| BackgroundHead and neck cancer(HNC)is one of the most important tumors that endanger human health.There are more than 870,000 new cases worldwide each year,resulting in more than 440,000 deaths.Over 90%of HNC is head and neck squamous cell carcinoma(HNSCC).Currently,the setting of the standard treatment plan for HNSCC is still based on the traditional clinicopathological staging.The data in the past 20 years shows that with the rapid development of tumor treatment technology,although the survival rate of many malignant tumor patients has been significantly improved,the 5-year survival rate of HNSCC patients did not improve significantly.The revelation of tumor heterogeneity raises new questions and research areas for accurate prognosis prediction and treatment settings of tumors.Studies have confirmed that HNSCC is a highly heterogeneous tumor.Patients with similar clinicopathological characteristics such as age,gender,and tumor stage have different prognosis and sensitivity to radiotherapy,chemotherapy,and immunotherapy.Therefore,based on traditional clinicopathological factors,it can no longer meet the individualized treatment needs of patients.With the rapid development of tumor molecular biology and bioinformatics technology,the molecular typing of many tumors such as lung cancer,gastric cancer,and breast cancer has been applied to the setting of clinical treatment strategies and significantly improved the curative effect,making patients’ 5-year survival rate has been significantly improved.However,the prognosis prediction and treatment plan of HNSCC are still mainly based on traditional imaging and histological techniques.Therefore,exploring new molecular stratification markers to accurately evaluate patient prognosis and treatment response and improve patient survival is an urgent clinical problem in the field of HNSCC research.A large number of studies have shown that the heterogeneity of the tumor microenvironment(TME)composed of tumor cells and surrounding area components(vascular,immune cells,cytokines,chemokines,fibroblasts,extracellular matrix,etc.)is the key determinant of variability in patient prognosis and response to treatment.Therefore,analyzing TME heterogeneity and its relationship with clinical prognosis and treatment response settings at the molecular and cellular levels is still a hot area of cancer research.Many factors have been found to play an important role in the composition of TME heterogeneity.Among them,the research on the cell death state of tumor tissue and the heterogeneity of tumor progression,prognosis and treatment response has attracted much attention,and it is one of the hot spots in tumor research.Cell death is a basic process necessary for the growth and development of the body.The body’s normal biological responses,external stimuli,or responses to treatment can initiate the cell death process in the tumor microenvironment.There are many types of cell death,among which apoptosis,pyroptosis and autophagy are the important types of cell death.Existing research data show that these types of cell death are related to the formation of TME,tumor progression,prognosis and therapeutic response,but the research data in HNSCC is limited.To analyze the relationship between these death types and TME,prognosis and treatment response and its clinical significance is an urgent need for accurate prognosis prediction and precise treatment settings in HNSCC.Based on this,our study analyzed the expression of apoptosis,autophagy and pyroptosis-related genes based on the clinical pathology and transcriptome sequencing(RNA sequencing,RNA-seq)data of the HNSCC cohort in the TCGA database.According to univariate Cox regression analysis,LASSO regression analysis and multivariate Cox regression analysis,16 survival-related cell death genes were screened out,and a cell death index(CDI)was constructed.The relationship between CDI and clinical prognosis,treatment response,and TME was analyzed,and the constructed prediction marker was verified and analyzed using the HNSCC transcriptome data from Qilu Hospital of Shandong University and cohorts of GSE42743 and GSE65858.The research results of this part were published in"Frontiers in Immunology" and applied for an invention patent(The application of immunogenic cell death-related genes in survival prognosis and radiotherapy response of head and neck squamous cell carcinoma,review stage).Analysis of the 16 screened cell death-related genes found that 10 of them were autophagy-related genes,and the prognostic predictive value of autophagy index was better than that of apoptosis index and pyroptosis index.Comparative analysis of CDI