The Orphan Nuclear Receptor LRH-1 Potentiates the Sterol-mediated Induction of the Human CETP Gene by Liver X Receptor* 210

The human cholesteryl ester transfer protein (CETP) transfers cholesteryl esters from high density lipoproteins to triglyceride-rich lipoproteins, indirectly facilitating cholesteryl esters uptake by the liver. Hepatic CETP gene expression is increased in response to dietary hypercholesterolemia, an effect that is mediated by the activity of liver X receptor/retinoid X receptor (LXR/RXR) on a direct repeat 4 element in the CETPpromoter. In this study we show that the orphan nuclear receptor LRH-1 also transactivates the CETP promoter by binding to a proximal promoter element distinct from the DR4 site. LRH-1 potentiates the sterol-dependent regulation of the wild typeCETP promoter by LXR/RXR. Small heterodimer partner, a repressor of LRH-1, abolishes the potentiation effect of LRH-1 but not its basal transactivation of the CETP promoter. Since this mode of regulation of CETP is very similar to that recently reported for the bile salt-mediated repression of Cyp7a(encoding the rate-limiting enzyme for conversion of cholesterol into bile acid in the liver), we examined the effects of bile salt feeding on CETP mRNA expression in human CETPtransgenic mice. Hepatic CETP mRNA expression was repressed by a diet containing 1% cholic acid in male mice but was induced by the same diet in female mice. Microarray analysis of hepatic mRNA showed that about 1.5% of genes were repressed, and 2.5% were induced by the bile acid diet. However, the sexually dimorphic regulatory pattern of the CETP gene was an unusual response. Our data provide further evidence for the regulation ofCETP and Cyp7a genes by similar molecular mechanisms, consistent with coordinate transcriptional regulation of sequential steps of reverse cholesterol transport. However, differential effects of the bile salt diet indicate additional complexity in the response of these two genes.

The cholesteryl ester transfer protein (CETP) 1 catalyzes the transfer of cholesterol ester from HDL to triglyceride-rich lipoproteins (1). CETP is expressed in liver, intestine, and a number of peripheral tissues, such as adipose (1). In humans and animals, plasma CETP and tissue mRNA levels are increased in response to high fat, high cholesterol diets or endogenous hypercholesterolemia. These increases are due to elevated CETP gene transcription especially in the liver (2,3). Transgenic mice expressing human CETP, controlled by its natural flanking region, also increase expression of CETP in response to hypercholesterolemia (4). The mechanism of this effect was recently shown to involve the transcription factor LXR, binding as a heterodimer with RXR to a site in the CETP proximal promoter, a direct repeat of a nuclear receptor binding sequence separated by 4 nucleotides (DR4 element, Ϫ384 to Ϫ399). This response was seen with both LXR␣ and LXR␤, related nuclear hormone receptors that bind and are activated by specific hydroxylated sterols at physiological concentrations (5,6).
LXR␣ also transactivates the Cyp7a gene, encoding cholesterol 7␣-hydroxylase, the rate-limiting enzyme in the pathway converting cholesterol to bile acids (7). Furthermore, disruption of LXR␣ in mice abolished the induction of Cyp7a expression by dietary cholesterol (8). Based on these data, we proposed that LXRs may coordinate the regulation of genes involved in different steps of reverse cholesterol transport (9). This idea was further supported by the demonstration that ABCA1 is upregulated by sterols in an LXR-dependent fashion, due to the interaction of LXR with a DR4 element in the proximal promoter of the ABCA1 gene (10). ABCA1, the gene that is mutated in Tangier disease, mediates phospholipid and cholesterol efflux from macrophages to apoA-I (11)(12)(13)(14)(15). Recently, additional LXR-regulated ABC transporters were shown to be mutated in sitosterolemia, implying a role for these molecules in excretion of sitosterol and cholesterol from intestinal cells (16). In addition to stimulating reverse cholesterol transport and cholesterol excretion, LXRs activate the promoter of SREBP-1c, indicating a role in the regulation of fatty acid synthesis (17,18).
