1. Volume 39, Issue 4, Pages 480-488 (October 2003)
Role of nuclear bile acid receptor, FXR, in adaptive ABC transporter regulation by cholic and ursodeoxycholic acid in mouse liver, kidney and intestine 
Gernot Zollner, Peter Fickert, Andrea Fuchsbichler, Dagmar Silbert, Martin Wagner, Silvia Arbeiter, Frank J Gonzalez, Hanns-Ulrich Marschall, Kurt Zatloukal, Helmut Denk, Michael Trauner 
Journal of Hepatology 
Volume 39, Issue 4, Pages (October 2003)
DOI: /S (03)

2. Fig. 1
Effects of CA and UDCA feeding on hepatic Mrp2 expression in FXR+/+ and FXR−/− mice. Total RNA and liver membranes were isolated from control diet-fed, CA- (1% w/w, 7 days), and UDCA- (1% w/w, 7 days) fed FXR+/+ and FXR−/− mice and analyzed by competitive RT–PCR (A) and Western blotting (B). In addition, Mrp2 tissue localization was assessed by confocal laser scanning microscopy (C–H) as described in Section 2. *CA- or UDCA-fed mice versus standard diet-fed mice (P<0.05). #UDCA-fed FXR−/− mice versus UDCA-fed wild-type mice (P<0.05). (A) Steady-state mRNA levels (means±SEM) are reported as % of control diet-fed FXR+/+ mice. Control diet-fed (Control, open bars), CA-fed (CA, hatched bars) and UDCA-fed (UDCA, closed bars) mice (n=5 in each group). CA as well as UDCA show a trend towards increased Mrp2 mRNA levels in both genotypes without reaching statistical significance. (B) Representative immunoblots. Densitometry data are expressed as the fold change relative to standard diet-fed FXR+/+ mice. Values are the average obtained from 3–4 animals in each group. There are no significant differences in baseline Mrp2 protein levels for standard diet-fed FXR−/− mice and wild-type controls. CA and UDCA significantly increase Mrp2 protein levels in both genotypes. (C–H) Representative immunofluorescence images. Canalicular Mrp2 tissue localization (in red) was assessed by immunofluorescence microscopy. For better visualization of the hepatic architecture, the cytokeratin network (in green) was counterstained. Overlap of pericanalicular cytokeratin sheet with canalicular Mrp2 immunolabeling results in a yellow signal. Similar canalicular Mrp2 staining pattern in control diet-fed FXR−/− (F) and wild-type mice (C). Enhanced canalicular staining pattern in CA- and UDCA-fed FXR+/+ (D,E) and FXR−/− mice (G,H). Bar: 20 μm.
Journal of Hepatology  , DOI: ( /S (03) )

3. Fig. 2
Effects of CA feeding on hepatic Mrp3 expression in FXR+/+ and FXR−/− mice. Total RNA and liver membranes were isolated from control diet-fed, CA- (1% w/w, 7 days), and UDCA- (1% w/w, 7 days) fed FXR+/+ and FXR−/− mice and analyzed by competitive RT–PCR (A) and Western blotting (B). In addition, Mrp3 tissue localization was assessed by confocal laser scanning microscopy (C–H) as described in Section 2. *CA- or UDCA-fed mice versus standard diet-fed mice (P<0.05). #CA-fed FXR−/− mice versus CA-fed wild-type mice (P<0.05). (A) Steady-state mRNA levels (means±SEM) are reported as % of controls (control diet-fed FXR+/+ mice). Control diet-fed (Control, open bars), CA-fed (CA, hatched bars) and UDCA-fed (UDCA, closed bars) mice (n=5 in each group). CA as well as UDCA induce Mrp3 mRNA expression in both genotypes. (B) Representative immunoblots. Densitometry data are expressed as the fold change relative to standard diet-fed FXR+/+ mice. Values are the average obtained from 3–4 animals in each group. There are no significant differences in baseline Mrp3 protein levels between standard diet-fed FXR−/− mice and wild-type controls. CA and UDCA significantly increase Mrp3 protein levels in both genotypes. (C–H) Representative immunofluorescence images. Mrp3 tissue localization (in white) was assessed by immunofluorescence staining. Similar basolateral Mrp3 staining in control diet-fed FXR−/− and wild-type mice (C,F). Enhanced basolateral staining pattern in CA- and UDCA-fed FXR+/+ (D,E) and FXR−/− mice (G,H). Note also the positive staining of bile ducts (bd). Bar: 20 μm.
Journal of Hepatology  , DOI: ( /S (03) )

