Diet-induced gene expression of isolated pancreatic islets from a polygenic mouse model of the metabolic syndrome

Abstract

Aims/hypothesis: Numerous new genes have recently been identified in genome-wide association studies for type 2 diabetes. Most are highly expressed in beta cells and presumably play important roles in their function. However, these genes account for only a small proportion of total risk and there are likely to be additional candidate genes not detected by current methodology. We therefore investigated islets from the polygenic New Zealand mouse (NZL) model of diet-induced beta cell dysfunction to identify novel genes and pathways that may play a role in the pathogenesis of diabetes.

Methods: NZL mice were fed a diabetogenic high-fat diet (HF) or a diabetes-protective carbohydrate-free HF diet (CHF). Pancreatic islets were isolated by laser capture microdissection (LCM) and subjected to genome-wide transcriptome analyses.

Results: In the prediabetic state, 2,109 islet transcripts were differentially regulated (>1.5-fold) between HF and CHF diets. Of the genes identified, 39 (e.g. Cacna1d, Chd2, Clip2, Igf2bp2, Dach1, Tspan8) correlated with data from the Diabetes Genetics Initiative and Wellcome Trust Case Control Consortium genome-wide scans for type 2 diabetes, thus validating our approach. HF diet induced early changes in gene expression associated with increased cell-cycle progression, proliferation and differentiation of islet cells, and oxidative stress (e.g. Cdkn1b, Tmem27, Pax6, Cat, Prdx4 and Txnip). In addition, pathway analysis identified oxidative phosphorylation as the predominant gene-set that was significantly upregulated in response to the diabetogenic HF diet.

Conclusions/interpretation: We demonstrated that LCM of pancreatic islet cells in combination with transcriptional profiling can be successfully used to identify novel candidate genes for diabetes. Our data strongly implicate glucose-induced oxidative stress in disease progression.

Fig. 1

A CHF diet protects from hyperglycaemia and diabetes. a Body weight of 17-week-old male NZL mice fed either HF or CHF. b Body fat content. c Blood glucose levels at week 17. d Plasma insulin levels at week 22. Values were derived from 14 to 16 animals per group. e Intraperitoneal glucose tolerance test at week 8, five animals per group. Values are mean ± SE. White circles, CHF plasma glucose; black circles, HF plasma glucose; white triangles, CHF plasma insulin; black triangles, HF plasma insulin

Fig. 2

Enrichment of selected islet genes in LCM samples. mRNA expression genes was determined by quantitative TaqMan RT-PCR for each gene and for Actb from LCM islets (white bars) and total pancreas (black bars) of 8-week-old mice. Expression levels are presented as change in C _t or the cycle number of each gene subtracted from the cycle number of beta-actin (C _t value 20.15) for the same cDNA sample. Lower numbers thus correspond to higher expression. Data are means ± SE from five CHF-fed NZL mice. *p ≤ 0.05

Fig. 3

Validation of differential expression in islets by quantitative real-time PCR. TaqMan quantitative PCR validation of genes differentially regulated between CHF (white bars) and HF (black bars). The change in expression is shown for selected genes involved in islet growth and development, protein processing and secretion, metabolism, and signalling. Values are normalised to Actb. Data are mean ± SE of five separate mice per group. *p ≤ 0.05

Fig. 4

Islet levels of ChREBP and regulation of lipogenic target genes. Immunohistochemical detection of ChREBP in pancreatic sections from male NZL mice. HF (a) and CHF fed (b) animals at week 8, with (c, d) respective controls with blocking peptide. e TaqMan quantitative PCR validation of ChREBP target genes in islets. Values are normalised to Actb. Data are mean ± SE of five separate mice per group; *p ≤ 0.05. White bars, CHF; black bars, HF; Grey bars, CHF liver; hatched bars, HF liver. Srebp-1 (also known as Srebf1)

Fig. 5

TaqMan quantitative PCR validation of OXPHOS genes in islets. The changes in expression of selected OXPHOS genes are shown. Values are normalised to Actb. Data are mean ± SE of five separate mice per group; *p ≤ 0.05. White bars, CHF; black bars, HF