Binding of VEGF-A to canine cancer cells with preferential expression of VEGFR 1

Aim: Despite encouraging results in syngeneic and xenografts cancer models with various inhibitors of vascular endothelial growth factor (VEGF) or its receptors (VEGFRs), beneficial effects have not been consistently translated to the clinic, underscoring the need to develop strategies that go beyond the inhibition of these targets. The purpose of this study was to generate data to support the hypothesis that VEGF may be used as “bait” to selectively deliver therapeutics to VEGFRexpressing cancer cells. Materials and Methods: VEGFR1 and VEGFR2 expression was characterized using real time quantitative reverse transcriptase polymerase chain reaction (RT-qPCR) in canine hemangiosarcoma (Grace-HSA, Emma-HSA), melanoma (TLM-1), and thyroid adenocarcinoma (CTAC) cell lines. TLM-1 and Grace-HSA were identified as representative cell lines that selectively expressed high levels of VEGFR1. Flow cytometry was performed to examine binding of a single VEGF molecule (biotinylated VEGFA and avidin conjugated to fluorescein isothiocyanate (FITC)) by these chemoresistant cell


Introduction
. VEGFR1 is primarily expressed in monocytes and macrophages, VEGFR2 in vascular endothelial Vascular endothelial growth factors (VEGF) are cells and their precursors, and VEGFR3 in lymphatic key regulators of angiogenesis and of lymphatic and endothelial cells, vascular endothelial cells, and nonblood vessel function in both health and disease endothelial compartments such as neuronal progenitors, condition in the adult [1].VEGF-A (the prototype) is osteoblasts, and macrophages [6].VEGFR3 is overwidely expressed by nearly all human malignant expressed during angiogenesis, but its expression in tumors [1], as well as by various canine tumors, such as cancer cells remains a matter of debate [6].VEGFR2 is mammary carcinoma [2], fibrosarcoma, osteosarcoma, the canonical endothelial cell receptor that transmits melanoma, carcinoma [3], and hemangiosarcoma growth and survival signals and is considered the main (HSA) [2,4,5].Physiologically, VEGF stimulates the transducer of VEGFA effects on endothelial cell differenformation of new blood vessels and regulates their tiation, proliferation, migration, and neovascularization structure and function.This regulatory effect is lost in during physiologic and pathologic conditions [6].VEGF-induced tumor blood vessels, which are Conversely, the function and biology of VEGFR1 is disordered, tortuous, and leaky, resulting in high tumor more elusive.The physiological ligand of VEGFR1 is interstitial pressure.The increase in vascular placental-derived growth factor (PGF).VEGFR1 has permeability that results from increased VEGF been described as a possible negative regulator of signaling leads to poor perfusion and hypoxia, which in VEGFR2 biology via binding and sequestering of turn, further stimulates VEGF production.VEGF is VEGF; it has also been reported to regulate monocyte believed to act through three receptor tyrosine kinases migration during inflammation [6].A variety of pathologic (RTK), VEGFR1 (Flt1), VEGFR2 (Flk1), and VEGFR3 conditions, including tumor growth (glioma), metastasis, and inflammatory diseases (e.g.rheumatoid arthritis) are sensitive to loss or inhibition of VEGFR1 function [7,8].Possible positive and negative roles for VEGFR1 in the transmission of growth and survival signals also nanoparticle showed staining of VEGFR-expressing have been suggested [6].cells during angiogenesis in an ischemic mouse model, Various drugs targeting VEGF-A or VEGFRs, confirming successful binding by the VEGF moiety to including