The murine and human cholesterol 7alpha-hydroxylase gene promoters are differentially responsive to regulation by fatty acids mediated via peroxisome proliferator-activated receptor alpha.

We determined if fatty acids can regulate the murine Cyp7a1 and human CYP7A1 gene promoters via peroxisome proliferator-activated receptor alpha (PPARalpha)/9-cis-retinoic acid receptor alpha (RXRalpha). In transfected cells, the murine Cyp7a1 gene promoter displayed markedly lower basal activity, but greater sensitivity to fatty acid- or WY 14,643-activated PPARalpha/RXRalpha when compared with the human CYP7A1 gene promoter. PPARalpha/RXRalpha can bind to a site (Site II) located within the region at nucleotides -158 to -132 of both promoters. Mutagenesis of the human CYP7A1 Site II element abolished the response to activated PPARalpha/RXRalpha. The murine Cyp7a1 gene promoter contains an additional PPARalpha/RXRalpha-binding site (Site I) located within nucleotides -72 to -57. Replacement of a single residue in human CYP7A1 Site I with that found in the murine Cyp7a1 Site I sequence enabled PPARalpha/RXRalpha binding, and this mutation resulted in reduced basal activity, but substantially improved the response to activated PPARalpha/RXRalpha in transfected cells. We conclude that fatty acids can regulate the cyp7a gene promoter via PPARalpha/RXRalpha. The differential response of the murine Cyp7a1 and human CYP7A1 gene promoters to PPARalpha activators is attributable to the additional PPARalpha/RXRalpha-binding site in the murine Cyp7a1 gene promoter.

The liver plays an important role in maintaining whole body cholesterol homeostasis by regulating the biosynthesis of cholesterol, its uptake from plasma, storage, and catabolism to bile acids (16). The conversion of cholesterol to bile acids in the liver represents an important route by which cholesterol is eliminated from the body. Cholesterol 7␣-hydroxylase (cyp7a) is the rate-controlling enzyme of the major pathway involved in this process (17). The gene encoding cyp7a has been cloned from the rat (18,19), mouse (20,21), human (22,23), and hamster (24). Several studies have shown that cyp7a gene expression is regulated by a variety of hormonal (25)(26)(27)(28) and nutritional (29 -36) factors. Although phosphorylation has been suggested to regulate cyp7a (30,37,38), the abundance of cyp7a mRNA remains the best predictor of cyp7a enzyme activity.
The proximal promoter region of the cyp7a gene contains sequences that resemble nuclear hormone response elements. It was shown recently that induction of the rat Cyp7a1 gene in response to oxysterols involves the nuclear receptor LXR␣ (39), whereas repression of cyp7a gene transcription by bile acids appears to be mediated via the farnesoid X receptor (40). Other nuclear hormone receptors have also been shown to interact with the cyp7a gene promoter (41)(42)(43)(44). Given the number of nuclear hormone receptor-binding sites in the cyp7a gene promoter, it would seem that this class of transcription factors plays a major role in regulating cyp7a gene expression.
We previously demonstrated that dietary fats influence the response of the murine Cyp7a1 gene to dietary cholesterol (35,36). It is well known that fatty acids regulate gene transcription via PPAR␣/RXR␣ (2, 7). Here we evaluated the role of fatty acids in regulating the cyp7a gene expression and whether PPAR␣/RXR␣ is involved in the process.

Cloning and Construction of Recombinant Plasmids-Genomic
clones containing the human CYP7A1 gene were isolated from a human chromosome 8 library (American Type Culture Collection, Manassas, VA). A genomic clone containing the entire murine Cyp7a1 gene was described previously (21). Gene chimeras containing the proximal promoter region of the human CYP7A1 gene (nt Ϫ372 to ϩ61) or the murine Cyp7a1 gene (nt Ϫ412 to ϩ77) were generated from the cloned cyp7a genomic genes and linked to the chloramphenicol acetyltransferase (CAT) structural gene sequence in pCAT-Basic (Promega Corp., Madison, WI). The primary structures of the in vitro amplified human CYP7A1 and murine Cyp7a1 gene fragments were confirmed by sequencing. The rabbit antiserum to PPAR␣ and plasmids encoding murine PPAR␣ and RXR␣ were described previously (45).
