Over the past decade, the well-documented role of vascular endothelial growth factor (VEGF) in tumor angiogenesis has led it to become one of the leading therapeutic targets for the treatment of cancer. Emerging evidence from genetically modified animal models, however, suggests that elevated levels of VEGF, or a proangiogenic phenotype, may impede, rather than promote, early tumor development and progression. For example, hypermorph VEGF transgenic mice display delayed progression of a retroviral-induced murine leukemia, and knockdown of VEGF expression within the myeloid compartment accelerates tumor progression. Several mechanisms have been proposed to explain this paradox, whereby VEGF induces changes within the hematopoietic compartment and tumor microenvironment through recruitment of tumor inhibitory monocytic cells and the negative regulation of tumor angiogenesis. Thus, it is apparent that the levels of VEGF expression in both tumor and nontumor tissues, as well as the context and timing of its modulation relative to cancer induction, play an important role in determining the effects of VEGF expression on tumorigenicity. In light of these recent findings, the various mechanisms underlying the negative role of VEGF during early tumor development, progression, and metastasis will be discussed. Cancer Res; 70(3); 863–7

A growing body of experimental and clinical evidence has emerged in the last decade to support the concept that mediators of angiogenesis, particularly vascular endothelial growth factor (VEGF), contribute to the development and progression of various malignancies and the formation of metastases. Numerous studies have reported high levels of VEGF expression in a wide variety of tumors that have been associated with poor prognosis (1, 2). This evidence has led to the principle that inhibition of VEGF can block tumor angiogenesis, remove the main nutrient supply, and thereby incapacitate tumor growth. However, evidence from recent clinical trials targeting VEGF challenges this well-accepted principle because monotherapy with anti-VEGF agents has only modestly improved overall survival of cancer patients and tumor resistance is often observed (35). Accordingly, recent experimental evidence from Stockmann et al. and Greenberg et al. suggests that, although anti-VEGF therapy has been modestly successful at inhibiting tumor growth, VEGF can act as a negative regulator of angiogenesis and tumor progression (6, 7). This work questions the idea that maximally sustained levels of VEGF and potentially other proangiogenic factors are required for tumor growth (6, 7). From these and other findings, it is apparent that manipulation of the host toward a proangiogenic phenotype, either systemically or within the hematopoietic compartment, manifests into a protective role against tumor growth in vivo. Although evidence exists in the literature to support the idea that VEGF is a positive regulator of tumor growth, what we would like to elaborate on here is the concept that VEGF also acts as a negative regulator of tumor growth.

Emerging evidence from genetically modified animal models suggests that elevated levels of VEGF, or a proangiogenic phenotype, may impede, rather than promote, early tumor development and progression. Intriguingly, Stockmann and colleagues (7) observed that deleting VEGF in myeloid cells accelerates tumorigenesis in a mouse model of breast and lung cancer, depicting the protective role of VEGF. This study suggests that VEGF, derived from innate immune cells, plays an intricate role in orchestrating angiogenesis and tumorigenesis. The study by Greenberg et al. (6) also suggests that VEGF, on binding to its receptor, VEGF-R2, acts as a negative regulator of angiogenesis by impeding the function of vascular smooth muscle cells (VSMC) and pericytes through induction of a receptor complex with the platelet-derived growth factor receptor (PDGF-Rβ), leading to the loss of vessel stabilization. Consistent with these works, our group reported a tumor inhibitory role for VEGF by showing that a 2-fold overexpression of systemic levels of VEGF in mice heterozygous for a VEGF “hypermorphic” allele decelerates tumorigenesis in a retroviral-induced, spontaneous murine leukemia model (8). Alterations in the innate immune function, specifically enhanced natural killer cell activity, and increased hematopoietic progenitor cell survival were identified as acquired phenotypes that strongly correlated with and were likely responsible for this leukemic inhibition (8). Similarly, in the same study, the capacity of erythropoietin to confer deceleration of leukemia progression was also shown (8). This correlation was not unexpected because Epo, like VEGF, is known to be a proangiogenic hormone and exhibits pleiotropic properties (911). Recently, Nagy et al. (12) have similarly used VEGF transgenic mice to show an inhibitory role for VEGF in a murine breast cancer model. Taken together, these studies have provided evidence to support several mechanisms whereby VEGF inhibits the growth and progression of various cancer types through recruitment of tumor inhibitory monocytic cells and the negative regulation of tumor angiogenesis.

