Inherited β-Cell Dysfunction in Lean Individuals With Type 2 Diabetes

The syndromes comprising diabetes mellitus share the common feature of β-cell failure. Inheritance of β-cell failure is clear in patients diagnosed with the MODY (maturity-onset diabetes of the young, also referred to as monogenic diabetes in youth) syndromes (1), as well as in permanent neonatal diabetes mellitus (2) where mutations in genes essential for insulin secretion associate with elevated blood glucose. Causative alleles for β-cell dysfunction have been identified for the monogenic forms of diabetes. Likewise, a role for genes impacting glucose-stimulated insulin secretion has been proposed for some polygenic forms of diabetes mellitus. Genome-wide association studies have not associated single nucleotide polymorphisms (SNPs) in genes required for optimal insulin secretion with autoimmune type 1A diabetes. In contrast, latent autoimmune diabetes of adults has been associated with TCF7L2 (3), a gene associated with reduced insulin secretion (4). TCF7L2 is one of the most significant loci for type 2 diabetes (5) and, along with HNF1B/HNF4A (6) and KCNJ11 (7), tie genes involved in β-cell dysfunction to risk for developing type 2 diabetes in overweight individuals. In this issue of Diabetes, a new report (8) proposes that a recently identified type 2 diabetes risk gene is linked to dysfunctional glucose-stimulated insulin secretion (GSIS) in lean Japanese patients with type 2 diabetes.

Lean type 2 diabetes is highly prevalent in Japan where more than half of the patients with this condition are considered normal weight (BMI <25) (9,10). The contrasting general clinical phenotypes of obese individuals with type 2 diabetes compared with lean individuals with type 2 diabetes have recently been detailed (9). In this report of 40 case subjects with a diabetes duration of 10–12 years, patients had an average age of onset of 56 ± 2 years, low BMI (21 kg/m²), HbA_1c of 6.7 ± 0.1%, and fasting blood glucose of 144 ± 9 mg/dL. Diabetes in lean Japanese patients has been associated with a low level of fasting insulin (32 ± 17 pmol/L) in addition to reduced peripheral glucose uptake (2.3 ± 0.6 mg/kg/min) and elevated endogenous glucose production (13.8 ± 1.8 µmol/kg/min) (11). These data suggest that this disorder in normal-weight individuals results from impaired insulin secretion and action.

Studies to identify the genes contributing to type 2 diabetes in lean individuals identified a region on chromosome 21q (12). Follow-up examination of this region led to the identification of KCNJ15 (potassium inwardly-rectifying channel, subfamily J, member 15) as a susceptibility gene (13). KCNJ15, also known as Kir4.2, is expressed in a wide variety of tissues and cells including peripheral blood, kidney, and pancreatic islet cells (13). The risk allele mRNA is present at elevated levels due to increased mRNA stability. Overexpression of KCNJ15 in the rat β-cell line INS1 resulted in a significant reduction in GSIS. These data support a role for the KCNJ15 risk allele as causative in β-cell dysfunction observed in lean type 2 diabetic patients.

In this issue of Diabetes, Okamoto et al. (8) examined the mechanism whereby KCNJ15 inhibits GSIS using both in vitro and in vivo approaches and made several important observations. First, expression of KCNJ15 is positively regulated by glucose. Second, the expression of KCNJ15 was higher in the pancreatic islets of patients with type 2 diabetes. Additionally, knockdown of KCNJ15 in INS1 cells resulted in an approximate twofold increase in GSIS. Likewise, using a small inhibitory RNA approach, the authors were able to convincingly improve GSIS in both euglycemic and diabetic mouse models through knockdown of KCNJ15. However, it should be noted that while the suppression of this gene in diabetic mice did increase insulin secretion, it did not fully correct blood glucose levels to within the normal range. Finally, it is reported that KCNJ15 associates with calcium-sensing receptor in the plasma membrane, and the authors proposed that this interaction blocks the KCNJ15 activity allowing for stabilization of “post-prandial” membrane potential.

Unlike the MODY syndromes in which a single genetic defect results in decreased insulin secretion, the pathogenesis of type 2 diabetes in normal-weight individuals likely requires the contribution of multiple genetic regions (12). The data by Okamoto et al. (8) provide an important step forward in our understanding of β-cell function/dysfunction and offer a paradigm to study the singular impact of a risk gene within the pathology of diabetes. Once a mechanistic role has been elucidated, this provides a further platform to investigate interactions with other identified risk alleles. Moving toward expression of the specific risk allele in human β-cells, primary or a cell line that secretes insulin (such as the recently developed EndoC-βH1 [14]), will provide a further understanding of how a synonymous SNP modifies protein function in a human system. It should also be pointed out that many of the SNPs associated with polygenic forms of diabetes are in noncoding regions or induce synonymous changes. Therefore, these studies provide a model for studying polymorphisms that impact protein activity or function but do not alter protein sequence or a promoter region.

In light of these new data (8), the authors suggest KCNJ15 as a new pharmacological target for individuals with reduced insulin secretion. KCNJ15 depletion or inhibition with a small molecule may be an attractive therapeutic in any system where GSIS is decreased. This is strongly supported by the data provided by whole animal systems where knockdown of Kcnj15 led to elevated insulin secretion in mice without diabetes and in mice with diabetes induced by a mutation in the insulin 2 (Ins2) gene, that is presumably independent of Kcnj15. However, caution should be exercised. KCNJ15 appears to have a major role in renal function (15,16), and diabetic patients represent a population at increased risk for renal failure (17).

In conclusion, Okamoto et al. (8) present compelling evidence that KCNJ15 plays a role in reduced insulin secretion and the C566T SNP is likely causative in type 2 diabetes. However, further work is required for a systematic understanding of how KCNJ15 blocks GSIS.