distribution in different clinicopathological feature subgroups found that oral squamous cell carcinoma(OSCC)had a higher proportion of high CDI than pharyngeal squamous cell carcinoma,and the prognosis of OSCC patients with high CDI was worse.However,there are few studies on the relationship between autophagy and OSCC progression,prognosis prediction,and treatment settings.A comprehensive and in-depth analysis of OSCC autophagy and its significance is particularly important for improving its therapeutic efficacy.Autophagy is a"self-digestion" process characterized by the appearance of a double-membrane structure in the cytoplasm surrounding proteins or organelles to form autophagosomes.On the one hand,it maintains the metabolism and survival of cells under starvation and stress,and on the other hand,it eliminates damaged proteins and organelles to maintain the quality and quantity of proteins and organelles.Existing studies have shown that autophagy plays a key role in various pathological processes including tumors,aging,autoimmune diseases,and infections.Although tumor treatment strategies based on autophagy interference have entered clinical trials,their efficacy is not satisfactory.Studies by us and other research groups have found that interfering with autophagy can lead to the death of tumor cells and inhibit their proliferation,it can also lead to changes in the cytokine secretion profile of tumors,which in turn affects TME,tumor progression and treatment resistance.Therefore,the research on secretory autophagy of tumor and its influence on TME,tumor progression and treatment response has become a new hotspot in the field of tumor autophagy research.Toll-like receptors(TLRs)is a key pathway to regulate the level of autophagy,which can mediate the activation of autophagy.Among them,the activation of TLR3 pathway affects the surv ival of tumors and can affect the sensitivity of tumors to radiotherapy and chemotherapy by remodeling TME.Double-stranded RNA(dsRNA)overload is one of the characteristics of tumors and a key factor affecting the biological behavior of tumors.Poly(I:C),a synthetic dsRNA present in some viruses that can be used to mimic the effects of extracellular dsRNA,is a TLR3 agonist and its analog has been evaluated as a tumor vaccine.Therefore,in this study,we used transcriptome sequencing technology,ELISA technology,etc.to analyze the transcriptome and inflammatory cytokine secretion of poly(I:C)-treated OSCC cell line CAL27,and further using technologies such as transcriptome sequencing,ELISA and reverse phase protein array(RPPA),the effects of autophagic supernatants from OSCC cells on the gene expression and secretome of THP-1-derived macrophages and their relationship to reactive oxygen species(ROS)were analyzed.In this study,we used whole transcriptome sequencing analysis for the first time to describe the genomic changes of poly(I:C)-intervened CAL27 cells,and found that poly(I:C)significantly inhibited the proliferation of CAL27 cells and induced apoptosis,and also and significantly altered the secretion of inflammatory cytokines IL-6 and CSF2 associated with tumor progression.It is suggested that in the future,the tumor treatment strategy based on autophagy interference needs to solve the effect of secretory autophagy,so as to design a more reasonable autophagy intervention therapy.In this study,we also explained for the first time the effect of the autophagy-inhibiting supernatant of CAL27 cells on the secretory autophagy of macrophages,and confirmed that reactive oxygen species scavenging can reverse the cytokine changes of macrophages.In this study,we used RPPA technology for the first time to establish the secretome map of macrophages in CAL27 autophagy-inhibiting supernatants,and found that the autophagy secretome of OSCC cells can affect the changes of macrophage secretome.Future therapeutic research based on the tumor autophagy microenvironment will provide new data and clues.Finally,we used the TCGA database and the transcriptome sequencing data of the OSCC cohort of Qilu Hospital of Shandong University to analyze and verify the significance of ATG5 and VDAC1,two genes with independent predictive value among the autophagy genes in CDI,in predicting prognosis and treatment response.Molecular markers for OSCC prognosis prediction based on autophagy genes were optimized.Based on the results of in vitro studies,it was confirmed that SPP1high-TSG101low OSCC patients have a better prognosis and a significant survival benefit after radiotherapy.Our research results provide new perspectives and data for the study of autophagy-based OSCC intervention strategies,and further