help button home button Endocrine Society Endocrine Reviews
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Ferrara, N.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Ferrara, N.
Endocrine Reviews 25 (4): 581-611
Copyright © 2004 by The Endocrine Society

Vascular Endothelial Growth Factor: Basic Science and Clinical Progress

Napoleone Ferrara

Department of Molecular Oncology, Genentech, Inc., South San Francisco, California 94080

Correspondence: Address all correspondence and requests for reprints to: Napoleone Ferrara, M.D., Department of Molecular Oncology, Genentech, Inc., 1 DNA Way, South San Francisco, California 94080. E-mail: nf{at}gene.com


    Abstract
 Top
 Abstract
 I. Introduction
 II. Historical Note on...
 III. Identification of VEGF
 IV. Activities of VEGF
 V. VEGF Isoforms
 VI. Regulation of VEGF...
 VII. VEGFRs
 VIII. Role of VEGF...
 IX. Role of VEGF...
 X. VEGF and Therapeutic...
 XI. Perspectives
 References
 
Vascular endothelial growth factor (VEGF) is an endothelial cell-specific mitogen in vitro and an angiogenic inducer in a variety of in vivo models. Hypoxia has been shown to be a major inducer of VEGF gene transcription. The tyrosine kinases Flt-1 (VEGFR-1) and Flk-1/KDR (VEGFR-2) are high-affinity VEGF receptors. The role of VEGF in developmental angiogenesis is emphasized by the finding that loss of a single VEGF allele results in defective vascularization and early embryonic lethality. VEGF is critical also for reproductive and bone angiogenesis. Substantial evidence also implicates VEGF as a mediator of pathological angiogenesis. In situ hybridization studies demonstrate expression of VEGF mRNA in the majority of human tumors. Anti-VEGF monoclonal antibodies and other VEGF inhibitors block the growth of several tumor cell lines in nude mice. Clinical trials with various VEGF inhibitors in a variety of malignancies are ongoing. Very recently, an anti-VEGF monoclonal antibody (bevacizumab; Avastin) has been approved by the Food and Drug Administration as a first-line treatment for metastatic colorectal cancer in combination with chemotherapy. Furthermore, VEGF is implicated in intraocular neovascularization associated with diabetic retinopathy and age-related macular degeneration.

I. Introduction
II. Historical Note on Angiogenic Factors
III. Identification of VEGF
IV. Activities of VEGF
A. Mitogenesis, angiogenesis, and endothelial survival
B. Effects of VEGF on bone marrow cells and hematopoiesis
C. Enhancement of vascular permeability and hemodynamic effects

V. VEGF Isoforms
VI. Regulation of VEGF Gene Expression
A. Oxygen tension
B. Growth factors, hormones, and oncogenes

VII. VEGFRs
A. VEGFR-1 (Flt-1)
B. VEGFR-2 (KDR, human; Flk-1, mouse)
C. Neuropilin (NP)1 and NP2

VIII. Role of VEGF in Physiological Angiogenesis
A. Embryonic and postnatal development
B. Skeletal growth and endochondral bone formation
C. Angiogenesis in endocrine glands

IX. Role of VEGF in Pathological Conditions
A. Solid tumors
B. Hematological malignancies
C. Intraocular neovascular syndromes
D. Inflammatory disorders and brain edema
E. Pathology of the female reproductive tract

X. VEGF and Therapeutic Angiogenesis
XI. Perspectives


    I. Introduction
 Top
 Abstract
 I. Introduction
 II. Historical Note on...
 III. Identification of VEGF
 IV. Activities of VEGF
 V. VEGF Isoforms
 VI. Regulation of VEGF...
 VII. VEGFRs
 VIII. Role of VEGF...
 IX. Role of VEGF...
 X. VEGF and Therapeutic...
 XI. Perspectives
 References
 
THE CARDIOVASCULAR SYSTEM is the first organ system to develop and reach a functional state in an embryo (1). The initial steps consist of "vasculogenesis," the in situ differentiation of endothelial cell precursors, the angioblasts, from the hemangioblasts (2). The juvenile vascular system evolves from the primary capillary plexus by subsequent pruning and reorganization of endothelial cells in a process called "angiogenesis" (3). More recent evidence suggests that incorporation of bone marrow-derived endothelial precursor cells contributes to the growing vessels, complementing the sprouting of resident endothelial cells (4), although the precise contribution of these elements in various pathophysiological circumstances has been a matter of debate (5, 6, 7, 8, 9).

The development of a vascular supply is essential also for tissue repair and reproductive functions in the adult (10). Angiogenesis is also implicated in the pathogenesis of a variety of disorders: proliferative retinopathies, age-related macular degeneration (AMD), tumors, rheumatoid arthritis (RA), and psoriasis (10, 11).

In endocrine glands, vascularization serves a unique exchange role for secretory products between interstitial fluid surrounding the parenchymal cells and plasma. Endothelial cells of endocrine glands frequently display fenestrae, which are highly permeable to fluid and small solutes, thus facilitating bidirectional transport (12).

For more than a decade, the role of vascular endothelial growth factor (VEGF) in the regulation of angiogenesis has been the object of intense investigation (13). Recent evidence indicates that new vessel growth and maturation are highly complex and coordinated processes, requiring the sequential activation of a series of receptors [e.g., Tie1, Tie2, and platelet-derived growth factor (PDGF) receptor-ß (PDGFR-ß)] by numerous ligands in endothelial and mural cells (for recent reviews see Refs.14, 15, 16). However, VEGF signaling often represents a critical rate-limiting step in physiological angiogenesis. VEGF (referred to also as VEGF-A) belongs to a gene family that includes placenta growth factor (PlGF) (17, 18), VEGF-B (19), VEGF-C (20, 21), and VEGF-D (22, 23). Additionally, homologs of VEGF have been identified in the genome of the parapoxvirus, Orf virus (24), and shown to have VEGF-like activities (25, 26). Importantly, VEGF-C and VEGF-D regulate lymphatic angiogenesis (27, 28), emphasizing the unique role of this gene family in controlling growth and differentiation of multiple anatomic components of the vascular system.

The main focus of this review is the progress in the biology and clinical applications of the prototype member, VEGF-A. For additional reviews on this topic, see Refs.29, 30, 31, 32, 33, 34, 35 .

Importantly, very recent data have shown that inhibiting VEGF results in a clinical benefit, including increased survival, in patients with advanced malignancies, providing the first clinical validation of the hypothesis that blocking angiogenesis is a strategy to treat cancer (36, 37).


    II. Historical Note on Angiogenic Factors
 Top
 Abstract
 I. Introduction
 II. Historical Note on...
 III. Identification of VEGF
 IV. Activities of VEGF
 V. VEGF Isoforms
 VI. Regulation of VEGF...
 VII. VEGFRs
 VIII. Role of VEGF...
 IX. Role of VEGF...
 X. VEGF and Therapeutic...
 XI. Perspectives
 References
 
The observation that tumor growth can be accompanied by increased vascularity was reported more than one century ago (for review, see Ref.13). In 1939, Ide et al. (38) postulated the existence of a tumor-derived blood vessel growth-stimulating factor on the basis of the strong neovascular response induced by tumors transplanted in transparent chambers. These authors proposed that such a factor may be responsible for inducing a neovascularization and thus for delivery of nutrients to the growing tumor (38). In 1945, Algire et al. (39) advanced this concept, proposing that "the rapid growth of tumor transplants is dependent upon the development of a rich vascular supply," and speculated that capillary proliferation elicited by tumor cells is mediated by chromatin breakdown products, in agreement with a view prevalent at that time that such products have growth-promoting activity. These investigators also suggested that the acquisition by tumor cells of the ability to promote vascular proliferation is a critical step in tumorigenesis, because it is expected to confer on the tumor cells a growth advantage relative to normal cells (39).

