Received 20 September 2009 | Accepted 4 December 2009 | Published 10 December 2009
1Department of Ophthalmology and Visual Sciences Washington University School of Medicine, St. Louis, MO; 2Barnes Retina Institute, St. Louis, MO
Correspondence to: Milam A. Brantley Jr., Department of Ophthalmology and Visual Sciences/Campus Box 8096, Washington University School of Medicine, 660 S. Euclid Avenue, St. Louis, MO, 63110; Phone: (314) 747-5606, FAX: (314) 747-4121, email: Brantley@vision.wustl.edu
Purpose: Genetic factors influence an individual’s risk for developing neovascular age-related macular degeneration (AMD), a leading cause of irreversible blindness. Previous studies on the potential genetic link between AMD and vascular endothelial growth factor (VEGF), a key regulator of angiogenesis and vascular permeability, have yielded conflicting results. In the present case-control association study, we aimed to determine whether VEGF or its main receptor tyrosine kinase VEGFR-2 is genetically associated with neovascular AMD.
Methods: A total of 515 Caucasian patients with neovascular AMD and 253 ethically-matched controls were genotyped for polymorphisms in the VEGFA and VEGFR-2 genes. A tagging single nucleotide polymorphism (tSNP) approach was employed to cover each gene plus two kilobases on each side, spanning the promoter and 3′ untranslated regions. SNPs with a minimum allele frequency of 10% were covered by seven tSNPs in VEGFA and 20 tSNPs in VEGFR-2. Two VEGFA SNPs previously linked with AMD, rs1413711 and rs3025039, were also analyzed.
Results: The 29 VEGFA and VEGFR-2 SNPs analyzed in our cohort demonstrated no significant association with neovascular AMD. A single rare haplotype in the VEGFR-2 gene was associated with the presence of neovascular AMD (p=0.034).
Conclusions: This study is the first to investigate the association of VEGFR-2 polymorphisms with AMD and evaluates VEGFA genetic variants in the largest neovascular AMD cohort to date. Despite the angiogenic and permeability-enhancing effects of VEGF/VEGFR-2 signaling, we found minimal evidence of a significant link between polymorphisms in the VEGFA and VEGFR-2 genes and neovascular AMD.
Age-related macular degeneration (AMD) is a leading cause of irreversible blindness in individuals over the age of 50 in the Western world . In addition to dietary and environmental risk factors, genetics influences AMD susceptibility . Recently, single nucleotide polymorphisms (SNPs) in the complement factor H (CFH) gene [3-5] and the ARMS2/HTRA1 locus [6,7] have been strongly linked to AMD. Common variants in the complement factor B (CFB)/complement component 2 (C2) [8,9] and complement component 3 (C3) [10-13] genes have also been associated with presence of AMD. Vascular endothelial growth factor (VEGF), coded for by the VEGFA gene, is a key mediator of angiogenesis and vascular permeability. Thus, VEGF is a logical candidate for genetically influencing AMD susceptibility, based on its functional relevance to AMD pathophysiology.
VEGF plays a key role in the angiogenesis, vascular leakage, and inflammation that is characteristic of the neovascular form of advanced AMD . In recent years, the VEGF signaling pathway has been targeted for inhibition therapy to treat choroidal neovascularization secondary to AMD . VEGF regulates angiogenesis in the vascular endothelium through the high-affinity receptor tyrosine kinases VEGFR-1 (Flt-1) and VEGFR-2 (Flk-1, KDR) . Of these two receptors, VEGFR-2 is responsible for the majority of the angiogenic and permeability-enhancing effects of VEGF [17,18]. VEGFR-1 regulates VEGF activity in the vascular endothelium by preventing VEGF/VEGFR-2 binding .
A relationship between the VEGFA and VEGFR-2 genes and neovascular AMD seems plausible, given the role of choroidal neovascularization in the pathophysiology of late-stage AMD and the importance of the receptor VEGFR-2 in the VEGF-signaling pathways that modulate angiogenesis and vascular permeability. Several small-scale case-control studies have reported associations between various SNPs in VEGFA and AMD [18-21], but these results have not been confirmed by subsequent studies [22-24]. While a recent analysis of VEGFA SNPs showed no associations with neovascular AMD, it did suggest that a three-SNP VEGFA haplotype increased the risk of neovascular AMD . To our knowledge, the potential link between VEGFR-2 and AMD has not been investigated. In this study, we focused on neovascular AMD, as its positive response to anti-VEGF therapy implicates the VEGF pathway in the pathogenesis of this AMD subtype. Using a tagging SNP (tSNP) approach in a large cohort of neovascular AMD patients, we aimed to determine if a genetic association exists between the VEGFA or VEGFR-2 genes and neovascular AMD.
