Objective:

To delineate the biological and clinical consequences of APOBEC activation in colorectal cancer (CRC) and its impact on the tumor microenvironment (TME).

Approach:

• Data Integration: Integrated whole-exome sequencing, bulk and single-cell transcriptomics from multi-cohort datasets, supported by functional assays and in vivo xenograft models.
• Machine Learning: Utilized a machine-learning framework to derive an APOBEC activation–associated transcriptional signature (AAS).

Key Findings:

• AAS defined a colorectal cancer subtype with enriched TCW mutagenesis and significantly worse survival.
• High AAS tumors showed increased fibroblast and endothelial signatures and reduced cytotoxic immune infiltration.
• Single-cell analyses revealed high AAS tumors enriched for endothelial states with arterial and pro-angiogenic features.
• Conditioned-medium experiments suggested a link between APOBEC3B activation and endothelial transcriptional remodeling.
• PCDH9 identified as a candidate APOBEC-associated target linked to Hippo, Wnt/β-catenin, and TGF-β signaling alterations.

Interpretation:

Limitations:

• Mutation-specific causality remains to be experimentally validated.

Conclusion: