Liver receptor homologue-1 mediates species- and cell line-specific bile acid dependent negative feedback regulation of the apical sodium-dependent bile acid transporter

Intestinal reclamation of bile salts is mediated in large part by the apical sodium-dependent bile acid transporter (ASBT). The bile acid responsiveness of ASBT is controversial. Bile acid feeding in mice results in decreased expression of ASBT protein and mRNA. Mouse but not rat ASBT promoter activity was repressed in Caco-2, but not IEC-6, cells by chenodeoxycholic acid. A potential liver receptor homologue-1 (LRH-1) cis-acting element was identified in the bile acid-responsive region of the mouse but not rat promoter. The mouse, but not rat, promoter was activated by LRH-1, and this correlated with nuclear protein binding to the mouse but not rat LRH-1 element. The short heterodimer partner diminished the activity of the mouse promoter and could partially offset its activation by LRH-1. Interconversion of the potential LRH-1 cis-elements between the mouse and rat ASBT promoters was associated with an interconversion of LRH-1 and bile acid responsiveness. LRH-1 protein was found in Caco-2 cells and mouse ileum, but not IEC-6 cells or rat ileum. Bile acid response was mediated by the farnesoid X receptor, as shown by the fact that overexpression of a dominant-negative farnesoid X-receptor eliminated the bile acid mediated down-regulation of ASBT. In addition, ASBT expression in farnesoid X receptor null mice was unresponsive to bile acid feeding. In summary cell line- and species-specific negative feedback regulation of ASBT by bile acids is mediated by farnesoid X receptor via small heterodimer partner-dependent repression of LRH-1 activation of the ASBT promoter.


Introduction
Reclamation of bile salts by the intestine is primarily mediated by the apical sodium dependent bile acid transporter (ASBT) located in the terminal ileum (1,2). The bile acid responsiveness of ASBT expression, which is of fundamental importance to both cholesterol metabolism and cholestatic liver disease, is a matter of on-going controversy (3). Conflicting results have been observed in a number of laboratories, where experimental methodologies and animal species have been variable (4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15).
The molecular mechanisms of bile acid responsiveness have been elucidated in recent elegant investigations from a number of laboratories (16). In general two transcriptional activation of the ileal lipid binding protein (ILBP) and the canalicular bile salt excretory pump promoters by bile acids (17)(18)(19)(20)(21). Negative feedback regulation is more complex and involves an indirect effect. The first step is dependent upon activation of the expression of the small heterodimer partner (SHP) by a complex of bile acids and the farnesoid X-receptor. SHP then mediates down-regulation of target genes by inhibiting the activity of endogenously expressed positive trans-acting factors.
The aim of the current study was to apply these new paradigms of bile acid responsiveness to the unanswered questions revolving around the regulation of the apical sodium-dependent bile acid transporter. Analysis in our own laboratory has indicated that in the rat, ileal ASBT does not respond to alterations in lumenal bile acid concentrations (10). In contrast mouse studies conducted by Torchia and confirmed in our own laboratory (reported herein) reveal that ASBT is under negative feedback regulation (8,12). Recent cloning and characterization of the rat (25)

