Introduction

Hypertension is a complex multifactorial disorder that is thought to result from an interaction between an individual's genetic background and various environmental factors.1 Given that hypertension is a major risk factor for coronary heart disease, stroke and chronic kidney disease, personalized prevention of hypertension is an important public health goal. An approach to personalized prevention of, and selection of the most appropriate treatment for, hypertension is to identify genes that confer susceptibility to this condition. Although genetic linkage analyses,2, 3, 4, 5 genome-wide mapping with microsatellite markers,6 genome-wide association studies7 and candidate gene association studies8, 9, 10, 11, 12 have implicated various loci and genes in predisposition to hypertension, the genes that confer susceptibility to this condition remain to be identified definitively. In addition, ethnic divergence of lifestyle and environmental factors as well as of genetic background necessitates examination of polymorphisms related to hypertension in each ethnic group.

With the use of a candidate gene approach, we have been attempting to identify genes associated with blood pressure (BP) and the prevalence of hypertension in Japanese women and men recruited to a population-based prospective cohort study. In this study, we have selected 22 candidate gene polymorphisms that might be expected to contribute to the regulation of BP and the development of hypertension (Table 1) and have examined the relationships of these polymorphisms with systolic and diastolic BP and the prevalence of hypertension, even though there is no apparent biological link among these genes. Our aim was to identify one to two genetic markers significantly associated with hypertension for each gene. Among the 22 polymorphisms examined in the study, the 682A → G (Thr228Ala) polymorphism of SORBS1, the −30G → A polymorphism of GCK and the 2364A → G polymorphism of WISP1 were significantly associated with hypertension in community-dwelling Japanese individuals. Given that among the various polymorphisms identified earlier, these polymorphisms were shown to be related to phenotypes,13, 14, 15 they might be expected to affect gene function. We now show the relationships of the polymorphisms of SORBS1, GCK and WISP1 with BP and the prevalence of hypertension in community-dwelling Japanese individuals.

Table 1 The 22 gene polymorphisms examined in this study
Full size table

Methods

Study population

The National Institute for Longevity Sciences—Longitudinal Study of Aging is a population-based prospective cohort study of aging and age-related diseases, the details of which have been described earlier.16, 17, 18, 19 We examined the relationships of genetic variants with BP and the prevalence of hypertension in 2233 individuals (1106 women, 1127 men) recruited to the National Institute for Longevity Sciences—Longitudinal Study of Aging. Individuals whose genotypes were not successfully determined were excluded from the analysis. The study protocol complies with the Declaration of Helsinki and was approved by the Committee on Ethics of Human Research of the National Center for Geriatrics and Gerontology. Written informed consent was obtained from each subject.

Measurement of BP

Blood pressure was measured with an automatic sphygmomanometer (BP-203RV-II; Colin, Tokyo, Japan) in subjects who had rested in the sitting position for at least 15 min. BP in each subject was confirmed by measurement with a mercury manometer performed by a physician according to the guidelines of the American Heart Association.20 Hypertension was defined as a systolic BP of ⩾140 mm Hg, a diastolic BP of ⩾90 mm Hg or the taking of antihypertensive medication. Normal BP was defined as both a systolic BP of <140 mm Hg and a diastolic BP of <90 mm Hg without the taking of antihypertensive medication.

Genotyping of polymorphisms

Genotypes for polymorphisms of SORBS1, GCK and WISP1 were determined by melting curve analysis (intercalater-mediated fluorescence resonance energy transfer probe method) (Table 2). The polymorphic region of each gene was amplified by PCR in a reaction mixture (25 μl) containing 20 ng of DNA, 5 pmol of each primer, 0.2 mmol l−1 of each deoxynucleoside triphosphate, 2 mmol l−1 (for SORBS1 and GCK) or 3 mmol l−1 (for WISP1) MgCl2 and 1.25 U of rTaq DNA polymerase (Toyobo, Osaka, Japan) in polymerase buffer. The amplification protocol comprised initial denaturation at 95°C for 5 min; 45 cycles of denaturation at 95°C for 30 s, annealing at 65°C for 30 s and extension at 72°C for 30 s; and a final extension at 72°C for 2 min. A mixture (2 μl) of 10 pmol of probe labeled at the 5′-end with Texas red and a 1:400 dilution of SYBR Green I was added to the PCR products, which was then transferred to a Prism 7700 instrument (Applied Biosystems, Foster City, CA, USA) for measurement of melting temperature. The program for analytical melting comprised incubation at 95°C for 30 s, 40°C for 1 min and temperatures increasing to 80°C over 10 min. The fluorescence signals were detected at excitation and emission wavelengths of 485 and 612 nm, respectively.