in-depth research will provide new ideas for the design of clinical autophagy-targeted OSCC treatment programs.Part Ⅰ Constructing the prognosis prediction model of HNSCC based on bioinformatics analysis of cell death-related genesResearch contents1.Construct a cell death index and analyze its predictive value for HNSCC survival1.1 Based on the RNA-seq data of the TCGA database,the expression of autophagy,apoptosis and pyroptosis-related genes in HNSCC was analyzed,the survival-related genes were screened,and the cell death index was constructed.1.2 The KM survival curve was used to analyze the predictive effect of the cell death index on the survival of HNSCC,and two GEO databases(GSE65858 and GSE42743)and the HNSCC cohort of Qilu Hospital of Shandong University were selected for verification analysis.Time-dependent ROC curves evaluated the predictive value of cell death index for survival.2.Analysis of the predictive effect of cell death index on HNSCC treatment responseUsing the clinicopathological data of the TCGA database,the patients were grouped according to whether they received radiotherapy or chemoradiotherapy,and whether there was survival benefit after receiving treatment was analyzed.3.Differences in the immune microenvironment of different cell death index subgroupsIn the TCGA database,HNSCC patients were divided into high and low cell death index subgroups according to the cell death index,and the proportion of immune cell infiltration,immune checkpoint gene expression,and differentially expressed gene enrichment analysis were performed in the two groups of samples.The immune microenvironment status of the two groups of patients was evaluated.Research purposes1.Construct a cell death index in HNSCC using autophagy,apoptosis,and pyroptosisrelated genes,and evaluate its predictive value for survival and therapeutic response.2.Analyze the differences in the immune microenvironment between the high and low cell death index groups.Materials and methods1.Clinicopathological data and specimen collection.The cohort included 28 newly diagnosed HNSCC patients diagnosed from March 2010 to March 2015 in Qilu Hospital of Shandong University,and followed up until May 2019.Inclusion criteria:patients with complete clinicopathological information(age,sex,smoking,drinking,TNM stage,tumor differentiation,radiotherapy,chemotherapy)and survival information(survival and follow-up time,survival status).Patients with previous history of malignant tumors or receiving antitumor therapy and those with distant metastasis were excluded.2.The tumor paraffin tissues of 28 HNSCC patients were selected for transcriptome sequencing analysis,and the FPKM(Fragments Per Kilobase of exon model per Million mapped fragments)value was the gene expression level.Survival,immune infiltration,treatment response,etc.were analyzed based on gene expression and clinicopathological data.3.The gene expression data of head and neck cancer patients based on RNA sequencing(RNA-seq)in the TCGA database was downloaded from the database UCSC Xena(https://xenabrowser.net/datapages/).Clinicopathological information and follow-up data of head and neck cancer patients in the TCGA database were obtained from the GDC database(https://portal.gdc.cancer.gov/).The HPV status of patients with head and neck squamous cell carcinoma was determined with reference to the study by Cao et al.Patients with missing clinicopathological information,survival information,gene expression data,and HPV status,as well as patients with non-squamous cell carcinoma and those with a history of malignancy or receiving anti-tumor therapy were excluded.We finally obtained a HNSCC cohort of 379 cases.We downloaded two head and neck squamous cell carcinoma cohorts,GSE65858 and GSE42743,from the GEO public database(https://www.ncbi.nlm.nih.gov/geo/).and the gene expression data of both cohorts were microarray data.4.From the GSEA database(https://www.gsea-msigdb.org/gsea/index.jsp),the apoptosis-related gene set including 273 genes and the autophagy-related gene set including 153 genes were retrieved,the pyroptosis-related gene set containing 42 genes was obtained from the GSEA and GO databases(http://gepeontollogy.org/).5.Based on the HNSCC cohort data in the TCGA database,survival-related genes were screened through univariate Cox regression analysis,LASSO regression analysis,and multivariate Cox regression analysis in sequence,and a prognosis prediction model was constructed based on this.6.Use the Kaplan-Meier survival curve and log-rank test to analyze the predictive effect of the constructed model on survival,and verify it in Qilu Hospital of Shandong University and