The function of vascular endothelial growth factor

Abstract

Vascular endothelial growth factor (VEGF) is considered the master regulator of angiogenesis during growth and development, as well as in disease states such as cancer, diabetes, and macular degeneration. This review details our current understanding of VEGF signaling and discusses the benefits and unexpected side effects of promising anti-angiogenic therapeutics that are currently being used to inhibit neovacularization in tumors.

1. Introduction

All tissues require a continual supply of oxygen and nutrients coupled with the removal of carbon dioxide and toxic cellular byproducts. These essential needs are met by two processes, the de novo formation of blood vessels, termed vasculogenesis and by angiogenesis, the process by which new blood vessels sprout and grow from pre-existing vessels. While vasculogenesis is restricted to embryonic development, angiogenesis occurs during embryonic development and in select instances in the adult including female reproduction, wound healing, and following bouts of exercise. Unfortunately, dysregulation of angiogenesis also contributes to progression of diseases including cancer, macular degeneration, and diabetic retinopathy.

While angiogenesis is a normal physiological process, our current understanding of angiogenesis originated with the observation that newly formed blood vessels infiltrated tumors and wounds [1–5]. Soon after, the term “angiogenic factor” was coined to describe some activity derived from tumors that could promote blood vessel formation [6], and by the late 1960’s and early 1970’s it was understood that a diffusible factor was released from tumors that attracted new vessels to the tumor site. This observation began to redefine our understanding of tumor progression in terms of angiogenesis [7–9]. Indeed, early studies demonstrated that living tumors do not increase in size when they are suspended in the anterior chamber of the eye, where they cannot become vascularized [10], revealing that angiogenesis is essential for tumor growth beyond a defined volume. In 1983, a secretory protein termed vascular permeability factor, later to be renamed vascular endothelial growth factor (VEGF), was discovered on the basis of its ability to increase vascular permeability [11], and subsequently determined to be the master regulator of angiogenesis [12,13]. Specific receptors for this protein were soon discovered [14–16] and numerous VEGF-mediated signaling pathways were identified that regulate angiogenesis by influencing endothelial proliferation, migration, survival, extracellular matrix degradation, and cell permeability [17].

In 1993, the first example of anti-VEGF therapy was reported whereby a monoclonal antibody against VEGF was shown to inhibit the growth of xenograft tumors in mice [18]. Following this report, exciting new second generation anti-angiogenic therapies targeting VEGF signaling pathways have been developed and are being used against diseases such as cancer and macular degeneration [19]. While these new therapeutics have demonstrated remarkable efficacy, especially in macular degeneration, a number of unexpected side effects have occurred in a subset of patients taking these drugs. Consistent with these observations, several recent studies suggest novel non-angiogenic roles for VEGF in both the embryo and adult, including neuronal survival, muscle development, wound healing, and bone and cartilage formation.

This review will examine the major components, signaling pathways, and aberrant regulation of the VEGF pathway. Moreover we will discuss novel non-angiogenic roles for VEGF signaling that have been recently reported. Finally, we hypothesize how these novel roles for VEGF could be responsible for the unexpected side effects in patients who undergo anti-VEGF therapeutic treatment.

2. VEGF-A

The VEGF family is composed of six secreted cysteine knot glycoproteins: VEGF-A, B, C, D, E and placental growth factor (PlGF), each associated with numerous developmental as well as pathological processes [20]. The scope of this article, however, remains limited to VEGF-A (referred to in this article as VEGF) and its specific receptors, and the research concerning its fundamental role in angiogenesis. The human VEGF gene produces eight isoforms generated through alternative splicing, whose most studied members include VEGF₁₂₁, VEGF₁₆₅, and VEGF₁₈₉ [21], each designated according to the number of amino acids following the signal sequence cleavage site (mouse and rat proteins are one amino acid shorter and are numbered accordingly). Exons 6 and 7 encode heparin sulfate proteoglycan binding domains, thus isoforms containing these exons are sequestered on the cell surface and/or in the extracellular matrix and exhibit decreased diffusibility [22]. In contrast, the absence of exons 6 and 7 in VEGF₁₂₁ result in its being freely diffusible. Interestingly, the abundance of VEGF isoforms varies among tissues according to the function and specific needs of the tissue, suggesting that different VEGF isoforms may play distinct roles in vascular development [23], yet how each isoform specifically contributes to individual tissue function is largely unknown.

