Cholestasis is associated with retention of bile acids and reduced expression of the Na+/taurocholate cotransporter (Ntcp), the major hepatocellular bile acid uptake system. This study aimed to determine whether downregulation of Ntcp in obstructive cholestasis 1) is a consequence of bile acid retention and 2) is mediated by induction of the transcriptional repressor short heterodimer partner 1 (SHP-1). To study the time course for the changes in serum bile acid levels as well as SHP-1 and Ntcp steady-state mRNA levels, mice were subjected to common bile duct ligation (CBDL) for 3, 6, 12, 24, 72, and 168 h and compared with sham-operated controls. Serum bile acid levels were determined by radioimmunoassay. SHP-1 and Ntcp steady-state mRNA expression were assessed by Northern blotting. In addition, Ntcp protein expression was studied by Western blotting and immunofluorescence microscopy. Increased SHP-1 mRNA expression paralleled elevations of serum bile acid levels and was followed by downregulation of Ntcp mRNA and protein expression in CBDL mice. Maximal SHP-1 mRNA expression reached a plateau phase after 6-h CBDL (12-fold; P < 0.001) and preceded the nadir of Ntcp mRNA levels (12%, P < 0.001) by 6 h. In conclusion, bile acid-induced expression of SHP-1 may, at least in part, mediate downregulation of Ntcp in CBDL mice. These findings support the concept that downregulation of Ntcp in cholestasis limits intracytoplasmatic accumulation of potentially toxic bile acids.
- bile acids
- proinflammatory cytokines
- orphan nuclear receptors
- transcription factors
cholestasis is associated with reduced expression of hepatocellular transport systems, which contributes to impaired hepatic uptake and biliary excretion of bile acids and other biliary constituents (e.g., bilirubin) (24, 44). However, it is unclear whether these changes in transporter expression are primary events resulting in elevated bile acid levels or consequences of bile acid retention.
The Na+/taurocholate cotransporter (Ntcp; official gene code Slc10a1) is the major hepatocellular uptake system for bile acids (23) and is profoundly downregulated at a transcriptional level in animal models of cholestasis such as common bile duct-ligated (CBDL) or endotoxin-treated rats (14,42). Hepatic steady-state mRNA levels of Ntcp are inversely related to serum bile acid levels in CBDL rats (15) and humans with various cholestatic disorders (47), suggesting that accumulating bile acids may suppress Ntcp gene expression (15).
Bile acids stimulate expression of short heterodimer partner 1 (SHP-1; official nuclear receptor name NR0B2), a transcriptional repressor inhibiting expression of several genes (17, 30,38-40). As such, downregulation of cholesterol-7α-hydroxylase (cytochrome P-450 7a1,Cyp7a1), the rate limiting enzyme of bile acid synthesis, by bile acid-induced SHP-1 provides the molecular mechanism for feedback inhibition of bile acid synthesis through bile acids (7, 8, 17, 30). Furthermore, bile acid-induced SHP-1 also inhibits retinoid transactivation ofNtcp promoter activity in vitro (10). Mice with targeted disruption of the farnesoid X receptor (FXR, NR1H4), now known as the nuclear bile acid receptor (31, 35, 45), lack bile acid-mediated induction of SHP-1 and concurrent reduction of Ntcp in response to cholic acid (CA) feeding (40). These findings further suggest that SHP-1 may a play a pivotal role in the regulation of Ntcp expression in vivo. However, the role of endogenous bile acids (accumulating during cholestasis) and SHP-1 in mediating reducedNtcp expression during cholestasis is unknown.
Therefore, the aim of the present study was to test the hypotheses that1) downregulation of Ntcp in obstructive cholestasis is a consequence of elevated bile acid levels and 2) is mediated by induction of the transcriptional repressor SHP-1. To address these questions, changes in serum bile acid levels and SHP-1 steady-state mRNA levels were studied in bile duct-ligated and sham-operated mice and compared with the time course of changes in Ntcp expression. The present study clearly demonstrates, that accumulation of bile acids and induction of SHP-1 precedes the downregulation of Ntcp by several hours, indicating that reduced Ntcp expression is a secondary event rather than the cause for elevated bile acid levels in obstructive cholestasis.
