| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Institute of Epidemiology (T.I., F.B., A.S., B.T., C.V.) and Institute of Health Economics and Health Care Management (R.H.), GSF National Research Centre for Environment and Health, D-85764 Neuherberg, Germany; German Diabetes Centre (S.M.-S., H.K., C.H.) and Department of Biometrics and Epidemiology (W.R.), German Diabetes Research Institute at the University of Düsseldorf, D-40225 Düsseldorf, Germany; and Department of Internal Medicine II-Cardiology, University of Ulm, Medical Centre (W.K.), D-89081 Ulm, Germany
Address all correspondence and requests for reprints to: Wolfgang Koenig, M.D., Department Internal Medicine II-Cardiology, University of Ulm, Medical Centre, Oberer Eselsberg, Robert-Koch-Strasse 8, D-89081 Ulm, Germany. E-mail: wolfgang.koenig{at}medizin.uni-ulm.de.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
28.7 kg/m2, i.e. study median) but not in women or more obese persons. Circulating IL-6 levels were not associated with the IL-6 polymorphisms, but significantly elevated levels of the chemokine monocyte chemoattractant protein-1/CC chemokine ligand 2 in carriers of the protective genotypes indicated an indirect effect of these single nucleotide polymorphisms on the innate immune system. Our findings confirm that immune gene polymorphisms can be considered as independent risk factors in the etiology of type 2 diabetes and suggest that their contribution may be indirect, by influencing the levels of other immune mediators like monocyte chemoattractant protein-1. |
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Biochemical analyses
Concentrations of C-reactive protein (CRP), serum amyloid A (SAA), and fibrinogen in plasma were determined by nephelometric assays as described earlier (3, 10, 11). Serum concentrations of cytokines were determined by rigidly evaluated sandwich ELISA; lipid diagnostics were done by standard procedures (3).
Genotyping
Genomic DNA was extracted from leukocytes with a commercial DNA isolation kit (Gentra Systems, Minneapolis, MN) according to the manufacturer’s recommendation. Genotyping analyses were carried out by using the MassARRAY system (Sequenom, San Diego, CA). Briefly, genomic DNAs were amplified by PCR using HotStarTaq DNA Polymerase (Qiagen, Hilden, Germany). Genotyping assays were carried out using 5 ng genomic DNA. PCR primers were used at 167-nM final concentrations for a PCR vol of 6 µl. The PCR condition was 95 C for 15 min for hot start, followed by denaturing at 95 C for 30 sec, annealing at 56 C for 30 sec, extension at 72 C for 1 min for 44 cycles, and finally incubation at 72 C for 10 min. PCR products first were treated with shrimp alkaline phosphatase (Amersham, Freiburg, Germany) for 20 min at 37 C to remove excess deoxynucleotide triphosphates and afterward for 10 min at 85 C to inactivate shrimp alkaline phosphatase. ThermoSequenase (Amersham) was used for the base extension reactions. Extension primers were used at a final concentration of 5.4 µM in 10-µl reactions. The base extension reaction condition was 94 C for 2 min, followed by 94 C for 5 sec, 52 C for 5 sec, and 72 C for 5 sec for 40 cycles. The final base extension products were treated with SpectroCLEAN resin (Sequenom) to remove salts in the reaction buffer. This step was carried out with a 96-channel autopipette, and 16 µl resin-water suspension was added into each base extension reaction, resulting in a total vol of 26 µl. After quick centrifugation, 10 nl reaction solution was dispensed onto a 384 format SpectroCHIP (Sequenom) prespotted with a matrix of 3-hydroxypicolinic acid using a SpectroPoint nanodispenser (Sequenom). A modified Bruker Biflex matrix-assisted laser desorption ionization-time-of-flight mass spectrometer (Sequenom) was used for data acquisition from the SpectroCHIP. Genotyping calls were made in real time with MASSARRAY RT software (Sequenom).
