Abstract
OBJECTIVE
To explore the association between epistasis among related genes of the renin-angiotensin system (RAS) and type 2 diabetes.
RESEARCH DESIGN AND METHODS
Gene polymorphisms were genotyped in 394 type 2 diabetic patients and 418 healthy control subjects in this case-control study. We used the multifactor dimensionality reduction method to identify gene-gene interactions.
RESULTS
No single locus was associated with type 2 diabetes, except for the insert/deletion (I/D) polymorphism of the ACE gene in female subjects. In multi-locus analyses, in male subjects the model of rs2106809 (ACE2), rs220721 (Mas), rs699 (AGT), and I/D (ACE) was significant (P = 0.043). This combination was associated with a 4.00 times (95% CI 2.51–6.38; P < 0.0001) greater prevalence of type 2 diabetes. In female subjects, the model of rs2106809 (ACE2), I/D (ACE), and rs1403543 (AGTR2) was significant (P = 0.012). This three-locus combination was associated with a 2.76 times (1.91–3.97; P < 0.0001) greater prevalence of type 2 diabetes.
CONCLUSIONS
Interactions among RAS-related genes were associated with type 2 diabetes in a Chinese population.
Over the past few years, a number of new genetic loci associated with type 2 diabetes have been uncovered based on genome-wide association scans. Investigations into gene-gene interactions, however, are uncommon because the method is computationally challenging (1). In type 2 diabetes, attempts to elucidate possible epistasis have provided only a few examples (2–4).
Recently, studies have supported the idea that using our understanding of biology, including that of cytokine networks and hormone systems, may help guide analysis of epistasis (5,6). Clinical evidence suggests that the renin-angiotensin system (RAS) is associated with the etiology of type 2 diabetes (7–9). However, the influence of genetic interaction within the RAS on type 2 diabetes susceptibility is still unknown. The aim of our study was to explore the contribution of epistasis among RAS-related genes.
RESEARCH DESIGN AND METHODS
The study subjects were selected from an ongoing large-scale population-based cohort (10). Participants without previously known diabetes were selected from the 2,826 registered individuals. Written informed consent was obtained from each participant. Subjects' fasting plasma glucose (FPG) was obtained from our previous study, and the subjects with FPG >5.6 mmol/l performed a 75-g oral glucose tolerance test. Diabetes was diagnosed according to the 1999 World Health Organization criteria. All subjects in both groups were matched for blood pressure, serum creatinine, and age.
Eight single nucleotide polymorphisms (SNPs) from seven RAS-related genes were then assessed; these were as follows: rs699 (AGT), insert/deletion (I/D) polymorphism (ACE), rs2106809 and rs2074192 (ACE2), rs5186 (AGTR1), rs1403543 (AGTR2), rs220721 (Mas), and rs1799722 (BDKRB2). Detection was completed using a MassARRAY platform (Sequenom, San Diego, CA).
We used the multifactor dimensionality reduction (MDR) software 2.0-β (http://www.multifactordimensionalityreduction.org) to identify gene-gene interactions.
RESULTS
A total of 394 unrelated type 2 diabetic patients and 418 healthy control subjects were enrolled in this case-control study. Demographic and clinical characteristics of the subjects are given in supplementary Table A1 in the online appendix available at http://care.diabetesjournals.org/cgi/content/full/dc10-0349/DC1. Age, blood pressure, and serum creatinine were comparable between the two groups. Because of the ACE2 and AGTR2 genetic presence in the X chromosome, the analysis was performed separately for male and female subjects.
Association analyses of male subjects
No single-locus analysis showed a significant association with type 2 diabetes (supplementary Table A2). According to the MDR analysis through five-locus comparisons, a significant interaction was observed for variant alleles in the following three loci: rs2106809 (ACE2), rs220721 (Mas), and I/D (ACE) (Table 1). This combination had the maximum cross-validation consistency of 10 that was significant at the 0.01 level of P value, as calculated using the sign test. The four-locus model of rs2106809 (ACE2), rs220721 (Mas), rs699 (AGT), and I/D (ACE) scored nine of cross-validation consistency that was significant at the 0.05 level of P value. The four-locus model was also significant in the 1,000 permutation test (Table 1). In the χ2 test, the odds ratio (OR) of high-risk combination of four loci increased the risk of type 2 diabetes by 4.00 times (95% CI 2.51–6.38, P < 0.0001).
Table 1.
