Comprehensive analyses of the microRNA–messenger RNA–transcription factor regulatory network in mouse and human renal fibrosis

Abstract

Objective: The aim of this study was to construct a microRNA (miRNA)–messenger RNA (mRNA)–transcription factor (TF) regulatory network and explore underlying molecular mechanisms, effective biomarkers, and drugs in renal fibrosis (RF).Methods: A total of six datasets were downloaded from Gene Expression Omnibus. “Limma” and “DESeq2” packages in R software and GEO2R were applied to identify the differentially expressed miRNAs and mRNAs (DEmiRNAs and DEmRNAs, respectively). The determination and verification of DEmiRNAs and DEmRNAs were performed through the integrated analysis of datasets from five mouse 7 days of unilateral ureteral obstruction datasets and one human chronic kidney disease dataset and the Human Protein Atlas (http://www.proteinatlas.org). Target mRNAs of DEmiRNAs and TFs were predicted by prediction databases and the iRegulon plugin in Cytoscape, respectively. A protein–protein interaction network was constructed using STRING, Cytoscape v3.9.1, and CytoNCA. Functional enrichment analysis was performed by DIANA-miRPath v3.0 and R package “clusterProfiler.” A miRNA–mRNA–TF network was established using Cytoscape. Receiver operating characteristic (ROC) curve analysis was used to examine the diagnostic value of the key hub genes. Finally, the Comparative Toxicogenomics Database and Drug-Gene Interaction database were applied to identify potential drugs.Results: Here, 4 DEmiRNAs and 11 hub genes were determined and confirmed in five mouse datasets, of which Bckdha and Vegfa were further verified in one human dataset and HPA, respectively. Moreover, Bckdha and Vegfa were also predicted by miR-125a-3p and miR-199a-5p, respectively, in humans as in mice. The sequences of miR-125a-3p and miR-199a-5p in mice were identical to those in humans. A total of 6 TFs were predicted to regulate Bckdha and Vegfa across mice and humans; then, a miRNA–mRNA–TF regulatory network was built. Subsequently, ROC curve analysis showed that the area under the curve value of Vegfa was 0.825 (p = 0.002). Finally, enalapril was identified to target Vegfa for RF therapy.Conclusion: Pax2, Pax5, Sp1, Sp2, Sp3, and Sp4 together with Bckdha-dependent miR-125a-3p/Vegfa-dependent miR-199a-5p formed a co-regulatory network enabling Bckdha/Vegfa to be tightly controlled in the underlying pathogenesis of RF across mice and humans. Vegfa could act as a potential novel diagnostic marker and might be targeted by enalapril for RF therapy.

Pioglitazone Protects Tubular Epithelial Cells during Kidney Fibrosis by Attenuating miRNA Dysregulation and Autophagy Dysfunction Induced by TGF-β

“…For instance, miR-199a-5p was reported to inhibit endoplasmic reticulum stress in renal ischemia-reperfusion (I/R) injury [45] and to downregulate PPARγ, exacerbating renal interstitial fibrosis in rats with hyperuricemia [28]. Elevated miR-199a-5p has been reported in mouse and human CKD as well, underscoring its potential importance [30]. On the other hand, PPARγ might regulate the expression of miR-199a-5p by influencing the processing or stability of pre-miR-199a at the posttranscriptional level [46].…”

“…Both miR-199a-3p and miR199a-5p are potential regulatory factors of TGF-β-mediated pathological renal processes [28][29][30]. TGF-β administration markedly increased the expressions of both miR-199a-3p and miR-199a-5p in primary mouse tubular epithelial cells, which were inhibited by pioglitazone (Figure 5C,E).…”

“…The miR-199a forms a pre-miR that undergoes RNA processing to yield miR-199a-5p and miR-199a-3p. Elevated renal miR-199a-3p can trigger STAT3 activation [29], while increased miR-199a-5p was reported in both murine and human CKD [30]. Nevertheless, the possible effect of PPARγ agonist treatment on the renal expression of the miR-130 or miR-199a family has not been investigated yet.…”

LncRNA‐associated competing endogenous RNA network analysis uncovered key lncRNAs involved in temozolomide resistance and tumor recurrence of glioblastoma

“…After obtaining significantly positively correlated lncRNA‐mRNA pairs, we adopted a comprehensive computational approach to identify significant ceRNA candidates 31 . Several methods have been used to predict the ceRNA network, such as the hypergeometric test, which predicts the ceRNAs based on the number of common miRNAs binding to mRNA or lncRNA 32–34 . These studies considered only a single parameter.…”

“…31 Several methods have been used to predict the ceRNA network, such as the hypergeometric test, which predicts the ceRNAs based on the number of common miRNAs binding to mRNA or lncRNA. [32][33][34] These studies considered only a single parameter. We used several attributes, such as expression levels of F I G U R E 1 Workflow of the study to obtain long non-coding RNA (lncRNA)-associated competing endogenous RNAs (ceRNAs) possibly involved in temozolomide (TMZ) resistance and tumor recurrence.…”