MATERIALS AND METHODS
Male Swiss albino mice (strain Him OF1 SPF), 25–30 g body wt, were obtained from the Institute of Laboratory Animal Research, University of Vienna, School of Medicine, Himberg, Austria. They were housed with a 12:12 h light-dark cycle and permitted ad libitum consumption of water and a standard mouse diet (Marek, Vienna, Austria). The experimental protocols were approved by the local Animal Care and Use Committee, according to criteria outlined in the Guide for the Care and Use of Laboratory Animals (National Institutes of Health publication 86–23, revised 1985).
The following reagents were used: pCR II vector and One Shot competent cells (Invitrogen, Groningen, The Netherlands); Quantum Prep Plasmid Miniprep and Maxiprep kits (Bio-Rad Laboratories, Hercules, CA); deoxycytidine 5′-triphosphate ([32P]dCTP) and Random Prime II DNA Labeling Kit (Amersham, Little Chalfont, UK); avian myeloblastosis virus reverse transcriptase and restriction enzymes (Boehringer Mannheim, Mannheim, Germany); AmpliTaq DNA Polymerase (Perkin-Elmer, Branchburg, NJ). Cholate was obtained from Aldrich (Steinheim, Germany). Lipopolysaccharide (LPS) from Salmonella typhimurium, Escherichia coli O26:B6, and E. coliO55:B5 was purchased from Sigma (Steinheim, Germany). All other chemicals used were of the highest purity commercially available and purchased from Sigma.
All surgical procedures were performed under sterile conditions. The common bile duct was ligated close to the liver hilum immediately below the bifurcation and dissected between the ligatures as described previously (43). In addition, cholecystectomy was performed after ligation of the cystic duct. Control animals underwent a sham operation with exposure, but without ligation of the common bile duct and removal, of the gallbladder. The livers were excised 3, 6, 12, 24, 72 h (3 days), and 168 h (7 days) after CBDL under general anesthesia, immediately snap-frozen, and stored in liquid nitrogen until RNA extraction and preparation of liver membranes.
Bile acid feeding and endotoxin treatment.
To compare the effects of bile acids (retained during cholestasis) and proinflammatory cytokines [induced during biliary obstruction (4, 16)], mice were fed a diet supplemented with CA (1% wt/wt) or treated with intraperitoneal injections of LPS from S. typhimurium, E. coli O26:B6, and E. coliO55:B5 (at doses ranging from 0.5 to 15 mg/kg body wt). The CA and LPS doses were previously shown to be biologically active and to reduce Ntcp expression (13, 18). Control animals received a standard diet or were injected with vehicle (saline), respectively. Furthermore, CA feeding at this dose (1% wt/wt) results in enrichment of the bile acid pool with taurocholate by >95% (13). Livers were harvested after 7 days of CA-feeding or 16 h after LPS administration.
Serum bile acid measurements.
Blood was collected at the time of death, and serum samples were stored at −70°C until analysis of total serum bile acid levels by a commercially available radioimmunoassay for conjugated bile acids (ICN Pharmaceuticals, New York, NY). Tests were performed in duplicate.
Isolation of total RNA was performed according to a procedure described by Krieg et al. (22). RNA was quantified spectrophotometrically at 260 nm, and the quality of total RNA was controlled by denaturing formaldehyde agarose gel electrophoresis.
Northern blot analysis.
