VEGF121b and VEGF165b are weakly angiogenic isoforms of VEGF-A
© Catena et al; licensee BioMed Central Ltd. 2010
Received: 18 May 2010
Accepted: 31 December 2010
Published: 31 December 2010
Different isoforms of VEGF-A (mainly VEGF121, VEGF165 and VEGF189) have been shown to display particular angiogenic properties in the generation of a functional tumor vasculature. Recently, a novel class of VEGF-A isoforms, designated as VEGFxxxb, generated through alternative splicing, have been described. Previous studies have suggested that these isoforms may inhibit angiogenesis. In the present work we have produced recombinant VEGF121/165b proteins in the yeast Pichia pastoris and constructed vectors to overexpress these isoforms and assess their angiogenic potential.
Recombinant VEGF121/165b proteins generated either in yeasts or mammalian cells activated VEGFR2 and its downstream effector ERK1/2, although to a lesser extent than VEGF165. Furthermore, treatment of endothelial cells with VEGF121/165b increased cell proliferation compared to untreated cells, although such stimulation was lower than that induced by VEGF165. Moreover, in vivo angiogenesis assays confirmed angiogenesis stimulation by VEGF121/165b isoforms. A549 and PC-3 cells overexpressing VEGF121b or VEGF165b (or carrying the PCDNA3.1 empty vector, as control) and xenotransplanted into nude mice showed increased tumor volume and angiogenesis compared to controls. To assess whether the VEGFxxxb isoforms are differentially expressed in tumors compared to healthy tissues, immunohistochemical analysis was conducted on a breast cancer tissue microarray. A significant increase (p < 0.05) in both VEGFxxxb and total VEGF-A protein expression in infiltrating ductal carcinomas compared to normal breasts was observed. A positive significant correlation (r = 0.404, p = 0.033) between VEGFxxxb and total VEGF-A was found.
Our results demonstrate that VEGF121/165b are not anti-angiogenic, but weakly angiogenic isoforms of VEGF-A. In addition, VEGFxxxb isoforms are up-regulated in breast cancer in comparison with non malignant breast tissues. These results are to be taken into account when considering a possible use of VEGF121/165b-based therapies in patients.
Angiogenesis is a process by which new blood vessels are formed from pre-existing ones . In physiological conditions, this process is strictly controlled by a set of molecules that can either activate the process (proangiogenic factors) or inhibit it (antiangiogenic factors) . During the last decades, it has been widely established that solid tumors have abnormal hyperactivation of angiogenesis . Among the factors that can trigger angiogenesis, the lack of oxygen (hypoxia) is of special importance. Virtually all solid tumors eventually activate angiogenesis in order to overcome lack of oxygen and nutrients after reaching a certain burden [3, 4]. One of the most important mediators of hypoxia-activated angiogenesis is the Vascular Endothelial Growth Factor (VEGF-A), produced by tumor cells after sensing low oxygen levels [5, 6]. VEGF-A expression can also be induced by non-hypoxia mediated activation, such as Ras signalling .
VEGF-A is a key player in tumor-induced angiogenesis, and its overexpression has been found in most solid tumor types . VEGF-A acts through its cognate receptors VEGFR1 (Flt-1) and VEGFR2 (Flk-1/KDR), in endothelial and bone marrow-derived cells [6, 8]. The VEGF pathway has been used as a major target to block tumor angiogenesis. A set of molecules that bind and inhibit different components of the VEGF-A pathway have been developed during the past years. Some of them have already reached the clinical practice, such as bevacizumab (Avastin®, Genentech), a monoclonal antibody that binds and inactivates VEGF-A, or sunitinib (Sutent®, Pfizer), a tyrosine-kinase inhibitor that blocks phosphorylation of several tyrosine-kinase receptors including VEGFR1 and VEGFR2 [9, 10].
The VEGF-A gene contains 8 exons, which can give rise to 5 main alternatively spliced isoforms (VEGF121, VEGF145, VEGF165, VEGF189 and VEGF206) . Alternative translation codons upstream of the canonical ATG codon can be used, so that longer isoforms can also be generated . However, the relative importance of these members is still undetermined. Recently, a novel set of isoforms, the so-called "b-isoforms" or "VEGFxxxb" isoforms, have been described. These transcripts of the VEGF-A gene code for polypeptides with the same length as the classical ones, because exon 8 (present in all the formerly known isoforms) is substituted by an alternatively spliced exon of the same size (exon 8b) . These isoforms were therefore named VEGF121b, VEGF165b, VEGF189b etc. In the classically studied isoforms, exon 8 is known to be important for receptor activation . Thus, the "b-isoforms", where exon 8 is substituted by another peptide sequence, were hypothesized to act as potential antagonists of VEGF-A receptors . Several reports have indeed shown that VEGF165b may have anti-angiogenic properties [14, 15], while others cast some doubts about such activities, suggesting that it may act as a VEGF-A receptor agonist [16, 17].
Another interesting issue is the possible differential expression between "angiogenic" vs. "antiangiogenic" isoforms in pathologies where development of aberrant vasculature is involved, including cancer. Previous studies have shown in a limited number of samples, using semiquantitative RT-PCR, that VEGFxxxb isoforms are highly expressed in normal prostate, colon and kidney compared to their malignant counterparts [14, 18, 19]. It was proposed that formation of neovasculature in pathological conditions would modify alternative splicing of VEGF-A, thus promoting the expression of the "b-isoforms" (supposedly anti-angiogenic) at the expense of the classical angiogenic family of isoforms. This would also be extremely interesting because expression of the ratio VEGFxxxb/VEGF could be utilized as a biomarker of angiogenic disease .
Since a therapeutic approach using recombinant VEGFxxxb proteins is very attractive, but the biological activity of such transcripts is not yet clear, we sought to produce recombinant VEGF121b and VEGF165b proteins in the yeast Pichia pastoris, and constructed expression vectors to overexpress these isoforms, in order to further elucidate their role in cancer models. In addition, we analyzed protein expression of VEGFxxxb and total VEGF in normal mammary glands and 50 breast cancer samples, using specific antibodies previously characterized.