Mutant derivatives of the human CYP7A1 gene promoter were con-structed using mutagenic primers by in vitro DNA amplification and the plasmid containing the wild-type gene chimera as template. The 3Ј-half-site of the human CYP7A1 Site II regulatory element was converted into an XhoI site, 2 and a point mutation was introduced into the 3Ј-half-site of the human CYP7A1 Site I element. The primary structures of the mutagenized promoters were verified by sequencing. Cell Culture and Transfections-McArdle RH-7777 rat hepatoma cells (46) were maintained in Dulbecco's modified Eagle's medium supplemented with 20% fetal calf serum at 37°C in a humidified atmosphere of 5% CO 2 and 95% air. Cells were plated (60-mm culture dishes) at equal densities 16 -18 h prior to transfection and grown in medium containing 20% delipidated fetal calf serum. The cells were transfected when cultures were at 50 -60% confluency using the calcium phosphate coprecipitation procedure (47) with a mixture of plasmids that contained 2.5 g of plasmid pCMV-␤-gal (encoding ␤-galactosidase) as a control for transfection efficiency, cyp7a-CAT gene chimeras, and expression vectors encoding PPAR␣ and RXR␣. All samples were complemented with sonicated salmon sperm DNA to equalize the mass of total DNA (20 g) added to each dish. One hour after the introduction of DNA, the transfected cells were treated with either fatty acids (stearic acid (18:0), oleic acid (18:1), and linoleic acid (18:2); Sigma-Aldrich, Oakville, Ontario, Canada) complexed to lipid-free bovine serum albumin (BSA) or WY 14,643 (BIOMOL Research Labs Inc., Plymouth Meeting, PA) dissolved in dimethyl sulfoxide at a final concentration of 100 or 50 M, respectively. The control cells received either lipid-free BSA or Me 2 SO alone. In initial experiments, we found that fatty acids dissolved in ethanol were toxic when added to the culture medium of transfected McArdle RH-7777 cells at concentrations Ն100 M. However, the equivalent concentrations of fatty acids were well tolerated when added to cell culture medium as a complex with BSA and were therefore used in all subsequent experiments. Extracts of transfected cells were prepared 36 -48 h after the introduction of DNA and assayed for both CAT and ␤-galactosidase activities by standard methods. The CAT reaction products were separated by TLC, and radioactivity of acetylated [ 14 C]chloramphenicol was quantitated using a Fuji BAS1000 phosphoimager. The CAT activity values were normalized to the ␤-galactosidase activity, which was not affected by the treatments.