MDR results of multi-locus interaction
| Best model | Testing balance accuracy | Cross-validation consistency | P | P* |
|---|---|---|---|---|
| Male subjects | ||||
| Mas | 0.4714 | 8 | 0.9893 | 0.804 |
| Mas, BDKRB2 | 0.4759 | 5 | 0.8281 | 0.375 |
| ACE2-A, ACE, Mas | 0.6195 | 10 | 0.0107 | 0.172 |
| ACE2-A, ACE, Mas, AGT | 0.5624 | 9 | 0.0547 | 0.043 |
| ACE2-A, ACE, Mas, AGT, BDKRB2 | 0.5399 | 6 | 0.1719 | 0.6010 |
| Female subjects | ||||
| ACE2-B | 0.5364 | 9 | 0.1719 | 0.516 |
| ACE2-A, ACE | 0.5567 | 9 | 0.1719 | 0.068 |
| ACE2-A, ACE, AGTR2 | 0.5892 | 10 | 0.0010 | 0.012 |
| ACE2-B, ACE, Mas, AGTR2 | 0.5319 | 6 | 0.1719 | 0.341 |
| ACE2-A, ACE, Mas, AGT, AGTR2 | 0.5586 | 9 | 0.0547 | 0.148 |
*P based on 1,000 permutations. Data in bold represent statistical significance. Genotypes: AGT (rs699) C/T; ACE (I/D); ACE2-A (rs2106809) G/A; ACE2-B (rs2074192) C/T; AGTR1 (rs5186) A/C; AGTR2 (rs1403543) A/G; Mas (rs220721) A/G; BDKRB2 (rs1799722) C/T.
Association analyses of female subjects
The results of the single-locus analyses showed that only I/D from ACE was associated with type 2 diabetes (P = 0.039) (supplementary Table A2). According to the MDR analysis, the most significant combination was the three-locus model, rs2106809 (ACE2), I/D (ACE), rs1403543 (AGT2R), which had the maximum cross-validation consistency of 10 and was significant at the 0.012 level of P value, as calculated using permutation test. In the χ2 test, the OR of high-risk combination of three loci increased the risk of type 2 diabetes by 2.76 times (95% CI 1.91–3.97, P < 0.0001).
The logistic regression model suggested a nonsignificant gene-gene interaction in a multiplicative manner in the male and female participants.
CONCLUSIONS
The results from our study evidenced that although main effects of the individual loci may not be observed, the interaction among RAS-related genes is directly correlated with the susceptibility of type 2 diabetes. It is thus possible that loci contribute to some complex diseases only by their interaction with other genes, although the main effects of the individual loci may be too small to be observed (11).
Identifying genes in multi-factorial diseases is difficult. There is no consensus as to the best strategy for detecting epistatic interactions in humans (12). In the present study, with MDR analysis we found interactions among RAS-related genes. These interactions make mechanistic sense because these genes are involved in the same biological pathways (13). However, the susceptibility interaction was not confirmed by the logistic regression analysis. A possible reason for these inconsistent results is that MDR did not detect the interaction defined by “deviation from the multiplicative” as in the logistic regression model. Only the significant results from MDR showed that the combination of different loci may increase or decrease the risk of disease (14). Based on the association OR of 1.3 (typical for type 2 diabetes) and allele frequency of 0.49, this study showed over 65% power to detect interactions of genes.
A limitation of our study is that by representing each gene locus with a single SNP, multiple association signals for a given gene might be missed (15). Furthermore, our findings need to be replicated in further studies with larger samples and different populations.
In conclusion, we were the first to show that the interactions among RAS-related genes are associated with type 2 diabetes.
Supplementary Material
Acknowledgments
This work was supported by two grants from the National Natural Science Foundation of China (30671001 and 30871887) and a grant from the National 863 Program of China (2006AA02A409) to J.-K.Y.
No potential conflicts of interest relevant to this article were reported.
J.-K.Y. researched data, contributed to the discussion, and reviewed/edited the manuscript. J.-B.Z. wrote the manuscript. Z.X., L.Z., M.Y., J.-P.F., H.Y., and Y.-H.M. researched data.
