Prediabetes blunts DPP4 genetic control of postprandial glycaemia and insulin secretion

Abstract

Aims/hypothesis: Imbalances in glucose metabolism are hallmarks of clinically silent prediabetes (defined as impaired fasting glucose and/or impaired glucose tolerance) representing dysmetabolism trajectories leading to type 2 diabetes. CD26/dipeptidyl peptidase 4 (DPP4) is a clinically proven molecular target of diabetes-controlling drugs but the DPP4 gene control of dysglycaemia is not proven.

Methods: We dissected the genetic control of post-OGTT and insulin release responses by the DPP4 gene in a Portuguese population-based cohort of mainly European ancestry that comprised individuals with normoglycaemia and prediabetes, and in mouse experimental models of Dpp4 deficiency and hyperenergetic diet.

Results: In individuals with normoglycaemia, DPP4 single-nucleotide variants governed glycaemic excursions (rs4664446, p=1.63x10^-7) and C-peptide release responses (rs2300757, p=6.86x10^-5) upon OGTT. Association with blood glucose levels was stronger at 30 min OGTT, but a higher association with the genetic control of insulin secretion was detected in later phases of the post-OGTT response, suggesting that the DPP4 gene directly senses glucose challenges. Accordingly, in mice fed a normal chow diet but not a high-fat diet, we found that, under OGTT, expression of Dpp4 is strongly downregulated at 30 min in the mouse liver. Strikingly, no genetic association was found in prediabetic individuals, indicating that post-OGTT control by DPP4 is abrogated in prediabetes. Furthermore, Dpp4 KO mice provided concordant evidence that Dpp4 modulates post-OGTT C-peptide release in normoglycaemic but not dysmetabolic states.

Conclusions/interpretation: These results showed the DPP4 gene as a strong determinant of post-OGTT levels via glucose-sensing mechanisms that are abrogated in prediabetes. We propose that impairments in DPP4 control of post-OGTT insulin responses are part of molecular mechanisms underlying early metabolic disturbances associated with type 2 diabetes.

Keywords: CD26/DPP4; Dysglycaemia; Genetic association; Hyperenergetic diet; Hyperinsulinaemia; Insulin secretion; Postprandial glucose; Prediabetes.

Fig. 1

DPP4 genetic association with post-OGTT plasma glucose and C-peptide levels in NGT and prediabetic participants. (a, b) The AUC of the glucose excursion (a) and the C-peptide response AUC (0–120 min) (b) were calculated for each participant. The violin plots represent the probability density, the median, the IQR and the 95% CI of the phenotype distributions in 736 NGT (blue) and 233 prediabetic (orange) participants. ***p<0.001, unpaired t test (Mann–Whitney). (c) Correlation of plasma glucose AUC and C-peptide AUC (0–120 min) during OGTT, in NGT participants (blue circles; Pearson’s correlation, r²=0.193, p<0.001) and in prediabetic participants (orange circles; Pearson’s correlation, r²=0.055, p=0.0002). (d, e) Plots of quantitative trait locus analysis for 33 SNPs in the DPP4 gene region, testing for association with (d) plasma levels of glucose AUC and (e) C-peptide AUC (0–120 min) for NGT participants (blue circles) and prediabetic participants (orange circles). Results represent the nominal –log10 (p value) for allelic association and the SNP position in chromosome 2 is represented in Mb and a bar representing the DPP4 gene region is shown above the plots. (f, g) Violin plots of rs2909449 genotypic effects on plasma glucose AUC (f) and C-peptide AUC (0–120 min) (g) during OGTT in NGT participants (green, ancestral allele homozygotes; blue, heterozygotes; yellow, minor allele homozygotes). *p<0.05, **p<0.01 and ***p<0.001 by Kruskal–Wallis test with Dunn’s correction for multiple comparisons. The plots represent the probability density, the median, the IQR and the 95% CI of the phenotypic distributions per genotype class totalling 736 NGT participants. Chr, chromosome

Fig. 2

Dpp4 is implicated in mouse experimental post-OGTT responses under normal diet or HCD. (a) Blood glucose levels before and at 30 min (Gluc T30) after gavage with 1.5g/kg glucose (Gluc). (b, c) mRNA expression of the mouse Dpp4 gene at T30 after gavage with 1.5g/kg glucose (Gluc T30) or with water was quantified in portions of the duodenum (Duod.), jejunum (Jeju.) and ileum (Ile.) (b) and the liver (c) by real-time PCR. Expression levels were normalised to those of the mouse endogenous control Gapdh and the mean value for duodenum (b) or liver (c) from control mice gavaged with water. Values are represented on a logarithmic scale, n=5–25 mice per group. (d) Serum C-peptide levels were measured by ELISA at baseline (0) and 15, 30, 60 and 120 min after OGTT in C57BL/6 (B6) and Dpp4 KO mice on regular chow or after 6 weeks of HCD. The plot represents the AUC (0–120 min), n=4–12 mice per group. *p<0.05, **p<0.01 and ***p<0.001 (one-way ANOVA using Tukey’s correction)