Statistical analysis
Association of IL-6 genotypes and type 2 diabetes was analyzed by conditional logistic regression of diabetes patients vs. age- and sex-matched controls. Conditional logistic regression was carried out with the SAS procedure PROC PHREG (SAS Institute, Cary, NC), which performs an automatic frequency matching of the data. Genotypic data entered the regression model in the form of two dummy variables representing the effect of genotype –174C/G and genotype –174G/G vs. the reference category –174C/C, respectively. The same was done for A-598G, where –598A/A was the reference group. In a second approach, genotype data were introduced as a three-categorical variable with the proportional odds assumption that reflects the additive allele effect. Interactions were modeled with interaction terms as well as with subgroup analyses. The components of the metabolic syndrome that showed no major deviation from the normal distribution [waist circumference, body mass index (BMI), total cholesterol, low-density lipoprotein (LDL) cholesterol, high-density lipoprotein (HDL) cholesterol, and leukocyte count] were described by means ± SD in Table 1
and by means with confidence intervals adjusted for the covariables age, sex, and diabetes status (see Table 4). The P values reflecting differences between groups of diabetes, IGT, and control probands and between groups of different genotypes resulted from ANOVA models. Components of subclinical inflammation [IL-6, CRP, SAA, fibrinogen, monocyte chemoattractant protein-1 (MCP-1)/CC chemokine ligand 2 (CCL2), and macrophage inflammatory protein-1
/MIP-1/CCL3] were described by median and interquartile range, and differences between groups were analyzed by the nonparametric Kruskal-Wallis test. The remaining components of the metabolic syndrome [fasting insulin, homeostasis model assessment (HOMA)-insulin resistance, and fasting triglycerides] were treated likewise because of clear deviation from the normal distribution. A P value < 0.05 was considered statistically significant. All calculations were performed with the software package SAS 8.
|
|
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
shows the association of diabetes status with parameters of the metabolic syndrome and with inflammatory markers. For the C-174G SNP, the frequencies of the CC, CG, and GG genotypes were 0.174, 0.508, and 0.318, respectively, with allele frequencies of 0.428 for C-174 and 0.572 for –174G. For the A-598G SNP, almost identical results were obtained with frequencies of the AA, AG, and GG genotypes of 0.180, 0.507, and 0.313, respectively, and allele frequencies of 0.433 for A-598 and 0.567 for –598G. Analysis of linkage disequilibrium (LD) revealed that the C-174G and A-598G SNPs were in LD in 97.5% of subjects. A bias due to the genotyping rates of 92.3% (C-174G) and 86.2% (A-598G) appears unlikely because both SNPs were in Hardy Weinberg equilibrium and the data are consistent with previously published results for comparable Caucasian populations regarding allele frequencies (12, 13, 14) and LD (15).
Conditional logistic regression demonstrated that both C-174G and A-598G SNPs exhibited a statistically significant association with diabetes. Because of the high degree of LD and the higher call rate for the C-174G SNP, only data for the C-174G SNP are shown. The frequencies for the genotypes –174C/C, C/G, and G/G were 0.196, 0.536, and 0.268, respectively, for the controls; 0.181, 0.476, and 0.343, respectively, for subjects with IGT; and 0.144, 0.509, and 0.347, respectively, for patients with type 2 diabetes. The genotype frequencies for the A-598G SNP were almost identical.
With the reference category of –174C/C, conditional logistic regression yielded an odds ratio (OR) of 1.32 with a 95% confidence interval (CI) of 0.78–2.23 (P = 0.31) for genotype C/G and an OR of 1.81 (CI, 1.02–3.21; P = 0.044) for genotype G/G (Table 2). The additional consideration of BMI as covariable increased the OR in both genotypes C/G and G/G vs. C/C (genotype C/G: OR = 1.61, CI = 0.89–2.94, P = 0.12; genotype G/G: OR = 2.34, CI = 1.22–4.47, P = 0.010) (Table 2).
|
).
Subanalyses showed that the association of the C-174G SNP with diabetes status reached statistical significance in men (OR, 1.46; P = 0.042) but not in women (OR, 1.20; P = 0.42) and also in leaner subjects (BMI median, 28.7 kg/m2 of KORA Survey 2000) compared with subjects with BMI > 28.7 kg/m2 (OR, 1.67; P = 0.022 vs. OR, 1.36, P = 0.14). When physical activity was considered as a covariable in addition to age, sex, and BMI, the significant association of the C-174G SNP with type 2 diabetes persisted (OR, 1.44; CI, 1.05–1.99; P = 0.023). Further subdivision of the study group by sex and BMI confirmed that this association remained significant for men with BMI 28.7 kg/m2 (OR, 1.71) but not for men with BMI > 28.7 kg/m2 or for all women regardless of BMI.
However, the interaction between BMI and genotype that might be proposed by the subgroup analysis of Table 2 cannot be verified on a statistically significant basis (OR, 0.83; CI, 0.46–1.50; P = 0.54; Table 3). A model containing an interaction term for sex and IL-6 genotype also shows no significant interaction (data not shown).
|
The groups with genotypes –174C/C, C/G, and G/G showed a marginally significant impact of genotype on BMI but no difference regarding other key parameters characterizing the metabolic syndrome, such as waist circumference, total cholesterol, LDL cholesterol, HDL cholesterol, leukocyte count, insulin, insulin resistance (as estimated by the HOMA, and fasting triglycerides (the latter two parameters determined only in individuals without a history of diabetes) (Table 4). Conditional logistic regression showed that there was no association of the C-174G SNP with hypertension (defined as blood pressure 
140/90 or antihypertensive treatment) or smoking habits (data not shown).