Experimental deletion of even a single VEGF allele results in abnormal blood vessel development and embryonic lethality by E9.5 in murine models, indicating the critical role of this growth factor in embryonic development [24,25]. Moreover, modest overexpression of VEGF leads to embryonic lethality by E12.5–14, resulting from disorganization of blood vessels and malformation of the heart and coronary vasculature [26]. These studies indicate strict dosage dependence between VEGF expression and normal cardiovascular development.

3. VEGF receptors

VEGF ligands bind specifically to two receptor tyrosine kinase membrane-bound proteins—VEGFR1 (also called Flt-1) and VEGFR2 (also called Flk-1 or KDR), which are expressed in most endothelial cells. VEGFR1 and 2 contain three functional regions including seven Ig-like extracellular domains, of which domains 2 and 3 mediate ligand binding, a single membrane spanning region, and a cytoplasmic tyrosine kinase domain [27–29]. Upon VEGF binding, receptors homo-or hetero-dimerize and subsequent autophosphorylation of several tyrosine residues in the cytoplasmic region of the VEGF receptor occurs [30,31]. Additionally, VEGF receptors reportedly heterodimerize with the semaphorin receptors neuropilin (NRP) 1 and 2; NRP receptors are believed to enhance, but not directly participate in, VEGFR2-mediated VEGF signaling [16].

VEGFR2 is considered the primary signaling receptor for VEGF, integrating VEGF ligand-mediated extracellular signals to activation of numerous transduction cascades, including mitogen activated protein kinase (MAPK), phosphatidyl inositol 3-kinase (PI3-K), protein kinase C (PKC), AKT, and PLC pathways, leading to increases in endothelial proliferation, migration, MMP expression, survival, and cell permeability [32] (Fig. 1). The critical role of this receptor is apparent given that a homozygous knockout of VEGFR2 in mice leads to embryonic lethality at E8.5–9.5 due to defects in the development of hematopoietic and endothelial cells and the absence of yolk sac vasculature [26,33].

Fig. 1.

VEGF signaling. VEGF stimulates the VEGF receptor in endothelial cells to activate numerous signal transduction pathways essential for cell proliferation, survival, cytoskeletal rearrangements, migration, gene transcription, and endothelial permeability.

Several lines of evidence suggest that VEGFR1 does not perform the principal role in mediating VEGF-induced signal transduction cascades in endothelial cells, but instead acts as a decoy to modulate VEGF levels [34]. For instance, VEGFR1 demonstrates strong binding affinity for VEGF; however the kinase activity of this receptor is one order of magnitude lower than that of VEGFR2 [35]. Antisense knockdown of VEGFR1 does not affect endothelial cell proliferation, migration, and platelet activating factor expression, while knockdown of VEGFR2 severely impairs these processes [36]. Moreover, null mutations in the murine VEGFR1 gene result in embryonic lethality by E8.5 due to overgrowth of endothelial cells and disorganized blood vessel formation [14]. The hypothesis that VEGFR1 serves as a VEGF trap is complicated by a handful of studies suggesting that R1 is essential for some cellular processes. Although 10-fold weaker activity than VEGFR2, VEGFR1 does possess tyrosine kinase activity in its cytoplasmic domain. There is evidence to indicate that VEGFR1 stimulates MAPK signaling in breast cancer [37] and modulates macrophage and monocyte migration [38], suggesting that VEGF-mediated VEGFR1 signaling may play an important signaling function in non-endothelial cells.