In 1948, Michaelson (40) proposed that a diffusible angiogenic "factor X" produced by the retina is responsible for retinal and iris neovascularization that occurs in proliferative diabetic retinopathy and other retinal disorders, such as central retinal vein occlusion.

In 1968, the first experiments to directly test the hypothesis that tumors produce angiogenic factors were performed. Greenblatt and Shubik (41) and Ehrmann and Knoth (42) demonstrated that transplantation of melanoma or choriocarcinoma cells promoted blood vessel proliferation even when a Millipore filter is interposed between the tumor and the host, thus providing evidence that tumor angiogenesis was mediated by diffusible factor(s) produced by the tumor cells.

In 1971, Folkman (43) proposed that antiangiogenesis might be an effective approach to treat human cancer. Folkman et al. (44) initiated initial efforts aimed to isolate a "tumor angiogenesis factor" from human and animal tumors. Subsequently, the angiogenic effects of various factors, including epidermal growth factor, TGF-{alpha}, TGF-ß, TNF-{alpha}, angiogenin, etc., were reported. These molecules all were shown to have activity in angiogenesis bioassays—either directly, by promoting endothelial cell proliferation or indirectly, via recruitment of inflammatory cells that could, in turn, release endothelial mitogens (45). However, much of the attention was directed toward two related potent endothelial cell mitogens and angiogenic factors, acidic and basic fibroblast growth factors (aFGF and bFGF) (46). In 1985, the purification to homogeneity and sequencing of both aFGF (47) and bFGF (48) were reported, and the subsequent year their cDNAs were cloned (49, 50). An unexpected finding was that the genes for both aFGF and bFGF do not encode for a conventional secretory signal peptide. Accordingly, it became clear that these molecules are not efficiently secreted and are mostly cell associated (46). Yet, as previously noted, earlier reports had pointed toward the involvement of diffusible factors in tumor angiogenesis (41, 42). This requirement appeared to be true also for physiological angiogenesis such as that associated with corpus luteum (CL) development (51). Vlodavsky et al. (52) suggested that the FGFs are sequestered and stored in the extracellular matrix (ECM) bound to heparan sulfate-containing proteoglycans and can be released in a soluble form when the ECM is degraded. However, several studies suggested that immunoneutralization of bFGF had little or no effect on tumor angiogenesis (53, 54). Furthermore, bFGF-null mice, and even double knockout mice with disruptions in aFGF and bFGF genes, do not develop vascular defects (55, 56). A plausible explanation is that several soluble members of the FGF family compensate for the absence of the cell-associated forms. Thus, the ability of growth factors to promote angiogenesis in in vitro or in vivo bioassays does not necessarily predict a role for such factors in physiological or pathological angiogenesis (57, 58).


    III. Identification of VEGF
 Top
 Abstract
 I. Introduction
 II. Historical Note on...
 III. Identification of VEGF
 IV. Activities of VEGF
 V. VEGF Isoforms
 VI. Regulation of VEGF...
 VII. VEGFRs
 VIII. Role of VEGF...
 IX. Role of VEGF...
 X. VEGF and Therapeutic...
 XI. Perspectives
 References
 
Independent and unrelated lines of research converged toward the identification of VEGF (see Ref.13 for additional details).

In 1983, Senger et al. (59) described the partial purification from the conditioned medium of a guinea-pig tumor cell line of a protein able to induce vascular leakage in the skin, which was named "tumor vascular permeability factor" (VPF). The authors proposed that VPF could be a mediator of the high permeability of tumor blood vessels. Because VPF was not isolated and sequenced, this factor remained molecularly unknown at that time. Senger et al. (60) reported the purification and NH2-terminal amino acid sequencing of guinea pig VPF in 1990.

In 1989, Ferrara and Henzel (61) reported the isolation of a diffusible endothelial cell-specific mitogen from medium conditioned by bovine pituitary follicular cells, which they named "vascular endothelial growth factor" to reflect the restricted target cell specificity of this molecule. NH2-terminal amino acid sequencing of purified VEGF proved that this protein was distinct from the known endothelial cell mitogens such as aFGF or bFGF and indeed did not match any known protein in available databases (61). Subsequently, Connolly et al. (62) followed up on the work by Senger et al. and independently reported the isolation and sequencing of human VPF from U937 cells. cDNA cloning of VEGF (63) and VPF (64), reported also in 1989, demonstrated that VEGF and VPF were the same molecule. This was surprising, considering that other endothelial cell mitogens such as FGF do not increase vascular permeability. The finding that VEGF is potent, diffusible, and specific for vascular endothelial cells led to the hypothesis that this molecule might play a role in the regulation of physiological and pathological growth of blood vessels (61, 63, 65).


    IV. Activities of VEGF
 Top
 Abstract
 I. Introduction
 II. Historical Note on...
 III. Identification of VEGF
 IV. Activities of VEGF
 V. VEGF Isoforms
 VI. Regulation of VEGF...
 VII. VEGFRs
 VIII. Role of VEGF...
 IX. Role of VEGF...
 X. VEGF and Therapeutic...
 XI. Perspectives
 References
 
A. Mitogenesis, angiogenesis, and endothelial survival
A well-documented in vitro activity of VEGF is the ability to promote growth of vascular endothelial cells derived from arteries, veins, and lymphatics (for review see Ref.30). VEGF promotes angiogenesis in tridimensional in vitro models, inducing confluent microvascular endothelial cells to invade collagen gels and form capillary-like structures (66, 67). Also, VEGF induces sprouting from rat aortic rings embedded in a collagen gel (68). VEGF also elicits a pronounced angiogenic response in a variety of in vivo models including the chick chorioallantoic membrane (63, 69), the rabbit cornea (70), the matrigel plug in mice (71), the primate iris (72), etc. VEGF delivery also induces lymphangiogenesis in mice, at least in some circumstances (73). Ergun et al. (74) recently proposed that induction of carcinoembryonic antigen-related cell adhesion 1, a membrane glycoprotein expressed in some microvascular endothelial cells, mediates some of the angiogenic effects of VEGF.

VEGF is also a survival factor for endothelial cells, both in vitro and in vivo (75, 76, 77, 78, 79). In vitro, VEGF prevents endothelial apoptosis induced by serum starvation. Such activity is mediated by the phosphatidylinositol 3-kinase (PI3 kinase)/Akt pathway (77, 80). Also, VEGF induces expression of the antiapoptotic proteins Bcl-2, A1 (76), XIAP (81), and survivin (82) in endothelial cells. In vivo, the prosurvival effects of VEGF are developmentally regulated. VEGF inhibition results in extensive apoptotic changes in the vasculature of neonatal, but not adult, mice (83). Furthermore, a marked VEGF dependence has been demonstrated in endothelial cells of newly formed but not of established vessels within tumors (78, 79). Coverage by pericytes has been proposed to be one of the key events resulting in loss of VEGF dependence (78).

Although endothelial cells are the primary targets of VEGF, several studies have reported mitogenic effects also on certain nonendothelial cell types, such as retinal pigment epithelial cells (84), pancreatic duct cells (85), and Schwann cells (86). Compernolle et al. (87) have also shown that VEGF stimulates surfactant production by alveolar type II cells, resulting in a protective effect from respiratory distress syndrome in mice. Recent studies have emphasized the potential role of VEGF as a neuronal protective factor, and a haplotype in the VEGF gene promoter associated with reduced VEGF expression has been reported to be is a risk factor for amyotrophic lateral sclerosis (88).