Cloning of VEGF121b and VEGF165b
Oligonucleotides were purchased to Sigma-GenoSys (Sigma, St. Louis, MO, USA). The primers VF (5'GAAACCATG AACTTTCTGCTGTCTT3') and V121bR (5' TTAAGCTT TCAGTCTTTCCTGGTGAGAGATTTTTCTTGTCTTGCTCTATC3') were used to clone the VEGF121b isoform by PCR into pCR2.1 vector (Invitrogen). VF and V165bR (5' TTAAGCTT TCAGTCTTTCCTGGTGAGAGATCTGCAAGTAC
GTTCGTTTAACTC 3') were used to clone VEGF165b. Note that initiation codon is underlined in VF and both reverse oligonucleotides contain Hind III restriction sites (bold). VEGF121b and VEGF165b coding sequences were then subcloned into the pCDNA3.1(-)Neo expression plasmid. The primer VPPF (5' GGTCTCGAGAAAAGAGAGGCTGAAGCTGCACCCATGGCAGAAGG 3'), together with V121bR or V165bR, was used to clone VEGF121b and VEGF165b coding sequences lacking the signal peptide (ΔPSVEGFxxxb) into the pPICZalphaC vector (Invitrogen) for production of recombinant proteins in the yeast Pichia pastoris, where the alpha-factor signal peptide is used to achieve extracellular expression of the VEGFxxxb sequences.
Recombinant protein production and purification
The pPICZalphaC plasmids carrying the ΔPS-VEGFxxxb sequences were linearized, gel-purified, and measured for concentration. 80 μL of Pichia pastoris cells were mixed with 5 μg of linearized-plasmid in 1 mm-wide electroporation cuvettes (Bio-Rad). Electroporation was carried out in a Gene-pulser II (Bio-Rad) using the preset yeast conditions. After electroporation, 1 mL of 1 M sorbitol was added to the cuvettes and the electroporated cells were transferred to sterile microtubes. Yeasts were incubated at 30°C for 2 h and then spread in YPDSZ plates (1% yeast extract, 2% peptone, 2% sorbitol, 2% agar and 100 μg/mL zeocin) and incubated for 9 days at 29°C. Zeocin-resistant colonies were picked and grown in YPD medium. Yeast clones were transferred to BMGY medium (1% yeast extract, 2% peptone, 100 mM potassium phosphate pH 6.0, 1.34% yeast nitrogen base, 0.00004% Biotin, 1 U/mL gentamycin sulphate, and 1% glycerol) to allow cells to grow exponentially for 30 h at 29°C and thorough shaking. Yeasts were then centrifuged and resuspended in BMMY medium (the same composition as BMGY but containing 1% methanol instead of glycerol) to induce protein production. Supernatants were collected 24 hours after incubation at 29°C and thorough shaking, and analyzed by SDS-PAGE to determine the best clone producing each of the VEGF121/165b isoforms. Selected clones were grown at 29°C in 2 L of BMGY for 2 days and then changed to BMMY inducing medium, in order to produce large amounts of recombinant products. For purification, nickel-affinity chromatography was used. A Hi-Trap chelating column (Amersham) was connected to an AKTÄ High Pressure Liquid Chromatography (HPLC) device (Amersham). Pichia pastoris supernatants containing recombinant VEGF121/165b proteins were diluted in binding buffer (0.02 M sodium phosphate, 1 M NaCl, pH 7.2; all reagents from Sigma) and loaded into the HPLC device. Elution buffer (0.02 M sodium phosphate, 1 M NH4Cl, 500 mM Imidazole, pH 7.2; all reagents from Sigma, except for imidazole, purchased from Merck) was loaded, and gradually mixed with binding buffer with an increasing proportion of elution buffer. Fractions of 1 mL were collected throughout the process.
Purified proteins after affinity chromatography were depleted from eluting medium and changed to PBS through dialysis, using Slide-A-lyzer cassettes (Pierce) with a 10 KDa threshold pore. Cassettes were filled with eluted protein and drawn in 3L of PBS, overnight. This step was repeated with new PBS for 6 more hours. Dialyzed proteins were extracted from the cassettes with syringes and snap frozen.
Protein extraction, electrophoresis and western blot
For protein extraction, cultured cells were lysed for 30 min at 4°C in RIPA buffer (50 mM Tris pH 7.4, 150 mM NaCl, 1 mM PMSF, 1 mM EDTA, 1% sodium deoxycholate and 0.1% sodium dodecyl sulphate; all reagents from Sigma) plus a protease inhibitor cocktail (Roche, Switzerland). Samples were then centrifuged at 13000 rpm. Protein concentration was determined by the bicinchoninic acid protein assay (Pierce, Rockford, IL). In the case of conditioned culture media, supernatants were centrifuged to get rid of any cell debris and 20-fold concentrated by centrifugation for 45 min, using 15-KDa Amicon Ultra centricons (Millipore, Billerica, MA, USA).
Proteins were electrophoresed in Bis-Tris buffered gels (Novex gels, Invitrogen) in either reducing or non-reducing conditions, following standard procedures. 20 μg protein solution (in RIPA buffer) were mixed with Laemmli sample buffer and boiled for 5 min. Electrophoresis was carried out in 1X running buffer for 90 min at 130V and room temperature. Proteins were directly stained in the gel with Coomassie blue or transferred to PVDF membranes for immunodetection. Deglycosylation analysis of VEGFxxxb proteins (90 μM) was treated with 0.8 mM Endo F1 and incubated for 1 h at 37°C; cleavage was monitored by SDS-PAGE.
