Spatial profiling of gastric cancer reveals distinct transcriptional meta-programs driving immunosuppressive microenvironment

Abstract Objective To systematically characterize tumor microenvironmental features and spatial cellular interactomes associated with distinct transcriptional meta-program (MP) in gastric cancer, and explore their clinical implications. Methods We integrated 2 gastric cancer single-cell RNA sequencing (scRNA-seq) datasets to identify malignant cell-derived transcriptional MP. The prognostic value was validated in 3 independent bulk transcriptomic cohorts: The Cancer Genome Atlas (TCGA), GSE15459, and GSE66229. Using scRNA-seq, we dissected MP-specific shifts in microenvironmental cell composition and inferred dominant ligand-receptor interactions. Spatial transcriptomics confirmed the co-localization of implicated cell subsets and signaling pairs. Survival analyses were used to access the associations between key cellular signatures/interaction signatures and patient outcomes. Results Four MP from MP1 to MP4 were identified in malignant cells. High expression of MP2 and MP4 significantly correlated with worse overall survival (P