| Research purpose:The purpose of this study was to investigate the metabolites associated with poor prognosis of colorectal cancer and the metabolic pathways involved.Differentially expressed mRNAs were screened based on colorectal cancer and normal colorectal tissues with bioinformatic analyses to explore the prognostic significance of differentially expressed mRNAs involved in signaling pathways.Methods:Two sections were used to elucidate this study.In the first section of metabonomic research,236 colorectal patients with different pathological stages admitted to gastrocolorectal department of Berthune first hospital of Jilin University were recruited in this research.Preoperative blood samples collected from these patients were examined by high performance liquid chromatography-time-of-flight mass spectrometry and a blood metabolic profile of the patients will be obtained.Firstly,236 colorectal patients were grouped by K-means clustering analysis on the basis of their original clinical date and pathological data.Principal component analysis,partial least squares discriminant analysis and t-test analysis were used to analyze the metabolic patterns differences in different groups of colorectal cancer patients to find differential metabolites with statistical significance(P<0.05),which were utilized to verify discriminant ability of different groups of patients with colorectal cancer and signaling pathways making between-group differences.HMDB online database was used to retrieve lipid metabolites,and the related data of lipid metabolites were conducted with single-factor and multi-factor Cox regression analysis to establish the prognostic model based on lipid metabolites,and the prognostic value of the model was evaluated by Kaplan-Meier(KM)survival analysis,independent prognostic analysis,and time-dependent receiver operating characteristics(ROC)curves.Section 2:Based on the first section of this study,we found that the colorectal metabolites were mainly enriched in glycerophospholipid metabolism pathway,so in this section,we further explored the prognostic significance of glycerophospholipid metabolism in colorectal cancer from molecular level.glycerophospholipid metabolism-related mRNAs obtained from GO,KEGG and SMPDB databases were integrated with survival data(survival time and survival status)and differential mRNAs expression matrix of colorectal cancer patients obtained from TCGA database to obtain the glycerophospholipid metabolism-related mRNAs expression matrix of colorectal cancer patients including survival data.Single Cox regression analysis was performed to identify the survival-related glycerophospholipid metabolism-related mRNAs of colorectal cancer,and multiple Cox regression analysis for the survival-related glycerophospholipid metabolism-related mRNAs was performed to establish a prognostic model.Patients with colorectal cancer were divided into high-risk group and low-risk group based on risk score of the prognostic model,that is,patients with greater than or equal to the risk score were classified into high-risk group,and patients with less than the risk score were classified into low-risk group.The prognostic significance of this model was further evaluated by KM survival analysis,ROC curve,and single and multiple independent prognostic analyses.Clinical correlation analysis was conducted to further explore the statistical relationship between the survival-related glycerophospholipid metabolism-related mRNAs and clinical characteristics,and the nomogram based on independent prognostic factors presented the overall survival rates of colorectal patients intuitively.Gene set enrichment analysis(GSEA)was used to determine the genetic characteristics of the colorectal patients in the high-and low-risk group.Moreover,relationship between glycerophospholipid metabolism-related mRNAs and immune infiltration further broadened the view of the relationship between metabolism and immunity.Results:Section 1:(1)236 patients with colorectal cancer were divided into two groups by clustering analysis,K-M survival analysis found that there was a difference in survival between the two groups.We found that the survival rate of patients with colorectal cancer in group 1 was worse compared with that in group 2 and mucinous adenocarcinoma was more common.Through metabonomic analysis,we found 175 different metabolites,most of which were fatty acids,glycosylthioesters and glycosolipids and so on,mainly involved in the glycerophospholipid metabolism pathways and sphingolipid metabolic pathways.(2)Single Cox regression analysis was further performed on the 175 differential metabolites to obtain 14 survival related metabolites,which were then performed multiple Cox regression analysis to obtain the prognostic model based on six metabolites(high-risk metabolites: Cer,Ganglioside GT3,Lysope,PA and PS;low-risk metabolites: Substance P).According to the median value(0.875)of the risk score of the prognostic model,patients divided into high-risk and low-risk groups,K-M survival analysis suggested there was a significant difference in overall survival rate between high-and low-risk group(P<0.0001).Single and multiple independent prognostic analysis indicated that the prognostic model based on six metabolites could be deemedd as independent prognostic factor to predict survival,1-year,3-year and 5-year area under curve(AUC)of ROC curve were respectively 0.769,0.711 and 0.723,reflecting the favourable accuracy in the prediction of patients' survival.