Abstract
B cells can promote liver fibrosis, but the mechanism of B cell infiltration and therapy against culprit B cells are lacking. We postulated that the disruption of cholangiocyte-B-cell crosstalk could attenuate liver fibrosis by blocking the CXCL12-CXCR4 axis via a cyclooxygenase-2-independent effect of celecoxib. In wild-type mice subjected to thioacetamide, celecoxib ameliorated lymphocytic infiltration and liver fibrosis. By single-cell RNA sequencing and flow cytometry, CXCR4 was established as a marker for profibrotic and liver-homing phenotype of B cells. Celecoxib reduced liver-homing B cells without suppressing CXCR4. Cholangiocytes expressed CXCL12, attracting B cells to fibrotic areas in human and mouse. The proliferation and CXCL12 expression of cholangiocytes were suppressed by celecoxib. In CXCL12-deficient mice, liver fibrosis was also attenuated with less B-cell infiltration. In the intrahepatic biliary epithelial cell line HIBEpiC, bulk RNA sequencing indicated that both celecoxib and 2,5-dimethyl-celecoxib (an analog of celecoxib that does not show a COX-2-dependent effect) regulated the TGF-β signaling pathway and cell cycle. Moreover, celecoxib and 2,5-dimethyl-celecoxib decreased the proliferation, and expression of collagen I and CXCL12 in HIBEpiC cells stimulated by TGF-β or EGF. Taken together, liver fibrosis can be ameliorated by disrupting cholangiocyte-B cell crosstalk by blocking the CXCL12-CXCR4 axis with a COX-2-independent effect of celecoxib.