Establishment Of Autophagy-related Gene Regulatory Network And Prognosis Model Of Head And Neck Squamous Cell Carcinoma Based On TCGA Database

Background and ObjectiveHead and neck squamous cell carcinoma(HNSCC)is the sixth most common malignant tumor in the world.According to statistics,there are more than 500,000 cases of head and neck squamous cell carcinoma(HNSCC)reported in the world every year,more than 90 percent of them are squamous cell carcinoma.Some HNSCC patients with surgery or radiation therapy such as nasopharyngeal carcinoma or oral cancer have good therapeutic effect at an early stage,And patients with local recurrence or distant metastasis can be treated with chemotherapy combined with radiotherapy.No matter what kind of treatment method,it will have a significant impact on the patient’s physical functioning and quality of life.Significant progress has been made in the diagnosis and treatment of HNSCC in the past decades.However,more research on head and neck tumors is needed to improve the adverse effects and collateral damage caused by specific tumors in HNSCC patients.Although several biomolecular markers,including clinical parameters,have been identified,single molecular markers and clinical parameters are very limited in predicting patients prognosis in specific sites.Molecular markers that use gene expression to predict the prognosis of HNSCC from genomics have been attempted.There have been many studies using multiple molecules to predict the prognosis of HNSCC,such as glycolysis genes,hypoxia genes,metabolism genes and ferroptosis related genes.Autophagy is a pathway of intracellular self-degradation,it is a process by which cells transport damaged,deformed,senescent and nonfunctional proteins and some organelles to lysosomes for digestion and degradation.Autophagy has two sides in the process of tumor genesis.Autophagy has different effects in different stages of tumor genesis and can provide an environment that is conducive to the survival and development of tumor cells.In this respect,it has a certain protective effect on tumor cells.Autophagy is also an important way for normal cells to clear themselves.However,further studies on the role of autophagy-related genes in the prognosis of HNSCC and their comprehensive analysis are needed.For HNSCC,various treatments have greatly improved the prognosis of patients,but there are still some patients who cannot benefit from treatment.Therefore,it is an urgent task to study and find effective markers for the prognosis of HNSCC patients to guide clinical practice.In this study,bioinformatics methods were used to analyze the expression of autophagy-related genes in HNSCC transcriptomics from the TCGA database,and through screening,we tried to construct a prognostic model of HNSCC based on the expression of autophagy-related genes.And the mutual regulatory network of autophagy-related genes and the potential signal pathways are analyzed.In addition,by analyzing the correlation between autophagy-related genes and clinical parameters,we try to provide theoretical support and potential therapeutic targets for the clinical treatment of HNSCC.Material and methods1.This study obtained the transcriptome and clinical data of HNSCC patients from the TCGA public database(The Cancer Genome Atlas,TCGA).2.The limma package of R to analyze the difference in autophagy-related gene expression between normal samples and tumor samples;using Gene Ontology(GO)and KEGG pathway analysis enrich autophagy-related genes;Single-factor cox regression and LASSO regression were used to screen autophagy genes that related to the prognosis of HNSCC;multi-factor cox regression analysis is used to screen the prognostic model of autophagy-related genes of HSCNN.3.Calculate the risk score based on the prognostic model.According to the risk score,the study subjects are divided into high-risk groups and low-risk groups.Kaplan-Meier survival analysis is used to compare the survival curves of the high-risk group and the low-risk group.4.Use the survival package of R software to perform multi-factor adjustment analysis on the relationship between risk score and prognosis of HNSCC.The adjusted factors include: age,gender,tumor grade,pathologic stages tumour,T stage,M stage and N stage.Calculate the corresponding hazard ratio(HR)and 95%confidence interval(CI).5.Use the survival ROC software package of R software to evaluate the constituted risk model,calculate time-dependent receiver operating characteristic curve(ROC)of HNSCC for