Genetic and Epigenetic Risk of T2D in the UAE

The general aim of my research will be to investigate the contribution of environmental, genetic, and epigenetic factors to the increased risk of T2D in the Emirati population compared to Europeans. My specific aims will include:

1. Establish a cohort of Emirati T2D cases and controls to enable investigation of genetic and environmental risk factors for T2D in the UAE.

2. Use the data collected to examine the contribution of adiposity, physical inactivity, and dietary habits to the increased risk of T2D amongst the Emiratis compared to the Europeans living in the UK.

3. Investigate whether known and novel genetic variants account for the high risk of T2D in Emiratis.

4. Investigate DNA methylation profiles of subjects with T2D and non-diabetic controls in Emiratis and Europeans, and identify the differences in methylation profiles between the two populations.

5. Identify candidate methylation regions that are associated with, and could account for the higher risk of T2D amongst Emiratis.

1. ESTABLISH A COHORT OF EMIRATI T2D CASES AND CONTROLS TO ENABLE INVESTIGATION OF GENETIC AND ENVIRONMENTAL FACTORS FOR T2D IN THE UAE.

STUDY DESIGN: Case-control study SUBJECTS: 600 Emirati adults with T2D, and 600 age-sex matched control free from T2D, recruited from Imperial College London Diabetes Centre (ICLDC) based in Abu Dhabi, UAE. Recruitment started in April 2016. Participants include men and women ≥18 years old. They are provided with a complete information letter about the study, and are asked to sign a written consent to be included in the study.

METHODS: Participants will undergo health examination, which will be conducted by nurses and interviewers at the ICLDC. Examinations will include a brief medical history, anthropometric measures including weight, height, waist and hip circumferences, and blood pressure. Blood samples will be collected after an 8-hour fast for biochemical analysis, which includes measuring glucose, insulin, HDL and total cholesterol, and triglycerides levels. Diabetes will be defined using 1999 WHO criteria. Obesity will be assessed by BMI, where subjects will be classified as obese if BMI ≥30 kg/m2 and overweight if 25 kg/m2 ≤ BMI < 30 kg/m2. Additionally, participants will be asked to fill a physical activity questionnaire to assess their activity at work, transport, and recreational times during a typical week. A subset of 200 participants, 100 free from T2D and 100 with T2D, will be asked to wear Actigraph activity monitors that will measure their physical activity levels for 7 days. These participants will also be asked to fill out a food frequency questionnaire to assess their dietary habits.

ANALYSIS: Data will be examined for completeness, consistency, and response rates and for evidence of responder bias. I will carry out descriptive analyses to define the age and sex-specific prevalence of conventional risk factors for T2D amongst Emiratis, and their relationship with T2D in this population.

JUSTIFICATION: This project was planned to start in March 2016; however, due to lack of funding, and delays in college enrolment, I was not able to continue with my analyses. During April 2016 - Feb 2017, I have completed recruiting 1000 Emirati subjects; 500 with T2D and 500 free from diabetes. I will now continue recruiting 200 more subjects for physical activity monitoring and dietary habits data collection.

2. EXAMINE THE CONTRIBUTION OF ENVIRONMENTAL RISK FACTORS TO THE INCREASED RISK OF T2D IN THE UAE.

STUDY DESIGN: Case-control study SUBJECTS: The cohort of Emirati T2D cases and controls collected (specific aim 1), as well as ~5,000 Europeans.

ANALYSIS: Assess differences in BMI, waist and hip circumference, physical activity levels, and dietary intake among Emiratis and Europeans who are free from diabetes. This will allow me to identify whether differences in the prevalence rates of T2D in Emirati and European subjects reported in the literature is paralleled with different rates of environmental risk factors. Furthermore, the strength of the association of environmental risk factors with T2D will be assessed using logistic regression. Odds ratios (OR, Confidence Interval (CI) 95%) of age, BMI, waist-to-hip ratio (WHR), physical inactivity, and dietary intake will be calculated amongst Emiratis and compared to Europeans. This will assess whether environmental risk factors can explain the difference in T2D magnitude of risk between the two populations.

3. CARRY OUT GENOME-WIDE ASSOCIATION ANALYSIS TO IDENTIFY NOVEL AND KNOWN GENETIC LOCI INFLUENCING, AND CONTRIBUTING TO INCREASED RISK OF T2D IN THE UAE.

