Vascular endothelial growth factor (−460 C/T, +405 G/C, and +936 C/T) polymorphisms and endometriosis risk in Tunisian population

Abstract

The vascular endothelial growth factor (VEGF), a major angiogenic factor, is known to play an important role in the development of endometriosis. The aim of this study was to investigate the association of three VEGF (−460 C/T, +405 G/C, and +936 C/T) polymorphisms with the risk of endometriosis in the Tunisian population. This study includes 105 women with endometriosis and 150 women with no laparoscopic evidence of disease. Genotyping of the VEGF −460 C/T, +405 G/C, and +936 C/T polymorphisms were performed by polymerase chain reaction restriction fragment length polymorphism (PCR-RFLP). The distribution of genotypes (P = 0.006) and allele (P = 0.0009) frequencies of the +936 C/T polymorphism was significantly different between patients and controls. Patients with stages III-IV endometriosis showed a higher VEGF + 936T allele frequency than controls (P = 0.0001). However, the distribution of genotypes and allele frequencies of the VEGF −460 C/T and +405 G/C polymorphisms did not differ significantly between endometriosis patients and controls. These findings suggest that the VEGF +936 C/T polymorphism may be a risk factor for endometriosis development and the VEGF +936 T allele is associated with an increased risk of stages III-IV endometriosis in the Tunisian population.

• Endometriosis
• gene
• polymorphism
• vascular endothelial growth factor
• VEGF

Introduction

Endometriosis is a common gynecological disorder characterized by the presence of endometrial tissue outside the uterus and associated with both pelvic pain and infertility [Bulun 2009]. It affects 6 to 10% of women of reproductive age and 25–50% of women with infertility [Bulletti et al. 2010]. Endometriosis is defined as a multifactorial and polygenic disease, in which angiogenesis may be implicated in implantation and development [Taylor et al. 2002]. Endometrium undergoes cyclical growth and regression during the menstrual cycle, which depends on ovarian steroid levels and angiogenic growth factors, including the vascular endothelial growth factor (VEGF) [McLaren 2000]. The VEGF, a major mediator of angiogenesis, induces endothelial cell proliferation, migration, differentiation, capillary formation, and would contribute to the pathogenesis and progression of endometriosis [Ferrara 2004].

The human VEGF gene is located on chromosome 6p21.3 and consists of eight exons exhibiting alternate splicing which form a family of proteins. Several single nucleotide polymorphisms (SNPs) in the VEGF gene were described [Ferrara et al. 2003; Watson 2000]. Among them, the −460 C/T (rs833061) located in the VEGF promoter region, the +405 G/C (rs2010963) located in exon 1 of the VEGF gene, and the +936 C/T (rs3025039) located in exon 8, corresponding to the 3′ untranslated region (UTR) of the gene were extensively investigated in the literature and reported to be associated with differential VEGF expression and the risk of endometriosis [Cosín et al. 2009; Hsieh et al. 2004; Renner et al. 2000].

Based upon the genetic predisposition and the important role of VEGF in the pathogenesis of endometriosis, it is necessary to investigate whether these three functional VEGF (−460 C/T, +405 C/G, and +936 C/T) polymorphisms are associated with endometriosis risk in the Tunisian population. We also assessed the association between the VEGF gene polymorphisms and serum levels and performed a VEGF haplotype-based analysis in association with endometriosis risk.

Results

The genotyping of the −460 C/T, +405 G/C, and +936 C/T polymorphisms of VEGF gene by PCR-RFLP was successfully achieved for all subjects using the Table 1 primers. The genotype distributions of the VEGF −460 C/T, + 405 G/C, and + 936 C/T polymorphisms were in Hardy–Weinberg equilibrium in both patient and control groups. There was no significant difference between the patients and controls in mean age (34.46 ± 4.20 and 34.68 ± 4.40 years, respectively). Demographic data were similar among groups.

Table 1. Restriction fragment length polymorphism conditions for identification of VEGF gene polymorphisms.

