Influence of age on the association of GIRK4 with metabolic syndrome

Abstract

Background

G protein-coupled inward rectifier K⁺ channel 4 (GIRK4) gene expressions have been implicated in the development of obesity, a key feature of metabolic syndrome (MetS). We investigated whether sequence variants of GIRK4 may represent metabolic risk factors for the Uygur Chinese population.

Methods

The entire GIRK4 gene, including all exons, the promoter and untranslated regions from 48 MetS individuals, was studied in order to identify genetic variations associated with the disorder. Targeted genotyping of four common single nucleotide polymorphisms (SNPs: rs11221497, rs6590357, rs4937391 and rs2604204) and one novel missense mutation (M210I) was performed using the TaqMan polymerase chain reaction method for 443 MetS and 786 non-MetS subjects.

Results

When all MetS cases were compared against all non-MetS controls, no significant association was found between the three SNPs (rs2604204, rs4937391 and rs6590357) and MetS status or metabolic traits. After adjustment, rs11221497 was associated with MetS (odds ratio (OR) [95% CI] = 0.731 [0.551–0.968], P= 0.029). Interestingly, when the MetS group was stratified into subclasses by age, an association was found for the three SNPs (rs2604204, rs4937391 and rs6590357) having estimated false discovery rates <0.001 and age of <50 y. After adjustment, the SNPs rs2604204, rs4937391 and rs6590357 were also associated with MetS in younger subjects: ORs [95% CI]: 1.678 [1.149–2.450], 1.839 [1.204–2.809] and 0.602 [0.379–0.958], respectively. All of the four SNPs showed a trend towards lower or higher metabolic traits (P< 0.05) in younger subjects. In addition, a newly identified missense mutation (M210I) was not specifically related with MetS.

Conclusions

GIRK4 sequence variants appear to associate with MetS in the Uygurian population, and this association may be influenced by age.

Introduction

Worldwide prevalence of obesity has significantly increased among adults and children over the past few decades. This trend is particularly troubling since the presence of excessive abdominal fat has been strongly associated with several metabolic abnormalities, including insulin resistance, dyslipidaemia, hypertension and impaired glucose tolerance. Together, these clinical features comprise a profile of risk factors know as metabolic syndrome (MetS), which leads to diabetes mellitus and cardiovascular disease (CVD).^1,2 Studies into each of the components of MetS have indicated that disease outcome is influenced by complex interactions involving environmental elements and genetic variations,^3–6 thus complicating our attempts to attain a detailed understanding of the aetiology of MetS.

Recent studies into MetS-induced CVD have revealed a functional role for the obesity-related genes encoding the G protein-coupled inward rectifier K+ channels (GIRKs). GIRKs act to govern the resting membrane voltage in many cells, and play important roles in controlling heart rate and action potentials of neurotransmitters in the central nervous system.^7,8 Four GIRK genes (GIRK1–4) have been identified in the human genome.⁹ GIRK4 (also known as KCNJ5) has been implicated in hypothalamic control of energy homeostasis, and GIRK4 homozygous knockout led to a late-onset obesity phenotype in mice.¹⁰ The human GIRK4 gene consists of three exons that encode 420 amino acid (aa) residues and is located on chromosome 11q24. So far, the relationship between genetic variations in human GIRK4 and MetS or metabolic traits remains uncharacterized.

Epidemiological studies of MetS in China's largest province, Xinjiang, have shown that the Uygur ethnic group experiences much higher rates of MetS than other ethnic populations residing in that same area.^11–13 Since genome-based research has demonstrated that common and rare genetic variations can collectively contribute to variation in complex disease in the general population,^14–16 we hypothesized that rare mutations and common single nucleotide polymorphisms (SNPs) in GIRK4 could contribute to MetS. Thus, we carried out sequencing of the GIRK4 promoter and all exons using DNA from 48 Uygur MetS individuals to identify the representative variations for this population. In addition, we investigated the association between these genetic polymorphisms of GIRK4 and the MetS phenotype.

Materials and methods

Study population

Uygur adults, 18 years of age or older and with no intermarriages within the previous three generations, were recruited from the Hetian area of Xinjiang Uygur Autonomous Region during January and February 2006. By using a random multistage cluster sampling approach, 1352 participants were enrolled in the study. Written informed consent was obtained from all subjects prior to the start of study-related procedures.

Demographic and lifestyle characters were collected by a self-reported questionnaire, including sex, age, cigarette smoking and alcohol consumption. Smoking habits were coded as current daily smoker, former smoker or lifetime non-smoker. Similarly, alcohol intake was recorded as current alcohol intake, former alcohol intake or no history of alcohol intake. The questionnaire also determined present intake of alcohol by type (beer, wine, spirits and others). Daily intake of all alcohol was averaged and used to categorize the individual among one of three drinking groups: no alcohol (0 g/day); <30 g/day; or ≥30 g/day. Waist circumference (WC) was measured by a single observer with subjects removing any unnecessary clothing. The measurer faced the subject and placed an inelastic tape around him/her in a horizontal plane at the level of the natural waist, which is the narrowest part of the torso, as seen from the anterior aspect.

