Autosomal recessive insulin-dependent diabetes

Background and hypothesis: Type 1 diabetes (T1DM) is due to the autoimmune destruction of the insulin-producing pancreatic beta-cells. T1DM is a multifactorial disease with a strong genetic (polygenic) component. However, not all cases of insulin-dependent diabetes starting early in life are of autoimmune and polygenic etiology. Indeed, neonatal diabetes and MODYs (Maturity-Onset Diabetes of the Young) are both monogenic diseases that can still be misdiagnosed as T1DM. Our hypothesis is that a new form of monogenic, autosomal recessive, non-autoimmune, insulin-deficient, young-onset diabetes exists but has escaped identification because of lack of strongly suggestive family or clinical history. Objective: To discover genes whose mutations are responsible for this type of diabetes by exome sequencing of selected multiplex families diagnosed as T1DM. Materials and Methods: Data from 2,345 sibling pairs affected with T1DM, from the T1DM Genetics Consortium (T1DGC) were analyzed for the presence of autoimmune markers (3 available autoantibody tests (GAD65, ZNT8 and IA2)). From 167 families where both affected siblings were autoantibody-negative, we found 37 families in which both were also negative for any of the known predisposing HLA haplotypes (e.g. HLA-DRB1*04/DQB1*0302 or HLA-DRB1*0301/DQB1*0201). Although some of these 37 families may still have autoimmune T1DM, we posited that most could have diabetes for a reason other than autoimmunity. DNA from all members of these 37 families was obtained from the NIDDK repository and DNA from 14 affected children was sent to McGill/Génome Québec Innovation Centre for whole-exome sequencing (WES). Results: Before filtering WES results to select new candidate genes, we identified 3 families with mutations in genes already known to be involved in monogenic diabetes: KCNJ11 and WFS1. Thereafter, we selected 20 new candidate genes according to 3 criteria: 2 or more protein-altering variants, positive LOD (Logarithm of Odds) score in the nuclear family, and significant expression of the gene in the pancreatic islets (RPKM (reads per kb per million) higher than 3). We genotyped probands and their families for SNVs found in 16 of those candidate genes and we found 3 genes carrying variants that segregated with the disease under a recessive model (NCKAP5L, HSPBAP1 and SCAMP2). We sequenced all families linked to those 3 loci but we did not find any other family with mutations inherited on a recessive Mendelian fashion within one of those 3 genes. Based on our first findings (KCNJ11 and WFS1 diabetes misclassified as T1DM), we posited that some patients with MODY3, the most frequent type of MODY, must also be misdiagnosed as T1DM. Among all samples listed in the NIDKK repository, we selected families where one parent was diabetic and where the affected children and the affected parent were negative for autoimmune markers. Nine families fulfilled these criteria. All protein-coding exons of HNF-1α were sequenced in 2 families that showed linkage at the HNF-1a locus and whose DNA was available. We found c.599 G>A missense mutation (p.R200Q) within exon 3 in the affected mother and in the 3 affected children of the first family, confirming MODY3 diagnosis. Conclusions: Different cases of monogenic diabetes are still misdiagnosed as T1DM. Usually, clinical and/or family history should help us making the right diagnosis. Some monogenic diabetes (e.g. KCNJ11) represent life-changing diagnosis as they can be treated with sulfonylureas for perfect glycemic control without risk of hypoglycemia. In addition to missed diagnosis, some mutations appear to alter to a lesser extent the protein function which can influence clinical evolution of the disease and explain misdiagnosis. In the same way, we have identified three highly likely candidate genes for a new form of autosomal recessive form of diabetes that remain to be confirmed in at least 1 additional unrelated proband.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.