Analysis and identification of potential key genes in hepatic ischemia-reperfusion injury.

Background: Hepatic ischemia-reperfusion injury (HIRI) is an unavoidable surgical complication after liver transplantation, but current HIRI treatments cannot achieve satisfactory clinical outcomes. Thus, safer and more effective prevention and treatment methods need to be explored.

Methods: Transcriptome messenger ribonucleic acid (mRNA) and long non-coding RNA (lncRNA) sequencing data were obtained from male Sprague-Dawley rats, and these data were used to identify the differentially expressed genes (DEGs) and differentially expressed lncRNAs (DE-lncRNAs) between the HIRI and control samples. A protein-protein interaction (PPI) network was also constructed for the DE-mRNAs to identify candidate genes, and the receiver operating characteristic curves of the 21 candidate genes were plotted to evaluate the diagnostic value of the candidate genes for HIRI. A random forest (RF) model, support vector machine model and generalized linear model were constructed based on the candidate genes. A gene set enrichment analysis (GSEA) of the key genes was conducted to determine the enriched pathways in the high expression groups. The miRWalk and miRanda database were used to constructed the lncRNA-miRNA-mRNA network. Finally, the expressions of the key genes were verified by quantitative real-time polymerase chain reaction (qRT-PCR).

Results: A total of 256 DEGs and 67 DE-lncRNAs were identified in the HIRI and control samples. To explore the interactions between the DE-mRNAs, a PPI network of 130 DEGs was constructed. Further, 21 genes were selected as the candidate genes. Subsequently, 6 genes [i.e., Keratin-14 (), Uroplakin 3B (), Keratin 7 (), Cadherin 3 (), mesothelin (), and Glypican 3 ()] in the RF model were defined as the key genes. The GSEA results indicated that these key genes were enriched in the terms of extracellular structure organization, and extracellular matrix organization. Moreover, a lncRNA-miRNA-mRNA network was constructed with 4 lncRNAs, 5 mRNAs, and 11 miRNAs. Finally, the results indicated that the expression of , and were more highly expressed in the control samples than the HIRI samples.

Conclusions: A total of 6 key genes (i.e., and ) were identified. Our findings provide novel ideas for the diagnosis and treatment of HIRI.