Liver receptor homologue-1 (LRH-1) is a mouse homologue of the orphan nuclear receptor fushi tarazu F1 (Ftz-F1) from Drosophila. CYP7A promoter binding factor (CPF), the human homologue of LRH-1, was found to transactivate the human CYP7A promoter (19). LRH-1 and CPF bind to an extended nuclear hormone receptor-binding site as monomers. LRH-1 has also been shown to act as a competence factor, enhancing the ability of LXR␣/RXR␣ to mediate a sterol response on the Cyp7a gene (20). This interaction is abolished by small heterodimer partner 1 (SHP-1), just as SHP-1 negates the inter-* 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. action of DAX1 with SF-1 (21) and represses the activity of several nuclear receptors (22)(23)(24). Since CETP and Cyp7a are both regulated by LXR/RXR, we investigated the role of LRH-1 and SHP-1 in the regulation of CETP gene expression. Our study shows a marked similarity between the effects of these factors on the CETP and Cyp7a promoters. This supports the idea of coordinate regulation of the sequential metabolic steps mediated by CETP and CYP7A. However, divergent responses of these two genes to a cholic acid-containing diet indicate additional complexity in the in vivo response to bile acids.

MATERIALS AND METHODS
Cell Culture and Transfection-HEK293 cells were maintained in Dulbecco's modified Eagle's medium with 10% fetal bovine serum at 37°C. 70 -80% confluent cells were transfected using LipofectAMINE transfection reagent (Life Technologies, Inc.) as described previously (10). 0.2 g of reporter DNA, 25 ng of pCMV-RL (Renilla) (Promega), and 100 ng of receptors (CMX-hRXR␣, CMX-hLXR␣, CMV-LXR␤, CMV-LRH, and CMV-SHP) were used in each transfection experiment. Empty pcDNA3.1 expression vector (Invitrogen) was used as control and to maintain equal amounts of DNA (0.625 g per well in 24-well plate) for each transfection. The transfected cells were cultured in 5% LPDS medium (consists of Dulbecco's modified Eagle's medium with 5.1 lipoprotein deficient serum) in the presence of 2 g/ml 22(R)-hydroxycholesterol (Sigma) or vehicle alone for 24 h. The luciferase activities were measured using Promega Dual Luciferase assay system. Reporter constructs were constructed as described previously (9). Each experiment was carried out in duplicate.
Plasmid Construction-LRH-1 and SHP-1 were cloned by reverse transcriptase-polymerase chain reaction from mouse liver RNA and subcloned into pcDNA3.1 (Invitrogen). The 10-amino acid epitope from human c-MYC (EQKLISEEDL), which can be recognized by the monoclonal antibody 9E10, was used for tagging the LRH-1 gene to produce Myc-LRH fusion protein. Myc-LRH was subcloned into the pcDNA3.1 expression vector.
Dietary and mRNA Studies in Human CETP Transgenic Mice-Human CETP transgenic mice (C57BL/6J background, 10 -15 weeks old), expressing CETP controlled by its native flanking region (3.4-kb NFR) (4), were fed standard rodent chow diets or chow diets supplemented with 1% cholic acid (CA) (TD00548, Harlan Teklad) for 5 days. The animals were housed under a 12-h light/dark cycle. Total RNA was isolated using Trizol Reagent (Life Technologies, Inc.) and pooled from four female or four male mice for each diet. Poly(A) ϩ RNA was prepared using Qiagen Oligotex mRNA purification Kit. Northern blotting was carried out as described (25). The blots were exposed to a PhosphorImager screen and visualized with Molecular Dynamics PhosphorImager system. The intensity of the bands was quantified using an Image-Quant tool IQMac version 1.2.
