Construction and validation of a prognostic model for colorectal cancer based on bioinformatics and basement...

Basement membrane-related genes (BMRGs) play a crucial function in the progression of various malignancies. This study aimed to explore the relationship between BMRGs and colon cancer.

Data from Colorectal Cancer (CRC) were extracted from the Xena and GEO databases, with 224 BMRGs identified from CRC sequencing data. A single-cell data set (GSE132465) was analyzed to compute BM scores, followed by correlation analysis. Differentially expressed genes (DEGs) were derived from the TCGA-COAD colon cancer data set. Prognosis-associated characteristic genes were identified using univariate Cox, Lasso-Cox, and random survival forest (RSF) methods. These genes were incorporated into a multivariate Cox regression model to calculate risk scores. A clinical predictive model was constructed and validated using the GSE103479 data set. Key gene expression differences between tumor and normal tissues were validated through immunohistochemical (IHC) staining.

Three BMRGs have been identified as prognostic markers for colorectal cancer: GPC2, UNC5D, and SPOCK3. These genes exhibited higher expression levels in cancer tissues compared to adjacent normal tissues. Risk scores for each sample were calculated, categorizing them into high-risk and low-risk groups. The model was validated by constructing an optimal predictive model. Analysis of immune infiltration demonstrated a notable elevation in CD4 + T cell populations and a substantial reduction in CD8 + T cell populations within high-risk groups when contrasted with low-risk groups. Immune function analysis indicated reduced scores for APC co-stimulation, CCR, MHC class I, Th1 cells, and Th2 cells in high-risk groups. IHC confirmed elevated expression of GPC2, UNC5D, and SPOCK3 in colorectal cancer tissues, consistent with bioinformatics findings.

The expression levels of GPC2, UNC5D, and SPOCK3 serve as prognostic biomarkers for colorectal cancer, as demonstrated by bioinformatics analysis and IHC. These findings provide valuable insights for prognosis prediction and the development of novel therapeutic targets for colorectal cancer.