Total RNA (20 μg) was electrophoresed on a 1.2% agarose-formaldehyde gel, transferred to Hybond N membranes (Amersham) by overnight capillary transfer blotting and cross-linked by ultraviolet light (Stratalinker 1800; Stratagene, La Jolla, CA). Membrane prehybridization and hybridization were performed at 45°C for 1 h and overnight, respectively, following a standard protocol (2). Probes were labeled with [32P]dCTP by a random primed method according to the manufacturer's instructions (Random Prime II DNA Labeling Kit; Amersham). Membranes were washed under medium-stringency conditions (two washing steps with 2× SSC/1% SDS followed by two washing steps with 0.2× SSC-1% SDS at room temperature, 10 min each). mRNA levels were detected by exposure to Kodak BioMax films (Kodak, Rochester, NY) and quantified using video-densitometry software (RFLP-Scan or ZeroD-Scan; Scanalytics, Billerica, MA). Membranes were stripped and reprobed for GAPDH to determine equal loading. The size of mRNA was estimated by a 0.24– to 9.5-kb RNA ladder (GIBCO-BRL, Gaithersburg, MD). Specific probes were generated for SHP-1 (NR0B2), Ntcp (Slc10a1), and GAPDH using RT-PCR with following primer pairs: Ntcp (Gene Bank Acc. No. U95131) (5) upstream 290–307, downstream 905–886; SHP-1 (GenBank accession no. NM-011850) (38) upstream 290–309, downstream 782–763; GAPDH (GenBank accession no.M32599) (37) upstream 428–445, downstream 551–535. The PCR products were cloned into pCR II vectors (Invitrogen) and their specificity was checked by sequencing with an Abi Prism automatic sequencer (Perkin-Elmer).
Preparation of liver membranes.
Crude liver membranes were isolated, and protein concentrations were determined according to Bradford, as described previously (13). Protein yields were similar in controls and CBDL (data not shown).
Western blot analysis.
Similar amounts of protein (150 μg) were loaded onto 10% SDS-polyacrylamide gels, without boiling, and subjected to electrophoresis (25). Equal protein loading was confirmed by Coomassie staining of gels. After electrotransfer onto nitrocellulose membranes (Bio-Rad, Richmond, CA), the blots were incubated with a polyclonal antibody against Ntcp (kindly provided by Drs. Peter Meier and Bruno Stieger, Zurich, Switzerland), and immune complexes were detected using horseradish-conjugated goat anti-rabbit IgG F(ab′)2 fragments as described (13).
Immunofluorescence staining for Ntcp was performed as described previously (13). In brief, cryosections of liver tissue were fixed in a 4% paraformaldehyde solution in PBS for 20 min, rinsed three times in 50 mM NH4Cl in PBS, and treated with 5% Triton X-100 in 50 mM NH4Cl in PBS for 5 min, followed by incubation with a polyclonal antibody against Ntcp (dilution, 1:25). A tetramethylrhodamine isothiocyanate-conjugated goat anti-rabbit antibody (Dako, Glostrup, Denmark) was used as secondary antibody. Negative controls were performed by omitting the primary antibody. Fluorescent staining was visualized using a MRC 600 laser scanning confocal device (Bio-Rad) attached to a Axiophot (Zeiss, Oberkochen, Germany). The fluorescent images were collected using the confocal photomultiplier tube as full frame (768 × 512 pixels). Only samples prepared in parallel in all steps were compared.
In each group, five animals were studied for each given time point. Data are reported as arithmetic means ± SE. Differences among experimental groups were analyzed by unpaired t-test and ANOVA with Bonferroni posttesting using the SigmaStat statistic program (Jandel Scientific, San Rafael, CA). A P value <0.05 was considered significant.
Bile acid levels in CBDL mice.
To establish a time course for the accumulation of serum bile acids compared with SHP-1 and Ntcp mRNA expression in biliary obstruction, mice were subjected to CBDL for 3, 6, 12, 24, 72 h (3 days), and 168 h (7 days) and compared with naive mice (0 h) and sham-operated controls. Compared with naive mice (1.3 ± 0.3 μmol/l), serum bile acid levels already were 270-fold increased in 3-h CBDL mice (343 ± 101 μmol/l) with a first peak around 6-h CBDL (370-fold; 464 ± 36 μmol/l), followed by a second and even higher peak around 24-h CBDL (690-fold; 1011 ± 83 μmol/l) and decreasing levels, thereafter (570-fold at 72 h; 180-fold at 168 h) (Fig.1 B). Sham-operated controls showed no significant differences compared with naive mice (Fig.1 B).
Induction of SHP-1 expression parallels elevation of serum bile acid levels in CBDL mice.
Because SHP-1 is induced by bile acids and represents a well-defined suppressor of Ntcp expression in vitro (10), SHP-1 steady-state mRNA expression was determined in CBDL mice at the time points indicated above (Fig. 1, A and B). SHP-1 mRNA levels paralleled the increase of bile acid levels. Maximal expression was observed 6 h after CBDL (coinciding with the first peak of serum bile acid levels) followed by a plateau despite ongoing bile acid accumulation culminating in a second peak at 24 h after CBDL. The decline of SHP-1 mRNA levels after 72 h paralleled the reduction of serum bile acid levels beyond this time point. Sham-operated controls showed no significant differences compared with naive mice (Fig. 1 B).