Membranes for western blots were rinsed twice with PBS-tween, blocked with PBST plus 5% skim milk for 30 min at room temperature, and incubated with primary antibodies. Antibodies against VEGF (MAB293, R&D; and sc-152, Santa Cruz), VEGFxxxb (MAB3045, R&D), pKDR, total KDR, pERK1/2, total ERK1/2, and GAPDH (all of these latter ones from Cell Signalling) were used. Then, horseradish peroxidase-labelled secondary antibodies (GE Healthcare) against the corresponding primary antibodies were added. Immunoreactive bands were visualized by a chemoluminescent method using the Lumi-lightPLUS kit (Roche).
HUVECs, PC3 and A549 cell lines were obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA). PC3 and A549 cells were maintained in complete medium, consisting of: RPMI-1640 growth medium (Invitrogen) with Glutamax®, supplemented with 10% heat-inactivated FBS, 100 U/mL penicillin and 100 μg/mL streptomycin (both antibiotics from Invitrogen). HUVECs were maintained in EGM-MV2 medium (Lonza) containing human recombinant EGF, VEGF, FGF, IGF-1, hydrocortisone, ascorbic acid and 2% FBS. Cell culture medium with 1% serum was used to analyze cells supernatants by western blot.
Purified plasmidic DNA was introduced into mammalian cells through cationic lipid-based transfection with the reagent Lipofectamine 2000, according to manufacturer's recommendations. Transfected cells were selected and maintained with complete medium plus 300 μg/mL (PC3) or 500 μg/mL (A549) G418 .
Cell proliferation assays
Two different methods were used to assess cell proliferation. The first method consisted of the MTT (Sigma-Aldrich, Italy) assay. Cells were seeded in 50 μL 2% FBS-containing growth medium in 96-well culture plates and allowed to attach overnight. Two-fold concentrated recombinant human VEGF165 (rhVEGF165, Apollo Cytokine Research), recombinant human VEGF121b and VEGF165b produced in Pichia pastoris (VEGF121b(pp) and VEGF165b(pp)), recombinant human VEGF165b produced in CHO cells (VEGF165b(hs)), kindly provided by Dr. David O. Bates (Microvascular Research Laboratories, Department of Physiology, University of Bristol, UK), or the VEGFRs inhibitor GW654652 (GlaxoSmithKline) were added. In each time point, 10 μL 5% MTT solution was added to each well. Plates were incubated for additional 3 h at 37°C. The resulting formazan crystals were finally solubilized by administration of 100 μl 10% SDS in 50% N-N-Dimethylformamide to each microplate well. Absorbance at 550 nm was measured using a TECAN Sunrise microplate reader. Wells containing only complete medium were used as controls. Each experiment was performed three times using six replicates for each drug concentration.
The second method consisted of analysis of DNA synthesis by incorporation of the modified nucleotide EdU, using the Click-it reaction according to the manufacture's instructions (Invitrogen). Briefly, cells were plated at 50% confluence and treated with 50 or 100 ng/mL of rhVEGF165, VEGF121b(pp), VEGF165b(pp), or bFGF overnight. Cells were then incubated for 1 h with 5 mM EdU solution, washed, trypsinized, fixed, permeabilized, and incubated with Alexa-Fluor-647 dye in the presence of copper for catalysis of the Click-it reaction. Cells were analyzed with a FACScalibur flow cytometer to determine EdU incorporation.
Tumor xenograft and Matrigel plug assays
Nu/Nu mutant athymic mice (Balb/C genetic background) were purchased from Harlan Laboratories (Barcelona, Spain) and maintained in SPF (Specific Pathogen Free) standard conditions. One million PC3 or five million A549 cells and their corresponding transfectants in exponential growth phase were resuspended in 200 μL PBS and injected subcutaneously in the flanks of Nu/Nu mice. Tumor measurements were done with precision callipers and animals were sacrificed before tumors reached 1.7 cm in diameter. Experiments were conducted according to the guidelines for ethical use of animals of our Institution (CIMA-University of Navarra) under an approved protocol. Tumors were harvested and fixed overnight in 10% buffered formalin, embedded in paraffin, and sectioned. Primary tumor volumes were calculated with the formula: V = length × (width)2/2.
For Matrigel plug assays, 400 μL Growth Factor Reduced Matrigel (BD) were mixed with 100 ng of rhVEGF165, VEGF121b(pp), VEGF165b(pp), or bFGF (as positive control) in 100 μL PBS and injected subcutaneously in Nu/Nu mice. One week after cell inoculation, mice were injected retro-orbitally with 100 mL Fluorescein-labelled dextran (3 mg/mL) or with Alexa-647-labelled isolectin B4 (100 μg/mL). After 15 min, mice were sacrifized, and the Matrigel plugs were explanted and analyzed under a Zeiss Axiovert confocal microscope.
Immunohistochemistry, fluorescence analysis and quantification
Tissues (xenografted tumors or matrigel plugs) were obtained from the in vivo experiments, fixed in 10% buffered formalin and embedded in paraffin. Tissue Microarray (TMA) slides were obtained from AccuMax (Seoul, Korea; catalogue # A202(I)). This TMA contains 100 breast tissue cores from 50 patients and 8 tissue cores from 4 normal breast tissue obtained by mammoplasty. Breast cancer types include 33 infiltrating ductal carcinomas, 7 papillary carcinomas, 3 phyllodes tumors, 4 infiltrating lobular carcinomas, and 3 samples corresponding to other breast cancer tumor types.
