Loss of HNF6 results in normal apical-lateral localization of ZO-1, whereas loss of HNF1β results in very low levels of ZO-1 on the parenchymal side of forming bile duct lumen and improper apical localization of ZO-1 on the portal side. The postnatal consequences of these embryonic phenotypes that suggest loss of a cholangiocyte apical pole also differ between these two mouse models. Partial restoration of apical-basal polarity is observed when HNF6 is absent, but is not the case with HNF1β deficiency. Indeed, in patients with HNF1β mutations, ZO-1 was irregularly expressed in dysplastic ducts, and the observed DPM did not express ZO-1.

Absence of cystin-1 did not influence the apical marker osteopontin, but ZO-1 expanded to the apical surface of cholangiocytes, indicating that the basal and lateral poles were not established correctly. These phenotypes correspond PD0325901 ic50 to liver samples examined from ARPKD CT99021 datasheet fetuses. Because of the polarity defects observed during biliary tubulogenesis, the authors investigated whether cholangiocyte ciliogenesis was disrupted in these mouse models. Previously, HNF6 and HNF1β were implicated in either control of cilia formation or regulation of genes involved in cilia function in the pancreas and kidney, respectively.10, 11 Because of the

random distribution of centrioles ALOX15 observed in the absence of HNF6 or HNF1β, it is not surprising that embryonic cilia formation on cholangiocytes was significantly disrupted. The postnatal partial restoration of the apical-basal polarity in deficient HNF6 mice correlates with a few cilia present on cholangiocytes. However, the lack of cilia present on cholangiocytes remains as a postnatal defect in liver deficient for HNF1β. To determine if either HNF6 or HNF1β are involved in regulating the formation or function of cholangiocyte cilia, expression levels of candidate genes were examined in these two mouse models. Cystin-1 was the

only gene with reduced expression in both mouse models. Interestingly, in the cystin-1–deficient (cpk−/−) mouse model, the presence of a cilium is observed on some cholangiocytes. Therefore, reduced expression of cystin-1 in liver deficient in HNF6 and HNF1β is not the explanation for the reduced or absence of cilia in these two mouse models. Notably, and an avenue for further research, is the observation that the HNF1β targets in the kidney (Pkhd1, polycystic kidney and hepatic disease 1; Pkd2, polycystic kidney disease 2; Nphp1, nephronophthisis 1; IFT88, intraflagellar transport 88 homolog; and Kif12, kinesin family member 12) were not changed in HNF1β deficient liver, which indicates that HNF1β regulates a divergent transcriptional landscape for cilia in cholangiocytes versus kidney.