Variation in the Interleukin-6 Receptor Gene Associates With Type 2 Diabetes in Danish Whites

The Asp358Ala polymorphism was genotyped in the Inter99 cohort of middle-aged subjects, in a population-based sample of elderly glucose-tolerant subjects, and in a group of type 2 diabetic patients recruited from the outpatient clinic at Steno Diabetes Center. Also, a subgroup of glucose-tolerant subjects who was age and sex matched to type 2 diabetic patients was selected from the total group of glucose-tolerant subjects. All subjects who participated in the present protocol were living in the same area of Copenhagen.

The Inter99 cohort is a population-based, randomized, nonpharmacological intervention study for the prevention of cardiovascular disease undertaken at the Research Centre for Prevention and Health in Glostrup, Copenhagen County. Details of this cohort have been reported (22). Glucose tolerance status was evaluated according to the 1999 World Health Organization criteria (23). Oral glucose tolerance and genotype data were available on 5,702 white subjects. Of those, 4,251 subjects had normal glucose tolerance (NGT), 653 had impaired glucose tolerance (IGT), 474 subjects had impaired fasting glucose (IFG), 231 patients had screen-detected type 2 diabetes, and 93 patients had known type 2 diabetes.

The case-control study of type 2 diabetes involved 1,349 case subjects and all glucose-tolerant subjects from the Inter99 cohort, and the group of elderly glucose-tolerant subjects served as control (n = 4,596). Clinical characteristics of the study groups included in the present investigation are given in Table 1. Patients with diabetes due to known chronic pancreatitis, hemochromatosis, severe insulin resistance, maturity-onset diabetes of the young, or maternally inherited diabetes and deafness; patients with a family history of first-degree relatives with type 1 diabetes; patients with insulin requirement within the 1st year after diabetes diagnosis; or patients with a fasting serum C-peptide level ≤150 pmol/l at the time of recruitment were excluded in the present study from the category of clinically defined type 2 diabetes.

All study participants were Danish whites by self-report. Informed written and oral consent were obtained before participation. The studies were approved by the Ethical Committee of Copenhagen in accordance with the Helsinki Declaration II.

Blood samples for analyses of biochemical variables were drawn in the morning after an overnight fast. Plasma glucose was analyzed by a glucose oxidase method (Granutest; Merck, Darmstedt, Germany) and serum-specific insulin [excluding des(31,32) and intact proinsulin] and serum C-peptide were measured by time-resolved fluoroimmunoassay (AutoDELFIA; Perkin Elmer-Wallac, Turku, Findland). HbA_1c was analyzed by principles of ion exchange high-performance liquid chromatography using Bio-Rad variant HbA_1c (normal range 4.1–6.4%). The concentration of serum triglycerides was determined by the triglycerides GPO-PAP (glycerol phosphate oxidase para amino phenazone) method, HDL cholesterol by the HDL cholesterol plus method, and the serum total cholesterol by the cholesterol CHOD-PAP (cholesterol oxidase para amino phenazone) method (Roche Diagnostics, Mannheim, Germany).

The insulinogenic index was calculated as fasting serum insulin (in picomoles per liter) subtracted from 30-min post–oral glucose tolerance test (OGTT) serum insulin (in picomoles per liter) and divided by 30-min post-OGTT plasma glucose (in millimoles per liter). Homeostasis model assessment for insulin resistance was calculated as fasting plasma glucose (in millimoles per liter) multiplied by fasting serum insulin (in picomoles per liter) and divided by 22.5.

DNA was isolated from leukocytes using standard procedures. The genotyping method used for detection of the IL6R Asp358Ala polymorphism (rs8192284) was a chip-based matrix-assisted laser desorption/ionisation time-of-flight mass spectrometry (DNA MassARRAY) analysis of PCR-generated primer extension products as previously described (24). The genotyping success rate was 94%, and among 89 replicate samples there were no mismatches. The genotypes observed in both the Inter99 cohort and in the type 2 diabetic group were in Hardy-Weinberg equilibrium.

Fisher’s exact test was applied to test for significance of differences in allele frequencies, and logistic regression analyses (with sex and age as covariates) were applied to evaluate differences in genotype distributions. Odds ratios (ORs) were calculated after adjustment for sex and age. Phenotypic differences between the genotype groups among the NGT subjects were tested with a general linear model including sex and genotype as fixed factors and age and BMI as covariates. For genotype quantitative trait studies, normal distribution of the residuals was verified and logarithmically transformed if appropriate. A two-sided P value <0.05 was considered significant in case-control studies. Because the examination for the pre-diabetic quantitative traits had an explorative character, all significant levels were considered after Bonferroni correction for multiple testing (25). All analyses were done using SPSS for Windows, version 12.0.