VEGF SNPs	PCR Primers	PCR product	Restriction enzymes	Restriction fragment length
−460 C/T (rs833061)	5′-TGTGCGTGTGGGGTTGAGCG-3′ (F) 5′-TACGTGCGGACAGGGCCTGA-3′ (R)	175 bp	Bsh12361 (62°C)	175 bp (T) 155 + 20 bp (C)
+405 G/C (rs2010963)	5′-ATTTATTTTTGCTTGCCATT-3′ (F) 5′-GTCTGTCTGTCTGTCCGTCA-3′ (R)	304 bp	FaqI (57°C)	304 bp (C) 203 + 101 bp (G)
+936 C/T (rs3025039)	5′-AGGGTTCGGGAACCAGATC-3′ (F) 5′-CTCGGTGATTTAGCAGCAAG-3′ (R)	266 bp	Hin1II (66°C)	266 bp (C) 211 + 55 bp (T)

VEGF: vascular endothelial growth factor; SNP: single-nucleotide polymorphism; PCR: polymerase chain reaction; bp: base pairs; F: forward; R: reverse.

As shown in Table 2, the distribution of genotypes and allele frequencies of the − 460 C/T and + 405 G/C polymorphisms did not differ significantly between patients and controls.

Table 2. Distribution of genotypes and alleles frequencies of VEGF polymorphisms in endometriosis patients and controls.

	Population
VEGF polymorphisms	Patients n = 105	Controls n = 150	P value	OR (95% CI)
−460C/T Genotype
CC	15 (14.28%)	29 (19.33%)		1.00^a
CT	53 (50.48%)	67 (44.67%)	0.32	1.52 (0.74–3.14)
TT	37 (35.24%)	54 (36%)	0.58	1.32 (0.62–2.8)
TT + CT	90 (85.72%)	121 (80.67%)	0.37	1.43 (0.72–2.84)
−460C/T Allele
C	83 (39.52%)	125 (41.67%)		1.00^a
T	127 (60.48%)	175 (58.33%)	0.69	1.09 (0.72–1.56)
+405G/C Genotype
GG	37 (35.24%)	61 (40.67%)		1.00^a
GC	44 (41.9%)	68 (45.33%)	0.93	1.06 (0.61–1.86)
CC	24 (22.86%)	21 (14%)	0.11	1.88 (0.92–3.84)
GC + CC	68 (64.76%)	89 (46.47%)	0.45	1.26 (0.75–2.11)
+405G/C Allele
G	118 (56.19%)	190 (63.33%)		1.00^a
C	92 (43.81%)	110 (36.67%)	0.12	1.34 (0.93–1.93)
+936C/T Genotype
CC	62 (59.05%)	117 (78%)		1.00^a
CT	33 (31.43%)	26 (17.33%)	0.006*	2.39 (1.31–4.36)
TT	10 (9.52%)	7 (4.67%)	0.08	2.69 (0.97–7.42)
TT + CT	43 (40.95%)	33 (22%)	0.001*	2.45 (1.42–4.25)
+936C/T Allele
C	157 (74.76%)	260 (86.67%)		1.00^a
T	53 (25.24%)	40 (13.33%)	0.0009*	2.19 (1.39–3.46)

VEGF: vascular endothelial growth factor; OR: odds ratio; CI: confidence interval; *significant (P value) < 0.05; ^agenotype or allele category accepted as reference group.

However, the endometriosis group showed a higher VEGF + 936T allele frequency than the control group (P = 0.0009), which resulted from an increased proportion of heterozygote CT genotype carriers among endometriosis patients as compared to controls (31.43 versus 17.33 %, P = 0.006 and OR = 2.39 (1.31−4.36)).