4. VEGF expression

Secretion of VEGF by epithelial, mesenchymal, and tumor cells activates VEGF receptors on the nearby endothelium, thus initiating angiogenesis and stimulating recruitment of newly formed blood vessels to the source tissue (Fig. 2). This process occurs extensively during embryonic development and in the adult during wound healing responses, the female reproductive cycle, and following bouts of exercise [39]. Surprisingly, it has been convincingly demonstrated that VEGF is expressed in a cell-specific manner in nearly all vascularized adult tissues including select endothelium, and VEGFR2 is constitutively phosphorylated across several tissues in vivo [40]. These observations suggest that VEGF performs a physiological function in the quiescent vasculature in addition to simply promoting angiogenesis. Indeed, several studies have indicated that VEGF may serve a maintenance role in vascularized tissues by inducing capillary fenestrations, as loss of VEGF has been shown to modulate fenestration stability in liver endothelial cells, pancreatic islet endothelial cells, and choroid plexus [41–43] and in promoting endothelial cell survival, as inhibition of VEGF leads to loss of alveolar, tracheal, and peritubular capillary endothelium [44–46].

Fig. 2.

VEGF recruits endothelial cells to its source. (A) In normoxic (normal oxygen levels) conditions, HIF1α associates with von Hippel-Lindau tumor suppressor protein (VHL), an E3 ubiquitin ligase that targets HIF1α for ubiquitin (UB)-mediated proteosomal degradation. Hypoxia (low oxygen levels) leads to the activation of HIF1α prolyl hydroxylases (HPH) and factor inhibiting HIF1α hydroxylase (FIH), which hydroxylate HIF1α protein. Hydroxylation of HIF1α promotes the association of the transcriptional coactivators CREB binding protein (CBP) and p300 leading to binding of HIF1β to HIF1α. This transcriptional complex then binds to hypoxia response elements in the promoter region of the VEGF gene to enhance VEGF transcriptional expression. (B) Epithelial, mesenchymal, and tumor cells secrete VEGF ligand in response to hypoxia and other stimuli. Upon binding to VEGF receptors on the nearby endothelium, VEGF initiates angiogenesis and stimulates recruitment of newly formed blood vessels to the source tissue. (triangle represents VEGF ligand).

The molecular control of VEGF expression is best understood in terms of hypoxic regulation of angiogenesis. In this textbook example of VEGF regulation, hypoxia stimulates the activation of the helix-loop-helix transcriptional regulator hypoxia inducible factor-1 alpha (HIF1a), which then directly promotes the transcription of the VEGF gene at consensus hypoxia response elements located in the VEGF promoter [47]. Interestingly, while mice lacking the HIF1a gene show severe cardiac and vascular malformations leading to embryonic lethality [48], deletion of the hypoxia response elements results in viable offspring with only minor neurological defects in motor neurons [49,50], suggesting that the classical understanding of VEGF transcription may not be the primary means by which its expression is controlled. In fact, it has been demonstrated that, at least in some cell types, hypoxic upregulation of VEGF mRNA occurs via stabilization of the otherwise very short-lived VEGF transcripts [51].

In addition, recent reports have suggested that a number of transcriptional regulators may control VEGF expression in both hypoxia-dependent and -independent pathways. For instance, VEGF expression is regulated in skeletal muscle cells through hypoxia-dependent binding of peroxisome proliferator activated receptor-gamma coactivator-1alpha (PGC1a) and its orphan nuclear receptor oestrogen-related receptor-alpha (ERRa) directly to the VEGF promoter and first intron [52]. Moreover, the metabolic fuel sensor AMP-activated protein kinase (AMPK) reportedly promotes VEGF transcriptional expression and enhances stabilization of VEGF mRNA in response to exercise [53]. The transcriptional regulator MyoD, which serves as the master regulator of myogenic specification and differentiation, directly stimulates VEGF expression during the differentiation of myogenic progenitors into skeletal muscle cells [54]. Insulin-like growth factor-1 (IGF1) has been shown to upregulate VEGF expression via hypoxia-dependent signaling in retinal pigment epithelial cells [55] and osteoblasts [56]. Both hypoxia-dependent and -independent regulation of VEGF by SP1 has been reported. In these cases, PI3K/Akt pathway-dependent upregulation of VEGF is mediated in a hypoxia-independent manner through Sp1 binding sites located in the proximal promoter [57] and, in contrast, through a combinatorial activation of Sp1, HIF1a, and signal transducer and activator of transcription 3 (Stat3) [58]. Though numerous positive regulators of VEGF expression have been identified, relatively little is known regarding the negative regulation of VEGF expression. One recent study identified a novel repressor of VEGF expression called ZNF24, whereby, in breast cancer cells, the steady state expression of ZNF24 is inversely correlated with oxygen availability, suggesting that ZNF24 inhibits VEGF expression during normoxic conditions [59].

