A role for strain-based mechanical feedback in establishing the myosin distribution driving Drosophila germ-band extension
The problem of understanding the complex sequence of morphogenetic processes that transform a single cell zygote into a fully formed and functional organism is at the heart of the field of developmental biology. Early progress in the field relied on detailed observations of the developing embryo to produce qualitative descriptions of the various morphogenetic processes, establishing a rough timeline for development and the relative sequence of events. Later, extensive genetic screens allowed researchers to identify critical genes in morphogenetic processes based on the observed defects, and with many iterations carefully piece together the genetic cascades that direct development. Nevertheless, shape is inherently a problem of mechanics, and therefore a thorough view of morphogenesis requires an understanding of the intimate interplay between genetics and mechanics. Although an appreciation for this role of the physical aspects of developmental biology predate our knowledge of genetics, its characterization has been much slower that its genetic counterpart. Perhaps best characterized genetically is the example of the embryogenesis of Drosophila melanogaster, which therefore offers an ideal context within which to study the role of mechanics. In this dissertation, we aim to take advantage of the detailed genetic foundation to better understand the role of mechanics in development, in particular during the period of early gastrulation in the fruit fly. Extensive studies of the genetics and preliminary characterization of the mechanics have been undertaken for the morphogenetic processes associated with this developmental stage, namely ventral furrow formation and germ-band extension. We aim to build off this previous work utilizing in toto multi-view light sheet microscopy to image global dynamics of the developing embryo and take advantage of image processing techniques to allow rigorous quantitative analysis of strain and myosin rates. With these methods, we identify a strong correlation between strain and myosin dynamics across the embryo surface. We confirm the relationship at the level of single junctions with high spatiotemporal resolution confocal microscopy and demonstrate that a simple physical model with mechanical feedback as a key feature recapitulates the single junction observations. We next utilize an optogenetic construct to pattern actomyosin contractility in order to induce physiologically relevant and minimally invasive ectopic strain rates and measure the resulting changes to the myosin dynamics. By altering the pattern of activation, we show that not only does a strain based mechanical feedback mechanism recruit myosin in vivo, but we further show that this mechanism is isotropic within the cell but spatially modulated along the D-V axis. We then adapt our model to predict the myosin distribution at the tissue scale and show that incorporating this D-V modulation of feedback strength provides an accurate prediction of the myosin profile. Finally, we demonstrate the importance of this feedback mechanism for normal development by analyzing genetic mutants defective for a major strain generating process: ventral furrow formation. We find that these mutants have significantly reduced strain rates and proportional reductions in the myosin distribution. Notably, these embryos have an equivalent reduction in the rate of germ-band extension, which is known to depend on the myosin distribution, indicating the developmental relevance of this mechanism.