Constructing prognostic models for colorectal cancer through dynamic profiling of tumor-associated neutrophils via single-cell and bulk RNA sequencing

Abstract Background: This study elucidates tumor-associated neutrophils (TANs) role in the colorectal cancer (CRC) microenvironment and develops a predictive risk model. Methods: Single-cell sequencing data from CRC patients in the Gene Expression Omnibus (GEO) database was utilized to identify subgroups of TANs and their marker genes. These subgroups' prognostic and diagnostic abilities were assessed using Kaplan-Meier survival analysis and receiver operating characteristic (ROC) curve analysis. Additionally, CIBERSORT and TIDE scores were employed to evaluate tumor immune cell infiltration and predict the effectiveness of immunotherapy. Pseudotime analysis and CellChat were used to explore the roles of TANs within the tumor microenvironment. Molecular docking of candidate genes with PD-L1 inhibitors was performed using PyMOL software. Candidate gene expression in CRC and adjacent tissues was confirmed by real-time quantitative PCR. Results: We identified 26 cell clusters and 226 TAN markers, establishing a six-gene prognostic model (CD36, SLC2A3, YBX3, ZFB36L1, TIMP1, ASAH1). The immune infiltration differences between risk groups were notable, suggesting better immunotherapy outcomes for the low-risk group. Cell communication analysis showed TANs affecting monocytes, epithelial, and endothelial cells via ANNEXIN, MIF, and VISFATIN pathways. Molecular docking revealed strong PD-L1 inhibitor affinity to SLC2A3. Conclusion: Our study highlights TANs' intricate interactions within the CRC microenvironment and the effectiveness of a TAN-based prognostic gene model in predicting immunotherapy responses, offering new directions for targeted CRC treatments and overcoming immune resistance.