This retrospective case-control association study was approved by the Washington University Human Research Protection Office and the Barnes Retina Institute Study Center. Research adhered to the tenets of the Declaration of Helsinki and was conducted in accordance with Health Insurance Portability and Accountability Act regulations. All participants were enrolled from the clinical offices of the Barnes Retina Institute, and informed consent was obtained before participation. Participants were identified through chart review, and only Caucasian patients with a diagnosis of neovascular AMD were included. Controls were identified as having no signs of AMD and were recruited from the same locations.
A total of 768 patients were genotyped for SNPs in the VEGFA and VEGFR-2 genes using DNA extracted from mouthwash samples. Each participant provided buccal tissue samples by expectorating into 50 ml conical tubes (Falcon; BD Biosciences, San Jose, CA) after vigorously rinsing for 30 s with approximately 20 ml Scope mouthwash (Procter & Gamble, Cincinnati, OH). Genomic DNA was prepared from the collected buccal cells using the Gentra Puregene Buccal Cell Kit (Qiagen, Valencia, CA).
We employed a tSNP approach to cover the VEGFA and VEGFR-2 genes plus two kilobases (kb) on each side, including the promoter and 3′ untranslated regions. Using HapMap Project Build 36, we identified the SNPs in each gene with a minimum allele frequency (MAF) of 10% [18,22]. The minimum r2 value was set to 0.80, and genotypes for the Centre d’Etude du Polymorphisme Humain (CEPH) from Utah (CEU) population (Utah residents with ancestry from northern and western Europe) were used. We selected VEGFA tSNPs and VEGFR-2 tSNPs covering the HapMap-identified SNPs through the Tagger algorithm. Using Sequenom MassARRAY technology (Sequenom Inc., San Diego, CA), study participants were genotyped for these tSNPs, as well as for two additional SNPs (rs1413711 and rs3025039) previously associated with AMD [19,20].
Descriptive statistics for all demographic and clinical variables were calculated, and comparisons were made using the ANOVA test for means with continuous data (e.g., age) and the chi-square test for categorical data (e.g., gender). Genotyping data was loaded into Haploview in linkage format to generate allele frequencies, ratios, and p values based on a chi-square test for association of alleles . The Hardy–Weinberg equilibrium test was used to confirm that genotypes fell within a standard distribution. Genotype distributions were analyzed by logistic regression, incorporating adjustments for age and gender. Haplotype blocks encompassing the tested SNPs were defined by 95% confidence intervals and were predicted using Haploview. Statistics were performed with SPSS (version 17; SPSS Inc., Chicago, IL). For all statistical analyses, p<0.05 was considered to be statistically significant, and multiple comparisons were adjusted for by the Bonferroni correction.
Using HapMap and Tagger for tSNP selection, we identified seven VEGFA tSNPs and 22 VEGFR-2 tSNPs to cover both genes. Because two assays failed design, 20 tSNPs in VEGFR-2 were selected for genotyping. The two failed tSNPs correlated poorly with other SNPs, and alternative SNPs could not be chosen. The 20 genotyped tSNPs in VEGFR-2 had a mean r2 of 0.97 and captured 44 of the 46 SNPs with a MAF greater than 10%.
A total of 515 Caucasian patients with neovascular AMD and 253 ethnically-matched controls were genotyped for this study. Cases and controls had mean ages of 79.2 and 69.3 years, respectively (p<0.001). The case cohort consisted of a lower percentage of males (33.6%) than did the control cohort (43.9%; p=0.005). Adjustments for age and gender were incorporated in the genotype analysis of both VEGFA and VEGFR-2. All analyzed SNPs conformed to Hardy–Weinberg equilibrium in both the case and control populations. Allele distributions did not differ significantly between cases and controls for any of the nine VEGFA SNPs (Table 1). According to the genotype distribution for these SNPs (Table 2), one tSNP (rs699947) and one SNP previously associated with AMD (rs1413711) demonstrated significant uncorrected p values, but no significant association was found for any VEGFA SNP after correcting for multiple comparisons. Allele distributions for the 20 VEGFR-2 tSNPs showed significant uncorrected p values for three tSNPs, but none remained significant after correcting for multiple comparisons (Table 3). No VEGFR-2 tSNPs were associated with neovascular AMD by genotype distribution (Table 4).
