Human Usher syndrome is a severe and congenital form of syndromic deafness that affects 1 person in 25,000 people in the world population. Normally the stereocilia, microvillar protrusions of the apical membrane of inner ear hair cells, are organized into coherent bundles. This precise organization is critical for mechanosensing, i.e. for hearing. Mutation in any of the five known Usher syndrome genes is sufficient to alter the precise organization of stereocilia, a condition that results in deafness. To date, however, the molecular mechanisms responsible for the splaying of stereocilia and genesis of the disease are not well understood. Here, I identified Drosophila melanogaster genes related to human Usher syndrome and characterized some of them (Cad99C, DSANS and crinkled) during Drosophila development, in the processes of microvilli morphogenesis in the follicular and wing imaginal disc epithelia. Cadherin Cad99C is a transmembrane protein with putative cell adhesion properties. Similar to its human ortholog Protocadherin 15, Drosophila Cad99C localizes to microvillar protrusions in the follicular epithelium. In this epithelium, Cad99C is required for the proper morphogenesis and organization of microvilli into bundles, similar to human Protocadherin 15. Further, overexpression of the full-length Cad99C or of a deleted version, devoid of the cytoplasmic region, promotes microvilli bundling. This finding suggests that Cad99C establishes adhesive interactions between microvilli via its extracellular region. Interestingly, morphological alteration of follicle cell microvilli associates with defective deposition of the vitelline membrane, an extracellular matrix that protects the embryo from osmotic stresses. These findings suggest that microvilli are normally required for the even deposition of the extracellular matrix. In order to test whether Cad99C is involved in microvilli morphogenesis and bundling in other tissues, I analyzed the function of Cad99C in a larval tissue, the wing imaginal disc. Cad99C overexpression, but not Cad99C removal, is sufficient to alter microvilli morphology and organization in the columnar epithelium of the wing imaginal disc. Likely, other molecules can compensate for Cad99C loss of function in this tissue. To possibly get some insights on the molecular function of other Usher syndrome proteins, I analyzed the function of Drosophila SANS and crinkled in the follicular epithelium, where both these genes are expressed. crinkled is the ortholog of myosinVIIa, that encodes a motor protein of the actin cytoskeleton. DSANS is related to human SANS and encodes a cytoplasmic protein of unknown function. It has been puzzling how removal of SANS, a cytoplasmic protein, could impair adhesion and bundling of stereocilia. To study the function of DSANS, I generated null mutant flies and observed that, in the absence of DSANS, delivery of Cad99C to microvilli is impaired. Cad99C localization is however unperturbed in crinkled mutant follicle cells. By immunostaining, DSANS immunoreactivity was detected diffusively in the cytoplasm and in dot-like structures, possibly corresponding to vesicles. In conclusion, DSANS is a cytoplasmic protein that is required for the efficient delivery of Cad99C to microvilli protrusions. Taken together, the analysis that I here performed of Drosophila Usher syndrome related genes indicates two novel molecular mechanisms of function for the corresponding human Usher syndrome proteins. First, human Protocadherin 15, like Drosophila Cad99C, could be involved in establishing adhesive interactions between microvilli protrusions of the inner ear (stereocilia). Removal of Protocadherin 15 would then cause splaying of stereocilia due to lack of inter-stereocilia adhesive links. Second, the analysis here performed suggests that SANS is involved in the efficient delivery of Protocadherin 15 to stereocilia. Mutations in SANS would then lead to splaying of stereocilia and deafness due to poor localization of Protocadherin 15 to stereocilia.

Developing biological systems can be approximately described as complex, three dimensional cellular assemblies that change dramatically across time as a consequence of cell proliferation, differentiation and movements. The presented project aims to overcome problems of limited resolution in both space and time of classical analysis by in situ hybridization on fixed tissue. The employment of the newly developed Single Plane Illumination Microscopy (SPIM) combined with new approaches for in vivo data acquisition and processing promise to yield high-resolution four-dimensional data of the complete Drosophila embryogenesis. We developed a toolkit for high-throughput gene engineering in flies, that provides means for creating faithful in vivo reporters of gene expression during Drosophila melanogaster development. The cornerstone of the toolkit is a fosmid genomic library enabling high-throughput recombineering and φC31 mediated site-specific transgenesis. The dominant, 3xP3-dsRed fly selectable marker on the fosmid backbone allows, in principle, transgenesis of the fosmid clones into any non-melanogaster species. In order to extend the capabilities of the gene engineering toolkit to include “evo-devo” studies, we generated genomic fosmid libraries for other sequenced Drosophilidae: D. virilis, D.simulans and D. pseudoobscura. The libraries for these species were constructed in the pFlyFos vector allowing for recombineering modification and φC31 transgenesis of non-melanogaster genomic loci into D. melanogaster. We have developed a PCR pooling strategy to identify clones for a specific gene from the libraries without extensive clone sequencing and mapping. The clones from these libraries will be primarily used for cross-species gene expression studies. As another application, transgenes originating from closely related species can be used to rescue D. melanogaster RNAi phenotypes and establish their specificity. Together with SPIM microscopy, the toolkit will allow to visualize gene expression patterns throughout Drosophila development.