Table 2 Primers, probes and other PCR conditions for genotyping
Full size table

Statistical analysis

Age, body mass index, and systolic and diastolic BP were compared between subjects with hypertension and controls by the unpaired Student's t-test, and the prevalence of smoking was compared between the two groups by the χ2 test. BP values were analyzed in individuals who were not taking antihypertensive drugs. We examined the effects of genetic variants of SORBS1, GCK and WISP1 on the prevalence of hypertension and on systolic and diastolic BP based on a 6-year longitudinal cohort study. The data for examination of each subject in the first wave (November 1997 to April 2000), second wave (April 2000 to May 2002), third wave (May 2002 to May 2004) and fourth wave (June 2004 to July 2006) were pooled and analyzed. Systolic and diastolic BP and the prevalence of hypertension were evaluated for both sexes combined as well as for women and men separately. Longitudinal changes in the prevalence of hypertension were compared between two groups (dominant or recessive genetic model) by a generalized estimating equation,21 with adjustment for age and sex in all subjects or for age in women or men. Longitudinal changes in systolic and diastolic BP in individuals not taking antihypertensive medication were compared between two groups (dominant or recessive model) in a mixed-effect model,22 with adjustment for age and sex in all subjects or for age in women or men. Age-related changes in the prevalence of hypertension or in systolic or diastolic BP were estimated with quadratic curves controlling for the observation year during which the subjects attended at least one follow-up examination. Allele frequencies were estimated by the gene-counting method, and the χ2 test was used to identify significant departure from Hardy–Weinberg equilibrium. A P-value of <0.05 was considered statistically significant. Statistical analysis was performed with SAS software release 9.13 (SAS Institute, Cary, NC, USA).

Results

The baseline characteristics (first wave) of the 2233 study subjects are shown in Table 3. Age, body mass index, as well as systolic and diastolic BP were greater in subjects with hypertension than in controls for both women and men, whereas the prevalence of smoking was greater in controls than in hypertensive subjects.

Table 3 Baseline characteristics (first wave) of women and men with hypertension and corresponding controls
Full size table

The relationships of the three polymorphisms with the prevalence of hypertension were analyzed with a generalized estimating equation and adjustment for age and sex in all subjects or for age in women or men examined separately (Table 4). The 682A → G (Thr228Ala) polymorphism of SORBS1 and the −30G → A polymorphism of GCK were significantly associated with the prevalence of hypertension among women in a dominant model. The variant G allele of the SORBS1 polymorphism was a risk factor for hypertension, whereas the variant A allele of the GCK polymorphism was protective against this condition. The 2364A → G polymorphism of WISP1 was significantly associated with the prevalence of hypertension among men in a dominant model, with the variant G allele representing a risk factor for this condition. The genotype distributions of the three polymorphisms in control subjects were all in Hardy–Weinberg equilibrium. The relationships between the prevalence of hypertension and age analyzed longitudinally with a generalized estimating equation are shown in Figure 1. The prevalence of hypertension was greater in the combined group of all subjects with the AG or GG genotype of SORBS1 than in those with the AA genotype from 40 to 80 years of age (Figure 1a), was lower in the combined group of women with the GA or AA genotype of GCK than in those with the GG genotype (Figure 1b) and was greater in the combined group of men with the AG or GG genotype of WISP1 than in those with the AA genotype (Figure 1c).

Table 4 Relations of three polymorphisms to the prevalence of hypertension analyzed with a generalized estimating equation (first wave to fourth wave)
Full size table
Figure 1
figure 1

Longitudinal analysis of relationships between the prevalence of hypertension and age according to genotype for SORBS1 (AA vs. AG+GG) in women and men combined (a), for GCK (GG vs. GA+AA) in women (b) or for WISP1 (AA vs. AG+GG) in men (c), with a generalized estimating equation and adjustment for either age and sex (a) or age (b and c).

