Heterozygous RFX6 protein truncating variants are associated with MODY with reduced penetrance

Abstract

Finding new causes of monogenic diabetes helps understand glycaemic regulation in humans. To find novel genetic causes of maturity-onset diabetes of the young (MODY), we sequenced MODY cases with unknown aetiology and compared variant frequencies to large public databases. From 36 European patients, we identify two probands with novel RFX6 heterozygous nonsense variants. RFX6 protein truncating variants are enriched in the MODY discovery cohort compared to the European control population within ExAC (odds ratio = 131, P = 1 × 10^-4). We find similar results in non-Finnish European (n = 348, odds ratio = 43, P = 5 × 10^-5) and Finnish (n = 80, odds ratio = 22, P = 1 × 10^-6) replication cohorts. RFX6 heterozygotes have reduced penetrance of diabetes compared to common HNF1A and HNF4A-MODY mutations (27, 70 and 55% at 25 years of age, respectively). The hyperglycaemia results from beta-cell dysfunction and is associated with lower fasting and stimulated gastric inhibitory polypeptide (GIP) levels. Our study demonstrates that heterozygous RFX6 protein truncating variants are associated with MODY with reduced penetrance.Maturity-onset diabetes of the young (MODY) is the most common subtype of familial diabetes. Here, Patel et al. use targeted DNA sequencing of MODY patients and large-scale publically available data to show that RFX6 heterozygous protein truncating variants cause reduced penetrance MODY.

Fig. 1

Extended pedigree of non-Finnish European patients identified in the discovery cohort. a Pedigree of family 1 that were identified with heterozygous RFX6 variant (NM_173560.3:c.875-T > G,p.Leu292Ter) from the discovery cohort. b Pedigree of family 2 from the discovery cohort with heterozygous RFX6 variant (NM_173560.3:c.1051-A > T, -p.Lys351Ter). Genotype is shown underneath each symbol; M and N denote mutant and wild-type alleles, respectively. Directly below the genotype is the age of diabetes onset in years, duration in years, BMI and treatment at study entry. Squares represent male family members, and circles represent female members. Black-filled symbols denote patients with diabetes. An arrow denotes the proband in the family. OHA, oral hypoglycaemic agents. *age at recruitment. One of the daughters of patient III.1 in family 2 had a history of gestational diabetes

Fig. 2

Penetrance of diabetes in people with MODY. Heterozygous RFX6 PTV (n = 18), pathogenic HNF1A variant (n = 1265) or HNF4A variant (n = 427)

Fig. 3

Phenotypic characteristics of the Finnish RFX6 p.His293Leufs heterozygotes without diabetes. Figure shows the phenotypic characteristics at fasting and during an oral glucose tolerance test (OGTT) of the Finnish RFX6 p.His293Leufs heterozygotes without diabetes (filled symbols; n = 11) compared with population controls from the PPP-Botnia Study (open symbols; N = 55) matched for age, sex and BMI. a Median and interquartile range of plasma glucose concentrations during an OGTT (N = 11 vs. 55). b Median and interquartile range of serum insulin concentrations during an OGTT (N = 11 vs. 55). c Individual fasting GIP concentration for RFX6 heterozygotes and controls (N = 6 vs. 55). d Individual 120-minute GIP concentration during an OGTT (N = 8 vs. 55). e Individual fasting GLP-1 concentration (N = 6 vs. 55). f Individual 120 min GLP-1 concentration during an OGTT (N = 6 vs. 55). The P-values (Mann-Whitney U test) under 0.05 for the difference in a Glucose 0 min, P = 0.02; b Insulin 30 min, P = 0.015; c fasting GIP, P = 8.6 × 10⁻³ d 120-minute GIP, P = 0.029; e 120-minute GLP-1, P = 0.047;  +/− , heterozygous for RFX6 variant; −/−, controls without RFX6 variant. Horizontal black line in panels c-f represents median for each group