Electrophoretic Mobility Shift Assays-The murine PPAR␣ and RXR␣ proteins were synthesized in vitro using a coupled transcription/ translation system (Promega Corp.) with a standard scale of synthesis consisting of 25 l of reaction mixture and 0.5 g of DNA template. Fragments produced by restriction enzyme digestion of the human CYP7A1 gene promoter were first tested for binding to PPAR␣/RXR␣ by electrophoretic mobility shift assays. The sequences of the promoter fragments that showed binding to PPAR␣/RXR␣ were analyzed, and regions that bore similarity to the consensus PPRE sequence were further characterized. The following synthetic double-stranded oligonucleotides bear natural sequences from the cyp7a gene promoters and were used in electrophoretic mobility shift assays: nt Ϫ150 to Ϫ123 of the human CYP7A1 gene (human Site II, 5Ј-TACCTGTGGACTTAGT-TCAAGGCCAGTT-3Ј), nt Ϫ73 to Ϫ54 of the human CYP7A1 gene (human Site I, 5Ј-CTTTGTCAACCAAGCTCAAG-3Ј), and nt Ϫ73 to Ϫ55 of the murine Cyp7a1 gene (murine Site I, 5Ј-CTGGTCACCCAAGT-TCAAG-3Ј). The altered residues in the mutagenized derivatives of these sequences are described in the figure legends. The oligonucleotides were synthesized with a 4-nt 5Ј-extension (5Ј-AATT-3Ј) to allow for radiolabeling of annealed oligonucleotides by fill-in reaction catalyzed by the Klenow fragment of Escherichia coli DNA polymerase I. Radiolabeled double-stranded oligonucleotides (0.1 pmol, Ϸ15,000 cpm) were incubated in a total volume of 40 l for 15 min at 25°C with 4 g of poly(dI⅐dC), 4 g of BSA, and 1 l of recombinant PPAR␣ and RXR␣ in binding buffer (10 mM HEPES, pH 7.9, 150 mM NaCl, 1 mM EDTA, 7% glycerol, 150 M phenylmethylsulfonyl fluoride, and 1 mM dithiothreitol). Double-stranded oligonucleotides containing the rat acyl-CoA oxidase PPRE (45) and the rat tyrosine aminotransferase glucocorticoid response element (48) were used as competitors for PPAR␣/RXR␣ binding. In competition experiments, increasing amounts (5-50-fold molar excess) of unlabeled acyl-CoA oxidase PPRE, murine Cyp7a1 Site I, or human CYP7A1 Site I were added to reaction mixtures and incubated at room temperature for 10 min prior to the addition of radiolabeled double-stranded oligonucleotides. After an additional 15 min of incubation at room temperature, DNA-protein complexes were separated by electrophoresis on a 4% polyacrylamide gel in 0.5ϫ Tris borate electrophoresis buffer at 4°C and visualized by autoradiography or phosphoimaging. In some experiments, 1-2 l of rabbit anti-PPAR␣ antiserum (45) was added after incubation of the radiolabeled probes with the reticulocyte lysates. The reaction was left on ice for an additional 30 -60 min prior to electrophoresis.

RESULTS
Fatty Acids Stimulate the Murine Cyp7a1 and Human CYP7A1 Gene Promoters via PPAR␣/RXR␣-We observed that mice consuming a semipurified high fat (20%) diet in the absence of exogenous bile acids had a significantly higher (2-3fold, p Ͻ 0.01) abundance of Cyp7a1 mRNA compared with mice consuming a chow diet (5% fat content). To investigate whether fatty acids directly regulate cyp7a gene transcription through PPAR␣/RXR␣, gene chimeras containing the murine Cyp7a1 or human CYP7A1 proximal promoter regions linked to the CAT structural gene were transfected into rat hepatoma cells in the absence or presence of plasmids encoding PPAR␣ and RXR␣. In the presence of PPAR␣/RXR␣, the basal activity of the murine Cyp7a1 gene promoter was significantly lower (p Ͻ 0.001) than that of the human CYP7A1 gene promoter (Fig. 1A). Addition of stearic acid (18:0), oleic acid (18:1), or linoleic acid (18:2) complexed to lipid-free BSA to the culture medium at a concentration of 100 M increased the activities of both the murine Cyp7a1 and human CYP7A1 gene promoters as reflected by the rise in CAT activity in lysates of treated cells (Fig. 1B). These results indicate that fatty acids stimulate the cyp7a gene promoter of both species.
Interestingly, the response of the murine Cyp7a1 gene promoter to fatty acids was consistently greater than that of the human CYP7A1 gene promoter. To verify the observed difference in the response of the murine Cyp7a1 and human CYP7A1 gene promoters to fatty acids, cells transfected with the chimeric plasmids were treated with WY 14,643, a potent and well characterized synthetic activator of PPAR␣ (6,8,10). As shown in Fig. 1C, the response of the murine Cyp7a1 gene promoter to WY 14,643 was significantly greater (11-fold activation) compared with the human CYP7A1 gene promoter (1.5-fold activation). These results show that although the human CYP7A1 gene promoter has a higher basal activity compared with the murine Cyp7a1 gene promoter, the human CYP7A1 gene is less responsive to regulation by activated PPAR␣/RXR␣.