Downregulation of Ntcp expression follows bile acid-induced SHP-1 expression in CBDL mice.
Downregulation of Ntcp transcription has been previously shown in several rodent models of cholestasis (14, 42). However, it has remained an open question whether this is a cause or consequence of cholestasis (44). Therefore, the time course of Ntcp steady-state mRNA expression was compared with serum bile acid levels (as an indicator of bile acid retention) and SHP-1 mRNA expression in CBDL mice. Downregulation of Ntcp mRNA expression followed the elevation of serum bile acid levels and induction of SHP-1 mRNA expression with a lag period of several hours (Fig. 1,A and B). Of note, Ntcp mRNA levels remained almost unchanged after 3-h CBDL, a time point where serum bile acid levels and SHP-1 mRNA expression were already increased 270- and 6-fold, respectively. The nadir of Ntcp mRNA levels was reached at 12 h, representing a lag period of 6 h after the plateau of SHP-1 mRNA expression was reached (Fig. 1, A andB). Sham-operated controls showed no significant changes in Ntcp mRNA levels compared with naive mice (Fig. 1 B).
Ntcp protein levels and localization in CBDL mice.
To determine whether the observed changes of Ntcp mRNA levels affect Ntcp protein expression, Western immunoblotting was performed in crude liver membranes obtained from livers from sham-operated and CBDL mice (Fig. 2, A and B). Ntcp protein levels were still unchanged after 6- and 12-h CBDL (time points when SHP-1 mRNA expression was already induced 6- and 12-fold, respectively), but decreased significantly to 67 ± 5% after 24 h and reached the minimum with 44 ± 22% after 168 h (7 days) CBDL (Fig. 2 B). Sham-operated controls showed no significant changes in Ntcp protein levels compared with naive mice (Fig. 2 A). Immunofluorescence staining showed a regular basolateral staining pattern of Ntcp at 3-, 6-, and 12-h CBDL, but a disrupted and reduced staining pattern beyond 24-h CBDL (Fig.3, A–F). Reduced Ntcp protein expression after the elevation of serum bile acid levels further indicates that downregulation of Ntcp expression in CBDL mice is consequence rather than cause of elevated serum bile acid levels.
CA feeding, but not LPS administration, induces SHP-1 mRNA, whereas both reduce Ntcp mRNA expression.
In addition to retention of bile acids, CBDL also results in activation of proinflammatory cytokines (4, 16), which represent established inhibitors of Ntcp expression (9, 42). To distinguish between the impact of cytokines and bile acids on SHP-1 expression in CBDL, the effects of LPS administration (a potent stimulator of cytokine production in vivo) were compared with CA feeding [known to induce SHP-1 concomitantly with reducing Ntcp (40)]. In line with previous findings (40), CA feeding over 1 wk resulted in elevated serum bile acid levels (71.5 ± 20.9 μmol/l vs. 3.9 ± 0.8 μmol/l in pair-fed mice; P < 0.05), induced SHP-1 mRNA by 4.2-fold (P < 0.01), and reduced Ntcp mRNA and protein expression to 73 ± 15% (P < 0.05) and 52 ± 8% (P < 0.05), respectively (Figs.4 A and 5A), suggesting that induction of SHP-1 by bile acids may be linked to changes in Ntcp expression. The moderate changes in SHP-1 and Ntcp mRNA levels are consistent with the mild elevations of serum bile acid levels (when compared with the more profound changes after CBDL). To induce proinflammatory cytokines in vivo, mice were injected with various doses and types of LPS. In line with previous findings (18,42), Ntcp steady-state mRNA levels were reduced to 28 ± 10% of controls (P < 0.01) 16 h after injection of LPS from S. typhimurium (15 mg/kg body wt) (Fig.4 B). At this time point, Ntcp protein levels were not yet significantly reduced (Fig. 5 B), consistent with only minimally elevated serum bile acid levels (2.1 ± 1 μmol/l vs. 1.2 ± 0.4 μmol/l in saline injected mice; P < 0.05). In contrast to the effects of CA feeding and despite the profound LPS effects on Ntcp mRNA expression, SHP-1 mRNA levels were not induced by LPS (Fig. 4 B), suggesting that SHP-1 may not mediate the downregulation of Ntcp by LPS-induced proinflammatory cytokines. LPS from E. coli O26:B6 and E. coliO55:B5 had similar effects on SHP-1 and Ntcp expression (data not shown).
