Potential conflict of interest: Nothing to report.
Hepatocytes as a source of cholangiocytes in injured liver
Article first published online: 18 DEC 2013
© 2013 by the American Association for the Study of Liver Diseases
Volume 59, Issue 2, pages 726–728, February 2014
How to Cite
Lemaigre, F. P. (2014), Hepatocytes as a source of cholangiocytes in injured liver. Hepatology, 59: 726–728. doi: 10.1002/hep.26673
- Issue published online: 29 JAN 2014
- Article first published online: 18 DEC 2013
- Accepted manuscript online: 8 AUG 2013 08:16AM EST
- Manuscript Accepted: 11 JUL 2013
- Manuscript Revised: 10 JUL 2013
- Manuscript Received: 11 MAY 2013
Robust cellular reprogramming occurs spontaneously during liver regeneration. Genes Dev 2013;27:719-724. (Reproduced with permission.), , , , , , et al.
Cellular reprogramming—the ability to interconvert distinct cell types with defined factors—is transforming the field of regenerative medicine. However, this phenomenon has rarely been observed in vivo without exogenous factors. Here, we report that activation of Notch, a signaling pathway that mediates lineage segregation during liver development, is sufficient to reprogram hepatocytes into biliary epithelial cells (BECs). Moreover, using lineage tracing, we show that hepatocytes undergo widespread hepatocyte-to-BEC reprogramming following injuries that provoke a biliary response, a process requiring Notch. These results provide direct evidence that mammalian regeneration prompts extensive and dramatic changes in cellular identity under injury conditions.
A few years ago differentiation of hepatic epithelial cells appeared as a relatively simple process: in homeostatic condition hepatocytes and cholangiocytes were considered to maintain their differentiation status throughout life; following injury the restoration of liver mass and function was believed to result from proliferation of preexisting cells or from differentiation of stem cells.
The notion that mature hepatocytes and biliary cells retain their identity throughout life is no longer an option. Michalopoulos et al. challenged the phenotypic stability of mature hepatocytes in liver: severe biliary injury in rats with chimeric livers carrying the hepatocyte marker dipeptidyl peptidase IV (DPPIV) led to an increased number of DPPIV-positive duct cells, suggesting that hepatocytes can transdifferentiate to cholangiocytes. Using a dynamic lineage tracing approach Yanger et al. now detected hepatocyte-to-cholangiocyte conversion in injured liver, thereby adding an important contribution to earlier work.
Several sources of cholangiocytes can be considered (Fig. 1). In homeostatic condition, cholangiocytes derive from the embryonic ductal plate and display little proliferation except during the first weeks of life in the mouse.[3, 4] Upon injury, cholangiocytes proliferate and self-renew, and when proliferation is inhibited progenitor cells give rise to cholangiocytes.[5-8] There is also evidence that peribiliary glands and stellate cells constitute cell reservoirs that can be stimulated to produce cholangiocytes.[9, 10]
Notch signaling is essential in embryonic liver for bile duct development and is critical in adult liver to drive progenitor cell differentiation to biliary cells.[7, 11] Yanger et al. now show that activation of Notch in adult hepatocytes is sufficient to convert the latter to cholangiocyte-like cells. Moreover, lack of functional Notch signaling inhibits the production of hepatocyte-derived biliary cells after injury. Importantly, in mouse models of cholangiocarcinoma Notch-dependent conversion of hepatocytes to the biliary lineage has been detected too, and combined activation of Notch1 and AKT signaling transdifferentiated hepatocytes to cholangiocarcinoma. Similarly, thioacetamide-induced cholangiocarcinoma originated from hepatocytes; this occurred without transgene-dependent activation of Notch signaling, and was repressed in the absence of the Notch mediator Hes1. Interestingly, Yanger et al. noticed that Notch-induced transdifferentiation did not occur in hepatocytes located near the central vein (zone 3), indicating that only hepatocytes located in zones 1 and 2 are specifically sensitive to Notch. Such specificity might rely on zonation-dependent gene expression but more subtle phenotypic cell-to-cell variations will need to be considered in the future.
How phenotypic conversion takes place is not yet clear. The transdifferentiating hepatocytes do not express α-fetoprotein, indicating that the cells do not transit by way of a hepatoblast-like stage before redifferentiating to cholangiocytes. Nevertheless, some dedifferentiation may occur: constitutive activation of Notch signaling in liver stimulates proliferation, and proliferation is normally associated with reduced expression of differentiation genes.[15, 16] The transcription factor SOX9 is a direct target of Notch signaling and may be a critical player in cell fate conversion. In pancreas, inflammation induces SOX9 expression in acinar cells and this contributes to repression of acinar genes and to acinar-to-duct metaplasia.[17, 18] Therefore, a similar mechanism of SOX9-mediated repression of hepatocyte genes and up-regulation of biliary genes may function as well in injured liver.
Yanger et al. resorted to the increasingly popular approach of lineage tracing which uses genetic marking as a way to permanently label a cell population and to track its descendants. Notably, Malato et al. followed a nearly identical approach but did not detect phenotypic conversion of hepatocytes. Yanger et al. calculated that 14% of CK19-positive cells derived from hepatocytes after a 6-week 3,5-diethoxycarbonyl-1,4-dihydrocollidine (DDC) treatment, and it is unlikely that such a percentage escaped the attention of Malato et al. Cell lineage analysis in liver has led to other discrepancies: the continuous and homeostatic cell supply of hepatic cells from SOX9-expressing progenitors, as proposed by Furuyama et al., was not supported by three distinct lineage tracing experiments[3, 6, 19]; several studies concluded that exposing the liver to DDC does not lead to detectable production of hepatocytes from progenitors.[5, 6] Genetic lineage tracing is a powerful tool and Yanger et al. handled it with care. The field eagerly awaits confirmation of the data and looks forward to translating these findings to the benefit of patients.
Frédéric P. Lemaigre, M.D., Ph.D.
Université Catholique de Louvain
de Duve Institute