B. Effects of VEGF on bone marrow cells and hematopoiesis
The earliest evidence that VEGF can affect blood cells came from a report describing its ability to promote monocyte chemotaxis (89). Subsequently, VEGF was reported to have hematopoietic effects, inducing colony formation by mature subsets of granulocyte-macrophage progenitor cells (90). Interestingly, VEGF delivery to adult mice inhibits dendritic cell development (91, 92), leading to the hypothesis that VEGF facilitates tumor growth by allowing escape of tumors from the host immune system. Also, VEGF increased production of B cells and the generation of immature myeloid cells (93). Recently, conditional gene knockout technology has been employed to achieve selective VEGF gene ablation in bone marrow cell isolates and hematopoietic stem cells (HSCs) (94). VEGF-deficient HSCs and bone marrow mononuclear cells failed to repopulate lethally irradiated hosts, despite coadministration of a large excess of wild-type cells. These studies elucidated an internal autocrine loop, not blocked by extracellular inhibitors such as antibodies, whereby VEGF controls HSC survival during hematopoietic repopulation (94).

Interestingly, a VEGF-dependent pathway has been shown to play an important role in hematopoiesis even in Drosophila, where it controls migration (95) and proliferation (96) of blood cells. Three VEGF-like ligands and a single receptor, known as PDGF/VEGF receptor or PVR, have been identified in Drosophila (97). Because Drosophila is devoid of a vascular system, these findings indicate that one ancestral, conserved role of VEGF is indeed the regulation of blood cell function (97).

C. Enhancement of vascular permeability and hemodynamic effects
As previously noted, VEGF is known also as VPF, based on its ability to induce vascular leakage (59, 98). Such permeability-enhancing activity underlies important roles of this molecule in inflammation and other pathological circumstances (see Section IX.D). Bates and Curry (99) have shown that VEGF induces an increase in hydraulic conductivity of isolated microvessels, an effect that is mediated by increased calcium influx (100). Consistent with a role in the regulation of vascular permeability, VEGF induces endothelial fenestration in some vascular beds (101) and in cultured adrenal endothelial cells (102).

Several studies have pointed to the critical role of nitric oxide (NO) in VEGF-induced vascular permeability, as well as angiogenesis (103, 104, 105). Recently, Fukumura et al. (106) assessed the relative contribution of the NO synthase (NOS) isoforms, inducible NOS and endothelial NOS (eNOS) to these processes. Angiogenesis, vessel diameter, blood flow rate, and vascular permeability were proportional to NO levels and were most impaired in eNOS–/– mice. VEGF significantly increased permeability in both wild-type and inducible NOS–/– mice, but not in eNOS–/– mice. VEGF-induced angiogenesis was markedly reduced in eNOS–/– mice, although the mice develop normally and have no apparent defect in the vasculature. These findings suggest that, although eNOS plays a predominant role in angiogenesis and vascular permeability in response to exogenous VEGF, this pathway is dispensable for developmental angiogenesis.

An issue that has been long debated is whether a correlation exists between vascular permeability and angiogenesis. It has been proposed that increase in microvascular permeability is a step necessary and sufficient for angiogenesis, by providing extravasation of fibrin, which represents a scaffold for endothelial cell proliferation and migration (107). However, as previously mentioned, factors such as bFGF are not known to induce vascular permeability and yet potently induce angiogenesis. Furthermore, vascular leakage is not necessarily followed by angiogenesis. For example, in background diabetic retinopathy, vascular leakage and fibrin deposition (108) may occur in the retina for decades before the onset of angiogenesis in the proliferative phase (109, 110). The report by Eliceiri et al. (111) that members of the Src family are differentially involved in mediating VEGF-dependent permeability and angiogenesis showed that the permeability-enhancing activity specifically depends on Src, or Yes. Mice lacking src and Yes display a normal angiogenic response to VEGF without any overt defects in the vasculature, suggesting that enhanced vascular permeability is not a requirement for VEGF-dependent angiogenesis, at least in the circumstances examined to date. Recently, Gratton et al. (112) have reported that a peptide that prevents the association of eNOS with caveolin inhibits vascular permeability and tumor progression in mice. However, additional studies have emphasized the complexity of the role of eNOS in tumorigenesis, including a role in the recruitment of endothelial progenitor cells (113) as well as a requirement for angiogenesis (114). Clearly, further studies are needed to fully elucidate this complex issue.

VEGF induces vasodilatation in vitro in a dose-dependent fashion (115, 116) and produces transient tachycardia, hypotension, and a decrease in cardiac output when injected iv in conscious, instrumented rats (116). Such effects appear to be caused by a decrease in venous return, mediated primarily by endothelial cell-derived NO (116). Hypotension was a dose-limiting side effect in human trials in which VEGF was systemically administered (117). Conversely, administration of anti-VEGF monoclonal antibodies to cancer patients resulted in elevation of blood pressure (36), indicating that VEGF signaling plays a tonic homeostatic role in the regulation of blood pressure. The mechanism is likely to involve eNOS, but remains to be fully elucidated.


    V. VEGF Isoforms
 Top
 Abstract
 I. Introduction
 II. Historical Note on...
 III. Identification of VEGF
 IV. Activities of VEGF
 V. VEGF Isoforms
 VI. Regulation of VEGF...
 VII. VEGFRs
 VIII. Role of VEGF...
 IX. Role of VEGF...
 X. VEGF and Therapeutic...
 XI. Perspectives
 References
 
VEGF has significant homology to PDGF, and all the eight cysteines found in the A and B chains of PDGF are conserved in VEGF (63, 64). The human VEGF-A gene is organized in eight exons, separated by seven introns (118, 119) and is localized in chromosome 6p21.3 (120). Alternative exon splicing results in the generation of four different isoforms, having 121, 165, 189, and 206 amino acids, respectively, after signal sequence cleavage (VEGF121, VEGF165, VEGF189, VEGF206) (118, 119). VEGF165, the predominant isoform, lacks the residues encoded by exon 6, whereas VEGF121 lacks the residues encoded by exons 6 and 7. Less frequent splice variants have been also reported, including VEGF145 (121), VEGF183 (122), VEGF162 (123), and VEGF165b, a variant reported to have paradoxically an inhibitory effect on VEGF-induced mitogenesis (124).

VEGF is a heparin-binding homodimeric glycoprotein of 45 kDa (61). Such properties closely correspond to those of VEGF165, which is indeed the major VEGF isoform (125).

Solution of the crystal structure of VEGF has shown that VEGF forms an antiparallel homodimer covalently linked by two disulfide bridges between Cys-51 and Cys-60 (126). This mode of dimerization is similar to that of the PDGF monomers. The dominant feature within the VEGF monomer is the cystine knot motif that is found in other growth factors (126). Although the VEGF monomer resembles that of PDGF, its NH2-terminal segment is helical rather than extended (126).

VEGF121 is an acidic polypeptide that fails to bind to heparin (125). VEGF189 and VEGF206 are highly basic and bind to heparin with high affinity (125). VEGF121 is a freely diffusible protein. In contrast, VEGF189 and VEGF206 are almost completely sequestered in the ECM. VEGF165 has intermediate properties, because it is secreted, but a significant fraction remains bound to the cell surface and ECM (127). The ECM-bound isoforms may be released in a diffusible form by heparin or heparinase, which displaces them from their binding to heparin-like moieties, or by plasmin cleavage at the COOH terminus, which generates a bioactive fragment consisting of the first 110 NH2-terminal amino acids (125). Given the important role of plasminogen activation during physiological and pathological angiogenesis processes (128), this proteolytic mechanism can be particularly important in regulating locally the activity and bioavailability of VEGF.

Plouet et al. (129) have proposed a role for urokinase in the generation of bioactive VEGF. Recombinant VEGF189 from insect cells infected with a recombinant baculovirus was purified as a nonmitogenic 50-kDa precursor that binds to the receptor VEGFR-1 but not to VEGFR-2. However, it could be matured by urokinase as a 38-kDa fragment able to promote endothelial cell proliferation (129). Figure 1Go illustrates the properties of the VEGF isoforms.



