Background
Continuing investigations into the biology of tumor-stromal interactions have identified a number of pathways and events critical to the development and maintenance of tumors and their metastatic spread. Tumor neovascularization is a critical, robust process dependent on the interplay between numerous soluble cytokines, growth factors and their receptors. Targeted therapy focusing on the tumor neovascularization process appears to be a promising approach in this regard [
1]. The VEGF-A family of cytokines and their cognate receptors have been identified as key mediators of angiogenesis and endothelial cell proliferation, migration and survival [
2‐
6], and play a central role in the organization of solid tumor vasculature [
7,
8].
The smallest of the VEGF isoforms, VEGF
121 binds to two receptors designated VEGFR-1 (Flt-1/FLT-1) and VEGFR-2 (Flk-1/KDR), both of which are over-expressed on the endothelium of tumor vasculature but virtually undetectable in the vascular endothelium of adjacent normal tissues. We have previously characterized a novel fusion construct of VEGF
121 and the plant toxin Gelonin (rGel). Gelonin is a 28.5 kDa single-chain protein belonging to the family of Type 1 plant Ribosome-Inactivating Proteins (RIPs) that can hydrolyze the glycosidic bond of a highly conserved adenosine residue in the largest RNA in the 28S ribosome, resulting in irreversible inhibition of protein synthesis.
In vivo, VEGF
121/rGel targets and destroys tumor neovasculature in solid tumors [
9,
10], reduces breast cancer metastatic spread and dramatically reduces neovascularization of pulmonary breast metastases [
11], prevents tumor growth in bone in osteolytic and osteoblastic bone metastasis models [
12,
13], and blocks retinal and choroidal neovascularization in studies of experimental ocular neovascular disease [
14]. The binding of VEGF
121/rGel to both VEGFR-1 and VEGFR-2 has been demonstrated
in vivo using non-invasive bioluminescence imaging (BLI), magnetic resonance imaging (MRI) and positron-emission tomography (PET) [
15]. Thus, VEGF
121/rGel appears to be a promising candidate for targeting its cognate receptors in various disease states.
Interestingly, VEGF
121/rGel demonstrates targeted toxicity
in vitro to endothelial cells which over-express VEGFR-2 (IC
50 = 0.5 - 1 nM) but not to cells which over-express VEGFR-1 (IC
50 = 300 nM) compared to gelonin alone (IC
50 = 300 nM) [
10]. This is surprising since VEGF
121 binds to both receptors with affinity in the picomolar range [
16]. There are several possibilities that may account for this difference in toxicity: (a) the binding affinity of VEGF
121/rGel to VEGFR-1 may be reduced, (b) binding affinity is not affected but the rate of internalization of VEGF
121/rGel bound to VEGFR-1 is reduced compared to VEGFR-2 and (c) different access to the ribosomal machinery following cell entry due to being trapped in the endosomal compartment. In addition, while the molecular effects of VEGF
121-treatment of endothelial cells have been studied [
17], the effects of VEGF
121/rGel on endothelial cells have yet to be elucidated. This information is critical in the context of
in vivo targeting because of the potential role that stimulation by VEGF
121 can have on cell survival and rGel-mediated toxicity. For example, VEGF
121 may activate particular signal transduction pathways early in the process that can result in increased toxicity of the rGel component even prior to complete inhibition of protein synthesis. The biochemical process of drug action, and its off-target effects can best be studied under controlled conditions
in vitro. In this report, we focus on understanding the mechanism of action of VEGF
121/rGel on endothelial cells by determining its binding profile to VEGFR-1 and VEGFR-2, identifying its effects on angiogenesis models
in vitro and
ex vivo, and exploring its intracellular effects on a number of molecular pathways using microarray analysis.
Discussion
Tumor neovascularization is highly dependent upon numerous cytokines and signaling events critical for the growth and organization of the vascular tree. A number of agents targeting tumor neovascularization and which interfere with one or several steps in this robust process have demonstrated significant clinical efficacy and have received FDA approval [
24]. These include agents which block angiogenesis signaling events by inhibiting various growth factor receptor kinases [
25]; interfere with VEGF physical interaction with its receptors such as anti-VEGF antibodies (bevacizumab and ranibizumab) and anti-receptor antibodies (IMC-1121B and DC101) [
26,
27]; and strategies that trap growth factor ligands (VEGF-Trap) [
28]. These have all shown antitumor efficacy alone and in combination with conventional antitumor modalities [
29,
30].
