Objective
This study aimed to assess the long noncoding RNA (lncRNA)‐microRNA (miRNA)‐messenger RNA (mRNA) regulatory network in clear cell renal cell carcinoma (ccRCC) by gene expression analyses.
Materials and Methods
LncRNA, miRNA, and mRNA expression profiles in ccRCC were obtained from The Cancer Genome Atlas. Differentially expressed lncRNAs, mRNAs (cut‐off: |log 2 [fold change, FC])| > 2.0 and adjusted P < 0.01) and miRNAs (|log 2FC| > 1.5 and adjusted P < 0.01) were unveiled using R. Cox regression analysis was performed to identify prognostic factors of ccRCC related to overall survival (OS). A protein‐protein interaction (PPI) network was constructed for differentially expressed mRNAs (DEmRNAs) by Search Tool for the Retrieval of Interacting Genes (STRING). Key hub genes were screened from top 300 DEmRNAs. LncRNA‐miRNA and miRNA‐mRNA regulatory network were constructed and combined into the competing endogenous RNA regulatory network. Gene ontology biological terms were screened by STRING; Kyoto Encyclopedia of Genes and Genomes pathways were identified using the “clusterProfiler” package in R.
Results
A total of 2331, 1517, and 83 DEmRNAs, lncRNAs, and miRNAs were identified, respectively. Eleven lncRNAs (AC016773.1, HOTTIP, LINC00460, NALCN‐AS1, PVT1, TRIM36‐IT1, WT1‐AS, COL18A1‐AS1, LINC00443, LINC00472, and TCL6), three miRNAs (hsa‐mir‐21, hsa‐mir‐144, and hsa‐mir‐155), and three mRNAs (COL4A4, NOD2, and GOLGA8B) were associated with OS. Specifically, four lncRNAs (PVT1, LINC00472, TCL6, and WT1‐AS1) and one mRNA (Collagen Type IV Alpha 4 Chain) were verified as independent prognostic factors by Gene Expression Profiling Interactive Analysis. Eleven key hub genes were obtained by PPI analysis. “Cell adhesion molecules (CAMs),” “chemical carcinogenesis,” and “cytokine‐cytokine receptor interaction” were significantly enriched in the network.
Conclusion
The findings clarify the pathogenesis of ccRCC and might provide potential therapeutic targets.