For immunohistochemistry, slides were deparaffinized, hydrated, and incubated for 10 min with 3% H2O2 in water to quench the endogenous peroxidase activity. An antigen retrieval method was used for detection of the antibodies. Dilutions of primary antibodies were as follows: 1:200 for Caspase 3 (Cleaved Caspase-3 Asp 175, Cell Signaling); 1:20 for CD-31 (Dianova); 1:100 for PDGFRβ (Cell Signaling); 1:200 for VEGF (Santa Cruz); 1:50 for VEGFxxxb (R&D). Primary antibodies were incubated at 4°C overnight or for 1 h at RT in the case of CD31. Tissues were washed in TBS and incubated with the appropriate secondary antibody. Afterwards, slides were incubated for 30 min with the EnVision™ anti-rabbit detection system (Dako). Peroxidase activity was carried out with DAB (Dako). Finally, slides were counterstained with hematoxylin, dehydrated, and mounted. For quantifications of immunohistochemistry in xenografted tumor sections, 10 random images (200×) per mouse were captured with a microscope (Leica, Wetzlar, Germany) equipped with the Analysis™ software. Positive cells were quantified with Image J (NIH, Bethesda, MD, USA). Measurements are given as relative area occupied by the stained tissue with respect to the reference area (for CD-31 and PDGFRβ) or percentage of positive cells (for active caspase-3). For the TMA, a semiquantification procedure was performed according to a score that takes into account the extension and intensity of the immunoreactivity signal. The "0" score was assigned when the sample was completely negative. Score "1" meant weak signal; Score "2", strong and widespread signal; and Score "3", very strong and widespread signal. Signal was considered positive when at least 10% of the cells were positive.
For quantification of FITC-dextran and Alexa-647 Isolectin B4 in sections of Matrigel plugs, slides were analyzed with an Axiovert (Carl Zeiss, Germany) epifluorescence microscope and 10 random pictures of each Matrigel were taken. Labelled area was measured with the ImageJ software.
Statistical analysis of data
Data sets were tested for normal distribution with Shapiro-Wilks and Kolgomorov-Smirnoff tests. Levene's test was also performed to verify homogeneity of variances. In those cases where tests displayed normal distribution, ANOVA was performed to test for possible differences among groups. Bonferroni correction was used for post-hoc comparison in the case of variance homogeneity, whereas Tamhane's correction was the choice in the case of Levene's positive test. For non-normal distributed data sets, non-parametric tests were applied: Kruskal-Wallis for multiple comparisons and Mann-Whitney's U-test with significance correction for double comparisons of independent samples. Wilcoxon's test was carried out for dependent sample data. To run these tests, The SPSS software was used. Results with a p-value < 0.05 were considered significant (*), and those with a p-value < 0.01 (**), or <0.001 (***), very significant.
Generation of Pichia pastoris-derived clones with VEGF121/165b overexpression and purification of recombinant proteins
Electrophoresed culture media from Pichia pastoris clones obtained after electroporation with the linearized VEGF121b or VEGF165b sequence containing pPICZaphaC plasmids and selection with zeocin for one week is shown in Figure 2C. Bands of the expected molecular masses were detected in supernatants from Pichia pastoris clones. The amount of ectopic protein was clearly visible among the total secreted proteins by this yeast. The clones overexpressing higher amounts of recombinant VEGFxxxb proteins were chosen for large-scale production. Importantly, Pichia pastoris-derived recombinant proteins were immunoreactive to the currently available (and previously validated) commercial antibody (R&D) against VEGFxxxb (Figure 2D).
Recombinant VEGF121/165b isoforms expressed in Pichia pastoris form dimers
The band pattern of recombinant human VEGF121/165b isoforms expressed in Pichia pastoris was similar to that of the native VEGFxxx isoforms, as previously described . VEGF121b forms dimers that can be detected in the gel as three bands under non-reducing conditions (Figure 2C). The three bands probably correspond to dimers of glycosylated-glycosylated, glycosylated-non-glycosylated and non-glycosylated-non-glycosylated proteins, as described for the VEGF-A classic isoforms . The same culture supernatants run under reducing conditions showed only two bands (Figure 2C), which would be concordant with the ability of these proteins to dimerize. The same pattern can be seen for VEGF165b (Figure 2C), although the bands are not as clearly seen as for VEGF121b. Recombinant VEGFxxxb proteins also formed larger complexes, such as tetramers and octamers, especially in the case of VEGF165b (Figure 2D).
To assess the glycosylation status of the VEGFxxxb recombinant proteins produced in Picha pastoris, endoglycosidase F1 was used both in reducing and non-reducing conditions (Additional file 1 Figure S1). VEGF121/165b deglycosylation produced an electrophoretic shift for both isoforms. Molecular weights for both glycosylated and deglycosylated proteins are compatible with those observed for VEGF121 and VEGF165 [20, 21].
Recombinant VEGF121/165b proteins induce proliferation of HUVEC cells
To confirm our results on proliferation we used an alternative method, based on DNA incorporation into the cells. As shown in Figure 3B, administration of bFGF and VEGF165 increased DNA incorporation into HUVECs by 3-fold (p < 0.001), compared to untreated controls. Exposure to VEGF165b(pp) and VEGF121b(pp) also increased significantly (p < 0.01) DNA incorporation into HUVECs by almost 2-fold as compared to controls. Therefore, these experiments parallel those obtained by MTT showing that VEGFxxxb isoforms increase proliferation, although less potently than VEGF165.
VEGF121/165b induces phosphorylation of Flk-1/KDR and its downstream effector p42-44/ERK
VEGF121/165b isoforms stimulate angiogenesis in vivo
VEGF121/165b overexpression stimulates tumor growth and angiogenesis in vivo
Expression of VEGFxxxb proteins and total VEGF-A in a set of normal breast and breast cancer samples
Previous studies had suggested that VEGFxxxb isoforms can be differentially expressed in normal vs. pathological conditions . To study if this was the case for breast cancer, we used a TMA containing core biopsies from breast cancer patients and normal breasts. To this end, we used a well validated antibody that recognizes total VEGF-A (including all VEGF-A isoforms) (R&D systems) and the only available and validated anti-VEGFxxxb antibody that recognizes all the VEGFxxxb proteins (R&D systems) . No available antibody specific for the "non-b" isoforms (VEGFxxx) is currently available.
The importance of VEGF-A in normal and pathological angiogenesis has been widely explored and documented in the last decades . VEGF-A has become a major target for cancer therapy in solid tumors, and VEGF-targeted drugs, such as bevacizumab or sunitinib are currently being used in patients [9, 10]. The biology of VEGF-A spliced isoforms is, however, poorly understood yet. In particular, understanding how the different VEGF-A isoforms are generated through alternative splicing may be important to design more specific and effective molecular therapies.
