Identification of a lactylation-related gene signature in microsatellite stable gastric cancer based on bulk and single-cell RNA-seq.

Microsatellite stable (MSS) gastric cancer (GC) responds poorly to immunotherapy and exhibits heterogeneous outcomes. Histone lactylation plays a critical role in cancer progression. However, the prognostic and therapeutic potential of a lactylation-related gene signature (LRGS) in MSS GC remains largely unexplored. Data of MSS GC patients were obtained from The Cancer Genome Atlas and Gene Expression Omnibus databases. Consensus clustering based on lactylation-related gene expression profiles was performed to stratify patients. We further constructed the LRGS using machine learning algorithms. We also assessed its correlations with clinicopathological parameters, the tumor microenvironment, and chemosensitivity. The Tumor Immune Dysfunction and Exclusion (TIDE) algorithm was employed to predict potential responses to immunotherapy. Single-cell analysis, cell-cell communication analysis, in silico knockout, and immunohistochemical validation were integrated to explore the functional roles of key genes. Consensus clustering uncovered two clusters with significantly different overall survival outcomes. A nine-gene LRGS was established and effectively stratified patients into high- and low-risk groups. The high-risk group displayed an immunosuppressive microenvironment, reduced chemosensitivity, and higher TIDE scores. Single-cell analysis revealed that LRGS scores were highest in cancer-associated fibroblasts (CAFs), with APOD and SERPINE1 highly expressed in PDGFRA+ CAFs and CA9+ CAFs, respectively. Cell-cell communication analysis showed that these two CAF subtypes exhibited distinct signaling patterns via the COLLAGEN and MIF pathways. In silico knockout further validated that APOD and SERPINE1 are essential for matrix remodeling and hypoxia adaptation in their respective CAF subtypes. Immunohistochemical analysis confirmed elevated protein levels of APOD and SERPINE1 in MSS GC. This study developed and validated a robust LRGS for MSS GC. This signature facilitates accurate prognosis prediction and shows potential to predict responses to both chemotherapy and immunotherapy.