one,three and five years respectively,and calculate the the area under the ROC curve and its 95% confidence interval.6.Use the rms package of R software to draw a nomograph including clinical parameters as a tool for predicting the prognosis of HNSCC.7.Use the beeswarm package of R software to analyze the relationship between autophagy-related genes that constitute prognostic labels and clinical parameters.8.Divide the samples into high-risk groups and low-risk groups according to the risk score,and use GSEA software to analyze the enrichment of signaling pathways in the high-risk group and the low-risk group.Result1.After gene name conversion,the obtained expression of matrix includes 44 normal tissue samples and 502 tumor tissue samples,with 56753 cases of gene data.According to preliminary statistics,there are 527 patients,385 males and 142 females,aged 19-90 years old,with an average age of 60.9 years,199 deaths and 328 survivors.2.Compare the expression of 232 autophagy-related genes in tumor samples and normal samples.The expression ratio was greater than 1,and there were 48 genes with statistical difference,38 of which were up-regulated and 10 of which were down-regulated.3.GO enrichment analysis showed that the molecular functions were mainly concentrated in neuronal necrosis,positive regulation and emergency regulation of cytokine-mediated signaling pathway,apoptosis regulatory signaling pathway,mitochondrial regulation,regulation of protein localization on membrane and autophagy regulation.The composition of cell is mainly concentrated in integrin complex,autophagosome and cell adhesion protein complex.Biological processes are mainly concentrated in receptor-ligand activity,receptor-priming activity,factor activity,protein binding.4.KEGG pathway analysis showed that it was mainly involved in apoptosis,human cytomegalovirus infection,HPV infection,herpes virus infection,cisplatin resistance,EGFR tyrosine kinase inhibitor resistance,HIF-1 signaling pathway,PDL1 and PD-1 checkpoint pathway,IL-7 signaling pathway,PI3K-Akt signaling pathway and so on.5.Univariate cox regression analysis was performed to analyze the prognosis of the HNSCC remaining,and the results determined a total of 48 autophagy genes related to the prognosis of HNSCC.The expression of 25 genes out of 48 genes is conducive to the prognosis of HNSCC patients,and the expression of 23 genes is not conducive to the prognosis of HNSCC patients.6.Lasso regression analysis combined with multivariate Cox regression analysis showed that a total of 11 autophagy-related genes were included in the final model.Overexpression of five genes(BAG3,ST13,ATIC,CALCOCO2,and LAMP1)was associated with poor prognosis of HNSCC,and the low expression of four genes(MAP2K7,IKBKB,CFLAR,and EEF2K)was associated with poor prognosis of HNSCC,and two genes(ATG5 and NKX2-3)were not relevant.7.Divide the total sample into a high-risk group and a low-risk group according to the mean value of the risk score.The prognosis of the high-risk group is significantly worse than that of the low-risk group(P<0.0001).8.Multivariate cox regression analysis showed that risk score was an independent predictor of poor prognosis of HNSCC(HR=1.546,95%CI :1.379-1.734,P<0.001).The one-year,three-year and five-year predicted survival ability of the prognostic label is evaluated,and the area under the ROC curve is calculated,respectively: 0.737(95%CI:0.627-0.774)、0.754(95%CI:0.618-0.761)and 0.696(95%CI:0.526-0.718),the predicted AUC area is larger than each clinical parameter.9.The nomogram has a good agreement between the one-year and three-year forecasts,and there is a slight deviation in the forecast of the five-year survival rate.10.GSEA enrichment analysis showed that the high risk group enriched the pathway of proteasome,pathogen infection,glycosaminoglycan synthesis,steroid synthesis and polysaccharide synthesis.The low-risk group was enriched in pathways Of ABC transporter,FcεRI signaling,taste transduction,B-cell receptor signaling,immunodeficiency,pantothenic acid,and coenzyme A synthesis.ConclusionThis study identified the HNSCC prognostic risk model consisting of 11autophagy-related genes(BAG3,ST13,ATG5,EEF2 K,MAP2K7,ATIC,IKBKB,CFLAR,CALCOCO2,NKX2-3,LAMP1).The 11 autophagy-related genes were identified as an independent prognostic factor for HNSCC prognosis.The constructed prognostic model showed good performance in 1-,3-,and 5-year survival.These identified prognostic genes may become new potential biomarkers and therapeutic targets.