No significant differences were observed in the allele frequencies of the −460 C/T, +405 G/C, and +936 C/T polymorphisms between the endometriosis patients at stages I-II and controls, (P = 0.46, P = 0.21 and P = 0.07, respectively). However, a significant difference was observed in the frequency of the VEGF 936T allele in the subgroup of endometriosis patients at stages III–IV compared with the control group (32.15% versus 13.33%, P = 0.0001 and OR = 3.07 (1.74–5.42; Table 3). Furthermore, we found no significant differences in the allele frequencies and genotype distributions of the − 460 C/T, +405 G/C, and +936 C/T polymorphism between patients at stages I-II and those at stages III-IV.

Table 3. Distribution of genotypes and alleles frequencies of VEGF polymorphisms in stages I-II and stages III-IV endometriosis patients and controls.

VEGF polymorphisms	Endometriosis stages I-II n = 63	Endometriosis stages III-IV n = 42	P^b	OR^b (95% CI)	P^c	OR^c (95% CI)
−460C/T Genotype
CC	9 (14.28%)	6 (14.28%)		1.00^a		1.00^a
CT	29 (46.04%)	24 (57.14%)	0.58	1.39 (0.58–3.31)	0.39	1.73 (0.64–4.68)
TT	25 (39.68%)	12 (28.58%)	0.50	1.49 (0.61–3.61)	0.88	1.07 (0.36–3.15)
CT + TT	54 (85.72%)	36 (85.72%)	0.49	1.43 (0.63–3.24)	0.60	1.43 (0.55–3.73)
−460C/T Allele
C	47 (37.3%)	36 (42.85%)		1.00^a		1.00^a
T	79 (62.7%)	48 (57.15%)	0.46	1.2 (0.78–1.84)	0.94	0.95 (0.58–1.55)
+405 G/C Genotype
GG	21 (33.33%)	16 (38.1%)		1.00^a		1.00^a
GC	29 (46.03%)	15 (35.71%)	0.63	1.23 (0.64–2.39)	0.81	0.84 (0.38–1.84)
CC	13 (20.64%)	11 (26.19%)	0.25	1.79 (0.76–4.21)	0.21	1.99 (0.8–4.98)
GC + CC	42 (66.67%)	26 (61.9%)	0.39	1.37 (0.73–2.54)	0.9	1.11 (0.55–2.24)
+405 G/C Allele
G	71 (56.35%)	47 (55.95%)		1.00^a		1.00^a
C	55 (43.65%)	37 (44.05%)	0.21	1.33 (0.87–2.04)	0.27	1.36 (0.83–2.32)
+936C/T Genotype
CC	40 (63.5%)	22 (52.38%)		1.00^a		1.00^a
CT	20 (31.74%)	13 (30.95%)	0.03*	2.25 (1.13–4.46)	0.02*	2.65 (1.18–5.95)
TT	3 (4.76%)	7 (16.67%)	0.95	1.25 (0.3–5.08)	0.005*	5.31 (1.69–16.67)
CT + TT	23 (36.5%)	20 (47.62%)	0.04*	2.03 (1.07–3.87)	0.002*	3.22 (1.57–6.61)
+936C/T Allele
C	100 (79.37%)	57 (67.85%)		1.00^a		1.00^a
T	26 (20.63%)	27 (32.15%)	0.07	1.69 (0.98–2.91)	0.0001*	3.07 (1.74–5.42)

VEGF: vascular endothelial growth factor; OR: odds ratio; CI: confidence interval; *significant (P value) <0.05; ^agenotype or allele category accepted as reference group; ^bstage I-II endometriosis patients versus controls; ^cstage III-IV endometriosis patients versus controls.

Analysis of VEGF serum levels between endometriosis patients and controls

VEGF serum levels were estimated and detectable in all endometriosis patients and controls. As shown in Figure 1, the VEGF serum levels were statistically significantly higher in endometriosis patients with the mean of 593.97 ± SD 434.70 pg/ml compared to the control group (82.11 ± 49.91 pg/ml), (P < 0.0001).

Figure 1. Comparison of VEGF serum levels between endometriosis patients and controls.

