A genetic screen identifies Crat as a regulator of pancreatic beta-cell insulin secretion

Glucose-stimulated insulin secretion is a critical function in the regulation of glucose homeostasis, and its deregulation is associated with the development of type 2 diabetes. Here, we performed a genetic screen using islets isolated from the BXD panel of advanced recombinant inbred (RI) lines of mice to search for novel regulators of insulin production and secretion.

Pancreatic islets were isolated from 36 RI BXD lines and insulin secretion was measured following exposure to 2.8 or 16.7 mM glucose with or without exendin-4. Islets from the same RI lines were used for RNA extraction and transcript profiling. Quantitative trait loci (QTL) mapping was performed for each secretion condition and combined with transcriptome data to prioritize candidate regulatory genes within the identified QTL regions. Functional studies were performed by mRNA silencing or overexpression in MIN6B1 cells and by studying mice and islets with beta-cell-specific gene inactivation.

Insulin secretion under the 16.7 mM glucose plus exendin-4 condition was mapped significantly to a chromosome 2 QTL. Within this QTL, RNA-Seq data prioritized Crat (carnitine O-acetyl transferase) as a strong candidate regulator of the insulin secretion trait. Silencing Crat expression in MIN6B1 cells reduced insulin content and insulin secretion by ∼30%. Conversely, Crat overexpression enhanced insulin content and secretion by ∼30%. When islets from mice with beta-cell-specific Crat inactivation were exposed to high glucose, they displayed a 30% reduction of insulin content as compared to control islets. We further showed that decreased Crat expression in both MIN6B1 cells and pancreatic islets reduced the oxygen consumption rate in a glucose concentration-dependent manner.

We identified Crat as a regulator of insulin secretion whose action is mediated by an effect on total cellular insulin content; this effect also depends on the genetic background of the RI mouse lines. These data also show that in the presence of the stimulatory conditions used the insulin secretion rate is directly related to the insulin content.