antibodies, bioengineered molecules that VEGFR2 [14].While VEGFR1 and VEGFR2 expression mimic VEGFRs, and small molecule RTK inhibitors, has been confirmed in various human tumors and can prevent tumor growth in syngeneic and xenograft neoangiogenic vascular cells, little is known about mouse models [1].Nevertheless, these beneficial their expression in canine tumors.Our group previously effects have not been consistently replicated in the showed a heretofore-unrecognized pattern of VEGFR clinic.For example, despite success in certain, such as expression in canine HSA that appears to be modulated metastatic colorectal cancer [9], adjuvant bevacizumab by dogs' genetic background [4].Cultured HSA cells (Avastin; Genentech) did not provide a survival were resistant to VEGFR2 inhibition, with no change advantage to patients with metastatic breast cancer in magnitude or kinetics of cell growth.In contrast, compared to those treated with chemotherapy [10].In when VEGFR1 was inhibited, tumors that over-2011, the Food and Drug Administration (FDA) expressed VEGFR1 showed greater proliferation, revoked the approval of bevacizumab for breast cancer supporting the role of VEGFR1 as an active growth patients due to potentially life-threatening side effects inhibitor.This suggests VEGFR1-expressing cells without proof of survival or quality of life benefits.
should be targetable with VEGF-conjugates.Thus, in Other FDA-approved VEGFA/VEGFR inhibitors have this study, we preliminarily investigated the feasibility similarly yielded disappointing results [11].
of our proposed strategy that VEGF may be used as Furthermore, VEGFR inhibitors suppressed early "bait" to deliver therapeutics to VEGFR1-expressing tumor growth in transgenic models of islet cell adenocancer cells.The hypothesis of our study was that carcinoma, but this was not achieved in advanced tumors, canine cells that preferentially express VEGFR1 could presumably due to increased pericyte coverage in the be efficiently bound by VEGF-A.To the authors' maturing vessels [1].Notably, the lack of benefit of knowledge, this is the first time that binding of VEGFthese strategies in the clinic goes in parallel with their A to VEGFR1 has been tested in canine cancer cells possible detrimental consequences.In fact, cancer cells and similar studies have not been previously conducted that are not killed may be rendered hypoxic, leading to in humans.While evaluating the efficacy of this further expression of VEGFA and other pro-angiogenic possible therapeutic strategy is beyond the scope of this growth factors [1].
hypothesis-generating study, the effective binding of Noteworthy, there are striking structural similarities VEGF-A to VEGFR-1 suggests that if therapeutics shared amongst canine and human VEGF and could be selectively delivered to VEGFR-1 expressing VEGFRs.The canine and human sequences for VEGF, cancer cells, this may result in greater anticancer VEGFR1 and VEGFR2 are 95%, 90%, and 93% activity than what has been achieved with drugs identical at the amino acid level, and all retain 95-96% inhibiting VEGF receptors.Importantly, given the homology when conserved substitutions are considered.
aforementioned similarities shared amongst canine These similarities are based on published sequences and human VEGF and VEGFRs, the successful available through the National Center for Biotechapplication of this strategy in dogs may establish the nology Information, National Library of Medicine and foundation to extend this approach to human cancer Ensemble.Gene IDs (NCBI) are as follows: canine patients.VEGFA: 403802, human VEGFA: 7422; canine VEGFR1: 403727, human VEGFR1: 2321; canine

