Abstract
BACKGROUND: Colorectal cancer (CRC) is the third most common cancer globally, and the identification of novel biomarkers remains an urgent priority in clinical practice and research. Although the Golgi apparatus plays a crucial role in tumorigenesis and cancer progression, its role in CRC remains unclear. METHODS: A comprehensive bioinformatics analysis of the transcriptome was conducted using the TCGA and GSE87211 datasets. A reliable risk signature model based on Golgi apparatus - related genes (GARGs) was constructed. Subsequently, we performed an in - depth single - cell RNA sequencing analysis. Finally, based on the consistent evidence across all analytical layers, GDI1 was selected for subsequent functional validation in cellular assays. RESULTS: By integrating univariate Cox regression and LASSO analysis, we identified a five-GARG signature to construct a prognostic model. This model stratified patients into high- and low-risk groups with significantly different overall survival. The high-risk group was associated with elevated stromal and ESTIMATE scores, along with increased infiltration of monocytes and M0 macrophages, suggesting a potentially more immunosuppressive and tumor-promoting microenvironment. Bioinformatic predictions indicated that the high-risk group might show enhanced sensitivity to Dasatinib, Sapitinib, and SB216763; this was preliminarily supported by our subsequent in vitro drug sensitivity assays. Cell-cell communication analysis suggested that macrophages might primarily influence the microenvironment through the tumor necrosis factor (TNF) signaling pathway. Critically, functional experiments demonstrated that silencing GDI1 significantly inhibited cell proliferation and migration and induced apoptosis, supporting its potential oncogenic role in COAD. CONCLUSION: Collectively, our work links GARG expression to CRC prognosis and immune features, and functionally implicates GDI1 in tumor cell aggressiveness, supporting its further study as a potential therapeutic target.