A misexpression genetic screen in Drosophila identifies candidate axon guidance factors with differential expression

摘要

We are interested in the fundamental question of how genetic programs control the development of synaptic wiring in the brain. The precision with which growth cones navigate suggests that there is a heavy genetic component to this process, especially early in development. We have taken a genetic approach to identify genes involved in axon guidance. More specifically, we are interested in identifying those factors that provide us with an explanation of how it is that two seemingly similar growth cones can display disparate behaviors. The theme that has emerged from many areas of developmental biology is that differential expression of genes is key to creating diversity of behavior among cells. We designed a misexpression screen that would have a bias towards identifying differentially expressed genes over ubiquitously expressed genes. We have reported here a number of new candidate axon guidance molecules. Two of these, dTU3A and phat, we have examined in greater detail. Such analysis has shown that dTU3A is differentially expressed in the nervous system, but is unlikely to function in axon guidance based on experiments we performed using a loss-of-function allele of dTU3A and an antibody to dTU3A. phat (PH domain Axon Targeting molecule) gives a potent and specific axon guidance phenotype when misexpressed. Molecular examination of the locus revealed that phat encoded a protein with several splice variants, all containing a PH domain. PH domains are thought to interact with phosphoinositides in the plasma membrane. The primary sequence of phat's PH domain suggested that it may bind specifically to phosphatidylinositol (3,4,5)P3 (PiP3), a phosphoinositide generated by phosphatidylinositol 3-kinase (Pi3Kinase). We determined that the PH domain of Phat does localize to the plasma membrane, but not in response to Pi3Kinase signaling. However, we observe significant genetic interactions between phat and Pi3Kinase suggesting that at some level, there is an interaction. We generated phat loss-of-function alleles and have yet to find an axon guidance phenotype, possibly due to redundant factors or from not using a sensitive enough assay. However, we observe striking phenotypes in the early embryonic mesoderm, where phat is also expressed.