Clinical baseline measurements were taken for concentrations of electrolytes, blood glucose, total cholesterol (TC) and triglycerides (TG), and assessments were made for renal function, diabetes status and use of antihypertensive drug therapy. Subjects with secondary hypertension, excessive drinking, cancer or history of stroke were excluded from this study, as were all women currently using hormone-based contraceptives. A total of 123 individuals were removed based on these criteria. The final study population of 1229 individuals was composed of 463 men and 766 women ranging in age from 30 to 70 years.

This study was approved by the Ethical Committee of the People's Hospital of Xinjiang and was conducted in accordance with the World Medical Association Declaration of Helsinki regarding ethical conduct of research involving human subjects. All study methods were carried out using the Strengthening the Reporting of Observational Studies in Epidemiology case-control guidelines.¹⁷

Diagnostic criteria and metabolic measurements

Diagnosis of MetS was based on the criteria defined by the China Diabetes Society. The presence of any three or more of the following factors indicated MetS: overweight or obesity, body mass index (BMI) ≥25 kg/m²; hyperglycaemia, fasting blood glucose (FBG) ≥6.1 mmol/L and/or a two-hour postprandial glucose (2HPG) ≥7.8 mmol/L and/or drug treatment for diabetes mellitus; hypertension, systolic blood pressure (SBP) ≥140 mmHg and/or diastolic blood pressure (DBP) ≥90 mmHg or antihypertension treatment; lipid disorders, TG ≥1.7 mmol/L and/or reduced high-density lipoprotein-cholesterol (HDL-C) <0.9 mmol/L in men and <1.0 mmol/L in women.¹⁸

Diabetes mellitus was defined by an FBG of ≥7.0 mmol/L or a 2HPG of ≥11.1 mmol/L. Blood pressure was measured in the left arm with a standard sphygmomanometer while the subject was seated and after at least 10 min of rest; blood pressure measurements were taken twice over two consecutive days.

Screening genetic variations of GIRK4 in MetS patients

Whole blood samples (5 mL) were collected in PAXgene DNA tubes (PreAnalytiX; Qiagen, Hombrechtikon, Switzerland) and stored at −80°C until use for genomic DNA extraction by using the PAXgene Blood DNA Kit. DNA samples from 48 typical MetS Uygur people were randomly chosen for sequencing of the entire coding region of the GIRK4 gene, including the 5′- and 3′-untranslated regions (UTRs). Sequencing was performed on an ABI 3130Xl genetic analyser (Applied Biosystems, Foster City, CA, USA) using 10 sets of primers. The obtained sequences were analysed for polymorphisms by using Sequencher version 4.7 software (Gene Codes Corp., Ann Arbor, MI, USA), followed by visual inspection. The A nucleotide of the ATG initiator Met codon was designated as +1, after which all numbering followed. The human GIRK4 nucleotide sequence in GenBank (accession number NM_000890) was used as the reference sequence.

Targeted genotyping of rare mutations and common SNPs

Considering the function and linkage disequilibrium (LD) relationships among the identified genetic variations, four representative common SNPs and one rare missense mutation in the GIRK4 gene were selected for genotyping. TaqMan-based amplification was carried out in the ABI 7900HT Fast Real-Time PCR system using DNA samples from all 1229 subjects. Selected representative common SNPs were based on LD with an r ² cut-off less than 0.8 and minor allele frequency (MAF) of greater than 10%. The primers and probes for the known SNPs were chosen based on information available on the ABI website (http://myscience.appliedbiosystems.com). For the novel 630G>A (M210I) SNP, the primers used were 5′-CGGAGACCCTCATGTTTTCCAA-3′ (forward) and 5′-GGAACATGAGGCACAGCTTCT-3′ (reverse), and the sequencing probes were CATCTCCATGCGGGACG (for the G allele) and TCATCTCCATACGGGACG (for the A allele). The success rate of genotyping reached 96%. The samples were also analysed for genotyping errors and the concordance of duplicates was determined to be 95.49%.

Statistical analysis

All values are expressed as either means ± standard deviation (SD) or as median, according to normal or skewed distribution patterns, respectively. The distribution of subjects’ characteristics or genotype frequencies of SNPs between MetS and non-MetS groups was analysed by using Student's t-test or χ ² analysis. To evaluate quantitative MetS-related traits for differences between genotype groups, one-way analysis of variance was used. Logistic regression adjusted for confounding factors, including age, gender and lifestyle (smoking and drinking), was used to evaluate the risk between different genotypes. LD, Hardy–Weinberg equilibrium (HWE) and case-control association analyses were assessed using the SNPAlyze version 7.0 Pro (Dynacom Co. Ltd, Mozart, Japan). The final P values were adjusted by the false discovery rate (FDR) correction for multiple comparisons.¹⁹ The SPSS statistical software package for Windows (version 16; SPSS Inc., Chicago, IL, USA) was used for all analyses. P values <0.05 were considered statistically significant.

