Sex-Specific Effects of NLRP6/AVR and ADM Loci on Susceptibility to Essential Hypertension in a Sardinian Population

Coronary artery disease, heart failure, fatal arrhythmias, stroke, and renal disease are the most common causes of mortality for humans, and essential hypertension remains a major risk factor. Elucidation of susceptibility loci for essential hypertension has been difficult because of its complex, multifactorial nature involving genetic, environmental, and sex- and age-dependent nature. We investigated whether the 11p15.5 region syntenic to rat chromosome 1 region containing multiple blood pressure quantitative trait loci (QTL) detected in Dahl rat intercrosses harbors polymorphisms that contribute to susceptibility/resistance to essential hypertension in a Sardinian population. Initial testing performed using microsatellite markers spanning 18 Mb of 11p15.5 detected a strong association between D11S1318 (at 2.1 Mb, P = 0.004) and D11S1346 (at 10.6 Mb, P = 0.00000004), suggesting that loci in close proximity to these markers may contribute to susceptibility in our Sardinian cohort. NLR family, pyrin domain containing 6/angiotensin-vasopressin receptor (NLRP6/AVR), and adrenomedullin (ADM) are in close proximity to D11S1318 and D11S1346, respectively; thus we tested single nucleotide polymorphisms (SNPs) within NLRP6/AVR and ADM for their association with hypertension in our Sardinian cohort. Upon sex stratification, we detected one NLRP6/AVR SNP associated with decreased susceptibility to hypertension in males (rs7948797G, P = 0.029; OR = 0.73 [0.57–0.94]). For ADM, sex-specific analysis showed a significant association between rs4444073C, with increased susceptibility to essential hypertension only in the male population (P = 0.006; OR = 1.44 [1.13–1.84]). Our results revealed an association between NLRP6/AVR and ADM loci with male essential hypertension, suggesting the existence of sex-specific NLRP6/AVR and ADM variants affecting male susceptibility to essential hypertension.


Introduction
Essential hypertension is a highly prevalent disorder and remains a major risk factor for the most common causes of mortality, including coronary artery disease, heart failure, fatal arrhythmias, stroke, and renal disease [1,2]. As a complex, multifactorial disorder, elucidation of susceptibility loci remains difficult. Previous studies have emphasized the challenges of genetically analyzing essential hypertension [3][4][5]. Genomewide association studies have failed to detect major hypertension susceptibility genes that may contribute greater than 5 mmHg to the blood pressure (BP) effect [6][7][8][9]. These studies reported that several loci were significantly associated with increased BP by analyzing tens of thousands of patients in a multi-center, meta-analysis paradigm with BP effects ranging from 0.5-1 mmHg [6][7][8][9]. This implies that either hundreds of hypertension susceptibility genes exist that account for essential hypertension in humans, with each locus contributing a small fraction of the increase in BP (0.5-1 mmHg effect), or the studies have failed to detect major hypertension susceptibility loci due to major confounders, such as inherent genetic heterogeneity of human populations, great variability in trait measurements, the existence of several factors that are not accounted for, and factors in the analytical paradigm (e.g., sex-specific effects, gestational risk factors, and presence of hypertension subtypes with unique pathogenetic mechanisms).
For complex multifactorial diseases with clinical heterogeneity such as hypertension, genetic studies of inbred rat models of polygenic (essential) hypertension are instrumental for identifying BP-quantitative trait loci (QTL) and candidate susceptibility genes for subsequent testing in human essential hypertension. We have successfully employed this approach to identify ATP1A1 (α1 Na,K-ATPase) [10,11] and DEspR (dual endothelin1/vascular endothelial growth factor signal peptide receptor) [12,13] as candidate hypertension genes in a Dahl salt-sensitive rat model of polygenic hypertension. Subsequent studies detected an association of ATP1A1 [14,15] and DEspR [15] with essential hypertension in humans and revealed a functionally significant DEspR T/ CATAAAA-box promoter variant associated with 7.7 mmHg in increased systolic BP in a male Sardinian population [16]. We recently reported two closely-linked BP QTLs on chromosome-1 (chr1) (BP-m1 at 144.3 Mb and BP-m2 at 208.8 Mb) affecting salt-sensitive hypertension in Dahl rats [12]. These two chr1-BP QTLs contain some candidate genes, which has been supported by experimental evidence. Briefly, molecular genetic evidence demonstrates that a functionally significant N119S/C163R variant of the angiotensinvasopressin receptor (NLRP6/AVR at 201.08 Mb), which exhibits sodium-induced dysfunction [17], may underlie chr1 BP-m1 QTL of salt-sensitive hypertension in F2 (Dahl S x Dahl R]-intercross male rats [12]. Another candidate gene in this chr1-BP QTL region includes adrenomedullin (ADM; at 168.38 Mb), which has been implicated in hypertension pathogenesis [18,19] and may underlie BP-m2 QTL [12].
Analyses of BP-m1 and BP-m2 corresponding syntenic regions in humans localized this chromosomal segment to the 11p15.5 region. We therefore investigated whether this 11p15.5 region harbors polymorphisms contributing to susceptibility/resistance to essential hypertension in our Sardinian population.

