Liver receptor homolog-1 is critical for adequate up-regulation of Cyp7a1 gene transcription and bile salt synthesis during bile salt sequestration

Authors

  • Carolien Out,

    Corresponding author
    1. Department of Pediatrics, Center for Liver, Digestive, and Metabolic Diseases, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands
    • Laboratory of Pediatrics, Center for Liver, Digestive and Metabolic Diseases, University Medical Center Groningen, Hanzeplein 1, 9713 EZ Groningen, The Netherlands
    Search for more papers by this author
    • These authors contributed equally to this work.

    • fax: 0031(0)50-3611746

  • Jurre Hageman,

    1. Department of Pediatrics, Center for Liver, Digestive, and Metabolic Diseases, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands
    Search for more papers by this author
    • These authors contributed equally to this work.

  • Vincent W. Bloks,

    1. Department of Pediatrics, Center for Liver, Digestive, and Metabolic Diseases, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands
    Search for more papers by this author
  • Han Gerrits,

    1. Merck Sharp & Dohme Research Laboratories, Oss, The Netherlands
    Search for more papers by this author
  • Maarten D. Sollewijn Gelpke,

    1. Merck Sharp & Dohme Research Laboratories, Oss, The Netherlands
    Search for more papers by this author
  • Trijnie Bos,

    1. Department of Pediatrics, Center for Liver, Digestive, and Metabolic Diseases, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands
    Search for more papers by this author
  • Rick Havinga,

    1. Department of Pediatrics, Center for Liver, Digestive, and Metabolic Diseases, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands
    Search for more papers by this author
  • Martin J. Smit,

    1. Merck Sharp & Dohme Research Laboratories, Oss, The Netherlands
    Search for more papers by this author
  • Folkert Kuipers,

    1. Department of Pediatrics, Center for Liver, Digestive, and Metabolic Diseases, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands
    2. Department of Laboratory Medicine, Center for Liver, Digestive, and Metabolic Diseases, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands
    Search for more papers by this author
  • Albert K. Groen

    1. Department of Pediatrics, Center for Liver, Digestive, and Metabolic Diseases, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands
    2. Department of Laboratory Medicine, Center for Liver, Digestive, and Metabolic Diseases, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands
    Search for more papers by this author

  • Potential conflict of interest: Nothing to report.

  • Supported within the framework of Top Institute Pharma, The Netherlands, project T2-110.

Abstract

Liver receptor homolog-1 (LRH-1) is a nuclear receptor that controls a variety of metabolic pathways. In cultured cells, LRH-1 induces the expression of CYP7A1 and CYP8B1, key enzymes in bile salt synthesis. However, hepatic Cyp7a1 mRNA levels were not reduced upon hepatocyte-specific Lrh-1 deletion in mice. The reason for this apparent paradox has remained elusive. We describe a novel conditional whole-body Lrh-1 knockdown (LRH-1-KD) mouse model to evaluate the dependency of bile salt synthesis and composition on LRH-1. Surprisingly, Cyp7a1 expression was increased rather than decreased under chow-fed conditions in LRH-1-KD mice. This coincided with a significant reduction in expression of intestinal Fgf15, a suppressor of Cyp7a1 expression, and a 58% increase in bile salt synthesis. However, when fecal bile salt loss was stimulated by feeding the bile salt sequestrant colesevelam, Cyp7a1 expression was up-regulated in wildtype mice but not in LRH-1-KD mice (+593% in wildtype versus +9% in LRH-1-KD). This translated into an increase in bile salt synthesis of +272% in wildtype versus +21% in LRH-1-KD mice. Conclusion: Our data provide mechanistic insight into a missing link in the maintenance of bile salt homeostasis during enhanced fecal loss and support the view that LRH-1 controls Cyp7a1 expression from two distinct sites, i.e., liver and ileum, in the enterohepatic circulation. (HEPATOLOGY 2011;)

