Hepatoblasts are bipotential progenitors that are destined to become either hepatocytes or cholangiocytes in the liver. Hepatocytes exhibit a unique cell polarity, called multiaxial polarity, which means that these cells have multiple apical and basal surfaces. This multiaxial polarity enables hepatocytes to form multiple apical connections with adjacent hepatocytes, allowing hepatocytes to form the liver's three-dimensional network of tubular bile canaliculi (BC). Knockdown of the small GTPase Rab35 leads to the loss of the hepatocyte-specific multiaxial polarity, causing the hepatocytes to resemble cholangiocytes. Here, Marino Zerial and colleagues examine whether cell polarisation and/or lumen morphogenesis impact the genetic programmes underlying the decision between the hepatocyte or cholangiocyte cell fate. They first confirm that knockdown of Rab35, or of its associated factors, impairs hepatocyte polarity. They find that Rab35 knockdown also impairs hepatoblast-to-hepatocyte differentiation through the downregulation of hepatocyte-specific genes and the upregulation of cholangiocyte-specific genes. Furthermore, these transcriptional changes can be recapitulated by perturbing the morphology of the BC lumens. Overall, this work demonstrates that cell polarity and anisotropic expansion of BC provide feedback to hepatocyte differentiation.
