Liao et al. BMC Genetics 2014, 15:113 http://www.biomedcentral.com/1471-2156/15/113 RESEARCH ARTICLE Open Access Identified single-nucleotide polymorphisms and haplotypes at 16q22.1 increase diabetic nephropathy risk in Han Chinese population Li-Na Liao1†, Ching-Chu Chen2,3†, Fang-Yang Wu1, Cheng-Chieh Lin4,5, Jen-Hao Hsiao6, Chwen-Tzuei Chang2, Sharon LR Kardia7, Tsai-Chung Li8,9* and Fuu-Jen Tsai3,10,11* Abstract Background: Diabetic nephropathy (DN) has become one of the most common causes of end-stage renal disease (ESRD) in many countries, such as 44.5% in Taiwan. Previous studies have shown that there is a genetic component to ESRD. Studies attempting to determine which genetic variants are related to DN in Han Chinese are limited. A case–control study was conducted to identify DN susceptibility variants in Han Chinese patients with type 2 diabetes. Results: We included 574 unrelated type 2 diabetes patients (217 DN cases and 357 controls), who were genotyped using Illumina HumanHap550-Duo BeadChip. In single-SNP association tests, the SNPs rs11647932, rs11645214, and rs6499323 located at 16q22.1 under the additive-effect disease model were significantly associated with an approximately 2-fold increased risk of DN. In haplotype association tests, identified haplotypes located in the chromosome 16q22.1 region (containing ST3GAL2, COG4, SF3B3,andIL34 genes) raised DN risk. The strongest association was found with haplotype rs2288491-rs4985534-rs11645214 (C-C-G) (adjusted odds ratio [AOR] 1.93, 95% confidence interval [CI] 1.83-2.03, p = 6.25 × 10−7), followed by haplotype rs8052125-rs2288491-rs4985534-rs11645214 (G-C-C-G) (AOR 1.92, 95% CI 1.82-2.02, p = 6.56 × 10−7), and haplotype rs2303792-rs8052125-rs2288491-rs4985534-rs11645214 (A-G-C-C-G) (AOR 1.91, 95% CI 1.81-2.01, p = 1.15 × 10−6). Conclusions: Our results demonstrate that the novel SNPs and haplotypes located at the 16q22.1 region may involve in the biological pathways of DN in Han Chinese patients with type 2 diabetes. This study can provide new insights into the etiology of DN. Keywords: Diabetic nephropathy, Single-nucleotide polymorphism, Haplotype, Han Chinese Background genetic susceptibility plays an important role in the devel- Diabetic nephropathy (DN) has become one of the most opment and progression of DN [5,6]. Epidemiologic stud- common causes of end-stage renal disease (ESRD) in many ies have shown that 35% of patients with diabetes develop countries. In Taiwan, diabetes accounted for 44.5% of all nephropathy, irrespective of glycemic control [7,8]. new cases of ESRD [1], and more than 99% diabetes pa- Mooyaart et al. performed a meta-analysis to evaluate the tients were with type 2 diabetes [2]. Diabetic ESRD patients pooled effect of each genetic variant reproducibly associated had worse survival than non-diabetic ESRD patients [3]. with DN [9]. They reported that 21 of 34 replicated genetic It has been reported that genetic predisposition is one of variants remained significantly associated with DN. These 34 the main risk factors for the development of DN [4]. Nu- variants were in or near the following genes: ACE, ELMO1, merous familial aggregation studies have suggested that PPARG, etc. Recently, several genome-wide association stud- ies (GWASs) have attempted to detect genetic variants asso- * Correspondence: [email protected]; [email protected] †Equal contributors ciated with the risk of DN or diabetic ESRD in those of 8Graduate Institute of Biostatistics, College of Management, China Medical Japanese [10], Pima Indian [11], and African American [12] University, No. 91 Hsueh-Shih Road, Taichung 40402, Taiwan with type 2 diabetes, as well as European ancestry [13,14] 3School of Chinese Medicine, College of Chinese Medicine, China Medical University, Taichung, Taiwan with type 1 diabetes. Full list of author information is available at the end of the article © 2014 Liao et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Liao et al. BMC Genetics 2014, 15:113 Page 2 of 11 http://www.biomedcentral.com/1471-2156/15/113 Studies attempting to determine which genetic vari- using above two equations were similar. A spot urine dip- ants are related to DN in Han Chinese patients with type stick test was used for detecting proteinuria. Patients with a 2 diabetes are limited. In the present genetic association positive dipstick test (>1+) were classified as with protein- study, a case–control study was carried out to identify uria [18]. The sociodemographic and lifestyle characteristics DN susceptibility variants in Han