mRNA. A, upper panel, oxysterol 7-hydroxylase mRNA; lower panel, cholesterol 7-hydroxylase mRNA. Poly(A)⁺-enriched mRNA (15 µg) was first subjected to blot hybridization using radiolabeled fragments from the oxysterol 7-hydroxylase and cyclophilin cDNAs. The washed filter was subjected to autoradiography for 18 h, stripped, and reprobed with a radiolabeled fragment from the cholesterol 7-hydroxylase cDNA. The washed filter was subjected to autoradiography for 18 h. The positions of size standards are shown on the left of the autoradiograms. Numbers below the autoradiograms indicate the levels of 7-hydroxylase mRNAs in each liver pool relative to those in normal diet-fed mice. B, upper panel, oxysterol 7-hydroxylase enzyme activity. Aliquots of microsomal protein (100 µg) were incubated with 25-[³H]hydroxycholesterol (0.072 µM) and 1.5 mM NADPH for 15 min at 37 °C. Radiolabeled sterols were separated by thin layer chromatography and visualized by autoradiography. The positions to which enzyme substrate and product migrated are shown to the left of the autoradiogram. Numbers below the autoradiogram represent the diet induced changes in enzyme activity relative to control mice. Lower panel, immunodetection of oxysterol 7-hydroxylase protein. Liver microsomal proteins (100 µg) were separated on a 10% (w/v) polyacrylamide-SDS gel, transferred to a PolyScreen membrane (NEN Life Science Products), and blotted with an antipeptide antibody. Size standards are indicated on the left. Numbers below the lumigram indicate amounts of oxysterol 7-hydroxylase protein in each liver pool relative to that in control mice.

All of these findings were consistent with Cyp7bl encoding an oxysterol 7-hydroxylase that participates in bile acid biosynthesis. To determine the extent to which this enzyme activity, and hence the mitochondrial pathway leading to 7-hydroxylated bile acids, was present in other vertebrates, liver microsomal membranes isolated from 14 different species were assayed for their ability to convert 25-[³H]hydroxycholesterol into [³H]7-hydroxylated 25-hydroxycholesterol (Fig. 7). Oxysterol 7-hydroxylase activity was detected in all species tested, including humans (lane 1), ungulates (lanes 2-8), carnivores (lanes 9 and 10), lagomorphs (lane 11), rodents (lanes 12-16), and aves (lanes 17 and 18). Among these, the highest levels of enzyme activity were detected in mouse, hamster, rabbit, and pig livers. In the sheep, enzyme activity was present in both sexes and in young and adult animals (lanes 6-8). In each species, the initial 7-hydroxylated oxysterol product was further converted into a spectrum of metabolites that presumably reflected downstream intermediates of bile acid biosynthesis (Fig. 7).

Hepatic expression of oxysterol 7-hydroxylase in vertebrates.

DISCUSSION

In the current paper, we present several lines of evidence to support the hypothesis that the murine oxysterol 7-hydroxylase of bile acid biosynthesis is identical with Cyp7b1, a microsomal cytochrome P450 enzyme expressed in multiple tissues. First, transfection of an expression vector containing the mouse Cyp7b1 cDNA into cells produces oxysterol 7-hydroxylase enzyme activity. Second, the nafimidone inhibition profile of the mouse oxysterol 7-hydroxylase activity in microsomes is identical to that of the recombinant enzyme produced in transfected cells. Third, oxysterol 7-hydroxylase enzyme activity and Cyp7b1 mRNA and protein show concordant age-dependent induction patterns in mouse liver. Fourth, expression of oxysterol 7-hydroxylase enzyme activity and Cyp7b1 mRNA and protein are down-regulated by dietary bile acid. We further show that oxysterol 7-hydroxylase enzyme activity is widely distributed among vertebrate species and that the regulation of this enzyme is different from that of cholesterol 7-hydroxylase.