View larger version (32K):
[in this window]
[in a new window]
 
FIG. 1. VEGF isoforms and their interaction with VEGFRs. The diffusible VEGF isoforms, VEGF121 and VEGF165, are released by a variety of normal and transformed cells (the figure shows tumor cells) and may bind to VEGFR-1 (R1) and VEGFR-2 (R2). VEGF165, but not VEGF121, interacts also with NP1 and NP2. This binding results in enhancement of VEGFR-2-dependent signaling in endothelial cells. After plasmin generation and ECM breakdown, VEGF189 is cleaved at the COOH terminus, and the resulting 110-amino acid NH2-terminal fragment is diffusible and bioactive.

 
Importantly, loss of the heparin-binding domain results in a reduction in the mitogenic activity of VEGF (130). These findings suggest that VEGF165 has optimal characteristics of bioavailability and biological potency. In agreement with such conclusions, only VEGF164 (murine VEGF is shorter by one amino acid) is able to fully rescue a tumorigenic phenotype in mouse VEGF–/– cells (131). The significance of the heparin-binding VEGF isoform(s) is also emphasized by the finding that 50% of the mice expressing exclusively VEGF120 (VEGF120/120) die shortly after delivery, whereas the rest die within 2 wk (132). Recent studies have also evidenced a deficit in the distribution of endothelial cells and impaired filopodia extension in VEGF120/120 mice, suggesting that the heparin-binding VEGF isoforms provide essential stimulatory cues to initiate vascular branch formation (133).


    VI. Regulation of VEGF Gene Expression
 Top
 Abstract
 I. Introduction
 II. Historical Note on...
 III. Identification of VEGF
 IV. Activities of VEGF
 V. VEGF Isoforms
 VI. Regulation of VEGF...
 VII. VEGFRs
 VIII. Role of VEGF...
 IX. Role of VEGF...
 X. VEGF and Therapeutic...
 XI. Perspectives
 References
 
A. Oxygen tension
Oxygen tension plays a key role in regulating the expression of a variety of genes (134). VEGF mRNA expression is induced by exposure to low pO2 in a variety of pathophysiological circumstances (135, 136). Earlier studies indicated similarities between the mechanisms leading to hypoxic regulation of VEGF and erythropoietin (Epo) (137). Hypoxia inducibility is conferred on both genes by homologous sequences. A 28-base sequence has been identified in the 5'-promoter of the rat and human VEGF gene, which mediates hypoxia-induced transcription (138, 139). Such sequence reveals a high degree of homology and similar protein binding characteristics as the hypoxia-inducible factor 1 (HIF-1) binding site within the Epo gene (140). HIF-1 is a basic, heterodimeric, helix-loop-helix protein consisting of two subunits, HIF-1{alpha} and aryl hydrocarbon receptor nuclear translocator, known also as HIF-1ß (141). It is now well established that HIF-1 is a key mediator of hypoxic responses (142). In response to hypoxia, HIF-1 binds to specific enhancer elements, resulting in increased gene transcription. A gene highly homologous to HIF-1, HIF-2, also forms heterodimers with aryl hydrocarbon receptor nuclear translocator and regulates VEGF expression (143). Recent studies have uncovered the critical role of the product of the von Hippel-Lindau (VHL) tumor suppressor gene in HIF-1-dependent hypoxic responses (for review see Ref.144). The VHL gene is inactivated in patients with von Hippel-Lindau disease, an autosomal-dominant neoplasia syndrome characterized by capillary hemangioblastomas in retina and cerebellum, and in most sporadic clear cell renal carcinomas (145). Also, the mitogenic activity for endothelial cells in the conditioned medium of renal cell carcinoma cells expressing a mutant VHL was largely neutralized by anti-VEGF antibodies (146). Earlier studies indicated that a function of the VHL protein is to provide negative regulation of VEGF and other hypoxia-inducible genes (147). The spectrum of activities of the VHL protein remains to be fully elucidated, and multiple functions have been proposed, including interaction with fibronectin (148). However, the VHL protein is known to interact with a series of proteins including elongins B and C and CUL2, a member of the Cullin family (149), suggesting homology to yeast ubiquitin ligase complexes known as "SCF complexes." HIF-1 was shown to be constitutively activated in VHL-deficient renal cell carcinoma cell lines (150). More recent studies demonstrated that, indeed, one of the functions of VHL is to be part of a ubiquitin ligase complex that targets HIF subunits for proteasomal degradation after covalent attachment of a polyubiquitin chain (151, 152). Oxygen promotes the hydroxylation of HIF at a proline residue, a requirement for the association with VHL (151, 152). Recently, a family of prolyl hydroxylases related to Egl-9 Caenorhabditis elegans gene product were identified as HIF prolyl hydroxylases (134, 153, 154).

Importantly, other studies have implicated the PI3 kinase/Akt pathway in the regulation of HIF-mediated responses in a hypoxia-independent manner. Zundel et al. (155) have shown that mutations resulting in loss of function of the tumor suppressor PTEN, which negatively regulates effectors of PI3 kinase/Akt and is mutated in glioblastoma and other tumors (156, 157), result in increased activation of HIF-1 and increased VEGF transcription. Tang and Lasky (158) have recently elucidated the role of the Forkhead transcription factor (FOXO4) in this pathway. Nuclear localization of Forkhead, which is inhibited by PI3 kinase activation, normally down-regulates the HIF-1 protein. These findings emphasize the multiple advantages conferred on tumor cells by PI3 kinase/Akt activation.

B. Growth factors, hormones, and oncogenes
Several major growth factors, including epidermal growth factor, TGF-{alpha}, TGF-ß, keratinocyte growth factor, IGF-I, FGF, and PDGF, up-regulate VEGF mRNA expression (159, 160, 161), suggesting that paracrine or autocrine release of such factors cooperates with local hypoxia in regulating VEGF release in the microenvironment. Also, inflammatory cytokines such as IL-1-{alpha} and IL-6 induce expression of VEGF in several cell types, including synovial fibroblasts, in agreement with the hypothesis that VEGF may be a mediator of angiogenesis/permeability in inflammatory disorders (162, 163).

Hormones are also important regulators of VEGF gene expression. TSH has been shown to induce VEGF expression in several thyroid carcinoma cell lines (164). Shifren et al. (165) have also shown that ACTH is able to induce VEGF expression in cultured human fetal adrenal cortical cells, suggesting that VEGF may be a local regulator of adrenal cortical angiogenesis and an important mediator of the tropic action of ACTH. Gonadotropins have been shown to be potent inducers of VEGF transcription in the ovary, both in vivo (166, 167) and in vitro (168). Also, human chorionic gonadotropin results in increased VEGF mRNA transcription and protein levels in cultured Leydig cells (169). Several studies have implicated sex steroids as an important stimulus for VEGF regulation in hormone-sensitive tissues. In vitro, androgen deprivation of LnCaP prostate cancer cells led to decreased VEGF mRNA and protein expression as well as a 5-fold destabilization in VEGF mRNA transcripts. In mice bearing LnCaP tumors, castration resulted in a rapid decrease in mRNA expression and markedly reduced tumor neovascularization (170). Mueller et al. (171) have reported that estradiol is a direct transcriptional activator of VEGF, mediated by a variant estrogen response element located 1.5 kb from the transcription start. Progestins have also been reported to induce VEGF gene transcription in endometrial carcinoma cells (172).