Association between 11p15.1-11p15.5 region and essential hypertension
To enhance robustness of the population under study, we ascertained first a limited genetic diversity by restricting the cohort under analysis to a relatively isolated genetic population of northern Sardinia [20,21] and second , we minimized subtype heterogeneity by focusing on the extreme of the population to distinguish hypertensive patients and normotensive controls. Thus, we compared ascertained hypertensive patients (n = 433) with group means for systolic BP (SBP) of 174.4 ± 14.7 mmHg and diastolic BP (DBP) of 110.5 ± 9.9 mmHg against control normotensives (n = 279) with group means for SBP of 127.6 ± 11.3 mmHg and DBP of 77.6 ± 7.2 mmHg (Table 1). Both groups contained equivalent representation of both sexes (Table 1).
Initial testing of possible loci associated with essential hypertension within the 11p15.1-11p15.5 region was performed with microsatellite markers spanning 18 Mb of this chromosome 11 region. Since microsatellite markers are highly variable, they typically reveal a significant amount of information; therefore, they may be useful for examining large genomic segments in linkage and/or association studies [22,23]. As shown in Table 2, the D11S1318 and D11S1346 microsatellite markers showed a strong association with hypertension susceptibility after adjusting for multiple testing (D11S1318, P = 0.004; D11S1346, P = 0.00000004). This suggests that loci in closed proximity to D11S1318 and D11S1346 contribute to hypertension susceptibility in our Sardinian cohort.

Single-point association analysis of NLRP6/AVR and ADM loci
Closed examination of genes in the vicinity of D11S1318 and D11S1346 identify NLRP6/AVR (at 278570) and ADM (at 10327248) as candidates for the corresponding associated regions marked by D11S1318 and D11S1346, respectively ( Table 2). To assess the possible association of NLRP6/AVR and ADM with essential hypertension in our Sardinian cohort, we examined single-point associations between NLRP6/AVR and ADM SNPs ( Figure 1, Table 3) with hypertension susceptibility. We used 4 tagged SNPs for the NLRP6/AVR  Table 3. None of the tagged SNPs deviated significantly from Hardy-Weinberg equilibrium in both hypertensive and normotensive cohorts ( Table 3). One of the NLRP6/AVR SNPs, rs7937440, demonstrated a significant association with hypertension in the total cohort after adjustment for multiple testing (P = 0.032, Table 4). Upon sex stratification, one NLRP6/AVR SNP (minor allele rs7948797G, Table 4) was associated with decreased susceptibility to hypertension only in the male population (P = 0.029; OR = 0.63 [0.45-0.88], Table 4). For ADM, sex-specific analysis detected a significant association between rs4444073C minor allele and increased susceptibility to hypertension exclusively in the male cohort as well (P = 0.013; OR = 1.62 [1.14-2.29], Table 4).

