No association between polymorphisms/haplotypes of the vascular endothelial growth factor gene and preeclampsia

Background Preeclampsia (PE) is the first worldwide cause of death in pregnant women, intra-uterine growth retardation, and fetal prematurity. Some vascular endothelial grown factor gene (VEGF) polymorphisms have been associated to PE and other pregnancy disturbances. We evaluated the associations between VEGF genotypes/haplotypes and PE in Mexican women. Methods 164 pregnant women were enrolled in a case-control study (78 cases and 86 normotensive pregnant controls). The rs699947 (-2578C/A), rs1570360 (-1154G/A), rs2010963 (+405G/C), and rs25648 (-7C/T), VEGF variants were discriminated using Polymerase Chain Reaction - Restriction Fragment Length Polymorphism (PCR-RFLP) methods or Taqman single nucleotide polymorphism (SNP) assays. Results The proportions of the minor allele for rs699947, rs1570360, rs2010963, and rs25648 VEGF SNPs were 0.33, 0.2, 0.39, and 0.17 in controls, and 0.39, 0.23, 0.41, and 0.15 in cases, respectively (P values > 0.05). The most frequent haplotypes of rs699947, rs1570360, rs2010963, and rs25648 VEGF SNPs, were C-G-C-C and C-G-G-C with frequencies of 0.39, 0.21 in cases and 0.37, 0.25 in controls, respectively (P values > 0.05) Conclusion There was no evidence of an association between VEGF alleles, genotypes, or haplotypes frequencies and PE in our study.


Background
Pre-eclampsia (PE) is a multi-organ disorder defined as elevated blood pressure (>140/90 mmHg) and proteinuria that may develop from 20 weeks gestation. PE is the first cause of maternal death in the western world and affects 5 to 8% of pregnant women, depending on the studied population and definitions of PE [1][2][3]. The Mexican Secretary of Health Registry reported 278 deaths by PE in 2007 in the country [4].
The etiopathology of PE remains unknown, but improper placentation has been suggested as a possible etiologic factor. This process involves the abnormal infiltration of the cytotrophoblast into the decidual endometrium and development of the spiral arteries originated from the myometrium. As a result, the diameter of the spiral arteries enlarge while arteries lose elasticity to allow an increased demand of blood flow to satisfy the intensifying demands of the growing fetus. Improper cytotrophoblast invasion may affect vascular remodelling and blood supply to the fetus, leading to local placental hypoxia [5][6][7]. Expression of vascular endothelial growth factor (VEGF) has been implied as a hypoxia-induced factor possibly associated to PE, as has been suggested previously [8]. Gu et al. observed increased mRNA and protein levels of VEGF in response to hypoxia in cultured cytotrophoblasts [9]. Zhou et al. found that cytotrophoblast differentiation and invasion during pregnancy is regulated through the VEGF receptor-2. They also found that VEGF regulates cytotrophoblast survival and that expression is deregulated in severe * Correspondence: margaritamf@uaz.edu.mx 1 Laboratorio de Medicina Molecular, Unidad Academica de Medicina Humana y Ciencias de la Salud, Universidad Autonoma de Zacatecas, Carretera Zacatecas-Guadalajara Km.6. Ejido la Escondida, Zacatecas, Mexico Full list of author information is available at the end of the article forms of PE. [10] Several authors reported an association between some VEGF polymorphisms and PE in selected populations, with some discrepancies [3,[11][12][13][14]. No association studies for gene polymorphisms and PE have been reported for the Mexican population, despite the prevalence of this pregnancy disorder in the region. The purpose of the present study was to analyze the association of rs699947 (-2578C/A), rs1570360 (-1154G/A), rs2010963 (+405G/C), and rs25648 (-7C/T) VEGF polymorphisms in a sample of Mexican pregnant women enrolled during 2009-2010.  [15]. In general, these guidelines include two recordings of diastolic blood pressure of 90 mmHg at least 4 h apart in previously normotensive women and proteinuria of 300 mg or more in 24 h, or two readings of at least ++ on dipstick analyses. PE was considered severe if blood pressure of 160 mm Hg systolic or higher or 110 mm Hg diastolic or higher on two occasions at least 6 hours apart while the patient was on bed rest, and proteinuria of 5 g or higher in a 24-hour urine specimen or 3+ or greater in two random urine samples collected at least 4 hours apart were present [16]. Exclusion criteria included multiple pregnancies, pregnancies with structural or chromosomic fetus malformations detected by ultrasound, chronic hypertension, chronic renal disease, gestational diabetes mellitus, and underlying medical diseases such as rheumatoid arthritis, and systemic lupus erythematosus.

Patients and samples collection
A questionnaire, consisting of demographic and gynecologic data and risk factors (including family history of PE, cardiovascular disease, type 2 diabetes mellitus or autoimmune diseases) was administered to each patient by the participant gynecologist, and a venous blood sample was obtained using EDTA collection tubes. Blood samples were maintained at 4°C until their processing.

