Mitochondrial uncoupling protein 2 (UCP2) gene polymorphism − 866 G/A in the promoter region is associated with type 2 diabetes mellitus among Kashmiri population of Northern India

The study aimed to evaluate the association of UCP2 gene polymorphism − 866 G/A and its expression with diabetes predisposition in the North Indian population. The study involved 850 subjects, including 425 each T2DM and control subjects. The serum metabolic and clinical parameters were estimated using standard protocols. The PCR–RFLP based genotyping was performed to determine UCP2 gene polymorphism, while the expression was measured by real-time quantitative PCR. The genotypic and allelic frequencies showed a significant difference in cases compared to controls (p < 0.05). The diabetes patients had a 4.2-fold decrease in UCP2 gene expression. The expression was 29.8 and 8.4 fold lower in diabetes patients with homozygous (AA) and heterozygous (GA) mutation at − 866 locus of UCP2 nucleotide sequence, respectively. When categorized according to age and BMI, the T2DM subjects with age ≥ 50 and BMI ≥ 25 had a 5.53 and 8.2-fold decrease in UCP2 expression, respectively. The diabetes subjects with homozygous and heterozygous mutation demonstrated a pathological increase in serum metabolic and clinical parameters, which corroborated with UCP2 gene expression, indicating a strong association between the two. Intriguingly, we did not find any association between − 866 G/A polymorphism of UCP2 with serum insulin levels. Our investigation is the first among the studies conducted in Jammu and Kashmir to work on adipose tissue and UCP2 gene polymorphism. The association of − 866 G/A SNP of the UCP2 gene with its expression in diabetes patients appears to be an important genetic determinant in the progression of T2DM. Moreover, age ≥ 50 years and BMI ≥ 25 could be considered risk factors for developing T2DM in the studied population.


Introduction
Current global estimates evidenced over 537 million young adults are living with diabetes worldwide [1]. Diabetes mellitus is a metabolic disorder characterized by disruption of glucose homeostasis and is associated with hyperglycemia. It primarily occurs due to defects in insulin secretion and action. World Health Organization reported that 9% of the world population are diabetes, of which 90% cases have type-2 Diabetes Mellitus (T2DM) [2]. T2DM is a condition of impaired glucose regulation due to dysfunctional pancreatic β-cells and insulin resistance [3]. Studies on genetic linkage analysis and the T2DM association have shown a dynamic interplay between genetic and environmental components [5]. Genetic factors associated with impaired insulin secretion, insulin resistance, or environmental factors are classic components of T2DM development [4].
The patients suffering from T2DM have severe mitochondrial abnormalities in their β-cells [5]. Failure of β-cell function or their mitochondria is closely associated with T2DM pathogenesis [6] UCP2, a mitochondrial protein, is highly expressed in β-cells of patients with hyperglycemia or hyperlipidemia [7,8] and its genetic deficiency has been shown to improve β-cell function [9,10]. It plays a major role in guarding cells against oxidative stress, modulation of insulin secretion and metabolism of fatty acids. It thus may play an important role in linking obesity to T2DM [6]. UCP2 being an integral membrane protein that upon activation, leads to proton leak across the inner mitochondrial membrane and uncouples oxidative metabolism from ATP production [11]. Because of its proton leak activity, UCP2 is considered a negative regulator of glucose-stimulated insulin secretion [7]. Insulin secretion by β-cells is an ATP-dependent process, so impaired ATP production by increased UCP2 activity could be essential in regulating glucose-stimulated insulin secretion in these cells [6]. Thus, an increase in UCP2 activity either by increased glucose, lipids and/or ROS levels, results in increased stimulation of proton leak across the mitochondrial membrane, leading to decreased ATP production. This decreased ATP production impairs insulin secretion by β-cells, eventually leading to T2DM development.
Genome-wide scan has explored many SNPs as a highresolution marker for accelerating the pace of gene mapping related to diseases or traits. Regulatory polymorphisms at DNA level can potentially result in different gene expressions [12]. The − 866 G/A SNP is a common functional variant that exists in the promoter of the human UCP2 gene. Several studies have demonstrated A allele [13][14][15] and G allele [16,17] as important determinants of T2DM development, while others have reported that GG genotype carrying subjects had an increased risk of developing T2DM other than subjects with GA and AA genotype [17,18] though several others have reported no significant association of − 866 G/A SNP with T2DM in their respective populations [19][20][21]. An association of − 866 G/A SNP with obesity has been evident from previous research [22]. In the present study, we evaluated the association of the common − 866 G/A SNP in the promoter of the UCP2 gene with T2DM in the Kashmiri population. Besides the effect of UCP2 gene polymorphism on gender and age, we also analyzed the expression analysis of the UCP2 gene in diabetes patients based on genotype, age and BMI.