The analysis of inflammatory mediators demonstrated that the protective –174C/C genotype showed a trend for an association only with increased levels of MCP-1 (P = 0.065), whereas no differences could be found for IL-6, CRP, SAA, fibrinogen, and MIP-1/CCL3 (Table 4). Because diabetes and diabetes-related factors, such as poor glycemic control and antidiabetic therapy, might affect these variables, the analyses were repeated after exclusion of all patients with type 2 diabetes. Median levels of MCP-1 were almost identical for the genoypes –174C/C, C/G, and G/G with 299.0 pg/ml (n = 81), 291.5 pg/ml (n = 218), and 274.6 pg/ml (n = 131), respectively (P = 0.14).
We investigated the serum levels of MCP-1/CCL2 separately in both men and women and also in leaner and more obese subjects. As shown in Table 5, the association of the C-174G SNP with type 2 diabetes closely reflected the association of this SNP with MCP-1 levels. For the protective –174C/C genotype, both men and subjects with BMI 28.7 kg/m2 had increased MCP-1 levels. This increase was most pronounced when both variables were combined, i.e. in men with BMI 28.7 kg/m2 (median, 382.6 pg/ml compared with 292.5 pg/ml in all –174C/C carriers). These findings correlated with the diabetes prevalence among carriers of the –174C/C genotype, which was 27.4% for the whole study group, but was reduced to 14.3% in men with BMI 28.7 kg/m2. The respective values for –174G/G carriers were 36.2% (study group) and 35.0% (men with BMI 28.7 kg/m2). After exclusion of individuals with type 2 diabetes, median MCP-1 levels in men with BMI 28.7 kg/m2 were not altered significantly and reached 399.2 pg/ml for the genotype –174C/C (n = 30), 290.2 pg/ml for C/G (n = 74), and 296.8 pg/ml for G/G (n = 39) (P = 0.051).
|
and 5, were also performed for the A-598G SNP. Due to the high degree of LD, results comparable with the C-174G SNP would have to be expected and were indeed found. |
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Therefore, we used the population-based KORA Survey 2000 to reevaluate the relevance of IL-6 SNPs for type 2 diabetes. We extended our analysis by testing for influence of sex, BMI, and impact on parameters of the metabolic syndrome and subclinical systemic inflammation, which might indicate a potential mediator that links the IL-6 gene SNPs with diabetes. First of all, our results demonstrate the association of the C-174G and A-598G SNPs with type 2 diabetes in a German Caucasian population. We did not find an association of the IL-6 SNPs with insulin resistance (HOMA) and IGT, which might indicate that the association with type 2 diabetes may have been a false positive. However, this appears very unlikely because our main finding is confirmed by the study by Vozarova et al. (7), which describes the association of the –174G allele with type 2 diabetes independently in smaller populations of Native Americans and Spanish Caucasians. Our study considerably extends the previous findings by subgroup analyses that indicate that the association of IL-6 genotype and diabetes is statistically significant in men and lean to moderately overweight subjects, i.e. that the G alleles of the C-174G and A-598G SNPs might be associated with an elevated diabetes risk, whereas for women and more obese individuals, the SNPs do not appear to affect diabetes risk. Interaction models, however, cannot confirm the association of genotype and BMI or sex. Concerning the association of IL-6 genotype and type 2 diabetes, the outcome of our study differs from the DPS (16). However, it should be noted that the DPS participants were selected for IGT, which led to a study group being more obese than the KORA participants, which therefore might have possessed additional risk factors that could have obscured the impact of the SNP. In addition, the comparison between subjects with IGT who progressed or did not progress to diabetes might not be as suitable to detect subtle risk factors as our comparison of diabetes patients, subjects with IGT, and age- and sex-matched healthy controls.
We were not able to demonstrate an association of the SNPs with insulin resistance and IGT tolerance, which precede the development of overt type 2 diabetes, both in the whole study group and in the subgroups divided by sex and/or BMI. Thus, the results from our large, population-based study do not confirm the aforementioned associations of –174 genotypes with insulin resistance (17, 18). Importantly, the subanalyses in our study emphasize that immune gene polymorphisms may be considered as risk factors only in certain population segments and that comprehensive studies with a high number of participants may indeed be required to detect SNP-related increased risks in these groups, whereas smaller studies might include patient selection-inherent biases that can skew the outcome and do not reflect the general risk of a population.