Results

Table 1 shows the demographic and anthropometric parameters of the 1229 study subjects. As expected, there were significant differences between the MetS and non-MetS groups for many of the variables, including BMI, WC, SBP, DBP, FBG, 2HPG, TC, TG, HDL-C and low-density lipoprotein-cholesterol (LDL-C) (all P< 0.001). No significant differences were found between two groups for gender or age (P= 0.989 and 0.189, respectively). In addition, the proportion of smoking and drinking was similar between the two groups (P= 0.538 and 0.102, respectively).

Table 1

Basic characteristics of subjects in non-MetS and MetS groups

	Non-MetS	MetS	P
Gender (male/female, n)	296/490	167/276	0.989*
Age (y)	49.89 ± 11.45	52.22 ± 9.23	0.189^†
BMI (kg/m²)	25.10 ± 3.97	29.95 ± 3.68	<0.001^†
WC (cm)	80.39 ± 10.01	92.82 ± 9.49	<0.001^†
SBP (mmHg)	122.41 ± 22.51	146.56 ± 25.79	<0.001^†
DBP (mmHg)	73.82 ± 13.02	87.67 ± 15.07	<0.001^†
FBG (mmol/L)	5.08 (4.61–5.76)	5.47 (4.74–6.47)	<0.001^†
2HPG (mmol/L)	6.34 (5.20–7.94)	7.31 (5.66–10.02)	<0.001^†
TC (mmol/L)	4.25 ± 1.02	4.99 ± 1.60	<0.001^†
HDL-C (mmol/L)	1.13 ± 0.33	1.06 ± 0.34	0.001^†
LDL-C (mmol/L)	2.26 (1.64–3.02)	2.61 (1.87–3.35)	<0.001^†
TG (mmol/L)	1.08 (0.79–1.43)	1.85 (1.24–2.52)	<0.001^†
Lifestyle
Alcohol consumption (%)	5.7	5.9	0.538*
Smoking (%)	10.6	11.3	0.102*

Metabolic syndrome (MetS) was defined by the China Diabetes Society

BMI, body mass index; WC, waist circumference; SBP, systolic blood pressure; DBP, diastolic blood pressure; FBG, fasting blood glucose; 2HPG, 2-hour postprandial glucose; TC, total cholesterol; TG, triglyceride; HDL-C, high-density lipoprotein-cholesterol; LDL-C, low-density lipoprotein-cholesterol

Values were mean ± SD for normal distribution or median (25–75 percentiles) for non-parametric distribution or percentage

*P was calculated by χ ² test

^† P was calculated by Student's t-test

Identification of polymorphisms and selection of representative SNPs in GIRK4

After sequencing GIRK4 in 48 typical MetS subjects, a total of 21 genetic variations were identified, including two in the promoter, eight in intron regions, four in the exon 2 regions and seven in the 3′-UTR. Nine of these genomic mutations were novel and 12 were previously known. The novel 630G>A transition in exon 2 would lead to an amino acid substitution from a methionine to an isoleucine at position 210 (M210I). Our assessment of all 21 variants’ functions, LD relationship and MAF led to four representative common SNPs (−20559C>G [rs11221497], 171C>T [rs6590357], 5126A>G [rs4937391], 6262A>C [rs2604204]) and one rare missense mutation (630G>A [M210I]) in GIRK4 being selected for targeted genotyping in a larger Uygur population (Table 2).

Table 2

Identified variations of the GIRK4 gene in 48 MetS subjects

SNPs	LD	Region	Amino acid substitution	Allele 1 freq	Allele 2 freq	Flanking sequence	ddb SNP ID
−20559C>G	*	Promoter		0.89	0.11	cagcgagccc(c/g)gggacagcta	rs11221497
−20395C>T		Promoter		0.94	0.06	cccatctgct(c/t)catccctctc	rs12419455
−19494C>T	a	Intron1		0.42	0.58	ctggagggca(c/t)gatgacaccc	rs7941582
171C>T	a*	Exon2	Ser57Ser	0.13	0.87	tggagaagag(c/t)ggcaagtgca	rs6590357
630G>A	*	Exon2	Met210Ile	0.98	0.02	tcatctccat(g/a)cgggacgaga
810T>G	a	Exon2	Leu271Leu	0.11	0.89	gcctcttcct(t/g)gtgtctcctc	rs7118824
833T>C	a	Exon2	His278His	0.13	0.88	tcatctccca(t/c)gagatcaacc	rs7118833
5008G>A	a	Intron2		0.75	0.25	atggatggat(g/a)aacagatgac
5010C>T	a	Intron2		0.25	0.75	gatggatgaa(c/t)agatgactgg
5011A>G	a	Intron2		0.60	0.40	atggatgaac(a/g)gatgactgga
5015G>T		Intron2		0.90	0.10	atgaacagat(g/t)actggatggc
5024G>A	a	Intron2		0.94	0.06	tgactggatg(g/a)tggatggatg
5025A>C	a	Intron2		0.94	0.06	tgactggatg(a/c)tggatggatg
5126A>G	b*	Intron2		0.41	0.59	tgtaacttcc(a/g)tttccccagg	rs4937391
6237G>T		3′-UTR		0.84	0.16	ccaccaaggg(g/t)gaaccccagc	rs4373934
6262A>C	*	3′-UTR		0.71	0.29	tgggcttttc(a/c)agtgcttttt	rs2604204
6364T>G		3′-UTR		0.75	0.25	gttttgtttt(t/g)ttcacaattc
6386C>T	a	3′-UTR		0.04	0.96	agccagcctg(c/t)ggtgtgacct	rs7925056
6423A>G		3′-UTR		0.97	0.03	gtcaggggac(a/g)tgatgagcca
6521G>A	b	3′-UTR		0.13	0.88	gatggctcac(g/a)cctgtgatcc	rs2846696
6626C>G	b	3′-UTR		0.77	0.23	actgaaaata(c/g)aaaaaattag	rs1137937