Bile salts are synthesized from cholesterol exclusively in the liver by a complex multienzyme process. Crucial steps in the synthesis pathway comprise the addition of one or two hydroxyl groups to the sterol nucleus and the oxidative cleavage of the side chain of cholesterol, resulting in a highly amphipathic class of bile salt molecules. Bile salts are potent surfactants that solubilize phosphatidylcholine and cholesterol in bile and promote lipid absorption in the small intestine. Next to being the primary driving force for hepatic bile formation, the role in intestinal lipid digestion has long been thought to be the most important function of bile salts.1 The landmark discovery of bile salts as endogenous ligands for the nuclear hormone receptor farnesoid X-receptor (FXR) and, more recently, for the G-protein-coupled receptor TGR5 has completely transformed the field of bile salt research. In addition to mediating the feedback control of bile salt synthesis, FXR influences many pathways involved in lipid metabolism and has recently also been implicated in control glucose metabolism.2 TGR5 seems particularly important in regulating energy metabolism.3

Accordingly, it is essential to fully understand the factors that regulate synthesis of the various types of bile salts. Liver receptor homolog-1 (LRH-1) has been implicated herein, but its exact role has remained elusive so far. LRH-1 belongs to the NR5A family of nuclear receptors together with steroidogenic factor-1 and the Drosophila melanogaster ortholog Fushi tarazu factor-1.4-6 In contrast to most other nuclear receptors, members of the NR5A subfamily bind DNA as monomers.7, 8 LRH-1 is essential for embryogenesis, as targeted gene disruption results in early embryonic lethality.9 In the adult mouse, LRH-1 is expressed predominantly in the ovaries, the exocrine pancreas, and the organs that constitute the enterohepatic axis, i.e., liver and small intestine.9-11 In the small intestine, LRH-1 has been shown to stimulate cell proliferation in intestinal crypts12 and to regulate extra-adrenal glucocorticoid synthesis13 that protects against inflammatory bowel disease.14 In line with these antiinflammatory effects, hepatic LRH-1 acts as a potent suppressor of the acute phase response.15, 16 Functional LRH-1 binding sites have been found within the promoter regions of several genes implicated in lipid metabolism and transport such as Abcg5/Abcg8, APOA1, and SR-B1.17-19

LRH-1 has been proposed to function as an important transcription factor in control of bile salt synthesis. The first and rate-controlling step in the classic pathway of bile acid synthesis is catalyzed by the enzyme cholesterol 7α-hydroxylase (CYP7A1).20 Subsequently 7α-hydroxycholesterol is converted into cholic acid by 12α-hydroxylase (CYP8B1), which determines the ratio in which the primary bile salt species cholate (3α,7α,12α-trihydroxy-5β-cholate) over chenodeoxycholate (3α,7α-dihydroxy-5β-cholate) are being produced.21

Hepatic bile salt synthesis is tightly regulated by complex feedback mechanisms involving the consecutive and/or simultaneous actions of a number of hepatic nuclear receptors and transcription factors such as LXR, SREBPs, and HNF4.3, 22-25 In addition, LRH-1 binding sites have been identified in the proximal promoter parts of CYP7A1 and CYP8B1.8, 26 Data from cell studies showed that LRH-1 is able to induce the expression of CYP7A18, 22, 23 and CYP8B1.26 Therefore, LRH-1 has been proposed to function in feedback regulation of CYP7A1 expression as part of the FXR-SHP-LRH-1 cascade, in which bile acids can inhibit their own synthesis. In this cascade bile salt-activated hepatic FXR induces the expression of small heterodimer partner (SHP) that functions as a potent repressor of hepatic LRH-1 activity,27 which then results in less activation of CYP7A1 by LRH-1. In addition, upon activation of intestinal FXR, the endocrine growth factor FGF15 is produced and transported to the liver, where it binds its receptor FGFR4 and represses CYP7A1 expression in the liver.28, 29 Thus, bile salt synthesis is under negative feedback control from at least two distinct sites in the enterohepatic system.

Although the results from the initial cell studies8, 22, 23 were consistent with respect to the regulation of Cyp7a1 by LRH-1, they were in apparent contrast with those of subsequent in vivo studies using conditional Lrh-1 deletion.30, 31 Two independent studies showed that Cyp7a1 messenger RNA (mRNA) levels and protein activity were not reduced upon hepatocyte-specific Lrh-1 knockout, whereas, as expected, Cyp8b1 levels were.30, 31 These studies hence suggest that LRH-1 regulates composition and thus physicochemical properties of the bile salt pool but does not control bile salt synthesis rate in mice. Furthermore, heterozygous Lrh-1 knockout mice exhibited 5-7-fold higher Cyp7a1 expression levels and increased total bile acid pool sizes.32 Therefore, the proposed role of LRH-1 in the FXR-SHP-LRH-1 cascade, regulating Cyp7a1 expression, remained uncertain.