The analysis of type 2 diabetes-related inflammatory markers could not reveal a direct mechanism of the IL-6 SNPs on circulating IL-6 levels. Although some data indicate that there might be an association between IL-6 –174C/G genotype and circulating IL-6 levels (5, 15), our study is in line with the majority of studies on IL-6 SNPs, which do not observe a significant impact of –174G/C genotypes on IL-6 levels in serum (12, 14, 18, 19). Interestingly, it has been shown that the regulatory effect of SNPs within the IL-6 promoter on IL-6 expression is cell-type specific (6). It is therefore reasonable to hypothesize that the impact of IL-6 SNPs on IL-6 expression by various cell types might depend on parameters such as acute inflammation or obesity (and therefore also on the study group), which exert different effects on IL-6-producing cells. The view that as-yet-undefined parameters can enhance or suppress the impact of SNPs on IL-6 levels is supported by the observation that the –174C allele was found to be associated with significantly elevated IL-6 expression after coronary artery bypass graft surgery; whereas, regarding the baseline IL-6 levels, no difference between the genotypes could be detected before surgery (15). It cannot be excluded that the presence of carrier proteins or soluble receptors in the blood, such as soluble glycoprotein 130 (20), might have obscured a subtle, but biologically relevant SNP-mediated genetic predisposition to slightly augmented IL-6 expression (17).
Our investigations, however, demonstrate that the stronger association of the C-174G SNP in subgroups of the study cohort is closely mirrored by elevation of MCP-1/CCL2 blood concentration in carriers of the protective genotype –174C/C. Because elevated levels of MCP-1 have recently been shown to be associated with a lower risk of diabetes (H. Kolb, C. Herder, S. Müller-Scholze, unpublished observations), MCP-1 appears a potential candidate to link IL-6 gene polymorphisms with diabetes status by an as-yet-unknown mechanism. This mechanism of MCP-1 up-regulation seems to be prone to disturbances by obesity and sex-related influences and might thus be suppressed in women and obese subjects. In this context, it should be noted that individuals taking antiinflammatory drugs were not excluded from our study because, in the age group studied, a large proportion has been taking statins and antihypertensive and antidiabetic compounds. Their impact appears limited, because systemic levels of inflammatory mediators are elevated in individuals with type 2 diabetes, metabolic syndrome, or atherosclerosis (1, 21) despite their more frequent use of these drugs.
Taken together, this study demonstrates that the IL-6 gene SNPs C-174G and A-598G are associated with type 2 diabetes and indicates that the chemokine MCP-1 may act as potential protective mediator, although the mechanism still remains to be elucidated. The SNP-related association reached statistical significance in men and in the absence of obesity, thereby underlining the relevance of comprehensive, population-based studies that allow for subgroup analysis.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
The KORA Group consists of H. E. Wichmann (speaker), H. Löwel, C. Meisinger, T. Illig, R. Holle, J. John, and their co-workers who are responsible for the design and conduct of the KORA studies.
Abbreviations: BMI, Body mass index; CCL, CC chemokine ligand; CI, 95% confidence interval; CRP, C-reactive protein; DPS, Diabetes Prevention Study; HDL, high-density lipoprotein; HOMA, homeostasis model assessment; IGT, impaired glucose tolerance; KORA, Kooperative Gesundheitsforschung im Raum Augsburg/Cooperative Research in the Region of Augsburg; LD, linkage disequilibrium; LDL, low-density lipoprotein; MCP-1, monocyte chemoattractant protein-1; MIP-1, macrophage inflammatory protein-1; OR, odds ratio; SAA, serum amyloid A; SNP, single nucleotide polymorphism.
Received February 24, 2004.
