Article Text
Abstract
Background and aims Vascular endothelial growth factor (VEGF) gene expression has been linked to cancer progression. Here we hypothesise that the polymorphism and protein expression of VEGF are correlated with the pathogenesis and therapy response of pterygium.
Methods 60 pterygial and 121 normal conjunctival samples were collected to determine the genotypes and protein expression of VEGF. Primary pterygium cells (PECs) were used to confirm the effect of the VEGF polymorphism on the angiogenesis of pterygium.
Results 48 (83.3%) pterygial specimens tested positive for VEGF protein expression, which was significantly higher than in the control groups (16.7%, p<0.0001). The frequency of the 936 C>T variant, but not the −2578C>A variant, was significantly higher in the pterygium group compared with the control group. VEGF protein expression was significantly higher in the 936 C/C group than in the 936 C/T and T/T groups (p=0.001). The results of our cell model showed that PECs with the C/C genotype had a higher angiogenesis ability and higher response to the antiangiogenesis drug bevacizumab than cells with the C/T and T/T genotypes.
Conclusions We suggest that VEGF could be used as a target for pterygium therapy in patients with the 936C>T genotype.
- Angiogenesis
- Conjunctiva
- Cornea
- Genetics
Statistics from Altmetric.com
Introduction
Angiogenesis is an important mechanism involved in tumour growth. One important angiogenic signalling factor is the vascular endothelial growth factor (VEGF) gene, located on chromosome 6q21.1 The coding region of VEGF spans approximately 14 kb and consists of eight exons.1 The gene product, VEGF, acts on vascular endothelial cells within minutes to increase vascular permeability and, over a longer term, reprogrammes gene expression, which is seen as proliferation and migration of endothelial cells and as generation of new blood vessels.2
Genetic variants are common, displaying a frequency of 1% or more in the population, which classifies them as polymorphisms rather than mutations. Polymorphisms are related to biodiversity, genetic variation and adaptation, and usually function to retain a variety of forms in a population living in a varied environment.3 In the case of VEGF, expression of the variant alleles results in increased levels of VEGF mRNA.3–6 Numerous genetic variants are recognised in the promoter region of VEGF, including −2578C>A, −2489C>T, −1498C>T, −1154 G>A, −634G>C, 936 C>T and 1612G>A.3–6 VEGF 936 C>T polymorphism is located in the 3′-UTR of the gene. The T variant of this polymorphism was associated with significantly lower VEGF plasma levels.5 In addition, the −2578C>A polymorphism was associated with reduced serum VEGF concentrations on epithelial ovarian cancer.7 Thus in this study, we focused on the −2578 C>A and 936 C>T polymorphisms.
A previous report also indicated that overexpression of the VEGF protein may contribute to the progression of pterygium in the eye by increasing angiogenesis and the growth of the primary pterygium.8 This disease represents an epithelial hyperpiesia associated with fibro-vascular growth.8–10 It is an active process, associated with cellular proliferation, remodelling of the connective tissue, angiogenesis and inflammation.8–10 However, the therapeutic response of pterygium to the antiangiogenesis drug ranibizumab has been inconsistent. For example, Mandalos et al11 showed that ranibizumab had no effect on the extent of vascularisation of primary pterygium while Shenasi et al12 indicated that subconjunctival injection of bevacizumab immediately after surgical excision of a primary pterygium was well tolerated but did not significantly prevent recurrence of this condition.
In the present study, we hypothesised that expression levels of the VEGF protein and the therapeutic response to antiangiogenesis drugs are correlated with VEGF allele variants in pterygium patients. We therefore analysed the correlation of VEGF gene polymorphisms and protein expression using real time PCR and immunohistochemistry methods in 60 pterygial specimens and 121 normal conjunctival samples. We also used cell models to determine whether the therapeutic response of pterygium to the antiangiogenesis drug bevacizumab is correlated with VEGF allele variants.