In 2002, a new family of VEGF-A isoforms generated by alternative splicing was reported [14, 19], which were denominated VEGFxxxb isoforms. These transcripts incorporate a different exon (called exon 8b, sequence SLTRKD) from the classical exon 8a (sequence CDKPRR) found in the angiogenic transcripts . Inclusion of exon 8b was thought to endow VEGFxxxb isoforms with the capacity to bind VEGF-A receptors without strong downstream signalling activation . Because of this property, the VEGFxxxb proteins were hypothesized to be antiangiogenic . Although results have to be interpreted carefully, it was shown that overexpression of VEGF165b or VEGF121b in tumor cells xenotransplanted into nude mice results in growth inhibition [14, 18, 23–25]. Another interesting finding that was reported is that VEGFxxxb isoforms may be differentially expressed in pathological tissues compared to normal tissues [13, 25, 26]. Changes in the natural splicing would favor the amount of the VEGFxxx transcripts in some aberrant angiogenesis-linked diseases, at the expense of reducing the amount of the VEGFxxxb isoforms (mainly expressed in normal tissues).
Because of the great potential importance of these findings, we aimed in the present study to produce VEGF121b and VEGF165b recombinant proteins in the yeast Pichia pastoris, with the goal of testing their antiangiogenic/antitumor properties in vitro and in vivo. Generation of VEGF121/165b-overexpressing cancer cells using the PCDNA3.1 plasmid was also used for this purpose. The yeast expression system was chosen for several reasons. First, this system allows protein glycosylation, and purification with little secretion of yeasts-derived endogenous proteins that may contaminate the recombinant protein of interest. Another advantage is that protein production is easier, faster and cheaper than that generated in mammalian systems. In addition, we made sure with this expression system that no exon 8-containing VEGF-A contamination is present, since yeasts do not code for any form of VEGF. Producing VEGF-A in mammalian cells has the drawback of possible formation of exons 8 and 8b heterodimers. This may be of special importance, since VEGF-A is secreted and active in its dimeric form .
We have successfully produced recombinant VEGF121b and VEGF165b proteins that share similar structural characteristics to those of the classical VEGF121 and VEGF165 proteins: Ability to form dimers or multimers, and reactivity with commercial antibodies that were raised against exons 1 to 5 (common to all VEGF-A isoforms). In addition, both recombinant VEGF121b and VEGF165b proteins were immunoreactive with a validated antibody that recognizes exon 8b (from R&D ).
To test the functionality of these isoforms in vitro, HUVECs were treated with recombinant proteins produced in yeasts, VEGF165b produced in mammalian cells, or the "classical" VEGF165 angiogenic protein (as control). All treatments were carried out in serum-free media. We found that VEGF121/165b isoforms induced proliferation of HUVECs and phosphorylation of VEGFR2 and its downstream transducer ERK. The effect of VEGF121/165b isoforms was milder than that found for VEGF165: Proliferation of HUVECs was ~50% less stimulated with VEGF121/165b isoforms than that using recombinant VEGF165. However, the degree of ERK activation was similar for all the proteins tested, 10 min after stimulation. This intracellular effector was phosphorylated specifically by VEGFRs, as demonstrated by the inhibition of such process in the presence of the VEGFR1 and VEGFR2 tyrosine-kinase inhibitor GW654652.
Kawamura et al.  have demonstrated that VEGF165b displays unique functional characteristics. They showed VEGF165b to be a weak agonist of VEGFR2 and to induce its phosphorylation in HUVECs with a similar potency as VEGF145. Similarly to VEGF121, VEGF165b does not bind neuropilin-1 (NRP1) and failed to induce complexes between NRP1 and VEGFR2. VEGF165b promotes cell migration (in PAE cells transfected with VEGFR2), similar to VEGF165, VEGF145, and VEGF121 isoforms. Resembling the activity of VEGF121 and VEGF145, but unlike VEGF165, VEGF165b does not induce endothelial sprout. But unlike the other isoforms, VEGF165b was unable to phosphorylate the mouse VEGFR2 at the Y1052 position. Unfortunately, downstream signalling activation was not tested in this study . Authors hypothesized that VEGF165b is not able to fully rotate the receptor's intracellular tail upon binding, thus blocking to some extent transphosphorylation . In addition, Glass et al.  have shown that VEGF165b activates transiently VEGFR1 in order to elicit vascular permeability in the mesenteric vein. Collectively, these and our results suggest that VEGF165b may exert weak downstream signalling in endothelial cells. However, depending on the cellular context (for instance, the presence of NRP1 or the amount of VEGFR1 or VEGFR2), VEGF165b may block transphosphorylation to produce a non-angiogenic response.
To further validate the observation of VEGF121/165b angiogenic properties seen in classical in vitro assays, we conducted in vivo experiments with Growth Factor Reduced Matrigel, in which VEGF, among other angiogenic cytokines, have been significantly depleted. Addition of recombinant VEGF121b, or VEGF165b, or VEGF165 caused recruitment of blood vessels to the Matrigel compared to PBS-loaded controls, thus demonstrating an angiogenic effect. It is worth mentioning that VEGF121b-containing Matrigels were highly enriched in dextran-FITC signal, which seemed to be within and out of the vessels (similar to results found for bFGF). This suggests an increase in vascular permeability, in keeping with vascular permeability described for VEGF121 .