VEGF and the recruitment of tumor inhibitory myeloid cells

Over the recent past, the contribution of various bone marrow–derived cells, primarily from the myeloid lineage, to tumor angiogenesis has been the subject of intense investigation. In tumors, infiltrating monocytes differentiate into tumor-associated macrophages (TAM) that are persuaded by the tumor microenvironment to secrete proangiogenic and prometastatic cytokines stimulating cell migration, invasion, and metastasis (13, 14). Additionally, these recruited monocytes are capable of undergoing trans-differentiation into endothelial-like cells, expressing the angiopoietin receptor Tie2, that directly participate in the formation of blood vessels required for tumor growth (15, 16). Mouse tumor models that are refractory to anti-VEGF therapy display a CD11b+/Gr1+ myeloid cell population within the tumor microenvironment (17). These cells, which are associated with tumor refractoriness, carry alterations in the gene expression of both proangiogenic and antiangiogenic factors, such as downregulation of the angiogenic inhibitor thrombospondin-1 (17).

Interestingly, unlike the tumor promoting ability of the above-mentioned monocytic cells or TAMs, in a murine leukemia model it was recently identified that induction of a proangiogenic phenotype, achieved through in vivo Epo administration, results in the accumulation of a different monocytic cell population (Sca1+/cKit/Mac1+) capable of leukemia inhibition in a nitric oxide (NO)–dependent manner (18). These tumor inhibitory monocytic cells differentiate into macrophages and dendritic cells capable of NO production (18). The production of NO has been implicated in the destruction of tumor cells by macrophages (19) as well as induction and accumulation of hypoxia-inducible factor 1α (HIF-1α; refs. 20, 21). These findings suggest that recruitment of different subsets of myeloid-lineage cells into the tumor microenvironment leads to the acceleration or inhibition of tumor growth.

It is clear that tumor-infiltrating monocytes can have separate and contradictory functions, although the mechanisms regulating the recruitment and activation of these cells in the tumor microenvironment remain to be fully understood. Indeed, several proangiogenic cytokines such as granulocyte-colony stimulating factor and monocyte chemoattractant protein-1 have been implicated in the mobilization of TAMs and myeloid cells to the tumor microenvironment (14, 17). However, additional studies are necessary to provide evidence of direct correlation between the activity of these chemoattractants and the recruitment of tumor-infiltrating myeloid cells. Conceivably, the recruitment of monocytic cells with opposing functions is dependent on the concentration of oxygen within the tumor microenvironment. This notion can be supported by studies reporting that TAMs, which exert proangiogenic and prometastatic phenotypes, are attracted to and accumulate under hypoxic conditions (14). The state of hypoxia can be induced by the modulation of VEGF, such as anti-VEGF therapy. Conversely, normoxic conditions, maintained by a proangiogenic phenotype, may explain the accumulation of tumor inhibitory myeloid cells (8, 18).

As mentioned above, we provided evidence to suggest that the tumor inhibitory role of this myeloid population is attributed, at least in part, to its ability to produce NO (18). In turn, this induces the activation of HIF-1α (20) and prevents the degradation of these tumor inhibitory myeloid cells (21). Recent studies have discovered that HIF-1α expression promotes myeloid cell survival and inflammatory and immune functions (22, 23), potentially enhancing the antitumoral phenotype. Moreover, HIF-1α has been implicated in the regulation of growth factor–induced VEGF expression as well as Epo expression (24). The accumulation of HIF-1α thereby sustains the proangiogenic conditions in the tumor microenvironment by promoting the expression of VEGF or other mediators of angiogenesis, which in turn maintains the normoxic conditions, therefore leading to further recruitment of tumor inhibitory myeloid cells. The interpretation of these findings has led us to propose this intriguing positive feedback loop model, which may explain the inhibitory role of tumor-associated myeloid cells recruited to the tumor microenvironment under normoxic conditions maintained by increased levels of VEGF or a proangiogenic phenotype (Fig. 1).

Figure 1.

Positive and negative regulation of tumor growth. In an antiangiogenic phenotype, provided by anti-VEGF therapies, monocytes, capable of promoting tumor progression, are recruited to the hypoxic tumor microenvironment. This myeloid population responds to hypoxia by upregulating prometastatic and proangiogenic factors, thereby enhancing tumor progression. Conversely, in a proangiogenic phenotype, maintained by increased levels of angiogenic factors, tumor inhibitory cells are recruited to the normoxic tumor microenvironment. The tumor inhibitory role of these cells is partly attributed to their ability to produce NO. NO production leads to the accumulation of HIF-1α, which enhances immune cell function, and sustains the proangiogenic phenotype by promoting the expression of mediators of angiogenesis, such as VEGF. This, in turn, maintains the normoxic conditions and leads to further recruitment of immunoregulatory cells.