Animal Studies
All animal studies were performed under protocols approved by the animal care and use committees of the Mount Sinai School of Medicine, Rockefeller University and the National Institutes of Health. The effect of bile acids on the expression of ASBT in the mouse ileum was determined in male C57Bl/6J mice that were fed a diet supplemented with 0.5% cholic acid or 0.5% taurocholic acid for a period of seven days.
Ileal protein homogenates and brush border membrane vesicles were prepared from the distal 1/3 of mouse ileum using methods that have been previously developed for the analysis of fetal and neonatal rat ileum (26). Brush border membrane enrichment was assessed by measuring alkaline phosphatase in homogenates and membranes.
Quantitative Western blot analysis of the apical sodium-dependent bile acid transporter and the ileal lipid binding protein was performed as previously described (10).
Antibodies against ILBP were generously provided by Dr. Michael Crossman (Children's Medical Center, Cincinnati, OH). Beta actin was used as a loading control (Sigma Chemicals, St. Louis, MO). Northern blot analysis was used to assess changes in steady-state mRNA levels of ASBT and ILBP as previously described (10). Twenty micrograms of total RNA was analyzed from the distal one third of the small intestine from 5 separate samples in each group (chow fed and cholic acid fed). Signal intensity for 28S rRNA was used as a loading control. Cloning and characterization of the mouse ASBT promoter A partial murine ASBT cDNA was obtained by PCR using oligonucleotide primers corresponding to amino acid sequences conserved between the ileal and liver sodium/bile acid cotransporters; the sense and antisense oligonucleotide primers corresponded to amino acids 75-81 and 261-266, respectively, of the human ASBT.
Oligonucleotide primers were then designed based on this mouse sequence and the predicted exon 1 region and used to screen a mouse (strain Sv129/OLA) bacteriophage P1 library (Genome Systems; St. Louis, MO). Two mouse genomic clones (12605, 12606) containing slc10A2 sequences were obtained and mapped. The structure of the mouse slc10a2 gene was elucidated using a combination of subcloning, restriction enzyme analysis, Southern blotting, long PCR amplification, and DNA sequencing. The nucleotide sequences have been submitted to the GenBank with accession number AF266724. Specific elements of the ASBT promoter were cloned and sequenced from several representative mouse (C57Bl/6J, SJL/J and BALB/cByJ, Jackson Laboratories, Bar Harbor, ME) and rat (Long evans, SHR, Fischer-344, Taconic Laboratories, Germantown, NY) strains. Genomic DNA from each of the species was amplified with by guest on July 9, 2020 http://www.jbc.org/ Downloaded from Negative Feedback Regulation of the Ileal Bile Acid Transporter 7 oligonucleotide primers extending from -385 to -95 (sense 5'-GCCCTAGAAGTCTGTG-3" antisense 5'-GCTGGGAATAATTTTAG-3'). PCR products were then sequenced directly.
Western analysis of ASBT expression in FXR null mice FXR null mice or control litter-mates were fed a diet of 1% cholic acid or normal chow for a period of five days as previously reported (30). ASBT and ILBP protein expression were measured in the distal ileum as described above.

Statistical Analysis
Statistical analysis was performed using InStat software (GraphPad Software, Inc. San Diego, CA). Means were compared by student's t-test or the Tukey-Kramer multiple comparisons test. Unless otherwise stated all values are mean + standard deviation.

Results
Bile acid response in the mouse Bile acid feeding for seven days was tolerated without adverse effect. Ileal Cloning and characterization of the mouse ASBT promoter The murine slc10A2 gene is organized in 6 exons spanning approximately 24 kb of DNA sequence and its structure is identical to the human and rat SLC10A2 genes (25,31,32). Exon/intron boundaries conform to consensus motifs; the introns all began with a GT at the 5' splice donor sites and ended with an AG at the 3'-splice acceptor sites. The first exon encompasses the 5'-untranslated region and encodes amino acids The truncation studies above indicated that bile acid responsiveness was localized in the region of the mouse promoter P2 but not P3. Sequence analysis of the region of the mouse ASBT promoter between nucleotides -378 and -208, that is located upstream of P3 and is part of P2, revealed a potential LRH-1 binding site between mouse base pairs -336 and -327 ( Figure 1). The same potential LRH-1 element was observed in all strains of mice that were examined. Two nucleotide differences were noted in the mouse and rat sequences in this region. Sequence analysis in this region was identical in all rat strains that were examined. In light of the role that LRH-1 plays in negative feedback regulation of cholesterol 7-alpha hydroxylase, initial efforts were directed at analysis of this element in the mouse ASBT promoter. The mouse P2 promoter activity was markedly enhanced by co-transfection of an LRH-1 expression construct ( Figure 5). The enhanced activity could be partially off-set by simultaneous co-transfection with the short heterodimer partner (SHP).
Protein binding by the mouse but not rat potential LRH-1 cis-acting element was seen after electrophoretic mobility shift assay using nuclear extracts from Caco-2 cells  Figure 6C). SHP expression was increased four-fold after treatment with CDCA ( Figure 6C). LRH-1 and FXR were unchanged.
Site directed mutagenesis was performed in order to convert the mouse LRH-1 site into the rat site (P2 -> P2µ) and vice versa. Basal activity of the mouse P2 promoter was higher than the rat and the relationship was reversed after the LRH binding sites were interchanged (Figure 7). The mouse but not rat promoter could be stimulated by co-transfection of LRH-1. The response to LRH-1 was reversed after inter-conversion of the LRH-1 binding sites between the mouse and rat promoters.
Mutation of the rat LRH-1 site led to binding of LRH as demonstrated by gel shift analysis ( Figure 6A, lanes 10 -12). Similarly, mutation of the mouse LRH-1 site lead to loss of LRH-1 binding ( Figure 6A, lanes 4 -6). The mouse but not rat promoter could be inhibited by the addition of 100 µM chenodeoxycholic acid and the responsiveness to bile acids was also reversed after interchange of the mouse and rat LRH-1 element sequences ( Figure 7). The bile acid responsiveness of the mouse P2 promoter was not  Figure 11A and 11B). As expected cholic acid feeding in wild type mice lead to reduction in ASBT protein expression (12,600 + 6,000, n=3, p<.01). In contrast cholic acid feeding in the FXR null mice resulted in no significant change relative to chow fed FXR null mice (61,500 + 3,200, n=3, p<.01). Overall ASBT expression was nearly five-fold higher in cholic acid fed FXR null mice relative to wild type littermates. ILBP expression was markedly reduced in FXR null mice (groups 2 and 4, Figure 11A and B) independent of feeding status, indicating that FXR was essential for basal expression of ILBP. As expected cholic acid feeding resulted in