The apparent linkage disequilibrium (LD), defined by r ² cut-off of 0.8. Single nucleotide polymorphisms (SNPs) indicated by ‘a’ or ‘b’ are in tight LD: a, r _a ² = 0.8; b, r _b ² = 1

*These SNPs were used for genotyping experiments. The A of the initiator Met codon is denoted nucleotide +1. The genome sequence retrieved from GenBank (accession ID: NM_000890) was used as a reference sequence

Among the five mutations examined in 1229 individuals with and without MetS, 630G>A was identified in only seven. Even though several of these 630G>A carrying subjects did not meet the diagnostic criteria of MetS, all presented with mid-section obesity (having WC >90 cm [men] and >80 cm [women]).²⁰ The complete clinical profiles for these seven carriers are presented in Table 3.

Table 3

Profiles of seven individuals with missense mutations 630G>A (M210I) in GIRK4

Case	1	2	3	4	5	6	7
Gender	Female	Female	Female	Female	Female	Female	Female
Age (years)	46	54	54	65	60	55	66
BMI (kg/m²)	28.60	31.60	36.60	25.90	24.00	25.00	27.90
WC (cm)	85	98	100	90	86	92	94
SBP (mmHg)	111	116	116	121	178	117	210
DBP (mmHg)	69	82	70	70	97	66	120
FBG (mmol/L)	6.6	5.3	8.7	4.2	4.9	4.2	6.5
2HPG (mmol/L)	6	9.7	9.7	4.8	6.8	8.1	7.2
TC (mmol/L)	3.7	5.6	8.2	3.3	4.6	4.4	7.4
HDL-C (mmol/L)	1.6	1.2	2.5	0.7	1	1.1	1.7
LDL-C (mmol/L)	1.5	3.3	5.9	1.6	1.9	2	3.1
TG (mmol/L)	2	4.1	3.3	0.7	1.3	1.1	1.9
Lifestyle
Alcohol consumption (%)	No	No	No	No	No	No	No
Smoking (%)	No	No	No	No	No	No	No
Present illness
Hypertension	No	Yes*	No	Yes*	Yes	No	Yes
Hyperglycaemia	Yes	Yes	No	No	No	No	No
Hyperlipidaemia	No	Yes	Yes	Yes	No	No	Yes

BMI, body mass index; WC, waist circumference; SBP, systolic blood pressure; DBP, diastolic blood pressure; FBG, fasting blood glucose; 2HPG, 2-hour postprandial glucose; TC, total cholesterol; TG, triglyceride; HDL-C, high density lipoprotein-cholesterol; LDL-C, low density lipoprotein-cholesterol

*Cases with diagnosed hypertension, blood pressure is well-controlled

Uygurians' MetS association with four common SNPs

An association was found to exist between MetS in Uygurians and the four selected representative SNPs in GIRK4. The genotype distribution of each SNP in MetS and non-MetS subjects was in HWE (P> 0.05). In particular, rs11221497 was found to be associated with MetS in a dominant model (P= 0.026); however, no significant difference was detected between MetS and either rs6590357, rs4937391 or rs2604204 in the total population. Notably, when the entire population was stratified according to median age (49 y), the associations to MetS were observed in younger (30–49 y) rather than in older (50–70 y) subjects (P= 0.014 for rs2604204, P= 0.032 for rs4937391 and P= 0.009 for rs6590357) Moreover, the effect of rs2604204, rs4937391 or rs6590357 genotype achieved the significant level after correction by FDR for multiple comparisons (FDR – P< 0.001) (Table 4). After adjusting for age, gender and lifestyle (smoking and drinking), multivariate logistic regression analysis indicated that rs11221497 was significantly associated with MetS in the total population (OR [95% CI] = 0.731 [0.551–0.968], P= 0.029); moreover, the associations were observed in younger subjects as well (OR [95% CI] = 1.625 [1.008–2.619] for rs6590357, 1.678 [1.149–2.450] for rs4937391 and 0.621 [0.396–0.973] for rs11221497); it is possible that only rs6590357 was associated with MetS in older subjects (OR [95% CI] = 0.675 [0.462–0.985], P= 0.042) (Table 5).