Based on our results which showed that the VEGF + 936 C/T polymorphism was strongly associated with endometriosis risk (P = 0.0009), we further evaluated the impact of the different genotypes of this SNP on the serum VEGF levels in the endometriosis patients and controls (Figure 2). We found that VEGF serum levels were significantly higher in endometriosis patients with VEGF + 936 CT genotype (1114.0 pg/mL ± 359.20) compared to the control group (157.03 pg/mL ± 65, 97) (P < 0.0001) but not with +936 CC and +936 TT genotypes (P > 0.05).

Figure 2. Comparison of VEGF serum levels with VEGF genotypes in endometriosis patients and controls.

Haplotype analysis

There was strong linkage disequilibrium between the VEGF − 460 C/T and +405 G/C polymorphisms with a standardized disequilibrium coefficient (D′) of 0.91 between the two loci. However, the +936C/T polymorphism was not linked to the other polymorphisms (D′ < 0.22). Therefore, haplotype analysis was only conducted between VEGF −460 T/C and +405G/C polymorphisms, and four haplotypes were inferred (Table 4). The frequencies of the −460/+405 haplotypes TG, CG, TC, and CC in the control group were 38.24, 25.09, 20.08, and 16.59%, respectively. There were no significant differences in the haplotypes frequencies of the VEGF −460/ +405 between the endometriosis patients and controls.

Table 4. Haplotype frequencies of VEGF −460C/T and +405G/C polymorphisms in endometriosis patients and controls.

VEGF haplotypes		Haplotype frequency
−460	+405	Endometriosis (%)	Control (%)	OR (95% CI)	P value
T	G	34.04	38.24	1.00^a	–
C	G	22.16	25.09	0.96 (0.55–1.68)	0.90
T	C	26.45	20.08	1.39 (0.84–2.27)	0.19
C	C	17.35	16.59	1.16 (0.68–1.98)	0.56

VEGF: vascular endothelial growth factor; OR: odds ratio; CI: confidence interval; Significant (P value) < 0.05; ^areference group.

Discussion

Association has been reported between the VEGF 5′ or 3′-UTR polymorphisms and endometriosis development. In the present study, two SNPs in the 5′-UTR (−460 C/T and +405 G/C) and one in the 3′-UTR (+936 C/T) of the VEGF gene were investigated to verify whether these three polymorphisms are associated with endometriosis susceptibility in the Tunisian population. We demonstrated a statistically significant association between the risk of endometriosis and the VEGF +936 C/T polymorphism (P = 0.0009). In addition, endometriosis women showed a higher VEGF +936T allele frequency with an increased risk of stages III–IV compared to the controls (P = 0.0001 and OR = 3.07 (1.74–5.42)), which suggest that the VEGF +936T allele may play an important role in the pathogenesis of endometriosis. However, no significant association was found between the VEGF −460 C/T (P = 0.69) and +405 G/C (P = 0.12) polymorphisms and endometriosis development.

To the best of our knowledge this is the first study in the Tunisian population that evaluated these three functional VEGF polymorphisms and revealed a positive association between endometriosis and +936 C/T polymorphism of the VEGF gene. A summary of association studies between VEGF gene polymorphisms and endometriosis risk is presented in Table 5. In keeping with previous findings reported by Cosín et al. [2009] in a Caucasian population, the +936 T allele was significantly more represented in endometriosis patients compared with controls (P = 0.02; OR = 1.75, 95% CI = 1.12–2.74), whereas the VEGF −460 C/T and +405 G/C polymorphisms were not associated with the risk of endometriosis. Similarities were also observed with the results obtained by Ikuhashi et al. [2007] in the Japanese population, which suggest that the +936T allele is associated with an increased risk of stages III–IV of endometriosis (P = 0.018; OR = 1.57, 95% CI = 1.08–2.29). In contrast with our data, other studies conducted in Chinese, Estonian, and Korean populations showed no evidence for an association between the +936 C/T polymorphism of VEGF gene and endometriosis [Kim et al. 2008; Lamp et al. 2010; Liu et al. 2009]. Moreover, Zhao et al. [2008] evaluated four SNPs (−2578 A/C, −460 C/T, +405 G/C, and +936 C/T) in Australian women and concluded no significant association between the genotyped VEGF polymorphisms and endometriosis. However, in their study, the control group might not be representative because the women that were selected have never undergone laparoscopy, whereas endometriosis should be a surgically and histologically confirmed diagnosis. The differences within the literature may be due to the dissimilar genetic and environmental backgrounds.

