Dr. Constantin Polychronakos obtained his medical degree from the Aristotelian University in Greece, with subsequent training and career in Canada. He is a professor in the Departments of Paediatrics and Human Genetics at the McGill University Health Centre, Montreal and Editor-in-Chief of the Journal of Medical Genetics. His main research interest is in the genetics of complex traits, where he published the first genome-wide association studies for type 2 and type 1 diabetes (Sladek et al., Nature 2007, Hakonarson et al., Nature 2007), as well as investigations in transcriptional and translational eQTLs (Vafiadis et al., Nat Genet 1997, Li et al., Nat Commun 2012). His research has also addressed monogenic diabetes, with the discovery of RFX6 as the gene mutated in the Mitchell-Riley syndrome (Smith et al., Nature 2010). A more recent interest is in the genetics of diabetes in East Asian populations, with extensive collaborations in China, where he is Chief Scientific Officer of MaiDa Gene Tech Inc. His expertise and interest in both Mendelian and complex genetics led him to investigate questions of penetrance in monogenic diabetes, which will be the focus of his talk at this conference.
Penetrance as a complex trait
The broad penetrance-expressivity spectrum in the phenotype of monogenic diseases must be due, at least in part, to modifier genes in trans. These are likely to behave as complex traits, requiring large sample sizes to dissect genome-wide, a logistic impossibility. Monogenic diabetes, due to mutations in any one of 18 genes, may offer a window into strong effects by candidate modifiers because of the well-defined physiology of the pancreatic beta cells and the plethora of the genes known to disrupt it.
As part of a large-scale exome sequencing of autoantibody-negative Chinese patients diagnosed with type 1 diabetes to detect misdiagnosed monogenic cases, we examined co-existence of mutations in more than one of the known genes. We used variant frequencies in the East Asian ExAC cohort (8.5 k subjects) to calculate the probability of protein-altering variants in more than one of 18 monogenic diabetes genes. For variants not in the database, we arbitrarily (and conservatively) assumed one instance in the 8.5 k cohort.
Among 20 patients with ACMG category 1-3 variants in the target genes (MAF<0.0001), we discovered no fewer than 9 with variants in a second gene (p=3.9 x 10-9), by relaxing the MAF cut-off to 0.001.
Our findings are consistent with digenic inheritance of monogenic diabetes, i.e. a phenotypic modulation by modifier variants in genes that can cause the disease when mutated.