Table 4

Distribution of genotypes for GIRK4 variable in MetS and non-MetS groups

Polymorphisms	Genotypes	Total case/control	*FDR-P	Power	Subgroup 1 case/control	*FDR-P	Power	Subgroup 2 case/control	*FDR-P	Power
rs2604204	AA	217/408			65/202			152/206
	AC	191/315	0.93	—	85/149	<0.001	0.957	106/166	0.56	—
	CC	33/57			13/29			20/28
rs4937391	AA	219/398			66/198			153/200
	AG	192/327	0.93	—	87/157	<0.001	0.544	105/170	0.56	—
	GG	33/57			13/28			20/29
rs6590357	CC	337/608			112/307			225/301
	CT	99/169	0.93	—	50/73	<0.01	0.798	49/96	0.52	—
	TT	7/9			3/4			4/5
rs11221497	CC	353/577			136/279			217/298
	CG	87/196	0.93	0.756	30/29	<0.001	0.883	57/97	0.6	—
	GG	4/8			0/3			4/5

*P values were modified by the false discovery rate (FDR) correction for multiple comparisons. Subgroup 1, age <50 y; subgroup 2, age ≥50 y

MetS, metabolic syndrome

Table 5

Odds ratios and 95% confidence intervals for four variations of GIRK4 gene associated with MetS

		Total	Subgroup 1		Subgroup 1
Polymorphisms	Genotype	OR (95 CI %)	OR (95 CI%)	P*	OR (95 CI%)	P*
rs2604204	AA	1	1	0.004	1	0.401
	AC+CC	1.100 (0.912–1.326)	1.749 (1.195–2.559)		0.876 (0.643–1.193)
rs4937391	AA	1	1	0.007	1	0.207
	AG+GG	1.059 (0.878–1.277)	1.678 (1.149–2.450)		0.819 (0.601–1.116)
rs6590357	CC	1	1	0.005	1	0.042
	CT+TT	1.086 (0.823–1.431)	1.839 (1.204–2.809)		0.675 (0.462–0.985)
rs11221497	CC	1	1	0.032	1	0.253
	CG+GG	0.731 (0.551–0.968)	0.602 (0.379–0.958)		0.808 (0.561–1.164)

CI, confidence interval; OR, odds ratio

*Conditional logistic analysis, adjusted for age, gender and lifestyle (smoking and drinking). Subgroup 1, age <50 y; subgroup 2, age ≥0 y

Biochemical parameters of MetS-afflicted Uygurians

Considering the significant association with MetS that was found in people aged below 50 y, we further investigated potential relationships between the polymorphisms and various metabolic traits in the younger population (Table 6). The quantitative phenotypes of WC, 2HPG and TC were found to be significantly associated with the genotypes of rs11221497 (P= 0.022, P= 0.004, P= 0.007, respectively) and after multivariate adjustment (P= 0.006, P= 0.037, P= 0.003, respectively). For rs6590357, the polymorphisms and the quantitative traits of WC, LDL-C and TG were meaningfully related before (P= 0.017, P= 0.011, respectively) and after multivariate adjustment (P= 0.029, P= 0.0041, respectively). Also, rs4937391 was significantly associated with 2HPG and LDL-C before (P< 0.001, P = 0.019, respectively) and after adjustment (P= 0.004, P = 0.048, respectively). The quantitative traits of 2HPG were also notably associated with rs2604204 after adjustment (P= 0.017). However, no SNP was associated with blood pressure (P> 0.05).

Table 6

Comparison of MetS quantitative phenotypes in genotypes of GIRK4 in Uygur younger subjects

	WC	FBG	2HPG	TC	LDL-C	TG	SBP	DBP
Genotypes	(cm)	(mmol/L)	(mmol/L)	(mmol/L)	(mmol/L)	(mmol/L)	(mmHg)	(mmHg)
rs2604204
AA (290)	85.06 ± 11.62	5.13	6.14	4.35 ± 1.23	2.41	1.35	123.33 ± 22.15	77.43 ± 14.10
AC (258)	86.00 ± 11.98	5.12	6.96	4.30 ± 1.22	2.27	1.31	124.34 ± 26.16	77.56 ± 15.23
CC (48)	86.21 ± 9.92	5.57	6.68	4.65 ± 1.36	2.81	1.41	117.79 ± 18.40	76.21 ± 14.00
P value (*/^†)	0.59/0.45	0.011/0.188	0.08/0.017	0.194/0.566	0.016/0.074	0.274/0.195	0.214/0.447	0.839/0.73
rs4937391
AA (289)	85.51 ± 11.59	5.15	6.14	4.34 ± 1.23	2.36	1.34	123.37 ± 22.26	77.43 ± 14.08
AG (266)	85.78 ± 12.07	5.13	6.89	4.33 ± 1.23	2.35	1.31	124.27 ± 26.05	77.62 ± 15.31
GG (48)	85.98 ± 10.47	5.55	6.9	4.59 ± 1.41	2.92	1.49	117.04 ± 18.39	75.17 ± 13.91
P value (*/^†)	0.945/0.884	0.655/0.056	<0.001/0.004	0.391/0.479	0.019/0.048	0.273/0.144	0152/0.379	0.557/0.293
rs6590357
CC (459)	85.01 ± 11.53	5.14	6.22	4.31 ± 1.26	2.27	1.29	123.17 ± 23.42	76.94 ± 14.37
CT (135)	87.84 ± 12.26	5.22	6.96	4.52 ± 1.22	2.57	1.45	123.96 ± 25.39	78.39 ± 15.52
TT (9)	85.22 ± 9.78	5.1	7.08	4.44 ± 0.72	2.99	2.57	120.67 ± 19.14	80.50 ± 13.36
P value (*/^†)	0.047/0.1	0.054/0.077	0.109/0.055	0.212/0.308	0.017/0.029	0.011/0.041	0.893/0.887	0.486/0.562
rs11221497
CC (454)	86.47 ± 12.06	5.19	6.62	4.44 ± 1.28	2.41	1.36	124.24 ± 25.01	77.84 ± 15.02
CG (142)	83.39 ± 10.22	5.07	6.01	4.15 ± 1.10	2.31	1.32	121.08 ± 19.36	76.19 ± 13.23
GG (5)	83.80 ± 9.04	5.01	6.43	3.28 ± 0.83	2.85	1.17	109.80 ± 15.35	69.00 ± 12.63
P value (*/^†)	0.022/0.006	0.057/0.052	0.004/0.037	0.007/0.003	0.189/0.102	0.157/0.062	0.17/0.058	0.22/0.106

