A cautionary tale: the non-causal association between type 2 diabetes risk SNP, rs7756992, and levels of non-coding RNA, CDKAL1-v1

• Published in: Diabetologia Vol. 58; no. 4; pp. 745 - 748
• Main Authors: Locke, Jonathan M., Wei, Fan-Yan, Tomizawa, Kazuhito, Weedon, Michael N., Harries, Lorna W.
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.04.2015; Springer Nature B.V
• Subjects: Case-Control Studies; Cyclin-Dependent Kinase 5 - blood; Cyclin-Dependent Kinase 5 - genetics; Diabetes; Diabetes Mellitus, Type 2 - diagnosis; Diabetes Mellitus, Type 2 - ethnology; Diabetes Mellitus, Type 2 - genetics; European Continental Ancestry Group - genetics; Genetic Predisposition to Disease; Genome-Wide Association Study; Genomes; Genotype & phenotype; Human Physiology; Humans; Insulin; Internal Medicine; Introns; Islets of Langerhans - chemistry; Linear Models; Medicine; Medicine & Public Health; Metabolic Diseases; MicroRNAs; Phenotype; Polymorphism, Single Nucleotide; Regression analysis; Risk Factors; RNA, Untranslated - blood; RNA, Untranslated - genetics; Short Communication; tRNA Methyltransferases; United Kingdom - epidemiology; United Kingdom > UK; United States > US; Type 2 diabetes; Expression quantitative trait loci; Genome-wide association studies; Non-coding RNA
• ISSN: 0012-186X; 1432-0428
• DOI: 10.1007/s00125-015-3508-9

Aims/hypothesis Intronic single nucleotide polymorphisms (SNPs) in the CDKAL1 gene are associated with risk of developing type 2 diabetes. A strong correlation between risk alleles and lower levels of the non-coding RNA, CDKAL1-v1 , has recently been reported in whole blood extracted from Japanese individuals. We sought to replicate this association in two independent cohorts: one using whole blood from white UK-resident individuals, and one using a collection of human pancreatic islets, a more relevant tissue type to study with respect to the aetiology of diabetes. Methods Levels of CDKAL1-v1 were measured by real-time PCR using RNA extracted from human whole blood ( n  = 70) and human pancreatic islets ( n  = 48). Expression with respect to genotype was then determined. Results In a simple linear regression model, expression of CDKAL1-v1 was associated with the lead type 2 diabetes-associated SNP, rs7756992, in whole blood and islets. However, these associations were abolished or substantially reduced in multiple regression models taking into account rs9366357 genotype: a moderately linked SNP explaining a much larger amount of the variation in CDKAL1-v1 levels, but not strongly associated with risk of type 2 diabetes. Conclusions/interpretation Contrary to previous findings, we provide evidence against a role for dysregulated expression of CDKAL1-v1 in mediating the association between intronic SNPs in CDKAL1 and susceptibility to type 2 diabetes. The results of this study illustrate how caution should be exercised when inferring causality from an association between disease-risk genotype and non-coding RNA expression.