Global miRNA expression reveals novel nuclear and mitochondrial interactions in Type 1 diabetes mellitus

Abstract

Background: Considering the potential role of miRNAs as biomarkers and their interaction with both nuclear and mitochondrial genes, we investigated the miRNA expression profile in type 1 diabetes (T1DM) patients, including the pathways in which they are involved considering both nuclear and mitochondrial functions.

Methods: We analyzed samples of T1DM patients and control individuals (normal glucose tolerance) by high throughput miRNA sequencing (miRNome). Next, five miRNAs - hsa-miR-26b-5p, hsa-let-7i-5p, hsa-miR-143-3p, hsa-miR-501-3p and hsa-miR-100-5p - were validated by RT-qPCR. The identification of target genes was extracted from miRTarBase and mitoXplorer database. We also performed receiver operating characteristic (ROC) curves and miRNAs that had an AUC > 0.85 were considered potential biomarkers.

Results: Overall, 41 miRNAs were differentially expressed in T1DM patients compared to control. Hsa-miR-21-5p had the highest number of predicted target genes and was associated with several pathways, including insulin signaling and apoptosis. 34.1% (14/41) of the differentially expressed miRNAs also targeted mitochondrial genes, and 80.5% (33/41) of them targeted nuclear genes involved in the mitochondrial metabolism. All five validated miRNAs were upregulated in T1DM. Among them, hsa-miR-26b-5p showed AUC>0.85, being suggested as potential biomarker to T1DM.

Conclusion: Our results demonstrated 41 DE miRNAs that had a great accuracy in discriminating T1DM and control group. Furthermore, we demonstrate the influence of these miRNAs on numerous metabolic pathways, including mitochondrial metabolism. Hsa-miR-26b-5p and hsa-miR-21-5p were highlighted in our results, possibly acting on nuclear and mitochondrial dysfunction and, subsequently, T1DM dysregulation.

Keywords: miRNAs; miRnome; mitochondrial target; nuclear target; type 1 diabetes.

Figure 1

Experimental workflow of the analyses in the study.

Figure 2

DE miRNAs in the T1DM patients in comparison to controls. Blue dots are considered DE miRNAs under the conditions of adjusted values of p < 0.05 and |log2 fold change| ≥ 1.5. Red dots are non-DE miRNAs. Note that miRNAs on right of figure are up-regulated, and on the left are down-regulated.

Figure 3

Heatmap with hierarchical clustering analysis of DE miRNAs in T1DM. Blue color in top bar represents control individuals and red colors represents T1DM patients. In the heatmap, dark-red color corresponds to high miRNA expression, and dark blue corresponds to low expression.

Figure 4

Quantitative of target genes of DE miRNAs in T1DM. (A) Number of target genes per DE miRNA. (B) Number of interacting miRNAs for the ten most frequently found genes.

Figure 5

Relative expression of five RT-qPCR validated miRNAs. All of them showed to be upregulated in T1DM group.

Figure 6

Analysis of biomarker sensitivity of five validated miRNAs in T1DM.

Figure 7

Target genes of validated miRNAs (blue) with strong evidence in T1DM. Those regulated by at least two of the miRNAs are pictured in green and those solely regulated by one of the validated miRNAs are picture in yellow.

Figure 8

Analyses of the 14 DE miRNAs that target mitochondrially-encoded genes. (A) Network of these miRNAs (in blue or green, the latter being those validated in the current study) and the mitochondrial genes they target (in yellow); (B) GTEx data of mitochondrially-encoded gene expression in the pancreas and whole blood; (C) KEGG pathways of mitochondrial functions and some multifactorial conditions; (D) Reactome pathways of mitochondrial mechanisms.