The cyp7a Gene Promoter Contains a PPAR␣/RXR␣-binding Site-Analysis of the human CYP7A1 and murine Cyp7a1 gene promoter fragments produced by restriction enzyme digestion revealed that PPAR␣/RXR␣ can bind to the cyp7a promoter somewhere in the region between nt Ϫ255 and Ϫ47 (data not shown). Inspection of the sequence revealed that this region of the cyp7a gene promoter contains a DR1 motif that resembles a PPAR␣/RXR␣-binding site. This DR1 motif spans nt Ϫ144 to Ϫ132 and nt Ϫ158 to Ϫ146 of the human CYP7A1 and murine Cyp7a1 gene promoters, respectively. The sequence of this site, which we have designated as Site II, is invariant among the promoter regions of the human, murine, rat, and hamster cyp7a genes ( Fig. 2A). To determine if Site II could function as a PPRE, we tested its ability to interact with PPAR␣/RXR␣ by electrophoretic mobility shift assay. Double-stranded oligonucleotides containing the Site II sequence displayed strong binding to PPAR␣/RXR␣ heterodimers (Fig. 2B, left arrow). Interaction of the probe with the recombinant transcription factors was not observed when either PPAR␣ or RXR␣ was omitted from the binding reactions (data not shown). Addition of anti-PPAR␣ antibodies to the binding reaction containing PPAR␣/ RXR␣ resulted in the formation of a complex with greatly reduced mobility (Fig. 2B, right arrow), demonstrating that the complex contains PPAR␣. The binding of Site II to PPAR␣/ RXR␣ was competed efficiently by the well characterized acyl-CoA oxidase PPRE (45), but not by a glucocorticoid response element (48) (Fig. 2C, left panel). Mutagenesis of the 3Ј-half-site of Site II drastically reduced binding to PPAR␣/RXR␣ (Fig.  2C, right panel). These data demonstrate that PPAR␣/RXR␣ can interact with the cyp7a gene promoter.
Next, we evaluated the role of Site II in the regulation of the cyp7a gene promoter by ligand-activated PPAR␣/RXR␣. The 3Ј-half-site of Site II (a DR1 element) was mutagenized, and then CAT gene chimeras containing the human CYP7A1 gene promoter with either the wild-type or mutagenized Site II element were transfected into hepatoma cells along with expression vectors encoding PPAR␣ and RXR␣. The activities of the wild-type and mutant Site II elements were compared in the presence and absence of 50 M WY 14,643. As shown in Fig.  2D, the cells transfected with the gene chimera containing the wild-type human CYP7A1 Site II showed a small but reproducible increase (1.5-fold, p Ͻ 0.05) in CAT activity, whereas cells transfected with the CYP7A1 gene chimera containing mutagenized Site II showed no response to WY 14,643. These results demonstrate that PPAR␣/RXR␣ can regulate the cyp7a gene promoter through Site II. As the sequence of Site II is identical in both human CYP7A1 and murine Cyp7a1 gene promoters, the basis for the more robust response of the murine Cyp7a1 gene promoter to fatty acids was unclear.
The Murine Cyp7a1 Gene Promoter Contains an Additional PPAR␣/RXR␣-binding Site-The LXR␣/RXR␣ heterodimers bind to a tandem repeat of a nuclear hormone-binding sequence motif separated by four unconserved residues (i.e. a DR4 element termed the LXR response element) (49). Region Ϫ72 to Ϫ57 of the murine Cyp7a1 gene promoter contains a DR4 element that is similar to the LXR response element found in the rat Cyp7a1 gene promoter (39). Alignment of murine and rat DR4 elements shows a single nucleotide difference in the spacer region, but the 5Ј-and 3Ј-half-sites are identical (Fig.  3A). Inspection of the murine DR4 sequence (designated as Site I) suggests the existence of a possible DR1 element embedded within DR4. The DR1 element is configured in such a way that it shares a common 3Ј-half-site with the DR4.