WC, waist circumference; FBG, fasting blood glucose; 2HPG, 2-hour postprandial glucose; TC, total cholesterol; TG, triglyceride; HDL-C, high-density lipoprotein-cholesterol; LDL-C, low-density lipoprotein-cholesterol; SBP, systolic blood pressure; DBP, diastolic blood pressure

Data were mean ± SD for normal distribution or median for non-parametric distribution

*P, analysis of variance

^† P, analysis of covariance adjusted for sex, age and lifestyle (smoking and drinking)

Discussion

Our investigations into the GIRK4 genotypes revealed that each of the four common polymorphisms examined (rs2604204, rs4937391, rs6590367 and rs11221497) are significantly associated with MetS in the Uygurian population. After adjustment for confounding factors, one polymorphism (rs11221497) was strongly associated with MetS. Moreover, the rs6590367 and rs4937391 were clearly connected with MetS in Uygur subjects under 50 years of age. Finally, we found that the G allele of rs11221497 acted as a protective factor against MetS, and the T allele of rs6590367 and G allele of rs4937391 may contribute to development of MetS.

A key finding of the present study was the association of some SNPs with MetS specifically in a younger population. It is interesting to consider that hyperactivity of the sympathetic nervous system is mainly observed in younger hypertensive persons. Increased peripheral sympathetic nerve activity and higher levels of circulating catecholamines characterize the hypertensive state, and these clinical features appear to be associated with MetS development.^21,22 Although Ford et al.²³ have demonstrated that MetS is more common in people ≥50 years of age, different from our findings here, we have previously observed that Uygurians experience lower WC and BMI after 50 years of age.²⁴ We propose then that the natural aging process may obscure the influence of GIRK4 polymorphisms on the development of MetS. We plan to address this possibility in future studies on the GIRK4 SNPs using a larger population of MetS individuals <50 years old.

Previous studies have provided strong evidence to indicate that MetS is a complex disease associated with several genetic factors which may modulate an individual's risk.^3–6 Current research strategies to elucidate the underlying molecular mechanisms of MetS are largely genome-oriented. A candidate gene may be selected based upon a previously established biological association or consanguinity to a known disease-related gene. Similarly, we employed a candidate gene strategy using a sequencing approach to investigate genetic variations affecting the Uygurian's apparent susceptibility to MetS. The Uygurians are considered a relatively pure ethnic group, since they have been geographically stable and have shunned the practice of intermarriage with other ethnic groups. We chose to use the exon sequencing approach since this high-throughput method is highly sensitive and cost-effective for identification of rare and common genetic variants, as opposed to genome-wide scans, and since rare variants play important roles in a broad spectrum of diseases.^14–16

The SNPs that we discovered are associated with MetS in Uygurians (rs11221497, rs6590357, rs4937391, rs2604204) are located throughout the GIRK4 gene (respectively, in the promoter, exon 2, intron 2 and 3′-UTR). The promoter acts as a functional DNA element, harbouring the site of transcription initiation. Coding exons are not only crucial to proper translation but are affected by alternative splicing and are important regions of functional variation in disease.^17,25–27 Likewise, the non-coding portions of genomes modulate gene expression.^28,29 For example, UTRs are involved in many post-transcriptional regulatory pathways related to mRNA localization, stability and translation efficiency, as well as influencing initiation of protein synthesis. Introns contain several short sequences that are important for efficient splicing, such as acceptor and donor sites. The mechanisms by which the four SNPs might contribute to MetS are currently unknown. It is possible that LD with other functional variations within the GIRK4 gene, or other functional polymorphisms of different genes, play more important roles in MetS. A greater density of genotyping around the GIRK4 gene is needed to address these issues.