Table 5. A summary of association studies between VEGF gene polymorphisms and endometriosis risk.

				VEGF SNPs
				−460 C/T	+405 G/C	+936 C/T
Country	Ethnicity	Patients	Controls	rs833061	rs2010963	rs3025039	Reference
China	Asian	122	131	<0.05	–	–	Hsieh et al. [2004]
India	Asian	215	210	NS*	<0.05	–	Bhanoori et al. [2005]
China	Asian	215	219	NS	<0.05	–	Kim et al. [2005]
Japan	Asian	147	181	NS	NS	<0.05	Ikuhashi et al. [2007]
Korea	Asian	105	101	–	–	NS	Kim et al. [2008]
Australia	Caucasian	958	959	NS	NS	NS	Zhao et al. [2008]
Spain	Caucasian	186	180	NS	NS	<0.05	Cosín et al. [2009]
China	Asian	344	360	NS	–	NS	Liu et al. [2009]
Turkey	Caucasian	52	60	NS	<0.05	–	Attar et al. [2010]
Estonia	Caucasian	150	199	–	NS	NS	Lamp et al. [2010]
Turkey	Caucasian	98	94	NS	<0.05	–	Altinkaya et al. [2011]
Iran	Asian	480	600	NS	<0.05	–	Emamifar et al. [2012]
Tunisia	African	105	150	NS	NS	<0.05	Our study

VEGF: vascular endothelial growth factor; SNP: single nucleotide polymorphism; *significant p value <0.05.

NS*: non significant P value.

Analyzing the VEGF serum levels in this present study, we demonstrated that the VEGF serum levels were statistically significantly higher in endometriosis patients compared to the control (P < 0.0001) which is in agreement with the findings reported by Matalliotakis et al. [2003] in the Caucasian population (P < 0.001) but in contrast with those conducted by Kianpour et al. [2013]. Additionally, we reported a significant association of the +936 CT genotype with an increased VEGF serum levels than the CC and TT genotypes (P < 0.0001).

Moreover, Renner et al. [2000] studied several VEGF polymorphisms, analyzed their relationship to VEGF lasma levels, and found that the +936 C/C genotype was associated with higher plasma VEGF levels than the +936 C/T and T/T genotypes. However, in their study, VEGF plasma levels were determined in 23 samples obtained from men in 16 samples with the +936 C/C genotype and only 7 samples with the +936 C/T or T/T genotype. Thus, their findings may not be relevant. In comparison, Awata et al. [2003] studied VEGF serum levels and VEGF genotypes in the Japanese population and showed no significant correlation between the VEGF +936 C/T polymorphism and serum levels in diabetic retinopathy. The possible reasons for this discrepancy could be due to a different ethnic background of study subjects, sample preparations, and method of measurement. Whether or not the VEGF +936C/T polymorphism relates to different VEGF expression remains unclear. The mechanism by which the +936 C/T polymorphism affects the VEGF levels is currently unknown. Furthermore, the analysis of potential transcription factor binding sites showed that the +936 C→T mutation led to the loss of a potential binding site for AP-4 (activating enhancer binding protein 4) [Quandt et al. 1995]. AP-4 is a helix-loop-helix transcription factor enhancing expression of several viral and cellular genes by binding to specific enhancer sites [Hu et al. 1990; Mermod et al. 1988]. Nevertheless, it is currently unclear if the loss of this AP-4 binding site is of any relevance for the expression of VEGF. Another possible explanation for the association between the + 936 C/T polymorphism and the VEGF levels could be in a linkage disequilibrium between this polymorphism and another yet unknown functional polymorphism in the VEGF gene sequence.