It has been speculated that the reason for the discrepancy between in vitro and in vivo approaches could be a redundant factor that maintains Cyp7a1 transcription in mice in the absence of LRH-1.31 As previous in vivo experiments were all performed under normal physiological feeding conditions, it is at this stage unclear whether LRH-1 functions as an important transcriptional regulator for Cyp7a1 expression under conditions in which bile salt synthesis rates must be enhanced to maintain homeostasis, such as during increased fecal bile salt loss.

In this study we describe a novel conditional systemic LRH-1 knockdown mouse model (LRH-1-KD) to evaluate the dependency of bile salt synthesis on LRH-1 under normal, chow-fed conditions, and under conditions of high fecal bile salt loss. Our data show that under physiological (low flux) conditions, LRH-1 determines pool composition rather than bile salt synthesis rate: bile salt synthesis is even slightly increased rather than decreased in LRH-1-KD mice likely due to suppressed ileal Fgf15 expression. However, using bile salt sequestrants to deplete the bile salt pool by enhancing their fecal excretion, we found that LRH-1 does function as a critical factor in the compensatory induction of hepatic Cyp7a1 expression and bile salt synthesis. Our data provide mechanistic insight in a missing link in the maintenance of bile salt homeostasis and support the view that LRH-1 functions in a compensatory safeguard mechanism for adequate induction of bile salt synthesis under conditions of high bile salt loss.

Abbreviations

alpha-MCA, alpha-muricholate; beta-MCA, beta-muricholate; CYP7A1, cholesterol 7-alpha-monooxygenase; CYP8B1, sterol 12-alpha-hydroxylase; CA, cholate; DCA, deoxycholate; CDCA, chenodeoxycholate; FXR, farnesoid X-receptor; HDCA, hyrodeoxycholate; LRH-1, liver receptor homolog-1; omega-MCA, omega-muricholate.

Materials and Methods

Standard methods and assays can be found in the Supporting Information.

Animals.

LRH-1-KD mice were obtained from Taconic Artemis (Cologne, Germany). Details can be found in the Supporting Experimental Procedures. Twenty to 27-week-old male (n = 8) and female (n = 4) LRH-1-KD mice on the C57BL/6J background and wildtype (WT) male (n = 5) and female (n = 3) littermates were housed in individual cages in a temperature- and light-controlled facility with 12 hours light-dark cycling. All mice were fed commercially available laboratory chow (RMH-B; Hope Farms, Woerden, The Netherlands) containing 200 mg/kg doxycycline (Sigma, St. Louis, MO) and supplemented with colese velam HCl 2% (w/w) (Daiichi Sankyo, Parsippany, NJ) when indicated. All experiments were approved by the Ethical Committee for Animal Experiments of the University of Groningen. All animals received humane care according to the criteria outlined in the “Guide for the Care and Use of Laboratory Animals” prepared by the National Academy of Sciences and published by the National Institutes of Health.

Genotyping.

Detailed information for genotyping can be found in the Supporting Experimental Procedures.

Experimental Animal Procedures.

All mice were fed chow with 200 mg/kg doxycycline for 4 weeks. Thereafter, mice were transferred for 14 days to chow with doxycycline only or to chow with doxycycline supplemented with 2% (w/w) colesevelam HCl. Weight gain was followed during the course of the study. Mice were anesthetized by intraperitoneal injection of Hypnorm (1 mL/kg) (fentanylcitrate 0.315 mg/mL and fluanisone 10 mg/mL, VetaPharma, Leeds, UK) and diazepam (10 mg/kg) (Centrafarm, Etten-Leur, The Netherlands) and subjected to gallbladder cannulation for 20 minutes as described.35 During bile collection, body temperature was stabilized using an incubator. Bile was stored at −20°C until analyzed. Directly following bile collection, heart puncture was performed under isoflurane anesthesia and animals were sacrificed by cervical dislocation.

