Search

Adult thorax formation inDrosophilabegins during pre-pupal development by fusion of its two contralateral progenitor halves, the heminotal epithelia (HE). HEs migrate and replace an underlying cell layer of thoracic larval epidermal cells (LECs) during a morphogenetic process called thorax closure. The LEC layer has so far been proposed to be a passive substrate over which HEs migrate before their zipping. By contrast, here we show that the pull forces generated within the LEC layer drive HE migration. During thorax closure, the LECs display actomyosin-mediated contraction, via enrichment of non-muscle myosin-II and actin, besides squamous-to-pseudostratified columnar epithelial transition and tissue shrinkage. This shrinkage of the LEC layer is further accompanied by cell extrusion and death, that prevent overcrowding of LECs, thereby promoting further shrinkage. The pull forces thus generated by the shrinking LEC layer are then relayed to the HEs by their mutual adhesions via βPS1 (Mys) and αPS3 (Scb) integrins. Suppression of cell death in the LEC layer by a gain of p35 leads to cell overcrowding, which impedes HE migration and zipping. Further, knockdown ofsqh, the light chain of non-muscle myosin II, in LECs or integrins (mysorscb) in either the LEC layer or in the HEs, or both abrogate thorax closure. Mathematical modeling also reveals the biophysical underpinnings of the forces that drive this tissue closure process wherein a degenerating LEC layer mediates its succession by the future adult primodia. These essential principles of thorax closure appear ancient in origin and recur in multiple morphogenetic contexts and tissue repair.

Culture of Drosophila expressing the steroid-dependent GeneSwitch transcriptional activator under the control of the ubiquitous α-tubulin promoter was found to produce extensive pupal lethality, as well as a range of dysmorphic adult phenotypes, in the presence of high concentrations of the inducing drug RU486. Prominent among these was cleft thorax, seen previously in flies bearing mutant alleles of the nuclear receptor Ultraspiracle and many other mutants, as well as notched wings, leg malformations, and bristle abnormalities. Neither the α-tubulin-GeneSwitch driver nor the inducing drug on their own produced any of these effects. A second GeneSwitch driver, under the control of the daughterless promoter, which gave much lower and more tissue-restricted transgene expression, exhibited only mild bristle abnormalities in the presence of high levels of RU486. Coexpression of the alternative oxidase (AOX) from Ciona intestinalis produced a substantial shift in the developmental outcome toward a wild-type phenotype, which was dependent on the AOX expression level. Neither an enzymatically inactivated variant of AOX, nor GFP, or the alternative NADH dehydrogenase Ndi1 from yeast gave any such rescue. Users of the GeneSwitch system should be aware of the potential confounding effects of its application in developmental studies.