Specific transforming events also result in induction of VEGF gene expression. Oncogenic mutations or amplification of ras leads to VEGF up-regulation (173, 174). These studies indicate that mutant ras-dependent VEGF expression is necessary, albeit not sufficient, for progressive tumor growth in vivo (173, 174, 175). Mutations in the Wnt-signaling pathway, which are frequently associated with premalignant colonic adenomas, result in up-regulation of VEGF (176). K-ras activation appeared to enhance Wnt signaling, which suggests an interaction between these two pathways (176). Interestingly, VEGF is up-regulated in polyps of Apc knockout (Apc{delta}716) mice, a model for human familial adenomatous polyposis (177). In both benign and malignant mouse intestinal tumors, stromal expression of cyclooxygenase 2 results in elevated PGE2 levels that stimulate, in turn, cell surface receptor EP2, followed by induction of VEGF and angiogenesis (177, 178, 179). In this context, Amano et al. (180) have recently shown that PGE2-EP3 receptor signaling also plays a significant role in up-regulating VEGF in stromal cells and thus potentially in tumor angiogenesis.


    VII. VEGFRs
 Top
 Abstract
 I. Introduction
 II. Historical Note on...
 III. Identification of VEGF
 IV. Activities of VEGF
 V. VEGF Isoforms
 VI. Regulation of VEGF...
 VII. VEGFRs
 VIII. Role of VEGF...
 IX. Role of VEGF...
 X. VEGF and Therapeutic...
 XI. Perspectives
 References
 
Initially, VEGF binding sites were identified on the cell surface of vascular endothelial cells in vitro (181, 182) and in vivo (183, 184). Subsequently, VEGFRs were shown to exist also on bone marrow-derived cells such as monocytes (185). VEGF binds two highly related receptor tyrosine kinases (RTKs), VEGFR-1 and VEGFR-2. Both VEGFR-1 and VEGFR-2 have seven Ig-like domains in the extracellular domain, a single-transmembrane region, and a consensus tyrosine kinase sequence that is interrupted by a kinase-insert domain (186, 187, 188). A member of the same family of RTKs is VEGFR-3 (Flt-4) (189), which, however, is not a receptor for VEGF, but instead binds VEGF-C and VEGF-D (27). In addition to these RTKs, VEGF interacts with a family of coreceptors, the neuropilins (NP). Figure 2Go summarizes the interaction of the members of the VEGF gene family with the VEGF RTKs.



View larger version (29K):
[in this window]
[in a new window]
 
FIG. 2. Role of the VEGFR tyrosine kinases in endothelial cells. VEGFR-1 (R1) and VEGFR-2 (R2) are expressed in the cell surface of most blood endothelial cells. In contrast, VEGFR-3 (R3) is largely restricted to lymphatic endothelial cells. VEGF-A binds both VEGFR-1 and VEGFR-2. In contrast, PlGF and VEGF-B interact only with VEGFR-1. VEGF-C and VEGF-D bind VEGFR-2 and VEGFR-3. There is much evidence that VEGFR-2 is the major mediator of endothelial cell mitogenesis, survival, and microvascular permeability. In contrast, VEGFR-1 does not mediate an effective mitogenic signal in endothelial cells and it may, especially during early embryonic development, perform an inhibitory role by sequestering VEGF and preventing its interaction with VEGFR-2. Such a "decoy" role could be also performed by the alternatively spliced soluble VEGFR-1. EC, Endothelial cell; uPA, urokinase-type plasminogen activator; tPA, tissue-type plasminogen activator.

 
A. VEGFR-1 (Flt-1)
Although Flt-1 (fms-like tyrosine kinase) was the first RTK to be identified as a VEGFR more than a decade ago (190), the precise function of this molecule is still the object of debate. Recent evidence indicates that the conflicting reports may be due, at least in part, to the fact that VEGFR-1 functions and signaling properties can be different depending on the developmental stage and the cell type, e.g., endothelial vs. hematopoietic cells. VEGFR-1 expression is up-regulated by hypoxia via a HIF-1-dependent mechanism (191). VEGFR-1 binds not only VEGF-A but also PlGF (192) and VEGF-B (193), which fail to bind VEGFR-2. An alternatively spliced soluble form of VEGFR-1 (sFlt-1) has been shown to be an inhibitor of VEGF activity (194). The binding site for VEGF (and PlGF) has been mapped primarily to the second Ig-like domain (195, 196, 197). The crystal structure of a VEGF-Flt-1 domain 2 complex has shown the poles of the VEGF dimer to be in a predominantly hydrophobic interaction with domain 2 (198). Figure 3Go illustrates the complex VEGF-VEGFR-1 domain 2 in a ribbon format.



View larger version (31K):
[in this window]
[in a new window]
 
FIG. 3. Ribbon representation of the VEGF-VEGFR1-domain 2 complex. The two VEGF monomers are shown in blue and yellow, and the two receptor molecules are depicted in green. This is a "top down" view and shows the complex looking toward the membrane. The termini are labeled. [Reproduced with permission from C. Wiesmann.]

 
Flt-1 reveals a weak tyrosine autophosphorylation in response to VEGF (190, 199). Park et al. (192) initially proposed that VEGFR-1 may be not primarily a receptor transmitting a mitogenic signal, but rather a "decoy" receptor, able to regulate in a negative fashion the activity of VEGF on the vascular endothelium, by sequestering and rendering this factor less available to VEGFR-2 (see Fig. 2Go). Thus, the observed potentiation of the action of VEGF by PlGF could be explained, at least in part, by displacement of VEGF from VEGFR-1 binding (192). Not only the full-length membrane-bound form of VEGFR-1, but also sFlt-1, could perform such a decoy function (200). Recent studies have shown that, indeed, a synergism exists between VEGF and PlGF in vivo, especially during pathological situations, as evidenced by impaired tumorigenesis and vascular leakage in Plgf–/– mice (200). Gille et al. (201) have identified a repressor motif in the juxtamembrane region of VEGFR-1 that impairs PI3 kinase activation and endothelial cell migration in response to VEGF. Zeng et al. (202) have proposed that VEGFR-1 activation results in inhibition of VEGFR-2-dependent endothelial cell proliferation and that this inhibitory pathway is PI3 kinase dependent. However, other studies indicated that VEGFR-1 is able to interact with various signal-transducing proteins and generate, in some circumstances, a mitogenic signal (203, 204). Very recently, Autiero et al. (205) have proposed that PlGF regulates inter- and intramolecular cross-talk between the VEGF RTKs. Activation of VEGFR-1 by PlGF resulted in transphosphorylation of VEGFR-2, thus amplifying VEGF-driven angiogenesis through VEGFR-2. According to these studies, although VEGF and PlGF both bind VEGFR-1, PlGF uniquely stimulated the phosphorylation of specific VEGFR-1 tyrosine residues, and this results in the expression of distinct target genes (205). This finding is somewhat surprising, considering that PlGF and VEGF bind to the same binding interface of VEGFR-1 in a very similar fashion (206).

Irrespective of the conflicting evidence on the role of VEGFR-1 as a signaling receptor, gene-targeting studies have demonstrated the essential role of this molecule during embryogenesis. Flt-1–/– mice die in utero between d 8.5 and d 9.5 (207, 208). Endothelial cells develop but fail to organize in vascular channels. Excessive proliferation of angioblasts has been reported to be responsible for such disorganization and lethality (208), indicating that, at least during early development, VEGFR-1 is a negative regulator of VEGF action. More compelling evidence in support of this view stems from the report that a targeted mutation resulting in a VEGFR-1 lacking the tyrosine kinase (TK) domain, but able to bind VEGF, does not result in lethality or any overt defect in vascular development (209). Nevertheless, one specific biological response, the migration of monocytes in response to VEGF (or PlGF) has been shown to require the tyrosine kinase domain of VEGFR-1 (209, 210) (Fig. 4Go). Selvaraj et al. (211) have shown recently that PlGF binding to VEGFR-1 in monocytes results in activation of PI3 kinase/AKT and ERK-1/2 pathways, leading to chemotaxis as well as to the induction of a series of inflammatory cytokines. Furthermore, Lewis lung carcinoma cells overexpressing PlGF grow in wild-type mice faster than in VEGFR-1 tyrosine kinase-deficient mice, suggesting that VEGFR-1 may be a positive regulator under pathological conditions when a VEGFR-1-specific ligand is highly expressed (212). These findings suggest that VEGFR-1 has a dual function in angiogenesis, acting in a positive or negative manner in different circumstances. Recently, VEGFR-1 signaling has been also linked to the induction of matrix metalloproteinase 9 (MMP-9) in lung endothelial cells and to the facilitation of lung metastases (213). Recent studies have emphasized the effects of VEGFR-1 in hematopoiesis and recruitment of endothelial progenitors. Hattori et al. (214) have shown that VEGFR-1 activation by PlGF is able to reconstitute hematopoiesis by recruiting VEGFR-1+ HSC. In addition, Gerber et al. (94) have shown that VEGFR-1 activation by enforced expression of PlGF rescues survival and ability to repopulate in VEGF–/– HSC. Furthermore, PlGF can promote collateral vessel growth and arteriogenesis in models of myocardial and limb ischemia through the recruitment of bone marrow cells such as monocytes (215, 216).