cohorts of GSE42743 and GSE65858.The predictive value of the constructed models on survival was evaluated using time-dependent ROC curves,univariate and multivariate Cox regression models.7.According to the radiotherapy and chemoradiotherapy response and survival information in the TCGA database,the Kaplan-Meier survival curve and log-rank test were used to analyze the predictive effect of the constructed model on the response to HNSCC treatment.8.Use integrated bioinformatics methods(immune cell infiltration and immune checkpoint gene expression analysis,differentially expressed gene analysis,GO and KEGG enrichment analysis,etc.)to explore the immune inflammatory microenvironment status of the samples in the prognosis prediction model.Results1.Based on the TCGA database HNSCC cohort,univariate Cox regression analysis found 48 survival-related apoptosis,autophagy and pyroptosis-related genes,followed by LASSO regression analysis and multivariate Cox regression analysis to obtain 16 survival-related cell death genes,including 5 apoptosis-related genes(BCAP31,CDKN2A,FNTA,STK24 and YWHAQ),10 autophagy-related genes(ATG5,CSNK2A2,DYNC1I1,MVB12A,MVB12B,RRAGA,TSG101,PRKN,VDAC1 and VPS37C)and 1 pyroptosis-related gene(NLRP1).We constructed a cell death index(CDI)based on 16 cell death genes.2.KM survival analysis in the TCGA HNSCC cohort showed that patients with high CDI had a poorer prognosis(P<0.001),the median survival time of the high CDI group was 20 months,and the 95%confidence interval was 18 to 29 months,while the median survival time of patients in the low CDI group was 110 months,the lower limit of the 95%confidence interval was 90 months,and the upper limit was not obtained.Both cohorts of GSE65858 and GSE42743 and cohort of Qilu Hospital of Shandong University showed that patients with high CDI had a poorer prognosis(median 11 months vs 35 months in cohort of GSE42743,57 months vs 62 months in cohort of GSE65858 and 24 months vs not reached in cohort of Qilu Hospital of Shandong University,respectively).3.Time-dependent ROC curve analysis of the sensitivity and specificity of CDI prediction,the area under the curve of 1-year,3-year,5-year and 10-year were 0.773,0.779,0.739 and 0.791,which was better than age,gender,tumor site,tumor TNM stage,tumor differentiation and HPV status.Multivariate Cox regression analysis showed that CDI was an independent risk factor for overall survival in HNSCC patients,HR was 3.80,95%confidence interval was 2.70-5.40,P<0.001.4.Based on the multivariate Cox regression model,we constructed a nomogram survival prediction model to help predict the 1-year,3-year and 5-year survival rates of HNSCC patients.Among them,CDI was the most predictive factor,followed by age,TNM stage and tumor differentiation.The Calibration curve shows that the consistency index predicted by the model is 0.735.5.Analyze the distribution of CDI and its predictive effect on survival in different clinicopathological subgroups.Except for HPV positive HNSCC,CDI has a good predictive effect on overall survival in different clinicopathological subgroups,patients with high CDI have poorer survival.6.To analyze the predictive effect of CDI on treatment response in HNSCC and HNSCC with different HPV status.Patients with high CDI can benefit overall survival after radiotherapy,while patients with low CDI have no survival benefit after radiotherapy or chemoradiotherapy.However,in HPV negative HNSCC,the low CDI group had higher immune scores and microenvironmental scores,more infiltration of immune cells,the immune checkpoint-related genes were highly expressed,and the main enrichment pathway of the up-regulated genes was the immune response-related pathway in the low CDI group.Conclusions1.We analyzed apoptosis,autophagy,and pyroptosis-related gene expression in HNSCC based on sequencing data from the TCGA database,and constructed a cell death index(CDI).The overall survival of patients with high CDI was poor,and the analysis was validated using HNSCC cohort transcriptome data from Qilu Hospital of Shandong University and cohorts of GSE42743 and GSE65858.The predictive value of CDI on survival is superior to clinicopathological features such as age,gender,tumor site,TNM stage,tumor differentiation and HPV status,and is an independent risk factor for overall survival in HNSCC patients(HR=3.80,95%CI:2.70-5.40,P<0.001).2.We also found a survival benefit after radiotherapy in patients with high CDI in HNSCC and in different subgroups of HPV status.However,in HPV negative HNSCC,the low CDI group showed more immune cell infiltration,high expression of immune checkpoint genes,and up-regulated genes mainly