Blood obtained by heart puncture was collected in ethylenediaminetetraacetic acid (EDTA)-containing tubes. Plasma was stored at −20°C until analyzed. The liver was removed, weighed, and snap-frozen in liquid nitrogen. The intestine was excised, flushed with phosphate-buffered saline, and placed in a Z-form. Three samples of ≈1 cm were removed from the proximal, medial, and distal part of the intestine, representing duodenum, jejunum, and ileum, and snap-frozen in liquid nitrogen. Liver and intestinal samples were stored at −80°C until RNA isolation or biochemical analysis. Fecal excrement was collected from individually housed mice over a continuous 48-hour period. After air-drying, feces were kept at room temperature until analysis.

Analysis in Plasma and Liver.

Triglycerides, cholesterol, free fatty acids were determined with commercial kits. Activities of alanine and aspartate aminotransferases were measured using commercial kits. Quantification of bile salt and neutral sterol species was performed by gas chromatography. Details on analytical measurements can be found in the Supporting Experimental Procedures.

RNA Isolation and Polymerase Chain Reaction (PCR) Procedures.

Gene expression was measured using quantitative PCR (qPCR) performed with a 7900HT FAST system using FAST PCR master mix, Taqman probes, and MicroAmp FAST optical 96-well reaction plates (Applied Biosystems Europe, Nieuwekerk ad IJssel, The Netherlands). Primer and probe sequences can be obtained at RTprimerDB (http://www.rtprimerdb.org) (see Supporting Experimental Procedures for details).

Statistics.

All values are presented as Tukey's Box-and-Whiskers plot using median with 25th to 75th percentile intervals (P25-P75). Plots were created using the GraphPad Prism 5 software package. Statistical analyses were performed using SPSS 16.0 (Chicago, IL). Differences between the groups were analyzed by the nonparametric Mann-Whitney U test. When multiple comparisons were made (wildtype versus knockdown and chow versus colesevelam), the Kruskal-Wallis H test was performed, which was followed by the Conover Posthoc Test using Brightstat.36 Differences were considered statistically significant when P < 0.05.

Results

Metabolic Parameters in Chow-Fed Conditional Systemic LRH-1 Knockdown Mice.

A conditional short hairpin RNA (shRNA) knockdown strategy was utilized to obtain an inducible and reversible whole body Lrh-1 knockdown model. The model is based on an shRNA sequence targeting Lrh-1 (NR5A2) cloned behind a doxycycline-responsive promoter. The construct is targeted at the Rosa26 locus along with the enhanced tet-repressor (Fig. 1A). The resulting C57BL/6J mice were bred to be heterozygous for the knockdown cassette and WT littermates lacking the targeting construct were used as controls. Lrh-1 gene knockdown was induced by doxycycline administration by way of the food for 5 weeks. As shown in Fig. 1B, hepatic Lrh-1 mRNA levels were reduced by ≈90%-95%, whereas the reduction of Lrh-1 expression in small intestine was ≈60%-70% in male and female mice (Fig. 1B). The expression of Shp, a well-established Lrh-1 target gene,22, 23 was robustly reduced in liver (Fig. 1B). In contrast, levels of steroidogenic factor-1, the closest paralog of LRH-1, in the ovaries were unaltered upon expression of the shRNA (data not shown), indicating that knockdown is specific for Lrh-1. There were no overt abnormalities noticed in either group. Plasma aspartate aminotransferase and alanine aminotransferase activities were unchanged (Fig. 1C), implying that knockdown of hepatic Lrh-1 has no detrimental effect on hepatocyte cell integrity. As our model is fundamentally different from two previously reported ones,30, 31 we first analyzed a number of general metabolic parameters. As shown in Supporting Table 1, plasma cholesterol and triglyceride levels were unaltered and plasma lipoprotein profiles were found to be unchanged between wildtype and knockdown animals (data not shown).

Figure 1.

Efficient knockdown of LRH-1 in a novel LRH-1 transgenic mouse model. (A) Schematic representation of the Tet-inducible LRH-1-KD system. (B) Gene expression shows a strong reduction in mRNA levels of Lrh-1 and the LRH-1 target-gene Shp. (C) Plasma aspartate aminotransferase (ASAT) and alanine aminotransferase (ALAT) levels were not altered in LRH-1-KD mice compared to WT littermates (n = 3-8 animals per group) (*P < 0.05).

LRH-1 Knockdown Affects Bile Salt Composition and Bile Salt Synthesis Under Chow-Fed Conditions.