Chinese patients with and the self-reported health status for each subject were re- type 2 diabetes, which can provide new insights into the corded using self-administered questionnaires. etiology of DN. Genotyping and quality control Methods Genomic DNA from peripheral blood was prepared using Study subjects the Puregene DNA isolation kit (Gentra Systems, Individuals with type 2 diabetes and aged over 20 years Minneapolis, MN, USA). Our samples were genotyped were recruited using the American Diabetes Association using Illumina HumanHap550-Duo BeadChip, which (ICD-9-CM, Diagnosis code 250) criteria for diagnosis of was performed by deCODE Genetics (Reykjavík, type 2 diabetes. Individuals with type 1 diabetes, gesta- Iceland). Genotypes were called using the standard pro- tional diabetes, and maturity-onset diabetes of the young cedureinBeadStudio(Illumina,Inc.,SanDiego,CA, were excluded. The dataset used is part of the whole data- USA), with the default parameters recommended by the set for the published paper titled “A Genome-wide Associ- platform manufacturer. The genotyping quality control ation Study Identifies Susceptibility Variants for Type 2 procedures used to identify and remove poor-quality Diabetes in Han Chinese” [15]. A total of 995 type 2 dia- data were described previously [15]. Individual SNPs betes subjects recruited from China Medical University were excluded if they had a total call rate <95% across Hospital were included in the current study. All patients all individuals, a minor allele frequency <5% and a total with type 2 diabetes were of Han Chinese origin, including call rate <99%, or had significant deviation from Hardy- − Minnan, Hakka, and Mainland Chinese. Individuals with Weinberg disequilibrium (p-value <10 7) in these subjects. significant aboriginal ancestry were excluded by using self- Further details of genotyping quality control procedures administered questionnaires with six items regarding the are available in the published study [15]. After SNPs qual- ancestral origin of their parents and grandparents. ity control, 429,018 SNPs were used and their total call Chronic kidney disease (CKD) was determined by esti- rate was 99.9%. We also examined population stratifica- mated glomerular filtration rate (eGFR) and urine protein. tion by using multidimensional scaling (MDS) analysis Diabetic patients with eGFR <60 mL/min/1.73 m2 or with as implemented in PLINK. The results of MDS analysis proteinuria determined by a spot urine dipstick of >1+ showed that there was no evidence for population were defined as DN cases. A total of 217 DN cases were stratification. eligible for the study. To increase comparability between DN cases and controls, 357 controls were randomly Statistical analysis selected based on frequency-matching of their age and Demographic and clinical characteristics of study subjects durations of diabetes. This study was approved by the were examined, including sex, age, diabetes-related variables, Human Research Committee of China Medical University behaviors, biochemical variables, and history of diseases. Hospital. All patients signed informed consent forms. Continuous variables are reported as mean ± standard devi- ation (SD), and categorical variables are reported as number Measurements and percentage. Two-sample t tests and Chi-square tests Blood samples were collected in the morning after a 12-h were used for the bivariate analyses. Because the distribution overnight fast and were sent for analysis within 4-h of col- of the triglycerides was skewed, the data were normalized lection. Spot morning urine samples were collected. Triglyc- using a natural log-transformation, and the geometric erides, total cholesterol, high-density lipoprotein cholesterol mean ± SD was calculated. To identify the DN susceptibil- (HDL-C), low-density lipoprotein cholesterol (LDL-C), cre- ity variants, single-SNP association tests using Cochran- atinine, blood urea nitrogen (BUN), and uric acid were Armitage trend test were performed. Then, multiple measured by a biochemical autoanalyser (Beckman logistic regression analysis using an additive-effect disease Coulter, Synchron LX20, Fullerton, CA, USA). Renal model (an ordinal genotype model: 0, 1 and 2 of a minor function was evaluated by eGFR, which was estimated by allele) was performed for each SNP with adjustment of the using the Modification of Diet in Renal Disease Study subject's sex, age, BMI, and durations of diabetes. The equation for Taiwanese: eGFR (ml/min/1.73 m2)=175× Bonferroni correction was used to