STUDY DESIGN: Case-control study. SUBJECTS: The cohort of Emirati T2D cases and controls collected (specific aim 1), as well as ~5,000 Europeans.

METHODS: Whole blood samples will be collected from participants after an 8-hour fast for DNA extraction and genotyping. DNA will be extracted using PureLinkTM Genomic DNA extraction kit (Invitrogen, Life Technologies), following manufacturer's instructions. Genotyping will be carried out on the Illumina Infinium II SNP genotyping array, which scales genotyping of up to 500,000 SNP loci. Amplified and fragmented DNA samples will be applied to BeadChips, where they anneal to locus-specific 50-mers (covalently linked to one of over 500,000 bead types). One bead type corresponds to each allele per SNP locus. Allelic specificity is conferred by enzymatic base extension, and subsequently identified by measuring beads' fluorescence intensities.

ANALYSIS: Identify variants with higher allele frequencies among Emiratis compared to Europeans. Allele frequencies will be compared between Emirati subjects with T2D and healthy controls to identify known and novel genetic loci that are associated with T2D in the UAE. The strength of association of genetic variants with T2D will be analysed with logistic regression under an additive genetic model to estimate effect sizes (ORs, CI 95%). Effect sizes will be compared to data from Europeans to assess whether identified genetic variants can explain part of the increased risk of T2D among Emiratis.

In addition, I will test SNPs for association with fasting glucose levels, fasting insulin levels, triglycerides levels, insulin sensitivity or β-cell function, under an additive genetic model, and effect sizes will be compared to data from Europeans. Furthermore, the combined effect of genetic loci influencing similar T2D- related phenotypic traits will be tested across the two populations. This will be through evaluating the effect of carrying increasing number of risk alleles on a T2D intermediate phenotype in Emiratis compared to Europeans.

JUSTIFICATION: Due to recruitment and funding constraints, GWA analyses will be carried out in only 600 T2D cases and 600 controls free from diabetes. With this sample size, I will be able to identify Emirati-specific genetic risk variants that have 10% minor allele frequency and an effect size of 1.3, with 80% power. This fulfils our aim of identifying Emirati-specific variants with possible clinical relevance. Table 1 summarizes the sample size required for a range of minor allele frequencies and odds ratios in a case-control study design.

4. INVESTIGATE DIFFERENCES IN DNA METHYLATION PROFILES BETWEEN EMIRATIS AND EUROPEANS, AND IDENTIFY CANDIDATE METHYLATION REGIONS THAT CAN ACCOUNT FOR THE HIGHER RISK OF T2D AMONG EMIRATIS AND UNDERLIE DISEASE PATHOGENESIS.

STUDY DESIGN: Case-control study. SUBJECTS: A subset of the cohort collected (specific aim 1), including 100 Emirati T2D cases, 100 insulin resistant controls, and 100 insulin sensitive controls. In addition to 100 European adults with T2D, 100 insulin resistant controls, and 100 insulin sensitive controls. All subjects included will be age and sex matched.

METHODS: Whole blood samples will be collected from participants. DNA will be extracted from isolated blood leukocytes for DNA methylation assays. Genomic DNA will be treated by bisulphite conversion, and bisulphite converted DNA will be analysed using Illumina Infinium Human Methylation 450K BeadChip Array technology. The BeadChip interrogates 485,577 cytosine positions in the human genome; the majority are CpG (cytosine-guanine nucleotide pairs) dinucleotides, with sites located within known gene regions, including promoter, gene body, intergenic, and untranslated regions.

ANALYSIS: For a given CpG site, methylation status is defined by a beta value, which is calculated as the intensity of methylated channel divided by total intensity. Methylation for genomic regions will be calculated as the mean beta for all probes located within the region. Identification of differentially methylated regions (DMRs) will be performed using the probe lasso method implemented in the R package ChAMP. DNA methylation profiles of Emiratis and Europeans free from diabetes will be compared to identify differences in methylation profiles across populations. DNA methylation profiles in Emiratis and Europeans with T2D and/or insulin resistance will be compared to that of insulin sensitive controls to identify DMRs associated with T2D and/or insulin resistance in each population. The strength of association of DMRs with T2D and/or insulin resistance will be analysed with logistic regression to estimate effect sizes (ORs, CI 95%). Effect sizes will be compared across the two populations to assess whether identified DMRs can explain part of the increased risk of T2D among Emiratis.