This is the first study demonstrating bile acid-mediated induction of a transcriptional repressor (SHP-1), which precedes downregulation of the major hepatic bile acid uptake system (Ntcp) in a mouse model of cholestasis (CBDL) in vivo.
Ntcp was the first cloned bile acid transport system (20) and is among the best characterized hepatocellular transport systems under physiological and cholestatic conditions (23). Previous studies have demonstrated profound reduction ofNtcp gene transcription in various animal models of cholestasis including CBDL and LPS-treated rats (14, 42), although the mediators and molecular mechanisms responsible for these changes have not been entirely clarified (44). Furthermore, it was not clear whether changes in transporter expression are primary (causing elevations of bile acid levels) or secondary (a consequence of bile acid retention). Using Ntcp as a model system, the present study demonstrates that reduction of Ntcp expression is secondary to bile acid retention, resulting in induction of the transcriptional repressor SHP-1.
The recent discovery of bile acids as ligands for orphan nuclear receptors allows a better understanding of bile acid effects on gene transcription (6, 36). SHP-1 is an atypical member of the orphan nuclear receptor family lacking a DNA-binding domain and inhibits gene transcription by interaction with other nuclear transcription factors (29, 38, 39). Bile acids stimulateSHP-1 transcription via activation of farnesoid X receptor (FXR) (17, 30, 40); in addition, activation of the c-Jun NH2-terminal kinase pathway resulting in formation of activating protein-1 (AP-1) may also contribute to the effects of bile acids on SHP-1 expression (19). Interaction of bile acid-induced SHP-1 with liver receptor homolog 1, a monomeric orphan receptor required for hepatic expression of Cyp7a1, plays a critical role in the negative feedback repression of Cyp7a1 catalyzing the rate limiting step in bile acid biosynthesis (17, 30).
What could be the molecular mechanisms for suppression ofNtcp gene expression in CBDL mice? Constitutive expression of Ntcp critically depends on transactivation by a heterodimer (formerly known as footprint B binding protein) composed of the orphan nuclear receptors retinoic X receptor-α (RXRα; NR2B1) and retinoic acid receptor-α (RARα; NR1B1) (9, 21). SHP-1 inhibits DNA binding of RXR:RAR heterodimers (38) and competes with coactivators for binding to ligand-activated RXR (29). Decreased activity of RXR:RAR has been demonstrated in animal models of cholestasis and is thought to play a critical role for reduced Ntcp transcription under these conditions (9, 42). Bile acid-induced SHP-1 also inhibitsNtcp promotor activity in vitro by inhibition of RXR:RAR transactivation (10). In line with previous findings (40), CA feeding to mice in the present study induced SHP-1 concurrently with reducing Ntcp expression. These changes are absent in FXR−/− mice (40), further suggesting a role for bile acid-induced SHP-1 in the regulation of Ntcp expression in vivo. Therefore, bile acid-mediated induction of SHP-1 in CBDL mice as demonstrated in the present study is likely to be responsible for reduced Ntcp gene expression in mechanical cholestasis (Fig. 6). This concept is further supported by the time course of events with induction of SHP-1 steady-state mRNA levels paralleling serum bile acid levels and preceding the reduction of Ntcp by several hours as demonstrated in this study. Future studies will have to explore the role of SHP-1 in the downregulation of other genes (e.g., the canalicular conjugate export pump), which also critically depends on transactivation by RXR:RAR (9), to explain the coordinated downregulation of several basolateral and canalicular transporter genes in cholestasis (23, 44).