Materials and methods
Study subjects
Pterygial samples were harvested from 60 patients undergoing pterygium surgery. All patients gave written informed consent and the study was approved by the institutional review board of China Medical University Hospital. Patients in whom the apex of the pterygium had invaded the cornea by more than 1 mm were included in the study. The pterygia were classified as grade 1, 2 or 3 based on slit lamp biomicroscopic evaluation. Grade 1 (‘atrophic’) had clearly visible episcleral vessels under the body of the pterygium. Grade 2 (‘intermediate’) had partially visible episcleral vessels under the body of the pterygium. Grade 3 (‘fleshy’) had totally obscured episcleral vessels under the body of the pterygium. The controls included normal conjunctival samples collected from the superior conjunctiva of 62 patients and the medial conjunctiva of 58 patients without pterygium and pinguecula; all of these patients were undergoing cataract or vitreoretinal surgery. There were 40 men and 20 women in the pterygium group (age range 55–82 years, mean 65.7 years), and 60 men and 60 women in the control group (age range 55–75 years, mean 62.8 years). Normal conjunctival samples were collected from bulbar conjunctivas. All pterygial specimens came from primary pterygia.
VEGF genotyping analyses
Polymorphisms in the VEGF gene at positions −2578C>A (rs699947) and 936 C>T (rs3025039) in the 5′-UTR region were determined on an ABI 7900HT using allelic discrimination and previously developed assays (Applied Biosystems, Birkerod, Denmark). Reactions of 5 μL contained approximately 50 ng of DNA, 2.5 μL of Mastermix (Applied Biosystems) and the predesigned assays. Controls were included in each run, and a repeat analysis of a subset (10%) of the samples yielded 100% identical genotypes. The pair of primers and two probes, which were designed and validated by Applied Biosystem, of each polymorphism site were used. The primers were used for identification of the DNA sequence that contains the polymorphism, and the two probes were categorised as ‘normal probe’ (used to amplify the wild type-allele) and ‘polymorphic probe’ (to amplify the allele with the polymorphism). These reagents allow identification of samples from normal homozygous individuals (both alleles without polymorphism), polymorphic homozygous (both alleles with polymorphism) and heterozygous (one of the alleles with a polymorphism). Allele discrimination can be determined when the fluorescent probe hybridises in the complementary target region that should be amplified. The results were analysed on a 7900 real-time PCR system using the allelic discrimination assay program Sequence Detection Software V.2.4 (Applied Biosystems).
Immunohistochemistry
Formalin fixed, paraffin embedded specimens were sectioned at a thickness of 3 μm. All sections were deparaffinised in xylene, sequentially rehydrated through serial dilutions of alcohol and washed in phosphate buffered saline. Sections used for VEGF detection were immersed in citrate buffer (pH 6.0) and heated in a microwave oven twice for 5 min. Mouse anti-VEGF monoclonal antibodies (at a dilution of 1:200) (Santa Cruz) were used as the primary antibodies. The detailed protocol used has been described in our previous report.13 Negative controls that did not include the primary antibodies were also prepared. Lung tumour tissue was used as a positive control. The results were evaluated independently by three observers and were scored for percentage of positive expression. The VEGF protein was expressed only in the cytoplasm. We counted the number of epithelial cells and VEGF positive cells of the pterygium or normal conjunctiva in three independent fields using a high power field (objective lens 40×). Cells that positively stained for the anti-VEGF antibody were noted by their labelling index as a percentage in each specimen, and the measurements were averaged. The results regarding VEGF in pterygial tissues are presented as means. Scores were indicated as follows: score 0, no positive staining; score +, 1–10% staining; score ++, 11–50% staining; and score +++, >50% positive cells. In this study, scores of +, ++ and +++ were considered to represent positive immunostaining, and a score of 0 was classified as negative immunostaining.
Pterygium cell lines
Pterygium cell lines (PECs) used in this study were established previously.12 Fresh pterygial specimens were cut into small pieces (1–2 mm in diameter) under a stereomicroscope, washed in a solution of Dulbecco's modified Eagle's medium (DMEM) and placed in a culture dish. Serum free DMEM was added to cover the explants, and the culture dish was placed into a CO2 regulated incubator with a 5% CO2 atmosphere overnight. The culture medium was replaced three times a week after the appearance of an outgrowth of cells from the explants. To confirm whether the established PECs were epithelial cells, the cell type was further confirmed via staining with p63 and pan cytokeratin antibodies (Santa Cruz Biotechnology, Santa Cruz, California, USA).14
VEGF protein expression evaluated by western blot analysis
Total proteins were extracted from PECs with a lysis buffer (100 mM Tris, pH 8.0, 0.1% sodium dodecyl sulfate) and the concentrations of recovered proteins were determined using the Bio-Rad protein assay kit (Bio-Rad Laboratories, Hercules, California, USA), followed by separation with sodium dodecyl sulfate-polyacrylamide gel electrophoresis (12.5% gel, 1.5 mm thick). After electrophoretic transfer to a polyvinylidene difluoride membrane, non-specific binding sites were blocked with 5% non-fat milk in Tris buffered saline-Tween 20. Detection of VEGF and β-actin was conducted by incubating the membrane with anti-VEGF (Santa Cruz) and β-actin antibodies (Sigma, Missouri, USA). The detailed protocol used has been described in our previous reports.13 The protein bands were visualised using enhanced chemiluminescence (NEN Life Science Products Inc, Boston, Massachusetts, USA). Densitometer analysis of the films was performed using a computerised image analysis (AlphaImager HP; Alpha Innotech, California, USA) program. Protein expression levels were established by calculating the target molecule/β-actin ratio (all cases were scored for band intensity compared with that of the internal control).