Comparison of the murine, human, rat, and hamster cyp7a gene promoters revealed that the sequence of Site I is not conserved (Fig. 3A). Electrophoretic mobility shift assays using oligonucleotides corresponding to the murine Site I element demonstrated a strong interaction with PPAR␣/RXR␣, whereas the corresponding Site I element of the human CYP7A1 gene promoter was unable to bind PPAR␣/RXR␣ (Fig. 3B, left arrow). Binding of PPAR␣ was confirmed by the formation of a complex with greatly reduced mobility in the presence of anti-PPAR␣ antibodies (Fig. 3B, right arrow). Consistent with these results, the binding of PPAR␣/RXR␣ to murine Cyp7a1 Site I could be competed with the acyl-CoA oxidase PPRE, but not with the human CYP7A1 Site I element (Fig. 3C). These results indicate that the Site I element of the murine Cyp7a1 gene promoter is configured to enable binding of PPAR␣/RXR␣.
Alteration of a Single Nucleotide in the Human CYP7A1 Site I Regulatory Element Allows PPAR␣/RXR␣ Binding-The human CYP7A1 Site I differs from the murine Cyp7a1 Site I by three nucleotide residues (Fig. 4A). We designed three different double-stranded oligonucleotides containing the human CYP7A1 Site I sequence in which each of the variant residues was replaced with the corresponding residue in the murine Cyp7a1 Site I sequence. Changes from T-to-G and from A-to-C in the 5Ј-half-site of the DR4 element and in the spacer region, respectively, did not promote binding of human CYP7A1 Site I to PPAR␣/RXR␣ (Fig. 4B, left panel). In contrast, the C-to-T substitution in the 3Ј-half-site of human CYP7A1 Site I enabled binding to PPAR␣/RXR␣ (Fig. 4B, left panel), and the binding was competed by both the acyl-CoA oxidase PPRE and murine Cyp7a1 Site I (Fig. 4B, right panel).
To demonstrate the functional significance of the human CYP7A1 Site I C-to-T mutation, a cyp7a-CAT gene chimera containing mutagenized human CYP7A1 Site I was created and analyzed for its response to WY 14,643-activated PPAR␣/ RXR␣ in hepatoma cells. Analysis of CAT activity in extracts of cells transfected with the gene chimera carrying this mutation revealed a marked reduction of the basal promoter activity as compared with cells transfected with the wild-type version of the human CYP7A1 promoter (Fig. 4C, left panel). This result, taken together with the results shown in Fig. 1A, suggests that the low basal activity of the murine Cyp7a1 gene promoter compared with the wild-type human CYP7A1 gene promoter is attributable to the sequence of the Site I element. As observed before, the wild-type human CYP7A1 gene promoter showed a small (1.5-fold) but consistent stimulation in the presence of WY 14,643. In contrast, the human CYP7A1 gene promoter containing the Site I C-to-T mutation responded significantly better (3.5-fold, p Ͻ 0.01) (Fig. 4C, right panel). These results illustrate that a single nucleotide difference in the 3Ј-half-site of Site I can alter the basal activity of the cyp7a promoter as well as determine the ability of Site I to interact with PPAR␣/RXR␣. DISCUSSION Both the murine Cyp7a1 and human CYP7A1 gene promoters were induced when the culture medium of cells transfected with gene chimeras containing these promoters was supplemented with fatty acids. The stimulatory effect was observed even with stearic acid, a fatty acid that interacts weakly with PPAR␣ (9,11). This could be explained by the fact that metabolism of fatty acids in the transfected cells can produce metabolites that are more potent activators of PPAR␣ than the parental fatty acids. Inspection of the proximal region of the cyp7a gene promoters from various mammalian species revealed the existence of an invariant DR1 motif that is a possible binding site for PPAR␣/RXR␣. Electrophoretic mobility shift assays using recombinant PPAR␣ and RXR␣ revealed that this sequence (termed Site II) was capable of binding PPAR␣/RXR␣ in vitro. Since mutagenesis of the human Site II element abolished the stimulatory effect of fatty acids on the human CYP7A1 gene promoter, we surmised that this site is capable of acting as a target for PPAR␣-mediated signaling.
