A transition of C to G (rs1800796) at promoter region of IL6 gene is associated with lowered risk of elevated blood glucose in rural Thais

Background Inammation has been associated with increased risk of type 2 diabetes mellitus (T2DM). Thus, polymorphisms of genes encoding pro-inammatory cytokines may inuence the condition. This study aimed to determine the association between genetic variations in inammation-related genes, including rs2794521 of the CRP gene, rs1800796 of the IL6 gene, and rs1799964 of the TNF gene, and risk of elevated blood glucose. Methods A total of 296 subjects were allocated to either the control group (n=135) or elevated blood glucose group (n=161). The studied single nucleotide polymorphisms were investigated among 296 subjects by real-time PCR technique. Logistic regression model was used to evaluate the association between the genetic variations and risk of elevated blood glucose. Results Both dietary intake and concentrations of inammatory cytokines (CRP, IL6, and TNF-α) in the subjects with different genotypes were analogous. AG carriers of rs2794521 CRP showed the highest levels of obesity indices (BMI, waist circumference, and body fat; p<0.05 for all) compared to wild type. Homozygous variants of rs1800796 (GG) and rs1799964 (CC) were associated with signicantly lower HbA1c level ( p =0.041) and fasting blood glucose concentration ( p =0.043), respectively. An association with decreased risk of elevated blood glucose was found among GG female carriers of rs1800796 0.23 (0.06-0.87), p =0.031]. glucose was in AGT carriers 0.49 (0.024-1.02), p =0.058).


Background
The prevalence of type 2 diabetes mellitus (T2DM) has continually increased around the world. In 2017, there were approximately 425 million diabetic patients, which is expected to rise to 627 million in 2045 [1]. T2DM has been reported to be associated with in ammation [2,3], and obesity is hypothesized to trigger this in ammation. The expansion of adipose tissue eventually activates chronic low-grade in ammation [3][4]. This results in increased secretion of an acute phase reactant protein, namely Creactive protein (CRP) [5][6], as well as in ammatory cytokines such as interleukin 6 (IL6) [6][7] and tumor necrosis factor alpha (TNF-α) [8], which are known to affect glucose homeostasis. Since the expression of cytokines and the degree of in ammation can be modulated by genetic polymorphisms, it is thought that the polymorphisms of genes encoding for these cytokines may also in uence the development of T2DM.
Polymorphisms located in promoter regions, including rs2794521 (-717 A/G) of CRP gene, -572 C/G (rs1800796) of IL6 gene and -1031 T/C (rs1799964) of TNF gene, have been associated with T2DM risk [7,[9][10][11][12][13]. Several studies demonstrated a notable in uence of the rs2794521 polymorphism on CRP secretion [14][15][16]. Alteration of CRP concentration may modify T2DM risk [5][6]9]. Similarly, the rs1800796 IL6 polymorphism was postulated to affect the expression of interleukin 6 [11,17] which can initiate insulin resistance and eventually T2DM [3]. In addition, previous studies have produced related evidence [7,11,18]. Consistent with rs2794521 and rs1800796, the exchange of T to C allele at -1031 locus in the TNF gene was associated with T2DM [12][13]. Nevertheless, study outcomes regarding the relationship between selected genetic polymorphisms and T2DM risk are in disagreement. Additionally, to the authors' knowledge, evidence of the effect of a combination of these polymorphisms in enhancing T2DM is inconclusive. Therefore, this study aimed to investigate whether speci c polymorphisms, namely rs2794521, rs1800796 and rs1799964, increase the risk of elevated blood glucose among Thais, and whether the haplotypes of the studied polymorphisms relate to the condition.

Subjects
Thais who were aged between 35 and 66 and living in Sun Noen district, Nakhon Ratchasima province, Thailand, were randomly recruited into the study. All subjects had been previously screened for T2DM at health promoting hospitals. Pregnant and lactating women and people who had been medically diagnosed with chronic diseases or had any type of infection were omitted. After written informed consent had been obtained, anthropometric data (weight, height, body mass index, waist circumference, percentage of body fat, percentage of visceral fat, and muscle mass), biochemical data (fasting blood glucose, 2-hour blood glucose, HbA1c, total-cholesterol, triglyceride, HDL-C, LDL-C, C-reactive protein, interleukin 6, and tumor necrosis factor-aloha) and dietary data were determined and presented in a previous publication [6]. Leftover blood samples were obtained from this previous study, and used meet the aim of the current study. The current study procedure was approved by the Ethics Committee of the Faculty of Tropical Medicine, Mahidol University (TMEC 19-050). For the purposes of the current study, the samples from subjects with normal blood glucose concentrations (fasting blood glucose <100 mg/dl, 2-hour blood glucose <140 mg/dl, and HbA1c <5.7%) were allocated to the control group (n = 135 [45 males and 90 females]) and samples from subjects who had elevated blood glucose concentrations (fasting blood glucose ≥100 mg/dl, 2-hour blood glucose ≥140 mg/dl, and HbA1c ≥5.7%) were assigned to the elevated blood glucose (EBG) group (n = 161 subjects [55 males and 106 females]).

