Tumors need the development of new vessels from the pre-existing vasculature to bring nutrients and oxygen to the whole tumor mass. The tumor vascular network is known to be poorly functional due to architectural and functional abnormalities. The end result is an inadequate and heterogeneous tumor perfusion leading to the development of tumor hypoxia. From a therapeutic perspective, hypoxia is a source of radioresistance and the dysfunctional perfusion hampers drug delivery. Historically, tumor hypoxia refers to chronic hypoxia (or diffusion-limited hypoxia) that results from the increasing distance between O2-consuming cells and blood vessels due to the high metabolic rate of tumor cells. Many studies have demonstrated the impact of chronic hypoxia on the clonal selection of tumor cells resistant to conventional anti-cancer therapies. Growing evidence for the existence of another form of hypoxia caused by heterogeneities in tumor perfusion, namely acute or perfusion-limited hypoxia, plead however for a non-genetic source of phenotype conversion reaching not only tumor cells but also the tumor vasculature and in particular endothelial cells. In the cardiovascular field, the cyclic exposure to different pO2 levels is known to precondition cardiac myocytes to resist more prolonged ischemic insults. We hypothesized that this concept of myocardium preconditionning to promote the resistance vs pro-apoptotic stresses could be translated in tumors. Indeed, intermittent hypoxia in tumors is nothing else than cyclic changes in pO2 and radio- and chemotherapy can be viewed as pro-apoptotic stresses that the tumor can face. In particular, in the case of the tumor vasculature, the resistance could be a capacity to re-initiate angiogenesis after treatment. Radioresistance would be further potentiated since low pO2 is per se associated to reversibility of the damages. Also, since intermittent hypoxia is thought to be due in part to fluctuations in tumor blood flow (TBF), access of chemotherapy to the tumor could also further participate to chemoresistance. To address the above hypotheses, we first aimed to explore the extent and the origin of TBF fluctuations in tumor mouse models and to determine whether therapeutic modulation of such potential TBF heterogeneities could improve the efficacy of chemotherapy. We then more directly examined whether and how intermittent hypoxia could influence endothelial cell survival and modulate resistance to radiotherapy. We also took advantage of this study to dissect the molecular mechanisms driving the phenotype conversion of endothelial cells exposed to intermittent hypoxia. Finally, because VEGF plays a major role in hypoxia-mediated angiogenesis but also regulates major pro-survival pathways in endothelial cells, we evaluated the potential role of caveolin as a new therapeutic target to tackle EC resistance. Caveolin is, indeed, a key structural protein recently documented to interact with many downstream targets of VEGF. 1. To explore the extent and the origin of TBF fluctuations in tumor mouse models and to determine whether therapeutic modulation of such potential TBF heterogeneities could improve chemotherapy. We focused this part of the work on the vascular tone modulator endothelin-1. Indeed, this peptide is over-expressed in many mouse and human tumors where it is documented to act as a mitogenic factor in both para- and autocrine manners. Endothelin-1 is also a potent vasoconstrictor acting through the ETA receptors located on VSMCs. In our lab, we previously showed that over-expression of endothelin-1 in tumors accounted for the development of a myogenic tone within the tumor vasculature. We have now documented that an ETA receptor antagonist induces the relaxation of microdissected tumor arterioles and selectively and quantitatively increases tumor blood flow in experimental tumor models. We also combined dye staining of functional vessels, fluorescent microsphere-based mapping, and magnetic resonance imaging to identify heterogeneities in tumor blood flow and to examine the reversibility of such phenomena. We showed that administration of an ETA receptor antagonist reduces the extent of underperfused tumor areas, proving the key role of vessel tone variations in tumor blood flow heterogeneity. We also provided evidence that ETA antagonist could improve the access of cyclophosphamide to the tumor compartment and thereby induces a significant tumor growth delay. 