Embryonic cholecystitis and defective gallbladder contraction in the Sox17-haploinsufficient mouse model of biliary atresia

Abstract

The gallbladder excretes cytotoxic bile acids into the duodenum through the cystic duct and common bile duct system. Sox17 haploinsufficiency causes biliary atresia-like phenotypes and hepatitis in late organogenesis mouse embryos, but the molecular and cellular mechanisms underlying this remain unclear. In this study, transcriptomic analyses revealed the early onset of cholecystitis in Sox17^+/- embryos, together with the appearance of ectopic cystic duct-like epithelia in their gallbladders. The embryonic hepatitis showed positive correlations with the severity of cholecystitis in individual Sox17^+/- embryos. Embryonic hepatitis could be induced by conditional deletion of Sox17 in the primordial gallbladder epithelia but not in fetal liver hepatoblasts. The Sox17^+/- gallbladder also showed a drastic reduction in sonic hedgehog expression, leading to aberrant smooth muscle formation and defective contraction of the fetal gallbladder. The defective gallbladder contraction positively correlated with the severity of embryonic hepatitis in Sox17^+/- embryos, suggesting a potential contribution of embryonic cholecystitis and fetal gallbladder contraction in the early pathogenesis of congenital biliary atresia.

Keywords: Biliary atresia; Gallbladder; Shh; Sox17.

Fig. 1.

Sox17^+/− gallbladder epithelium exhibits cystic duct-like phenotypes. (A) Transverse sections [Hematoxylin and Eosin (HE) staining] of the gallbladder and cystic duct from wild-type (wt) and Sox17^+/− mouse embryos at 17.5 dpc. The levels of the sections are indicated in A′ (arrowheads). Insets show high magnification images of the epithelium. (B) Quantitative analyses show a significant reduction in epithelial height and epithelial/cavity areas of the gallbladder in Sox17^+/− embryos at 17.5 dpc. *P<0.05, **P<0.01, ANOVA followed by Tukey's test. (C) Anti-SOX17 and anti-SOX9 double immunohistochemistry in the primordial gallbladder and cystic duct regions of the wild-type embryo at 13.5 dpc. (D) Sox17^+/− embryos display ectopic SOX9-positive cells in the distal edge of the gallbladder at 13.5 and 15.5 dpc. Bracket indicates the SOX9-positive domain. (E) qPCR analysis indicates a significant increase in Sox9 mRNA in the Sox17^+/− gallbladder at 15.5 dpc. *P<0.05, Student's t-test. In B and E, sample number is indicated within each bar. cd, cystic duct; duo, duodenum; gb, gallbladder; hd, hepatic duct; pd, pancreatic duct; pv, portal vein; D, distal; P, proximal. Scale bars: 100 µm.

Fig. 2.

Global expression profiles and cholecystitis of fetal Sox17^+/− gallbladders. (A,B) Microarray analysis identifies 279 downregulated genes (blue circle in A) and 501 upregulated genes (red circle in B) in the Sox17^+/− gallbladder at 15.5 dpc, as compared with gallbladder from wild-type (wt) littermates. Among these downregulated or upregulated genes, 207 (74.4%) or 221 (44.1%) were shared with the 5361 gallbladder-specific or 4886 cystic duct (cd)-specific genes, respectively, in wild-type embryos at the same stage. (C,D) qPCR analyses showing the upregulated profiles of the biliary disease marker genes Olfm4, Abcb4 and Pkhd1 (C) and the inflammatory marker genes Cxcl10 and Serpine1 (D) in Sox17^+/− gallbladder at 17.5 dpc. The y-axis is fold change in expression level of each gallbladder or liver sample relative to those of the wild type (mean value set to 1). Each solid or open circle indicates a sample from an embryo with (severe phenotype) or without (mild phenotype) gross anatomical hepatic lesions, respectively. Bar charts show the average mean values of all (severe and mild) samples. The expression level of each marker gene in the Alb-cre/Sox17^+/+ and Alb-cre/Sox17^flox/+ embryos at the same stage is also shown. *P<0.05, **P<0.01, Mann-Whitney U-test. (E) Spearman rank correlation analysis of the association between gallbladder (y-axis) and liver (x-axis) phenotypes for the indicated genes in Sox17^+/− embryos at 17.5 dpc. Both axes represent fold change in expression relative to wild-type littermates (mean value set to 1). In Sox17^+/− embryos, significant correlations between the two tissues were detected in both the severe phenotype and all (mild and severe phenotype) samples for Cxcl10 and Serpine1 mRNA levels, and in only the severe phenotype samples for Olfm4.

