Identification of hub genes predicting sensitivity to neoadjuvant chemoradiation in locally advanced rectal cancer

Preoperative neoadjuvant chemoradiation (NACR) benefits disease control in most locally advanced rectal cancer (LARC) patients. However, effective biomarkers predicting response to NACR are still not accessible. This study aimed to find potential biomarkers to assess therapy response and susceptibility to LARC.

Differentially expressed genes (DEGs) between NACR-sensitive and resistant patients were screened using GEO database. STRING and Cytoscape were utilized to construct PPI networks and identify hub genes. Based on CIBERSORT, TCGA, GTEx, GSEA and ROC curves, the connections between hub genes and specific signaling pathways, immune cell infiltration, prognosis value and miRNA-transcription factor (TF)-target network were investigated. Human Protein Atlas (HPA) database was used to visualize hub gene expression in clinical samples.

We identified 2619 up- and 2466 down-regulated genes between NACR-sensitive and resistant patients. The up-regulated DEGs were searched for highly expressed genes in the NACR-resistant, TCGA and GTEx-related datasets compared to the NACR-sensitive group, yielding six hub genes (RRM2, HNRNPL, EZH2, METTL1, NHP2L1 and ASF1B). ROC curves demonstrated the predictive utility of the six genes in NACR sensitivity. Immune infiltration research revealed no significant relationship between NACR sensitivity and immune cell infiltration extent. The miRNA-TF-target network of hub genes was established. Finally, HPA database results showed that six genes were expressed at variable levels in rectal cancer patients.

This study identified six hub genes (RRM2, HNRNPL, EZH2, METTL1, NHP2L1 and ASF1B) up-regulated in LARC and valuable for predicting patient susceptibility and response to NACR.