4. Fig. 3
Effects of CA and UDCA feeding on renal Mrp2 expression in FXR+/+ and FXR−/− mice. Total RNA and kidney membranes were isolated from control diet-fed, CA- (1% w/w, 7 days), and UDCA- (1% w/w, 7 days) fed FXR+/+ and FXR−/− mice and analyzed by competitive RT–PCR (A) and Western blotting (B). In addition, Mrp2 tissue localization was assessed by confocal laser scanning microscopy (C-H) as described in Section 2. *CA- and UDCA-fed mice versus standard diet-fed mice (P<0.05). #CA- or UDCA-fed FXR−/− mice versus CA- or UDCA-fed wild-type mice (P<0.05). (A) Steady-state mRNA levels (means±SEM) are reported as % of controls (control diet-fed FXR+/+ mice). Control diet-fed (Control, open bars), CA-fed (CA, hatched bars) and UDCA-fed (UDCA, closed bars) mice (n=5 in each group). CA significantly induces renal Mrp2 mRNA expression in both genotypes, while UDCA only shows a trend for increased levels. (B) Representative immunoblots. Densitometry data are expressed as the fold change relative to standard diet fed FXR+/+ mice. Values are the average obtained from 3–4 animals in each group. There are no significant differences between standard diet-fed FXR−/− mice and wild-type controls. CA and UDCA significantly increase renal Mrp2 protein levels in both genotypes. In addition, renal Mrp2 expression in bile acid-fed FXR−/− mice is significantly increased compared to bile acid-fed wild-type controls. (C–H) Representative immunofluorescence images. Tubular Mrp2 tissue localization (in white) was assessed by immunofluorescence staining. Similar apical Mrp2 staining in proximal renal tubules in control diet-fed FXR−/− and wild-type mice (C,F). Enhanced proximal tubular staining pattern in CA- and UDCA-fed FXR+/+ and FXR−/− mice (D,E,G,H). g, glomerula. Bar: 30 μm.
Journal of Hepatology  , DOI: ( /S (03) )

5. Fig. 4
Effects of CA and UDCA feeding on intestinal Mrp2 expression in FXR+/+ and FXR−/− mice. Total RNA and intestinal membranes were isolated from control diet-fed, CA- (1% w/w, 7 days), and UDCA- (1% w/w, 7 days) fed FXR+/+ and FXR−/− mice and analyzed by competitive RT–PCR and Western blotting. In addition, duodenal Mrp2 tissue localization was assessed by confocal laser scanning microscopy (C–H) as described in Section 2. *CA- or UDCA-feeding versus control diet-fed mice. Steady-state mRNA and protein levels (means±SEM) in FXR+/+ (A) and FXR−/− (B) are reported as % of controls (defined as duodenal expression of control diet-fed mice). Duodenum (closed bars), jejunum (open bars), ileum (hatched bars) and colon (squared bars) (n=3 in each group). CA and UDCA significantly induce Mrp2 mRNA and protein expression in duodenum, jejunum and ileum in FXR+/+ (A) and FXR−/− (B), while expression levels remain low in colon in both genotypes (A,B). (C–H) Representative immunofluorescence images. Duodenal Mrp2 tissue localization (in white) was assessed by immunofluorescence staining. Similar Mrp2 staining on apical membranes of enterocytes in control diet-fed FXR−/− and wild-type mice (C,F). Enhanced staining pattern in CA- and UDCA-fed FXR+/+ and FXR−/− mice (D,E,G,H). Bar: 20 μm.
Journal of Hepatology  , DOI: ( /S (03) )

6. Fig. 5
Effects of CA and UDCA feeding on intestinal Mrp3 expression in FXR+/+ and FXR−/− mice. Total RNA and intestinal membranes were isolated from control diet-fed, CA- (1% w/w, 7 days), and UDCA- (1% w/w, 7 days) fed FXR+/+ and FXR−/− mice and analyzed by competitive RT–PCR and Western blotting as described in Section 2. *CA- or UDCA-feeding versus control diet-fed mice. Steady-state mRNA and protein levels (means±SEM) in FXR+/+ (A) and FXR−/− (B) are reported as % of controls (defined as duodenal expression of control diet-fed mice). Duodenum (closed bars), jejunum (open bars), ileum (hatched bars) and colon (squared bars) (n=3 in each group). (A) CA and UDCA significantly induce Mrp3 protein expression in duodenum and colon in FXR+/+. (B) In FXR−/− however both bile acids were unable to further induce basally elevated Mrp3 levels.
Journal of Hepatology  , DOI: ( /S (03) )