The ability to inhibit tumor growth in vivo was tested in our study by VEGF121/165b overexpression in xenograft models. We chose two different cancer cell lines: The low-VEGF-expressing lung adenocarcinoma A549 cell line (~1-5 pg/mL (28)) and PC-3, a prostate cancer cell line that secretes ~800 pg/mL; . These cell types were chosen to ascertain possible differences in the VEGFxxxb behavior depending on the endogenous VEGF expression. Experiments were performed without any further exogenous VEGF stimulation, such as transfection of VEGF. In the PC3 xenografts, no statistical differences between controls and VEGF121/165b-expressing tumors were seen. However, a tendency of VEGF121b-overexpressing tumors to grow at a faster rate than the other groups was found. In agreement with these results, vascular density for VEGF121b-overexpressing PC-3 tumors was significantly higher than that observed for the rest of the groups, and apoptotic levels significantly lower. A549 xenografts grew slow and formed small tumors after subcutaneous implantation in nude mice, as previously described . Strikingly, overexpression of VEGF121/165b in these cells resulted in a significant increase in tumor development over the controls, although no differences in vascular density were found between groups, which may be due to the poor angiogenic potential of these tumors.
These results are in conflict with previous reports showing that VEGF165b causes anti-tumor effects [14, 15, 23, 24]. Our interpretation for these seemingly contradictory results is as follows: The in vivo models in which the VEGFxxxb isoforms have been proven to be efficacious in reducing tumor growth expressed high VEGF levels (either by the nature of the cells used or by transfection with VEGF-carrying plasmids). In these conditions, both VEGFxxx and VEGFxxxb proteins would compete equally for receptor binding. Because VEGFxxxb is only able to signal weakly, a reduced tumor volume will be observed when tumors overexpressing VEGFxxxb are compared to high VEGF-expressing tumors. However, in the case of tumors with low VEGF production, overexpression of VEGFxxxb would stimulate tumor growth to a certain extent. Thus, in CAKI cells, where endogenous total VEGF levels are ~900 pg/mL , when both parental and VEGF165b-overexpressing cells were injected into mice, no differences in tumor growth were found. Only when parental cells were transfected with VEGF165, a reduction in tumor volume was observed compared to VEGF165b-overexpressing tumors . Similarly, significant differences in tumor volumes were only obtained in Mel57 (<600 pg/mL, ) and A375 (~600 pg/mL, ) melanoma xenograft models when VEGF165- and VEGF165b-overexpressing clones were compared [14, 23]. On the contrary, in LS174t cells, which secrete high VEGF levels (~2000 pg/mL, ), overexpression of VEGF165b or VEGF121b , or administration of VEGF165b recombinant protein  led to tumor shrinkage compared to controls (parental cells with no further VEGF overexpression). CT71 Ewing sarcoma cell lines produce ~1200 pg/mL, . When CT71 cells overexpressing VEGF165b were xenotransplanted into nude mice, a significant reduction in tumor volume was found when compared with untransfected CT71 xenografts .
It is possible then that therapies based on the administration of VEGFxxxb proteins are effective in tumors with high endogenous VEGF expression. However, the use of this therapy in tumors with low VEGF levels, which likely rely on other angiogenic factors for their growth (such as bFGF, IL-8, etc.) may worsen the evolution of the tumor. Therefore, therapy with VEGFxxxb casts some doubts about its possible use in unselected patients. Caution must be taken in defining which possible patient will benefit from VEGFxxxb-based therapies. Stratifying the patients based on the amount of VEGF production may be critical for future possible clinical trials.
But before translating this type of therapy, many aspects of the VEGFxxxb biology should be clarified. For instance, it is quite surprising that VEGF121b, although found to inhibit endothelial cell migration, was cytoprotective for endothelial cells (in serum starvation experiments), in a similar way as VEGF165 . Such cytoprotection involved activation of VEGF receptors and downstream signalling . Even more surprising is the fact that in some xenograft models (for instance using CAKI cells), co-overexpression of VEGF165 and VEGF165b results in tumors that are smaller in size than those generated by overexpression of just VEGF165b . Other issues, such as whether or not VEGFxxxb proteins are able to heterodimerize with members of the VEGFxxx angiogenic family, should be clarified as well.
We have also addressed in this study the issue of whether the VEGFxxxb isoforms may be differentially expressed in malignant tissues (human breast cancer) compared to healthy tissues (normal mammary gland). For our experiments we used immunohistochemistry with validated antibodies specific for all VEGFxxxb isoforms and antibodies that recognize all VEGF transcripts. Our results revealed that both total VEGF and VEGFxxxb levels tended to increase in breast cancer samples (n = 50) compared to normal breast tissues (n = 8). This increase was statistically significant for intraductal carcinomas (IDC). Expression of both total VEGF and VEGFxxxb were significantly correlated, thus suggesting that both families of VEGF (VEGFxxx and VEGFxxxb) follow a similar pattern of expression.
Previous studies have analyzed, in a limited number of samples, expression of the VEGFxxxb isoforms. Whereas total VEGF mRNA levels were found significantly up-regulated in colon carcinoma samples (n = 6) compared to controls, no changes were observed for VEGFxxxb mRNA levels . This result shows that the increase in total VEGF is due to the VEGFxxx angiogenic isoforms . A similar result was found analyzing protein levels by isoform-specific ELISAs . RT-PCR analysis showed that VEGF165b was present in 17 of 18 normal kidney samples, but only in 4 of 18 matched malignant tissues . Immunohistochemical analysis of 19 melanoma samples (9 metastatic and 10 nonmetastatic) using the VEGFxxxb specific antibody, found a decrease in VEGFxxxb expression in the neoplastic tissue (especially in metastasis) compared to the normal skin .
The lack of appropriate antibodies specific for each of the VEGFxxx and VEGFxxxb proteins hinders the fine characterization of the pattern of expression in malignant and normal tissues. Future studies using larger number of samples, and quantitative real time RT-PCR and immunohistochemical analyses would be needed to assess whether VEGFxxxb expression could be used as a cancer biomarker.
Our results demonstrate that VEGF121/165b are not anti-angiogenic, but weakly angiogenic isoforms of VEGF-A that may foster tumor growth and angiogenesis in vivo. We also conclude that VEGFxxxb isoforms (as well as total VEGF levels) are up-regulated in breast cancer in comparison with non malignant breast tissues.
