|Home | About | Journals | Submit | Contact Us | Français|
Conceived and designed the experiments: CH KX WJ. Performed the experiments: RZ CW WY JL XM JW. Analyzed the data: CH YB. Contributed reagents/materials/analysis tools: FJ ST. Wrote the paper: CH.
Single nucleotide polymorphisms (SNPs) from GCK, GCKR, G6PC2 and MTNR1B were found to modulate the fasting glucose levels. The current study aimed to replicate this association in the Chinese population and further analyze their effects on biphasic insulin secretion.
SNPs from GCK, GCKR, G6PC2 and MTNR1B were genotyped in the Shanghai Chinese, including 3,410 type 2 diabetes patients and 3,412 controls. The controls were extensively phenotyped for the traits related to glucose metabolism and insulin secretion. We replicated the association between GCK rs1799884, G6PC2 rs16856187 and MTNR1B rs10830963 and fasting glucose in our samples (p=0.0003~2.0×10−8). GCK rs1799884 and G6PC2 rs16856187 showed association to HOMA-β, insulinogenic index and both first- and second-phases insulin secretion (p=0.0030~0.0396). MTNR1B rs10830963 was associated to HOMA-β, insulinogenic index and first-phase insulin secretion (p=0.0102~0.0426), but not second-phase insulin secretion (p=0.9933). Combined effect analyses showed individuals carrying more risk allele for high fasting glucose tended to have a higher glucose levels at both fasting and 2 h during OGTTs (p=1.7×10−13 and 0.0009, respectively), as well as lower HOMA-β, insulinogenic index and both first- and second-phases insulin secretion (p=0.0321~1.1×10−7).
We showed that SNPs from GCK, G6PC2 and MTNR1B modulated the fasting glucose levels in the normoglycaemic population while SNPs from G6PC2 and GCKR was associated with type 2 diabetes. Moreover, we found GCK and G6PC2 genetic variants were associated to both first- and second-phases insulin secretion while MTNR1B genetic variant was associated with first-phase insulin secretion, but not second-phase insulin secretion.
Fasting glucose plays a central role in the pathogenesis of diabetes and its complications , . Studies in twins and families had shown that genetic factors contributed to fasting glucose levels in the population , . However, the genes regulating fasting glucose levels are different from the genes affecting either type 1 or type 2 diabetes susceptibility, although fasting glucose is an important component of diabetes diagnosis. Recently, advance was made in identifying genes regulating fasting glucose through genome-wide association studies. The first wave of discovery of fasting glucose genes identified glucokinase (GCK), glucokinase regulatory protein (GCKR), glucose-6-phosphatase catalytic subunit 2 (G6PC2) and melatonin receptor 1B (MTNR1B) , , . More recently, the meta-analysis identified additionally 9 new loci implicated in fasting glucose homeostasis, including ADCY5, MADD, ADRA2A, CRY2, FADS1, GLIS3, SLC2A2, PROX1 and C2CD4B) . Although these genes were identified originally from the European descent populations, replication studies in the Chinese populations confirmed some of their effects on the traits related to glucose metabolism. Two studies, focused on GCK and GCKR, were performed in other Chinese populations. They found GCK rs1799884 was associated with high fasting glucose while GCKR rs780094 was associated with high triglyceride level and type 2 diabetes , . Another study replicated the effects of MTNR1B rs10830963 on fasting glucose and type 2 diabetes . We previously also studied G6PC2 using a tagging single nucleotide polymorphism (SNP) approach covering all the common variants spanning the gene. We identified the SNP rs16856187 was associated to the fasting glucose levels as well as type 2 diabetes risks in the Chinese . These genes were linked to beta cell function in various populations mainly because of their association to HOMA-β. However, HOMA-β cannot be used to quantify the biphasic insulin response after stimulated by food. Thus the effects of these genes on first- and second-phases of insulin secretion remained largely unknown. To investigate the effects of GCK, GCKR, G6PC2 and MTNR1B variants on biphasic insulin secretion, we tested the individual as well as combined effects of these variants on the traits related to glucose metabolism, especially multiple measurements on insulin secretion.
This study was approved by the institutional review board of Shanghai Jiao Tong University Affiliated Sixth People's Hospital in accordance with the principle of the Helsinki Declaration II. Written informed consent was obtained from each participant.
We recruited a total of 6,822 participants of Chinese Han ancestry residing in Shanghai, comprising 3,410 type 2 diabetes patients and 3,412 controls. The inclusion criteria were described previously . Briefly, all cases were unrelated type 2 diabetes patients defined according to 1999 WHO criteria and were treated with oral hypoglycemic agents and/or insulin . The controls were selected from community-based random sample epidemiological studies of diabetes and related metabolic disorders. All controls were unrelated subjects with normal glucose tolerance as assessed by a standard 75 g oral glucose tolerant tests (OGTTs), and negative family history of diabetes. Among them, 1,892 cases and 1,808 controls overlapped with previous studies , . The clinical characteristics were shown in Table 1.