It is well-established that mutations in coding exons can lead to phenotypic alterations. In our report, the rare missense mutation 610G>A was situated in the GIRK4 gene exon 2. Although no AA homozygote was verified in our study population, all carriers of the GA sequence presented with obvious abdominal obesity (mean BMI of 28.51 ± 4.38 kg/m²). Therefore, it is possible that this particular mutation may confer increased risk of developing mid-section obesity.

In order to gain a better understanding of our genetic findings, we considered the previous research reported for both GIRK4 and MetS. GIRK channels have been detected in many different mammalian cell types, where they have been found to participate in the maintenance of K⁺ homeostasis and regulation of membrane electropotential.^7–9,30,31 Inward rectifier K+ channels handling glucose metabolism have been proposed as important aetiological factors for diabetes.^32–34 Experiments in mice have shown that GIRK4 (KCNJ5) protein is highly expressed in the hypothalamus, the satiety control organ.^35,36 Genes expressed in the hypothalamus are considered as particularly promising candidates for susceptibility to obesity, which is a key component of MetS.^37–39 Furthermore, a GIRK4 gene knockout resulted in a murine phenotype of late-onset obesity.¹⁰ While Perry and colleagues found that GIRK4 ^−/− mice were approximately 25% heavier than age-matched wild-type controls, the precise mechanism by which GIRK4 gene expression leads to obesity is not clear. A previous study has suggested that GIRK4 may act to inhibit firing of NPY neurons, but further evidence to support this idea has yet to be reported.⁴⁰ Genetic variations of GIRK4 that lead to perturbed protein function could blunt the satiation response, resulting in more food intake and reducing energy expenditure through GIRK channels, thus having an inhibitory effect on hypothalamic neuron subtypes.

An important limitation of our study was the relatively small sample size. Unfortunately, our finding that none of the younger MetS subjects carried the rs11221497 GG genotype could not be considered as a definitive result. A larger cohort is needed to explore further the genetic associations with MetS in different age groups of Uygurians.

Nonetheless, the present study demonstrates that the genetic variations of GIRK4 are associated with the MetS phenotype in Uygur Chinese. Rs11221497 appears to be an independent risk factor for MetS in our study population. A newly identified mutation, 630G>A (M210I), might contribute to development of visceral obesity in Uygurians, and merits further study. Furthermore, our study emphasized the necessity of taking into account physiological changes related to aging when analyzing phenotypic effects of genetic variants.

In conclusion, the present study has suggested that GIRK4 sequence variants may contribute to MetS and demonstrated that age should be considered when interpreting the results of investigations of GIRK4 and features of MetS. An explanation as to why the relationship of GIRK4 and features of MetS are age-dependent is a worthy area of future investigation.

DECLARATIONS

Competing interests: None.

Funding: This research was funded by the Basic Conditions and Infrastructure for Science and Technology project in Xinjiang Uygur autonomous region (PT0903).

Ethical approval: The ethics committee of the People's Hospital of Xinjiang Uygur Autonomous Region approved this study.

Guarantor: NL.

Contributorship: NL, DZ and JZ researched the literature and conceived the study. YG, ZY, HW, LZ, JH, XW and ZA were involved in protocol development, gaining ethical approval, patient recruitment and data analysis. DZ and JZ wrote the first draft of the manuscript. All authors reviewed and edited the manuscript and approved the final version of the manuscript.

Acknowledgements: We would like to express our sincere gratitude to all the staff at the Center of Diagnosis, Treatment and Research of Hypertension in Xinjiang for support with the medical examinations and demographic data collection.

References

Grundy

. Metabolic syndrome pandemic. Arterioscler Thromb Vasc Biol 2008;28:629–36

Alberti

, Zimmet

, Shaw

. IDF Epidemiology Task Force Consensus Group. The metabolic syndrome: a new worldwide definition. Lancet 2005;366:1059–62

Edwards

, Hutter

, Wan

, Genome-wide linkage scan for the metabolic syndrome: the GENNID Study. Obesity 2008;16:1596–601

Bray

, Bouchard

. Obesity. In: King

, Rotter

J-I

, Motulsky

eds. The Genetic Basis of Common Disease. 2nd edn. Oxford: Oxford University Press, 2002:481–509

Joy

, Lahiry

, Pollex

, Genetics of metabolic syndrome. Curr Diabetes Rep 2008;8:141–8

Kraja

, Hunt

, Pankow

, Quantitative trait loci for metabolic syndrome in the Hypertension Genetic Epidemiology Network study. Obes Res 2005;13:1885–90

Reimann

, Ashcroft

. Inwardly rectifying potassium channels. Curr Opin Cell Biol 1999;11:503–8

Logothetis

, Kurachi

, Galper

, The beta gamma subunits of GTP-binding proteins activate the muscarinic K+ channel in heart. Nature 1987;325:321–6

Mark

, Herlitze

. G-protein mediated gating of inward-rectifier K⁺channels. Eur J Biochem 2000;267:5830–6

10.

Perry Cydne

, Marco

, Teske Jennifer

, Predisposition to late-onset obesity in GIRK4 knockout mice. Proc Natl Acad Sci 2008;105:8148–53

11.

Yan

, Yang

, Zheng

, The metabolic syndrome in Uygur and Kazak populations. Diabetes Care 2005;28:2554–5

12.