enriched in immune response-related pathways,suggesting that these patients may be more promising to benefit from immune checkpoint therapy.This index may be able to stratify HNSCC patients to select more effective treatment strategies,and has the potential to become a new effective molecular marker for adjuvant clinical treatment options.3.The sensitivity and specificity of the 1-year,3-year,5-year and 10-year survival prediction of the autophagy index were better than those of the apoptosis index and the pyroptosis index,and the impact on survival is higher than that of apoptosis and pyroptosis index.Clinical subgroup analysis showed that the proportion of high CDI in oral squamous cell carcinoma was significantly higher than that in pharyngeal squamous cell carcinoma,and the prognosis of high OSCC was poor.It is suggested that autophagy may become a new research target for clinical treatment of oral squamous cell carcinoma.Part Ⅱ Effects of TLR3 agonist poly(I:C)on oral squamous cell carcinoma cell CAL27 and effects of its conditioned supernatant on macrophage transcription and secretory autophagyResearch contents1.Study the effect of TLR3 agonist poly(I:C)on the cytotoxicity and secretion of inflammatory cytokines in OSCC cellPoly(I:C)acts on the OSCC cell line CAL27 to analyze the effect on cell apoptosis and proliferation;detect the effect on cell autophagy;conduct transcriptome sequencing analysis to study the regulation of cell gene expression;collect cell supernatants to analyze the effect on the role of inflammatory cytokine secretion.2.Autophagy-dependent tumor supernatant regulation of macrophage gene expression and secretory autophagy2.1 Collect the conditioned medium of CAL27 cells from different treatment groups to act on THP-1-derived macrophages,collect macrophage lysates for transcriptome sequencing to analyze the effect of autophagy-dependent tumor supernatant on macrophage gene expression.2.2 Reverse phase protein array(RPPA)technology was used to detect the protein expression of macrophages in different treatment groups,and the effect of secretory autophagy of tumor cells on the expression of macrophage autophagy proteins was analyzed.2.3 Collect the THP-1-derived macrophage supernatants,and detect the effect of autophagy-dependent tumor supernatants on the secretion of inflammatory cytokines from macrophages.2.4 Reactive oxygen scavenger NAC pretreated macrophages,and detected the correlation of poly(I:C)-stimulated CAL27 cell conditioned medium to regulate the macrophage protein and gene expression with the level of reactive oxygen species.Research purposes1.To study the effect of TLR3 agonist poly(I:C)on autophagy and inflammatory cytokine secretion of CAL27 cells.2.The effect of secretory autophagy of CAL27 cells on the expression of macrophage genes and proteins and the secretion of inflammatory cytokines.Materials and methods1.The effect of poly(I:C)on autophagy,apoptosis and proliferation of CAL27 cells.Poly(I:C)(25 μg/mL)was applied to CAL27 cells for 24 hours,and the cell proliferation was measured using Cell Counting Kit 8,apoptosis assay was assessed using the Annexin V-FITC/PI apoptosis detection kit,the expression of autophagy proteins(LC3B and p62)was detected by Western blot,the expression of autophagy genes in CDI was detected by qPCR.2.CAL27 cells were lysed with a sufficient amount of Trizol,and the cell lysate was used for transcriptome sequencing.Analyze the differential gene expression in each group of cells,and perform GO and KEGG enrichment analysis.3.The conditioned supernatant of CAL27 cells with poly(I:C)-treated was collected,and the contents of inflammatory cytokines such as IL-6,CXCL8,CCL3,CSF2 and CSF3 were detected by ELISA detection kit.4.Collect poly(I:C)or PBS-treated CAL27 cell-conditioned supernatants.THP-1 cells were cultured with 100 ng/mL PMA for 6 hours,and CAL27 cell conditioned medium was added at a ratio of 1:1,and cultured for 42 hours.Sufficient Trizol lysed macrophages for transcriptome sequencing analysis.5.Macrophages from different treatment groups were collected using cell-specific lysates for RPPA detection,and detect the expression of 384 protein targets in the cancer signal panorama analysis panel.6.Collect supernatants of macrophages in different groups,and use ELISA kits to detect the contents of IL-6,CXCL8,CCL3,CSF2,and CSF3 in the supernatants.Results1.TLR3 agonist poly(I:C)inhibits the expression of autophagy proteins(LC3B and p62)in CAL27 cells,promotes the expression of autophagy-related genes ATG5,TSG101,VDAC1 and VPS37C,inhibits the expression of PRKN,and induces apoptosis and inhibits cell proliferation.2.Analyze gene