Material and methods
This case-control study was conducted in the Department of Biochemistry, Govt. Medical College (Research Center, University of Kashmir) India. The study participants involved 425 newly diagnosed T2DM subjects and 425 healthy volunteers/controls. The diagnosis T2DM subjects were done as per the ADA criteria 2010 [23]. The study was approved by the Institutional Ethics Committee, Government Medical College, Srinagar. All methods were performed following the guidelines and regulations of the Indian Council of Medical Research (ICMR), New Delhi. All participants signed the informed consent.

BMI calculation
Body mass index (BMI) was calculated as the ratio of body weight in kg and height in meter square (kg/m 2 ). Height (cm) was noted using a measuring tape to the nearest 0.1 cm. Weight (kg) was measured to the nearest 0.1 kg using a weighing machine simultaneously.
The classification of T2DM subjects according to BMI in our study was done as per WHO: less than 18.5 kg/ m 2 = underweight; 18.5-24.9 kg/m 2 = normal weight; 25.0-29.9 kg/m 2 = overweight; and 30.0 kg/m 2 or more = obesity.

Sample collection
The blood samples were collected from 425 T2DM subjects and 425 controls in EDTA and clot activator vials. The blood samples in EDTA vials were used for isolation of genomic DNA using Gen ELUTE Blood Genomic DNA Kit (Sigma-Aldrich, USA). From clot activator vials, serum was separated by centrifuging blood samples at 4000 rpm for 15 min. The separated serum was gently transferred to 1.5 ml microcentrifuge tubes and stored at − 20 °C to estimate biochemical parameters, including blood sugar fasting, Glycated hemoglobin (HbA1c), insulin, triglycerides, cholesterol, HDL and LDL levels.
Adipose tissues were collected from 65 T2DM patients and 67 control subjects who underwent elective abdominal surgeries (cholecystectomy, appendectomy, inguinal hernia, epigastric hernia, incisional hernia) in the department of Surgery Govt. Medical College and Associated Hospitals, Srinagar. The tissue samples were immediately put in RNA later and stored at − 80 °C for further experiments.

Determination of Serum biochemical parameters
Serum fasting glucose was estimated by glucose oxidase method and serum HbA1c levels were estimated by enzymatic method/kit method as per the manufacturer's protocol. Further, the serum triglyceride, cholesterol, HDL, LDL and insulin levels were estimated by enzymatic method/kit as per the manufacturer's protocol. The observed values were then compared with reference range values as mentioned [23][24][25].

PCR amplification and restriction fragment length polymorphism for determining − 866 G/A SNP
DNA containing the UCP2 gene polymorphic site at − 866 position was amplified by polymerase chain reaction using the following primers: Forward-5ʹ-CAC GCT GCT TCT GCC AGG AC-3ʹ; Reverse 5ʹ-AGG CGT CAG GAG ATG GAC CG-3ʹ. Briefly, a total of 20 µl reaction volumes were carried out using 1 µl of each forward and reverse primers, 2.5-3 µl of DNA template, 10 µl of the ready-to-use reaction mixture (3B BlackBio Biotech India Ltd) & 5-6 µl of Milli-Q water.
The thermal cycler conditions standardized for − 866 G/A SNP were: initial denaturation 94 °C for 10 min, followed by 35 cycles of 95 °C for 30 s, 58.7 °C for 30 s, and 72 °C for 1 min. The reactions were given a final 10 min extension at 72 °C. The PCR amplified products so obtained were resolved on 1% agarose gel and visualized under gel doc (Supplementary Figure S1). The PCR products so obtained were enzyme digested using Mlu1 restriction enzyme (Sigma, USA) for determining G/A SNP at − 866 position of the UCP2 gene. The RFLP products so obtained were resolved on 3% agarose gel and three types of band patterns were observed, the homozygous wild (GG) yielded a single band of 363 bp uncut fragment, and the heterozygous mutant (GA) yielded three fragments of 363, 295 and 68 bp and the homozygous mutant (AA) yielded two fragments of 295 and 68 bp respectively (Fig. 1A).

