Background
Essential hypertension (EH) is a multifactorial disease caused by various environmental and genetic factors. The estimated genetic contribution to EH ranges from 25% to 60%, and high blood pressure occurring before the age of 55 years of age is 3.8 times more frequent in individuals having two or more first-degree relatives with high blood pressure (BP) [
1], indicating that EH has a strong genetic component.
The prevalence of Hypertension continues to rise despite recent advances in diagnosis and treatment. Approximately 40% of the global adult population aged 25 and above suffer from hypertension [
2]. In Sudan, the prevalence of hypertension was reported as 20% in 2011 [
3]. Hypertension is a major contributor to the growing global pandemic of CVD and stroke.
Nitric oxide (NO) is important for the anatomical and functional integrity of the vascular endothelium, which is essential for the prevention of atherosclerosis, hypertension, and other CVDs [
4]. NO mediates its protective effect on the endothelium by a number of mechanisms including regulation of vasodilatation (either flow-dependent or receptor-mediated vasodilatation), inhibition of leukocyte adhesion to vessels, inhibition of platelet aggregation, and control of muscle cell proliferation [
5‐
7].
NO is produced in endothelial cells by the enzyme, endothelial nitric oxide synthase (eNOS), which catalyzes the conversion of the amino acid arginine to NO and citrulline [
8]. Hypertensive patients have reduced levels of NO production, manifested as low levels of urinary and serum nitrate [
9].
The gene encoding eNOS,
NOS3, has been mapped to human chromosome 7q36 and spans approximately 23 kilobases of the genome [
10].
NOS3 represents an interesting candidate gene in relation to EH. The association between
NOS3 and EH has been widely studied, and the disruption of the gene leads to hypertension in mice [
11].
Several polymorphic variations of
NOS3 have been identified and investigated. The most clinically relevant
NOS3 variants are rs1799983 (G894 T; Glu298Asp); a variable number tandem repeat (VNTR) in intron 4; and rs2070744 (T-786C) in the promotor region. These variants are associated with CVD, including coronary artery disease [
12], myocardial infarction [
13,
14], hypertension [
15,
16], and stroke [
17].
Additional knowledge of NOS3 gene polymorphisms and their role in hypertension will improve understanding of EH, its common predisposing factors, and potential treatment options. The potential contribution of NOS3 polymorphisms to the development of hypertension in Sudan has received no attention to date, and no previous studies have addressed this subject. This study aimed to investigate the association of the three polymorphisms in the NOS3 gene with EH in the Sudanese population.
Methods
Patients and control samples
This case control study was conducted in Khartoum, and samples were collected between February 2014 and February 2015. Hypertensive patients (
n = 260) were enrolled in the study from Samir Health Centre, Soba Teaching Hospital and Fath El Rahman El Bashir Referral Clinics in Khartoum. Patients were selected according to the following criteria: (1) age ≥ 18 years; and (2) established hypertension, defined either by chronic therapy or by blood pressure ≥ 140/90 mmHg according to the National Institute of Health and Care Excellence guidelines [
18]. Patients with the following criteria were excluded from the study: (1) any secondary hypertension (excluded by history, clinical examination, creatinine levels in plasma, and urine testing for albuminuria); and (2) evidence of inflammatory processes assessed by the presence of two of the following: tachycardia, hypotension, tachypnea, and high or low temperature [
19]. Plasma levels of C-reactive protein were also measured.
Controls (n = 144; age ≥ 18 years, blood pressure < 140/90, and without evidence of disease) were recruited mainly from Faculty of Medicine - University of Khartoum and other different institutes in Khartoum and volunteered to participate in the study.
As patients and control groups differed significantly in their age (patients were much older), we omitted all participants with age less than 30 and more than 65 years old. Thereby, 157 patients and 85 controls were included in the statistical calculations.