7. Fig. 6
Effects of CA and UDCA feeding on hepatic Bsep expression in FXR+/+ and FXR−/− mice. Total RNA and liver membranes were isolated from control diet-fed, CA- (1% w/w, 7 days), and UDCA- (1% w/w, 7 days) fed FXR+/+ and FXR−/− mice and analyzed by competitive RT–PCR (A) and Western blotting (B). In addition, Bsep tissue localization was assessed by confocal laser scanning microscopy (C–H) as described in Section 2. *CA- or UDCA-fed mice versus standard diet-fed mice (P<0.05). ¶Control diet-fed FXR−/− versus control diet-fed FXR+/+. #CA- or UDCA-fed FXR−/− mice versus CA- or UDCA-fed wild-type mice (P<0.05). (A) Steady-state mRNA levels (means±SEM) are reported as % of control diet-fed FXR+/+ mice. Control diet-fed (Control, open bars), CA-fed (CA, hatched bars) and UDCA-fed (UDCA, closed bars) mice (n=5 in each group). CA significantly induces Bsep mRNA expression in FXR+/+ mice whereas FXR−/− mice lack induction of Bsep. UDCA has no significant effect on Bsep mRNA expression in either genotype. (B) Representative immunoblots. Densitometry data are expressed as the fold change relative to standard diet-fed FXR+/+ mice. Values are the average obtained from 3–4 animals in each group. CA increases Bsep protein levels in FXR+/+ mice compared to standard diet fed controls. Standard diet-fed FXR−/− mice show significantly lower Bsep protein levels compared to wild-type controls and lack induction of Bsep expression by CA. Comparable data are obtained in UDCA-fed FXR−/− mice, which also lacked the induction of Bsep observed in wild-type animals. (C–H) Representative immunofluorescence images. Canalicular Bsep tissue localization (in red) was assessed by immunofluorescence staining. For better visualization of the hepatic architecture, the cytokeratin network (in green) was counterstained. Overlap of pericanalicular cytokeratin sheet with canalicular Bsep immunolabeling results in a yellow signal. Reduced canalicular Bsep staining in control diet-fed FXR−/− mice (F) compared to wild-type controls (C). Enhanced canalicular staining pattern in CA- and UDCA-fed FXR+/+ mice (D,E) compared to wild-type controls on standard diet. Lack of Bsep induction in CA- and UDCA-fed FXR−/− mice (G,H). Note also the increase in cell size and density of the cytokeratin intermediated filament network in CA-fed mice (D,G) compared to controls (C,F). Bar: 20 μm.
Journal of Hepatology  , DOI: ( /S (03) )

8. Fig. 7
CA feeding induces liver cell necroses in FXR−/− mice. Representative liver histology in control diet-fed, CA- (1% w/w, 7 days), and UDCA- (1% w/w, 7 days) fed FXR+/+ (A-C) and FXR−/− mice (D-F). (A,C) Standard diet- and UDCA-fed FXR+/+ mice show unremarkable liver histology. (B) Hepatocytes of CA-fed FXR+/+ mice are enlarged. (D,F) Control diet- and UDCA-fed FXR−/− mice with mild steatosis. (E) CA-feeding in FXR−/− mice results in disseminated but also confluent hepatocyte necroses (arrows). Original magnifications: ×10. cv, central vein; pv, portal vein; bd, bile duct.
Journal of Hepatology  , DOI: ( /S (03) )

9. Fig. 8
Adaptive changes in hepatic and renal ABC transporter expression in bile acid-fed mice. (A) Bile acid feeding in FXR wild-type mice. (1) Bile acid feeding results in induction of canalicular Bsep and Mrp2 as well as basolateral Mrp3 expression. Basolateral shunting of bile acids (BA) from hepatocytes via Mrp3 is followed by (2) increased urinary bile acid excretion which may at least in part be mediated via induced renal tubular Mrp2 expression. (3) Luminal bile acids also induce intestinal Mrp2 and – to a lesser extent – Mrp3 expression providing a defense mechanism of enterocytes against bile acid toxicity. (B) Bile acid feeding in FXR−/− mice. (1) Bile acids accumulate within hepatocytes of FXR−/− mice as a result of inadequate Bsep expression, despite a significantly stronger induction of basolateral Mrp3 and (2) renal Mrp2 expression when compared to bile acid-fed wild-type controls. These findings may explain the significantly higher serum bile acid levels and urinary bile acid output in bile acid-fed FXR−/− mice [16]. (3) Bile acids induce intestinal Mrp2 in FXR−/−, while no further induction of basally elevated Mrp3 is observed. The greater susceptibility to cholic acid-induced liver toxicity in these animals may be due to absent induction of canalicular Bsep expression as the major adaptive response mechanism of hepatocytes against bile acid-induced cytotoxicity.
Journal of Hepatology  , DOI: ( /S (03) )