Real-time quantitative PCR
The total RNA was extracted from adipose tissues (200-300 mg) of 67 controls and 65 T2DM subjects using Trizol (Invitrogen, Canada), as per the manufacturer's protocol. The quantity and purification of RNA was measured on a nanodrop spectrophotometer (Thermo Fisher, IL, USA) [26]. An equal quantity of RNA (1 μg) from all the subjects was used for cDNA synthesis using cDNA synthesis kit (Fermentas, USA). The primer sequences used for the gene expression were: GAPDH forward:

Statistical analysis
Statistical analysis was conducted using SPSS software 17.0 version (SPSS, Chicago, IL, USA) and GraphPad InStat 3.0 (USA). The means ± SD between cases and control groups, were tested using student's t-test. The allelic and genotypic frequencies between cases and controls were tested using the Chi-square test. The associations between SNPs and T2DM risks were assessed using odds ratio (OR) with 95% confidence interval (95% CI) and the p-value < 0.05 was considered to be significant.

Evaluation of biochemical parameters and their categorization based on UCP2 genotypes
We categorized the serum metabolic parameters of the study population as per the UCP2 genotypes for − 866 G/A polymorphism (Table 1). It was noted that all the T2DM subjects with GA and AA genotypes had higher values of biochemical and clinical parameters compared to subjects with GG genotypes (p < 0.05). However, no relative change in HDL levels was noted.

Association of − 866 G/A SNP of UCP2 gene with T2DM
The association of − 866 G/A polymorphism of the UCP2 gene with T2DM in the North Indian population was evaluated. The genotypic frequency for − 866 G/A SNP in the said population is shown in Table 2. The frequency of genotype GG vs GA, AA, and GA + AA was statistically significant in T2DM cases as compared to controls. The frequency of variant allele "A" was also found to be statistically significant among cases compared to control subjects (OR 1.56; 95% CI 1.25-2.05; p < 0.005).

Measurement of UCP2 gene expression
We measured the expression of the UCP2 gene in adipose tissue of type 2 diabetes patients and control subjects. We observed a 4.2-fold decrease in the expression of the UCP2 gene compared to control subjects (Fig. 1B). We further categorized the expression levels of the UCP2 gene in cases as per genotype (GG, GA and AA). Interestingly, we noted that expression of UCP2 was 29.8 fold lower in patients with AA genotype, whereas it was 8.4 and 1.8 fold lower in patients with GA and GG genotypes respectively (Fig. 1C). Intriguingly, the levels of the biochemical and clinical parameters were corroborated with the expression pattern of the UCP2 gene in diabetes patients (Table 2; Fig. 1C).

UCP2 gene expression in T2DM subjects followed age and BMI pattern
The expression of UCP2 in diabetes subjects as per age and BMI is shown (Fig. 1D). The diabetes subjects with BMI ≥ 25 had a remarkable decrease in UCP2 gene expression (8.2 fold), while it was 2.76 fold lower in diabetes subjects with BMI < 25 compared to controls (Fig. 1D). Interestingly, we noted that patients ≥ 50 years of age had comparatively lower expression of the UCP2 gene as compared to patients with age < 50. We found that expression of UCP2 was 5.53 fold lower in diabetes subjects of age ≥ 50 years, while it was 3.47 fold lower in subjects < 50 years of age (Fig. 1E). As per our results, we also noted that control subjects with age ≥ 50 years also had a 1.2-fold decrease in the expression of the UCP2 gene (Fig. 1E).

Expression profiling of TCF7L2, ABCC8, GCGR and CAPN10 genes
We have also assessed the gene expression of four known genes (TCF7L2, ABCC8, GCGR and CAPN10) involved in obesity and diabetes. Based on the qPCR gene expression analysis in the adipose tissue of control and diabetic patients, we observed no significant association of these genes between case-control groups and among different genotype groups (Supplementary Figure S2).

