Increasing studies have highlighted the importance of ferroptosis in colorectal cancer (CRC). However, how to use ferroptosis-related genes (FRGs) to predict the prognosis and guide the treatment of CRC remains unknown. To build a prognostic prediction model using the GEO and TCGA databases and explored a therapeutic strategy for CRC patients based on FRGs. A total of 60 FRGs were identified and three of them including ACACA, GSS, and NFS1 were associated with the prognosis of CRC. Using Lasso regression analysis, an FRGs signature was constructed and validated as an independent prognostic predictor. Then we developed a nomogram based on the FRGs signature and clinical prognostic factors to predict the prognosis of CRC patients, which was better than the traditional TNM staging system. Single-sample gene set enrichment analysis (ssGSEA) was further performed for the functional analysis and suggested that JAK-STAT signaling, Ras signaling pathway, MAPK signaling pathway, and PI3K-Akt signaling pathway were significantly enriched in CRC patients with higher FRGs risk score. Interestingly, CRC cells with higher FRGs risk score were more sensitive to RSL3. Knocking down GSS and NFS1 increased the FRGs risk score and the sensitivity of CRC cells to RSL3. For the clinic use, we screened 75 FDA-approved cancer drugs and found that Fludarabine phosphate could decrease the expression of GSS and NFS1 most. Fludarabine phosphate, in combination with RSL3, showed a strong synergistic effect on CRC cells. Together, this study identified a potent prognostic model and provided an alternative individualized treatment for CRC patients.
Keywords: Ferroptosis-related genes; colorectal cancer; fludarabine phosphate; rsl3; signature.