Genotyping the VEGF +405 C/G polymorphism, showed no significant differences in the frequency and genotype distribution of this SNP between endometriosis patients and controls. Our results are consistent with some published studies [Cosín et al. 2009; Ikuhashi et al. 2007; Lamp et al. 2010; Zhao et al. 2008], but are discordant with other reports [Altinkaya et al. 2011; Attar et al. 2010; Bhanoori et al. 2005; Emamifar et al. 2012; Kim et al. 2005]. Bhanoori et al. [2005] investigated the VEGF −460 C/T and +405 C/G polymorphisms in South Indian women and reported a positive association between the −405 G/C polymorphism and endometriosis (P = 0.002) with the −405 G allele significantly more frequent in stage III-IV patients (81.7%) compared to controls (72.7%). Kim et al. [2005] also evaluated these two SNPs in the Korean population, however, they showed a higher frequency of the +405 CC genotype in patients with severe endometriosis implying that the +405 C, rather than the G allele may be involved in the risk of developing severe endometriosis. Discrepancies among studies could be caused by ethnic differences in study populations.

In our study, we examined the VEGF −460 C/T polymorphism and reported no significant association between this SNP and endometriosis development, which is in agreement with most of the previous studies [Li et al. 2013]. In contrast, Hsieh et al. [2004] showed an association between the VEGF −460 C/T polymorphism and endometriosis in the Chinese population. The discrepancy might be due to ethnic composition differences between the two studies or to the fact that the frequency of the −460 CC, −460 CT, and −460 TT genotypes of the VEGF gene polymorphism (0, 63.4 and 36.6%, respectively) in the control group of the Hsieh et al. [2004] study were not in Hardy–Weinberg equilibrium (P = 0.00006), most likely implying a genotyping problem. Therefore, the results of our study may be more reliable.

As discussed above, in most of the studied populations, there is a relationship between the risk of endometriosis and VEGF polymorphisms which may have an effect on the regulation of gene expression [Li et al. 2013]. The effect could be caused by a single polymorphism, or by the combination with other polymorphisms.

The findings obtained in the present study demonstrated that the VEGF +936C/T polymorphism is associated with the risk of endometriosis in the Tunisian population. In particular, the VEGF +936T allele might affect the pathogenesis of endometriosis with an increased risk of stages III–IV. However, the −460 C/T and +405 G/C polymorphisms of the VEGF gene seem not to be good markers for predicting susceptibility to endometriosis development. Obviously, further studies are required to confirm the functional significance of the +936 T allele and how it participates in the pathogenesis of endometriosis.

Materials and Methods

Subjects

The study group included 105 infertile women (mean age ± standard deviation, 34.46 ± 4.2) with endometriosis who were diagnosed after undergoing laparoscopy and diagnosis was later confirmed by histopathological examination. Endometriosis stages were stratified by the Revised American Society for Reproductive Medicine classification [ASRM 1997] according to the severity of disease into two groups as follows: the patients having minimal (stage I = 27) and mild (stage II = 36) endometriosis as stages (I–II) and the patients having moderate (stage III = 18) and severe (stage IV = 24) endometriosis as stage (III–IV).

In addition, 150 women (mean age ± standard deviation, 34.68 ± 4.4 years) who had undergone diagnostic laparoscopy and had no evidence of endometriosis served as the control group. Patients having rheumatoid arthritis [Han et al. 2004], giant cell arthritis [Boiardi et al. 2003], diabetic retinopathy [Awata et al. 2003], breast cancer [Krippl et al. 2003], and Behçet’s disease [Salvarani et al. 2004] were excluded from this study due to the possible association with a VEGF gene polymorphism. All women with leiomyoma, adenomyosis, fibroids, and invasive carcinoma of the uterine cervix or ovarian cancer were excluded from both cases and controls. All subjects involved in this study had regular menstrual cycles, no history of previous pelvic surgery, and had not received any hormonal medical treatment at least six months prior to the onset of the study. The design of this study was approved by the National Ethics Committee of Aziza Othmana Hospital and all participants signed informed consent.

