Multi-level analysis of regulation of EGFR signalling during Drosophila melanogaster leg proximal-distal axis patterning

A major pursuit of Developmental Biology is to determine how organisms composed of cells containing a single genome generate stereotyped body plans with diverse, complex morphologies. The development of these patterns is often determined by gradients of secreted factors known as morphogens, which activate cascades of gene expression to subdivide fields of cells into increasingly complex patterns. In many animals, including Drosophila, a rudimentary anterior-posterior (A-P) and dorsal-ventral (D-V) axes of the body plan are already established in the zygote, but the proximal-distal (P-D) axis of any appendages must be generated and patterned seperately. The spatio-temporal information responsible for activating gene expression and cell signalling that establishes this new axis is integrated at DNA regulatory elements often referred to as enhancers. The segmented leg of the insect Drosophila melanogaster offers an ideal system for studying how signalling pathways control P-D axis establishment and patterning. In addition to the fact that flies are a particularly genetically tractable model organism, many of the signals required for leg patterning have already been identified. A number of signalling pathways, including Wingless (Wg), Decapentaplegic (Dpp) and Epidermal Growth Factor Receptor (EGFR), are important for proper P-D axis patterning in a dynamic fashion during embryonic and larval development. The leg primordia are fist specified in the embryo and then patterned throughout development as intercalated circles and rings of gene expression are established in the leg imaginal disc. The radius of these domains corresponds to the P-D axis of the adult appendage. A rudimentary P-D axis is established in the embryo and the larval leg imaginal disc by the expression of the transcription factors Distalless, Dachshund and Homothorax in distal, medial and proximal domains, respectively. The P-D axis is further refined by activation of EGFR signalling in the presumptive tarsus, the distal-most portion of the fly leg, during the early third larval instar. As well as slightly later, in medial and proximal rings. EGFR signalling is a ubiquitous pathway with numerous roles throughout fly development as well as across metazoan taxa. Its activation produces diverse cellular outcomes such as growth, differentiation, or regulation of apoptosis depending on the precise regulation of its inputs and modulation of intracellular signalling components in a tissue-specific manner. The precise mechanism by which EGFR signalling is activated during tarsal patterning is the focus of this dissertation. As a crucial first step in the detailed characterization of EGFR activation in the leg, we have identified leg-specific enhancers of the genes encoding the neuregulin-like EGF ligand Vein and the ligand-activating protease Rhomboid and performed genetic and site-specific mutagenesis experiments to characterize the factors necessary to activate expression of vein and rho in the distal leg. While the enhancers of vein and rho (vnE and rhoE, respectively) employ similar transcriptional programs to activate target gene expression, there are some key differences. Both enhancers require Dll for their expression throughout leg development, however vnE requires Wg and Dpp only early and later becomes independent from these signals while rhoE requires them until much later in development. Further, vnE requires Sp1 while rhoE does not. These differences may be important for the precise timing of expression of these genes, with vn expression coming on several hours earlier than that of rho. It has been proposed that the distal source of EGFR ligand may act as a long-range morphogen to pattern the entire tarsus in a graded manner (Campbell, 2002; Galindo et al., 2005). Our analysis indicates that vnE and rhoE represent the only sources of EGFR ligand in the distal leg. Therefore, in order to determine the importance of distal of EGFR signalling for tarsal patterning we carried out CRISPR targeting to delete vnE and rhoE. Because these deletions produce only mild distal leg truncations and cannot be worsened by removal of other candidate EGFR inputs (for example the Rho homolog, Roughoid) we conclude that the long-range distal gradient model for P-D patterning by EGFR must be revised. Instead we propose that the tarsal segments are patterned by the combined action of a local, distal gradient of EGFR supplied by vnE and rhoE combined with secondary, more medial sources of EGFR signal. Our analysis of the mechanism by which EGFR patterns the distal leg segments improves our understanding not only of leg development, but also of how the EGFR pathway is regulated in general. Our conclusions have important evolutionary implications, as receptor tyrosine kinase signalling, of which EGFR is an example, is involved in limb patterning in taxa whose limbs themselves are not thought to be structurally homologous to fly legs (Panganiban et al., 1997; Pires-daSilva and Sommer, 2003). Further, the components of the EGFR pathway assessed in this work are highly conserved signalling molecules, involved in cell proliferation and are therefore often misregulated in tumors. A nuanced understanding of the ways in which EGFR signalling is activated, particularly via regulation at non-protein-coding loci, could motivate new therapeutic approaches