Serum bile acid levels showed a biphasic pattern, with a first peak around 6 h followed by a second and an even higher peak 24 h after CBDL. Interestingly, serum bile acid levels were already significantly elevated at time points when Ntcp protein levels were still normal and Ntcp was regularly localized to the basolateral membrane. The most likely explanation for this finding could be regurgitation of bile acids through leaky canalicular and ductular tight junctions (1), because bile infarcts as a result of biliary leakage are observed as early as 3 h after CBDL (P. Fickert and M. Trauner, unpublished observation). Alternatively, posttranslational modifications of Ntcp impairing its function (34) and basolateral efflux of bile acids from cholestatic hepatocytes via induced multidrug resistance protein 3 (11,41) could also contribute to increased serum bile acid levels, although this remains to be determined in future studies. Of interest, maximal SHP-1 mRNA expression already reached a plateau 6 h after CBDL, despite a further increase in serum bile acid levels. Because SHP-1 is known to repress its own gene transcription (30), one may speculate that the stagnation of SHP-1 expression represents a negative feedback mechanism limiting further induction of SHP-1. SHP-1-induced downregulation of Ntcp resulting in impaired bile acid uptake may, at least in part, contribute to the further increase of serum bile acid levels between 12 and 24 h after CBDL. Decreasing bile acid levels beyond 24-h CBDL, despite continued biliary obstruction, have also been described in CBDL rats (14). This phenomenon may be explained by partial recovery of the canalicular bile salt export pump (28) and renal elimination of bile acids facilitated by upregulation of the apical sodium-dependent bile acid transporter in cholangiocytes and reciprocal downregulation in proximal renal tubules (26).
In addition to retention of bile acids, induction of proinflammatory cytokines could also contribute to downregulation of Ntcp in biliary obstruction (4, 16). Cytokines could mediate at least part of the bile acid effects on gene expression, as suggested by absent bile acid-mediated feedback inhibition of Cyp7a1 in cytokine-resistant mice (32). Proinflammatory cytokines such as tumor necrosis factor-α (TNF-α) and interleukin-1β are established inhibitors of Ntcp transcription (18, 33, 42). These effects may, at least in part, be mediated by inhibition of retinoid transactivation of Ntcp, because LPS and proinflammatory cytokines inhibit RXR:RAR binding activity (9,42) and reduce RXRα transcription (3). In addition, proinflammatory cytokines could also be expected to inhibit Ntcp expression by induction of SHP-1 via a recently identified AP-1 response element in the SHP-1 promoter, because cytokines (similar to bile acids) activate the c-Jun NH2-terminal kinase pathway (19). However, our results demonstrate that LPS-induced cytokines reduce Ntcp expression without inducing SHP-1. These data further indicate that, in contrast to bile acids, proinflammatory cytokines may not contribute to the induction of SHP-1 mRNA in CBDL mice. Moreover, the administration of anti-TNF-α antiserum and suppression of cytokine synthesis by dexamethasone did not prevent downregulation of Ntcp in CBDL rats (16). In line with our findings, these data suggest that proinflammatory cytokines may not contribute to downregulation of Ntcp in biliary obstruction.
In summary, the present study demonstrates for the first time that bile acid-mediated induction of a transcriptional repressor (SHP-1) precedes downregulation of Ntcp as a model transport system during cholestasis. Our findings suggest that bile acid-induced SHP-1 expression mediates, at least in part, downregulation of Ntcp, which aggravates cholestasis with systemic bile acid retention but may protect hepatocytes from accumulation of potentially toxic bile acids.
This work was supported by Grant 8522 from Jubilee Funds of the Austrian National Bank (to M. Trauner), the Joseph Skoda Prize from the Austrian Society of Internal Medicine (to M. Trauner) and Grant S7401-MOB from the Austrian Science Foundation (to K. Zatloukal).
First published September 21, 2001; 10.1152/ajpgi.00215.2001
This study was presented, in part, at the 52nd Annual Meeting of the American Association for the Study of Liver Diseases, Dallas, TX, November 9–13, 2001, and published in abstract form (Hepatology; 34:469A, 2001.
Address for reprint requests and other correspondence: M. Trauner, Division of Gastroenterology and Hepatology, Dept. of Internal Medicine, Karl-Franzens Univ., Auenbruggerplatz 15, A-8036 Graz, Austria (E-mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2002 the American Physiological Society