Abstract
Liver metastasis is the most common site of metastasis in colorectal cancer, and the prognosis of colorectal cancer patients with liver metastasis is extremely poor. Revealing the key genes of CLM and implementing targeted interventions is of great significance for colorectal cancer patients. By using the weighted gene co-expression network analysis (WGCNA) algorithm, key gene modules related to metastasis in colorectal cancer were identified. Subsequently, immune-regulating and prognostic-influencing key gene sets were identified from these modules to construct a prognostic model related to colorectal cancer metastasis. Genetic background differences underlying this model were analyzed using colorectal cancer methylation and mutation data, followed by Gene Ontology (GO) analysis and Gene Set Enrichment Analysis (GSEA) analysis of the relevant biological processes associated with the model. The value of predicting tumor drug response through the model was assessed using drug half maximal inhibitory concentration (IC50) data from colorectal cancer cell lines. Subsequently, utilizing single-cell sequencing data about liver metastasis, the colorectal cancer immune microenvironment reflected in the predictive model was analyzed, and a key gene set of the model was identified. Lastly, experimental validation was conducted to investigate the regulatory effects of photodynamic therapy (PDT) on the key genes of the model, and the cytotoxic effect of PDT on colorectal cancer was confirmed. An immune-related gene prognostic model regulating CLM was constructed, consisting of HSPA1A, ULBP2, RBP7, OXT, SLC11A1, INHBB, and ICOS. This model can predict the clinical response of colorectal cancer patients to Oxaliplatin, Cisplatin, Irinotecan, and 5-Fluorouracil. Single-cell sequencing results demonstrate that the model is associated with an immunosuppressive microenvironment in CLM. The higher the model's riskscore, the weaker the MHC-I, MHC-II, and various tumor immune signaling pathway networks in the colorectal cancer microenvironment. Causal analysis reveals that SLC11A1, ICOS, and HSPA1A play key roles in this model. PDT can kill colorectal cancer cells, inhibit colorectal cancer cell metastasis, significantly influence the expression of genes such as SLC11A1, ICOS, and HSPA1A in these processes, and suppress the infiltration of macrophages in the colorectal microenvironment, inhibiting the immune escape process of PD-1/PD-L1. A prognostic model based on immunity regulated by PDT has been established for assessing the prognosis of CLM patients, as well as clinical responses to chemotherapy drugs and immunotherapy.