View larger version (20K):
[in this window]
[in a new window]
 
FIG. 4. Role of VEGFR-1 (R1) in monocyte chemotaxis. VEGFR-1 activation by VEGF-A, PlGF, or VEGF-B (data not shown) results in monocyte recruitment in normal or transformed tissues. Monocytes may, in turn, produce VEGF-A and other angiogenic factors, which may amplify the angiogenesis cascade.

 
LeCouter et al. (217) recently provided evidence for a novel function of VEGFR-1 in liver sinusoidal endothelial cells (LSECs). VEGFR-1 activation achieved with a receptor- selective VEGF mutant or PlGF resulted in the paracrine release of hepatocyte growth factor, IL-6, and other hepatotrophic molecules by LSECs, to the extent that hepatocytes were stimulated to proliferate when cocultured with LSECs. VEGF had no direct mitogenic effect on hepatocytes. A VEGFR-1 agonist protected the liver from CCl4-induced damage, in spite of the inability to induce LSEC proliferation (Fig. 5Go). These findings suggest that a key function of VEGFR-1 signaling in the vascular endothelium is not the regulation of angiogenesis but, rather, the paracrine release of tissue-specific growth/survival factors, possibly in a vascular bed-specific fashion (217). In this context, liver and pancreas morphogenesis is induced by endothelial cells before the establishment of a blood flow, indicating that a paracrine function of gut endothelial cells plays a critical role during organogenesis (218, 219).



View larger version (26K):
[in this window]
[in a new window]
 
FIG. 5. Differential effects of VEGFR-1 and VEGFR-2 in LSECs. In response to VEGFR-1 activation, LSECs are not stimulated to proliferate but are instructed to up-regulate a series of hepatotrophic genes, including hepatocyte growth factor (HGF), IL-6, and heparin-binding epidermal growth factor (HB-EGF). Thus, VEGFR-1 agonists may result in significant hepatocellular protection from hepatotoxins, without stimulation of angiogenesis. VEGFR-2 activation not only mediates LSEC proliferation, migration, and survival, but also results in induction of a subset of hepatotrophic genes. CTGF, Connective tissue growth factor.

 
B. VEGFR-2 (KDR, human; Flk-1, mouse)
VEGFR-2 binds VEGF, albeit with lower affinity relative to VEGFR-1 [dissociation constant (Kd) 75–250 pM vs. 25 pM] (220, 221, 222). The key role of this receptor in developmental angiogenesis and hematopoiesis is evidenced by lack of vasculogenesis and failure to develop blood islands and organized blood vessels in Flk-1 null mice, resulting in death in utero between d 8.5 and d 9.5 (223). Consistent with a role in hematopoiesis, VEGFR-2 has been identified on a subset of multipotent human HSCs (224). There is now general agreement that VEGFR-2 is the major mediator of the mitogenic, angiogenic, and permeability-enhancing effects of VEGF.

The binding site for VEGF has been mapped to the second and third Ig-like domains (225). VEGFR-2 undergoes dimerization and strong ligand-dependent tyrosine phosphorylation in intact cells and results in a mitogenic, chemotactic, and prosurvival signal. Several tyrosine residues have been shown to be phosphorylated (for review see Ref.226). Takahashi et al. (227) have shown that Y1175 and Y1214 are the two major VEGF-A-dependent autophosphorylation sites in VEGFR-2. However, only autophosphorylation of Y1175 is crucial for VEGF-dependent endothelial cell proliferation. Also, VEGF has been shown to induce the phosphorylation of at least 11 proteins in bovine aortic endothelial cells (228). Among these, VEGF induces phosphorylation of phospholipases C{gamma}, PI3-kinase, ras GTPase activating protein (228), src family (111), and several other signal transduction molecules (226). Byzova et al. (229) have reported that VEGFR-2 activation by VEGF results in PI3 kinase/Akt-dependent activation of several integrins. VEGF enhanced cell adhesion, migration, soluble ligand binding, and adenovirus gene transfer mediated by {alpha}vß3 and also activated other integrins known to be involved in angiogenesis, {alpha}vß5, {alpha}5ß1, and {alpha}2ß1 (229). VEGFR-2 activation induces endothelial cell growth by activating the Raf-Mek-Erk pathway. An unusual feature of VEGFR-2 activation of this pathway is the requirement for protein kinase C but not ras (230, 231). VEGF mutants that bind selectively to VEGFR-2 are fully active endothelial cell mitogens, chemoattractants, and permeability-enhancing agents, whereas mutants specific for VEGFR-1 are devoid of all three activities (232). Also, VEGF-E, a homolog of VEGF identified in the genome of the parapoxvirus Orf virus (24), which shows VEGF-like mitogenic and permeability- enhancing effects, binds and activates VEGFR-2 but fails to bind VEGFR-1 (25, 26). Interestingly, similar biological effects and receptor selectivity have been recently reported with snake-derived VEGF (233). Furthermore, VEGFR-2 (but not VEGFR-1) activation has been shown to be required for the antiapoptotic effects of VEGF for human umbilical vein endothelial cells (77). As previously noted, such a prosurvival effect of VEGF is mediated by the PI3 kinase/Akt pathway (77). This pathway is critical also for VEGF-dependent endothelial chemotaxis (201, 232, 234). Recent studies suggest, however, that at least in some circumstances, VEGFR-1 may transmit a prosurvival signal in endothelial cells, possibly mediated by induction of the antiapoptotic gene survivin (235).

C. Neuropilin (NP)1 and NP2
Earlier studies indicated that certain tumor and endothelial cells express cell surface VEGF-binding sites distinct in affinity and molecular mass from the two known VEGF RTKs (236). Interestingly, VEGF121 failed to bind these sites, indicating that exon 7-encoded basic sequences are required for binding to this putative receptor (236). Subsequently, Soker et al. (237) identified such isoform-specific VEGF receptor as NP1, a molecule that had been previously shown to bind the collapsin/semaphorin family and was implicated in neuronal guidance (for review see Ref.238). When coexpressed in cells with VEGFR-2, NP1 enhanced the binding of VEGF165 to VEGFR-2 and VEGF165-mediated chemotaxis (237). NP1 appears to present VEGF165 to the VEGFR-2 in a manner that enhances the effectiveness of VEGFR-2-mediated signal transduction (237). Fuh et al. (239) have shown that NP1 is able to directly bind VEGFR-1, suggesting that one of the mechanisms by which VEGFR-1 functions as a negative regulator of VEGF activity is competing for NP1 binding. Binding to NP1 may help to explain the greater mitogenic potency of VEGF165 relative to VEGF121. So far, there is no clear evidence that NP1 or the related NP2 signals after VEGF binding (238). In contrast, in response to semaphorin binding, NP1 and NP2 signals axon repulsion. Interestingly, collapsin 1 is able to inhibit the motility of porcine aortic endothelial cells expressing NP1 (240). Recent evidence indicates that the formation of complexes with plexins is a requirement for NP signaling in neurons (241, 242). The role of NP1 in the development of the vascular system has been demonstrated by gene-targeting studies, documenting embryonic lethality in null mice (243). Furthermore, Lee et al. (244) have shown that, in the zebrafish, NP1 is required for vascular development and mediates VEGF-dependent angiogenesis. Interestingly, recent studies have linked NP2 to lymphatic vessel development (245).