2. To examine whether and how intermittent hypoxia could influence endothelial cell survival and modulate resistance to radiotherapy. To dissect the mechanisms driving the phenotype conversion of endothelial cells exposed to intermittent hypoxia. This second part of our work, is a comprehensive investigation of the consequences of intermittent hypoxia, as caused by TBF heterogeneities, on the endothelial cell phenotype. First, we postulated that intermittent hypoxia (IH) favors endothelial cell (EC) survival, thereby extending the concept of hypoxia-driven resistance to the tumor vasculature. We showed that exposing EC to cycles of hypoxia/re-oxygenation reduces radiation-induced cell death and promotes angiogenesis. In contrast, prolonged hypoxia failed to achieve such protection and even appeared deleterious. We also observed that although HIF-1£ is completely degraded during each re-oxygenation, its abundance is paradoxically found higher at each new hypoxic challenge. Moreover, the use of siRNA targeting HIF-1£ pointed out that HIF-1ƓÑ accumulation account for the increased resistance of EC to radiotherapy. Finally, we extended this concept in vivo by forcing IH in tumor-bearing mice and found that it is associated with less radiation-induced apoptosis within both the vascular and the tumor cell compartments (vs normoxia or prolonged hypoxia). Next, we focused our work on the underlying mechanisms of EC phenotype conversion exposed to IH and particularly on potential actors that may favor HIF-1£ accumulation during IH. Prolylhydroxylases (PHD), MAPK and PI3K/Akt pathways as well as eNOS are known to regulate HIF-1£ abundance and transcriptional activity. We documented that PHD2 and PHD3 abundance are slightly decreased during IH, whereas prolonged hypoxia increases PHD3 expression in EC. We then showed that, ERK, Akt as well as eNOS were phosphorylated during reoxygenation periods of the IH protocol. We also used specific inhibitors of these cascades (i.e. PD98059, LY294002 and L-NAME, respectively), to evaluate their specific impact on HIF-1£ abundance and performed clonogenic assays to evaluate their consequences on EC survival. We showed that although, PD98059 and LY294002 sensitizes EC to pro-apoptotic stresses, only the PI3K/Akt inhibitor abrogates the HIF-1£ signal during IH. Conversely, L-NAME, a non-specific NOS-inhibitor, appears to potentiate the expression of HIF-1£ and to favor the EC survival. 3. To identify new therapeutic targets to prevent endothelial cell resistance by studying VEGF signaling, the major pro-survival and pro-angiogenic growth factor in endothelial cells. Because VEGF plays a central role in hypoxia-mediated angiogenesis and cell survival, the VEGF signaling cascade is a an obvious therapeutic target. To more specifically identify the pathways leading to cell survival and the resistance phenomena that we observed in response to intermittent hypoxia, a careful dissection of the downstream VEGF signaling cascades was performed. In this part of the work, we focused our attention on caveolin since it modulates the activity of eNOS, ERK and Akt that are major effectors acting downstream VEGF stimulation. We demonstrated the paradoxical role of caveolin-1 preventing signaling in basal conditions and ensuring the coupling between VEGFR2 and the downstream cascades upon VEGF stimulation. We used mice deficient for the caveolin-1 gene (Cav-/-) to examine the impact of caveolae suppression in a model of adaptive angiogenesis obtained after femoral artery resection. Evaluation of the ischemic tissue perfusion and histochemical analyses revealed that contrary to Cav+/+ mice, Cav-/- mice fails to recover a functional vasculature and actually loose part of the ligated limbs. We also isolated endothelial cells (ECs) from Cav-/- aorta and showed that on VEGF stimulation, endothelial tube formation is dramatically abrogated when compared with Cav +/+ ECs. The Ser1177 eNOS phosphorylation and Thr495 dephosphorylation but also the ERK phosphorylation were similarly altered in VEGF-treated Cav-/- ECs. Interestingly, caveolin transfection in Cav-/- ECs redirected the VEGFR-2 in caveolar membranes and restored the VEGF-induced ERK and eNOS activation. However, when high levels of recombinant caveolin are reached, VEGF exposure fails to activate ERK and eNOS. Altogether, these data identify caveolin as a new therapeutic target to alter VEGF signaling, in particular the cascades leading to angiogenesis and resistance to stresses.