Two previous reports showed that bile salt composition rather than synthesis rate was altered in liver-specific Lrh-1 knockout mice.30, 31 Consistent with this, hepatic Cyp7a1 mRNA levels remained unaltered or were even slightly induced, whereas those of Cyp8b1 were reduced. We also found that knockdown of LRH-1 resulted in a significant reduction of Cyp8b1 mRNA levels (Fig. 2A). Surprisingly, hepatic Cyp7a1 mRNA levels were increased upon LRH-1 knockdown (Fig. 2A). Several genes implicated in hepatic bile salt transport (e.g., Ntcp, Abcb11/Bsep, and Abcb4/Mdr2) were all mildly reduced upon LRH-1 knockdown (Fig. 2A), in agreement with previous findings.31

Figure 2.

LRH-1-KD in chow-fed animals reduces Cyp8B1 expression and changes the physicochemical properties of the bile salt pool. (A) LRH-1-KD altered gene expression of LRH-1 target genes. (B) LRH-1-KD caused a small but significant increase in fecal bile salt excretion. (C) LRH-1-KD induces a shift in the relative abundance of fecal CA versus CDCA-derived bile salts. (D) Relative abundance of fecal bile salt species in WT versus LRH-1-KD animals (n = 3-8 animals per group) (*P < 0.05). CA, cholate; DCA, deoxycholate; CDCA, chenodeoxycholate; HDCA, hyrodeoxycholate; alpha-MCA, alpha-muricholate; beta-MCA, beta-muricholate; omega-MCA, omega-muricholate.

We next tested whether the physicochemical properties of the neutral sterol fraction as well as the bile salt pool were affected upon LRH-1 knockdown. LRH-1 knockdown did not significantly alter amounts or relative abundances of each of the major neutral sterols in feces (Supporting Fig. 1A-C). In agreement with induced Cyp7a1 levels, the total amount of fecal bile salts secreted per day, reflecting hepatic synthesis, was slightly increased (males +57%, females +59%) (Fig. 2B). The primary bile salts cholate (CA) and chenodeoxycholate (CDCA) are the direct products of de novo bile salt synthesis. Modifications of these bile salts in liver and intestine give rise to differentially structured primary and secondary bile salts, respectively. Consistent with suppressed hepatic Cyp8b1 expression levels, the profile was shifted towards CDCA-derived bile salts relative to CA-derived bile salts (Fig. 2C). Specifically, fecal contents of deoxycholate (DCA) were greatly reduced (Fig. 2D), whereas the relative and absolute abundances of CDCA and α-muricholate were increased (Fig. 2D). These data show that bile salt synthesis is shifted towards the CDCA production upon LRH-1 knockdown, in agreement with previous findings.30, 31 For most of these observations, no gender differences were observed. However, fecal bile salt composition was slightly different between males and females under chow-fed conditions (Fig. 2D).

LRH-1 Is Critical for Up-regulation of Cyp7a1 Expression During Bile Acid Sequestration.

As LRH-1 seems to be dispensable for maintenance of Cyp7a1 expression under chow-fed conditions, we evaluated whether LRH-1 is essential for up-regulation of Cyp7a1 expression under conditions when high rates of bile salt synthesis are required to compensate fecal loss. Colesevelam-HCl is a widely used bile salt sequestrant and its administration massively induces fecal bile salt excretion in mice without affecting pool size.33 LRH-1-KD and WT littermates were fed chow with doxycycline for 4 weeks to induce LRH-1 silencing. Thereafter, mice were fed doxycycline-containing chow with or without colesevelam for 2 weeks. Also in this experiment, Lrh-1 mRNA levels were robustly reduced in livers of LRH-1-KD animals and reduced to about 60% to 40% along the small intestinal tract (Fig. 3A). Colesevelam results in enhanced conversion of hepatic cholesterol to bile salts that must be compensated for by induction of de novo cholesterol synthesis by way of up-regulation of HMG-CoA reductase (HMGCR), the rate-controlling enzyme of cholesterol synthesis. Indeed, robust Hmgcr induction was observed in the colesevelam-treated WT mice (Fig. 3B). Colesevelam treatment did not alter hepatic Lrh-1 expression but reduced hepatic Shp levels in wildtypes (Fig. 3C). Consistent with a previous report,31 we found a small but significant reduction in hepatic Fxr mRNA levels in LRH-1-KD mice (Supporting Fig. 2A), whereas small intestinal Fxr mRNA levels were unaltered (Supporting Fig. 2B). Colesevelam did not alter hepatic or intestinal Fxr expression (Supporting Fig. 2A,B). Hepatic Hnf4α transcript levels were also slightly reduced in LRH-1-KD mice, whereas those of the Liver X receptor (Lxrα), a nuclear receptor involved in Cyp7a1 transcription in mice,34 were found unchanged (Supporting Fig. 2A).

