Fibroblast growth factor 23 (FGF23) gene polymorphisms are associated with essential hypertension risk and blood pressure levels in Chinese Han population

In this case-control study, 246 EH patients and 157 healthy controls were selected from Chinese Han population to explore the associations between the fibroblast growth factor 23 (FGF23) gene polymorphisms and essential hypertension (EH).The SequenomMassarray system was used for the genotyping of three FGF23 gene Tag single-nucleotide polymorphisms, namely rs7955866, rs13312756, and rs3812822. The primers were designed by Assay Designer 3.1 software, and then the samples were added to a 384-well plate for the polymerase chain reaction amplification, shrimp alkaline phosphatase reaction, and desalting after extension. The distributions of the alleles, genotypes, and haplotypes were compared between the two groups. Confounding factors (sex, age, BMI, smoking, and drinking) were adjusted in the non-logistic regression, and the results showed that rs7955866 and rs3812822 polymorphisms were independently associated with the risk of developing EH (P < 0.05). The statistical analysis of the haplotype of rs7955866-rs13312756-rs3812822 showed that haplotype ACC could increase the risk of developing EH (P = 0.046; OR = 1.513, 95%CI: 1.005-2.278). The analysis of the control group showed that carrying rs7955866 A allele (P = 0.031) and rs3812822 C allele (P = 0.025) was associated with the increase of systolic blood pressure (SBP). The insulin (INS) level in the peripheral blood was significantly different between the case and control groups (P = 0.014). After confounding factors were excluded, the results showed that the serum INS level was also an independent risk factor of developing EH (P = 0.044; OR = 1.604, 95%CI: 1.014-2.539). In summary, our results suggest that FGF23 gene polymorphisms are associated with the risk of developing EH in Chinese Han population.

Fibroblast growth factor 23; single-nucleotide polymorphisms; essential hypertension; molecular genetics; haplotypes

Fibroblast growth factor 23 (FGF23) is a bone-derived hormone-like material that plays an important role in the metabolism of phosphate and vitamin D ([1] ). The concentration of serum FGF23 has been shown to be associated with cardiovascular diseases including coronary heart disease (CHD), atrial fibrillation, and heart failure in recent years ([2] -[5] ). In 2015, S. Masson et al. found in a clinical trial that FGF23 in the peripheral blood is not only associated with left ventricular mass index and left ventricle hypertrophy (LVH), but also independently associated with the mortality rate of patients with cardiovascular diseases after multiple cardiovascular markers including NT-proBNP and hs-cTnT were adjusted, suggesting that FGF23 could be used as a non-traditional risk factor of cardiovascular diseases ([6] ). However, no studies have investigated whether FGF23 gene variations are risk factors of cardiovascular diseases, and could be used for the genetic diagnosis of cardiovascular diseases to date.

The Report on Cardiovascular Disease in China in 2014 showed that the incidence of hypertension increased annually from 2003 to a high level (26.7%) in 2012, which was the highest incidence among all the cardiovascular diseases, and was closely related to the incidences of other cardiovascular diseases including CHD and heart failure ([7] ). The cases of essential hypertension (EH) account for over 90% of all hypertension cases, and the incidence is associated with multiple factors including sex, age, diet, and lifestyles ([8] ). Recent studies showed that 30%-50% of the blood pressure variations among EH patients were caused by genetic variations, which were mainly associated with the changes of the physiopathologic mechanisms in the kidney and adrenal glands. The FGF23 level increase in the peripheral blood could induce the increase of the expression and function of sodium-chloride cotransporters in the distal renal tubule, reduce the sodium excretion effects of the kidney, induce water-sodium retention, and thus cause hypertension and ventricular hypertrophy ([9] ,[10] ). In the light of previous findings, the authors hypothesized that FGF23 single-nucleotide polymorphisms (SNPs) could be associated with the development of EH, and verified this hypothesis in a case-control study to provide evidence to help expand the studies on the associations between FGF23 and cardiovascular diseases to genetic researches.

The present study was performed in accordance with the Helsinki Declaration and was approved by the Ethics Committee of the Daping Hospital, Third Military Medical University. Informed consents were obtained from all the subjects included.

