IMP is a conserved RNA binding protein involved in vertebrate cell migration and cancer metastasis. Drosophila IMP is required for border cell migration and follicle cell morphogenesis during oogenesis. Later on, imp null embryos do not complete embryogenesis and fail to undergo dorsal closure. One of the working models for IMP function is that it localizes actin mRNA to the leading edge of migrating cells. mRNA in situ for the two cytoplasmic actins in Drosophila (actin42A and actin5C) shows that both are localized to the front of the migrating border cell cluster as is IMP protein. Border cells mutant for either actin42A or actin5C are able to perform normal migration supporting the idea that these two genes are redundant. In order to study the localization of these mRNAs at the subcellular level in live samples I have used the MS2 system. Nevertheless the high level of background has hindered the study in border cells and led me to focus on the follicle cell phenotype. Follicle cells localize both IMP protein and actin5C mRNA at the apical cortex and this localization is microtubule dependent. The defect in the movement of imp null follicle cells to cover the oocyte has its root in their inability to respond promptly to Notch signaling at earlier stages. Consequently cells undergo a further round of replication. I hypothesize that this defect in responding to Notch is due to the smaller apical microvilli of imp null cells and hence fewer points of contact with the signal sending cell, the oocyte. Localization of actin mRNA at the apical side by IMP could play an important role in allowing for the extensive actin polymerization needed to form microvilli. During dorsal closure neither actin mRNA nor IMP protein are localized and instead show a ubiquitous distribution in all the tissues involved. Since IMP orthologs have been reported to play a role in regulating not only localization but also mRNA translation and stability of very many different RNAs it is likely that the mechanism used to regulate dorsal closure is distinct from the one used during oogenesis. 2