| ObjectiveImbalances in the tumor microenvironment(TME)can promote tumor growth,invasion,and spread,as well as immunosuppression and immune escape.Because hypoxia is a symptom of TME imbalance and is linked with a poor prognosis in cancer patients,it is critical to study the effect of hypoxia on the prognosis of gastric cancer(GC)patients.This research aimed to look into the phenomenon and mechanism by which the hypoxic microenvironment adds to a poor prognosis in gastric cancer by suppressing immunity.Methods1.Identification of genes with diagnostic potential for stomach cancer.The Cancer Genome Atlas(TCGA)database was used to acquire gastric cancer mRNA transcriptome data and related clinical information,and differential gene expressions(DEGs)were obtained using R language differential analysis.Protein-Protein Interaction Networks(PPI)were used to further filter the data,and a random forest(RF)algorithm was used to build a diagnostic model for gastric cancer.External validation of the diagnostic models was performed using the Gene Expression Omnibus(GEO)and Genotype-Tissue Expression(GTEx)databases,in addition to immunohistochemistry studies.2.Among the above-mentioned gastric cancer diagnosis-related genes,hypoxia genes linked with prognosis were sought.By downloading the hypoxic genes and intersecting them with the DEGs obtained in the prior phase,the hypoxic DEGs were obtained.The hypoxic DEGs were submitted to enrichment analysis using gene ontology(GO)and the Kyoto Encyclopedia of Genes and Genomes(KEGG),as well as a heat map of the DEGs and volcano maps.Finally,important hypoxic DEGs were identified and a prognostic risk model for hypoxia in gastric cancer was constructed using Cox regression models.The model was validated using the GSE84437 dataset and assessed using Kaplan-Meier survival analysis and the area under the curve(AUC)of the time-dependent receiver operating characteristic curve(ROC).3.Immune infiltration analysis of prognostic models was conducted using bioinformatics techniques to investigate the mechanisms of hypoxic prognostic genes in gastric carcinogenesis’ s development and immune infiltration.The CIBERSORT method was used to analyze immune infiltration in the prognostic model.Human Protein Atlas(HPA)and Cancer Cell Line Encyclopedia(CCLE)databases were used to verifying hypoxic genes.4.The gene BGN was chosen for this research to investigate its biological behavior and immunological impact on gastric cancer.BGN was subjected to macrophage immuno-infiltration and gene set enrichment analysis(GSEA).To confirm the differences in BGN protein expression in gastric cancer and paracancerous tissues,immunohistochemical assays using gastric cancer tissue microarrays were conducted.The study looked at the link between BGN gene expression levels and clinicopathological factors in gastric cancer patients.Prognostic nomograms containing BGN gene expression levels and multiple clinicopathological factors were created based on the findings of multifactorial Cox regression analysis.Results1.Differential analysis revealed 947 DEGs,of which 419 were up-regulated and 526 down-regulated.After PPI analysis,the 10 key genes(COL1A2,COL3A1,COL1A1,FN1,BGN,MAD2L1,COL4A1,SPARC,TTK,SPP1)that were substantially associated with a gastric cancer diagnosis were identified,and a 10-gene model for gastric cancer diagnosis was developed.The diagnostic model was evaluated using ROC curves,and the AUC value in the training set was 0.972 and0.914 in the external validation set,indicating the model’s decent predictive ability.Immunohistochemical experiments were carried out to investigate the protein expression levels of ten key genes,and it was discovered that,except for the FN1 and SPP1 genes,the expression of the remaining genes was significantly higher in gastric cancer tissues(P<0.05).2.The retrieved hypoxic genes were intersected with 947 DEGs to yield 57 hypoxic DEGs,of which 49 were up-regulated and 13 down-regulated.Cox regression analysis was used to identify three hypoxic core genes(BGN,CAV1,SERPINE1)that were substantially associated with prognosis to build a three-gene hypoxic prognostic risk model for gastric cancer.According to the median risk score,gastric cancer patients can be divided into high-risk and low-risk groups.The Kaplan-Meier survival analysis revealed that those in the high-risk group had a significantly lower overall survival(OS)rate than those in the low-risk group(P<0.01).The time-dependent ROC curve’s prediction of 5-year survival for patients with gastric cancer had an AUC of 0.899,and comparable outcomes were attained in the validation set,demonstrating the prognostic model’s excellent predictive power.3.Immune infiltration studies using the CIBERSORT method revealed a significant decrease in immune cell activity and an increase in the proportion of monocytes in high-risk patients;after GSEA analysis,we discovered that CD8+ T cell infiltration was significantly lower in the high-risk group.This was done in 37 different gastric adenocarcinoma cell lines,and the CAV1 gene was found to be strongly expressed in the FU97 cell line,while the BGN and SERPINE1 genes were found to be relatively more expressed in many gastric cancer cell lines.4.According to immuno-infiltration studies of the highlighted gene BGN,elevated BGN expression encourages M2 macrophage polarization and their differentiation into tumor-associated macrophages(TAM),leading to gastric cancer.The highly upregulated BGN was primarily enriched in the "Notch signaling pathway" and "adhesion junctions" signaling pathways,according to single gene enrichment analysis.BGN protein expression was considerably higher in gastric cancer than in para cancer(P<0.001),according to immunohistochemistry of human gastric cancer tissue microarrays,which was consistent with the bioinformatics analysis.BGN expression score,tumor size,depth of infiltration,lymph node metastasis,pathological stage,choroidal carcinoma thrombosis,and nerve infiltration were found to be distinct risk factors for gastric cancer mortality in a multi-factorial Cox regression prognostic analysis.These results were used to construct prognostic Nomograms for predicting individualized survival times in gastric cancer patients.Conclusion1.The machine learning-based 10-gene gastric cancer diagnostic model developed in this research is anticipated to be a useful tool for gastric cancer screening.2.This research suggests a 3-genetic hypoxia prognostic risk model that can more accurately predict survival status and the extent of immune cell infiltration in patients with gastric cancer.3.Immune infiltration studies have revealed a significant decrease in immune cell activity and a rise in the proportion of monocytes in high-risk patients in a hypoxia prognostic risk model.4.The BGN gene was discovered to be involved in both the diagnostic and prognostic models of gastric cancer in this research.High BGN expression supports M2 macrophage polarization and differentiation towards TAM,which in turn promotes gastric carcinogenesis,and BGN may become a key target for hypoxia-targeted therapy.5.Nomograms of independent factors associated with gastric cancer patients were created to predict the likelihood of OS in gastric cancer patients,which has possible clinical applications. |
PDF Full Text Request