From May 2014 to March 2016, 246 EH patients and 157 healthy subjects were randomly selected from the Department of Cardiology, Daping Hospital, Third Military Medical University. All the subjects were unrelated Chinese Han people with the ages over 30 years. The subjects with tumor, hematonosis, thyroid diseases, parathyroid diseases, hepatitis, liver dysfunction (ALT>2ULN;1ULN = 40U/L), renal diseases including renal vascular stenosis or renal dysfunction (Cr≥133 Umol/L), and autoimmune diseases were excluded. According to the USA 2014 Guidelines on Hypertension Management (JNC8) ([11] ), patients with the systolic blood pressure (SBP) ≥140 mmHg and/or diastolic blood pressure (DBP) ≥ 90 mmHg (1 mmHg = 0.133 kPa) with secondary hypertension excluded were included into the EH group. Subjects with the SBP<140 mmHg and DBP<90 mmHg that did not receive any antihypertensive therapies were included in the control group.

General characteristics of the subjects including the sex, age, smoking, drinking, drug use, previous medical history, and family history were collected with the questionnaire. Smoking was defined as smoking ≥100 cigarettes per year ([12] ). Drinking was defined as drinking ≥12 times per year ([13] ). The height and weight of the subjects were measured in the study, and then the body mass index (BMI) was calculated. For measuring the blood pressure, the sitting blood pressure at the upper arm was measured by a mercury sphygmomanometer. The SBP and DBP were recorded at the first and fifth Korotkoff sound, respectively. The subjects were asked to fast for 12 h, and then blood was collected from the antecubital vein. The levels of triglyceride (TG), total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), glycosylated hemoglobin (HbA1c), fasting blood glucose (FPG), serum creatinine (Scr), and uric acid (UA) were measured using BECKMAN DXC800 Automatic Biochemical Analyzer (Brea, USA). Serum levels of C-peptide (CP) and insulin (INS) were measured by ElectroChemiLuminescence (ECL, Cobas E601, Roche, Basel, Switzerland). The 2-hour postprandial blood glucose (2hPBG) of the subjects was also measured.

For all the subjects, 2 mL blood was collected from the antecubital vein in the early morning into EDTA-anticoagulant tubes and stored at −20°C. Genome DNA was extracted using the Blood Genome DNA Extraction Kit (DP332; Tiangen, Beijing, China), and stored at −70°C until use.

The genotypes of FGF23 gene in Chinese Han population were initially searched in the International HapMap Project Website (http://hapmap.ncbi.nlm.nih.gov/) and checked in the SNP database system of the National Center for Biotechnology Information (http://www.ncbi.nlm.nih.gov/snp) and the Sanger Institute Biobank (http://www.ensembl.org/index.html). Then Haploview 4.2 software was used to select the TagSNPs using the criteria of r2 ≥ 0.8 and the minor allele frequency (MAF) in Chinese Han population ≥0.1. Finally, three TagSNPs of the FGF23 gene were selected (Table 1).

Information of the SNPs selected.

Abbreviations: MAF (CHB), minor allele frequency in Chinese Han population; MAF (GLO): minor allele frequency in global people.

The Massarray system (Sequenom Company) was used for genotyping in the present study. The Massarray system has the following advantages: has high sensitivity (accuracy >99.7%), has low error, and can measure multiple sites in multiple samples simultaneously ([14] ). The Beijing Liuhe Genomics Institute (Shenzhen, China) provided the service of genotyping. For genotyping, the primers were designed by Assay Designer 3.1 software, and then the samples were added into a 384-well plate for the polymerase chain reaction (PCR) amplification, shrimp alkaline phosphatase (SAP) reaction, and desalting after extension. The data of the samples were input into the Typer 4.0 software and then the analysis plate was established. The samples were added again and the Massarray system was used for the last CHIP assay, and then the genotyping analysis was performed by the Typer 4.0 software.

Chi-square (χ2) test or Fisher exact test was used for the Hardy-Weinberg equilibrium test in each SNP for the patients in the control group. Independent t test, χ2 test, or non-parametric test was performed with SPSS 18.0 software for the comparison of age, sex, blood lipid, blood glucose, and INS between the EH and control groups. The online SHEsis software (http://analysis.bio-x.cn/myAnalysis.php) was used for the analysis of genotypes, alleles, and haplotypes between the EH and control groups, and for the linkage disequilibrium (LD) analysis. Non-conditional logistic regression was used for the multivariate regression of the positive SNPs. P < 0.05 was considered statistically significant; 95% confidential interval (CI) was obtained in the analyses performed with SPSS 18.0 and SHEsis software. The haplotypes with the frequencies <0.03 were excluded from the haplotype analysis in the SHEsis software.

