The field of axon guidance is concerned with deciphering how axons use temporal and spatial cues to pattern a precise trajectory. We use the fruit fly Drosophila melanogaster to study embryonic pathway decisions that are fundamental in connecting the bilaterally symmetrical halves of the nervous system at the ventral midline. At the midline choice point, axons make a binary decision regarding whether or not to cross to the opposite side. One signaling pathway controlling midline crossing is Slit-Roundabout repulsive signaling, whereby axons expressing the Roundabout (Robo) receptor bind the midline-expressed chemorepellant ligand, Slit, and sense the midline as inhibitory. Expression of Commissureless (Comm)—and its post-translational re-localization of Robo—effectively silences Slit-Robo signaling, allowing axons to cross the midline.
In addition to comm, there exist two other Comm family members: comm2 and comm3. We utilized cell culture and in vivo approaches to learn about comm2, whose spatiotemporal expression is similar to comm during embryogenesis. Like Comm, Comm2 re-localized full-length Robo in S2-based cell culture assays but required sequences in Robo’s extracellular (EC) and/or cytoplasmic (cyto) domain not required by Comm. Additionally, deletion of Comm2’s EC domain abrogated its Robo re-localization ability. In flies, both molecularly-defined deletion of Comm2 and Comm2 overexpression in wildtype had subtle phenotypes, but its expression in comm and comm,comm2 deletion backgrounds revealed a positive role in midline crossing. Interestingly, comparison of Comm2 expression in backgrounds deleted for comm, comm2, or both, revealed a possible inhibitory role for Comm2 at the midline.
We also created a series of chimeric fusion proteins among Comm family members. Analysis of these constructs supports an unexpected role for the Comm EC domain in regulating Robo. In the S2 cell culture assay, as little as the Comm EC domain was capable of transforming otherwise non-functional Comm2 and Comm3 transmembrane (TM) and cytoplasmic domains into Robo re-localizing proteins. This affect translated to Comm-like functionality when tested in vivo as well. Furthermore, although full-length Comm2 induced Robo re-localization in cell culture, the presence of its domains in various fusion proteins was unable to promote Robo re-localization. We also focused on Comm3, which has no Comm-like functionality, despite having a conserved TM domain and PY motifs. Its EC domain is not inhibitory to Robo re-localization; rather, it may lack the positive sequences present in Comm, making it a good tool to use for understanding Comm and Comm2 function. Its TM/cyto domains are, however, able to function when combined with a Comm EC domain.
Finally, we investigated the mechanism by which Comm regulates Robo, with a focus on ubiquitination of Comm and/or Robo. In contrast to other publications, we did not find a necessary role for Comm ubiquitination in Comm-mediated Robo re-localization in S2 cells or in vivo, as it behaved like wildtype Comm in both environments. On the other hand, in S2 cells, Comm less efficiently re-localized a short, lysine-free Robo construct, but some re-localization still occurred. When both were lysine-free, again, some re-localization still occurred, suggesting that ubiquitination may influence the process but is not absolute.