RUNX3+ basal-like malignant cells: architects of an immunosuppressive microenvironment in bladder cancer: a comprehensive bioinformatics analysis of single-cell and spatial transcriptomics

Background: Current single-cell and spatial transcriptomic studies on bladder cancer tissue samples are limited, highlighting the need for advanced molecular and single-cell approaches to enhance our understanding of these complexities and identify reliable prognostic indicators. Materials and methods: We integrated and analyzed five publicly available single-cell RNA-seq datasets, totaling over 530 000 cells from 62 bladder cancer and normal samples, conducting comprehensive analyses that were validated using single-nucleus and spatial transcriptomic datasets. Additionally, spatial transcriptomics findings were confirmed through RNA fluorescence in situ hybridization. Results: Our integrated single-cell analysis uncovered stage-dependent heterogeneity in bladder cancer, with malignant epithelial cells exhibiting dynamic trajectories and enrichment in the PI3K–AKT–mTOR pathway. A distinct subcluster of malignant epithelial cells characterized by the highest basal score (C8) was identified; this subpopulation appears to be specifically regulated by RUNX3 , is associated with poorer prognosis, and predominantly interacts with exhausted CD8 T cells, IL7R + naïve T cells, and FCN1 + monocytes via the MIF – CD74 signaling pathway. Elevated pathway activity scores for both the PI3K–AKT–mTOR and MIF – CD74 pathways were observed in single-nuclei and spatial tumor tissues, with notable colocalization of key markers mainly within bladder cancer regions. Conclusion: Our study demonstrates that RUNX3 + basal-like malignant epithelial cells, along with other immunosuppressive cell subtypes, their interactions, and pathway signatures, hold promise as prognostic biomarkers. These insights may facilitate the development of personalized therapeutic strategies for bladder cancer.