Discussion
Transcriptional activation of the apical sodium-dependent bile acid transporter is mediated by at least three different trans-acting factors. Hepatocyte nuclear factor-1_ trans-activates the ASBT promoter and the corresponding null mice have markedly reduced ASBT expression (33). The AP-1 protein c-Jun trans-activates ASBT and expression of c-Jun correlates with the developmental-stage and region-specific expression of ASBT (25). The current studies indicate that the liver receptor homologue-1 protein is also a transcriptional activator of ASBT. LRH-1 does not appear to be essential for basal activity, but plays a crucial role in mediating bile acid responsiveness. This is supported by the enhanced activity of the mouse ASBT promoter versus that of the rat promoter, the former of which harbors a functional LRH-1 cis-acting element.
The ASBT gene is under negative feedback regulation by bile acids in the mouse ileum. Our studies of ASBT protein using mouse ileal BBMV and mRNA confirmed the findings of a similar study of ASBT mRNA (8). In both studies ASBT expression was repressed after bile acid feeding. We have also observed the negative feedback regulation of ASBT in cholesterol 7-alpha hydroxylase knock-out mice, which have markedly reduced bile acid synthesis and secretion (34). The knock-out mice have enhanced expression of ASBT (12). In both the bile acid fed and cholesterol 7-alpha hydroxylase knock-out mice, regulated changes in the ileal lipid binding protein are opposite of ASBT. This is expected as the ileal lipid binding protein is positively regulated by bile acids. by guest on July 9, 2020 http://www.jbc.org/

Downloaded from
Bile acid responsiveness of the ASBT gene is mediated in an indirect fashion by farnesoid X-receptor dependent activation of the short heterodimer partner and subsequent inhibition of LRH-1 activity. This paradigm has been previously shown to be relevant for a similar negative feedback regulation of cholesterol 7-alpha hydroxylase (22,23). The cell-line and species-specific nature of the negative feedback regulation of ASBT can now be explained by the presence or absence of both the LRH-1 cis-and trans-acting elements. The mouse but not rat ASBT promoter has a functional LRH-1 cis-acting element. In addition, mouse but not rat ileum expresses the LRH-1 protein.
Similar differences in LRH-1 protein expression are found in Caco-2 and IEC-6 cells. It is fascinating that there has been a divergent and coordinate evolution of both the LRH-1 cis-and trans-acting elements between these two relatively closely related rodent species. The implications of the differences are profound in regard to the regulation of bile acid homeostasis. It is tempting to speculate that this difference may have evolved in conjunction with the presence (mouse) or absence (rat) of the gallbladder.
Preliminary sequence analysis of both the rabbit and human promoters does not give a clear indication as to whether these promoters will be bile acid responsive, and thus functional assays will be necessary.
The physiologic relevance of the bile acid response is apparent in the FXR null mice (30). The critical role of FXR in mediating the bile acid response of the ASBT promoter is evident from our transient transfection studies using the dominant negative        LRH-1) enhanced mouse P2, but not rat P2 transcriptional activity. Rat P2 failed to respond to addition of CDCA or introduction of exogenous LRH-1. To further study the role of LRH-1 in ASBT regulation, the cis element in mP2 and rP2 were point mutated as described in "Experimental procedures", resulting in loss and gain of the LRH-1 consensus sequence within mouse and rat P2 (P2µ) (Figure 2). Luciferase assays showed that not only the basal promoter activities interchanged between the two P2 mutants (P2µ) compared to the normal counterparts (P2), but also reactions to CDCA (P2µ + CDCA) and LRH-1 (P2µ + LRH) shifted to rat P2µ, and mouse P2µ lost these characteristics.    expression. This indicates that the FXR mutant acted as a dominant negative factor in the CDCA initiated pathway leading to the down-regulation of ASBT promoter activity.
These results also suggest that bile acid is required for the FXR inhibitory effect of ASBT expression.