Abstract
In order to explore the molecular mechanism of acute rejection after liver transplantation (ARLT) in rats, we employed the GSE36798 data set in the Gene Expression Omnibust (GEO) database to construct a related ceRNA network. This dataset contained a total of 16 samples (8 graft samples and 8 plasma samples). Each kind of sample was divided into acute rejection (AR) groups and non-acute rejection (NR) groups, and each group had 4 replicates. First, we performed principal component analysis (PCA) with downloaded data to compare the difference between samples in a macroscopic way. Then, we used the "limma" R package to screen out differentially expressed miRNAs among different groups and used the "pheatmap" R package to perform bidirectional hierarchical clustering analysis for these differentially expressed miRNAs. The miRWalk database and the LncBase V.2 database were applied to predict downstream target genes and upstream-related lncRNAs, respectively. Meanwhile, the String database was used to predict the relationship between target genes, and the aforementioned results were processed for visualization by Cytoscape software. In addition, we exhibited the ultimate ceRNA network, including two lncRNAs, two miRNAs, and 77 mRNAs. Finally, we constructed a rat model of ARLT and applied graft specimens to relevant experimental verification. We found that the lncRNA Snhg1/rno-miR-139-5p axis might be involved in the regulation of ARLT in rats. In short, we demonstrated the differentially expressed miRNA profile, constructed a related ceRNA network, and screened out a possible regulatory axis. In view of the conservation of genes among species, this work was expected to provide a new strategy for the treatment and prevention of ARLT in the clinical setting.