Figure 1.

Positive and negative regulation of tumor growth. In an antiangiogenic phenotype, provided by anti-VEGF therapies, monocytes, capable of promoting tumor progression, are recruited to the hypoxic tumor microenvironment. This myeloid population responds to hypoxia by upregulating prometastatic and proangiogenic factors, thereby enhancing tumor progression. Conversely, in a proangiogenic phenotype, maintained by increased levels of angiogenic factors, tumor inhibitory cells are recruited to the normoxic tumor microenvironment. The tumor inhibitory role of these cells is partly attributed to their ability to produce NO. NO production leads to the accumulation of HIF-1α, which enhances immune cell function, and sustains the proangiogenic phenotype by promoting the expression of mediators of angiogenesis, such as VEGF. This, in turn, maintains the normoxic conditions and leads to further recruitment of immunoregulatory cells.

Close modal

VEGF and the regulation of tumor angiogenesis

The key regulatory role of VEGF in angiogenesis that leads to suppression of early tumor growth may not arise from alterations within the hematopoietic compartment alone. Angiogenesis is dependent on the formation of vascular sprouts, which requires the proliferation and invasion of endothelial cells and subsequent vessel stabilization that is provided by vascular smooth muscle cells and pericyte coverage of the newly formed vessels (6). Interestingly, Stockmann et al. (7) reported that the cell lineage–specific targeted deletion of VEGF in myeloid cells accelerates tumor progression in multiple subcutaneous isograft models of lung cancer and an autochthonous transgenic model of breast cancer. In the absence of myeloid cell-derived VEGF, an atypical high-density vessel network is formed. Quantitative analysis showed that these mice display decreased intratumoral blood vessel length, tortuosity, and permeability, as well as increased levels of pericyte coverage. The lack of this malignancy-associated increased vascularization facilitates more rapid tumor development. This study suggests that myeloid-derived VEGF plays a unique role in the management of the tumor “angiogenic switch,” facilitating changes in tumor vessel function and normalization, thereby indicating the ability of VEGF to act as a negative regulator of tumor progression.

In an accompanying study published simultaneously with that of Stockmann et al., Greenberg and colleagues (6) presented intriguing evidence also suggesting that VEGF acts as a negative regulator of angiogenesis by disrupting the function of vascular smooth muscle cells and pericytes. Vessel stabilization and maturation is dependent on the migration of vascular smooth muscle cells and pericytes to vessel walls, which is mediated by binding of PDGF to its cognate receptor, PDGF-Rβ (6). The capacity of VEGF to act as a negative regulator of angiogenesis can be attributed to the formation of a previously undescribed VEGF-R2/PDGF-Rβ receptor complex resulting in the suppression of PDGF-Rβ signaling. On binding to its receptor, VEGF-R2, present on the surface of vascular smooth muscle cells and pericytes, VEGF blocks PDGF-mediated phosphorylation of PDGF-Rβ and in turn ablates vessel stabilization (6). The deletion of tumor cell–derived VEGF blocks the induction of this receptor complex and thereby increases pericyte coverage and tumor vessel maturation (6). Thus, although VEGF is known to promote sprouting angiogenesis by endothelial cell activation, it can also modulate vessel integrity and/or vascular tone by interfering with accessory cell function.

VEGF may act as a negative regulator by initiating changes in the microenvironment or tumor niche. Such changes include alterations in the host innate inflammatory response (7, 8) and even in the function of mesenchymal cells, such as pericytes (6), which are crucial for orchestrating tumor growth and angiogenesis. Accordingly, Ebos and colleagues (25) have recently shown that blockade of the VEGF pathway, before tumor induction, leads to a more aggressive metastasis and shortened survival, possibly by the resulting establishment of a “tumor-promoting niche.” Also in support of this notion, Xue et al. (26) showed that the function of multiple organs is systemically impaired by tumor cell–produced VEGF. This effect can be counteracted by treatment with VEGF antagonists, leading to a survival advantage. Thus, levels of VEGF in both tumor and nontumor tissues play important and differential roles in determining the effects of VEGF on tumorigenicity. However, additional analyses of the consequences of VEGF deletion or activation after tumor initiation, using conditional alleles, are necessary to provide critical insight into the effect of this angiogenic factor on tumor progression and metastasis.