For a SNP with minor allele frequency over 0.2, our case-control samples had over 80% power to detect the minimum OR of 1.13 and our control samples had over 90% power to detect the minimum effect size of 0.05 mmol/l per allele on fasting glucose at a level of significance of 0.05.
Phenotypes for anthropometric and biochemical traits related to glucose metabolism were extensively measured for both case and control subjects. OGTTs were performed in the controls in the morning after an overnight fast. Blood samples were obtained at the fasting and 2 h during OGTTs. Plasma glucose and serum insulin were measured. Basal insulin sensitivity and beta cell function were calculated from fasting plasma glucose and insulin using HOMA . First- and second-phase insulin secretions were estimated using the glucose and insulin levels at 0 and 120 min during the OGTT and BMI measurements . In a subgroup of controls, plasma glucose and serum insulin levels at 30 min during OGTTs were also measured. The insulinogenic index was calculated as the ratio of the increment in insulin concentration to the increment in glucose concentration ([30 min insulin – fasting insulin]/[30 min glucose – fasting glucose]).
We selected four previously reported SNPs from loci affecting fasting glucose, including GCK rs1799884, GCKR rs780094, G6PC2 rs16856187 and MTNR1B rs10830963 , , , , , , . The SNPs were genotyped using primer extension of multiplex products with detection by matrix-assisted laser desorption ionization–time of flight mass spectroscopy using a MassARRAY Compact Analyzer (Sequenom, San Diego, CA, USA). Detailed information on genotype data quality controls was mentioned previously . After all the quality control checks, 6,540 individuals (3,228 cases and 3,312 controls) and all four SNPs were analyzed.
The Hardy-Weinberg equilibrium test was performed in the cases and controls separately before association analysis. The allelic frequencies between the diabetic patients and controls were compared using χ2 tests, and ORs with 95% CIs were presented. Quantitative traits were analyzed by linear regression adjusted for age, gender and BMI, and the regression coefficients (βs) were presented. An additive genetic model was used for the analysis, unless specified otherwise. All skewly distributed quantitative traits were logarithmically transformed to approximate univariate normality. Permutations (100,000 times for fasting glucose and 10,000 times for other traits) were performed for each trait to assess empirical p values using PLINK  in order to adjust the multiple comparison. The statistical analyses were performed using SAS for Windows (version 8.0; SAS Institute, Cary, NC, USA) unless specified otherwise. A two-tailed p value of <0.05 was considered statistically significant.
All SNPs were in Hardy-Weinberg equilibrium. We analyzed the effects of these loci on glucose levels in the controls. GCK rs1799884, G6PC2 rs16856187 and MTNR1B rs10830963 showed associations to fasting glucose (p=7.2×10−5, 0.0003 and 2.0×10−8; empirical p=0.00015, 0.00099 and 0.00001) (Table 2). After further adjusting for age, gender and BMI as confounders, the effects of these SNPs were similar to the previous reports in European and Asian populations (β=0.048−0.06 mmol/L). No interaction effect was detected between GCK, GCKR and G6PC2 SNPs (data not shown). GCK rs1799884 and G6PC2 rs16856187 showed effects on the 2 h glucose levels during the OGTTs (GCK: β=0.11±0.04 per A allele, p=0.0089, empirical p=0.0255; G6PC2: β=0.07±0.03 per C allele, p=0.0153, empirical p=0.0517). For the lipid profiles, GCKR rs780094 and G6PC2 rs16856187 were associated with triglyceride (p=0.0002 and 0.0118, respectively).
To further investigate the impact of these variants on beta cell function, we analyzed their effects on HOMA-β, estimated first- and second-phases of insulin secretion in the controls with both fasting and 2 h insulin data available (n=2,376), as well as the insulinogenic index in the controls with fasting, OGTT 30 min glucose and insulin data available (n=960). As shown in Table 2, GCK rs1799884 was associated with HOMA-β, insulinogenic index and both first- and second phases of insulin secretion (p=0.0396, 0.0129, 0.0030 and 0.0223, respectively). G6PC2 rs16856187 showed evidence for association to first-phase insulin secretion (p=0.0108), and second-phase insulin secretion under a dominant genetic model (p=0.0431). MTNR1B rs10830963 was associated with HOMA-β, insulinogenic index and first-phase insulin secretion (p=0.0426, 0.0162 and 0.0102, respectively), but not second-phase insulin secretion (p=0.9933). GCKR rs780094 showed no association to the traits related to insulin secretion.
We then analyzed the association between these SNPs and the risk for type 2 diabetes susceptibility. As shown in Table 3, we found GCKR rs780094 and G6PC2 rs16856187 showed evidence for association to type 2 diabetes (p=5.25×10−5 and 0.0021, respectively). No association was detected between GCK rs1799884 and MTNR1B rs10830963.
The combined effects of the SNPs from GCK, G6PC2 and MTNR1B were then examined in our samples. We only included the samples without genotype missing (n=3,212). We found the individuals carrying more risk alleles tended to have a higher glucose levels at both fasting and 2 h during OGTT (fasting glucose: β=0.04mmol/l per allele, p=1.69×10−13; 2 h glucose: β=0.05mmol/l per allele, p=0.0009), and lower HOMA-β (p=0.0011), insulinogenic index (p=0.0017) and both first- and second-phases of insulin secretion (p=1.11×10−7 and 0.0321, respectively) (Figure 1).
