|M.Sc Student||Zidon Roni|
|Subject||The Role of Semaphorin 3C in Tumor Progression and the|
Mechanism by which Semaphorins Affect the
|Department||Department of Medicine||Supervisor||Professor Emeritus Gera Neufeld|
|Full Thesis text|
Semaphorin 3C (sema3C) is a member of the class 3 semaphorins. Other class 3 semaphorins such as semaphorin 3A (sema3A) and semaphorin 3F (sema3F) have been found to be secreted and function as potent anti-tumorigenic and anti-angiogenic agents. To study whether sema3C inhibits tumor progression, we expressed the sema3C cDNA in several different cell lines. Interestingly, despite being strongly expressed, sema3C failed to be secreted from these cells. Domain swapping experiments revealed that the signal sequence of sema3C is functional, suggesting that sema3C may contain retention signals. We have managed to force the secretion of sema3C only as a fusion protein in which alkaline-phosphatase (AP) is fused in frame to the N-terminus of sema3C (AP-sema3C). This fusion protein successfully repelled endothelial cells (ECs), similarly to other class 3 semaphorins. Finally, expression of recombinant non-secreted sema3C in HEK293 cells and HCT116 colon cancer cells significantly inhibited tumor development in nude mice compared to control cells. As both of these cell lines express neuropilins and several plexins, it is possible that the effects of sema3C on these cells are mediated via an autocrine pathway.
The cytoskeleton is a common downstream target of semaphorin signaling pathways. We examined the effect of various class 3 semaphorins on the actin cytoskeleton of ECs. Sema3A and semaphorin 3E (sema3E) showed the most potent and rapid effects on the endothelial cytoskeleton, while causing the formation of F-actin containing vacuoles. Sema3A-induced cytoskeletal collapse was dependent on the expression of neuropilin-1 and plexinA1, while sema3E-induced collapse depended only on the expression of plexinD1. These results indicate that different class 3 semaphorins may activate common signaling pathways albeit acting through different receptors. To determine the role of the Rho family GTPase Rac in class 3 semaphorin signaling, we have expressed constitutively active and dominant negative Rac mutants in ECs. Expression of constitutively active Rac in ECs completely abolished both sema3A and sema3E-induced cytoskeletal collapse. Conversely, expression of dominant negative Rac in these cells did not prevent sema3A and sema3E-induced collapse. Both of these experiments indicate that semaphorin-induced cytoskeletal collapse in ECs is not dependent on Rac activation. Since Rac was previously shown to mediate the cytoskeletal effects of semaphorins in neurons, our findings suggest that different signaling pathways mediate the effects of semaphorins in different cell types.