Identification of novel high-impact recessively inherited type 2 diabetes risk variants in the Greenlandic population

Niels Grarup,Oluf Pedersen,Allan Linneberg,Torben Hansen,Bert B Boyer,Peter Bjerregaard,Howard W Wiener,Ida Moltke,Emil Jørsboe,Anders Albrechtsen,Mette K Andersen,Christina V L Larsen,Scarlett E Hopkins,Marit E Jørgensen,Inger K Dahl-Petersen,Hemant K Tiwari

doi:10.1007/s00125-018-4659-2

Abstract

Aims/hypothesisIn a recent study using a standard additive genetic model, we identified a TBC1D4 loss-of-function variant with a large recessive impact on risk of type 2 diabetes in Greenlanders. The aim of the current study was to identify additional genetic variation underlying type 2 diabetes using a recessive genetic model, thereby increasing the power to detect variants with recessive effects.MethodsWe investigated three cohorts of Greenlanders (B99, n = 1401; IHIT, n = 3115; and BBH, n = 547), which were genotyped using Illumina MetaboChip. Of the 4674 genotyped individuals passing quality control, 4648 had phenotype data available, and type 2 diabetes association analyses were performed for 317 individuals with type 2 diabetes and 2631 participants with normal glucose tolerance. Statistical association analyses were performed using a linear mixed model.ResultsUsing a recessive genetic model, we identified two novel loci associated with type 2 diabetes in Greenlanders, namely rs870992 in ITGA1 on chromosome 5 (OR 2.79, p = 1.8 × 10−8), and rs16993330 upstream of LARGE1 on chromosome 22 (OR 3.52, p = 1.3 × 10−7). The LARGE1 variant did not reach the conventional threshold for genome-wide significance (p < 5 × 10−8) but did withstand a study-wide Bonferroni-corrected significance threshold. Both variants were common in Greenlanders, with minor allele frequencies of 23% and 16%, respectively, and were estimated to have large recessive effects on risk of type 2 diabetes in Greenlanders, compared with additively inherited variants previously observed in European populations.Conclusions/interpretationWe demonstrate the value of using a recessive genetic model in a historically small and isolated population to identify genetic risk variants. Our findings give new insights into the genetic architecture of type 2 diabetes, and further support the existence of high-effect genetic risk factors of potential clinical relevance, particularly in isolated populations.Data availabilityThe Greenlandic MetaboChip-genotype data are available at European Genome-Phenome Archive (EGA; https://ega-archive.org/) under the accession EGAS00001002641.

Highlights

Numerous genome-wide association studies (GWAS) have been performed to identify genetic factors predisposing to type 2 diabetes
The novel variants in ITGA1 and LARGE1 had a high impact on risk of type 2 diabetes in Greenlanders, evident by the estimated effect sizes (Table 1) and by the higher frequency of type 2 diabetes among homozygous carriers compared with non-carriers (24.9% and 28.9% vs 10%; Fig. 2)
Using a recessive genetic model, we investigated the association between markers on MetaboChip and type 2 diabetes in a Greenlandic study population

Summary

Introduction

Numerous genome-wide association studies (GWAS) have been performed to identify genetic factors predisposing to type 2 diabetes. GWAS of type 2 diabetes and glycaemic traits have almost exclusively been performed using an additive genetic model [1,2,3,4,5]. This model, has limited statistical power to detect associations with variants displaying recessive effects [9, 10], unless the effect is very large. Given the effective sample size of around 15,300, this study lacked statistical power to exclude the possibility that additional alleles with recessive effects predispose to type 2 diabetes in European populations

Objectives

Methods

Results

Conclusion