Yianying

, Kun

, Lei

, Epidemiology study on the metabolic syndrome in boertala mongol autonomous prefecture of Xinjiang. Chin J Intern Med 2006;45:227–8 [Chinese]

13.

Yan

, Zijin

, Abulikemu

. Epidemiological study for the metabolic syndrome in Uygur adults in Urumqi city. China J Mod Med 2004;14:155–6 [Chinese]

14.

Manolio

, Collins

, Cox

, Finding the missing heritability of complex diseases. Nature 2009;461:747–53

15.

Cohen

, Kiss

, Pertsemlidis

. Multiple rare alleles contribute to low plasma levels of HDL cholesterol. Science 2004;305:869–72

16.

, Leal

. Methods for detecting associations with rare variants for common diseases: application to analysis of sequence data. Am J Hum Genet 2008;83:311–21

17.

Little

, Higgins

, Ioannidis

, Strengthening the Reporting of Genetic Association Studies (STREGA) – an extension of the STROBE statement. PLoS Med 2009;3:151–63

18.

China Diabetes Society. Definition of metabolic syndrome for Chinese. Chin J Diabetes 2004;12:156–61 [Chinese]

19.

Benjamini

, Drai

, Elmer

, Kafkafi

, Golani

. Controlling the false discovery rate in behavior genetics research. Behav Brain Res 2001;125:279–84

20.

WHO Expert Consultation. Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. Lancet 2004;363:157–63

21.

Gavi

, Qurashi

, Stuart

, Influence of age on the association of retinol-binding protein 4 with metabolic syndrome. Obesity 2008;16:893–5

22.

Shi

, Gu

, Kraja

, Genetic effect on blood pressure is modulated by age: the Hypertension Genetic Epidemiology Network Study. Hypertension 2009;53:35–41

23.

Ford

, Giles

, Dietz

. Prevalence of the metabolic syndrome among US adults: ﬁndings from the Third National Health and Nutrition Examination Survey. JAMA 2002;287:356–9

24.

Tian

, Yitong

, Yining

, The prevalence of metabolic syndrome in Han and Uygur in adults from Xinjiang. J Xinjiang Med Univ 2011;34:129–32 [Chinese]

25.

, Turner

, Robertson

, Targeted capture and massively parallel sequencing of 12 human exoms. Nature 2009;461:272–6

26.

, Levy

, Huang

, Genetic variation in an individual human exome. PLoS Genet 2008;8:e1000160

27.

Collins

, Guyer

, Charkravarti

. Variations on a theme: cataloging human DNA sequence variation. Science 1997;278:1580–1

28.

Wang

, Yang

, Xie

. Gene variants in noncoding regions and their possible consequences. Pharmacogenomics 2006;7:203–9

29.

Conklin

, Jonassen

, Aasland

, Association of nucleotide patterns with gene function classes: application to human 3' untranslated sequences. Bioinformatics 2004;18:182–9

30.

Aguado

, Colón

, Ciruela

, Cell type-specific subunit composition of G protein-gated potassium channels in the cerebellum. J Neurochem 2008;105:497–511

31.

Wickman

, Karschin

, Brain localization and behavioral impact of the G-protein-gated K⁺channel subunit GIRK4. J Neurosci 2000;20:5608–15

32.

Gloyn

, Pearson

, Antcliff

, Activating mutations in the gene encoding the ATP-sensitive potassium-channel subunit Kir6.2 and permanent neonatal diabetes. N Engl J Med 2004;350:1838–49

33.

Hattersley

, Ashcroft

. Permanent neonatal diabetes due to mutations in KCNJ11 encoding Kir6.2: patient characteristics and initial response to sulfonylurea therapy. Diabetes 2004;53:2713–8

34.

Okamoto

, Iwasaki

, Nishimura

, Identification of KCNJ15 as a susceptibility gene in Asian patients with type 2 diabetes mellitus. Am J Hum Genet 2009;86:54–64

35.

Spauschus

, Lentes

, Wischmeyer

, A G-protein-activated inwardly rectifying K+ channel (GIRK4) from human hippocampus associates with other GIRK channels. J Neurosci 1996;16:930–8

36.

Rankinen

, Zuberi

, Chagnon

, The human obesity gene map: the 2005 update. Obesity 2006;14:529–644

37.

Hotta

, Nakata

, Matsuo

, Kamohara

, Kotani

, Komatsu

. Variations in the FTO gene are associated with severe obesity in the Japanese. J Hum Genet 2008;53:546–53

38.

Wynne

, Stanley

, McGowan

, Appetite control. J Endocrinol 2005;184:291–318

39.

Acuna-Goycolea

, van den Pol

. Peptide YY (3–36) inhibits both anorexigenic proopiomelanocortin and rexigenic neuropeptide Y neurons: implications for hypothalamic regulation of energy homeostasis. J Neurosci 2005;25:10510–9

40.

DeNino

, Tchernof

, Dionne

, Contribution of abdominal adiposity to age-related differences in insulin sensitivity and plasma lipids in healthy non obese women. Diabetes Care 2001;24:925–32