In this study, we analyzed the effects of GCK, GCKR, G6PC2 and MTNR1B SNPs on the traits related to glucose metabolism. We replicated previous findings that GCK, G6PC2 and MTNR1B SNPs had individual and combined effects on fasting glucose in the Chinese populations. The effect sizes of the risk alleles were similar to what they were reported in other ethnics . However, we failed to find any effect of GCKR on fasting glucose although this association had been replicated in different samples , , , . One possible explanation for the negative finding in our samples is that the effect size of GCKR rs780094 (0.029mmol/l per allele) was relatively smaller than the other three SNPs (0.06–0.07mmol/l per allele) . Thus the statistical power of our samples (~60%) may not be enough. Two previous studies reported an interaction effect between GCKR rs780094 and GCK rs1799884 on fasting glucose, which was not observed in our samples either. But considering the variant showed opposite interaction effect in those two studies , , whether there was an interaction between SNPs from GCK and GCKR still remained to be investigated in additional studies with larger samples.
GCK and G6PC2 are enzymes related to glucose phosphorylation, which is the committed step of glucose metabolism. MTNR1B is the receptor of melatonin which inhibits insulin secretion through its effect on the formation of cGMP , , . Knock-out mice of these genes demonstrated significantly lower fasting glucose levels , , . There is no doubt that GCK, G6PC2 and MTNR1B variants impaired beta cell function, which was shown by current study as well as previous reports , , , . However, our data additionally suggested heterogeneity may exist for how these genes impaired beta cell function. We found GCK and G6PC2 variants affected both first- and second-phases insulin secretion, but MTNR1B variants only showed an association to first-phase insulin secretion in the same samples. MTNR1B participates in insulin secretion through its inhibitory effect on cGMP formation when activated by melatonin . But the underlying mechanism how MTNR1B regulated first- and/or second- phases of insulin secretion is still elusive.
Among the four genes studied, we found only GCKR rs780094 and G6PC2 rs16856187 affected type 2 diabetes risk in our samples. Although GCKR was also associated to triglyceride levels, its association to type 2 diabetes was independent of triglyceride (p=0.0001 after adjusting age, gender, BMI and triglyceride levels as the confounding factors). MTNR1B rs10830963 was previously reported to be associated with type 2 diabetes susceptibility in the Chinese , but we failed to replicate this association even we had sufficient statistical power (over 95%) to detect the reported effect. A previous study  also reported these loci showed a combined effect on type 2 diabetes susceptibility and led to a younger age at diagnosis. However, neither the combined effect on type 2 diabetes susceptibility nor the association with younger age at diagnosis was replicated in our study (data not shown). It is interesting that type 2 diabetes genes (e.g., KCNQ1 and SLC30A8) showed limited impacts on fasting glucose in the normoglycemic populations, although they also impaired insulin secretion , . How these genes regulate glucose metabolism and type 2 diabetes susceptibility remained to be elucidated by functional studies.
Although we carried out this replication study in large samples and indicated GCK, G6PC2 and MTNR1B had different effects on first- and second-phases of insulin secretion, there are limitations in our study. First, the two phases of insulin secretion were derived from fasting and 120 min glucose and insulin levels, the estimation would be more accurate if OGTT 30 and 60 min blood samples were available. However, participants with 30 min blood samples were limited in our study. Second, we didn't adjust lifestyle (e.g., alcohol consumption and smoking) as confounding factors. Whether there is interaction between lifestyle and these genetic variants on glucose metabolism still remained unknown.
In conclusion, we showed that GCK, G6PC2 and MTNR1B variants modulated fasting glucose levels while G6PC2 and GCKR variants were associated with type 2 diabetes in the Shanghai Chinese. Moreover, we found GCK and G6PC2 genetic variants were associated to both first- and second-phases insulin secretion while MTNR1B genetic variant was associated to first-phase insulin secretion, but not second-phase insulin secretion.
We thank all the participants of this research. We particularly thank Huijuan Lu, Xiaosi Li and Jing Xu for their outstanding technical assistance. We thank all nursing and medical staff at Shanghai Clinical Center for Diabetes for their dedication in this study.
Competing Interests: The authors have declared that no competing interests exist.
Funding: This work was supported by the “Chen Guang” project (09CG07) from Shanghai Municipal Education Commission and Shanghai Education Development Foundation, National Natural Science Foundation of China (30630061 and 30800617), National 973 Program (2006CB503901), National 863 Program (2006AA02A409), Shanghai Rising-Star Program (09QA1404400), major program of Shanghai Municipality for Basic Research (08dj1400601), Shanghai Key Laboratory of Diabetes Mellitus (08DZ2230200) and Shanghai Municipal Hospitals' Project of Chronic Disease Prevention and Treatment (SHDC12007316). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.