Serum insulin levels in T2DM subjects were independent of − 866 G/A SNP
The serum insulin levels of diabetes and control subjects were measured. We observed a 2.6-fold increase in insulin levels in diabetes patients ( Fig. 2A) compared to controls. The diabetes subjects had increased serum insulin levels compared to controls (14.62 ± 4.29 and 39.18 ± 4.89; p < 0.001), respectively. While the UCP2 gene expression was inverse i.e., 4.2 fold lower in diabetes patients (Fig. 1B), indicating that the two parameters are inversely related. We sought to categorize the serum Insulin levels of diabetes patients and controls based on the UCP2 genotype for − 866 G/A SNP. It was observed that all the three genotypes had approximately similar increase in serum insulin levels, indicating that changes in insulin levels were independent of − 866 G/A polymorphism of the UCP2 gene ( Fig. 2B; p > 0.05), which indicated that − 866 G/A polymorphism of UCP2 gene is not associated with serum insulin levels.

Discussion
Mitochondrial uncoupling proteins (UCP) are members of the larger family of mitochondrial anion carrier proteins (MACP). The UCP2 gene is associated with BMI (Quantitative Trait Locus 4) and Hyperinsulinism (due to UCP2 protein deficiency). Figure 3 provides the structure and functional association of UCP2 protein in regulating various metabolic biological pathways. Several genetic studies have demonstrated a strong relationship between the UCP2 locus and susceptibility to type 2 diabetes or obesity, with specific attention being paid to − 866 G/A polymorphism in the promoter region [27]. Our findings strengthened the earlier reports that sequence variation in UCP2 might contribute to the development and pathogenesis of T2DM.
Our results demonstrated a significant association of − 866 G/A SNP with T2DM in the Kashmiri population. Fig. 2 A The insulin levels of controls and T2DM cases, estimated from serum B Insulin levels of T2DM cases and controls based on UCP2 genotype for − 866 G/A SNP. The data are shown as mean ± SD, analyzed by independent t-test. *indicate p < 0.05 as compared to control The frequency of variant allele "A" was found to be significantly higher when compared to controls (OR 1.56; 95% CI 1.25-2.05; p < 0.0001). The AA genotype showed a statistical increase of 3.45-fold risk for the development of diabetes when compared with controls (OR 3.45; 95% CI 1.58-7.6; p = 0.006). The AA genotype and A allele of the − 866 G/A polymorphism have been associated with an increased risk for diabetes in various populations. For instance, the AA genotype of UCP2-866 had higher fasting plasma glucose levels and HbA1c and lower plasma insulin levels compared to the GG and GA genotypes among T2D patients of the Chinese population [28]. A study found lower insulin secretion in response to glucose stimulation in UCP2 − 866 A/A homozygotes than compared with − 866 G/G and − 866 G/A carriers [29]. Several other studies have also reported that the subjects with GA genotype had an increased risk of T2DM than subjects with GG and AA genotypes. Interestingly results from another study depicted that the UCP2 − 866 G/A polymorphism affects the beta-cell function and enforces a risk for the development of T2D and obesity. At the same time, allele G provides relative protection for T2D [30]. A study from the Northern state of India (i.e., Punjab) found that a particular ethnic group has significantly higher chances of developing T2D based on the physical and biochemical parameters. The genetic profiling of the same population had the polymorphisms such as UCP2 − 866 G/A and SIRT1 − 1400 T/C, which directly affect the development of diabetes [31].
The UCP2 gene expression and translation are based on a few factors such as its half-life, transcriptional control and location. The UCP2 is an unstable protein with a half-life close to 30-60 min [32][33][34]. For its physiological functions, the transcription to translation equilibrium should be in place and regulated at several levels. The transcriptional control of UCP2 gene expression determines its mRNA levels and corresponding protein levels in several tissues of its importance, such as lung, kidney, pancreas and adipose tissues [35]. The present study is the first among the studies conducted in Jammu and Kashmir to work on adipose tissue and UCP2 gene polymorphism and expression. We sought to measure the expression of the UCP2 gene in adipose tissue of diabetes subjects. We found a 4.2-fold decrease in the expression of the UCP2 gene in diabetes patients compared to control subjects (p < 0.01). These results agree with previous studies on the Asian Indian populations that noted a significant decrease in the UCP2 gene expression among T2DM subjects. A previous study from the Indian population showed that decreased UCP2 gene expression had been associated with obesity parameters and lipid metabolism [36]. Furthermore, our results showed similar findings to another study where the UCP2 expression was significantly lower in T2D patients. However, they further noted that UCP2 expression was lower in first-degree relatives of these patients [37].