Microarray Printing and Hybridization-The Escherichia coli bacterial colonies harboring 1248 unique cDNA or EST clones were individually grown in 96-well microplates (Corning Costar Co.). The cDNA inserts were polymerase chain reaction-amplified with vector-derived primers directly using 5 l of bacterial lysates and subsequently purified in the microplates. A GMS 417 Arrayer (Genetic MicroSystems) was then used to deposit purified polymerase chain reaction products in 3ϫ SSC onto polylysine-coated glass slides (Sigma). The printed cDNA microarrays were then processed as described (26). The fluorescencelabeled cDNA probes for array hybridization were prepared from purified liver mRNA (4 g), pooled from 4 mice fed with chow diet or chow diet supplemented with 1% CA, by using Superscript II reverse transcriptase (Life Technologies, Inc.) and fluorescent Cy3-or Cy5-dUTP (Amersham Pharmacia Biotech). Hybridization of both Cy5-and Cy3labeled probes in 4ϫ SSC, 0.3% SDS, 50% formamide to the same microarray was carried out in a sealed, humid hybridization cassette (TeleChem International) for about 14 h at 42°C. After washing and drying, the slides were scanned with a confocal array scanner GMS 418 (Genetic MicroSystems). The fluorescence signals of Cy5-and Cy3tagged cDNA spots on arrays were quantified using an ArrayVision software package (Imaging Research). Two independent microarray hybridization experiments were carried out to analyze the gene expression profile in control transgenic mice or cholic acid-fed mice. The ratios of Cy5/Cy3 (ϩCA/ϪCA) fluorescence intensity of each cDNA sample were normalized with actin signal ratios obtained from the same measurements and were averaged from two hybridization results.
Electrophoresis Mobility Shift Assays-Myc epitope-tagged LRH-1 protein was translated in vitro using the TNT Quick-Coupled Tran-scription/Translation System (Promega). Double-stranded oligonucleotides with HindIII overhangs, corresponding to wild type CETP LRHbinding site (LRHBS) (5Јaggaagaccctgctgc3Ј) and mutated LRHBS (5ЈaggaagaGcAtgctgc3Ј) or Cyp7a LRH element (LRHE) (20), were used in gel shift experiments. 2 l of lysates expressing Myc-LRH or luciferase (control) were mixed with ϳ50 fmol of 32 P-end-labeled LRHBS fragment in a volume of 20 l of binding buffer (75 mM KCl, 20 mM HEPES, pH 7.9, 2 mM dithiothreitol, 10% glycerol, 2 g of poly(dI-dC), 30 pmol of nonspecific single-stranded oligonucleotides). Reactions were incubated at room temperature for 20 min, and protein-DNA complexes were resolved on 5% polyacrylamide gels at 140 V for 1 h. For competition experiments, ϳ50-fold molar excess of unlabeled competitor DNA relative to labeled DNA were added to the reaction mixture before the addition of the labeled probe. In antibody experiments, the protein lysates were first incubated with 0.4 g of Myc epitope antibody 9E10 (sc-40, Santa Cruz Biotechnology) or control monoclonal anti-actin antibody (sc-8432, Santa Cruz Biotechnology) for 10 min without the labeled DNA, followed by 20 min of incubation in the presence of labeled DNA.

Potentiation of LXR Function by LRH-1-Whereas
LXRs induce robust sterol activation of a deleted or multicopy version of the CETP promoter (9), activation of the intact, single copy CETP promoter by sterol was either modest (LXR␣) or nonexistent (LXR␤) (Fig. 1). To see if the competence factor LRH might increase the sterol activation, as reported recently for the Cyp7a promoter (20), we co-transfected LXR/RXR and LRH in 293 cells. For both LXR␣ and LXR␤, this resulted in a significant increase in sterol-dependent activation of promoter activity (Fig. 1). LRH alone increased basal promoter activity (about 3-fold) but did not provide a sterol response (Fig. 1). In contrast to these findings with the CETP promoter, the ABC1 promoter was well activated by sterols in the presence of LXR/ RXR, and the response was not further increased by co-expression of LRH (Fig. 2), showing that the LRH effect is promoter-specific.
Mutation of the LXR-binding site (DR4) in the CETP promoter abolished both the sterol-dependent increase in promoter activity, and the amplification of this effect by LRH (Fig.  3). However, this mutation did not affect the increase in basal promoter activity attributable to LRH.