The authors thank all the participants and staff in the study.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
References
- 1.Moore JH, Ritchie MD: STUDENTJAMA. The challenges of whole-genome approaches to common diseases. JAMA 2004;291:1642–1643 [DOI] [PubMed] [Google Scholar]
- 2.Wiltshire S, Bell JT, Groves CJ, Dina C, Hattersley AT, Frayling TM, Walker M, Hitman GA, Vaxillaire M, Farrall M, Froguel P, McCarthy MI: Epistasis between type 2 diabetes susceptibility Loci on chromosomes 1q21-25 and 10q23-26 in northern Europeans. Ann Hum Genet 2006;70:726–737 [DOI] [PubMed] [Google Scholar]
- 3.Zuniga J, Romero V, Azocar J, Stern JN, Clavijo O, Almeciga I, Encinales L, Avendano A, Fridkis-Hareli M, Pandey JP, Yunis EJ: Interaction of KIR genes and G1M immunoglobulin allotypes confer susceptibility to type 2 diabetes in Puerto Rican Americans. Hum Immunol 2006;67:907–914 [DOI] [PubMed] [Google Scholar]
- 4.Wang Y, Zhang D, Liu Y, Yang Y, Zhao T, Xu J, Li S, Zhang Z, Feng G, He L, Xu H: Association study of the single nucleotide polymorphisms in adiponectin-associated genes with type 2 diabetes in Han Chinese. J Genet Genomics 2009;36:417–423 [DOI] [PubMed] [Google Scholar]
- 5.Askland K, Read C, Moore J: Pathways-based analyses of whole-genome association study data in bipolar disorder reveal genes mediating ion channel activity and synaptic neurotransmission. Hum Genet 2009;125:63–79 [DOI] [PubMed] [Google Scholar]
- 6.Emily M, Mailund T, Hein J, Schauser L, Schierup MH: Using biological networks to search for interacting loci in genome-wide association studies. Eur J Hum Genet 2009;17:1231–1240 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Yusuf S, Sleight P, Pogue J, Bosch J, Davies R, Dagenais G: Effects of an angiotensin-converting-enzyme inhibitor, ramipril, on cardiovascular events in high-risk patients. The Heart Outcomes Prevention Evaluation Study Investigators. N Engl J Med 2000;342:145–153 [DOI] [PubMed] [Google Scholar]
- 8.Kjeldsen SE, Dahlöf B, Devereux RB, Julius S, Aurup P, Edelman J, Beevers G, de Faire U, Fyhrquist F, Ibsen H, Kristianson K, Lederballe-Pedersen O, Lindholm LH, Nieminen MS, Omvik P, Oparil S, Snapinn S, Wedel H: LIFE (Losartan Intervention for Endpoint Reduction) Study Group Effects of losartan on cardiovascular morbidity and mortality in patients with isolated systolic hypertension and left ventricular hypertrophy: a Losartan Intervention for Endpoint Reduction (LIFE) substudy. JAMA 2002;288:1491–1498 [DOI] [PubMed] [Google Scholar]
- 9.NAVIGATOR Study Group McMurray JJ, Holman RR, Haffner SM, Bethel MA, Holzhauer B, Hua TA, Belenkov Y, Boolell M, Buse JB, Buckley BM, Chacra AR, Chiang FT, Charbonnel B, Chow CC, Davies MJ, Deedwania P, Diem P, Einhorn D, Fonseca V, Fulcher GR, Gaciong Z, Gaztambide S, Giles T, Horton E, Ilkova H, Jenssen T, Kahn SE, Krum H, Laakso M, Leiter LA, Levitt NS, Mareev V, Martinez F, Masson C, Mazzone T, Meaney E, Nesto R, Pan C, Prager R, Raptis SA, Rutten GE, Sandstroem H, Schaper F, Scheen A, Schmitz O, Sinay I, Soska V, Stender S, Tamás G, Tognoni G, Tuomilehto J, Villamil AS, Vozár J, Califf RM: Effect of valsartan on the incidence of diabetes and cardiovascular events. N Engl J Med 2010;362:1477–1490 [DOI] [PubMed] [Google Scholar]
- 10.Xin Z, Yuan J, Hua L, Ma YH, Zhao L, Lu Y, Yang JK: A simple tool detected diabetes and prediabetes in rural Chinese. J Clin Epidemiol 2010;63:1030–1035 [DOI] [PubMed] [Google Scholar]
- 11.Culverhouse R, Suarez BK, Lin J, Reich T: A perspective on epistasis: limits of models displaying no main effect. Am J Hum Genet 2002;70:461–471 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Chen X, Liu CT, Zhang M, Zhang H: A forest-based approach to identifying gene and gene gene interactions. Proc Natl Acad Sci U S A 2007;104:19199–19203 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Tsai CT, Hwang JJ, Ritchie MD, Moore JH, Chiang FT, Lai LP, Hsu KL, Tseng CD, Lin JL, Tseng YZ: Renin-angiotensin system gene polymorphisms and coronary artery disease in a large angiographic cohort: detection of high order gene-gene interaction. Atherosclerosis 2007;195:172–180 [DOI] [PubMed] [Google Scholar]
- 14.Cordell HJ: Epistasis: what it means, what it doesn't mean, and statistical methods to detect it in humans. Hum Mol Genet 2002;11:2463–2468 [DOI] [PubMed] [Google Scholar]
- 15.Perry JR, McCarthy MI, Hattersley AT, Zeggini E: Wellcome Trust Case Control Consortium Weedon MN, Frayling TM: Interrogating type 2 diabetes genome-wide association data using a biological pathway-based approach. Diabetes 2009;58:1463–1467 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.