Fig. 3.

Embryonic hepatitis can be induced by conditional Sox17 deletion in primordial gallbladder but not in liver hepatoblasts. (A) Fluorescent images of gallbladder (gb), cystic duct (cd) and liver at 17.5 dpc, showing a liver-specific or gallbladder-specific fluorescent switch (from red to green) in Alb-cre/ROSA^mTmG or Pdx1-cre/ROSA^mTmG embryos, respectively. The Pdx1-cre/ROSA^mTmG embryo shows recombination in approximately half of the gallbladder and cystic duct epithelia by 17.5 dpc. (B) Caudal views of the fetal liver at 17.5 dpc. Peripheral inflammation of the liver lobules (arrowheads) is evident in two Pdx1-cre/Sox17^flox/flox embryos (top, type I; bottom, type II; see D), but not in the Alb-cre/Sox17^flox/flox embryo. Normal healthy livers (from wild-type and Sox17^flox/flox embryos) and a typical defective liver (Sox17^+/− embryo) are also shown. (C) qPCR analyses showing upregulation of the inflammatory marker genes Cxcl10 and Serpine1 in three Pdx1-cre/Sox17^flox/flox livers with appreciable gross anatomical lesions (solid circles). Expression data of wild type, Sox17^flox/flox and Alb-cre/Sox17^flox/flox are also shown. *P<0.05, Mann-Whitney U-test. (D) DBA staining showing phenotypic variation of the gallbladder and cystic duct in individual Pdx1-cre/Sox17^flox/flox embryos: type I (normal length of the gallbladder and cystic duct, 34/55 embryos), type II (half-length, 10/55) and type III (lack of the gallbladder and cystic duct, 11/55). duo, duodenum. Scale bars: 200 µm.

Fig. 4.

Expression profiles of Shh and its downstream genes in the developing gallbladder. (A) Whole-mount in situ hybridization (top) and section (bottom; at the level of the gallbladder, indicated by the dotted line) images, showing the expression of Sox17, Shh and Shh-related genes in wild-type embryos at 11.5 dpc. Sox17 and Shh are highly expressed in the gallbladder epithelia, in contrast to the mesenchymal enrichment for Ptch1, Gli1 and Hhip. (B) Whole-mount GFP fluorescence (left) and anti-GFP-stained sections (three right images) of Shh^+/−(GFP) and wild-type embryos at 16.5 dpc, showing GFP/Shh-positive epithelial signals in the gallbladder and cystic duct, but not in the hepatic ducts. Ducts are outlined. (C) qPCR analyses showing expression levels of Shh and related genes in Sox17^+/− embryos. *P<0.05, Student's t-test. Sample number is indicated within each bar. cd, cystic duct; duo, duodenum; gb, gallbladder; hd, hepatic duct. Scale bars: 100 µm.

Fig. 5.