Discussion
Identifying genes underlying susceptibility to essential hypertension has been difficult. Genome-wide association studies have failed to detect significant associations contributing greater than 5 mmHg in blood pressure despite the large number of subjects examined in different studied cohorts [6][7][8][9]. A number of reasons may account for these negative findings, including intrinsic genetic heterogeneity of human populations, differential accuracy and/or modality in trait measurements (blood pressure), exclusion of putative sexspecific effects on the phenotype in the analytical paradigm, and gestational risk factors.  Here, we present evidence suggesting that the NLRP6/AVR and ADM loci contribute to hypertension susceptibility in a Sardinian population. We first detected strong association of the 11p15.1-11p15.5 region with hypertension showing two distinct peaks marked by D11S1318 and D11S1346 SSLP markers. Their chromosomal location indicated NLRP6/AVR and ADM as putative loci underlying the observed associations. Subsequent analysis using SNPs spanning both the NLRP6/AVR and ADM transcription units confirmed their associations with essential hypertension in our Sardinian cohort. Both loci demonstrated sex-specific effects on hypertension susceptibility and were primarily associated with male essential hypertension.
ADM is a potent vasodilatory peptide important for blood pressure homeostasis, as well as cardiovascular and renal function [18,24,25]. Its role in blood pressure regulation has been substantiated by showing that heterozygous ADM +/knockout male mice exhibited elevated blood pressure when compared with wild-type littermates [26]. The ADM locus has been detected as a genome-wide significant finding in a recently reported study involving a large population [8]. In addition, the ADM SNP rs4399321 utilized in our study has been reported to be associated with proteinuria in subjects with essential hypertension [27] and with BP levels in normotensive subjects in a Chinese population [28] supporting a regulatory role for ADM in blood pressure homeostasis in humans. Moreover, recent studies have implicated ADM in hypertension pathogenesis in humans. A dose-response relationship between ADM and BP status was observed among age-and sex-matched normotensive, pre-hypertensive, and hypertensive subjects, detecting lower ADM levels in the hypertensive group than in the pre-hypertensive and normotensive groups [19]. Moreover, ADM levels were found to be associated with mean arterial pressure in men but not in women in the Framingham cohort, thus indicating sexspecificity [29]. This observation is consistent with our results showing that ADM is associated with hypertension susceptibility only in the male Sardinian population.
NLRP6/AVR has been linked to salt-sensitive hypertension in Dhal rats. Genetic analysis in an F2 (Dahl S × Dahl R)intercross rat population delineated NLRP6/AVR as a candidate gene underlying a chromosome 1 BP-m2 QTL [12]. BP-m2 has shown sex-specificity since it has been detected only in the F2 male population and not in females [12]. Additionally, functional androgen and estrogen response elements have been delineated within the NLRP6/AVR 5′regulatory region that may contribute to the sex-specific effects of NLRP6/AVR [30]. Moreover, an NLRP6/AVR structural variant has been identified in Dahl S rats exhibiting sodiuminduced dysfunction affecting ligand binding and hormonedependent signal transduction [17]. This supports a pathogenetic role of NLRP6/AVR in salt-sensitive hypertension in Dhal rats. The role of NLRP6/AVR in blood pressure regulation is further supported by the recent findings showing that NLRP6/AVR deficiency affects urinary concentrating ability and lowers blood pressure in mice [31]. Our data are consistent with the hypothesis that NLRP6/AVR contributes to hypertension susceptibility in humans.
In conclusion, our results revealed an association between NLRP6/AVR and ADM loci and hypertension susceptibility in a northern Sardinian population, primarily affecting male essential hypertension. Specific ADM and NLRP6/AVR variants affecting susceptibility to high blood pressure require further investigation.

Ethics Statement
This study was performed in strict accordance with the principles expressed in the Declaration of Helsinki. The protocol was approved by the local ethics committee of Local Health Unit-University of Sassari Medical School. Written informed consent was obtained and all clinical investigation was conducted according to the principles expressed in the Declaration of Helsinki.

Study population
The study cohort comprised 712 subjects, with 433 patients with essential hypertension and 279 normotensive controls enrolled at the Hypertension Center of the University of Sassari Medical School. All subjects were white, unrelated, born in different domains of North Sardinia previously ascertained to have a high degree of genetic homogeneity [20,21], ascertained to be Sardinian for at least 6 generations, and resided in Sardinia. Hypertensive patients with BP > 160/95

Genotyping
Genotyping of microsatellite markers was performed as described previously using 20 ng of DNA [11,14]. Alleles were identified using 6% non-denaturing polyacrylamide gel electrophoresis (1/2× Tris-Borate-EDTA, 750 V for 14 h). Gels were exposed to autoradiograms for genotype readings. SNP genotyping was carried out by the Molecular Genetics Core Facility at the Boston University School of Medicine on a Life Technologies 7900 Real-Time PCR System (Foster City, CA, USA). Chromosomal position of microsatellite markers and single nucleotide polymorphisms (SNPs) were based on the "NCBI build 37.1-Map". SNP assays (TaqMan assays) were procured from Life Technologies. The genotyping completeness rate was 87%.

Statistical analysis
Analysis of microsatellite allele frequencies between cases and controls was performed using χ 2 analysis (Sigma Plot 11.0) for D11S1318 2x9 χ 2 analysis (9 alleles), D11S1338 2×3 χ 2 analysis (3 alleles), D11S1323 2×4 χ 2 analysis (4 alleles), D11S1346 2×5 χ 2 analysis (5 alleles), D11S1315 2×5 χ 2 analysis (5 alleles), and D11S1310 2×3 χ 2 analysis (3 alleles). Bonferroni adjustment was applied to nominal P values for multiple testing corrections. For SNPs, single point association analysis comparing cases and control subjects was conducted using the HelixTree TM SNP & Variation Suite genetic analysis software (version 6.4.3, Golden Helix Inc., Bozeman, MT, USA). A basic allelic test (D vs d) was implemented using a Chi-squared test as statistical method, obtaining as well odds ratios with corresponding confidence limits. Missing genotypes were not included (imputed) in the association analysis. The false discovery rate was utilized for multiple testing corrections. A chi-squared test using our cohort with α = 0.005 predicts statistically significant detection of at least 0.1 in allele frequency differences between groups of 433 hypertensives and 279 normotensive subjects with a Power > 0.8.