Apicobasal polarity is a key feature of epithelial cells across the animal kingdom, and its establishment and maintenance are crucial for epithelial integrity. Over recent years, Drosophila has proved a fruitful model for investigating the molecular and cellular basis of epithelial polarity during development. Two papers in this issue of Development now provide further insights into how apicobasal polarity is established in the early fly embryo.

Both papers focus on the actin-binding protein Canoe (Cno), which is required for apicolateral restriction of the polarity protein Bazooka (the homologue of Par3), and hence for proper formation of adherens junctions, during cellularisation of the embryo. Previous work from the lab of Mark Peifer has shown that Cno localises to nascent adherens junctions in a Rap1-dependent manner. Now (dev157941), Peifer and colleagues provide further insights into how Cno localisation is controlled, and how this then affects polarity establishment and maintenance. They show that appropriate Cno localisation and Bazooka recruitment depend on apically restricted Rap1 activity, implicating a potential requirement for a localised guanine exchange factor (GEF) or GTPase-activating protein (GAP). Consistent with this, they uncover a role for the GEF Dizzy, which is itself localised to the apical side of the cell membrane. However, Dizzy appears to be responsible only for the localisation of Cno to tricellular junctions, and not for its recruitment to the subapical domain more generally. Moreover, neither Dizzy nor Rap1 seems to be essential for the maintenance of Cno at junctions after gastrulation. Thus, multiple GTPase regulators, as well as other factors, might be involved in the control of Cno localisation in this context.

Consistent with this model, a second study (dev157909) from Jörg Großhans and co-workers shows that the unconventional GEF complex ELMO-Sponge is required for Cno localisation during early cellularisation. Upon depletion of ELMO or sponge, Cno is no longer restricted to the subapical domain and spreads along the lateral furrow. Prior to cellularisation, Sponge and ELMO are observed apically in a disc-like pattern, but at the onset of cellularisation become redistributed into a subapical ring-like organisation. The authors propose that this transition in localisation is responsible for the rapid switch in Cno localisation at the same stage, whereby Cno segregates away from more lateral markers and into the subapical domain. Although not directly shown, the authors propose that ELMO-Sponge may act via Rap1 to achieve this localisation of Cno, suggesting that ELMO-Sponge might be (one of) the alternative regulator(s) proposed by the Peifer lab study.

Together, these two papers add to our understanding of how polarity is established in the early Drosophila embryo, and demonstrate the complexity of regulatory processes involved in setting up and maintaining apicobasal polarity in epithelia.