Type 2 diabetes is one of the major threats to global health. Around 10% of the world’s population either has type 2 diabetes, or is likely to develop it during their lives. Both genes and environment contribute to an individual’s risk of developing this disease. Preventative strategies currently available (for example, sustainable changes to diet or levels of exercise) are, in most people, of limited effectiveness. Our available treatment options, though they can improve health and reduce the risk of complications, rarely result in the restoration of a normal metabolism.

As a result, there is a pressing need for an improved understanding of the mechanisms involved in T2D and its complications. Human genetics is one of the most powerful ways of delivering such an understanding, because, when successful, it highlights fundamental biological differences that play directly into differences in risk of disease. Discoveries obtained through human genetics can therefore point the way towards the development of novel preventative and therapeutic approaches. Not only that, such discoveries can help to target those approaches more effectively to the individuals in whom they are most likely to provide benefit, and least likely to cause harm.

Manhattan plot from T2D Paper showing single-variant analysis.

Manhattan plot from T2D Paper showing single-variant analysis.

The paper in Nature, which has just been published, reports on a major international effort (involving over 300 scientists in 22 countries) to use human genetics to define mechanisms underlying the development of type 2 diabetes. The main technical and scientific advance deployed in this particular study has been to move discovery of DNA sequence differences influencing type 2 diabetes risk beyond the common (shared) variants examined in previous genome-wide association studies. Instead, this study used newly available DNA sequencing technologies to explore the full inventory of DNA sequence changes (shared and unique) in multiple individuals. We were then able to compare how those DNA sequence differences are distributed between individuals who have type 2 diabetes and those who do not.

This study of the genetics of type 2 diabetes is unprecedented in both scale and scope. It involves data generated from over 120,000 individuals (some with diabetes, some without) from a wide range of ethnic groups. It includes individuals with ancestral origins in Europe, South and East Asia, the Americas and Africa. In some of these individuals, the entire genome was sequenced. In others, there was a specific focus on variation within the parts of the genome (the “exome”) that codes directly for proteins: these changes in protein-coding sequence are likely to be especially informative from a biological perspective.

There are three main findings from this research.

First, we find over a dozen genes that contain T2D associated DNA variants which change amino acid sequence. Several of these provide new and important clues about the mechanisms underlying type 2 diabetes. For example, we find a DNA sequence variant in a gene called PAX4 which is powerfully associated with diabetes, but only in individuals from East Asia (including Korea, China, Singapore). PAX4 is involved in the development of the insulin-producing beta-cells in the human pancreas. We also implicate another gene, TM6SF2, already known to be involved in the development of hepatic steatosis (“fatty liver”) and which we show here also influences T2D risk. These finding matters because they provide many important new insights into the biology of diabetes: some of the genes and pathways implicated may represent novel avenues for drug development.

Second, there has been a longstanding debate (going back over a century) as to whether most of the genetic differences that influence individual predisposition to common diseases such as diabetes are ones that are widely shared within populations, or whether they are more often rare or unique events, specific to an individual and their family. Because, in this study, we were able to analyse both rare and shared variants, we have been able to show that the genetic contribution to diabetes risk lies predominantly at shared sites. This matters because it has implications for the ways in which we will be able to use genetic data to support personalised medicine.

Third, although most of the association signals we detect involve common, shared variants, we do find some rare (private, unique) variants that influence risk of diabetes, and show that these can also provide valuable insights into disease biology. We had already shown, a couple of years ago, using some of the same data, that rare variants that abrogate function at the SLC30A8 gene, are protective against type 2 diabetes. Here, we show that there is an excess of T2D association amongst rare coding variants in a set of around 30 genes that are already known to be involved in some rare familial forms of diabetes that mostly start in early life. This matters because it demonstrates that the same genes can harbour DNA sequence differences which result in very different types of diabetes: it highlights the need for careful interpretation when DNA sequence changes are detected in genes of medical significance, since it will not always be obvious what the impact of that variant is likely to be.

Because, we believe it is important that all researchers can benefit from the data we have generated in this project, data and discoveries are available to researchers and to the wider world through a variety of means. For example, much of the data from this and other studies is available on the freely-accessible T2D genetics portal developed as part of the Accelerating Medicines Partnership (www.type2diabetesgenetics.org).

Robert Turner Professor of Diabetic Medicine, Group Head, Wellcome Trust Centre for Human Genetics, Group Head / PI, Grant Holding Senior Scientist, Consultant Physician.