T2DM is a disease closely associated with aging, and its prevalence increases with age. The maximum prevalence of diabetes patients in India is above 50 years of age [38,39], which was true for our population also, as Fig. 3 A Network showing the functional association of UCP2 using STRING v11.0 (https:// string-db. org/). The proximity of the proteins via threads predicts the association with the UCP2 protein and their role in different biological processes. B Structural model of UCP2 protein (https:// www. unipr ot. org/). The UCP2 upon activation, leads to proton leak across the inner mitochondrial membrane and uncouples oxidative metabolism from ATP production we noted a significant association of − 866 G/A SNP with age ≥ 50 years (OR 2.28; 95% CI 1.55-3.55; p = 0.002). We found that the "AA" genotype was associated with 3.43 fold risk in the development of diabetes in T2DM subjects with age ≥ 50 years (OR 3.43; 95% CI 1.27-9.27; p = 0.0016). We sought to measure the expression of UCP2 gene expression as per age above and below 50 years; we noted that UCP2 was 5.53 and 3.47 fold lower in subjects of ≥ 50 and < 50 age when compared to controls of corresponding age (p < 0.05) respectively indicating a bias in the expression of UCP2 towards age. Interestingly, we also found that control subjects of age ≥ 50 years had 1.2 fold decrease in expression of the UCP2 gene. This data confirms a biased expression of the UCP2 gene towards age, indicating that aging somehow down-regulates the expression of the UCP2 gene and hence leads to the development of T2DM in our studied population.
Our results showed a significant association of − 866 G/A SNP with BMI ≥ 25 suggesting a link between increased BMI and diabetes development in the current population. The "AA" genotype was associated with a fivefold risk in the development of T2DM in patients with BMI ≥ 25 (OR 5.54; 95% CI 2.38-10.36; p < 0.001). Further, the results revealed that the UCP2 gene expression in T2DM subjects with BMI ≥ 25 was 8.2 fold lower. In contrast, it was 2.76 fold lower in T2DM subjects with BMI < 25 compared to controls (p < 0.05), demonstrating that increased BMI levels influence the expression of UCP2 gene.
An association of − 866 G/A SNP with obesity and increased BMI has already been established [22,31,40,41] and a decrease in UCP2 expression in adipose tissue has been reported in other studies discussed above [36,37,39,42]. Based on UCP2 gene expression data, it could be predicted that increased BMI affects the UCP2 gene expression so that it could be considered a marker for the risk of T2DM in our population.
We specifically demonstrated that UCP2 in adipose tissue plays an important role in controlling diabetes by regulating levels of most of the biochemical and clinical parameters. On categorizing the biochemical parameters and clinical characteristics of the study population as per the UCP2 genotype for − 866 G/A polymorphism. It was noted that all the parameters, except for HDL (p > 0.05) had significantly increased serum levels in diabetes patients compared to controls (p < 0.001). Intriguingly, the levels were considerably higher in subjects with GA and AA genotypes, indicating that AA mutation at − 866 loci is strongly associated with the pathophysiology of T2DM (p < 0.05). However, insulin didn't follow a similar trend; the diabetes subjects with GA and AA genotypes had similar insulin levels, depicting that serum insulin levels are independent of genotypic change.
Suggesting a strong association of AA alleles with age, BMI and altered biochemical and clinical parameters, we sought to categorize the UCP2 gene expression in diabetes subjects as per GG, GA and AA genotypes. Interestingly, we observed 29.8 fold lower UCP2 expression in patients with AA genotype; in contrast, it was 8.4 and 1.8 fold lower in patients with GA and GG genotypes, respectively, thereby indicating that subjects with AA allele were strongly predisposed towards diabetes development.

Conclusion
The polymorphism at − 886 G/A SNP and decreased expression of the UCP2 gene in diabetes subjects act synergistically towards the development of diabetes mellitus in the current population, particularly in subjects with age ≥ 50 and BMI ≥ 25. It could also be concluded from the present study that reduced UCP2 gene expression in diabetes patients confirms its role as a candidate gene in T2DM regulation in the North Indian population. Interestingly, targeting the UCP2 gene expression and/or regulating the genotypic development at the beginning of its positive involvement in T2DM may provide an excellent strategy to circumvent the complications associated with T2DM, obesity and aging.