SNP selection
Criteria for SNP selection included: SNPs for which an association with PE has been reported in other populations, and considering the potential impact on gene expression only SNPs located in promoter region of VEGF gene or inside its 5'UTR were considered. According to these criteria, rs699947 and rs1570360 variants (located in the VEGF promoter), and rs2010963 and rs25648 (located in 5'UTR of the VEGF gene) were included. Relative to translation start site, the variants rs699947 and rs1570360 are known as -2578C/A and -1154G/A, respectively [17,18]; while the rs2010963 SNP is called +405G/C (according to mRNA position) or -634G/C (relative to translation start site). In the same sense, rs25648 alternative names are 534C/T and -7C/T respectively.

Nucleic acid extraction and genotyping
DNA was isolated from 1 mL of blood sample using a standard fenol/chloroform method coupled to ethanol precipitation. Table 1 shows a summary of the genotyping methods employed for the variants screening. PCR coupled to Restriction Fragment Length Polymorphism (PCR-RFLP) method was used to discriminate the rs699947, rs1570360, and rs2010963 VEGF polymorphic variants, respectively. We used the protocols described by Banyasz I et al (rs699947 and rs2010963) [3] and Liu Q et al. (rs1570360) [19], respectively. Some modifications in the primer sequences were included; these modifications are indicated in bold letters in Table 1. All the PCR reactions were made in a volume of 10 μl containing 100 ng of genomic DNA, 0.5 μM of each primer, and 1× of GoTaq Colorless Master Mix (Promega: Madison, WI). PCR cycling conditions were as follows: 94°C for 5 min; 35 cycles of 94°C for 30 s, 60°C of annealing temperature for 30 s, and 72°C for 30 s; a final extension step of 72°C for 5 min was included. PCR products were digested with Bgl II (rs699947), Mnl I (rs1570360), and BsmF I (rs2010963) at 37°C (rs699947 and rs1570360) or at 65°C (rs2010963) overnight, respectively. The rs699947 and rs2010963 digested products were electrophoresed through 3.5% agarose gels and visualized by ethidium bromide staining. Digested products of rs1570360 VEGF variant were visualized using the DNA 1000 LabChip kit on the Agilent 2100 Bioanalyzer (Agilent: Palo Alto, CA). Length of PCR and PCR-RFLPs products are shown in Table 1.
The evaluation of the rs25648 VEGF variant was performed using a predesigned 5'-nuclease assay with TaqMan MGB probes (Applied Biosystems: Foster City, CA). All reactions were performed in 10 μL of volume and 10 ng of genomic DNA. Standard PCR cycling conditions were used and the allelic discrimination was made using the StepOne RT-PCR system with the StepOne v2.0 software (Applied Biosystems).
All the SNPs were validated by direct sequencing of PCR products obtained from four patients using BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems), and analyzed in an Avant 3100 Genetic Analyzer. To ensure reliability of the results, duplicate samples and known genotyped samples were included in the analysis as quality controls.

Data Analysis
General variables of the study groups such as maternal age, gestational age, and number of pregnancies, were compared using the Student's t test. In the control group, SNPs genotypes were tested for departures from Hardy-Weinberg equilibrium by exact test. To compare known risk factors, allele, and genotype frequencies between the study groups, a chi-square test was performed. Linkage disequilibrium statistics and haplotypes analysis, including haplotypes frequency estimation, as well as an analysis of association between haplotypes and PE, were made using SNPStats software http:// bioinfo.iconcologia.net/index.php?module=Snpstats. The Odds Ratio (OR) values were calculated for each allele, genotype, and haplotype. For the adjustment of gestational age, logistic regression analysis was used to compare allele and genotype frequencies between PE and the control group. Study power was calculated at the end of the study using Quanto software v1.2.4 [20]. For each SNP we considered its minor allele frequency as reported in the Mexican HapMap project. Depending on the SNP, in the analysis an effect of size 2 or greater was established.
The level of significance was set at P < 0.05. Statistical analysis was performed using Sigma Plot software v11.

