| 【Objective】To identify the gastric cancer(GC)associated metabolic biomarkers and metabolic quantitative trait loci(m QTL)and to exploration their mode of action,which is important for the prevention,early detection,diagnosis,and targeted treatment of gastric cancer.【Methods】1.Genome-wide and metabolome-wide datasets were generated from 233 GC patients and 233 healthy controls,which were matched by age,sex,smoking,drinking,HP infection status and blood sample collection time.Plasma metabolites were detected by liquid chromatography-mass spectrometry(LC-MS),and genotyping was performed using the Axiom(?) Precision Medicine Research Array.2.Those metabolic features with variable importance in the projection(VIP)values >1.0 in the orthogonal partial least squares discriminant analysis(OPLS-DA)model and FDR adjusted p values < 0.05 in the t-test were considered to be significantly different between GC cases and healthy controls.Then,logistic regression was performed to test the association between discriminant metabolites and incident GC.ROC curve analysis was performed to was used to assess the discriminative ability of the discriminant metabolites.Metabo Analyst conducts metabolite pathways to explore the metabolic pathways affected by differential metabolites in gastric cancer.3.After that,we analyzed the associations between discriminant metabolite levels and genome chip variants using a generalized linear model analysis in TASSEL software(version 5.0).The significant SNPs were annotated to the neighboring genes of 1000 Genomes Project(hg19/1000 Genomes ASN)downloaded from the University of California Santa Cruz(UCSC)genome browser.Further,we conducted Gene Ontology(GO)enrichment analysis to investigate possible biological,molecular,or cellular processes associated with significant SNPs related genes using Metascape.4.GLM was used to perform metabolome Genome-Wide Association Study(mGWAS)on the gastric cancer group and the healthy control group to locate the metabolism-related genes of the gastric cancer group and the control group.Then,the genes in the GC only group and the control only group outside the intersection were defined as differential metabolism-related genes.Then,metascape was used to perform GO and KEGG pathway enrichment analysis of metabolism-related genes in the gastric cancer only group and the control only group to compare the differences in the mode of action of metabolism-related genes between the two groups.5.Download the gene expression data of Asian gastric cancer patients from the TCGA database for differential analysis to externally verify the relationship between the located genes and gastric cancer.【Results】1.OPLS-DA results show that there is a significant difference in plasma metabolism profiles between gastric cancer patients and healthy controls;combined with VIP value> 1 and t-test FDR＿P value<0.05,22 discriminant metabolites in the plasma of GC patients and healthy controls have been screened out,including cytidine monophosphate 、 uridine 5’-monophosphate 、 uridine 5’-diphosphate 、 inosine triphosphate、guanosine、linoleic acid、 platelet activating factor,etc.After adjusting the intake of pickles,logistics regression analysis of 22 metabolites still showed that they were related to gastric cancer.The combined diagnostic efficiency of these 22 discriminant metabolites is as high as AUC: 0.897(95% CI: 0.869-0.924),showing high diagnostic value;the results of metabolic pathway enrichment analysis show that the 22 discriminant metabolites are mainly enriched in sub-four metabolic pathways including linoleic acid metabolism,pyrimidine metabolism,purine metabolism and folate biosynthesis.2.The genome-wide association analysis of 22 discriminant metabolites found that 57 SNPs site mutations were significantly related to 6 metabolites(FDR＿P value< 0.05).Among them,3 SNP sites have been identified in the previous mGWAS studies in the general population.In addition,bioinformatics analysis found that the60 genes corresponding to these 57 SNP sites are mainly involved in biological processes such as T cell receptor signaling pathways and immune response regulation signaling pathways.3.According to the mGWAS analysis results of the gastric cancer group and the healthy control group,a total of 9 metabolites and 233 SNPs(corresponding to 270genes)in the GC group and 16 metabolites and 150 SNPs(corresponding to 154genes)in healthy control group were found to have significant statistical associations(FDR＿P value<0.05),of which there were 10 overlapping genes.The enrichment analysis of GO and KEGG found that compared to the control only group,metabolism-related genes are mainly enriched in biological pathways such as glucose metabolism,nucleotide metabolism,while the GC only group metabolism-related genes are mainly involved in axon production,neuronal differentiation and other biological processes.4.The results of differential analysis of gene expression data in cancer and para-cancerous tissues of Asian gastric cancer patients downloaded from the TCGA database revealed that 12 of the 60 genes identified by the mGWAS analysis of 22 discriminant metabolites were different in expression in cancer and para-cancerous tissues(FDR＿P value<0.05).In the mGWAS analysis of the gastric cancer group and the healthy control group,93 genes out of the 357 genes outside the intersecting group were expressed differently in cancer tissues and para-cancerous tissues(FDR＿P value<0.05).【Conclusion】Our study found that 22 discriminant metabolites including cytidine monophosphate,uridine 5’-monophosphate,uridine 5’-diphosphate,inosine triphosphate,guanosine,platelet activating factor,etc.may become diagnostic biomarkers for gastric cancer,and 57 SNPs loci provide an important reference for the discovery of gastric cancer genetic biomarkers.At the same time,we provide a new perspective on the understanding of the biological mechanism of gastric cancer from the perspective of the mode of action of metabolism-related genes,suggesting that the role of the nervous system in the occurrence of gastric cancer should be paid attention to. |