expression based on transcriptome sequencing results,and screen differentially expressed genes based on P<0.05 and fold change greater than 1.5 times.GO enrichment analysis showed that genes with increased expression after poly(I:C)treatment were significantly enriched in cytokine response-related biological processes,including:response to cytokine,cytokine mediated signaling pathway,and response to type I interferon.The results of KEGG enrichment analysis showed that the top 10 enriched pathways of genes with increased expression after poly(I:C)treatment included TGF-beta signaling pathway,viral protein interaction with cytokine and cytokine receptor,and IL-17 signaling pathway.3.ELISA results showed that poly(I:C)inhibited the secretion of CXCL8,CSF2 and CSF3 in CAL27 cells.4.Analysis of macrophage transcriptome data showed that the biological processes and pathways of highly expressed genes enriched in the macrophage group of CAL27 cell supernatants treated were mainly related to cytokine response.For example,the top 10 GO-enriched pathways include response to cytokine,cytokine mediated signaling pathway,cytokine production,and responses to interferon gamma,and the top 10 KEGG-enriched pathways include cytokine-cytokine receptor interaction,viral protein interaction with cytokine and cytokine receptor,TNF signaling pathway and chemokine signaling pathway.5.RPPA proteomic analysis showed that the expression of autophagy proteins ATG3,ATG7,TSG101,Beclin-1,and LC3A/B in macrophages treated with CAL27 cell supernatants decreased,however,the expression of ATG3,TSG101,and LC3A/B proteins in the group treated with poly(I:C)-treated CAL27 cell supernatants was higher than that in the group treated with CAL27 cell supernatants.6.ELISA test found that the supernatant of CAL27 cells promoted macrophages to secrete IL-6,CXCL8,CCL3,CSF2,CSF3.Poly(I:C)further promote the secretion of IL-6,CXCL8,CCL3 and CSF2.7.Poly(I:C)-treated CAL27 cell conditioned supernatant acted on macrophages.After pretreatment with reactive oxygen species scavenger NAC,the expressions of autophagy proteins ATG4B,ATG5,ATG7,TSG101 and p62 in macrophages decreased,and the expression of LC3A/B increased,decreased secretion of inflammatory cytokines IL-6,CXCL8,CCL3,CSF2 and CSF3.Conclusions1.We used transcriptome sequencing technology to analyze the changes in gene expression of OSCC cell CAL27 treated with poly(I:C)and the effect of conditioned supernatant of CAL27 cells on the gene expression of macrophages.The obtained sequence information and expression information of almost all transcripts in different autophagic states of tumors and macrophages provide a complete gene expression profile for subsequent analysis of gene expression and screening of molecular markers.2.To study the regulation of secretory autophagy of tumor cells on macrophage gene expression and protein secretion in OSCC,and detect 384 tumor signaling target proteins based on RPPA technology,including drug targets and important upstream and downstream targets of cell signaling pathways,etc.Provide accurate protein target expression information for the study of tumor microenvironment and tumor therapy response,and promote the research progress of secretory autophagy.Part Ⅲ Analysis of autophagy-related genes in CDI on the prognosis and treatment response of OSCCResearch contents1.Autophagy-related gene expression and survival correlation analysis in OSCC1.1 In the OSCC cohorts of TCGA database and Qilu Hospital of Shandong University,the expression of 10 autophagy-related genes in the cell death index in different clinical subgroups of OSCC was analyzed.1.2 Univariate and multivariate Cox regression models were used to analyze the correlation between autophagy-related genes and OSCC survival and obtain survival-related autophagy genes,and analyze the predictive effect of these genes on treatment response and their correlation with immune cell infiltration and immune checkpoint gene expression.2.The predictive effect of TSG101 gene expression on OSCC survival and treatment response2.1 The macrophage markers CD68 and SPP1 were selected for macrophage infiltration ratio grouping,and the survival difference of OSCC patients with high and low autophagy gene TSG101 expression was analyzed between the two groups with high and low macrophage infiltration.2.2 Analyze the correlation between TSG101 expression and OSCC response to radiotherapy or chemoradiotherapy in the two groups with high and low expression of macrophage marker genes.Research purposes1.To analyze the expression profile of 10 autophagy-related genes in CDI in OSCC and their predictive value for survival and