Integrating RNA-seq and scRNA-seq to explore the prognostic features and immune landscape of exosome-related genes in breast cancer metastasis

This study aims to explore the role of exosome-related genes in breast cancer (BRCA) metastasis by integrating RNA-seq and single-cell RNA-seq (scRNA-seq) data from BRCA samples and to develop a reliable prognostic model. Initially, a comprehensive analysis was conducted on exosome-related genes from the BRCA cohort in The Cancer Genome Atlas (TCGA) database. Three prognostic genes (JUP, CAPZA1 and ARVCF) were identified through univariate Cox regression and Lasso-Cox regression analyses, and a metastasis-related risk score model was established based on these genes. Immune cell infiltration, immune escape and drug sensitivity disparities between high- and low-risk groups were assessed using CIBERSORT and single-sample gene set enrichment analysis (ssGSEA) methods. High- and low-risk cell populations were discerned based on the expression of prognostic genes in BRCA scRNA-seq data. M0 and M1 macrophages significantly promote the metastasis of breast cancer (BRCA). The developed prognostic model demonstrates good predictive performance for patient survival at 1, 3 and 5 years, with AUC values of 0.654, 0.602 and 0.635, respectively. Compared to the low-risk group, the high-risk group exhibits increased immune cell infiltration and higher levels of immune evasion. scRNA-seq data reveal that high-risk cells have significantly higher risk scores and exhibit notable differences in signalling pathways and intercellular communication patterns. This study presents a novel risk score model based on exosome-related genes, validated by comprehensive analyses including differential expression, survival analysis and external dataset validation. The model's clinical significance is reinforced through its ability to stratify patients into high- and low-risk groups with distinct survival outcomes and immune landscape characteristics. The integration of RNA-seq and scRNA-seq data highlights the predictive accuracy of the model and underscores its potential for identifying novel therapeutic targets and improving patient prognosis.