Genome-wide association study identifies new type 2 diabetes risk loci in Jordan subpopulations  Jin Li, Rana Dajani, Zhi Wei, Yousef Khader, Michael March, Nancy Hakooz, Raja Fatahallah, Mohammad El-Khateeb, Ala Arafat, Tarek Salah, Abdel Rahman Dajani, Zaid Al-Abbadi, Mohamed Abdul Qader, Abdel Haleem Shiyab, Anwar Bateiha, Kamel Ajlouni, Hakon Hakonarson RESULTS (cont’d) INTRODUCTION Diabetes is among the most common chronic diseases globally and the prevalence is increasing. In Jordan, Type 2 diabetes (T2D) affects 16% of the adults and another 23.8% with pre-diabetes. T2D accounts for 90% of the diabetes cases in adults, causing a significant public health burden. Previous genome-wide association studies (GWAS) have identified over 90 susceptibility loci, but none of these studies have specifically investigated the Jodan populations. Association statistics of top SNPs associated with T2D in Jordan subpopulations SNP Chr Pos (hg19) Closest gene A1/A2 MAF cases/ controls OR Chechen (95% CI) P Chechen OR Circassian (95% CI) P Circassia n Pmeta rs6134031 20 10752610 JAG1 T/C 0.51/0.23 9.84 (3.334, 29.02) 3.45E-05 9.48 (3.091, 29.07) 8.36E-05 1.12E-08 rs4758690 12 122610909 MLXIP G/A 0.60/0.38 3.89 (1.78, 8.47) 6.36E-04 2.41 (1.19, 4.91) 0.0152 4.20E-05 METHODS Study subjects recruitment and genotyping The study has been approved by the institutional review board committee at the National Center for Diabetes, Endocrinology and Genetics of Jordan. A total of 144 subjects including cases and controls from the Chechen population in Jordan and 140 subjects from the Circassian population in Jordan were recruited, with signed consent of agreeing to participate. All samples were recruited and genomic DNA was genotyped at Center for Applied Genomics (CAG) at the Children’s Hospital of Philadelphia (CHOP). Illumina Infinium II OMNI-Express BeadChip were used for genotyping. Phenotype confirmation Cases with diabetes mellitus were defined according to the following criteria: Diagnosis of diabetes was known to the patient or, if fasting serum glucose is equal to or greater than 7 mmol/l (126 mg/dl) based on the ADA definition. Impaired fasting glucose was defined as a fasting serum glucose level of between 6.1 mmol/L (100 mg/dl) and 7 mmol/l. HbA1c was used to evaluate glycemic control. Any previously diagnosed diabetic patient with HbA1c > 7% was classified as having ‘unsatisfactory’ glycemic control. Quality control A survey was completed by each participant and pedigree information was collected. Any Chechen subject whose parents, grandparents or great grandparents from maternal or paternal side are of non-Chechen heritage was excluded from the study. Similar exclusion criteria were applied to the Circassian population. We also remove samples with a SNP call rate < 98%, heterozygosity outliers , duplicate samples, population outliers; remove SNPs with a call rate <95%, MAF < 1% and markers that deviated from Hardy-Weinberg-Equilibrium (HWE test P-value < 10-5) in the controls. Association analysis logistic regression was performed in the two subpopulations separately, including the first three principal components as covariates, followed by meta-analysis. Analysis of Methylation Data Methylation status was ascertained for a subset of subjects recruited by the Center for Applied Genomics through the use of the Infinium HumanMethylation450 BeadChip Kit. Methylation data in gene MLXIP were analyzed as the response variable in a linear regression, with genotype at rs4758690 as the predictor variable. Sex, age, and 10 genotype-derived principle components were included as covariates. SNP rs4758690 correlates with methylation status in gene MLXIP The regional association plot at locus 12q24.31 P=2.97x10-5 SNP rs4758690 genotype associates with gene MLXIP expression level in tissue transverse colon SNP rs4758690 genotype associates with gene MLXIP expression level in tissue small intestine RESULTS Number of T2D-cases vs controls and their ethnic distribution following QC Ethnicity Cases  Controls Total Chechen N 34 109 143   ♂ 18 (53%) 65 (60%) 83 (58%)  Circassian 33 105 138 20 (61%) 58 (55%) 78 (57%) P=4.2x10-7 P=1.1x10-14 Principal component analysis of subjects in each cohort MLXIP expression in different tissue types Manhattan plot of the association statistics for T2D SUMMARY 1. We identified a novel T2D locus at chr20p12.2 at genome-wide significance in meta-analysis of GWAS conducted in Chechen and Circassian subpopulations of Jordan; It is located at an intergenic region, near gene JAG1; 2. Another locus at chr12q24.31 (rs4758690, P= 4.20e-05) was associated with T2D at suggestive significance level; 3. The locus has a significant eQTL for gene MLXIP expression in transverse colon and in small intestine-terminal Ileum; 4. This SNP is also significantly associated with methylation level in MLXIP. 5. MLXIP shows highest expression in colon and functions in transcriptional regulation of cellular glucose response. 6. Taken together, in the first GWAS of T2D cases constituting Jordan subpopulations, we identified novel susceptibility loci which may be unique risk factors in populations of the middle-east.