Expression of VEGF with different genotypes in PECs
Full length VEGF with different genotypes was amplified by PCR from conjunctival controls and the resulting PCR products were purified with GENECLEAN III kit and then cloned into a eukaryotic expression vector, pcDNA3.1/V5-His TOPO TA Expression Kit (Invitogen, California, USA). The recombinant plasmid was transfected into PECs. On the day prior to transfection, PECs were seeded at 1×105 cells per well and, after an overnight incubation, cells, at 30–50% confluent, were washed twice with phenol red free DMEM medium without fetal bovine serum. The 2 μg expression plasmids were transiently transfected into PECs using the Transfast reagent. Stable transfectants were selected by culturing those transfected cells in the medium containing antibiotic G418.
In vitro tube formation assay
The tube formation assay is one of the most widely used in vitro assays for modelling the reorganisational stage of angiogenesis. This assay measures the ability of endothelial cells, plated at subconfluent densities with the appropriate extracellular matrix, to form capillary-like structures (ie, tubes). Tube formation assays were run by coating each well of a 48 well plate with 250 μL/cm2 of Matrigel, which was allowed to solidify for 30 min at 37°C before addition of cells. PECs with different allele variants were seeded at 2×104 cells/well in the Matrigel coated wells and then incubated in endothelial basal medium supplemented with growth factors (0.5% fetal bovine serum, 1 μg/mL hydrocortisone, 10 ng/mL human epidermal growth factor, 100 ng/mL VEGF, 3 ng/mL basic fibroblast growth factor and 15 IU/mL heparin) for 18 h. Tube formation was assessed by counting the number of branch points in three randomly selected microscopic fields (×40 original magnification) per subject. All experiments were conducted in triplicate.
Statistical analysis
Statistical analysis was performed using the SPSS statistical software program (SPSS Inc, Chicago, Illinois, USA). Fisher's exact test and the χ2 test were applied for statistical analysis. A p value <0.05 was considered to be statistically significant.
Results
Comparison of genetic polymorphism and protein expression of VEGF between the pterygium and control groups
The association between risk and genetic change in the angiogenesis genes in pterygium development was verified by analysing the allelic variants of the VEGF gene in the pterygium and control groups. Table 1 presents the VEGF genotypes, −2578C>A (rs699947) and 936 C>T (rs3025039), of the pterygium and control groups. The −2578C>A (rs699947) genotype analysis showed no statistically significant difference between the pterygium and control groups (p=0.161). The 936 C>T (rs3025039) analysis showed that 40 (66.6%) were homozygous for the C/C genotype, 10 (16.7%) were homozygous for the T/T genotype and 10 (16.7%) were heterozygous for the C/T genotype in the pterygium group. In the control group, 79 (63.3%) were homozygous for the C/C genotype, three (2.5%) were homozygous for the T/T genotype and 39 (34.2%) were heterozygous for the C/T genotype. A statistically significant difference existed between the pterygium and control groups for the VEGF 936 C>T (rs3025039) genotype (p<0.001). Only the 936 C>T (rs3025039) genotype was associated with pterygium development. Therefore, in the following analysis, we focused on the VEGF 936C>T genetic polymorphism.
