LncRNA/miRNA/mRNA ceRNA network analysis in spinal cord injury rat with physical exercise therapy

Abstract

Noncoding RNAs have been implicated in the pathophysiology of spinal cord injury (SCI), including cell death, glial scar formation, axonal collapse and demyelination, and inflammation. The evidence suggests that exercise therapy is just as effective as medical treatment in SCI. However, studies of competing endogenous RNA (ceRNA)-mediated regulation mechanisms in the therapy of SCI with exercise are rare. The focus of this research was to investigate the effect of exercise therapy on the expression levels of long noncoding RNA (lncRNA), microRNA (miRNA), and mRNA in rats with SCI. The RNA-seq technology has been used to examine the differentially expressed circRNAs (DECs), lncRNAs (DELs), miRNAs (DEMs), and genes (DEGs) between SCI and exercise therapy rats. The ceRNA network was established using interactions between miRNAs and mRNAs, as well as between miRNAs and lncRNAs/circRNAs. The Database for Annotation, Visualization, and Integrated Discovery was used to anticipate the underlying functions of mRNAs. Our current study identified 76 DELs, 33 DEMs, and 30 DEGs between groups of SCI rats and exercise therapy rats. Subsequently, these newly discovered ceRNA interaction axes could be important targets for the exercise treatment of SCI.

Keywords: Competing endogenous RNA; Long noncoding RNA; Nucleus pulposus; Spinal cord injury; microRNA.

Figure 1. The inhibitory effect of exercise therapy on inflammatory response after SCI.

(A) The experimental procedures of the dual rat spinal lesion model and the time course in all rats are shown. TT, treadmill training, treadmill training were performed on rat receiving SCI after 7 days. (B–E) RT-PCR analyses (n = 3) of BDNF (P = 0.0476), TNF-α (P = 0.0181), IL-1β (P = 0.0263), IL-6 (P = 0.0265) in the tissue of lesion region were performed two weeks after treadmill training. Data are expressed as mean ± SEM. An asterisk (*) indicates P < 0.05 versus the SCI treatment groups by one-way ANOVA.

Figure 2. Hierarchical clustering and heat map analysis of differentially expressed, Differentially expressed genes (DEGs) determined by RNA-seq.

(A) The workflow of RNA-seq. (B) lncRNA, (C) miRNA and (D) mRNA in spinal cord injury. N = 3 rats for the SCI groups (SCI) versus SCI plus TMT groups (TMT).

Figure 3. Enriched GO functional categories.

GO terms are visualized in semantic similarity-based scatterplots. Bubble color indicates the different categories. Bubble size shows how much the GO term is represented in the GO database (right), Selected top 20 categories are shown on the left.

Figure 4. Enriched GO functional categories.

GO enrichment analysis shows the numbers of genes in different GO term.

Figure 5. KEGG pathway enrichment and reactome pathway enrichment analysis of DEGs.

(A) Circle plot demonstrating the KEGG pathway enriched by the DEGs. Inner plot color corresponds to the rich factor. The second ring displays up and down genes. The third ring displays Q-values. The outer ring displays GO terms. (B) Statistically significant pathways listed and their colors are shown by P-value. (C) Circle plot demonstrating the reactome pathway enriched by the DEGs. (D) Top 20 statistically significant pathways listed and their colors are shown by P-value.

Figure 6. Competing endogenous RNA interaction network of lncRNA-miRNA-mRNA.

Blue represents upregulated expression, whereas brown represents downregulated expression. Square nodes represent lncRNAs, triangular nodes represent miRNAs and oval nodes represent mRNAs. Hub genes are indicated by red boxes. FC, fold change; lncRNA, long noncoding RNA; miRNA/miR, microRNA.