5. Anti-VEGF cancer therapeutics

While the development and efficacy of new cytotoxic agents for cancer treatment has been disappointing, our increased understanding of the mechanism of VEGF signaling has equipped us with a novel class of anti-angiogenic therapies which have proven efficacy in a number of diseases including cancer and wet forms of macular degeneration. The first such therapy to be approved by the FDA was bevacizumab (Avastin), which is a humanized VEGF monoclonal antibody that binds with high affinity and sequesters the VEGF ligand, thus preventing its activation of VEGF receptors (Fig. 2). Bevacizumab, in conjunction with conventional chemotherapy, has been demonstrated to significantly decrease angiogenesis and primary and metastatic growth in colorectal cancer, breast cancer, and non-small cell lung carcinoma [60–63]. Moreover, the FDA approval of the intravitreal drug ranibizumab (Lucentis), and the off label usage of bevacizumab, significantly reduces disease progression in wet macular degeneration and has been shown to improve overall vision [64]. The strategy of denying tumors and other pathologies with an angiogenic component a blood supply shows exciting clinical promise, and several second generation anti-angiogenic drugs such as Sorafenib, Sunitinib, and Vandetanib, which inhibit VEGF receptors as well as other tyrosine kinase receptors that contribute to disease phenotypes, are currently being developed.

While the therapeutic efficacy of these agents is very promising, anti-angiogenic treatments are unfortunately not devoid of side effects. An array of expected, as well as unexpected, side effects have been observed including hypertension, gastrointestinal toxicity, hypothyroidism, proteinuria, coagulation disorders, general fatigue, and neurotoxicity [65]. Indeed, recent findings have shed new light on novel roles for VEGF that may explain some of these effects and assist clinicians in effectively managing these side effects. For instance, the neurological manifestations observed in some patients undergoing anti-angiogenic therapy [66] may be explained in part by a recent study which demonstrated that systemic blockage of VEGF in mice results in a disruption of the choroid plexus, leading to altered choroid plexus fenestrations and perfusion, as well as ependymal cell dysfunction, disruption of ventricular barrier function and altered periventricular homeostasis [67]. Moreover, bevacizumab and ranibizumab are both used clinically for the treatment of wet macular degeneration; however systemic VEGF neutralization in mice results in a significant increase in retinal apoptosis associated with reduced thickness of the inner and outer nuclear layers of the retina, and a decline in retinal function [68]. Retinal degeneration observed during this treatment was attributed to an autocrine role for VEGF in Müller cell survival and a paracrine neuroprotective effect on the photoreceptors. Indeed, VEGF autocrine signaling is reportedly essential for lens, skeletal muscle, and bone differentiation, as well as kidney podocyte and neuronal survival—all independent of effects on the vasculature [54,68–72]. Consistent with these observations, VEGF receptors are expressed at detectable levels in a large number of non-endothelial cells types including skeletal myocytes, dental odontoblasts, retinal neurons, keratinocytes, chondrocytes, and neurons [54,73–77], supporting the concept that VEGF signaling is important for more than endothelial cell function and suggesting that anti-VEGF therapies should be administered with caution.

Despite the occurrence of some side effects, newly developed anti-angiogenic therapeutics and combinations of existing therapeutics will likely lead to promising treatments for cancer and other vascular pathologies such as macular degeneration, psoriasis, and rheumatoid arthritis. As our ability to manipulate angiogenesis and its regulation expands, the identification of exciting new applications for VEGF therapy will undoubtedly increase.