LRH Binding to the CETP Promoter Is Required for Potentiation of Sterol Induction: Mapping the LRH-binding Site-The increase in basal activity by LRH, independent of the DR4 mutation (Fig. 3), suggests that LRH binds to a site distinct from the LXR-binding region. Consistent with this idea, mu-FIG. 1. Potentiation of LXR function by LRH-1. HEK293 cells were co-transfected with reporter construct NFR-luc in which the luciferase reporter gene is controlled by the proximal CETP promoter (from 1 to Ϫ570) and equivalent amounts (100 ng) of either control plasmid or LXR␣, LXR␤, RXR␣, and/or LRH-1. The transfected cells were cultured in LPDS ϩ vehicle or LPDS ϩ 2 g/ml 22(R)-hydroxycholesterol for 24 h, and luciferase activities were measured. Relative luciferase activity, which was obtained by normalizing to NFR-Luc basal activity, is shown. Results (mean Ϯ S.D.) are shown for three independent experiments, carried out in duplicate. The fold induction of reporter gene activity by sterol is shown above the filled bars.
tagenesis of the zinc finger DNA binding domain of LRH (C107S, C200S, and S128G) abolished both the increase in basal activity mediated by LRH and also the potentiation of sterol induction mediated through LXR (data not shown). There are several potential LRH-binding sites (YCA(A/ G)GGYCR) in the proximal CETP promoter (Fig. 4). Deletional mutagenesis localized the basal LRH response to a region between Ϫ60 to Ϫ88 (Fig. 5). Moreover, whereas a large deletion of the region between Ϫ90 to Ϫ370 did not affect the LRH potentiation of sterol induction by LXR (Supplemental Fig. 1), deletion of the Ϫ60 to Ϫ88 region abrogated this response (Fig.  6a). These findings indicate that the same LRH-binding region (Ϫ75 to Ϫ83) is involved in the induction of basal activity and in the amplification of the LXR response to sterols. To verify further this conclusion, the LRH site was mutated by changing nucleotides CCC to GCA (Fig. 4), which markedly reduced the impact of LRH on sterol induction of the CETP promoter activity (Fig. 6b).
We next showed binding of LRH to this site. Gel shift analysis showed specific binding of LRH to a DNA fragment con-taining the Ϫ75 to Ϫ83 region (LRHBS) of the CETP promoter (Fig. 7, arrow). In vitro translated, Myc epitope-tagged LRHs were used in the gel shift experiments. The Myc-tagged protein showed full potentiation of the sterol induction of the CETP promoter (not shown) and bound specifically to the CETP LRHBS (Fig. 7, a and b, lane 2, arrow). This specific band was reduced in competition assays, using either the CETP LRHBS or the Cyp7a LRHRE (20) (Fig. 7a, lanes 3 and 5), whereas the mutant CETP LRHBS was unable to compete for binding (Fig.  7a, lane 4). This same mutation abolished the functional effects of LRH (Fig. 6b). A factor from reticulocyte lysates also bound specifically to the CETP LRHBS (Fig. 7a, lane 1, indicated by asterisk). However, this gel shift band was not competed by the Cyp7a LRHRE. An antibody to the Myc epitope, but not a control antibody (anti-actin), specifically reduced the intensity of the shifted band resulting from the binding of Myc-LRH to the CETP LRHBS (Fig. 7b, lanes 3 and 4, arrow).
SHP Abolishes the Sterol Potentiation Effect of LRH-1 on LXR in the CETP Promoter-LRH-1 has been shown to interact with and be repressed by small heterodimer partner (SHP), an orphan nuclear receptor (20). Since LRH-1 also regulates CETP gene expression, we analyzed the effect of SHP on the expression of the CETP gene by co-transfecting SHP with RXR, LXR, and LRH in 293 cells. As shown in Fig. 8, LRH specifically increased the response to sterols but not the RXR ligand, 9-cisretinoic acid. SHP abolished the LXR-dependent induction of CETP promoter activity by sterols. Increasing the amounts of RXR␣, while other factors were held constant, did not eliminate the repression by SHP (data not shown), suggesting that the abolition of sterol induction by SHP is not mediated through RXR. These results are consistent with previous studies showing a direct interaction between SHP and LRH (20) and suggest that SHP might bind to a region of LRH involved in interactions with LXR. The 3.4-kb full-length CETP promoter was regulated similarly by LRH and SHP (Supplemental Fig. 2). The only difference was that expression of LXR␤/RXR␣ and LXR␣/RXR␣ led to about 2.5-and 3-fold sterol induction, respectively, in the 3.4-kb promoter, a more robust response than seen with the shorter promoter fragment (Fig. 1). LRH further increased the induction fold to about 4-fold for LXR␤/RXR␣ and 6-fold for LXR␣/RXR␣.