    VIII. Role of VEGF in Physiological Angiogenesis
 Top
 Abstract
 I. Introduction
 II. Historical Note on...
 III. Identification of VEGF
 IV. Activities of VEGF
 V. VEGF Isoforms
 VI. Regulation of VEGF...
 VII. VEGFRs
 VIII. Role of VEGF...
 IX. Role of VEGF...
 X. VEGF and Therapeutic...
 XI. Perspectives
 References
 
A. Embryonic and postnatal development
In 1996, two studies demonstrated an essential role of VEGF in embryonic vasculogenesis and angiogenesis in the mouse (246, 247). Inactivation of a single VEGF allele resulted in embryonic lethality between d 11 and d 12. The vegf+/– embryos exhibited a number of developmental anomalies, defective vascularization in several organs, and a markedly reduced number of nucleated red blood cells within the blood islands in the yolk sac, indicating that VEGF regulates both vasculogenesis and early hematopoiesis. Conditional VEGF gene inactivation in VEGF loxP mice, using a Nestin promoter-driven Cre-recombinase, has shown that the dosage of VEGF from neural progenitor cells is a critical determinant in the development and density of vascular plexus in the developing nervous system, to the extent that severe reductions in VEGF led to decreases in vascularity and subsequent hypoxia, resulting in the specific degeneration of the cerebral cortex and neonatal lethality (248, 249). Conversely, even modest increases in VEGF gene expression, achieved by the insertion of a LacZ cassette in the 3'-untranslated region of the VEGF gene, result in severe abnormalities in heart development and embryonic lethality at embryonic d 12.5 (E12.5)–E14 (250). These findings indicate a critical VEGF gene-dosage dependence during development. In contrast, inactivation of PlGF (200) or VEGF-B (251) genes did not result in any major development abnormalities, although VEGF-B inactivation in mice results in reduced heart size and impaired recovery from experimentally induced myocardial ischemia (251). So far, it appears that, among the other members of the VEGF gene family, only VEGF-C plays an essential role in development, because its inactivation results in embryonic lethality due to defective lymphatic development and fluid accumulation in tissues (252).

To determine the role of VEGF in early postnatal life, several strategies have been employed (83). Partial inhibition of VEGF achieved by Cre-loxP-mediated gene targeting resulted in increased mortality, stunted body growth, and impaired organ development. Administration of a soluble VEGFR-1 chimeric protein, which achieves a nearly complete VEGF inhibition, results in almost complete growth arrest, when the treatment is initiated at d 1 or d 8 postnatally. Endothelial cells isolated from the liver of VEGFR1-IgG-treated neonates demonstrated increased apoptotic index, indicating that VEGF is required not only for proliferation but also for survival of endothelial cells (83). Such treatment is also accompanied by rapid lethality, primarily due to inhibition of glomerular development leading to kidney failure (83). Defective glomerular endothelial development in neonates was also observed in studies using anti-VEGF antibodies (253). The pivotal role of VEGF in kidney development was also demonstrated by a very recent study showing that selective VEGF deletion in podocytes, using a Nephin promoter-driven Cre recombinase, leads to glomerular disease in a gene dosage-dependent fashion (254). Heterozygous mice developed renal disease by 2.5 wk of age, characterized by proteinuria and endotheliosis. Homozygosity resulted in perinatal lethality (254). However, VEGF neutralization in fully developed normal mice (83) or rats (255) had no significant effects on glomerular function. In contrast, VEGF inhibition in adult rats with mesangioproliferative nephritis led to a reduction of glomerular endothelial regeneration and an increase in endothelial cell death, indicating that VEGF may be important for glomerular endothelial cell repair after injury, but not for endothelial survival in a healthy animal (255). In apparent conflict with these conclusions, Sugimoto et al. (256) have recently reported that the administration of anti-VEGF antibodies or sFlt-1 to adult mice results in proteinuria accompanied by glomerular endothelial cell detachment and hypertrophy, in association with down-regulation of nephrin. The reason for such discrepancies is unclear.

Importantly, VEGF neutralization in juvenile primates using a humanized anti-VEGF monoclonal antibody (bevacizumab) did not result in any renal or other significant abnormalities, except the suppression of growth plate and ovarian angiogenesis, as described below (257). Furthermore, as discussed in Section IX.A, long-term administration of bevacizumab to cancer patients resulted in minimal kidney toxicity (36, 258).

B. Skeletal growth and endochondral bone formation
Endochondral bone formation is a fundamental mechanism for longitudinal bone growth. Cartilage, an avascular tissue, is replaced by bone in a process named endochondral ossification (259). VEGF mRNA is expressed by hypertrophic chondrocytes in the epiphyseal growth plate, suggesting that a VEGF gradient is needed for directional growth and cartilage invasion by metaphyseal blood vessels (260, 261). After VEGF blockade with a soluble VEGFR-1 chimeric protein or an anti-VEGF monoclonal antibody, blood vessel invasion is almost completely suppressed, concomitant with impaired trabecular bone formation, in developing mice and primates (257, 260). Although proliferation, differentiation, and maturation of chondrocytes were apparently normal, resorption of hypertrophic chondrocytes was inhibited, resulting in a marked expansion of the hypertrophic chondrocyte zone. Importantly, cessation of the anti-VEGF treatment is followed by capillary invasion, restoration of bone growth, and normalization of the growth plate architecture. Recent studies indicate that VEGF mRNA in osteoblasts is induced by bone morphogenetic proteins (BMPs), suggesting that VEGF produced by osteoblasts in response to BMPs may couple angiogenesis to bone formation (262). Conversely, VEGF may induce BMP-2 expression in endothelial cells, suggesting that endothelial cells may play also an osteogenic role by a BMP-2-dependent stimulation of osteoblasts (263).

Interestingly, a growth plate abnormality similar to that induced by VEGF inhibitors was observed in MMP-9–/– mice (264). Recent evidence indicates that a function of MMP-9 is to render VEGF bioavailable to its receptors (265). VEGF blockade inhibits bone repair (266); MMP-9–/– mice have delayed healing of fractures, and administration of exogenous VEGF corrects this defect (267). Furthermore, VEGF has direct chemotactic and other effects on osteoblasts (268) and osteoclasts (269). These findings indicate not only that VEGF-dependent blood vessel recruitment is essential for coupling cartilage resorption with bone formation, but also that the effects of VEGF on bone homeostasis are complex and involve direct effects on bone cells (270).

A similar, although less dramatic, phenotype was obtained, when VEGF was deleted in the cartilage of developing mice by means of Cre-loxP-mediated, tissue-specific gene ablation (271). Furthermore, examination of VEGF120/120 mice not only revealed a delayed recruitment of blood vessels into the perichondrium but also showed delayed invasion of vessels into the primary ossification center, demonstrating a significant role of heparin-binding VEGF isoform at both an early and later stage of cartilage vascularization (272).

C. Angiogenesis in endocrine glands
Angiogenesis is a key aspect of normal cyclical ovarian function. Follicular growth and the development of the CL are dependent on the proliferation of new capillary vessels (273). The process of selection of a dominant follicle in monovular species has been also associated with angiogenesis, as there is evidence that selected follicles possess a more elaborate microvascular network than other follicles (274). The angiogenesis that accompanies CL development also plays a key role in the delivery of cholesterol to luteal cells for progesterone biosynthesis (275). Subsequently, the blood vessels regress, suggesting the coordinated action of inducers as well as inhibitors of angiogenesis in the course of the ovarian cycle (276, 277).

