Identification of Key Prognostic Alternative Splicing Events of Costimulatory Molecule-Related Genes in Colon Cancer.

Objective: This study aimed to explore the key alternative splicing events in costimulatory molecule-related genes in colon cancer and to determine their correlation with prognosis.

Methods: Gene expression RNA-sequencing data, clinical data, and SpliceSeq data of colon cancer were obtained from The Cancer Genome Atlas. Differentially expressed alternative splicing events in genes were identified, Followed by correlation analysis of genes corresponding to differentially expressed alternative splicing events with costimulatory molecule-related genes. Survival analysis was conducted using differentially expressed alternative splicing events in these genes and a prognostic model was constructed. Functional enrichment, proteinprotein interaction network, and splicing factor analyses were performed.

Results: In total, 6504 differentially expressed alternative splicing events in 3949 genes were identified between tumor and normal tissues. Correlation analysis revealed 3499 differentially expressed alternative splicing events in 2168 costimulatory molecule-related genes. Moreover, 328 differentially expressed alternative splicing events in 288 costimulatory molecule-related genes were associated with overall survival. The prognostic models constructed using these showed considerable power in predicting survival. The ubiquitin A-52 residue ribosomal protein fusion product 1 and ribosomal protein S9 were the hub nodes in the protein-protein interaction network. Furthermore, one splicing factor, splicing factor proline and glutamine-rich, was significantly associated with patient prognosis. Four splicing factor-alternative splicing pairs were obtained from four alternative splicing events in three genes: TBC1 domain family member 8 B, complement factor H, and mitochondrial fission 1.

Conclusion: The identified differentially expressed alternative splicing events of costimulatory molecule-related genes may be used to predict patient prognosis and immunotherapy responses in colon cancer.