Full size image

The relationships of the three polymorphisms with systolic or diastolic BP in individuals not taking antihypertensive medication were analyzed with a mixed-effect model, with adjustment for age and sex in all subjects or for age in women or men examined separately (Table 5). The 682A → G (Thr228Ala) polymorphism of SORBS1 was related to diastolic BP for women in a recessive model, with the variant G allele being associated with increased BP. The −30G → A polymorphism of GCK was related to systolic and diastolic BP for women in a dominant model, with the variant A allele being associated with a reduced BP. The 2364A → G polymorphism of WISP1 was related to systolic and diastolic BP for men in a dominant model, with the variant G allele being associated with a higher BP. The relationships between systolic or diastolic BP and age analyzed longitudinally according to genotype for WISP1 in men with a mixed-effect model are shown in Figure 2. Systolic (Figure 2a) and diastolic (Figure 2b) BPs were greater in the combined group of men with the AG or GG genotype of WISP1 than in those with the AA genotype from 40 to 80 years of age.

Table 5 Relations of three polymorphisms to systolic and diastolic blood pressure (BP, mm Hg) in individuals not taking antihypertensive medication as analyzed with a mixed-effect model (first wave to fourth wave)
Full size table
Figure 2
figure 2

Longitudinal analysis of relationships between systolic (a) or diastolic (b) BP and age according to genotype for WISP1 (AA vs. AG+GG) in men with a mixed-effect model and adjustment for age.

Full size image

Discussion

The regulation of BP involves the integration of a variety of biological systems that control the structure and tone of the vasculature as well as the volume and composition of body fluid. It also involves the adaptation of these systems to constantly changing physiological needs.23 We have now examined the relationships of three candidate gene polymorphisms with the prevalence of hypertension and with systolic and diastolic BP in community-dwelling Japanese women and men. Our results show that the 682A → G (Thr228Ala) polymorphism of SORBS1 and the −30G → A polymorphism of GCK were related to both the prevalence of hypertension and BP in women and that the 2364A → G polymorphism of WISP1 was related to both these parameters in men. These observations thus suggest that SORBS1 and GCK are susceptibility loci for hypertension in Japanese women and that WISP1 is such a locus in men.

Sorbin and SH3 domain-containing 1 (SORBS1) is a human homolog of c-Cbl-associated protein (CAP),24 which is an important signaling molecule in insulin stimulation of glucose uptake in mouse adipocytes.24, 25, 26, 27 Phosphorylation of c-Cbl results in the dissociation of the c-Cbl–CAP complex from the insulin receptor and its translocation to a lipid raft domain of the plasma membrane. Subsequent interactions of the c-Cbl–CAP complex with proteins such as flotillin, Crkα, C3G and TC10 eventually leads to the translocation of vesicles containing the glucose transporter SLC2A4 (GLUT4) from the cytoplasm to the plasma membrane. The central role of the c-Cbl–CAP complex in the regulation of insulin-stimulated glucose uptake suggests that genetic variation of SORBS1 may be related to insulin resistance. Fourteen single-nucleotide polymorphisms have been identified in human SORBS1, among which, only 682A → G (Thr228Ala), which corresponds to a predicted phosphorylation site for mitogen-activated protein kinase, was found to be related to obesity and type 2 diabetes mellitus in Chinese adults, with the G allele being protective against these conditions.13 In contrast, the GG (Ala/Ala) genotype of the 682A → G (Thr228Ala) polymorphism of SORBS1 was associated with lacunar infarction in the Japanese population.28 The 682A → G (Thr228Ala) polymorphism (rs2281939) is located in exon 7 of SORBS1. A linkage disequilibrium (LD) block (standardized LD coefficient (r2)⩾0.3) containing this polymorphism includes ∼5 kb from exon 7 to the 3′ region of the gene (International HapMap Project, http://www.hapmap.org/index.html.ja). We have now shown that this polymorphism of SORBS1 was related to both the prevalence of hypertension and diastolic BP in Japanese women, with the G (Ala) allele representing a risk factor for hypertension. The risk allele for hypertension is thus consistent with that for lacunar infarction in the earlier study.28