Linkage disequilibrium-based haplotypes were defined for the VEGFA and VEGFR-2 SNPs, resulting in two haplotype blocks for VEGFA (Figure 1) and five haplotype blocks for VEGFR-2 (Figure 2). Haplotype analysis demonstrated no VEGFA haplotypes to be associated with neovascular AMD (Table 5) but did reveal a mild association of the rarest haplotype (TTT) in VEGFR-2 haplotype block 3 with neovascular AMD (p=0.034, Table 6).
We analyzed nine SNPs in VEGFA and 20 tSNPs in VEGFR-2 using a tSNP approach designed to cover both genes. The results of this study show little evidence of association with neovascular AMD in our cohort of 515 Caucasian patients.
Previous investigation of VEGFA and AMD has yielded conflicting results. A few smaller-scale case-control studies have reported significant associations for various SNPs in VEGFA [18-21], but the validity of these results remains unconfirmed [22-24]. The only one of these studies to investigate a large neovascular AMD cohort (n=342) using a comprehensive tSNP approach found no link between VEGFA tSNPs and development of neovascular AMD . More recently, a single haplotype was shown to be weakly associated with neovascular AMD in a moderately-sized cohort (n=211), although no associations were found with individual VEGFA SNPs . In the present study, we found no association between neovascular AMD and VEGFA SNPs or haplotypes. These discrepant findings may be due to inadequate sample size in the early studies or to the ethnic composition of study populations. Results could also be biased by genotyping error or confounding effects due to statistically significant variables such as age, gender, or other SNPs known to be associated with AMD. As a whole, the studies performed to date provide little evidence that VEGFA polymorphisms exert any significant influence on risk of neovascular AMD.
This is the first study to investigate the relationship between variants in VEGFR-2 and AMD. One of two receptor tyrosine kinases involved in VEGF signaling pathways, VEGFR-2 mediates the majority of the angiogenic and permeability-enhancing effects of VEGF . The importance of VEGFR-2 in developmental angiogenesis and hematopoiesis, as demonstrated by the abnormal vasculature of VEGFR-2 knockout mice , suggests a link between the VEGF/VEGFR-2 pathway and retinal pathology. While VEGF and its receptors play a key role in tumor angiogenesis and other pathological conditions [14,16,28,29], a limited number of gene association studies have been performed for VEGF receptor genes. Recently, polymorphisms in VEGFR-1 and VEGFR-2 were reported to be associated with sarcoidosis, an inflammatory condition with a hypothesized antigenic stimulus . Variants in VEGFR-2 have also been linked with heart disease and may influence the risk of developing breast cancer [31-33]. In our large neovascular AMD cohort, no associations were found for any VEGFR-2 tSNPs by allele or genotype analysis. Haplotype analysis, however, did show a single rare haplotype to be mildly associated with AMD.
This study is limited by its retrospective design, which did not allow for assessment of the predictive value of VEGFA and VEGFR-2. However, in light of our negative findings, it is doubtful that a prospective study would contribute significantly to our knowledge base. This study was designed to investigate common polymorphisms that might be associated with neovascular AMD risk, and it remains possible that rare variants with MAFs less than 10% could play a role in neovascular AMD development. Due to tSNP assay failure, two of the 46 VEGFR-2 SNPs with a MAF greater than 10% were poorly tagged. The 20 successful VEGFR-2 tSNPs in our study covered the polymorphisms previously found to be associated with human diseases: rs1870377, rs2071559 (covered by rs7667298), rs2125489, rs2305948, rs7667298, and rs7691507 [30-33].
In summary, this study is the first to investigate the association of VEGFR-2 polymorphisms with AMD and evaluates VEGFA genetic variants in the largest neovascular AMD cohort to date. Although both VEGF and its receptors have been implicated in the pathophysiology of diseases such as AMD, we found minimal evidence that polymorphisms in VEGFA and VEGFR-2 contribute significantly to risk of neovascular AMD.
The authors thank the Human Genetics Division Genotyping Core at the Washington University School of Medicine for assisting with genotyping and the physicians of the Barnes Retina Institute for generously contributing their time and patients to this study. This work was supported by the Jahnigen Career Development Award from the American Geriatrics Society (M.A.B.), the Carl M. & Mildred A. Reeves Foundation (M.A.B.), NEI Core Grant 5 P30 EY02687, and a grant from Research to Prevent Blindness to the Washington University School of Medicine Department of Ophthalmology and Visual Sciences.