Genomic DNA analysis

Peripheral blood was drawn from each patient and collected in an EDTA-containing tube. Genomic DNA was extracted from the peripheral blood using the salting-out method.

Genotyping of the −460 C/T (rs833061), +405 G/C (rs2010963), and +936 C/T (rs3025039) polymorphisms were assessed using polymerase chain reaction–restriction fragment length polymorphism (PCR– RFLP) analysis, as previously reported with minor modifications [Watson et al. 2000].

The primers used for amplification of VEGF −460C/T, +405 G/C, and +936C/T polymorphisms were listed in Table 1. Each PCR was carried out in a total volume of 25 µl containing 100 ng genomic DNA, 1 x PCR- Buffer (1.5 mM MgCl2), 0.5 mM of a dNTP-Mix, 0.5 Units of Taq DNA polymerase and 3 pmol of each primer. The amplification program consisted of an initial denaturation step at 95°C for 15 min and 35 cycles (denaturation at 95°C for 30 s, annealing for 30 s at 62°C for −460 C/T, 57°C for +405 G/C, and 66°C for +936 C/T, extension at 72°C for 30 s and final extension for 10 min at 72°C). The PCR products were digested at 37°C overnight with 3 IU of the appropriate restriction enzymes (Fermentas, Canada) listed in Table 1. Digested fragments were analyzed by 3% agarose gel electrophoresis followed by ethidium bromide staining and ultraviolet visualization.

For quality control, 10% of the samples were selected at random for repeated genotyping and concordance was 100%. Samples with ambiguous results were repeated.

Measurement of VEGF serum levels

The blood samples (5–10 ml) were collected from all study participants and centrifuged at 3000 rpm for 10 min at 4°C. The serum was collected, aliquoted, and stored at −80°C until analysis. VEGF serum levels were measured by a sandwich enzyme immunoassay (EIA) according to the manufacturer’s instructions (Neogen Corporation, Lansing, MI, USA). The EIA was reacted by using the mouse monoclonal antibody against human VEGF and then treated with biotinylated rabbit anti-human VEGF polyclonal antibodies to detect the VEGF in the serum sample. The assay was visualized with a streptavidin alkaline phosphatase conjugate and an ensuring chromogenic substrate reaction. Details of the assay were described elsewhere [Tsai et al. 2007; Hamzaoui et al. 2009].

Statistical analysis

Hardy–Weinberg equilibrium was tested in cases and controls separately. Association analysis was performed using standard Chi-squared test (Epistat Statistical Package, Epi Info Version 7) to detect differences in genotype and allele distribution among our groups. The strength of a gene association is indicated by the odds ratio (OR) with a 95% confidence interval (CI).

The haplotype frequencies were estimated using Thesias software v.3.1 which implements the stochastic-EM (expectation–maximization) algorithm. The degree of linkage disequilibrium was determined by using the normalized difference D′. A P value less than 0.05 was considered statistically significant. The OR and the 95% CI were calculated whenever applicable.

Abbreviations
ASRM	=	American Society for Reproductive Medicine
CI	=	confidence intervals
OR	=	odds ratio
PCR-RFLP	=	polymerase chain reaction restriction fragment length polymorphism
SNP	=	single nucleotide polymorphism
UTR	=	untranslated region
VEGF	=	vascular endothelial growth factor

Acknowledgments

This study was supported by a grant from the Ministry of Science and Technology and the Ministry of Health of Tunisia. Many thanks to Professor A. Hamzaoui, head of the unit research UR12SP15.

Declaration of interest

The authors declare no conflicts of interest.

Author contributions

Conceived and designed the experiments: HB, NA, KS; Performed the study, analyzed and interpreted the data: HB, KW; Collected the samples: HB, ZA; Wrote the manuscript: HB, HK; Revised the draft: HSF, HK.