Identification Of Molecular Subtypes And Establishment Of Prognostic Model Of Colorectal Cancer Based On Lactate-related Genes

Background: Lactic acid,as an energy-rich metabolite,plays an important role in the occurrence and development of Colorectal cancer(CRC),however,the specific mechanism of lactic acid in regulating CRC is not yet clear.The purpose of this study was to explore the potential mechanism of Lactate-related genes(LRGs)in CRC and construct a prognostic model to predict the prognosis of CRC patients.Methods: The m RNA expression profiles and clinical data of CRC patients were collected from The Cancer Genome Atlas(TCGA)database,GSE59382 and GSE17536 cohort,and lactic acid-related genes were collected from MSig DB database.Consensuscluster Plus package was used for consensus clustering of CRC patients and functional analysis of differential genes among subtypes was performed.Lactate-related gene signature(LRGS)was constructed by univariate Cox regression,lasso regression and multivariate Cox regression analysis.The differences in the expression of model-related genes were investigated by PCR using colorectal cancer tissues and paracancer tissues of 30 CRC patients.Survival analysis and Receiver Operating Characteristic(ROC)curves were used to evaluate the predictive effect of the model.The prognostic model and important clinicopathological features were combined to establish a nomogram.In addition,the relationship between the prognostic model and immune infiltration and drug sensitivity was investigated.Results:In this study,CRC patients were divided into two subtypes based on key lactate genes.Furthermore,a prognostic risk model was established based on 8 differential genes between the two subtypes: Risk score = 0.0068*(FABP4 expression level)-0.0049 *(NOX1 expression level)+ 0.0054 *(TM4SF4 expression level)-0.0120 *(NOS2expression level)-0.0160 *(GNL3 expression level)+ 0.0171*(AP2M1 expression level)+ 0.0104 *(KLK10 expression level)-0.0166 *(PAFAH1B3 expression level).The expressions of FABP4,TM4SF4,KLK10,AP2M1,NOS2 and PAFAH1 B were different in colorectal cancer tissues and adjacent tissues,while NOX1 and GNL3 showed no significant difference.CRC patients with high risk scores showed poorer outcomes overall.Meanwhile,there were differences in tumor microenvironment scores,immune infiltration and drug sensitivity among patients with high and low risk scores.In addition,the predictive performance of the line graph combined with LRGS and clinicopathological parameters was also higher.Conclusion: LRGS constructed in this study are a reliable biomarker that can be used to predict the clinical outcome of CRC patients,assess the immunoinvasive status of patients,and predict susceptibility to drugs.