Although both fatty acids and WY 14,643 reproducibly stimulated the human CYP7A1 gene promoter in our experimental system, the magnitude of the stimulation was modest. Fibrates have generally been shown to increase the concentration of cholesterol in human and rat bile, and this effect appears to be correlated with the reduction of cholesterol 7␣-hydroxylation activity in both species (50 -53). However, it should be noted that some fibrates have only minimal effects on the lithogenic index of bile or cyp7a activity (50,53). Significant changes in the activities of other cholesterol-metabolizing enzymes have also been demonstrated (50,51), and these effects likely also contribute to changes in the lithogenic index of bile.
Recently, it was discovered that human liver contains significantly lower levels of PPAR␣ mRNA compared with murine liver (54). Measurement of PPAR␣/RXR␣ DNA-binding activity in liver extracts suggested that human liver contains an order of magnitude less activity compared with murine liver (54). These observations raise the possibility that PPAR␣ may not play a significant role in regulating human CYP7A1 gene expression. The conserved Site II element of the cyp7a gene promoter is also capable of binding other transcription factors (42)(43)(44) and thus may not be available for PPAR␣/RXR␣ interaction in vivo. The reduction of cholesterol 7␣-hydroxylation activity observed in humans and rats after fibrate therapy may therefore be due to an indirect effect of PPAR␣-mediated signaling. A transgenic mouse strain bearing the human CYP7A1 gene should aid in clarifying the significance of PPAR␣ function in regulating human CYP7A1 gene expression.
In contrast, the murine Cyp7a1 gene promoter was highly responsive to fatty acids and WY 14,643 in transfected cells. Since the murine Cyp7a1 gene promoter did not contain other sequences that resemble the typical PPAR␣/RXR␣-binding site (a DR1 motif) (2,5,45), the basis for the enhanced response of the murine Cyp7a1 gene promoter in transfected cells was not immediately obvious. Mapping of the murine proximal gene promoter using restriction fragments as probes in gel mobility shift assays surprisingly revealed that the fragment containing the LXR␣/RXR␣-binding site (LXR response element, a DR4 motif) (49) could also bind PPAR␣/RXR␣ in vitro. Unlike Site II, the sequence that makes up the murine DR4 element (termed FIG. 4. Functional analysis of the human CYP7A1 Site I regulatory element. A, the murine Cyp7a1 and wild-type and mutant human CYP7A1 Site I sequences are shown. The dots represent residues identical to the human sequence, and the lowercase letters represent substituted residues. B, PPAR␣/RXR␣ binding to wild-type and mutant CYP7A1 Site I elements was evaluated by electrophoretic mobility shift assay (left panel, arrow). The specificity of PPAR␣/RXR␣ binding to human CYP7A1 Site I (C to T) (right panel, arrow) was tested by adding a 5-or 50-fold molar excess of either the rat acyl-CoA oxidase (AOx) PPRE or the murine Cyp7a1 Site I element relative to the radiolabeled probe in the binding reaction. C, the cyp7a-CAT gene chimeras with a wild-type or mutant Site I element were introduced into hepatoma cells along with plasmids encoding PPAR␣, RXR␣, and ␤-galactosidase. The left panel shows the relative CAT activity normalized to ␤-galactosidase activity encoded by a cotransfected ␤-galactosidase expression vector in the absence of PPAR␣ activators (white bars). The right panel shows the response of the cyp7a promoters to WY 14,643 (black bars). The normalized CAT activity values in the absence of PPAR␣/RXR␣ ligands (white bars) were assigned a value of 1. Results shown represent the means Ϯ S.E. of three independent experiments performed in triplicate.
