Abstract
BACKGROUND: Gastric cancer (GC) is among the most lethal malignancies worldwide. Due to the substantial heterogeneity of GC, more accurate molecular typing systems are desperately required to enhance the prognosis of GC patients. METHODS: The major immune cell subclusters in GC were identified by a single-cell RNA sequencing (scRNA-seq) dataset. High-dimensional weighted gene coexpression network analysis (hdWGCNA) and multiple bioinformatics methods were utilized to classify the molecular subtypes of GC and further investigate the differences among the subtypes. Based on the module genes and differentially expressed genes (DEGs), random survival forest analysis was applied to identify the key prognostic genes for GC, and the roles and functional mechanisms of the key genes in GC were explored by clinical samples and cellular experiments. RESULTS: Two distinct GC molecular subtypes (C1 and C2) associated with neutrophils were identified, with C1 associated with better prognosis. Compared with C2 subtype, C1 subtype has significant differences in immune infiltration, immune checkpoint expression, signaling pathway regulation, tumor mutation burden, and immunotherapy and chemotherapeutic drug sensitivity. Three new key genes (VIM, RBMS1 and RGS2) were revealed to be highly correlated with the prognosis of GC patients. In addition, the expression and cellular functions of key genes RBMS1 and RGS2 in gastric carcinogenesis were verified. CONCLUSION: We identified two neutrophil-related molecular GC subtypes with different prognostic outcomes and clinical significance. VIM, RBMS1 and RGS2 were identified as potential prognostic markers and therapeutic targets for GC. These findings provide a new perspective for the molecular typing and personalized treatment of GC.