Correlation of VEGF protein expression with genetic polymorphism in pterygium patients
The difference in VEGF protein expression between the pterygium and control groups was verified using immunohistochemistry to analyse VEGF protein expression (see online supplementary figure 1). A significantly higher expression frequency of the VEGF protein was found in the pterygium group compared with the control group (83.3% vs 16.7%, p<0.0001) (table 1). The associations between VEGF genetic polymorphism and protein expression were also statistically analysed. As shown in table 2, the frequency of positive expression of the VEGF protein was higher in the pterygium group for patients with the 936 C/C genotype (95.0%, 38 of 40) than with the C/T and T/T genotypes (60.0%, 12 of 20; p=0.001). No significant association was found in the control group (VEGF protein positive expression in the C/C group, 14 of 76, 18.4%; C/T and T/T groups, six of 44, 13.7%, p=0.615).
VEGF polymorphism as a risk factor for pterygium
We compared VEGF protein expression and the 936 C>T polymorphism in the pterygium and control groups. After adjustment for age and gender, multiple logistic regression analysis showed that the VEGF 936C>T genotype was related to pterygium. The prevalence of pterygium in subjects with VEGF protein expression appeared to be greater than the prevalence of pterygium in those without VEGF protein expression (OR 21.876, 95% CI 9.241 to 51.788, p<0.001) (table 3). In addition, a higher frequency of pterygium was apparent in subjects with VEGF 936 C/C genotypes than in those with C/T and TT genotypes (OR 2.387, 95% CI 1.040 to 5.495, p=0.040) (table 3). No statistically significant difference was noted in terms of gender (p=0.509) or age (p=0.319) (table 3).
Effect of VEGF genetic polymorphism on in vitro tube formation by PEC cells
We verified whether VEGF protein activity in pterygium was associated with genetic allelic variation by establishing PECs from pterygium patients with different VEGF 936C>T genotypes. The VEGF 936C>T genotype in these cell lines was detected by real-time PCR. To confirm that VEGF protein activity could be affected by genotype, VEGF protein expression and tube formation ability in the PECs with different genotypes were evaluated by western blot and an in vitro tube formation assay, respectively. As shown in figure 1A, there was no difference in VEGF protein expression in normal conjunctival cells with different genotypes (figure 1A). In addition, VEGF protein expression levels in PECs with the C/C genotype were significantly higher than in the C/T and T/T genotypes (figure 1B). We also found that tube formation was significantly higher in PECs with C/C genotypes than in cells with the C/T and the T/T genotypes (figure 1C). No any tube formation was detected in conjunctival cells. Transfection of the VEGF gene with the C allele into the PECs (with the C/T and the T/T genotypes) increased tube formation, indicating that this allelic variant was associated with VEGF protein activity (figure 1C).
Tube formation ability in bevacizumab treated PEC cells
The potential therapeutic response of pterygium to anti-VEGF targeting was verified by analysing tube formation by PECs following treatment with bevacizumab. As shown in figure 2, tube formation significantly decreased in PECs with the homogenous C/C (figure 2C) allele compared with the C/T (figure 2B) and T/T (figure 2A) genotypes, indicating that the therapeutic response of the antiangiogenesis drug was correlated with the VEGF allele variants. Furthermore, as shown in figure 2, after bevacizumab treatment, the tube formation inhibition ability of the PECs with the T/T alleles was lower than the C allele transfected cells (figure 2A). In addition, this association was also found in the C/C allele PECs transfected with the T allele. After bevacizumab treatment, the tube formation inhibition ability of the PECs with the C/C alleles was higher than the T allele transfected cells (figure 2C). These results indicate that the allele variant of the VEGF was associated with an anti-VEGF therapeutic response.
Discussion
In this study, we used immunohistochemistry to demonstrate that VEGF was expressed in pterygial tissues. Positive immunostaining clearly showed significantly higher levels of VEGF in pterygial epithelial cells than in normal conjunctival epithelium. These results are consistent with previous reports that expression of the VEGF protein is significantly higher in the epithelial and vascular endothelial component of pterygium than in normal conjunctiva and limbus.14–17
Genetic variants of VEGF have been associated with tumour progression and clinical outcomes in several types of cancer, including breast cancer, lung cancer, colorectal cancer and prostate cancer.7 Although overexpression of the VEGF protein has been shown in pterygium tissues,14–17 the relationship between VEGF gene polymorphism and pterygium has remained unclear. Previous studies indicated no significant differences in the VEGF-460 polymorphism in Taiwan between pterygium and control groups.18 ,19 However, these studies reported an approximately 2.5-fold increased risk of developing pterygium in women who carried at least 1 C allele (C/C and C/T genotypes) compared with those who carried the T/T genotype. This association was not found in male patients. The present study demonstrated a correlation between pterygium and VEGF polymorphism in the VEGF 936 C>T (rs3025039) but not in −2578C>A (rs699947) genotypes. Pterygium was also correlated with VEGF protein expression in the VEGF 936 C>T genotype (table 2). In addition, a higher risk of pterygium was apparent in subjects with VEGF 936 C/C genotypes than with the C/T and TT genotypes (OR 2.387, 95% CI 1.040 to 5.495, p=0.001) (table 3). Therefore, we believe that the polymorphism of VEGF in 936 C>T is an important risk factor for pterygium.