This work has been funded by "UTE project CIMA", ISCIII-RTICC RD06/0020/0066 grant; Ministerio de Educacion y Ciencia grant SAF2007-64184 (to AC), Gobierno de Navarra Salud (9/2007, to JA), ISCIII-RECAVA RD/0014/0008 (to JH). JA and ML were supported by an ISCIII fellowship, and LL by a Basque Government fellowship. We would also like to thank the Morphology and Image Analysis Unit at CIMA for technical assistance.
- Hanahan D, Folkman J: Patterns and emerging mechanisms of the angiogenic switch during tumorigenesis. Cell. 1996, 86: 353-364. 10.1016/S0092-8674(00)80108-7View ArticlePubMedGoogle Scholar
- Folkman J: Angiogenesis. Annu Rev Med. 2006, 57: 1-18. 10.1146/annurev.med.57.121304.131306View ArticlePubMedGoogle Scholar
- Ruan K, Song G, Ouyang G: Role of hypoxia in the hallmarks of human cancer. J Cell Biochem. 2009, 107: 1053-1062. 10.1002/jcb.22214View ArticlePubMedGoogle Scholar
- Semenza GL: HIF-1 inhibitors for cancer therapy: from gene expression to drug discovery. Curr Pharm Des. 2009, 15: 3839-3843. 10.2174/138161209789649402View ArticlePubMedGoogle Scholar
- Ferrara N: Role of vascular endothelial growth factor in the regulation of angiogenesis. Kidney Int. 1999, 56: 794-814. 10.1046/j.1523-1755.1999.00610.xView ArticlePubMedGoogle Scholar
- Ferrara N: Vascular endothelial growth factor: basic science and clinical progress. Endocr Rev. 2004, 25: 581-611. 10.1210/er.2003-0027View ArticlePubMedGoogle Scholar
- Kranenburg O, Gebbink MF, Voest EE: Stimulation of angiogenesis by Ras proteins. Biochim Biophys Acta. 2004, 1654: 23-37.PubMedGoogle Scholar
- Rafii S, Lyden D, Benezra R, Hattori K, Heissig B: Vascular and haematopoietic stem cells: novel targets for anti-angiogenesis therapy?. Nat Rev Cancer. 2002, 2: 826-835. 10.1038/nrc925View ArticlePubMedGoogle Scholar
- Jain RK: Lessons from multidisciplinary translational trials on anti-angiogenic therapy of cancer. Nat Rev Cancer. 2008, 8: 309-316. 10.1038/nrc2346View ArticlePubMedGoogle Scholar
- Ferrara N, Kerbel RS: Angiogenesis as a therapeutic target. Nature. 2005, 438: 967-974. 10.1038/nature04483View ArticlePubMedGoogle Scholar
- Tee MK, Jaffe RB: A precursor form of vascular endothelial growth factor arises by initiation from an upstream in-frame CUG codon. Biochem J. 2001, 359: 219-226. 10.1042/0264-6021:3590219PubMed CentralView ArticlePubMedGoogle Scholar
- Ladomery MR, Harper SJ, Bates DO: Alternative splicing in angiogenesis: the vascular endothelial growth factor paradigm. Cancer Lett. 2007, 249: 133-142. 10.1016/j.canlet.2006.08.015View ArticlePubMedGoogle Scholar
- Harper SJ, Bates DO: VEGF-A splicing: the key to anti-angiogenic therapeutics?. Nat Rev Cancer. 2008, 8: 880-887. 10.1038/nrc2505PubMed CentralView ArticlePubMedGoogle Scholar
- Woolard J, Wang WY, Bevan HS, Qiu Y, Morbidelli L, Pritchard-Jones RO, Cui TG, Sugiono M, Waine E, Perrin R, Foster R, Digby-Bell J, Shields JD, Whittles CE, Mushens RE, Gillatt DA, Ziche M, Harper SJ, Bates DO: VEGF165b, an inhibitory vascular endothelial growth factor splice variant: mechanism of action, in vivo effect on angiogenesis and endogenous protein expression. Cancer Res. 2004, 64: 7822-7835. 10.1158/0008-5472.CAN-04-0934View ArticlePubMedGoogle Scholar
- Rennel ES, Varey AH, Churchill AJ, Wheatley ER, Stewart L, Mather S, Bates DO, Harper SJ: VEGF(121)b, a new member of the VEGF(xxx)b family of VEGF-A splice isoforms, inhibits neovascularisation and tumour growth in vivo. Br J Cancer. 2009, 101: 1183-1193. 10.1038/sj.bjc.6605249PubMed CentralView ArticlePubMedGoogle Scholar
- Kawamura H, Li X, Harper SJ, Bates DO, Claesson-Welsh L: Vascular endothelial growth factor (VEGF)-A165b is a weak in vitro agonist for VEGF receptor-2 due to lack of coreceptor binding and deficient regulation of kinase activity. Cancer Res. 2008, 68: 4683-4692. 10.1158/0008-5472.CAN-07-6577View ArticlePubMedGoogle Scholar
- Glass CA, Harper SJ, Bates DO: The anti-angiogenic VEGF isoform VEGF165b transiently increases hydraulic conductivity, probably through VEGF receptor 1 in vivo. J Physiol. 2006, 572: 243-257.PubMed CentralView ArticlePubMedGoogle Scholar
- Varey AH, Rennel ES, Qiu Y, Bevan HS, Perrin RM, Raffy S, Dixon AR, Paraskeva C, Zaccheo O, Hassan AB, Harper SJ, Bates DO: VEGF 165 b, an antiangiogenic VEGF-A isoform, binds and inhibits bevacizumab treatment in experimental colorectal carcinoma: balance of pro- and antiangiogenic VEGF-A isoforms has implications for therapy. Br J Cancer. 2008, 98: 1366-1379. 10.1038/sj.bjc.6604308PubMed CentralView ArticlePubMedGoogle Scholar
