Gene expression profiles of sclerosing cholangitis activity in mice

Background. Abcb4 (-/-) mice secrete phosphatidylcholine-deficient bile and develop sclerosing cholangitis (SC). The cholangitis involves differential hepatic transcription of genes whose products govern inflammation, activation of hepatic stellate cells and fibrosis. This study was undertaken to test the hypothesis that several genes involved in regulation of tissue inflammation and fibrosis display transcription rates that reflect SC disease activity. Methods. Abcb4 (-/-) mice fed cholic acid (CA) display high SC activity and ursodeoxycholic acid (UDCA) fed mice display low SC activity. Differential hepatic transcription of genes was accordingly measured in abcb4 (-/-) mice maintained on CA- and UDCA-supplemented diets using cDNA microarrays. Abcb4 (+/+) mice served as controls. The differential transcription of selected genes was verified by real time polymerase chain reaction. Liver tissue pathology was quantified by histopathology scoring and immunohistochemistry to visualize bile duct cells and activated hepatic stellate cells. Results. Differential transcription of Ccl2, Ccl20, Cxcl10, Nfκb1, Nfκb2, Tgfβ1, Tgfβ2, Sparc, Ctgf, Lgals3, Elf3, Spp1, Pdgfa, Pdgfrb, Col1a1, Col1a2 and Col4a1 genes paralleled the differing SC activities of cholic acid- and UDCA-fed abcb4 (-/-) mice. Histopathology scores and immunohistochemistry showed greatly enhanced activation of hepatic stellate cells during high SC activity due to CA feeding. Conclusion. Differential transcription of several genes relating to tissue inflammation and hepatic stellate cell activation parallels SC activity in abcb4 (-/-) mice. Data on their differential transcription may be used to gauge SC disease activity. To achieve different degrees of disease activity we have fed the mice with either cholic acid (CA) or ursodeoxycholic acid (UDCA) for 9 weeks to increase and decrease disease activity, respectively. Each group uses wild type mice as controls. 5 different mice (livers) in each group was harvested. In 3 hybridizations wild-type livers were marked with Cy3, in 2 hybridizations wild-type livers were marked with Cy5. In addition to these 5 hybridizations one dye swap was performed from a randomly selected liver from the 5 and wild-type marked with Cy5. Thus, a total of 6 hybridizations was used as input for data-processing in each of the two groups. No pooling was done.