Genetic variations analysis
All leftover blood samples were stored at -80 °C in a laboratory of the department of Tropical Nutrition and Food Science, Mahidol University prior to analysis. Genomic DNA extraction was performed following the instructions of the FlexiGene DNA kit (Qiagen, Germany) using whole blood samples. The extracts were quanti ed and quali ed by a Nanodrop ND-1000 spectrophotometer. The selected genetic polymorphisms, which included -717 A/G (rs2794521) of the CRP gene, -572 C/G (rs1800796) of the IL6 gene and -1031 T/C (rs1799964) of the TNF gene, were analyzed using the LightCycler FastStart DNA Master HybProbe kit (Roche Diagnostics GmbH, Germany) based on the real-time polymerase chain reaction (q-PCR) method.

Statistical analysis
Statistical analysis was conducted using Statistical Package for the Social Sciences (SPSS), version 18. The distribution of data was examined using the Kolmogorov-Smirnov test. The Mann-Whitney U test was used to determine differences between the control group and EBG group, while the differences between genotypic groups were assessed using the Kruskal-Wallis test. The distribution of genetic variations was tested by the Hardy-Weinberg equilibrium equation. Odds ratio (ORs) with 95% con dence interval (CI) were computed through binary logistic regression, to investigate the association between genetic variations and risk of elevated blood glucose. A p-value less than 0.05 was considered statistically signi cant.

Results
Genetic polymorphisms including -717 A/G (rs2794521) of the CRP gene, -572 C/G (rs1800796) of the IL6 gene, and -1031 T/C (rs1799964) of the TNF gene were examined in 296 subjects. The frequencies of the selected single nucleotide polymorphisms are presented in Table 1. The studied genetic variations were compatible with Hardy-Weinberg equilibrium (p>0.05 for all). Most of the subjects (N = 296) had wild type genotypes, AA in CRP gene (n = 154, 52%), CC in IL6 gene (n = 163, 55%) and TT in TNF gene (n = 137, 46%). Genetic distribution patterns were similar in the control group and the EBG group. Table 2 compares the characteristics of subjects with different genotypic variants. Subjects with the heterozygous variant of the CRP gene (AG) had signi cantly higher BMI, WC and body fat than those with the wild type (AA) or homozygous variant (GG), p<0.025 for all. There was no signi cant difference in blood glucose indices or lipid pro les. Regarding rs1800796, a difference was only found in HbA1c. CC carriers had higher HbA1c levels compared to GG carriers, p<0.025. A noteworthy difference in a blood glucose index, FBG, was also present when analyzed within the TNF variant (rs1799964) groups. Subjects with the homozygous variant (CC) had lower concentrations of FBG compared to wild types (TT), p<0.025. Moreover, a signi cant difference was observed in the lipid index, triglyceride. Subjects with the heterozygous variant (TC) had greater triglyceride concentrations compared to wild types (TT), p<0.025. When investigating concentrations of in ammatory cytokines (CRP, IL6, and TNF-α), no notable differences were detected in subjects, p>0.05 for all. Dietary intake was similar among subjects with different genetic variants, p>0.05 for all.
The associations between the selected genetic variants and risk of elevated blood glucose among the subjects (N = 296) and within each sex (females: n = 196; males: n = 90) are shown in Table 3. Associations were determined in three genetic models including codominant, dominant, and recessive models. There was no signi cant association between rs2794521 or rs1799964 versus risk of elevated blood glucose. On the contrary, regarding rs1800796 in the IL6 gene, the codominant model (C/C vs. C/G vs. G/G) and the recessive model (C/C+C/G vs. G/G) revealed noteworthy associations in females. Those with a GG genotype were less likely to have an increased risk of elevated blood glucose [codominant model: OR 0.23 (0.06-0.88), p = 0.032; recessive model: OR 0.23 (0.06-0.87), p = 0.031] compared to those with other genotypes.
The relationship with risk of elevated blood glucose was also investigated within different haploid genotypes as displayed in Table 4. The most frequent pattern found in the study was ACT (37%), which was subsequently used as a reference haplotype to analyze the association. As a result, only the AGT pattern tended to have a protective effect on risk of elevated blood glucose [OR 0.49 (0.24-1.02), p = 0.058].