- Bates DO, Cui TG, Doughty JM, Winkler M, Sugiono M, Shields JD, Peat D, Gillatt D, Harper SJ: VEGF165b, an inhibitory splice variant of vascular endothelial growth factor, is down-regulated in renal cell carcinoma. Cancer Res. 2002, 62: 4123-4131.PubMedGoogle Scholar
- Birkenhager R, Schneppe B, Rockl W, Wilting J, Weich HA, McCarthy JE: Synthesis and physiological activity of heterodimers comprising different splice forms of vascular endothelial growth factor and placenta growth factor. Biochem J. 1996, 316 (Pt 3): 703-707.PubMed CentralView ArticlePubMedGoogle Scholar
- Claffey KP, Senger DR, Spiegelman BM: Structural requirements for dimerization, glycosylation, secretion, and biological function of VPF/VEGF. Biochim Biophys Acta. 1995, 1246: 1-9. 10.1016/0167-4838(94)00144-6View ArticlePubMedGoogle Scholar
- Pietras K, Ostman A: Hallmarks of cancer: interactions with the tumor stroma. Exp Cell Res. 2010, 316: 1324-1331. 10.1016/j.yexcr.2010.02.045View ArticlePubMedGoogle Scholar
- Rennel E, Waine E, Guan H, Schuler Y, Leenders W, Woolard J, Sugiono M, Gillatt D, Kleinerman E, Bates D, Harper S: The endogenous anti-angiogenic VEGF isoform, VEGF165b inhibits human tumour growth in mice. Br J Cancer. 2008, 98: 1250-1257. 10.1038/sj.bjc.6604309PubMed CentralView ArticlePubMedGoogle Scholar
- Rennel ES, Hamdollah-Zadeh MA, Wheatley ER, Magnussen A, Schuler Y, Kelly SP, Finucane C, Ellison D, Cebe-Suarez S, Ballmer-Hofer K, Mather S, Stewart L, Bates DO, Harper SJ: Recombinant human VEGF165b protein is an effective anti-cancer agent in mice. Eur J Cancer. 2008, 44: 1883-1894. 10.1016/j.ejca.2008.05.027PubMed CentralView ArticlePubMedGoogle Scholar
- Pritchard-Jones RO, Dunn DB, Qiu Y, Varey AH, Orlando A, Rigby H, Harper SJ, Bates DO: Expression of VEGF(xxx)b, the inhibitory isoforms of VEGF, in malignant melanoma. Br J Cancer. 2007, 97: 223-230. 10.1038/sj.bjc.6603839PubMed CentralView ArticlePubMedGoogle Scholar
- Schumacher VA, Jeruschke S, Eitner F, Becker JU, Pitschke G, Ince Y, Miner JH, Leuschner I, Engers R, Everding AS, Bulla M, Royer-Pokora B: Impaired glomerular maturation and lack of VEGF165b in Denys-Drash syndrome. J Am Soc Nephrol. 2007, 18: 719-729. 10.1681/ASN.2006020124View ArticlePubMedGoogle Scholar
- Zhang Y, Furumura M, Morita E: Distinct signaling pathways confer different vascular responses to VEGF 121 and VEGF 165. Growth Factors. 2008, 26: 125-131. 10.1080/08977190802105909View ArticlePubMedGoogle Scholar
- Shenberger JS, Zhang L, Powell RJ, Barchowsky A: Hyperoxia enhances VEGF release from A549 cells via post-transcriptional processes. Free Radic Biol Med. 2007, 43: 844-852. 10.1016/j.freeradbiomed.2007.05.033PubMed CentralView ArticlePubMedGoogle Scholar
- Fang J, Zhou Q, Liu LZ, Xia C, Hu X, Shi X, Jiang BH: Apigenin inhibits tumor angiogenesis through decreasing HIF-1alpha and VEGF expression. Carcinogenesis. 2007, 28: 858-864. 10.1093/carcin/bgl205View ArticlePubMedGoogle Scholar
- Akhtar S, Meeran SM, Katiyar N, Katiyar SK: Grape seed proanthocyanidins inhibit the growth of human non-small cell lung cancer xenografts by targeting insulin-like growth factor binding protein-3, tumor cell proliferation, and angiogenic factors. Clin Cancer Res. 2009, 15: 821-831. 10.1158/1078-0432.CCR-08-1901View ArticlePubMedGoogle Scholar
- Westphal JR, Van't Hullenaar R, Peek R, Willems RW, Crickard K, Crickard U, Askaa J, Clemmensen I, Ruiter DJ, De Waal RM: Angiogenic balance in human melanoma: expression of VEGF, bFGF, IL-8, PDGF and angiostatin in relation to vascular density of xenografts in vivo. Int J Cancer. 2000, 86: 768-776. 10.1002/(SICI)1097-0215(20000615)86:6<768::AID-IJC3>3.0.CO;2-EView ArticlePubMedGoogle Scholar
- Sumimoto H, Imabayashi F, Iwata T, Kawakami Y: The BRAF-MAPK signaling pathway is essential for cancer-immune evasion in human melanoma cells. J Exp Med. 2006, 203: 1651-1656. 10.1084/jem.20051848PubMed CentralView ArticlePubMedGoogle Scholar
- Dang DT, Chun SY, Burkitt K, Abe M, Chen S, Havre P, Mabjeesh NJ, Heath EI, Vogelzang NJ, Cruz-Correa M, Blayney DW, Ensminger WD, St Croix B, Dang NH, Dang LH: Hypoxia-inducible factor-1 target genes as indicators of tumor vessel response to vascular endothelial growth factor inhibition. Cancer Res. 2008, 68: 1872-1880. 10.1158/0008-5472.CAN-07-1589View ArticlePubMedGoogle Scholar
- Guan H, Zhou Z, Wang H, Jia SF, Liu W, Kleinerman ES: A small interfering RNA targeting vascular endothelial growth factor inhibits Ewing's sarcoma growth in a xenograft mouse model. Clin Cancer Res. 2005, 11: 2662-2669. 10.1158/1078-0432.CCR-04-1206View ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.