treatment response.2.Evaluate the differences in survival and treatment responsiveness of patients with different macrophage markers CD68 and SPP1 and TSG101 gene expression.Materials and methods1.Using the clinicopathological and gene expression data of the HNSCC cohorts of TCGA database and Qilu Hospital of Shandong University,select cases whose tumor site belongs to the oral cavity,and obtain OSCC cohorts containing 255 and 24 patients,respectively.2.Paraffin tissue sections were stained for p16 by immunohistochemistry.The p16 score was assessed based on staining intensity(0,1,2,3)and positive range(0,5%,25%,50%,and 75%).A staining intensity score of 3 and a positive proportion greater than 25%or a moderate staining intensity(intensity score of 2)and a positive range greater than 75%were defined as p16-positive.3.Analyze the expression of 10 autophagy-related genes in CDI in clinicopathological subgroups such as age,gender,smoking,drinking,tumor stage,tumor differentiation degree,HPV status,etc.Statistical analysis was performed using the Wilcox test.4.Using univariate and multivariate Cox regression models to analyze the relationship between 10 autophagy-related genes and OSCC survival.5.The CIBERSORT database was used to analyze the infiltration of immune cells,and the expression of immune checkpoint genes was analyzed based on RNA-seq data.6.Download the gene list of Cytokine-cytokine receptor interaction(hsa04060)from the KEGG database,and analyze the differences in the expression of cytokine response genes between the high and low expression of OSCC prognosis-related genes ATG5 and VDAC1.7.Based on the radiotherapy and drug treatment response data in the TCGA database,the differences in treatment response between the two groups with ATG5 and VDAC1 expression were analyzed.8.The patients were divided into different subgroups according to the median FPKM values of macrophage marker genes(CD68 and SPP1)and autophagy-related gene TSG101,respectively,and the R package survminer was used to analyze the relationship between TSG101 expression and OSCC survival and treatment response in the two groups with high and low CD68 and SPP1 gene expression using Kaplan-Meier survival curves and log-rank test,and the difference in 3-year survival rate was analyzed in the OSCC cohort of Qilu Hospital of Shandong University.Results1.Autophagy-related gene expression analysis of the OSCC cohort showed that male patients highly expressed CSNK2A2,low expression of TSG101 in smoking subgroup,high expression of MVB12B and low expression of ATG5 in stage Ⅰ-Ⅱ OSCC,high expression of MVB12B and DYNC1I1 in poorly differentiated group,high expression of TSG101 gene in HPV positive patients.2.Univariate Cox regression analysis showed that high expression of ATGS and VDAC1 and Ⅲ-Ⅳ stage were risk factors for OSCC survival.The HR for the ATG5 expression was 1.70,95%confidence interval(1.10-2.60),P=0.014.HR of VDAC1 expression was 1.80,95%confidence interval(1.20-2.70),P=0.005.The HR for stage Ⅲ-Ⅳ versus stage Ⅰ-Ⅱ was 2.00,with a 95%confidence interval(1.20-3.30),P=0.004.Multivariate Cox regression analysis of expression of ATG5 and VDAC1 and tumor TNM stage found that all three were independent risk factors for survival,with HR of 1.60,1.90 and 2.00,respectively.3.We further analyzed the relationship between the gene expression of ATG5 and VDAC1 and the response to radiotherapy and chemotherapy,and the results showed that the expression of VDAC1 in the radiotherapy response group of CR and PR was significantly lower than that of the radiotherapy response group of SD and PD groups,suggesting that patients with low VDAC1 expression were radiosensitive higher.4.Immune analysis found that patients with high ATG5 expression had less infiltration of CD8 T cells,activated memory CD4 T cells,follicular helper T cells,and regulatory T cells,and more infiltration of resting memory CD4 T cells.In OSCC with high expression of VDAC1,the infiltration of plasma cells,regulatory T cells,and M0 macrophages was less,and the infiltration of activated memory CD4 T cells,resting NK cells,and M1 macrophages was more.Expression of immune checkpoint genes of CD24 and PD1 was lower in OSCC patients with high ATG5 expression.Comparative analysis between the two groups with high and low VDAC1 expression found that OSCC patients with high VDAC1 expression had high expression of immune checkpoint genes such as CD47,IDO1,LAG3,PD-1,PD-L1,and PD-L2,and low expression of CD24.5.Analysis of the OSCC cohort in the TCGA database found that SPP/high-TSG101low OSCC patients have a better prognosis,and benefit from radiotherapy.Valida... |
PDF Full Text Request