VEGFR2: 482154, human VEGFR2: 3791. Their
Cell cultures: The canine hemangiosarcoma cell lines almost identical biological and cell-binding properties Grace-HSA and Emma-HSA, the melanoma cell line suggest that the identification of canine tumors TLM-1, and the thyroid adenocarcinoma cell line expressing VEGF and VEGFR and the optimization of o CTAC were maintained as adherent cultures at 37 C in targeted anti-cancer strategies in dogs will provide 5% CO2 atmosphere as described [4,[15][16][17].important insights in the treatment of humans with similar diseases.Because drugs that inhibit VEGF/ Real time quantitative reverse transcriptase VEGFR pathways might not improve patient outcomes polymerase chain reaction (RT-qPCR): Elimination of beyond what is achievable by standard of care therapy, genomic DNA and reverse transcription were both novel approaches that go beyond the inhibition of these carried out using QuantiTect Reverse Transcription Kit targets are necessary.
(Qiagen, Valencia, CA).RT-qPCR was done as A promising strategy is to use VEGF as a binding described using an Eppendorf Mastercycler ep realplex molecule ("bait") to selectively deliver therapeutic with FastStart SYBR Green Master Mix Protocol agents, such as de-immunized toxins, radionucleides, (Roche, Indianapolis, IN) [18].Primer sequences are provided in Table-1.GAPDH was used as the reference photodynamic therapy compounds, and chemotherastandard for normalization and relative levels of steady peutics, to VEGFR-expressing cancer cells and neostate mRNA were established using the comparative angiogenic endothelium [12,13].Work by Hamada and [delta][delta]Ct method (Livak) 2001 [19].colleagues using a VEGF-conjugated fluorescence and one thyroid adenocarcinoma cell line (CTAC).Expression of VEGF was remarkably consistent, and Conditions were performed in triplicate using 5,000 VEGFR expression was confirmed in all cell lines cells per well in 100 µl of culture medium.Cytotoxic tested.TLM-1 and Grace-HSA cells were identified as responses were assessed using a non-radioactive, representative cell lines that retained preferential colorimetric cell proliferation (MTD) assay.Viability expression of VEGFR1 with significantly reduced was assessed after 72 hr of culture using CellTiter96 expression of VEGFR2.These cell lines were selected AQueous kit (Promega, Madison, WI) as described by to verify their chemoresistance and to test VEGF the manufacturer.
binding in subsequent experiments as described below.Flow cytometry: Cells were harvested using Accutase TLM-1 and Grace-HSA cell lines are resistant to and resuspended in staining buffer (PBS with 2% paclitaxel (Taxol): Our group has previously shown FBS).VEGF binding was measured using the human that TLM-1 cells are highly resistant to various VEGF biotinylated fluorokine kit (R&D Systems, conventional chemotherapy agents.In fact, the half-Minneapolis, MN) as described by the manufacturer.maximal inhibitory concentrations (IC ) for TLM-1 50 Cells were analyzed using a FACSCalibur cytometer growth inhibition with doxorubicin, camptothecin, and (BD Biosciences, San Jose, CA) and data were 5-fluorouracil were all >1 µM in standard 72 hr analyzed with FlowJo (v9.5.2) software (Treestar, cytotoxicity assays (R. Weiss and J. Modiano, Ashland, OR).Viable cells were determined by unpublished data).Similarly, HSA cells are highly exclusion of 7-Aminoactinomycin D (7AAD).