(3)The results of clinical correlation analysis indicated that there were statistically significant(P < 0.05)for four metabolites(Cer,Ganglioside GT3,PS and PA)of the prognostic model in pathological stage,T Stage,N Stage and M Stage.Based on the pathological stage,it can be used to directly calculate colorectal cancer patients' 1-year,3-year and 5-year overall survival.Section 2:(1)91 glycerophospholipid metabolism-related mRNAs were retrieved from the glycerophospholipid metabolism pathway.After single Cox regression analysis being performed on differentially expressed glycerophospholipid metabolism-related mRNA expression matrix of in 576 patients with colorectal cancer and survival data,we identified the 12 survival-related glycerophospholipid metabolism-related mRNAs.Multiple Cox regression analysis was performed on 12 survival-related glycerophospholipid metabolismrelated mRNAs to obtain the prognostic model based on four glycerophospholipid metabolism-related mRNAs(CDIPT,GDE1,JMJD7-PLA2G4B and PLA2G2D)to predict the prognosis of patients with colorectal cancer.Patients were classified as high-risk or low-risk based on the median risk score(0.988).Moreover,K-M survival analysis showed that the survival rate of patients in the high-risk group were significantly lower than that in the low-risk group(P=0.00067).Single and multiple independent prognostic analysis showed that this model could be used as an independent prognostic factor.The AUC of the time-dependent ROC curve suggested that the model had an auspicious accuracy in predicting survival rates at 1 year(AUC=0.641),3 years(AUC=0.684)and 5 years(AUC=0.676).(2)Clinical correlation analysis revealed that 9 survival-related glycerophospholipid metabolism-related mRNAs(AGPAT5,CDS1,GDE1,GPD1 L,LCAT,AGPAT1,CDIPT,PLA2G2D and PLA2G5)showed statistical significance(P < 0.05)in pathological type,pathological stage,T Stage,N Stage and M Stage.Based on the independent prognostic factors(pathological stage and prognostic model),we plotted the nomogram to predict overall survival of patients with colorectal cancer,indicating favourable accuracy and consistency.(3)GSEA stratified the enrichment pathways of patients in the high-and low-risk group,in which the gene sets in the high-risk group were mainly enriched in 17 pathways(NOM P<0.05,FDR q<0.25),and the gene sets in the low-risk group were mainly enriched in 62 pathways(NOM P<0.05,FDR q<0.25).Moreover,immune infiltration analysis for glycerophospholipid metabolism-related mRNAs in the prognostic model indicated that CDIPT,GDE1 and PLA2G4B were linearly correlated with 8 immune cells(Regulatory T cells,M0 macrophages,memory activated CD4+ T cells,activated dendritic cells,eosinophils,memory resting CD4+ T cells,resting dendritic cells and neutrophils)(4)576 colorectal cancer patients were established according to the immune cell infiltration status of the patients,and then the Kaplan-Meier survival analysis was performed.There were survival differences between different immunotypes(p=0.019).We identified all the differential genes and used the expression of the differential genes to perform consistent clustering analysis on the mRNA data of 576 groups of colorectal cancer patients and perform genotyping.We then divided the colorectal cancer patients into the ICIscore-High and ICIscore-Low groups and analyzed the survival by using PCA according to the immunophenotyping and genotyping.The results showed that there was a difference in survival status between the high immune score group and the low immune score group(P=0.041).Then we divide the tumor data into high and low immune groups according to the level of immune scores,and perform differential gene analysis again between the two groups,and use the derived differential genes for GO enrichment analysis.It is found that in the two different enrichment pathways,there are also lipid metabolism pathways,and these pathways are mainly related to fatty acid and phospholipid metabolism.(5)We downloaded 56 colorectal cancer cell lines mRNA full gene expression matrix from the Encyclopedia of Tumor Cell Lines(CCLE),and found that there were 74co-expressed genes related to GDE1(|cor|>0.5,P<0.0001)),of which 47 are positively related genes and 27 are negatively related genes.Through the GO enrichment analysis of these 71 genes,we found that lipid pathways such as the glyceride pathway,the phospholipid inositol pathway,the phospholipid metabolism pathway,and the glycerophospholipid pathway were enriched.In addition,we also performed KEGG enrichment analysis on these 71 genes and found that lipid-related sphingolipid metabolism signaling pathways and immune cell-related T cell receptor signaling pathways,B cell receptor signaling pathways,and PD-L1 expression The PD-L1 checkpoint pathway is enriched in it.Conclusions:(1)Lipid metabolism,especially the glycerophospholipid metabolism pathway,may be involved in the occurrence and development of colorectal cancer and is related to the poor prognosis of patients.Among them Cer(d18:0/14:0),Ganglioside GT3(d18:0/18:1(9Z)),LysoPE(22:6(4Z,7Z,10 Z,13Z,16 Z,19Z)/0:0),PA(20: 3(5Z,8Z,11Z)/24:1(15Z)),PS(20:4(5Z,8Z,11 Z,14Z)/14:1(9Z)),Substance P these six types Metabolites are related to the prognosis of colorectal tumors and are risk factors that affect the prognosis of colorectal patients.(2)Among the genes in the glycerophospholipid metabolism pathway,the expression of CDIPT,GDE1,PLA2G4 B and PLA2G2 D genes are related to the prognosis of colon cancer patients,and may be used as predictors of colon cancer prognosis.And these four genes are related to immune cell infiltration.(3)Lipid metabolism pathway and immune cells act together on the tumor development of patients with colorectal cancer,and are related to the prognosis of patients with colorectal cancer.GDE1 may serve as a key tumor suppressor gene. |
PDF Full Text Request