Vascular endothelial growth factor A (VEGF-A) is one of the most potent and integral regulators of angiogenesis, which contributes to pathogenesis of many diseases, including cancer, rheumatoid arthritis and blinding eye diseases, such as age-related macular degeneration and proliferative diabetic retinopathy.1 In addition to driving angiogenesis, VEGF-A increases permeability of existing blood vessels, which also can lead to pathology. This is particularly relevant in eye diseases such as diabetic macular edema, in which VEGF-A-driven vascular leakage results in retinal edema, which compromises vision. Due to the essential role of VEGF-A in angiogenesis and permeability, anti-VEGF-A antibodies (bevacizumab), Fab fragments (ranibizumab) and several VEGF receptor tyrosine kinase inhibitors have been approved for clinical use in settings where angiogenesis or vascular leakage contributes to pathology.1,2
VEGF-A activates receptor tyrosine kinases (RTKs) VEGFR-1 and VEGFR-2, and the latter appears to be the main receptor responsible for the pro-angiogenic effects of VEGF-A. Activated VEGFR-2 engages various signaling enzymes, including PLCγ, Src, Erk, class Ia phosphatidylinositol 3-kinase (PI3K) and downstream effectors such as Akt.3 While much progress has been made cataloging the signaling events induced by VEGF-A, a major unanswered question is how signaling specificity is achieved. For example, since activation of the same signaling enzymes (e.g., PI3K/Akt) is associated with distinct VEGF-A-dependent cellular responses (migration and survival), what determines which cellular response prevails? Furthermore, since angiogenesis is a precise sequence of cellular events (destabilization of existing vessels, migration and proliferation of endothelial cells, maturation/stabilization of nascent vessels), it seems likely that VEGF-A-induced signaling events are coordinated with signaling events triggered by additional agents (integrins/matrix, soluble factors and cues arising from intercellular contacts). Thus, an open challenge is to determine how signaling events that are associated with the various components of an angiogenic response are integrated to achieve the appropriate outcome.
The PI3K/Akt pathway is essential for many VEGF-A-dependent effects, such as migration, survival and vascular permeability.4 Despite the importance of PI3K/Akt, the mechanism by which VEGF-A/VEGFR-2 activates PI3K/Akt is not well understood. Activation of class Ia PI3K by RTKs requires association between a protein phosphorylated within a Tyr-Xaa-Xaa-Met (YXXM) motif and the Src homology 2 (SH2) domains of the p85 regulatory subunit of PI3K.5 This interaction is triggered by either autophosphorylation (e.g., PDGFRs; Fig. 1A), or by receptor-mediated phosphorylation of a YXXM motif-containing adaptor protein [e.g., insulin receptor substrates (IRSs); Fig. 1B]. Since VEGFR-2 does not autophosphorylate within a YXXM motif, it is likely that a YXXM motif-containing adaptor protein serves as the liaison between PI3K and the activated VEGFR-2. While some groups reported that Gab1 serves this function in cells acutely stimulated with VEGF-A,6,7 other investigators failed to confirm this finding.8,9
Figure 1. Three mechanisms by which RTKs activate PI3K. (A) PDGF-dependent activation of PI3K. PDGF induces dimerization of the PDGFR and subsequent phosphorylation of many tyrosines including those that reside within its YXXM motifs. The SH2 domains of the p85 regulatory subunit of PI3K stably associate with the tyrosine phosphorylated YXXM motifs of PDGFR. This interaction relieves inhibition of the regulatory subunit on the p110 catalytic subunit, and localizes PI3K close to its substrates in the plasma membrane. Binding activated Ras (shown as a pentagon) to the p110 catalytic subunit fully activates PI3K. (B) Insulin-dependent activation of PI3K. Insulin promotes tyrosine phosphorylation of the YXXM motifs in IRS-1 and -2, which stably interact with the SH2 domains of the p85 regulatory subunit of PI3K. The IRS/PI3K complex localizes to the plasma membrane via an interaction of IRS with the tyrosine phosphorylated insulin receptor. (C) VEGF-A-dependent activation of PI3K. VEGF-A dimerizes and activates VEGFR-2, which results in TSAd-dependent activation of SFKs. SFKs promote phosphorylation of Axl and subsequently interact with Axl. This event is associated with autophosphorylation of Axl at the two YXXM motif tyrosines, and subsequent recruitment and activation of PI3K.
Our efforts to understand how VEGF-A/VEGFR-2 activates PI3K/Akt led to the discovery that the RTK Axl is an essential mediator of this process.8 Axl was found to be required for VEGF-A-dependent activation of PI3K/Akt as well as migration, in vitro tube formation, vascular permeability and corneal neovascularization. In addition, we found that VEGFR-2 acted through TSAd to activate Src family kinases (SFKs), which communicated with Axl via its juxtamembrane domain to promote autophosphorylation of the two YXXM motif tyrosines and thereby engaged the PI3K/Akt pathway (Fig. 1C). Other VEGF-A-activated signaling enzymes (e.g., Erk and Src) were independent of Axl.
Our findings may guide development of complementary anti-VEGF-A approaches to manage both VEGF-A-mediated angiogenesis and permeability. For instance, blocking Axl will interfere with only a subset of VEGF-A-dependent signaling events and may, therefore, be a safer and more selective strategy than anti-VEGF-A. It was recently reported that the Axl inhibitor R428 substantially reduces experimental breast tumor angiogenesis and metastasis.10 In addition to endothelial cells, Axl is highly expressed in tumor cells and vascular smooth muscle cells. Thus, disruption of Axl signaling may represent a promising therapy to treat tumors by simultaneously targeting angiogenesis and tumor growth.
Footnotes
Previously published online: www.landesbioscience.com/journals/cc/article/20535
References
- 1.Carmeliet P, et al. Nature. 2011;473:298–307. doi: 10.1038/nature10144. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Carmeliet P, et al. Nat Rev Drug Discov. 2011;10:417–27. doi: 10.1038/nrd3455. [DOI] [PubMed] [Google Scholar]
- 3.Koch S, et al. Biochem J. 2011;437:169–83. doi: 10.1042/BJ20110301. [DOI] [PubMed] [Google Scholar]
- 4.Olsson AK, et al. Nat Rev Mol Cell Biol. 2006;7:359–71. doi: 10.1038/nrm1911. [DOI] [PubMed] [Google Scholar]
- 5.Songyang Z, et al. Cell. 1993;72:767–78. doi: 10.1016/0092-8674(93)90404-E. [DOI] [PubMed] [Google Scholar]
- 6.Dance M, et al. J Biol Chem. 2006;281:23285–95. doi: 10.1074/jbc.M600987200. [DOI] [PubMed] [Google Scholar]
- 7.Laramée M, et al. J Biol Chem. 2007;282:7758–69. doi: 10.1074/jbc.M611327200. [DOI] [PubMed] [Google Scholar]
- 8.Ruan GX, et al. EMBO J. 2012;31:1692–703. doi: 10.1038/emboj.2012.21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Lu Y, et al. Proc Natl Acad Sci USA. 2011;108:2957–62. doi: 10.1073/pnas.1009395108. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Holland SJ, et al. Cancer Res. 2010;70:1544–54. doi: 10.1158/0008-5472.CAN-09-2997. [DOI] [PubMed] [Google Scholar]