Discussion
Genetic variants of in ammation-related genes are believed to play roles in the development of elevated blood glucose, due to their ability to modulate concentrations of in ammatory cytokines [9,[11][12] which may consequently affect the glucose pathway and eventually result in T2DM [2][3]. In this study, we investigated whether polymorphisms rs2794521 of the CRP gene, rs1800796 of the IL6 gene and rs1799964 of the TNF gene are related to elevated blood glucose among the Thai population. Regarding the results, the distribution of each variant was consistent with previous studies [9,12,15,[19][20]. The wild type of each of the selected gene polymorphisms was most frequent, followed by the heterozygous variant, and nally the homozygous variant.
Subjects with different genotypes of the selected gene polymorphisms were likely to have comparable consumption patterns, as there was no difference in dietary intake. Nevertheless, there was remarkable dissimilarity in body fat and blood glucose indices. The body mass index, waist circumference and body fat percentage of subjects with an AG genotype of the rs2794521 CRP polymorphism were notably greater than the others. Similarly, Martínez et al. revealed a signi cant association between the increase in BMI and rs2794521 at the haplotype level (rs1130864T-rs1205G-rs2794521G-rs3093062G) [9]. Thus, this genetic variant possibly has a substantial in uence on obesity, which is an important risk factor for T2DM. However, in the present study, the association of rs2794521 with either CRP concentration or T2DM was vague. Likewise, Sheu et al. did not detect an association between an A to G transition at rs2794521 and either CRP or blood glucose concentrations in Taiwanese subjects [21]. The CRP concentration of subjects from India with different genotypes of rs2794521 also did not differ [15]. In contrast, a study conducted in Spain reported a notable relationship between rs2794521 and CRP concentration. Subjects with the major A allele had signi cantly higher CRP concentrations than those with the G allele [14]. In addition, there was a signi cant increase in T2DM prevalence reported in Polish subjects with the AG+GG genotype [22] and black American subjects had a higher tendency to develop the disease [10]. Accordingly, the controversial association may be a consequence of differences in ethnicity.
In relation to the rs1800796 IL6 polymorphism, HbA1c level of subjects with the homozygous variant (GG genotype) was signi cantly lower than in those with the wild type (CC genotype Comparably, a meta-analysis of 11,681 participants of case-control studies conducted from 9 Asian studies and one European study, noted that the G allele was associated with an increased risk of T2DM [18]. Consistent with the GG genotype of rs1800796, study subjects who carried the homozygous variant (CC genotype) of rs1799964 TNF polymorphism revealed a lower level of fasting blood glucose compared to TT carriers. Thus, people who carry the homozygous variants may be less likely to develop high blood glucose levels. Nevertheless, the relationship between rs1799964 and the studied in ammatory cytokines or T2DM was unclear in the present study. Likewise, Chan et al. determined the in uence of rs1799964 genetic variants among postmenopausal women [26]. They found that the variants were not signi cantly associated with risk of T2DM. On the contrary, in a study of north Indian subjects, the C allele at the -1031 locus of the TNF gene was associated with diabetic nephropathy, a complication of T2DM [12].
Likewise, the study of Xu et al. also revealed the effect of -1031C-863C-857C of the TNF gene in elevating the risk of T2DM among Chinese subjects [13]. In addition, we evaluated the effect of the selected polymorphisms at the haplotype level. Subjects with AGT haplotype, a combination of rs2794521, rs1800796, and rs1799964, tended to exhibit a reduced risk of T2DM, compared to the reference haplotype ACT.

Conclusion
In conclusion, the studied polymorphisms, rs2794521 and rs1799964, tended to be associated with obesity and blood glucose indicators, both of which are substantial risk factors for the development of T2DM. The G allele of rs1800796 IL6 polymorphism strongly promoted a relatively low risk for developing T2DM. Subjects with similar dietary intakes, who carried the GG genotype of rs1800796 or AGT haplotype, were less likely to have increased T2DM risk factors, and the association was clearly seen in the females. This study suggested that rs2794521, rs1799964 and rs1800796 are involved in T2DM development in the Thai population. Thus, these polymorphisms may be an effective prognostic gene marker for the disease. As the association among males and the difference between males and females cannot be evaluated in this study, a larger study should be conducted to verify the results.

Availability of data and material
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.

Funding
This study was supported by the Faculty of Tropical Medicine, Mahidol University and was partially supported for publication by the China Medical Board (CMB), Faculty of Public Health, Mahidol University, Bangkok, Thailand. CP, PP, KK designed the study. CP, PP, KK obtained the data. CP and KK analyzed and interpreted the data. PP, HH, KK provided advice. CP wrote the rst draft. All authors read and approved the nal manuscript.   Odds ratio (OR) and 95% confidential Interval (CI) calculated by binary logistic regression.

Tables
*OR adjusted by age and gender. **OR adjusted by age.
P<0.05 was considered statistically significant. Odds ratio (OR) and 95% confidential Interval (CI) calculated by binary logistic regression.
OR adjusted by age and gender.
P<0.05 was considered statistically significant.