EH patients and normotensive individuals were consulted about their willingness to participate in the study, and written consent was obtained. Each study participant was interviewed about demographic data, duration of hypertension (for hypertensive patients), family history of hypertension, risk factors for hypertension, medication taken for EH, other medication, smoking, alcohol consumption, and other EH-associated chronic diseases, such as diabetes mellitus, MI, renal diseases, hypercholesterolemia, and stroke, using a structured questionnaire (Additional file
1).
Blood pressure was measured on two occasions in a quiet room after 15 min of resting in a supine position, using a recently calibrated sphygmomanometer.
DNA sample preparation
Venous blood (5 mL) was drawn from each subject. DNA extraction and genotyping were performed at the Institute for cardiogenetics, University of Lubeck, Germany. DNA extraction was performed using a Qiagen Gentra Puregene Blood Kit (Qiagen, Germany).
NOS3 variants and genotyping
rs1799983 (G894 T; Glu298Asp) is a missense variant in exon 7 of NOS3. Two alleles of the VNTR in intron 4 have been identified, the larger allele (b) consists of five tandem 27 bp repeats (GAAGTCTAGACCTGCTGCAGGGGTGAG), and the smaller allele (a) consists of four such repeats. Two other rare alleles have been reported, c with six, and d with three, 27 bp repeats. rs2070744 (T-786C) is a polymorphism in the promoter region of NOS3. Those three polymorphisms are the best studied in the NOS3 gene, so they were chosen for genotyping in this study. The literature review done to select these polymorphisms was conducted in 2011. Therefore, recent polymorphisms in NOS3 associated with EH were not included.
DNA genotyping was performed using TaqMan and polymerase chain reaction restriction fragment length polymorphism (PCR-RFLP) assays. rs1799983 were genotyped using PCR-RFLP as the samples were undetermined with the Taqman method possibly due to multiple variations within the sequence. Two hundred nineteen and 15 samples were genotyped for rs2070744 using Taqman and PCR-RFLP respectively. Genotyping of the VNTR was performed only by PCR-RFLP.
TaqMan assays were performed using a 7900HT Fast Real-Time PCR System (Applied Biosystems, USA), and Sequence Detection System (SDS) software was used to call alleles based on florescence measurements. Genotyping was run in duplicates and the duplicate concordance rate was 100%.
PCR reaction mix (for PCR-RFLP) of 10 μL was prepared. The touch down 61 program was used for the PCR in a Sensoquest thermocycler (Sensoquest GmbH, Germany).
The resulting amplification products were incubated with the restriction enzymes, BanII and MboI for the rs1799983 SNP, and MspI for the rs2070744 polymorphism. No restriction enzyme was used for the VNTR. Next, the products were mixed with loading dye and SYBR green mix and run in 1.2% (rs1799983 and rs2070744) and 1.8% (VNTR) agarose gels. Estimation of product sizes was carried out with peqGOLD 100 bp DNA ladder (PeqLab, Germany). A Bio-Rad gel documentation system (Bio-Rad Laboratories Inc., USA) was used for gel image capture. The results of the RFLP were determined by two persons reading the gels and in case they disagree about the result, genotyping was repeated. The results of the BanII and MboI restriction were consistent and concordant.
Primers, restriction enzymes and fragment lengths are presented in Table
1. The primers for rs1799983 and rs2070744 were designed using Primer 3 software [
20] and ordered from Eurofins Genomics (Germany). For the VNTR assay (alleles a and b), primers mentioned in the literature were used [
21]. For those samples not containing alleles a and b, PCR reactions were cleaned up using NucleoSpin® gel and PCR Clean-up kits (MACHEREY-NAGEL, Germany). Then, amplicons were cloned into the PCR 2.1-TOPO® plasmid vector and recombinants were transformed into competent
Escherichia coli cells, according to instructions of the TOPO® TA Cloning® kit (Invitrogen, Life Technologies, Germany). Next, PCR was performed using bacterial colonies directly as template. PCR products were visualized by agarose gel electrophoresis, and samples with different lengths were cleaned-up for sanger sequencing (Seqlab company, Germany). The FASTAQ files of the sequencing were aligned with a reference sequence of
NOS3 from Ensemble [
22].