Figure 3.

LRH-1-KD animals cannot adequately up-regulate Cyp7a1 expression. (A) Gene expression shows a strong reduction in mRNA levels of Lrh-1 in LRH-1-KD animals. (B) Colesevelam-mediated fecal bile salt excretion up-regulates HMG-CoA reductase expression. (C) Hepatic gene expression of WT versus colesevelam-treated animals. Colesevelam induces Cyp7a1 and Cyp8b1 in WT animals but not in LRH-1-KD animals (D) Ileal gene expression of WT versus colesevelam-treated animals. LRH-1-KD causes a reduction of ileal Fgf15 expression (n = 4-5 animals per group) (*P < 0.05).

In agreement with data from the previous experiment, knockdown of LRH-1 resulted in an increase of hepatic Cyp7a1 expression (Fig. 3C). Interestingly, whereas colesevelam treatment resulted in the expected and robust increase of Cyp7a1 transcription in wildtype mice, such an induction was not observed in the knockdown animals (Fig. 3C). Rather, hepatic Cyp7a1 mRNA levels were comparable in knockdown animals on and off colesevelam. The same pattern was seen for Hmgcr expression (Fig. 3B). As in the first experiment, Cyp8b1 mRNA levels were reduced in the LRH-1-KD animals. Transcription of the Cyp8b1 gene was tremendously induced upon colesevelam treatment in the wildtype but not in the knockdown animals (Fig. 3C).

These results show that LRH-1 is a critical transcription factor for adequate up-regulation of Cyp7a1 and Cyp8b1 transcription under conditions of bile salt sequestration. In addition, the apparent paradoxical behavior observed for Cyp7a1 transcription in the LRH-1-KD mice suggest that two LRH-1-dependent, but mechanistically different, mechanisms are involved in the transcriptional regulation of Cyp7a1 expression.

A previous study in mice deficient for intestinal Lrh-1 showed a reduction of intestinal Fgf15 mRNA expression, suggesting that intestinal LRH-1 directly or indirectly regulates Fgf15 expression.31 Colesevelam did not alter intestinal Lrh-1 expression in wildtype mice but did suppress Shp and Fgf15 expression (Fig. 3D), which is consistent with previous findings.33 Intestinal Shp levels were significantly reduced in LRH-1-KD mice on and off colesevelam (Fig. 3D). Interestingly, we also found a tremendous reduction in Fgf15 mRNA levels in Lrh-1-KD mice on and off colesevelam, indicating that (intestinal) Lrh-1 regulates the expression of the Fgf15 gene. To further support this relationship, we tested whether LRH-1 would increase expression of FGF19, the human ortholog of murine FGF15, in DLD cells. Transduction of DLD cells with increasing amounts of recombinant LRH-1 encoding adenoviral particles (Supporting Fig. 3A,B) caused a dose-dependent increase in FGF19 mRNA expression (Supporting Fig. 3C). These data indicate that LRH-1 indeed induces Fgf15/19 expression.

Alterations in Bile Salt Metabolism in LRH-1-KD Mice During Bile Acid Sequestration.

We tested whether altered Cyp7a1 expression in colesevelam-treated LRH-1-KD animals also had physiological consequences. Knockdown of LRH-1 did not cause significant alterations in bile flow rate and only tended to reduce biliary bile salt output (Fig. 4A,B). Treatment with colesevelam did not affect bile flow, but reduced biliary bile salt output in both WT mice and LRH-1-KD mice (Fig. 4A,B), in agreement with previous studies from our laboratory.33 In agreement with the observed increase in Cyp7a1 expression levels (Fig. 3C), knockdown of LRH-1 caused a modest increase (+10%) of fecal bile salt output (Fig. 4C). As expected, sequestrant treatment led to a massive induction (+272%) of fecal bile salt output in WT mice. Because colesevelam was given for 2 weeks, a new steady state is established in which fecal loss depicts enhanced bile acid synthesis. In LRH-1-KD mice there was no increase in fecal bile acid output after 2 weeks (Fig. 4C), indicating that LRH-1-KD mice cannot up-regulate bile acid synthesis during colesevelam treatment.