Overall, the evidence presented above suggests that VEGF plays an important role in governing the mechanisms that control tumor growth and confer a survival advantage to the host. Consequently, it may be an oversimplification that tumor growth is largely dependent on maximally sustained levels of VEGF and potentially other proangiogenic factors, particularly during the early stages of disease. Moreover, the ability of VEGF to act as either an enhancer or inhibitor of tumor growth is dependent on its concentration in the host microenvironment, as well as the context and timing of its modulation relative to cancer induction. This is apparent because the studies described above suggest that VEGF acts as an inhibitor of tumor growth when its levels are modulated through genetic modification before cancer induction. The study by Stockmann et al. (7) has established a critical role for VEGF derived from tumor-infiltrating immune cells in the formation of malignancy-associated high-density vasculature, signaling to the tumor endothelium, and the sensitization and efficacy of chemotherapeutic agents. The study by Greenberg et al. has confirmed the ability of VEGF to modulate tumor vessel maturation by interfering with pericyte function. Recently, preclinical studies have shown that mice treated with VEGF-targeted therapies, before tumor induction, display increased tumor invasion and metastasis and decreased overall survival (25, 27). Moreover, clinical evidence also challenges this well-accepted theory because monotherapy with VEGF-targeted agents has resulted in only modest objective responses with short-term survival benefits (4, 5), whereas the combination of antiangiogenic agents with chemotherapy improves overall survival rates (28, 29). As such, the combination of anti-VEGF therapy with antimetastatic therapy would likely suppress the role of VEGF in tumor growth to provide a more effective treatment option and ameliorate clinical outcomes.

These above findings also implicate that the hematopoietic compartment is a pivotal player in the control of tumor growth through the regulation of VEGF. A better understanding of the proangiogenic phenotype of the host, particularly with respect to the myeloid compartment and accessory cells of the vasculature, as presented herein, may likely lead to improved cancer therapies and/or preventative interventions in the future. Therefore, further experimental and clinical studies are required to clarify the controversy surrounding the dichotomous roles of VEGF in tumor angiogenesis.

No potential conflicts of interest were disclosed.