The Cyp7b1 cDNA utilized in this study was identified in a screen for genes that were preferentially expressed in the rodent hippocampus (18). DNA sequence analysis revealed that the encoded protein shared 39% sequence identity with cholesterol 7-hydroxylase, which suggested that Cyp7b1 would also be capable of 7-hydroxylating steroid substrates. This hypothesis has been confirmed with the findings that Cyp7b1 is capable of 7-hydroxylating 25-hydroxycholesterol (Fig. 1) and the neurosteroids dehydroepiandrosterone and pregnenolone (20). In addition, a study published while this paper was in review reported that recombinant Cyp7b1 converted 27-hydroxycholesterol to cholest-5-ene-3,7,27-triol (25). Here, we show that the murine enzyme, when overexpressed in transfected cells, also is an oxysterol 2-hydroxylase (Figs. 2 and 3). It is presently not clear whether the latter activity is peculiar to the mouse enzyme, is substrate-specific, or represents only an in vitro detectable side activity. Nevertheless, taken together, these studies indicate that Cyp7b1 is a diverse enzyme capable of participating in the synthesis of neurosteroids in the brain and of multiple bile acids in the liver.

The biosynthesis of 7-hydroxylated bile acids in the liver is accomplished by two pathways that differ in their initial steps (Fig. 8). A microsomal pathway begins with the 7-hydroxylation of cholesterol by the cytochrome P450 cholesterol 7-hydroxylase (5), whereas a mitochondrial pathway begins with hydroxylation of the side chain of cholesterol by sterol 27-hydroxylase or other enzymes (11). The second step of the mitochondrial pathway involves 7-hydroxylation of the oxysterol intermediate and is catalyzed by the oxysterol 7-hydroxylase characterized here. The cDNA-derived sequence of this cytochrome P450 enzyme suggests that it is localized to the endoplasmic reticulum (18). This enzyme is expressed in the liver of many species (Fig. 7; Refs. 12-14 and 16) and in peripheral tissues like the rodent brain (18) and possibly ovary (15). The enzyme participates in the synthesis of 7-hydroxylated bile acids in the liver of these species and in peripheral tissues; the oxysterol 7-hydroxylase may work in concert with the similarly expressed sterol 27-hydroxylase enzyme (26, 27) to facilitate excretion of oxysterols (28).

Two pathways of 7-hydroxylated bile acid biosynthesis.

The observation of redundancy in the enzymes that synthesize 7-hydroxylated sterols suggests that bile acids with this substituent are pivotal in lipid metabolism. Most mice that are deficient in cholesterol 7-hydroxylase die within the first 3 weeks of life, but those that survive this period thereafter have a normal life span (16). We previously proposed that another enzymatic pathway, specifically the oxysterol 7-hydroxylase pathway, was able to at least partially compensate for the absence of cholesterol 7-hydroxylase activity. This compensatory pathway is induced at approximately day 18 in the mouse, and expression of the oxysterol 7-hydroxylase enzyme leads to the synthesis of 7-hydroxylated bile acids (16). Our findings that oxysterol 7-hydroxylase activity is present in many different vertebrates (Fig. 7), further support the physiological importance of 7-hydroxylated bile acids.

The two hepatic pathways of 7-hydroxylated bile acid biosynthesis are differentially regulated (Fig. 6). Cholesterol 7-hydroxylase is induced by dietary cholesterol and by colestipol, which decreases the return of bile acids to the liver via the enterohepatic circulation. Dietary bile acid supplementation dramatically reduces cholesterol 7-hydroxylase. In contrast, oxysterol 7-hydroxylase remains unaffected by dietary cholesterol or colestipol and is only modestly down-regulated (~50%) by excess bile acids. Taken together, these data suggest that the liver is never without a capacity to synthesize 7-hydroxylated bile acids, even when presented with an excess of the end product of the pathways.

What then is the crucial physiological role of 7-hydroxylated bile acids? This stereochemistry may be important for some as yet undefined aspect of lipid uptake in the gut, or 7-hydroxylated bile acids may be less cholestatic (toxic) to the liver than bile acids without a 7-hydroxyl group. 7-Hydroxylated bile acids, but not 7-hydroxylated bile acids (29), mediate feedback regulation of cholesterol 7-hydroxylase; thus, the 7-configuration is of regulatory importance. We previously showed that mice deficient in cholesterol 7-hydroxylase, and therefore dependent on the oxysterol 7-hydroxylase/mitochondrial pathway for bile acid biosynthesis, contained abnormally low levels of vitamin E (16). Accordingly, chemical diversity arising from two pathways may be required to absorb the full complement of lipids and vitamins required for health. It is conceivable that as diets changed over time, the ability to synthesize a more diverse spectrum of bile acids was selected for, giving rise to the two pathways that are observed today in multiple vertebrates. In the future, we will search for additional pathways of bile acid biosynthesis and study the physiological roles of specific classes of bile acids in mice that are deficient in both cholesterol 7-hydroxylase and the oxysterol 7-hydroxylase.