br HIF Signaling in AA Therapy Resistance An important

HIF-α Signaling in AA Therapy Resistance An important result emerging from Phase 3 clinical trials testing AA therapies is improvement of progression-free survival (PFS) that is not accompanied by enhanced overall survival (OS) [26]. This dichotomy has been attributed to mechanisms of acquired resistance upon AA therapies. Indeed, VEGF-dependent inhibition of tumor vascularization increases IH and HIF-α, promoting resistance through the upregulation of alternative HIF-α-dependent angiogenic factors [7]. In addition, HIF-α activates paracrine loops between hypoxic tumor and stromal cells that further contribute to resistance [7,19,27,28]. However, AA therapy resistance also involves HIF-α-independent mechanisms, including, but not limited to, alteration in apoptotic signaling, drug metabolism/transport, and DNA damage/repair, which are beyond the scope of this article (reviewed in [29]). Here, we focus on the role of HIF-α signaling in RTKI treatment and on EC metabolism, recently shown to have a significant role in IH and therapy resistance. Several RTKIs have been approved by the FDA as first- or second-line AA therapies for a vestigated array of cancer types [8]. It has been argued that an advantageous aspect of RTKIs over VEGF-targeting antibody therapies is their ability to inhibit multiple VEGFRs and related receptors (i.e., c-KIT, c-MET, CSF1R, FGFRs, FLT3, PDGFRs, RET, and Tie2), although this broad-spectrum targeting activity increases toxicity compared with direct renin inhibitors [30]. Another less recognized aspect of RTKIs is the ability to block HIF-α signaling through the AKT→mTOR pathway or RTKs (reviewed in [6]). For example, whereas sorafenib (blocking VEGFRs, PDGFRs, and B/C-Raf) decreases HIF-1α levels via mTOR signaling, impairing HCC xenograft growth and angiogenesis [31], sunitinib (a dual PDGFR/VEGFR inhibitor) decreases HIF-1α→VEGF-A signaling upon ligand-induced activation of PDGFRβ or VEGFR1 in neuroblastoma cells and xenografts, thereby suggesting the presence of autocrine–paracrine loops between CCs and ECs [32]. Schito et al. extended this principle to lymphatic endothelial cells (LECs), wherein the RTKI imatinib impaired breast cancer orthograft growth, lymphangiogenesis, and lymphatic dissemination through the HIF-1α→PDGFB→PDGFRß axis [33]. Taken together, these data suggest that RTKIs inhibit RTKs upstream of HIF-α; however, caution is required in light of contrasting data from studies utilizing cabozantinib, a recently FDA-approved RTKI for medullary thyroid and advanced RCC (blocking Axl, c-Met, RET, and VEGFR2), wherein cabozantinib enhanced HIF-α-dependent transactivation of hepatocyte growth factor (HGF), c-MET, and VEGF-A, accompanied by impaired xenograft and patient-derived xenograft (PDX) growth [34,35]; moreover, combination therapy with the HIF-1α inhibitor 2-methoxylestradiol improved the response to cabozantinib [34]. These results indicate that the effects of RTKIs upon HIF-α are cell and cancer type-specific as well as dependent upon the selectivity of the inhibitor being studied. Another salient aspect illustrating these complexities is the observation that RTKI schedules and doses can have opposing effects on tumor growth depending on which tumor model is interrogated [36]. The modulation of malignant and EC metabolism (reviewed in [37–40]) represents another promising strategy to optimize AA therapy responses in hypoxic cancers, given the breadth of metabolic responses that are dependent upon HIF-α transcriptional activity. These include glycolysis, lipid ß-oxidation, pentose phosphate pathway, glutaminolysis, mitochondrial autophagy, and reactive oxygen species (ROS) generation, among others [6,7,41]. Several recent studies suggest that the tumor microenvironment differentially reprograms EC metabolism when comparing non-malignant versus malignant endothelia [37]; intriguingly, Wong et al. demonstrated that fatty acid β-oxidation mediates p300-dependent epigenetic activation of the master lymphangiogenic transcriptional regulator PROX1 during mouse development and tissue repair [42]. Accordingly, it will be necessary to determine the requirement of ß-oxidation for hypoxia-induced tumoral lymphangiogenesis and its potential cross talk with the HIF-α system, which heavily depends upon p300 histone acetylase activity for target gene transactivation, speculatively providing a molecular strategy to inhibit both PROX1 and HIF-α in tumoral LECs. A preclinical study by Navarro et al. in a spontaneous breast cancer model showed that the RTKI nintedanib (blocking VEGFRs, PDGFRs, and FGFRs) normalized tumor vasculature and decreased IH and HIF-1α while reprogramming CC metabolism, wherein mitochondrial respiration was enhanced at the expense of glycolysis (reversing the Warburg effect) [43]. In the same study, inhibition of mitochondrial respiration by phenformin or ME344 combined with nintedanib or the dual VEGFR2/Tie2 inhibitor regorafenib, resulted in synthetic lethality in two different cancer models [43]. These early results support the notion that tumor angiogenesis and AA therapy responses might rely upon distinct HIF-1α-driven metabolic adaptations that can be exploited to improve survival. It remains to be determined whether these metabolic effects can be explained by predominant modulation of CC or EC/LEC (lymph)angiogenic signaling pathways and whether the apparent clinical benefit of RTKIs over anti-VEGF antibodies is due to inhibition of RTK-dependent HIF-α expression.