Colorectal Cancer Core Gene Screening,Immunocorrelation Analysis And Clinical Prediction Model Construction

Background:Colorectal Cancer(CRC)accounts for about 10% of all diagnosed cancers and cancer-related deaths in the world every year.The epidemiological survey indicates that the number of CRC cases worldwide is expected to increase to 2.5 million by2035.The occurrence and development of CRC cannot be achieved without the accumulation of oncogene/oncosuppressor gene mutations.In recent years,the number of CRC-related genes reported at home and abroad has gradually increased,and the understanding of CRC has gradually deepened.As one of the malignant tumors that often develop symptoms in the late stage,it is particularly important to improve the early detection of CRC.Screening at the molecular level can help people identify early CRC,so as to reduce morbidity and mortality.Objective:Through bioinformatics analysis of colorectal cancer gene chip to screen are closely associated with colorectal cancer occurrence and development of core genes,further analysis the core gene and the correlation of immune cells and immune checkpoints.We attempted to use core genes to construct risk scores and clinical prediction models to evaluate the prognosis of patients with colorectal cancer.The expression of TIMP1 in colorectal cancer was verified by protein imprinting and immunohistochemistry.Methods:1.Through the screening of colorectal cancer gene expression chips in GEO database,the GSE106582 and GSE32323 datasets of GPL570 platform were determined,of which 77 cancerous tissues and 117 paracancerous tissues were in GSE106582 and 17 cancerous tissues and 17 paracancerous tissues were identified in GSE32323.Download the expression matrix,clinical information and survival information of colorectal cancer patients in the TCGA database through the XENA website.2.The downloaded datasets were processed separately by Rstudio,and the expression matrix and clinical information were extracted for further grouping and probe annotation.Gglpot2 and limma packages were used to analyze and visualize the different expression genes(DEG)of the two sets of datasets,and further intersect the two sets of different genes.The DEGs filter is:(1)|log FC| > 1;(2)P-value < 0.05。3.GO and KEGG analysis of DEGs were performed through Rstudio.The result filter is P<0.05.4.The common EDGs were further analyzed through the STRING website,we construct a protein interaction network,and the data were further processed by Cyto Hubba in Cytoscape to screen out the top 20 CRC core genes using the MCC algorithm.LASSO-COX analysis was performed on core genes to find the core genes most closely related to colorectal cancer and perform survival analysis.5.The invasion abundance of immune cells in colorectal cancer was evaluated by immune cell gene set,ss GSEA,and ESTIMATE algorithms,and the correlation of core genes with immune cells and immune checkpoints was analyzed by Hmisc pack.6.Based on core genes,risk scores were constructed through LASSO regression and multivariate COX analysis were constructed to construct clinical prediction models to evaluate the prognosis of colorectal cancer patients.7.The expression of TIMP1 in colorectal cancer was verified by western blot and immunohistochemistry.Results:1.A total of 293 DEGs were obtained after the intersection of the two groups of colorectal cancer chips,including 105 up-regulated DEGs and 188down-regulated DEGs.2.GO analysis showed that upregulated genes were mainly closely related to nuclear division,organelle lysis,extracellular structural organization and other functions.The downregulation of genes is mainly related to the process of cell metabolism of hormones,primary alcohol metabolism,and cell response to metal ions.KEGG analysis found that upregulated genes were mainly enriched in IL-17 signaling pathway,rheumatoid arthritis,and interaction of viral proteins with cytokines and cytokine receptors.The downregulated genes were mainly enriched in mineral absorption,nitrogen metabolism,aldosterone-regulated sodium reabsorption,and other pathways.3.The STRING website and the Cyto Hubba plugin screen the top 20 genes associated with colorectal cancer.LASSO-COX analysis finally determined that TIMP1 and CXCL8 were the core genes that most affected the prognosis of colorectal cancer.4.Immunocorrelation analysis showed that CXCL8 was significantly positively correlated with activated dendritic cells and activated CD4 T cells,while TIMP1 was significantly negatively correlated with activated B cells,eosinophils and immature B cells.CXCL8 was positively correlated with the immune checkpoint molecules PD-L2 and CTLA4,and the positive correlation between CXCL8 and PD-L2 was particularly significant.5.LASSO-COX analysis showed that risk score could be used as an independent risk factor for colorectal cancer patients,and the clinical prediction model could better predict the 3-year and 5-year mortality rates of colorectal cancer patients.6.Western blot analysis showed that the expression of TIMP1 in CRC tissues was significantly higher than that of paracancer normal tissue.The immunohistochemical results further verified the high expression of TIMP1 in colorectal cancer tissues.Conclusions:1.Bioinformatics analysis showed that TIMP1 and CXCL8 were highly expressed in colorectal cancer tissues,TIMP1 was associated with poor prognosis in colorectal cancer patients,and CXCL8 was associated with a better prognosis in colorectal cancer patients.2.Riskscores can be used as independent risk factors for colorectal cancer patients,and clinical prediction models can predict 3 and 5-year mortality in colorectal cancer patients.3.TIMP1 expression in colorectal carcinoma tissue was significantly higher than that of normal tissue adjacent to carcinoma,which may improve diagnosis and prognosis as a molecular target for colorectal cancer.