Results
One hundred and sixty four unrelated pregnant women (86 PE cases and 78 normotensive controls) were included in the study. The comparison among the general variables of the study population (Table 2) showed that gestational age was the only variable associated to PE (P = 0.028). We did not find an association between known risk factors and PE (P > 0.05). Genotype frequencies for all the SNPs included in the study were in Hardy-Weinberg equilibrium in the controls (P > 0.05). Table 3 displays the genotypes and minor allele frequencies for each SNP in cases and controls. The proportions for the minor allele for rs699947, rs1570360, rs2010963, and rs25648 VEGF SNPs were 0.33, 0.2, 0.39, and 0.17 in controls, and 0.39, 0.23, 0.41, and 0.15 in cases, respectively (Table 3). We did not observe significant differences between individual genotypes or allele frequencies of the studied VEGF SNPs in the cases and control groups (P values > 0.05).
In a statistical approximation, an association between genotype/allele frequencies and severity of the disease was not found (table S1 of Additional file 1). To test the association between PE onset and VEGF genotype variants, the group of severe PE patients was classified in early and late PE, with a cut off at 35 GW. Again, no association between VEGF genotypes and PE onset was found (P values > 0.05; data not shown).  To study the association between PE and possible combinations of the VEGF variants, we performed a haplotype analysis on for the four VEGF SNPs. A linkage disequilibrium (LD) test, showed a LD association between the rs699947, rs1570360, rs2010963, and rs25648 VEGF variants (D' > 0.87, range: 0.882 to 0.998 P < 0.001). The data for D'statistic of each pairwise of SNPs and its p values are shown in table S2 of Additional file 2. A comparison between haplotypes frequencies of VEGF SNPs and their association with PE is summarized in Table 4. Five major combinations between VEGF SNPs studied were determined; four of them with frequencies above 10% (C-G-C-C, C-G-G-C, A-A-G-C, and A-G-G-T for rs699947, rs1570360, rs2010963, and rs25648 VEGF SNPs, respectively) and one of them, A-G-G-C, with a frequency of 1.5%. The P values for individual and global haplotype score tests (Table 4) did not indicate a significant difference in VEGF haplotype frequency profiles between cases and controls (P > 0.05).
Considering the sample size of 164, the study power was calculated in a range of 45.19% to 65.96% to detect an effect of size 2 or greater depending on the SNP considered.

Discussion
Genetic predisposition is a well-documented risk factors for PE [21,22]; however, although several genes have been postulated as genetic susceptibility markers of the disease [21][22][23][24], there is not a recognized hereditary profile to explain the association [21]. Considering differences in genetic predisposition between populations and the important role of VEGF gene in the angiogenesis during the pregnancy, we conducted a case-control study, to test the association between the rs699947, rs1570360, rs2010963, and rs25648 VEGF genetic variants and PE. Consistent with the lack of association between -2578C/A (rs699947) VEGF genotypic/allelic frequencies in women with severe PE reported by Papazoglou D et al (2004) in Greece [13], we did not find an association between genotype/allele frequencies of the studied SNPs and PE or with severity of the disease (P > 0.05). However, our results contrasted with those reported by Sandrim VC et al. (2009) for Caucasic Brazilian women, who reported a positive association between -2578C/A (but not for -1154G/A) genotypes and PE [12]. Banyasz I et al (2006) suggested that carriers of the VEGF+405G (rs2010963) allele in an Hungarian population have a decreased susceptibility to PE and while carriers of the VEGF-2578A (rs699947) allele could be associated with an accelerated development of the disease in nulliparous pregnant women with severe PE [3]. In our study, carriers of the VEGF+405G (rs2010963) allele did not have a decreased susceptibility to PE and no association was found between allele/genotypes frequencies of the studied VEGF variants (including VEGF-2578 C/A) and PE onset. Discrepancies of VEGF allele/genotype frequencies and their association with PE, severity, or onset of disease, may be explained  in part by the ethnic variability between populations and by differences in selection criteria of the studied groups. For example, the lack of consensus in the controls selection, which has included post-menopausal women with a history of at least two uncomplicated pregnancies, nulliparous pregnant women, women with uncomplicated pregnancies, or normotensive pregnant women, makes difficult the comparison between studies and may include some bias in the association reports. PE susceptibility related to VEGF haplotypes has been considered in a very few studies; however, none of them has genotyped all the variants included in this work. We did not find differences in the identified haplotypes profiles between cases and controls. In contrast with our results, Sandrim VC et al. (2009) suggested a protective effect for the -2578C, -1154G and -634C haplotype (associated with higher VEGF gene expression) against the development of PE [12]. In our study, we did not include the -634G/C variant; however, although we performed a statistical analyses in which only the -2578C/A, -1154G/A SNPs were considered for this analysis, the lack of association between VEGF haplotypes frequencies and PE was consistent. Finally, some considerations should be highlighted: One of the major limitations of this work was the small sample size because it influenced the study power (calculated in a range of 45.19% to 65.96% depending on the SNP studied). In this same sense, the limited sample size affects the association analyses performed in sub-classified groups, as PE severity or PE onset, since sub-classifications reduced the numbers of study subjects and could introduce some bias in the statistical analysis. In our, study the nulliparous state was not considered as selection criteria or was subclassified for additional comparisons; however, this PE risk factor showed no difference between groups. Our control group included normotensive women, but their "current healthy pregnant state" does not discard the future presence of PE pregnancies.
In Mexico, there are no data about genetic susceptibility for PE associated to VEGF variants. Our study is the first protocol that evaluated the possible association between VEGF genotypes/haplotypes and PE, and the results obtained in the study suggest the presence of additional genetic variants different from rs699947, rs1570360, rs2010963, and rs25648 VEGF as markers for the genetic susceptibility to PE.

Conclusion
In our study, there was no evidence of an association between allele, genotype or haplotype frequencies and PE, its severity or onset of the disease.

Additional material
Additional file 1: Comparison of alleles, genotypes, and haplotypes frequencies between study groups, according to disease severity.
Additional file 2: Linkage Disequilibrium Analysis. D' statistics and P values for each parwise of VEGF variants included in the study.