Cytokeratin-19 positivity is acquired along cancer progression and does not predict cell origin in rat hepatocarcinogenesis.

Although the expression of the stem/progenitor cell marker cytokeratin-19 (CK-19) has been associated with the worst clinical prognosis among all HCC subclasses, it is yet unknown whether its presence in HCC is the result of clonal expansion of hepatic progenitor cells (HPCs) or of de-differentiation of mature hepatocytes towards a progenitor-like cell phenotype. We addressed this question by using two rat models of hepatocarcinogenesis: the Resistant-Hepatocyte (R-H) and the Choline-methionine deficient (CMD) models. Our data indicate that the expression of CK-19 is not the result of a clonal expansion of HPCs (oval cells in rodents), but rather of a further step of preneoplastic hepatocytes towards a less differentiated phenotype and a more aggressive behavior. Indeed, although HCCs were positive for CK-19, very early preneoplastic foci (EPFs) were completely negative for this marker. While a few weeks later the vast majority of preneoplastic nodules remained CK-19 negative, a minority became positive, suggesting that CK-19 expression is the result of de-differentiation of a subset of EPFs, rather than a marker of stem/progenitor cells. Moreover, the gene expression profile of CK-19-negative EPFs clustered together with CK-19-positive nodules, but was clearly distinct from CK-19 negative nodules and oval cells. i) CK-19-positive cells are not involved in the early clonal expansion observed in rat hepatocarcinogenesis; ii) CK-19 expression arises in preneoplastic hepatocyte lesions undergoing malignant transformation; iii) CK-19 positivity in HCCs does not necessarily reflect the cell of origin of the tumor, but rather the plasticity of preneoplastic cells during the tumorigenic process.