TLM-1 and Grace-HSA cells show preferential expression
the successful treatment of many solid tumors.It is of VEGFR1: Expression of VEGF, VEGFR1, and unclear if it is due to the use of conservative chemo-VEGFR2 has been confirmed in a variety of human therapy regimen, or to properties that are intrinsic to the tumors and neoangiogenic vascular cells, but little is specific cancer cells.To further examine the sensitivity known about their expression in canine tumors.A of TLM-1 and Grace-HSA cells to chemotherapy, we pattern of VEGFR expression modulated by dogs' determined in vitro cytotoxicity of paclitaxel, a potent genetic background was previously shown in canine anti-mitotic, against these two cell lines.As shown in HSA cells [4].Here, we used RT-qPCR to characterize

VEGF-A binds to cells expressing VEGFR1:
In contrast the clinical experience that HSA and melanoma cells to the limited efficacy of several chemotherapy are strongly chemoresistant.Despite low levels of regimens against many solid malignancies, HSA cells VEGFR2 expression, these cells could still be targeted are sensitive to cytotoxic compounds such as by VEGF, suggesting that tumors that preferentially genetically engineered bacterial toxins [20].If these express VEGFR1 can be viably targeted using VEGFcompounds could be conjugated to ligands such as conjugates.This effect could be additive with targeting VEGF, this may allow selective targeting of VEGFR-of stromal endothelial cells that express high levels of expressing cancer cells.Thus, to further investigate the VEGFR2.Interestingly, since VEGF-targeted compounds feasibility of targeting such compounds to VEGFR1-are not substrates for chemoresistance mechanisms, expressing tumors, we used flow cytometry to examine they may facilitate penetration of chemotherapeutics VEGF binding in TLM-1 and Grace-HSA cells.
into tumor and neoangiogenic cells, which may result Figure-3 illustrates that both TLM-1 and Grace-HSA in enhanced anti-tumor activity by means of selective cell lines showed specific binding to VEGF-A.Binding targeting rather than VEGFR inhibition.It has to be was inhibited by competition with the anti-VEGF acknowledged that normal cells expressing VEGFRs antibody, but intriguingly, there was significantly outside the tumor environment might be killed along greater receptor density (~1-log) in Grace-HSA cells with VEGFR-expressing cancer cells, leading to than in TLM-1 cells as determined both by the mean adverse effects of the treatment, including vascular fluorescence intensity at the same concentration of leak syndrome (VLS), hemolytic uremic syndrome, ligand, and by the blunted inhibition at the same liver, renal, or cardiac damage.Furthermore, binding concentration of antibody.This suggests that despite of targeted drugs might enhance the overall function of having lower steady state levels of VEGFR mRNA (in VEGFRs.Conversely, it cannot be completely relation to GAPDH) than melanoma cells, HSA cells excluded that binding with VEGF-A might neutralize might use post-translational mechanisms to stabilize the effect of those cytotoxic drugs.However, the latter expression of VEGFRs on the cell surface.seems unlikely given that previous studies using Indeed, the binding of biotinylated VEGF to VEGFR-targeting compounds as well as bioengineered Grace-HSA cells was significantly reduced when ligand-targeted agents showed potent antitumor unlabeled VEGF was used as a competitor at 100-fold activity both in vitro and in vivo in experimental mouse molar excess (500 ng/ml) but not at 10-fold excess models [12,13] with no loss of anti-proliferative molar excess.We believe this is due to the relatively potential due to binding by a specific targeting molecule.slow rate of ligand binding at 4C, so that even after 15 These studies also suggested that the structural changes min pre-incubation, unlabeled VEGF used at 50 ng/ml occurring in the ligands following conjugation with a did not occupy all available receptor molecules to binding molecule did not interfere with the ability of saturation.Endogenous VEGF produced by the cells the ligands to bind their respective receptors.For did not affect the assay, indicating the tumor cells were example, the bispecific ligand targeted toxin EGFuPA capable of binding exogenous VEGF and were thus made up of the epidermal growth factor (EGF) and the amenable to targeting.

Discussion
previously shown to have potent anticancer activity via Our cytotoxicity experiments showed that both inhibition of protein synthesis [21].Pre-clinical work TLM-1 cells and Grace-HSA cells were resistant to in vitro and in mice, respectively, showed highly paclitaxel at micromolar concentrations, corroborating efficient targeting and killing of canine hemangio- sarcoma (HSA) cells and HSA tumor initiating cells at for providing the funding that allowed the conduction clinically achievable concentrations of EGFuPA, and of our experiments, and Dr. Brenda Spangler for that EGFuPA targets chemotherapy-resistant tumor helpful discussions.cells and the associated tumor microvasculature [20][21][22].
Cytotoxicity assays: Cells were incubated in 96 well VEGF receptors by two HSA cell lines (Emma-HSA plates overnight prior to addition of paclitaxel at the and Grace-HSA), one melanoma cell line (TLM-1), concentrations indicated in the results section.
Figure-2, only minimal cytotoxicity was seen for both TLM1 and Grace-HSA cells following exposure to expression of VEGF, VEGFR1 and VEGFR2 genes in increasing concentrations of paclitaxel (IC >10 µM).four different canine cancer cells lines to identify 50 tumors that preferentially expressed VEGFR1.This corroborates the clinical experience that HSA and Figure-1 illustrates expression of VEGF and melanoma cells are strongly chemoresistant.

Figure- 1 .
Figure-1.VEGFR expression was present in all canine cancer cell lines tested.RT-qPCR was performed using an Eppendorf Mastercycler ep realplex with FastStart SYBR Green Master Mix Protocol.Expression of VEGF, VEGFR1, and VEGFR2 was evaluated in Emma-HSA, Grace-HSA, TLM-1, and CTAC cell lines.Relative expression for each of these genes is shown normalized to GAPDH.

Figure- 3 .
Figure-3.VEGF-A binds to both VEGFR1 and VEGFR2.One dimensional histograms illustrating fluorescence (x-axis) relative to cell number (y-axis) for both cell lines.The black peak represents the negative control.The grey peak represents human recombinant VEGF conjugated to fluorescein isothiocyanate (FITC).The white peak represents VEGF with blocking from anti-VEGF antibodies, to illustrate specificity.VEGF binding was measured using the human VEGF biotinylated fluorokine kit (R&D Systems, Minneapolis, MN) and cells were analyzed using a FACSCalibur cytometer (BD Biosciences, San Jose, CA).

Table 1 .
Primer Sequences for RT-qPCR