VEGF-A has been shown to play an important role in tube formation of endothelial cells
in vitro [
31] and in angiogenesis [
32]. In the present study, the effect of VEGF
121/rGel on tube formation of endothelial cells on Matrigel-coated plates was striking in that cells overexpressing VEGFR-2, but not cells overexpressing VEGFR-1, were affected. This result is consistent with our findings that VEGF
121/rGel is cytotoxic only to VEGFR-2-expressing endothelial cells [
10] and is internalized only into endothelial cells that express VEGFR-2 but not VEGFR-1 (this study). The inhibition by VEGF
121/rGel of tube formation
in vitro translates well to inhibition of both vascular endothelial growth and neovasculature
in vivo in the CAM membrane assays. The CAM assay also demonstrated that treatment with VEGF
121/rGel did not affect mature vessels. This critical finding supports our hypothesis that VEGF
121/rGel does not affect mature vessels in either normal tissues or tumors since both VEGFR-1 and VEGFR-2 are over-expressed on the endothelium of tumor neovasculature [
33‐
36] but are almost undetectable in the vascular endothelium of adjacent normal tissues and in mature tumor vessels. Therefore, small, newly vascularizing tumors and metastases may be the lesions most responsive to therapy with this agent.
The lack of internalization of VEGF
121/rGel into PAE/VEGFR-1 cells explains the difference in cytotoxicity compared to PAE/VEGFR-2. This also supports the hypothesis that VEGFR-1 is a decoy receptor, at least on endothelial cells, as it demonstrates weak tyrosine phosphorylation upon VEGF stimulation [
34]. However, we have demonstrated that mouse monocytes internalize VEGF
121/rGel via VEGFR-1 [
12], suggesting that other factors may influence VEGFR-1 receptor activity such as cell type, total receptor number and dimerization partner.
While the mechanism of rGel itself is to target the ribosomal machinery, the extent to which translation is inhibited will affect downstream cellular responses, such as other mechanisms of cell death. Information about these mechanisms may reveal additional pathways that can be targeted in combination with the fusion toxin to achieve optimal efficacy. Our study demonstrates that the cytotoxic effect of VEGF
121/rGel on VEGFR-2-overexpressing endothelial cells is not due to programmed cell death (apoptosis). Previous studies of a gelonin-based immunotoxin targeting tumor cells showed that intoxicated cells did not appear to display apoptotic characteristics [
37]. In contrast, gelonin coupled to BlyS induced apoptosis in B cells [
38] strongly supporting the idea that cell type differences can affect the mechanism of cytotoxicity.
A critical finding of this study is the identification of several genes that are regulated in response to treatment with the VEGF
121/rGel fusion construct. We observed an increase in the RNA levels of several genes that are involved in inflammation, chemotaxis, intermediary metabolism, and apoptotic pathways (Table
1). To our knowledge, this microarray analysis is the first to be performed on cells treated with a gelonin-based therapeutic. A previous report showed that only two of these genes, MKP-1 and CXCR4, were also upregulated in HUVECs after treatment with VEGF
165 for 24 h [
17]. The present study shows that VEGF
121/rGel is a member of the class of molecules that can prevent E-selectin-mediated metastasis because protein levels barely doubled in both PAE/VEGFR-2 and HUVECs after treatment with VEGF
121/rGel. We observed a similar pattern of induction of RNA but not protein levels with other genes as well. Several genes involved in the control of the apoptotic pathway were modulated in response to the fusion toxin even though the overall cytotoxic effect on target cells did not include an observable impact on the apoptotic pathway. Taken together, we conclude VEGF
121/rGel induces an increase in mRNA levels of genes that are important in cell adhesion, migration, and inflammatory response but generally does not induce a concomitant increase in protein expression. Since the rGel component of the fusion construct operates by inhibiting protein synthesis, VEGF
121/rGel could inhibit synthesis of critical proteins that are important for suppression of these specific genes. In our laboratory, current studies are under way in breast and prostate orthotopic and metastatic (i.e., lung and bone) tumor models to further characterize the effects of this drug
in vitro and
in vivo.
Acknowledgements
Research conducted, in part, by the Clayton Foundation for Research. Research supported by DAMD 17-02-1-0457 and NIH/NCI P30 CA16672. Work done at the Cancer Genomics Core Lab was supported by the Tobacco Settlement Funds appropriated by the Texas State Legislature, by a generous donation from the Michael and Betty Kadoorie Foundation, by a grant from the Goodwin Fund, and by Cancer Center Core Grant P30 CA016672 28 from the National Cancer Institute.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
KAM carried out the purification of VEGF121/rGel, performed the in vitro studies and drafted the manuscript. SR performed the cell surface binding of radiolabeled VEGF121/rGel, and helped draft the manuscript. CG-M performed the CAM assay, interpreted the results and helped draft the manuscript. LR performed the statistical analysis of the microarray data and helped draft the manuscript. JX performed the CAM assay and participated in data collection and interpretation. SK prepared, induced and harvested E. coli cells that expressed VEGF121/rGel. LHC participated in the design of the VEGF121/rGel construct. WNH performed the confocal microscopy and helped draft the manuscript. WZ performed the microarray analysis experiment and coordinated interpretation of the microarray data. JW provided the PAE/VEGFR-1 and PAE/VEGFR-2 cells and participated in the design of the study. PET participated in the design of the study. MGR participated in the design of the study and helped draft the manuscript. All authors read and approved the final manuscript.