To see if SHP also represses the basal transactivation activity of LRH-1 on the CETP promoter, increasing amounts of SHP plasmid were co-transfected with LRH. However, the LRH-1-dependent increase of the reporter activity was not repressed by expression of SHP (Fig. 9). Therefore, the inhibi-tory effect of SHP specifically abolished the potentiation effect of LRH-1, while not affecting the increase in basal promoter activity induced by LRH-1 ( Fig. 8 and 9). These results indicate that SHP only represses the LRH-1 competence effect on the CETP promoter in a sterol-dependent manner.

Regulation of CETP Gene by Bile Salt in CETP Transgenic
Mice-It has been proposed that the repression of LRH-1 activity by SHP may mediate the bile acid repression of Cyp7a (20,27). Since CETP expression seems to be regulated by LRH-1 and SHP in a similar fashion to Cyp7a, we analyzed the regulation of CETP gene expression by bile acid in CETP transgenic mice. CETP transgenic mice expressing CETP controlled A specific shifted band resulting from the binding of a factor from reticulocyte lysates was also obtained (lane 1, indicated by the asterisk). b, Myc epitope antibody 9E10 specifically blocks the binding of Myc-LRH to CETP LRHBS (lanes 3 and 4). by its native promoter (Ϫ3.4 kb) (25) were fed a chow diet with or without 1% cholic acid for 5 days. Northern blot analysis was carried out to analyze the expression of several genes that are regulated by bile acids (Fig. 10). As expected, Cyp7a expression was completely repressed by dietary cholic acid. Cyp8b (encoding 12␣-Hydroxylase), also regulated by LRH-1 (28), was repressed by cholic acid feeding as reported (29). The cholic acid-containing diet also led to an increase in SHP mRNA and a decrease in Cyp27 expression (20,29). For all of these genes a similar response was seen in male and female mice. Intriguingly, however, the response of CETP mRNA was sexually dimorphic. In male mice, CETP mRNA was decreased upon cholic acid feeding, whereas in female mice the bile salt diet led to a 2-fold induction of CETP mRNA.
Microarray Analysis of Hepatic mRNA from Mice Fed a 1% Cholic Acid Diet-In order to gain further insights into the effects of bile acids on gene expression, we carried out microarray analysis using mRNA from male and female mice. The microarrays contained about 1,200 cDNAs and ESTs, enriched for genes expressed in liver. Also, a substantial number of transcription factors expressed in liver were represented. About 2.5% of genes were induced, and 1.5% genes were repressed by dietary cholic acid, with parallel effects in both sexes (Supplemental Table 1). We and others (30) have found that the large majority of genes with altered expression on the microarrays show similar or larger changes in expression when assessed by Northern analysis. As an example, SR-BI, a high density lipoprotein receptor (31), showed 1.7-or 2.6-fold induction by the diet, in male and female mice, respectively. Northern blot analysis confirmed the results (Fig. 10).
Among the 1200 genes analyzed, 11 genes (0.9%) displayed sexually dimorphic changes in mRNA expression, in response to the bile acid-enriched diet ( Table I). The most common response was a higher induction in male than in female mice. The opposite pattern of response, i.e. female greater than male, was only demonstrated by the human CETP gene and by one EST. Notable among the dimorphically induced genes was a TNF␣-induced protein, suggesting a possible differential effect of TNF␣ on target genes. Hepatic TNF␣ is induced by the bile acid-enriched diet, and both CETP and Cyp7a are repressed by TNF␣ (32)(33)(34). DISCUSSION We have shown that the orphan nuclear receptor LRH-1 binds and transactivates the human CETP promoter. In 293 cells, the native CETP promoter showed modest sterol induction in the presence of LXR␣/RXR␣ or LXR␤/RXR␣ ( Fig. 1 and  Supplemental Fig. 2). The expression of LRH-1 enhanced the response of LXR/RXR to sterols. The principal tissue expressing LRH-1 is the liver, which also is a major site where CETP is expressed (4). Even though CETP is expressed in peripheral tissues, the sterol regulation in peripheral tissue is less pronounced than in liver (4). A requirement for LRH-1 in the sterol induction mediated by LXR might confer tissue specificity for the induction of genes by sterols, since the ubiquitously expressed LXR␤ could otherwise mediate sterol induction of CETP gene. Additional physiological significance might be that LRH-1 abolishes the requirement for RXR ligands, such as 9-cis-retinoic acid or docosahexanoic acid (35), in order to obtain significant LXR-mediated sterol induction of gene expression. Thus, in response to endogenous hypercholesterolemia, 24(S),25-epoxycholesterol, which is synthesized from a shunt pathway (36) and has been shown to act as an LXR ligand (6), may bind and activate LXR/RXR without the requirement for additional RXR ligands. This effect is clearly shown in Fig. 8 and Supplemental Fig. 2. The presence of LRH increased the induction of CETP by sterol but had no additional incremental effect as a result of the addition of 9-cis-retinoic acid.
