Background: The incidence of colon cancer is at the forefront of all cancers and poses great concerns for humans. Colon cancer is becoming more prevalent in young adults and therefore it is vital to find specific molecular markers for prognosis. This study aims to explore multidimensional prognostic targets of colon cancer from the perspective of the ceRNET and immune- infiltrating tumor cells.
Results: The lncRNA, miRNA and mRNA profiles and clinical data were downloaded from TCGA-COAD and 206 differential expression lncRNAs ,352 differential expression miRNAs and 2995 differential expression mRNAs were screened out. Interactions between miRNA–mRNA pairs and lncRNA–miRNA pairs were obtained in differentially expressed molecules through the Starbase database, thus constructing ceRNET. The relative content of 22 kinds of immune cells in COAD was estimated by “Cell type identification by estimating relative subsets of RNA transcripts (CIBERSORT)” algorithm. To predict the prognosis of colon cancer patients based on meaningful ceRNET or immune cells, two nomograms were constructed respectively. NRG1 had significant prognostic value(p=0.008) and was found to be highly correlated with Macrophages M1 (p = 6.5e−06) after co-expression analysis. In addition, there were two pairs of biomarkers that could affect the prognosis of colon cancer: Dendritic cells resting and UST (p =0.0012) and B cells naive and LINC00894 (p = 0.017). Finally, we found that MALAT1- NRG1 - has-miR-200c-3p and Macrophages M1 axis may have important prognostic significance for colon cancer.
Conclusions: In this research,we constructed ceRNET according to differential lncRNAs, miRNAs and mRNAs in the colon cancer and found ceRNAs and tumor-infiltrating immune cells valuable for colon cancer prognosis.