In addition to the SBP and DBP, the Scr, UA, FPG, CP, and INS levels were also significantly different between the EH and control groups (P < 0.05); however, the levels of TC, TG, HDL-C, LDL-C, HbA1c, and 2hPBG were not significantly different between the two groups (P > 0.05; Table 2). After the confounding factors including sex, age, BMI, smoking, and drinking were adjusted, the logistic regression showed that the INS level was independently associated with the development of EH (P = 0.044; OR = 1.604, 95%CI: 1.014-2.539).

General characteristics of the subjects in the EH and control groups.

Abbreviations: BMI, body-mass index; TC, total cholesterol; TG, triglyceride; HDL-C, high density lipoprotein cholesterol; LDL-C, low density lipoprotein cholesterol;SCr,serum creatinine;UA, uric acid; HbA1c, glycosylated hemoglobin; FPG, fasting blood-glucose;2hPBG, 2-hour postprandial blood glucose; INS, serum insulin.

The χ2 test and Fisher exact test showed that the frequencies of the genotypes of all the three SNPs in the control group were in Hardy-Weinberg equilibrium (P > 0.05), suggesting that the sample was with group representativeness.

The online analysis with the SHEsis software showed that the frequencies of rs7955866 A and rs3812822 C alleles were significantly higher in the EH group than in the control group (P < 0.05; Table 3); the distribution of the rs7955866 genotypes was also significantly different between the two groups (P = 0.025; Table 3). When the CC and CT genotypes of rs381282 were combined, the χ2 test showed that the frequency of CC+CT was significantly higher in the EH group than in the control group (P = 0.018; Table 3). Univariate logistic regression analysis results showed that the risk of developing EH in the subjects carrying rs7955866 AG genotype was 1.642-fold of the ones carrying GG genotype (P = 0.029; OR = 1.642, 95%CI:1.052-2.561; Table 4), and rs3812822 C allele gene (CC and CT genotypes) carriers also had a higher risk compared with the ones with TT genotype (P = 0.019; OR = 1.663, 95%CI:1.089-2.540; Table 4). Confounding factors including sex, age, BMI, smoking, and drinking were adjusted in the non-logistic regression, and the results showed that the SNPs at rs7955866 (P = 0.012; OR = 1.678, 95%CI: 1.119-2.516; Table 4) and rs3812822 (P = 0.039; OR = 1.519, 95%CI: 1.022-2.568; Table 4) were independently associated with the development of EH.

Distributions of the FGF23 genotypes in the EH and control groups (N = 403).

Univariate and multivariate logistic regression analysis results.

Haplotype analysis was performed with the online SHEsis software. According to the LD analysis results shown in Figure 1, rs7955866-rs13312756-rs3812822 was selected as the haplotype of the FGF23 gene. The results showed that the risk of developing EH in the subjects carrying ACC haplotype was 1.513-fold of the ones not carrying ACC haplotype. In addition, the frequency of ACC haplotype in the EH group was significantly higher than in the control group (P = 0.046; OR = 1.513, 95%CI: 1.005-2.278; Table 5).

Distribution of the haplotypes in the EH and control groups.

According to the genotypes, the subjects in the control group were categorized into different subgroups to further investigate the associations of FGF23 SNPs with SBP/DBP. As shown in Table 6 and Figure 2, for the rs7955866SNP, the subjects in the AA+AG subgroup were with significantly higher SBP than the ones in the GG subgroup (P = 0.031); while for the rs3812822SNP, the SBP was also significantly different in the ones in the CC+CT and TT subgroups (P = 0.025).

Associations between the genotypes and the blood pressure in the control group.

