Contribution of neural crest and GLAST + Wnt1 + bone marrow pericytes with liver fibrogenesis and/or regeneration.

Background and aims Liver fibrosis results from cycles of liver damage and scar formation. We herein aimed at analysing neural crest cells and/or bone marrow stromal cells contribution to the liver. Methods Two liver fibrosis and one hepatectomy model were applied on double-transgenic loxP-Cre mouse lines. Results Increased numbers of glia with more complex processes were found in fibrotic livers. During embryonic development, only few cells were traced in the liver and bone marrow, in a minor fraction of mice of different neural crest reporter strains analysed: therefore, a neural crest origin of such cells is doubtful. In the fibrotic liver, a significantly higher incidence of endothelial cells and hepatocyte-like cells expressing the reporter gene Tomato were found in Wnt1-Cre-Tom and GLAST-CreERT2-Tom mice. Consistently, during early fibrogenesis stromal Wnt1-traced cells, with progenitor (CFU-F) properties, get likely mobilized to peripheral blood. Circulating adult Wnt1-traced cells are stromal cells and lack from the expression of other bone marrow and endothelial progenitor cells markers. Furthermore, in a 70% hepatectomy model GLAST(+) Wnt1-traced pericytes were found to be mobilized from the bone marrow and the incidence of GLAST-traced hepatocyte-like cells was increased. Finally, GLAST-traced hepatocyte like-cells were found to maintain the expression of stromal markers. Conclusions Our data suggest a gliosis process during liver fibrogenesis. While neural crest cells probably do not contribute with other liver cell types than glia, GLAST(+) Wnt1-traced bone marrow pericytes are likely a source of endothelial and hepatocyte-like cells after liver injury and do not contribute to scarring.