Table 1
Primers, restriction enzymes, and fragment lengths of the major and minor alleles of rs1799983, rs2070744, and VNTR polymorphisms of NOS3 gene
Left primer (5′ to 3′) | AGCCTCGGTGAGATAAAGGA | AGGCCCTATGGTAGTGCCTT | CCCCTGTGGACCAGATGC |
Right primer (5′ to 3′) | TCTTGAGAGGCTCAGGGATG | TCTCTTAGTGCTGTGGTCAC | ACATTAGGGTATCCCTTCC |
Product size | 368 bp | | 379 bp |
Restriction enzyme |
BanII |
MboI | |
MspI |
Major allele fragment length (bp) | Cut, two fragments (251 and 117 bp) | Not cut | 420 (b) | Two fragments, 146 and 233 bp |
Minor allele fragment length (bp) | Not cut, 368 bp | Cut, two fragments (246 and 122 bp) | 394 (a) | Three fragments 46, 146, and 187 bp |
Heterozygous | | | 394 and 420 bp fragments | Four fragments 46, 146, 187, and 233 bp |
Statistical analysis
Statistical analyses were performed using Statistical Package for the Social Sciences for Windows (SPSS, version 12.0) software. Genotype and allele frequencies were compared between groups by χ2 test, and odds ratios (OR) with 95% confidence intervals (CIs) were calculated. Differences in genotype distributions under autosomal dominant and recessive models were also determined using χ2 test, and ORs were calculated. Multinomial and binary logistic regressions were performed in genotype comparisons with age, gender, BMI and smoking included as covariates. A χ2 test with one degree of freedom was used to determine deviation of genotype distributions from Hardy-Weinberg equilibrium. Values of P < 0.05 were considered statistically significant. To eliminate any confounding factors affecting the genetic association study, we also analyzed the distribution of genotypes according to the clinical and demographic characteristics of both groups, including age, gender, BMI, smoking, systolic and diastolic BP, and the presence of complications alongside hypertension (e.g., diabetes, MI, renal failure, hypercholesterolemia, and stroke), using ANOVA for continuous data and χ2 or Fisher’s Exact tests for categorical data. Linkage disequilibrium between the three polymorphisms was examined by χ2 analysis. The extent of disequilibrium was expressed as D′ = D/D
max and Pearson squared correlation coefficient (r2).
Discussion
In this study, three polymorphisms in the NOS3 gene were genotyped in hypertensive and control groups, to study their association with EH in the Sudanese population.
Regarding rs1799983, the observed minor T allele frequency (MAF) of 0.17 was similar to the global minor allele frequency at this SNP reported by the 1000 genomes project (0.18), although it differed from the reported frequency in the African population (0.07) [
23]. Our results were also inconsistent with a previous report by Thomas et al., indicating that the T allele is absent in an African population from Mali [
24]. A higher T allele frequency was found in the hypertensive group. However, the difference was not significant. This negative result may indicate a geographic difference within Africa in the distribution of alleles. Many investigations have examined associations between the rs1799983 variant and EH. However, the results have been controversial and inconclusive. Some studies have identified a higher T allele frequency in hypertensive patients and have reported that this allele is associated with resistance to conventional therapy [
15,
25]. In contrast, studies in Caucasian populations indicated a higher G allele frequency in the hypertensive group and an association of this allele with the outcome, all-cause mortality [
26,
27]. These discrepancies may indicate that either another mutation or SNP is linked to either of the two alleles, or that the observed associations are due to random errors. Conversely, other studies report a lack of evidence for linkage between this polymorphism and EH in the Japanese [
28] and Australian [
29] populations.
The cause of the association between the rs1799983 variant and CVDs is not well characterized; however, it has been suggested that the resulting replacement of Glu with Asp results in changes to the structure of the eNOS enzyme, increasing its susceptibility to proteolysis [
30]. It has also been suggested that this substitution affects the interaction of the enzyme with caveolin-1, thereby affecting its localization and activity [
31]. The GG genotype is associated with increased eNOS activity and higher NO levels that are toxic to cells, due to consequent increased superoxide anion production [
32,
33].
