The Identification of LA-tumor associated macrophages in immune modulation via amyloid-beta precursor protein/CD74 signal pathway in gastric cancer: a predictive module and machine learning

Background Tumor-associated macrophages (TAMs) play a critical role in cancer immune microenvironment, modulating immune evasion. The prognostic role of TAMs gives insights into the immune landscape and therapeutic targets in gastric cancer (GC). Methods GC microenvironment was analyzed via single-cell and bulk RNA-seq data from public databases. TAM subtypes were then identified via dimensionality reduction and annotation under quality control. TAM differentiation and function were evaluated by pseudo-time analysis, cell communication, molecular docking, and key gene enrichment. A predictive model based on LA-TAM was established. Amyloid-β precursor protein (APP) expression level and its effect on macrophage programmed death-1 (PD-1) expression was validated in vitro . Results In GC microenvironment, epithelial cells and fibroblasts were downregulated, while B cells, CD8 + T cells and myeloid cells were enriched. Among TAM subtypes, LA-TAM exhibited the potential of differentiation, metabolic reprogramming, and high plasticity. When LA-TAM interacts with endothelial cells, APP/Collagen pathway was activated, in which PD-1 expression was up-regulated by APP/CD74 activation. The LA-TAM-based predictive model showed significant performance among multiple cohorts (C-index >0.5, HR = 1.63, p