Diabetes genes identified by genome-wide association studies are regulated in mice by nutritional factors in metabolically relevant tissues and by glucose concentrations in islets

Abstract

Background: Genome-wide association studies (GWAS) have recently identified many new genetic variants associated with the development of type 2 diabetes. Many of these variants are in introns of known genes or between known genes, suggesting they affect the expression of these genes. The regulation of gene expression is often tissue and context dependent, for example occurring in response to dietary changes, hormone levels, or many other factors. Thus, to understand how these new genetic variants associated with diabetes risk may act, it is necessary to understand the regulation of their cognate genes.

Results: We identified fourteen type 2 diabetes-associated genes discovered by the first waves of GWAS for which there was little prior evidence of their potential role in diabetes (Adam30, Adamts9, Camk1d, Cdc123, Cdkal1, Cdkn2a, Cdkn2b, Ext2, Hhex, Ide, Jazf1, Lgr5, Thada and Tspan8). We examined their expression in metabolically relevant tissues including liver, adipose tissue, brain, and hypothalamus obtained from mice under fasted, non-fasted and high fat diet-fed conditions. In addition, we examined their expression in pancreatic islets from these mice cultured in low and high glucose. We found that the expression of Jazf1 was reduced by high fat feeding in liver, with similar tendencies in adipose tissue and the hypothalamus. Adamts9 expression was decreased in the hypothalamus of high fat fed mice. In contrast, the expression of Camk1d, Ext2, Jazf1 and Lgr5 were increased in the brain of non-fasted animals compared to fasted mice. Most notably, the expression levels of most of the genes were decreased in islets cultured in high glucose.

Conclusions: These data provide insight into the metabolic regulation of these new type 2 diabetes genes that will be important for determining how the GWAS variants affect gene expression and ultimately the development of type 2 diabetes.

Figure 1

Regulation of new diabetes genes by nutritional status in the liver. Data are shown as fold-change relative to those observed in the fasted chow-fed mice. The number of samples per group for the fasted chow-fed mice, fasted HFD-fed mice and non-fasted (fed) chow-fed mice, respectively are Adamts9 (7, 7, 4), Hhex (7, 5, 5), Jazf1 (7, 7, 8). P-values are shown for the indicated group compared to the fasted chow-fed controls. Data are shown as the fold change (2^ΔΔCt) ± 2^ΔΔCt±SE[87]. * P < 0.05, ** P < 0.01.

Figure 2

Regulation of new diabetes genes by nutritional status in adipose tissue. Data are shown as fold-change relative to those observed in the fasted chow-fed mice (n = 7 for the fasted chow-fed group, 6 for the fasted HFD-fed mice and 8 for the fed chow-fed mice). P-values are for the comparison between the indicated group and the fasted, chow-fed control mice. Data are shown as the fold change (2^ΔΔCt) ± 2^ΔΔCt±SE[87]. * P < 0.05.

Figure 3

Regulation of new diabetes genes by nutritional status in the hypothalamus. Data are shown as fold-change relative to those observed in the fasted chow-fed mice. The number of samples per group for the fasted chow-fed mice, fasted HFD-fed mice and non-fasted (fed) chow-fed mice, respectively are Adamts9 (7, 5, 4), Camk1d (7, 5, 5), Cdkal1 (10, 5, 9), Jazf1 (10, 8, 7), Thada (7, 5, 5). P-values are shown for the indicated group compared to the fasted chow-fed controls. Data are shown as the fold change (2^ΔΔCt) ± 2^ΔΔCt±SE[87]. * P < 0.05, ** P < 0.01.

Figure 4

Regulation of new diabetes genes by nutritional status in the remainder of the brain. Data are shown as fold-change relative to those observed in the fasted chow-fed mice. HFD = high fat diet. The number of samples per group for the fasted chow-fed mice, fasted HFD-fed mice and non-fasted (fed) chow-fed mice, respectively are (6, 5, 10) except for Adamts9 (6, 4, 10). P-values are shown for the indicated group compared to the fasted chow-fed controls. Data are shown as the fold change (2^ΔΔCt) ± 2^ΔΔCt±SE[87]. * P < 0.05, ** P < 0.01.

Figure 5

Regulation of new diabetes genes by glucose levels in pancreatic islets. Data are shown as fold-change, (2^ΔΔCt) ± 2^ΔΔCt±SE[87], relative to those observed in the islets incubated in low (5 mM) glucose. Each group is the average of three replicates, each of which was comprised of pooled islets from two mice. * P < 0.05, *** P < 0.001.