Grant Support: The Ben-David lab is supported by Canadian Institutes of Health Research grant MOP84460.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1
Schoenleber
SJ
,
Kurtz
DM
,
Talwalkar
JA
,
Roberts
LR
,
Gores
GJ
. 
Prognostic role of vascular endothelial growth factor in hepatocellular carcinoma: systematic review and meta-analysis
.
Br J Cancer
2009
;
100
:
1385
92
.
2
Podar
K
,
Anderson
KC
. 
The pathophysiologic role of VEGF in hematologic malignancies: therapeutic implications
.
Blood
2005
;
105
:
1383
95
.
3
Kerbel
RS
. 
Tumor angiogenesis
.
N Engl J Med
2008
;
358
:
2039
49
.
4
Mayer
EL
,
Lin
NU
,
Burstein
HJ
. 
Novel approaches to advanced breast cancer: bevacizumab and lapatinib
.
J Natl Compr Canc Netw
2007
;
5
:
314
23
.
5
Hurwitz
HI
,
Honeycutt
W
,
Haley
S
,
Favaro
J
. 
Long-term treatment with bevacizumab for patients with metastatic colorectal cancer: case report
.
Clin Colorectal Cancer
2006
;
6
:
66
9
.
6
Greenberg
JI
,
Shields
DJ
,
Barillas
SG
, et al
. 
A role for VEGF as a negative regulator of pericyte function and vessel maturation
.
Nature
2008
;
456
:
809
13
.
7
Stockmann
C
,
Doedens
A
,
Weidemann
A
, et al
. 
Deletion of vascular endothelial growth factor in myeloid cells accelerates tumorigenesis
.
Nature
2008
;
456
:
814
8
.
8
Cervi
D
,
Shaked
Y
,
Haeri
M
, et al
. 
Enhanced natural-killer cell and erythropoietic activities in VEGF-A-overexpressing mice delay F-MuLV-induced erythroleukemia
.
Blood
2007
;
109
:
2139
46
.
9
Chen
J
,
Connor
KM
,
Aderman
CM
,
Willett
KL
,
Aspegren
OP
,
Smith
LE
. 
Suppression of retinal neovascularization by erythropoietin siRNA in a mouse model of proliferative retinopathy
.
Invest Ophthalmol Vis Sci
2009
;
50
:
1329
35
.
10
Hardee
ME
,
Arcasoy
MO
,
Blackwell
KL
,
Kirkpatrick
JP
,
Dewhirst
MW
. 
Erythropoietin biology in cancer
.
Clin Cancer Res
2006
;
12
:
332
9
.
11
Ribatti
D
,
Presta
M
,
Vacca
A
, et al
. 
Human erythropoietin induces a pro-angiogenic phenotype in cultured endothelial cells and stimulates neovascularization in vivo
.
Blood
1999
;
93
:
2627
36
.
12
Sung
H-K
,
Park
J-G
,
Iacovos
M
, et al
. 
Dualistic effect of systemic vascular endothelial growth factor activity on tumor etiology. In preparation
.
13
Condeelis
J
,
Pollard
JW
. 
Macrophages: obligate partners for tumor cell migration, invasion, and metastasis
.
Cell
2006
;
124
:
263
6
.
14
Murdoch
C
,
Lewis
CE
. 
Macrophage migration and gene expression in response to tumor hypoxia
.
Int J Cancer
2005
;
117
:
701
8
.
15
Schmid
MC
,
Varner
JA
. 
Myeloid cell trafficking and tumor angiogenesis
.
Cancer Lett
2007
;
250
:
1
8
.
16
De Palma
M
,
Venneri
MA
,
Galli
R
, et al
. 
Tie2 identifies a hematopoietic lineage of proangiogenic monocytes required for tumor vessel formation and a mesenchymal population of pericyte progenitors
.
Cancer Cell
2005
;
8
:
211
26
.
17
Shojaei
F
,
Wu
X
,
Malik
AK
, et al
. 
Tumor refractoriness to anti-VEGF treatment is mediated by CD11b+Gr1+ myeloid cells
.
Nat Biotechnol
2007
:
911
20
.
18
Usenko
T
,
Li
Y-J
,
Haeri
M
, et al
. 
Enrichment of Sca1+ hematopoietic progenitors in polycythemic mice inhibits leukemogenesis
.
Blood
2009
;
114
:
1831
41
.
19
Ignarro
LJ
,
Cirino
G
,
Casini
A
,
Napoli
C
. 
Nitric oxide as a signaling molecule in the vascular system: an overview
.
J Cardiovasc Pharmacol
1999
;
34
:
879
86
.
20
Kasuno
K
,
Takabuchi
S
,
Fukuda
K
, et al
. 
Nitric oxide induces hypoxia-inducible factor 1 activation that is dependent on MAPK and phosphatidylinositol 3-kinase signaling
.
J Biol Chem
2004
;
279
:
2550
8
.
21
Metzen
E
,
Zhou
J
,
Jelkmann
W
,
Fandrey
J
,
Brune
B
. 
Nitric oxide impairs normoxic degradation of HIF-1α by inhibition of prolyl hydroxylases
.
Mol Biol Cell
2003
;
14
:
3470
81
.
22
Walmsley
SR
,
Chilvers
ER
,
Whyte
MK
. 
Hypoxia, hypoxia inducible factor and myeloid cell function
.
Arthritis Res Ther
2009
;
11
:
219
.
23
Zinkernagel
AS
,
Johnson
RS
,
Nizet
V
. 
Hypoxia inducible factor (HIF) function in innate immunity and infection
.
J Mol Med
2007
;
85
:
1339
46
.
24
Semenza
GL
. 
HIF-1, O2 and the 3PHDs. How animal cells signal hypoxia to the nucleus
.
Cell
2001
;
107
:
1
3
.
25
Ebos
JM
,
Lee
CR
,
Cruz-Munoz
W
,
Bjarnason
GA
,
Christensen
JG
,
Kerbel
RS
. 
Accelerated metastasis after short-term treatment with a potent inhibitor of tumor angiogenesis
.
Cancer Cell
2009
;
15
:
232
9
.
26
Xue
Y
,
Religa
P
,
Cao
R
, et al
. 
Anti-VEGF agents confer survival advantages to tumor-bearing mice by improving cancer-associated systemic syndrome
.
Proc Natl Acad Sci U S A
2008
;
105
:
18513
8
.
27
Pàez-Ribes
M
,
Allen
E
,
Hudock
J
, et al
. 
Antiangiogenic therapy elicits malignant progression of tumors to increased local invasion and distant metastasis
.
Cancer Cell
2009
;
15
:
220
31
.
28
Miller
K
,
Wang
M
,
Gralow
J
, et al
. 
Paclitaxel plus bevacizumab versus paclitaxel alone for metastatic breast cancer
.
N Engl J Med
2007
;
357
:
2666
76
.
29
Hurwitz
H
,
Fehrenbacher
L
,
Novotny
W
, et al
. 
Bevacizumab plus irinotecan, fluorouracil, and leucovorin for metastatic colorectal cancer
.
N Engl J Med
2004
;
350
:
2335
42
.