Accepted June 28, 2004.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
or its receptors. Diabetologia 45:805–812[CrossRef][Medline]
(G-308A) and IL-6 (C-174G) genes predict the conversion from impaired glucose tolerance to type 2 diabetes. Diabetes 52:1872–1876This article has been cited by other articles:
 |
|
 |
|
  C Holzapfel, N Klopp, H Grallert, C Huth, C Gieger, C Meisinger, K Strassburger, G Giani, H E Wichmann, H Laumen, et al. Genetic variants in the leukemia-associated Rho guanine nucleotide exchange factor (ARHGEF12) gene are not associated with T2DM and related parameters in Caucasians (KORA study) Eur. J. Endocrinol., September 1, 2007; 157(3): R1 - R5. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Qi, C. Zhang, R. M. van Dam, and F. B. Hu Interleukin-6 Genetic Variability and Adiposity: Associations in Two Prospective Cohorts and Systematic Review in 26,944 Individuals J. Clin. Endocrinol. Metab., September 1, 2007; 92(9): 3618 - 3625. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. J. Willer, L. L. Bonnycastle, K. N. Conneely, W. L. Duren, A. U. Jackson, L. J. Scott, N. Narisu, P. S. Chines, A. Skol, H. M. Stringham, et al. Screening of 134 Single Nucleotide Polymorphisms (SNPs) Previously Associated With Type 2 Diabetes Replicates Association With 12 SNPs in Nine Genes Diabetes, January 1, 2007; 56(1): 256 - 264. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Huth, I. M. Heid, C. Vollmert, C. Gieger, H. Grallert, J. K. Wolford, B. Langer, B. Thorand, N. Klopp, Y. H. Hamid, et al. IL6 Gene Promoter Polymorphisms and Type 2 Diabetes: Joint Analysis of Individual Participants' Data From 21 Studies. Diabetes, October 1, 2006; 55(10): 2915 - 2921. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Wernstedt, A. Edgley, A. Berndtsson, J. Faldt, G. Bergstrom, V. Wallenius, and J.-O. Jansson Reduced stress- and cold-induced increase in energy expenditure in interleukin-6-deficient mice Am J Physiol Regulatory Integrative Comp Physiol, September 1, 2006; 291(3): R551 - R557. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Qi, R. M. van Dam, J. B. Meigs, J. E. Manson, D. Hunter, and F. B. Hu Genetic variation in IL6 gene and type 2 diabetes: tagging-SNP haplotype analysis in large-scale case-control study and meta-analysis Hum. Mol. Genet., June 1, 2006; 15(11): 1914 - 1920. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Herder, S. Muller-Scholze, P. Rating, W. Koenig, B. Thorand, B. Haastert, R. Holle, T. Illig, W. Rathmann, J. Seissler, et al. Systemic monocyte chemoattractant protein-1 concentrations are independent of type 2 diabetes or parameters of obesity: results from the Cooperative Health Research in the Region of Augsburg Survey S4 (KORA S4) Eur. J. Endocrinol., February 1, 2006; 154(2): 311 - 317. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Herder, H. Kolb, W. Koenig, B. Haastert, S. Muller-Scholze, W. Rathmann, R. Holle, B. Thorand, and H.-E. Wichmann Association of Systemic Concentrations of Macrophage Migration Inhibitory Factor With Impaired Glucose Tolerance and Type 2 Diabetes: Results from the Cooperative Health Research in the Region of Augsburg, Survey 4 (KORA S4) Diabetes Care, February 1, 2006; 29(2): 368 - 371. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Justenhoven, U. Hamann, C. B. Pierl, S. Rabstein, B. Pesch, V. Harth, C. Baisch, C. Vollmert, T. Illig, T. Bruning, et al. One-Carbon Metabolism and Breast Cancer Risk: No Association of MTHFR, MTR, and TYMS Polymorphisms in the GENICA Study from Germany Cancer Epidemiol. Biomarkers Prev., December 1, 2005; 14(12): 3015 - 3018. [Full Text] [PDF]
|
|
|
|
|
|
C. Herder, B. Haastert, S. Muller-Scholze, W. Koenig, B. Thorand, R. Holle, H.-E. Wichmann, W. A. Scherbaum, S. Martin, and H. Kolb Association of Systemic Chemokine Concentrations With Impaired Glucose Tolerance and Type 2 Diabetes: Results from the Cooperative Health Research in the Region of Augsburg Survey S4 (KORA S4) Diabetes, December 1, 2005; 54(suppl_2): S11 - S17. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
O. P. Kristiansen and T. Mandrup-Poulsen Interleukin-6 and Diabetes: The Good, the Bad, or the Indifferent? Diabetes, December 1, 2005; 54(suppl_2): S114 - S124. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Cardellini, L. Perego, M. D'Adamo, M. A. Marini, C. Procopio, M. L. Hribal, F. Andreozzi, S. Frontoni, M. Giacomelli, M. Paganelli, et al. C-174G Polymorphism in the Promoter of the Interleukin-6 Gene Is Associated With Insulin Resistance Diabetes Care, August 1, 2005; 28(8): 2007 - 2012. [Abstract] [Full Text] [PDF]
|
|
Read all eLetters
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