The effect of LRH-1 was strikingly specific for the LXR-dependent sterol induction of CETP expression. The mechanism of the LRH-1 potentiation effect required the binding of LRH-1 to a site (Ϫ75 to Ϫ83) (LRHBS) on the CETP promoter, and the DNA binding ability is required for the potentiation function of LRH-1, since mutation of the LRH-1 DNA-binding zinc finger motif abolished this effect (not shown). The activity of LRH-1 on the CETP promoter might enhance ligand binding to LXR or cooperatively work with LXR/RXR to recruit ligand-dependent co-activators. The co-activator SRC-1 (nuclear receptor co-activator) has been shown to interact with LXR␤/RXR upon binding of 22(R)-hydroxycholesterol and/or 9-cis-retinoic acid (37). It is possible that sterol-dependent binding of co-activator is enhanced either by the LRH/LXR activation or by retinoid binding to RXR. This would explain why in the presence of sterols the expression of LRH or activation by 9-cis-retinoic acid achieved similar promoter activity (Fig. 8).
Recently, considerable evidence has been obtained to support the idea that LXR/RXR coordinates the regulation of genes involved in reverse cholesterol transport (9), such as ABCA1 Mice expressing the human CETP transgene, controlled by its native promoter, were fed with either rodent chow diet or chow diet containing 1% cholic acid for 5 days. Liver poly(A) ϩ mRNA was prepared from total RNA pooled from four mice per diet per sex. 2 g of poly(A) ϩ RNA was loaded, and gene expression was analyzed by Northern blot, using the cDNA probes shown. G3PDH, glyceraldehyde-3-phosphate dehydrogenase. (10,38), CETP (9), Cyp7a (7,8), and apoE (39). It appears that ABCA1 does not need LRH-1 as a competence factor for the regulation by LXR/RXR, since the LXR/RXR-binding site is sufficient to mediate sterol induction in cultured cells in the context of the native promoter (10). ApoE and ABC1 are involved in the initial step of cholesterol efflux from macrophage foam cells, resulting in enrichment of HDL particles with cholesterol and cholesteryl ester. CETP transfers cholesteryl esters between lipoproteins in the bloodstream, and thereby facilitates the transport of HDL cholesterol ester into the liver. Cyp7a converts cholesterol into bile acids in the liver, leading to excretion from the body.
The mechanism of regulation of CETP by LXR/RXR and LRH is strikingly similar to the regulation of Cyp7a by these factors. The sterol induction of Cyp7a has also been shown to require LRH-1 as a potentiation or competence factor (20). The competence effect requires binding of LRH-1 to both CETP and Cyp7a promoters. Similar to its effects on the human CETP promoter, LRH-1 (CPF) transactivates the human CYP7A promoter in 293 cells (19). However, little basal transactivation of the rat Cyp7a promoter by LRH-1 was seen in 293 cells (20). In the rat Cyp7a promoter, LRH-1 not only increased the induction by sterol but also the additive or synergistic induction by both sterols and 9-cis-retinoic acid. However, in the CETP promoter, LRH-1 only increased the induction by sterol. These results imply that the competence mechanism of LRH-1 on CETP and Cyp7a might be slightly different.
