Many epithelial cell types are polarized along two axes — the apical–basal (A–B) axis and a perpendicular axis, which establishes planar cell polarity (PCP). The coordination between these polarities determines the positioning of structural features on the cells and is essential for tissue development. A report in Cell now sheds light on the molecular connection between the A–B axis and PCP in the Drosophila melanogaster eye.

Factors that determine PCP are localized apically in fly imaginal-disc cells. Among them, Frizzled (Fz1) partially colocalizes with several determinants of A–B polarity, including the PALS-1-associated tight junction protein (Patj)-containing complex and a complex that comprises Bazooka (Baz), Par6 and atypical protein kinase C (aPKC). Marek Mlodzik and colleagues showed that one of these apical factors, aPKC, phosphorylates the C terminus of Fz1. Overexpression of wild-type Fz1 or a phosphorylation-defective mutant caused a gain-of-function PCP phenotype, whereas a mutant that mimics phosphorylated Fz1 had a reduced gain-of-function phenotype. So, aPKC phosphorylation seems to inhibit Fz1 activity.

Next, the authors asked how aPKC is recruited to the C terminus of Fz1. As it turned out, it was through another apical determinant of A–B polarity, Patj. Overexpression of a Fz1 mutant that lacked the Patj-binding site caused a stronger gain-of-function PCP phenotype compared with overexpression of wild-type Fz1, which indicates that aPKC and Patj are required for the inhibition of Fz1-mediated PCP.

It followed that Patj and aPKC should be absent, or at least downregulated, in R3 and R4 photoreceptor precursor cells that undergo active PCP establishment. Indeed, immunofluorescence studies showed the reduced presence of Patj and aPKC, and an increased signal for apically localized Fz1, in these cells. The opposite pattern was found in the neighbouring cells that lack PCP activity. Interestingly, Baz also showed a complementary immunofluorescence pattern compared with Patj and aPKC, implying it might antagonize the negative regulation of Fz1 by aPKC and Patj. Support for this model came from the finding that the removal of one copy of Baz suppressed the gain-of-function PCP phenotype of overexpressed Fz1.

The authors showed that aPKC phosphorylation did not affect the localization of Fz1 or the recruitment of another PCP factor, Dishevelled (Dsh). So, how does aPKC regulate Fz1 activity? The answer remains unknown for now, but Mlodzik and colleagues suggest that Fz1 phosphorylation might inhibit PCP-specific signalling to Dsh or, alternatively, promote the destabilization or turnover of Fz1.