Previous studies have shown that the VEGF mRNA expression is temporally and spatially related to the proliferation of blood vessels in the ovary (278, 279). Administration of VEGF inhibitors delays follicular development (280) and suppresses luteal angiogenesis in rodents (167, 281) as well as in primates (257, 282, 283, 284). These studies have established that VEGF is indeed the principal regulator of ovarian angiogenesis and that blockade of the VEGF pathway is sufficient to disrupt angiogenesis. Figure 6Go illustrates the experiments that demonstrated, for the first time, such a key role of VEGF, using VEGF-soluble receptors in a rat model of hormonally induced ovulation (167).



View larger version (69K):
[in this window]
[in a new window]
 
FIG. 6. VEGF is essential for CL angiogenesis (167 ). Ovulation was hormonally induced in rats using PMSG, followed by human chorionic gonadotropin (hCG), and this treatment resulted in a dramatic increase in ovarian weight and vascularity 5 d after PMSG administration. Animals were given human (h) or mouse (m) Flt(1–3)-IgG, which potently inhibits VEGF, or control proteins (CD4-IgG or mIgG). Note the complete suppression of ovarian angiogenesis and growth after administration of the VEGF inhibitors. PMSG, Pregnant mare’s serum gonadotropin.

 
More recent studies have indicated that endocrine gland-derived VEGF (EG-VEGF), a novel angiogenic factor that is selectively expressed in steroidogenic tissues, plays a cooperative role with VEGF in the regulation of angiogenesis in the human ovary (285). EG-VEGF is not structurally related to VEGF but belongs to a unique gene family having distant homology to Dickopf, an inhibitor of Wnt signaling (286, 287). A sequential activation of the two genes occurs in the human ovary (288). Whereas VEGF mRNA is strongly expressed in early-stage CL, coincident with the initial development of a capillary plexus, its expression is markedly reduced by midluteal phase. In contrast, EG-VEGF starts being expressed later than VEGF but persists throughout mid- and early-late luteal phase, suggesting that EG-VEGF may be important for the persistence and adequacy of luteal function (288). Thus, the ovary has apparently developed a highly specific local mechanism to complement the action of VEGF. Interestingly, such an acquisition seems to be, at least in part, a late event in evolution and may reflect a greater functional/morphological complexity of organs like the ovary. Although association of human EG-VEGF expression with steroidogenic cells is compelling, the mouse ortholog of this gene has a different expression pattern (289). In this context, a consensus binding site for the NR5A1 orphan nuclear receptor is present within the human EG-VEGF promoter (289). NR5A1, considered to be a key regulator of endocrine development and function (290, 291), regulates multiple target genes involved in gonadal and adrenal determination and development, steroidogenesis, and reproduction (for review see Ref.292). Although rodents have served as models for endocrinology and ovarian physiology, clear differences exist between the rodent and human ovary. The length of the ovarian cycle also distinguishes the human or primate from the rodent. In humans the cycle is 28 d, and in rodents the cycle is completed every 4 d (293). The primate CL is functional for 2 wk before its regression in the infertile cycle, whereas the rodent CL is active for less than 1 d (294).

Recently, the role of VEGF in the development of pancreatic islets has been investigated (295). Deletion of VEGF in the mouse pancreas reveals that endocrine cells signal back to the adjacent endothelial cells to induce the formation of a dense network of fenestrated capillaries in islets. Interestingly, VEGF is not required for the development of all islet capillaries. However, the remaining capillaries found in the VEGF-deficient islets were not fenestrated and contained an unusual number of caveolae. In addition, glucose tolerance tests reveal that the VEGF-induced capillary network is not strictly required for blood glucose control but is essential for fine tuning blood glucose regulation (295).


    IX. Role of VEGF in Pathological Conditions
 Top
 Abstract
 I. Introduction
 II. Historical Note on...
 III. Identification of VEGF
 IV. Activities of VEGF
 V. VEGF Isoforms
 VI. Regulation of VEGF...
 VII. VEGFRs
 VIII. Role of VEGF...
 IX. Role of VEGF...
 X. VEGF and Therapeutic...
 XI. Perspectives
 References
 
A. Solid tumors
Many tumor cell lines secrete VEGF in vitro, suggesting the possibility that this diffusible molecule may be a mediator of tumor angiogenesis (29). In situ hybridization studies have demonstrated that the VEGF mRNA is expressed in the vast majority of human tumors so far examined, including carcinoma of the lung (296, 297), breast (298, 299), gastrointestinal tract (300, 301, 302, 303), kidney (304, 305, 306), bladder (304), ovary (307, 308, 309), and endometrium (310) and several intracranial tumors including glioblastoma multiforme (311, 312, 313) and sporadic, as well as VHL syndrome-associated, capillary hemangioblastoma (314, 315). In glioblastoma multiforme and other tumors with significant necrosis, the expression of VEGF mRNA is highest in hypoxic tumor cells adjacent to necrotic areas (311, 312, 313).

VEGF mRNA is also expressed in endocrine tumors. Expression of VEGF has been demonstrated in a variety of pituitary tumors. Lloyd et al. (316) examined a series of 148 tumors and found that VEGF expression, as assessed by immunohistochemistry, although in general less intense than in normal pituitary tissue, is more prominent in certain adenoma subtypes, especially GH adenomas. Furthermore, carcinomas show increased VEGF expression relative to adenomas, suggesting an up-regulation of VEGF during pituitary tumor progression (316). Soh et al. (317) found the VEGF mRNA to be higher in thyroid cancer cell lines compared with primary cultures of normal thyroid cells and higher in thyroid cancers of follicular than those of parafollicular cell origin. Furthermore, Klein et al. (318) have shown that expression of VEGF by immunohistochemistry is a negative prognostic marker in papillary thyroid carcinoma. The distribution of VEGF and other angiogenic factors in endocrine tumors has been recently reviewed by Turner et al. (319).

In 1993, Kim et al. (320) reported that anti-VEGF monoclonal antibodies exert a potent inhibitory effect on the growth of several tumor cell lines in nude mice, whereas the antibody had no effect on the tumor cells in vitro. Subsequently, many other tumor cell lines were found to be inhibited in vivo by anti-VEGF monoclonal antibodies (321, 322, 323, 324, 325, 326, 327). Tumor growth inhibition was demonstrated also with other anti-VEGF treatments, including a retrovirus-delivered dominant negative Flk-1 mutant (328), small molecule inhibitors of VEGFR-2 signaling (329, 330, 331), antisense oligonucleotides (332, 333), anti-VEGFR-2 antibodies (334), and soluble VEGF receptors (335, 336, 337, 338, 339).

Tumors of endocrine origin are also substantially growth inhibited by anti-VEGF treatment. Treatment with an antihuman VEGF monoclonal antibody resulted in more than 90% inhibition of tumor growth in a model of thyroid cancer (340). Also, administration of PTK787, a small molecule VEGFR-2 kinase inhibitor, led to a 41% reduction of tumor volume in a nude mouse model of poorly differentiated thyroid carcinoma (341).

Although tumor cells usually represent the major source of VEGF, tumor-associated stroma is also an important site of VEGF production (337, 342, 343, 344). As illustrated in Fig. 7Go, chemotactic signals from tumor cells recruit stromals cells, which also produce VEGF and other angiogenic factors. The growth of a variety of human tumor cell lines transplanted in nude mice is substantially reduced, but not completely suppressed, by antihuman VEGF monoclonal antibodies (320). Administration of mFlt (1–3)-IgG, a chimeric receptor containing the first three Ig-like domains of VEGFR-1, that binds both human and mouse VEGF, results in