Multi-Omics Comparison of the Spontaneous Diabetes Mellitus and Diet-Induced Prediabetic Macaque Models

Abstract

The prevalence of diabetes mellitus has been increasing for decades worldwide. To develop safe and potent therapeutics, animal models contribute a lot to the studies of the mechanisms underlying its pathogenesis. Dietary induction using is a well-accepted protocol in generating insulin resistance and diabetes models. In the present study, we reported the multi-omics profiling of the liver and sera from both peripheral blood and hepatic portal vein blood from Macaca fascicularis that spontaneously developed Type-2 diabetes mellitus with a chow diet (sDM). The other two groups of the monkeys fed with chow diet and high-fat high-sugar (HFHS) diet, respectively, were included for comparison. Analyses of various omics datasets revealed the alterations of high consistency. Between the sDM and HFHS monkeys, both the similar and unique alterations in the lipid metabolism have been demonstrated from metabolomic, transcriptomic, and proteomic data repeatedly. The comparison of the proteome and transcriptome confirmed the involvement of fatty acid binding protein 4 (FABP4) in the diet-induced pathogenesis of diabetes in macaques. Furthermore, the commonly changed genes between spontaneous diabetes and HFHS diet-induced prediabetes suggested that the alterations in the intra- and extracellular structural proteins and cell migration in the liver might mediate the HFHS diet induction of diabetes mellitus.

Keywords: cynomolgus monkey (Macaca fascicularis); metabolomics; non-human primates; proteomics; spontaneous diabetes mellitus; transcriptomics.

FIGURE 1

The design of the experiments. (A) Experimental workflow. Three groups of Macaca fascicularis were sacrificed and their liver, PB, and HPVB were collected for metabolomic, transcriptomic, and proteomic profiling. (B–F) The comparison of urinary glucose (B), blood glucose (C), HbA1c (D), bodyweight (E), and insulin (F) between the groups. (G) and (H) The changes in blood glucose (G) and insulin (H) during IVGTT. (I) The representatives of the histochemical staining results of the macaque livers from the three groups.

FIGURE 2

The comparison among the metabolomics data. (A) Principal component analysis of the 1,082 metabolites identified from all 27 samples and 4 QC samples. (B) The metabolites of significantly different levels between the PB and HPVB samples of each group. (C–E) The metabolites of significantly different levels between the HFHS and NC groups (C), the sDM and NC groups (D), and the sDM and HFHS groups (E). (F) The enriched classes of molecules found from the metabolites altered in the HFHS and sDM groups. (G–I) The inter-group changes of acylcarnitines (Car), diacylglycerols (DG), and triacylglycerols (TG) in the liver (G), PB (H), and HPVB (I) samples.

FIGURE 3

The expression profile in the livers of three groups. (A–C) The volcano plots of the genes differentially expressed between the HFHS and NC groups (A), the sDM and NC groups (B), and the sDM and HFHS groups (C). (D) The altered transcripts between every two groups of the monkeys (E) The z-scored expression of the genes that significantly regulated commonly in both the sDM and HFHS macaques (upper panel) and only in the sDM group (lower panel). The genes that have known functions related to cell morphology and cell movement are underlined. (F) Fisher’s exact tested GO term enrichment (biological process) of the genes co-regulated in both sDM and HFHS groups.

FIGURE 4

Proteomic profiling in the liver of the three groups. (A) The histogram of sequence coverage of all identified proteins. (B) The numbers of significantly changed proteins between the groups. (C–E) KEGG pathway enrichment of the altered proteins between the NC and HFHS groups (C), between the sDM and HFHS groups (D), and between the sDM and NC groups (E). (F) The fold changes of the four differentially expressed proteins among all three groups. FABP4; Fatty acid binding protein 4; CNTFR, Ciliary neurotrophic factor receptor; HAL, Histidine ammonia-lyase; MVK, Mevalonate kinase. Mean ± SD are shown. *, p < 0.05; **, p < 0.01; ***, p < 0.001.

FIGURE 5

Joint analysis of the multiomics data. (A) The proportion of genes whose proteins and transcripts showed positive (> 0.5), negative (< −0.5), or no correlation (between −0.5 and 0.5). (B–D) the comparisons of log2 ratios of protein and transcript levels of differential genes for sDM versus HFHS (B), HFHS versus NC (C), and sDM versus NC (D). The black lines show the bisects y = x. (E) The pathway enrichment analysis using the metabolomics and proteomics data jointly. Plus symbol (+) annotates the enriched pathways of significance.