Site I) of the Cyp7a1 gene promoter is not conserved among the different mammalian species. The human and murine Site I sequences are non-identical; and unlike murine Site I, the human variant is not capable of binding PPAR␣/RXR␣. However, the substitution of a single residue at the 3Ј-half-site of the human Site I sequence with that found in the same position of the murine Site I element enabled PPAR␣/RXR␣ binding. In addition, this substitution substantially improved the response of the human CYP7A1 gene promoter to WY 14,643 in the presence of PPAR␣/RXR␣ in transfected cells. The magnitude of the response of the modified human CYP7A1 gene promoter did not reach that observed for the murine Cyp7a1 gene promoter, but this may be due to other sequence differences between the two promoters. Thus, we attribute the robust response of the murine Cyp7a1 gene promoter to fatty acids and WY 14,643 to the existence of a second PPAR␣/RXR␣-binding site. Studies done on mice that are homozygous for a disrupted Scp2 gene (encodes sterol carrier protein 2 and sterol carrier protein X) (55) provide additional support for the involvement of PPAR␣ in the regulation of murine Cyp7a1 gene expression. The expression of the Cyp7a1 gene in these mice is increased, coincident with the enhanced expression of hepatic genes known to be regulated by PPAR␣ (55). This effect has been ascribed to the accumulation of phytanic acid, a metabolite that can act as a ligand for and activator of PPAR␣ (56).
Previously, we demonstrated that the fat background of the diet had a significant impact on the response of the murine Cyp7a1 gene promoter to dietary cholesterol (35,36). The stimulatory effect of dietary cholesterol on the murine Cyp7a1 gene is likely mediated via LXR␣, a nuclear hormone receptor that is activated by oxysterols (39,57), since inactivation of the LXR␣ gene in mice abolishes the response (58). Based on the present study, it appears that PPAR␣/RXR␣ and LXR␣/RXR␣ bind to overlapping sequences in the murine Cyp7a1 gene promoter. The exclusive interaction of these transcription factors with the murine Site I element may explain why fats have an apparent influence on the stimulatory effect of dietary cholesterol. The exact requirements dictating the preferential binding of either of these transcription factors to the murine Site I element remain to be elucidated. In any case, the data suggest that fats are important physiological regulators of murine Cyp7a1 gene expression.
The inhibitory effect of bile acids on cyp7a gene expression is a well documented phenomenon (17). The discovery that bile acids serve as ligands for the farnesoid X receptor provides an explanation for the repressive effect of bile acids on the transcription of bile acid-responsive genes (40,59,60). A recent study demonstrated that ligand-bound farnesoid X receptor is capable of antagonizing transactivation by LXR␣ (60), although the precise mechanism by which this effect is accomplished remains to be described. It will be of interest to determine also if the stimulatory effect of ligand-activated PPAR␣ on murine Cyp7a1 gene expression is influenced by the farnesoid X receptor.
In summary, we demonstrated that the murine Cyp7a1 and human CYP7A1 gene promoters are differentially responsive to regulation by fatty acids via PPAR␣/RXR␣ in transfected cells. The basis for the differential sensitivity is apparently due to a variation in a critical residue in the Site I element of the cyp7a gene promoter. The murine Site I variant enables PPAR␣/ RXR␣ binding and allows the murine Cyp7a1 gene promoter to be more responsive to regulation by fats. Mice may have either maintained or acquired more sophisticated control of cyp7a gene expression through evolution to allow greater flexibility in regulating bile acid synthesis.