Glucokinase is expressed in pancreatic β cells and hepatocytes, with its expression being controlled by two tissue-specific gene promoters.29 Pancreatic glucokinase serves as the sensor for glucose in the regulation of insulin secretion.30 Mutations of GCK account for 10–50% of cases of maturity-onset diabetes of the young.30 The −30G → A polymorphism of GCK was shown to be associated with reduced β-cell function and impaired glucose tolerance in Japanese population.14, 31 The −30G → A polymorphism (rs1799884) is located in the β-cell-specific promoter region of GCK. An LD block containing this polymorphism includes ∼7.5 kb from 5′ region of the gene to intron 2. The A allele of this polymorphism was also found to increase the risk for coronary heart disease and was consistently related to this condition after adjustment for type 2 diabetes mellitus in a white population.32 The risk for coronary heart disease conferred by the A allele was even greater in individuals with type 2 diabetes mellitus than in non-diabetic individuals. We have shown earlier that the −30G → A polymorphism of GCK was associated with hypertension, with the A allele being protective against this condition, in a cross-sectional case–control study.11 We have now shown that this polymorphism was related to both the prevalence of hypertension and systolic and diastolic BP in Japanese women, with the variant A allele protecting against hypertension, consistent with our earlier study.11 The mechanisms responsible for the association of the A variant with an increased risk of coronary heart disease,32 with reduced β-cell function and impaired glucose tolerance,14, 31 and with a reduced risk of hypertension (present study) remain to be elucidated. The −258G → A polymorphism located in the liver-specific promoter of GCK was also shown to be associated with hypertension in a Taiwanese population.33

Wingless-type MMTV (mouse mammary tumor virus) integration site family member 1 (WNT1) belongs to a family of cysteine-rich, glycosylated signaling proteins that mediate diverse developmental processes, such as control of cell proliferation, adhesion, polarity and fate. WNT1-inducible signaling pathway protein 1 (WISP1) is induced by WNT1 and belongs to the CCN family, which includes CTGF (connective tissue growth factor), Cyr61 (cysteine-rich 61) and Nov (nephroblastoma-overexpressed gene).34 WISP1 is a target of the Wnt-β–catenin pathway, with its expression being regulated by β-catenin.35, 36 WISP1 activity and availability are modulated by its interaction with decorin and biglycan, two extracellular matrix-associated proteoglycans that are abundant in bone and cartilage.37 The 2364A → G polymorphism was shown to be related to spinal osteoarthritis in postmenopausal Japanese women, with the G allele protecting against this condition.15 The 2364A → G polymorphism (rs2929970) is located in the 3′-untranslated region of WISP1. An LD block containing this polymorphism includes ∼5.5 kb from intron 3 to 3′ region of the gene. We have now shown that this polymorphism was related to both the prevalence of hypertension and systolic and diastolic BP in Japanese men, with the variant G allele representing a risk factor for hypertension, although the underlying mechanism of the association remains to be elucidated.

Our study has several limitations. (i) Given the multiple comparisons of genotypes with BP or the prevalence of hypertension in this study, it is not possible to exclude completely potential statistical errors such as false positives. (ii) It is possible that one or more of the polymorphisms related to BP or the prevalence of hypertension in our study are in LD with other polymorphisms of the same or nearby genes that are actually responsible for the development of hypertension. (iii) The functional relevance of the identified polymorphisms to gene transcription or to protein structure or function was not determined in this study. (iv) Given the lack of replication, validation of our findings will require their replication with independent subject panels.

In conclusion, our results suggest that the 682A → G (Thr228Ala) polymorphism of SORBS1 and the −30G → A polymorphism of GCK may play a role in the development of hypertension in Japanese women and that the 2364A → G polymorphism of WISP1 may play such a role in Japanese men. Determination of genotypes for these polymorphisms may prove informative for the assessment of the genetic risk for hypertension. Given that multiple variants, each having a small effect, will likely ultimately be found to be responsible for a large fraction of the genetic component of essential hypertension, identification of additional hypertension susceptibility genes will allow a more accurate assessment of the genetic risk for this condition.

Conflict of interest

The authors declare no conflict of interest.