FGF23 gene is located at the human chromosome 12p13.3, and the full length is about 11.5 kBp. In recent years, researchers are becoming more and more interested in the association of FGF23 gene polymorphisms with diseases including rickets, ovarian cancer, and prostate cancer ([15] -[17] ). In the present study on the polymorphisms of FGF23 gene, the investigators found that rs7955866 and rs3812822 polymorphisms were independent risk factors of developing EH. Further investigation on the rs7955866-rs13312756-rs3812822 haplotype showed that carrying ACC haplotype was also an independent risk factor of developing EH. The sites of rs7955866 and rs3812822 are the exon and promoter of FGF23 gene, respectively. The rs7955866 (p.T239M) site is also a missense mutation site, and the variations of the allele could result in a change from threonine to methionine during the translation ([18] ). Ishii et al. ([19] ) found that the variations of the loci in rs7955866 could change the serum level of FGF23 and thus affect the risk of developing pulmonary emphysema. In addition, Pekkinen et al. ([16] ) found that the FGF23 gene variations were associated with serum P-PTH and U-Pi/U-Crea. These findings suggest that rs7955866 is a site with pathophysiological functions. No study focused on the rs3812822 variation in the promoter of FGF23 gene has been reported yet, and the pathogenic mechanisms are still unclear. Hoogendoorn et al. ([20] ) found that about one-third of all the SNPs in the promoter could significantly affect the level of encoding production. The level of the products was associated with the start timing of the gene transcription and the gene expression level. The pathogenic mechanisms of the rs3812822 variation in the promoter of FGF23 gene remain to be further explored.

To avoid the influence from the use of anti-hypertensive drugs in the patients in the EH group, the genotypes of the subjects in the control group were further investigated to explore the effects of the genotypes on the SBP/DBP. The results showed that the subjects in the rs7955866AA+AG subgroup and rs3812822CC+CT subgroup were with significantly higher SBP than the ones in the corresponding control group. These results further confirmed that rs7955866 A allele and rs3812822 C allele are susceptibility genes of EH. No significant association between FGF23 gene polymorphism and DBP was found in the present study, which could be associated with the facts that the mean age of the subjects in the control group was as high as 60.6 years, most of the controls were middle-aged or aged people, the vascular wall elasticity in these subjects was generally reduced, and the proportion of the subjects with simple SBP increase was increasing ([21] ). Epidemiological studies have shown that the association between cardiovascular diseases (CHD, LVH, stroke, and congestive heart failure) and SBP is significantly closer than with DBP ([22] -[24] ). Therefore, clinicians and researchers should seriously focus on the SBP-increasing effects of FGF23 gene polymorphisms on the cardiovascular system.

In addition, the investigators in the present study also found a significant difference in the peripheral blood INS level between the EH and control groups. After confounding factors were adjusted, the results showed that the serum INS level was also an independent risk factor of developing EH. There are ongoing studies investigating the association between hyperinsulinemia and EH. Soleimani([25] ) proposed in 2015 a novel idea that INS resistance could be associated with hypertension; he suggested that the hyperinsulinemia caused by INS resistance could enhance salt absorption in the proximal tubule, resulting in a state of salt overload and hypertension. The theoretical foundation for this hypothesis is that INS at the thick ascending limb of Henle could directly increase NaCl re-absorption via activating the Na-K-2Cl cotransporters and Na-K-ATPase ([26] ); while in the distal nephron and the connecting tubule, INS could activate the epithelial sodium channel (ENaC) and increase the ENaC channel density at the membrane to exert similar effects ([27] ). The findings of the present study were in agreement with the theories proposed by Soleimani. Ongoing studies by the authors of this study are further investigating the mechanisms involved in the associations of INS resistance and INS level with EH.

However, there are several limitations in the present study. For instance, not all the EH patients that were using anti-hypertensive drugs were excluded, thus the effects of different genotypes on SBP/DBP could only be investigated in the control group, which was with relatively small sample size. Clinical data including the coronary angiogram and echocardiography of the subjects will be collected in future studies by the authors of this study, to further investigate the associations between FGF23 gene polymorphisms and cardiovascular diseases including CHD and ventricular hypertrophy, and provide valuable molecular genetic evidence for the genetic diagnosis and individualized treatment of the patients with cardiovascular diseases.

We would like to thank the Beijing Liuhe Genomics Institute for the help in gene sequencing with the SequenomMassarray system.

The authors declare that they have no actual or potential conflicts of interest.

PHOTO (COLOR): Figure 1. Linkage disequilibrium (LD) analysis results.

PHOTO (COLOR): Figure 2. Comparison of the SBP/DBP of the subjects with different genotypes. A: SBP in the subjects with different rs7955866 genotypes; B: SBP in the subjects with different rs3812822 genotypes; C: DBP in the subjects with different rs7955866 genotypes; D:DBP in the subjects with different rs3812822 genotypes.