Sox17 and Shh coordinately regulate the proper formation of smooth muscles in the gallbladder. (A) Whole-mount DBA staining of wild-type, Shh^+/− and Shh^−/− liver with gallbladder and cystic duct. (B) Anti-SOX17 and anti-PCNA staining of serial transverse sections of wild-type and Shh^−/− gallbladder at 14.5 dpc. (C) Quantitative analysis of the PCNA index (percentage PCNA-positive cells among total cells) in the gallbladder epithelium shows no differences among wild-type, Shh^+/− and Shh^−/− gallbladders at 14.5 dpc. (D) HE and anti-SOX9 staining in transverse sections of fetal gallbladders of wild-type, Shh^+/−, Sox17^+/− and Sox17^+/−;Shh^+/− embryos at 17.5 dpc. (E) Quantitative data from wild-type and Sox17^+/− embryos at 17.5 dpc show no appreciable alteration in epithelial height of the gallbladder upon loss of one Shh allele. (F,G) Anti-SMA staining on transverse sections of the gallbladder in Sox17^+/−, Shh^+/−, Shh^−/− and Sox17^+/−;Shh^+/− embryos at 14.5 dpc (F) and 17.5 dpc (G). SMA-positive cells were patchy in the gallbladder mesenchyme and severely reduced in the Shh^−/− gallbladder at 14.5 dpc. At 17.5 dpc, proper formation of the SMA-positive muscle layer was evident in Shh^+/− embryos, whereas SMA-positive signals remained patchy in Sox17^+/−, Shh^+/− and Sox17^+/−;Shh^+/− gallbladders. (H) Quantitative data of SMA-positive signals show a gradual reduction in the Shh^+/−, Sox17^+/− and Sox17^+/−;Shh^+/− gallbladders. Insets are higher magnification images of SMA-positive cells in the gallbladder epithelia (B,D) and its surrounding mesenchyme (F,G). *P<0.05, **P<0.01, ANOVA followed by Tukey's test. Sample size is indicated within each bar. Scale bars: 100 µm.

Fig. 6.

Defective contraction of smooth muscles in Sox17^+/− gallbladder. (A,B) Whole-mount anti-SMA staining of wild-type and Sox17^+/− gallbladders at 17.5 dpc. (Top) Confocal z-projection from the area of the gallbladder indicated by the dashed box (left). (Bottom) Orthogonal projection in the xz plane made at the level of the arrowhead (above). The orientation of the discontinuous SMA-positive cells was randomized in the Sox17^+/− gallbladder, in contrast to the dense SMA-positive fibers arranged mainly in a circular direction in the wild-type gallbladder. D, distal; P, proximal. (C,D) Dissection microscope images of the fetal gallbladder before (top) and after (bottom) KCl treatment (C) and KCl-induced contraction levels (relative changes in luminal diameter; D) of wild-type and the Sox17^+/− gallbladder with (severe) or without (mild) gross anatomical hepatic lesions at 17.5 dpc. KCl treatment reduced luminal diameter considerably in the wild-type but not in the Sox17^+/− gallbladder, with the severe phenotype group being particularly unaffected. Number of samples (gallbladders) are indicated. **P<0.01, ANOVA followed by Tukey's test. Scale bar: 100 µm.

Fig. 7.

Defective smooth muscle development of Sox17^+/− gallbladders is rescued by exogenous SHH. (A,B) Three day organ culture of wild-type or Sox17^+/− gallbladder primordium initiated at 13.5 dpc (A). The gallbladder and bead are held inside a single cylindrical groove of agarose gel as schematically represented in B. D, distal; P, proximal. (C-F) HE, anti-SMA, anti-PCNA or anti-SOX9 staining on transverse sections of wild-type or Sox17^+/− gallbladder explants with PBS-soaked or SHH-soaked beads. Insets are higher magnification images of the epithelium or the signals. (G-I) Quantitative morphometric analysis of wild-type and Sox17^+/− gallbladder explants cultured with PBS-soaked or SHH-soaked beads. Exogenous SHH signaling rescued the defects in epithelial height and cavity area, but did not exert any significant influence on epithelial area (G), PCNA index (H) or relative SOX9-positive cell numbers (I) within the gallbladder epithelium. *P<0.05, **P<0.01, ANOVA followed by Tukey's test. Sample number is indicated with each bar. Scale bars: 100 µm.