Figure 4.

LRH-1-KD animals cannot adequately up-regulate bile salt synthesis. (A) Bile flow was unchanged whereas (B) biliary bile salt output was slightly reduced in colesevelam-treated animals. (C) Fecal bile salt excretion is strongly induced in WT animals but not in LRH-1-KD mice reflecting a lack in the up-regulation of bile salt synthesis in LRH-1-KD mice. (D) Synthesis of CA-derived bile salts was massively increased upon colesevelam treatment in WT but not LRH-1-KD mice (n = 4-5 animals per group) (*P < 0.05).

As Cyp8b1 expression was also dysregulated in LRH-1-KD mice, we expected that LRH-1 knockdown combined with sequestrant would have profound effects on bile salt composition. Supporting Fig. 4 provides details on both the relative and absolute fecal and biliary bile salt compositions. In agreement with previous findings,33 colesevelam treatment resulted in increased relative and absolute contents of fecal DCA (Supporting Fig. 4A,B). Under sequestrant-fed conditions, the loss of bile salts is mainly compensated by an increased hepatic synthesis of CA that results in an increased relative abundance of CA-derived bile salts in bile (Fig. 4D and Supporting Fig. 4C,D). However, LRH-1-KD animals cannot compensate for the sequestrant-induced loss of bile salts by up-regulating CA and DCA synthesis (Supporting Fig. 4B) and this results in a decrease in the relative abundance of CA-derived bile salts and an increase in the relative abundance of CDCA-derived bile salts in bile (Fig. 4D, Supporting Fig. 4C,D).

Discussion

LRH-1 is a nuclear receptor that regulates the expression of a variety of genes involved in cholesterol and bile salt metabolism. Cultured cell studies have shown that both CYP7A1 and CYP8B1, two key enzymes in bile salt synthesis, are regulated by LRH-1. Cyp7a1 was initially identified as an LRH-1 target gene in an unbiased screen.8 Subsequent cell studies showed that LRH-1 acts as a positive transcription factor as well as a docking site for the transcriptional repressor SHP.22, 23 Comprehensive analysis of the physiological importance of LRH-1 in vivo has been hampered by the embryonic lethality of Lrh-1 knockout mice. Two laboratories independently generated conditional liver-specific Lrh-1 knockout models.30, 31 Surprisingly, hepatocyte-specific deficiency of Lrh-1 had no significant effect on Cyp7a1 expression,30, 31 and heterozygous Lrh-1 knockout mice exhibited 5 to 7-fold higher Cyp7a1 expression levels.32 Proposed explanations for these surprising findings were that LRH-1 either does not regulate Cyp7a1in vivo, or that compensatory responses or redundant factors maintain Cyp7a1 expression in the absence of LRH-1.31 In this study we used conditional whole-body LRH-1 knockdown mice to establish the involvement of LRH-1 on Cyp7a1 transcription in vivo. Our data unequivocally demonstrate that LRH-1 is a critical transcription factor that is required for adequate up-regulation of Cyp7a1 expression under conditions associated with high fecal bile salt loss, as caused by sequestrant treatment. Hence, the inability to up-regulate Cyp7a1 expression translated into relatively low bile salt synthesis rates in LRH-1 knockdown animals compared to wildtypes during sequestrant treatment. Together, our data resolve the apparent discrepancy between the outcomes of in vitro cell studies8, 22, 23 and in vivo mouse studies.30, 31 This proves the previously predicted role of LRH-1 in CYP7A1 expression and complements the proposed mechanism of bile acid inhibition of CYP7A1 expression by way of the FXR-SHP-LRH-1 cascade. In this pathway bile acid activation of FXR leads to induction of SHP, which in turn inhibits CYP7A1 activation by LRH-1. In agreement with cell studies26 and previous in vivo studies,30, 31 our data demonstrate that LRH-1 is critical for maintenance of Cyp8b1 expression, also under normal feeding conditions. Our data also show that LRH-1 is critical for adaptation of Cyp8b1 expression during high bile salt loss. In physiological terms, the reduction of Cyp8b1 expression levels in the knockdown animals was accompanied by the anticipated proportions of CA-derived versus CDCA-derived bile salts in bile and feces.

