Background. Ischemic cardiovascular disease is the most common cause of death in the Western world and, despite advances in medical and surgical therapy, the morbidity and mortality remain very high. Therapeutic angiogenesis aims to induce the formation of new blood vessels to improve the perfusion of ischemic tissue in patients with end-stage coronary artery or peripheral arterial disease that are not amenable to other treatment options. Vascular Endothelial Growth Factor-A (VEGF) is the most potent and specific angiogenic factor and it has been tested in several clinical trials with a variety of delivery methods. However, preclinical studies clearly demonstrated that high doses of VEGF can also induce aberrant vascular structures which resemble cavernous hemangiomas and grow progressively. The results of placebo-controlled clinical trials have been disappointing and yielded mostly negative results. The main conundrum of VEGF delivery-based strategies for therapeutic angiogenesis is its apparently limited therapeutic window in vivo, such that low doses are safe, but mostly inefficient, and higher doses become rapidly unsafe. Rationale. Our previous results show that the transition between normal and aberrant angiogenesis takes place in an all-or-none fashion across a discrete threshold level of VEGF expression. However, we found that this threshold is not an intrinsic property of VEGF dose alone, but rather depends on the balance between angiogenic stimulation by VEGF and vascular maturation by PDGF-BB-mediated pericyte recruitment. Our results in the previous funding period show that 1) VEGF overexpression in skeletal muscle at therapeutic doses induces angiogenesis by circumferential enlargement followed by intussusception and not by sprouting; 2) Notch-1 regulates the initial stages of vascular enlargement by a different pattern of activation compared to sprouting, but is not involved in intussusceptive remodeling; 3) EphrinB2/EphB4 signaling controls the switch between normal and aberrant angiogenesis by different VEGF doses, by regulating the degree of endothelial proliferation and initial vascular enlargement. Specific aims. In the current funding period we propose to extend these results to investigate: 1) the role of Notch4 signaling in the switch between normal and aberrant angiogenesis by VEGF in skeletal muscle; 2) the cross-talk between EphrinB2/EphB4 signaling and Notch4 activation; and 3) whether stimulation of EphB4 signaling can improve the safety and therapeutic efficacy of VEGF gene delivery in a murine hind-limb ischemia model. Experimental design. Monoclonal populations of retrovirally transduced myoblast, which stably secrete different amounts of VEGF, or a highly controlled fibrin-based platform for protein delivery, will be used to deliver specific VEGF doses in skeletal muscle. Notch4-/- mice will be used to study loss of Notch4 function, while systemic treatment with specific molecules will be used to stimulate the Notch pathway, or block or activate EphB4 signaling. Expected value of the proposed project. The experiments proposed are expected to: 1) provide fundamental knowledge on the basic mechanisms by which VEGF dose determines a switch between normal and aberrant angiogenesis in skeletal muscle, and in particular the regulation of VEGF-induced intussusceptive angiogenesis, which is a poorly understood, but therapeutically relevant process; and 2) test the potential of targeting the EphrinB2/EphB4 pathway to improve both the efficacy and safety of therapeutic angiogenesis by VEGF delivery in a functional ischemia model and with a clinically relevant gene therapy approach.