Bevacizumab has been approved for use in oncology; it is a full length recombinant humanised monoclonal antibody that binds to VEGF and inhibits its biological activity. Bahar et al20 reported that subconjunctival bevacizumab injection did not have any effect on the formation of new vessels in recurrent pterygium. In a study of subconjunctival bevacizumab injection in corneal neovascularisation secondary to different aetiologies (ie, herpetic keratitis, chemical burn, failed graft), subconjunctival bevacizumab was found to be well tolerated and associated with a partial regression of corneal neovascularisation.21 However, the therapeutic response of pterygium to the antiangiogenesis drug ranibizumab has been inconsistent.11 ,12 Our present study demonstrated significantly higher tube formation by PECs with the C/C genotype than by cells with the C/T and T/T genotypes (figure 1C). In addition, after treatment with bevacizumab, the tube formation ability was significantly decreased in PECs with the homogenous T/T allele compared with the C/T and C/C genotypes (figure 2). Thus we believe that the therapeutic response of pterygium to this antiangiogenesis drug will depend on the VEGF allele variant that is present.
VEGF is known to be a ligand for the VEGF receptor (VEGFR). Previous studies reported that levels of VEGFR1 and VEGFR2 mRNA were lower in pterygia than in conjunctivas but similar in limbal and pterygium samples.9 Another report showed that expression of VEGF-C and its receptor VEGFR3 in pterygium were consistent and correlated with lymphatic vessel density. Expression levels of VEGFR were significantly higher in pterygium compared with control conjunctiva.22 A further study showed that the other angiogenesis factor, erythropoietin, and erythropoietin receptor, were highly expressed in pterygium compared with normal controls, and correlated with cell proliferation and angiogenesis in human pterygium.23 Thus we hypothesis that angiogenesis is involved in the pathogenesis of pterygium.
In present study, we found that the frequency of positive expression of the VEGF protein was higher in the pterygium group for patients with the 936 C/C genotype (95.0%, 38 of 40) than with the C/T and T/T genotypes (60.0%, 12 of 20; p=0.001) (table 2). We also found that PECs with the C/C allele was more responsive to bevacizumab treatment (figure 2). Previous reports showed the VEGF 936 T allele change led to loss of a potential binding site for the activating enhancer binding protein 4 (AP-4) transcription factor, which increases the expression of several viral and cellular genes.24 ,25 Thus we suggest that the VEGF 936 C>T polymorphism is associated with VEGF protein expression and protein activity, affecting the response to bevacizumab treatment.
In conclusion, we used pterygium tissues and primary PECs to demonstrate that genetic polymorphism of VEGF may play a role in pterygium formation and may contribute to the angiogenesis ability of pterygium. We also found that the therapeutic response of pterygium to an antiangiogenesis drug may depend on the type of VEGF allele variant. Based on our findings, the genetic polymorphism and protein expression of the VEGF gene may have potential use in the selective treatment of pterygium with anti-VEGF agents. Expression of VEGFR also needs to be further analysed.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
- Data supplement 1 - Online figure
Footnotes
-
M-LP and Y-YT contributed equally.
-
Contributors Y-WC designed the study and wrote the paper. M-LP, HL and J-NT designed the experiments, wrote the paper and prepared the figures. M-LP, Y-CH and Y-YT collected the pterygium and control samples. C-CC analysed the data. All authors gave final approval for the manuscript to be submitted for publication. M-LP and Y-YT contributed equally to this work.
-
Funding This work was supported by grants from the National Science Council (NSC 96-2314-B-039-009 -MY2) and Tungs’ Taichung MetroHarbor Hospital (TTM-TMU-102-01) of Taiwan.
-
Competing interests None.
-
Ethics approval The study was approved by the institutional review board of China Medical University Hospital.
-
Provenance and peer review Not commissioned; externally peer reviewed.