Together, the data clearly indicate that Cyp7a1 and Cyp8b1 expression are differentially regulated. LRH-1 appears to be critical for both Cyp7a1 and Cyp8b1 transcription under conditions of high bile salt loss yet dispensable for Cyp7a1 but not for Cyp8b1 expression under “normal” conditions. This strongly indicates that compensatory mechanisms or redundant transcription factors exist for maintenance of Cyp7a1 expression. Indeed, we and others showed that several transcription factors, including LXR/RXR, HNF4alpha and SHP contribute to Cyp7a1 transcription (Supporting Fig. 5). Unfortunately, several attempts to study Cyp7A1 and Cyp8B1 promoter occupancy by LRH-1 and HNF4alpha using chromatine immunoprecipitation analysis on liver material failed. Therefore, the nature of the differential regulation for Cyp7a1 and Cyp8b1 under normal conditions remains obscure and can even be mediated by epigenetic regulators such as GPS2.37

Careful examination of our data revealed that systemic knockdown of LRH-1 actually resulted in a significant up-regulation of hepatic Cyp7a1 expression that was accompanied by a small increase of bile salt synthesis. This indicates that two different pathways with a reciprocal outcome modulate Cyp7a1 expression in our model. Lrh-1 was significantly reduced in the small intestine of LRH-1-KD mice and, in agreement with the results from a conditional intestinal Lrh-1 knockout model,31 we also found that intestinal Fgf15 expression was significantly reduced. Experiments in DLD cells further support evidence that LRH-1 modulates FGF19 expression. However, it remains to be elucidated whether these effects result from a direct transcriptional induction by LRH-1, or by way of indirect mechanisms.

Surprisingly, Lee et al.31 reported that the reduction of intestinal Fgf15 expression in intestine-selective Lrh-1 knockouts did not result in an altered hepatic Cyp7A1 expression. However, the reduction of intestinal Fgf15 expression was relatively mild in these mice and these authors also found that hepatic Lrh-1 knockout resulted in a reduction of intestinal Fgf15 expression, possibly as a result of a reduction in FXR agonist activity in the hepatic Lrh-1 knockout mice.31 Thus, the separate deletion of either hepatic or intestinal Lrh-1, each reducing intestinal Fgf15 expression levels, appears not to alter hepatic Cyp7a1 expression levels. Yet when combined, as is the case in our LRH-1-KD mice, the reduction of Fgf15 expression is strong enough to affect hepatic Cyp7a1 expression. Indeed, Lrh-1 haploinsufficiency resulting in a whole-body reduction of LRH-1 showed higher Cyp7a1 levels compared to littermates harboring both alleles.9 This provides additional insights into the central role of FGF15 in bile acid homeostasis. Interestingly, our data show that only Cyp7a1 and not Cyp8b1 is induced upon LRH-1 knockdown. The involvement of Fgf15 herein is supported by data from Kim et al.,38 who showed that Cyp7a1 is suppressed much more efficiently compared to Cyp8b1 by FGF15 signaling.

In summary, our data demonstrate that LRH-1 is a critical transcription factor for up-regulation of Cyp7a1 expression and bile salt synthesis in vivo during bile salt sequestration. In addition, our data support the view that LRH-1 affects Cyp7a1 expression from at least two sites in the enterohepatic system. Hepatic LRH-1 together with other transcription factors positively regulates Cyp7a1 expression, whereas intestinal LRH-1 causes an opposing effect by stimulating the expression of Fgf15 expression in enterocytes resulting in a repression of CYP7A1 (Fig. 5). The finding that LRH-1 is indispensable for up-regulating bile salt synthesis indicates that it could serve as an attractive target to combat hypercholesterolemia.

Figure 5.

Lrh-1 regulates Cyp7a1 expression from two distinct sites in the enterohepatic circulation. Schematic representation of physiological pathways that regulate Cyp7a1 transcription. Hepatic LRH-1 positively contributes to Cyp7a1 transcription, whereas intestinal LRH-1 represses Cyp7a1 transcription by way of the induction of Fgf15 expression.

Acknowledgements

We thank Renze Boverhof for excellent technical assistance on GC/MS analyses.

Ancillary