In the case-control study we found significant differences in allele frequency and genotype distribution of the Asp358Ala polymorphism of the IL6R between 1,349 type 2 diabetic patients and 4,596 glucose-tolerant subjects applying a codominant as well as a recessive and dominant model of inheritance (Table 2). The Asp358-allele was more frequent among type 2 diabetic than control subjects (Table 2). The genotype and allele frequencies in the various examined subgroups are given in online appendix Table 1 (available from http://diabetes.diabetesjournals.org). For Asp/Asp carriers versus Ala/Ala carriers, the OR was 1.38 (95% CI 1.09–1.71) (Table 2). When the analysis was restricted to 1,365 glucose-tolerant subjects (794 men and 571 women) who were age and sex matched to the type 2 diabetic patients, a significant association of the Asp358 allele with type 2 diabetes was also found. Genotype distribution and allele frequencies of this subset of glucose-tolerant subjects were as follows: Asp/Asp 34.8%, Asp/Ala 47.8%, and Ala/Ala 17.4% (Ala-allele 41.3% [39.4–43.1]; P = 0.02).

The impact of the Asp358Ala variant on quantitative traits of metabolism was evaluated in a study involving the group of 4,251 glucose-tolerant subjects from the Inter99 cohort. After correction for multiple testing, the variant did not associate with levels of plasma glucose and serum insulin in the fasting state, during an OGTT, or with indexes of insulin resistance or insulin secretion. Similarly, the variant was not associated with alterations in serum lipids or indexes of body weight and body composition (Table 3).

Our finding that the Asp-allele of the Asp358Ala variant of the IL6R associates with type 2 diabetes among white Danes suggests that this polymorphism may contribute to the genetic susceptibility of common diabetes. The magnitude of this association, i.e., the OR, is, as reviewed by McCarthy (17), comparable with the assessed significance of other reported genetic variants conferring an increased risk for type 2 diabetes.

Whether our observation reflects a true diabetogenic impact of the Asp358Ala polymorphism of IL6R gene per se cannot be determined from the present study, and attempts for replication are obviously needed. In a preliminary communication of a study among U.S. whites, a similar association of the Asp358Ala variant of the IL6R to type 2 diabetes was reported (18), whereas the same variant was associated with obesity but not with diabetes among Pima Indians (20). Other IL6R functional variant(s) in high linkage disequillibrium with the Asp358Ala variant might explain the present association. However, previous molecular studies indicate that individuals carrying the Ala-allele at codon-358 may have reduced circulating levels of the soluble form of IL6R. This receptor, sIL6R, is released from the cell surface after proteolytic cleavage, called shedding, in the Gln357/Asp358 juncture (21,26). Thus, the Gln357/Asp358 juncture is obviously an important region for the IL6R cleavage, and interestingly, it has been reported that substitution with glycine at codon 358 reduces shedding to 26% (21). Furthermore, an Ala357/Ala358 double mutation attenuated shedding by 66% (21). These findings may indicate that carriers of the Ala358 allele have impaired proteolytic cleavage and thereby lower sIL6R production (21), which in turn might cause reduced IL-6 signal transduction. Hence, the diminished risk of type 2 diabetes linked with the presence of the Ala358 allele might be mediated by reduced IL-6 signal transduction or by reduced signaling of other cytokines that also mediate their biological impact via the IL6R (8).

To elucidate the potential patophysiological mechanisms underlying the association of the Asp358Ala variant and type 2 diabetes, we investigated the effect of the polymorphism on metabolic phenotypes in groups of NGT IFG, and IGT subjects. However, after Bonferonni correction we failed to show any significant associations.

In contrast to a recent study of 786 Pima Indians that showed an association of the codon 358 IL6R variant to obesity but not to diabetes (20), we failed to show any impact of the IL6R variant on estimates of obesity in any of our study groups. The reason for this inconsistency is unclear but might be explained by differences in study populations. In this respect, the marked difference in allele frequency of the Asp358Ala polymorphism in the two populations is also emphasized. Furthermore, complex interactions between ethnic-specific polymorphisms in the promoter regions of IL6R, IL-6, or in the complex of IL6-IL6R signaling proteins might contribute to this discrepancy.

Based on simulation studies, it was estimated that the present study has a power of 90% to detect a relative risk of 1.2 for an association with type 2 diabetes or obesity. Therefore, if the variant possesses an effect on obesity risk in the Danish population, the effect size is relatively subtle.

It is well known that spurious associations can even be found in well-designed case-control studies. However, we believe that the risk that our findings of an association with type 2 diabetes are due to population stratification is relatively modest, as all studied subjects are Danish whites and all have been collected from the same region of the western part of Copenhagen County. In addition, >20 other SNPs have been analyzed in the same study samples, and no associations with type 2 diabetes were found. Nevertheless, attempts to validate the present finding of an association to diabetes are essential.

In conclusion, the common Asp358Ala polymorphism of the IL6R gene associates with an increased risk of type 2 diabetes, with an OR for the Asp-allele carriers of 1.30 in Danish whites. This finding needs replication in other study populations, especially due to the lack in the present study of an association with metabolic phenotypes in middle-aged glucose-tolerant subjects.

The study was supported by grants from the Danish Diabetes Association, the Danish Medical Research Council, and The Danish Heart Foundation.

The authors thank Annemette Forman, Lene Aabo, Inge-Lise Wantzin, and Marianne Stendal for technical assistance and Grete Lademann for secretarial support.