Cyp7a is not only regulated by hydroxysterols, but also is under feedback control by bile salts in the entero-hepatic circulation. The nuclear receptor FXR was recently identified as a receptor for the bile acid, chenodeoxycholic acid (40 -42). Binding of FXR to the promoters of several bile salt-induced genes, such as ileal bile acid-binding protein and phospholipid transfer protein, resulted in bile acid-dependent promoter activation (40,43). It has been suggested that FXR/RXR may also mediate the bile acid repression of Cyp7a; however, since FXR/RXR does not bind the bile acid response element in the Cyp7a promoter, an indirect mechanism is implied (44). Disruption of FXR has provided strong evidence for a role of FXR in up-regulating expression of genes such as the ileal bile acid transporter and the basal bile salt exporter (29). FXR knock-out mice also failed to repress hepatic Cyp7a mRNA levels in response to a cholic acid-containing diet (29), suggesting a role for this factor in bile salt-mediated gene repression.
Recently, it has been proposed that FXR induces SHP, which then interacts with LRH-1, preventing its activation of the Cyp7a promoter (20,27). SHP, a transcriptional repressor, was induced by bile acid treatment as a result of activity of FXR/ RXR on the SHP promoter; furthermore, SHP directly interacts with LRH-1 and represses LRH-1 activity in cell culture (20,27). The abolition of LRH facilitation of the LXR-mediated sterol response by SHP was strikingly similar in the CETP and Cyp7a promoters (this study and (20)). A weak or absent effect of SHP on basal promoter activity in the presence of LRH also appears to be similar for CETP and Cyp7a (20). However, in previous studies (4) we observed that a high fat, high cholesterol diet containing bile salts induced hepatic CETP mRNA expression, whereas Cyp7a mRNA is repressed in response to the same diet. In the present study a somewhat different response of CETP and Cyp7a genes to a chow diet supplemented with 1% cholic acid was seen. Whereas Cyp7a mRNA was repressed by the bile salt-supplemented diet in both sexes, CETP mRNA was repressed in males but induced in female mice.
In an attempt to understand these different responses, we characterized the responses of about 1200 different cDNAs and ESTs to the bile salt diet. The majority of differentially expressed genes were either induced or repressed in parallel fashion in both sexes. However, 11 genes gave a sexually dimorphic response, with the majority showing an induction in males but not in females. Thus, the CETP response is unusual and is unlikely to be mediated by a simple mechanism, such as induction of a single transcription factor in one sex but not the other. The array results provide an intriguing hint for how the differential response could be mediated. Thus, the TNF␣-induced protein 2 was increased in males and decreased in females, suggesting that there could be a dimorphic effect of TNF␣ signaling in male versus female mice. Hepatic TNF␣ expression is induced by the cholic acid diet (it is probably made in Kupffer cells) (33), and TNF␣ has been shown to repress both CETP (34) and Cyp7a (32). If the effect of TNF␣ is larger in male mice, this could explain why CETP is repressed in this sex. One has to then hypothesize an additional factor that up-regulates CETP in response to bile acids and only becomes apparent in females because of a lessened TNF␣ effect. The related gene PLTP is up-regulated by the bile salt diet, through an FXR mechanism (43), and it is conceivable that CETP could be similarly regulated. Induction of TNF␣ and interleukin-1 by the bile salt diet has also been suggested as a possible mechanism of repression of Cyp7a (33,45), but this response was not dimorphic. There could be additional mechanisms of repression of Cyp7a, such as that proposed involving FXR, SHP, and LRH, affecting both sexes.
In summary, we have shown that the competence factor, LRH, enhances the LXR-mediated sterol response of the CETP promoter, probably contributes to the moderate tissue specificity of this response (46), and provides a way for the CETP promoter to respond to sterols, independent of RXR agonists. Both of these effects, i.e. tissue-specific responses and responses to LXR/RXR that are independent of RXR ligands, may be general properties of competence factors in transcriptional responses. The interaction of LXR and LRH in the sterol-dependent induction of the CETP and Cyp7a promoters, and the repression of this effect by SHP, is strikingly similar for CETP and Cyp7a, supporting the idea of coordinate regulation of these two genes in the liver (9). However, the divergent responses of these and other genes to diets containing bile salts highlights the complexity of the in vivo response to this challenge.