In the current study, no direct relationship was found between the VNTR and EH. Our results indicate that three hypertensive patients (1%) and five controls (3%) carried the c allele at this locus, which was previously reported by Thomas et al. to be present in Africans and African Americans, but not in Caucasians [
24]. In their report, the investigators indicated that this allele is rare; however, our results indicate that it is not rare (>1%) in the Sudanese population [
34].
Associations of the a allele of the VNTR in intron 4 of
NOS3 with coronary artery disease and renal disease have been reported [
35‐
37]. However, conflicting data have appeared in the literature concerning the association between this variant and hypertension, even within the same population [
15,
38‐
40]. Moreover, the effect of the a allele on eNOS expression is controversial, with studies reporting both reduced enzyme activity leading to lower NO levels [
41] and increased levels of NO production [
42] associated with this variant.
The lack of association between either the rs1799983 or VNTR with EH in the Sudanese population could be due either to the small sample size in this study or a genuine indication of a lack of association between these variants and EH in the Sudanese population.
In the present study, the hypertensive group had a significantly higher frequency of CC genotype of the rs2070744 polymorphism compared to the control group (
p = 0.02). The patients group also had higher frequency of TC + CC genotypes compared with the TT genotype considering a dominant model of inheritance (
P = 0.04; OR (95% CI) = 0.639 (0.420–0.973)) (Table
4). In addition, the C allele was more frequent in the patients group (
p = 0.03). There was no association between this genotype and the confounding factors, age, sex, BMI, smoking status, presence of diabetes, MI, stroke, renal failure, or hypercholesterolemia. The observed minor C allele frequency of 0.26 is similar to the global MAF reported by the 1000 genome project (0.23) but different from the MAF for the African population at this SNP (0.14) [
23]. The substitution of T to C in this variant at
NOS3 is associated with coronary spasm [
43], enhanced coronary vasoconstriction in response to acetylcholine, and impaired endothelium-independent vasodilatation in Caucasians [
44]. A study in a Canadian population showed that the CC genotype is associated with higher systolic BP in a healthy cohort and carriers of this allele have a relative risk of developing hypertension of 2.16 (95% CI, 1.3–3.7) compared with non-carriers [
45]. In contrast, a study in a Japanese population did not detect any association of this variant with EH [
46]. This T to C substitution in the promoter region of
NOS3 reduces its rate of transcription by 50%, both under baseline conditions and in response to hypoxia, and is associated with decreased serum levels of nitrite/nitrate [
41,
43]. These effects may be a reflection of the fact that the minor allele can be bound by replication protein A1, which acts as a transcription repressor protein [
47], ultimately leading to decreased levels of NO and endothelial dysfunction.
This study has number of limitations that are worth mentioning. The small samples size with potential lack of power to the study is a limitation. Furthermore, cases and controls were significantly different in their age, gender, BMI and smoking status. However, statistical analyses showed that none of these variables had an effect on any of the genotypes distribution.
In this study, we found that all three included
NOS3 polymorphisms are in linkage disequilibrium, and that this effect was strongest between rs1799983 and the VNTR in intron 4, which is expected as they are in close physical proximity to one another. There was also weaker linkage between (rs1799983 and rs2070744), and (rs2070744 and the VNTR). It is unlikely that the VNTR itself has a functional role, as it is in an intronic region. However, it has been suggested to act as a marker for other functional variants elsewhere in the gene. Nakayama et al. first reported that the VNTR is in linkage disequilibrium with rs2070744; hence, the effect of VNTR on
NOS3 mRNA expression, eNOS protein concentration, and enzyme activity is likely to be mediated by the differences in transcriptional efficiency associated with the rs2070744 polymorphism [
48‐
51]. The degree of linkage disequilibrium between the three polymorphisms may help in understanding of the evolutionary divergence of the
NOS3 gene.