G6PC2

Abstract: Recent genome-wide scans for association have identified a linkage disequilibrium block on chromosome 2 (q24.3), containing the G6PC2 and ABCB11 genes, in which common genetic variation has been shown to increase fasting glucose levels in the normal population [1,2]. The G6PC2 intergenic single nucleotide polymorphism (SNP) rs560887 has been reproducibly shown to alter fasting plasma glucose levels, with an increase of 0.06 mmol/l per copy of the variant [1,2]. Currently, studies have not conclusively established that the most strongly associated SNP is working through G6PC2 , but the consensus is that G6PC2 is a stronger candidate based on conditional analysis and biology [1–5]. G6PC2 encodes the islet-specific glucose-6-phosphatase catalytic subunit-related protein that is located within the endoplasmic reticulum of pancreatic B-cells [3]. The protein has been characterized as a B-cell-specific autoantigen target in Type 1 diabetes [6]. The function of G6PC2 in the pancreatic B-cell is unclear, but it is hypothesized that it may counteract glucose phosphorylation by glucokinase [4]. This hypothesis is supported by evidence from knockout mouse models, since mice have a mild metabolic defect with decreased fasting plasma glucose levels of 15% [4]. A common coding variant (P446L) in GCKR encoding glucokinase regulatory protein has recently been shown to be associated with decreased fasting glucose levels [7,8]. The only other published gene in which common genetic variation has been reproducibly shown to increase fasting glucose levels is glucokinase ( GCK ), where a common promoter variant increases levels by 0.06 mmol/l per copy of the variant [9]. Rare mutations in GCK cause a spectrum of monogenic forms of pancreatic B-cell dysfunction, including permanent neonatal diabetes (PNDM), maturity-onset diabetes of the young (MODY) and hyperinsulinaemia of infancy (HI) [10]. Therefore we hypothesized that rare mutations in G6PC2 may also play a role in the pathogenesis of monogenic forms of B-cell dysfunction, including PNDM, MODY and HI. To test this hypothesis we screened the five exons (including the intron–exon junctions), the 5 ′ and 3 ′ UTRs, and the promoter of G6PC2 using six pairs of M13 tailed primers to amplify the gene (Genbank reference NM_001081686). Primer details are available as supplementary online methods information. Amplicons were sequenced on an ABI3700 DNA analyser using standard protocols and analysis was performed using Mutation Surveyor v3.2 (Softgenetics, Cambridge, UK). Sequence variants were confirmed by re-amplifying a fresh dilution of DNA. Co-segregation in family members and the presence of variants in non-diabetic control subjects were used to assess the pathogenicity of genetic variants identified. We sequenced the four groups of patients with monogenic diabetes phenotypes in whom the genetic aetiology was unknown. These groups were: 72 unrelated subjects with permanent diabetes diagnosed before 6 months of age, with a median age of onset of 5 weeks (range 0–26); 116 subjects with a MODY-like phenotype diagnosed with diabetes before 25 years of life [median 18 years (3–25)], body mass index (BMI) ≤ 30 kg/m 2 , non-insulin dependent, and an affected first-degree relative (114/116); 42 subjects with a phenotype suggesting a defect in glucokinase but with no abnormality of the GCK gene, GCK-like MODY, diagnosed with diabetes before the age of 30 years [median 18 (1–30)], a mean fasting plasma glucose of 6.6 mmol/l and a mean BMI of 23.6 kg/m 2

Abstract: The endoplasmic reticulum enzyme glucose-6-phosphatase catalyzes the common terminal reaction in the gluconeogenic/glycogenolytic pathways and plays a central role in glucose homeostasis. In most mammals, different G6PC subunits are encoded by three paralogous genes (G6PC, G6PC2, and G6PC3). Mutations in G6PC and G6PC3 are responsible for human mendelian diseases, whereas variants in G6PC2 are associated with fasting glucose (FG) levels. We analyzed the evolutionary history of G6Pase genes. Results indicated that the three paralogs originated during early vertebrate evolution and that negative selection was the major force shaping diversity at these genes in mammals. Nonetheless, site-wise estimation of evolutionary rates at corresponding sites revealed weak correlations, suggesting that mammalian G6Pases have evolved different structural features over time. We also detected pervasive positive selection at mammalian G6PC2. Most selected residues localize in the C-terminal protein region, where several human variants associated with FG levels also map. This region was re-sequenced in ~560 subjects from Saudi Arabia, 185 of whom suffering from type 2 diabetes (T2D). The frequency of rare missense and nonsense variants was not significantly different in T2D and controls. Association analysis with two common missense variants (V219L and S342C) revealed a weak but significant association for both SNPs when analyses were conditioned on rs560887, previously identified in a GWAS for FG. Two haplotypes were significantly associated with T2D with an opposite effect direction. We detected pervasive positive selection at mammalian G6PC2 genes and we suggest that distinct haplotypes at the G6PC2 locus modulate susceptibility to T2D.

Abstract: Aims/hypothesis Autoimmunity to insulin, glutamic acid decarboxylase and the tyrosine-phosphatase-like protein IA-2 is associated with type 1 diabetes. The production of self-molecules in thymus and secondary lymphoid tissues is critical for self-tolerance; reduced levels may impair tolerance and predispose to autoimmunity, as shown for insulin. Alternative splicing causes differential expression of IA-2 gene (PTPRN) transcripts and IA-2 protein in human thymus and spleen compared with pancreas. IA-2 sequences not present in lymphoid tissues become autoimmune targets in type 1 diabetes. The beta cell molecule islet-specific glucose-6-phosphatase catalytic subunit-related protein (IGRP) is an autoantigen in the non-obese diabetic (NOD) mouse, a model of type 1 diabetes. IGRP is a candidate autoantigen in the human disease, but robust assays for IGRP autoantibodies and/or autoreactive T cells are not available. Both full-length and IGRP splice variants encoded by the G6PC2 gene are expressed in the pancreas. In this study we tested the hypothesis that IGRP splice variants could be differentially expressed in thymus and spleen compared with the pancreas.