Construction Of A Pancreatic Cancer Prognosis Model Based On Tumor Metastasis-related Genes And The Regulation Of RNF138 On The Malignant Phenotype Of Pancreatic Cancer And Its Molecular Mechanism

Part I:Development and validation of a metastasis-related gene signature for predicting the overall survival in patients with pancreatic ductal adenocarcinomaBackgroundPancreatic ductal adenocarcinoma(PDAC)is highly malignant with a very poor prognosis.Local invasion and distant metastasis frequently occur in the early stage.The highly invasive and metastatic characteristics of PDAC cells are one of the key factors result in the poor prognosis of PDAC.However,the molecular mechanism of the invasion and metastasis of PDAC has not yet been fully elucidated.In-depth study of the molecular mechanism of metastasis and metastasis-related genes(DE-MTGs)in PDAC is expected to elucidate its high invasive and metastatic characteristics,and provide new targets for molecular therapy.In addition,DE-MTGs can directly reflect the metastatic characteristics of PDAC,and have important potential to predict the prognosis of PDAC patients.There is an urgent need for an evaluation system that can more effectively and individually predict the prognosis of PDAC patients to assist clinical decision-making.Currently,staging systems based on clinical and pathological parameters are commonly used in the evaluation of PDAC prognosis.However,the traditional staging system cannot dynamically evaluate the progression of PDAC,nor can it reflect the malignant biological behavior of the patient's tumor on the molecular level.Therefore,it is necessary to establish novel models that can more accurately and individually assess the prognosis of patients with PDAC.In recent years,with the continuous accumulation of cancer genomics data,the use of cancer gene transcriptome data,combined with effective bioinformatics analysis methods,to establish gene signature has been proven to be an effective alternative in the field of clinical evaluation.In view of the correlation between MTGs and the malignant biological behavior and prognosis of PDAC,the identification of prognosis related MTGs in PDAC and the establishment of a gene signature for predicting the prognosis may be an effective strategy for the development of a more accurate prognostic evaluation system.ObjectiveThe main purpose of this part of the study is to identify the differentially expressed MTGs(DE-MTGs)that affect the prognosis of PDAC based on transcriptome and clinical data,Key DE-MTGs was further screened to establish a predictive gene signature.e prognostic-related clinical pathological parameters were further integrated to establish a better prognostic evaluation system for PDAC.MethodsIn this part of the study,in order to obtain reliable and representative expression profile data of PDAC,we first downloaded and integrated 4 PDAC chip expression data sets based on the GEO database,GSE15471,GSE16515,GSE32676 And GSE22780,and used R language to process the data to identify differentially expressed genes in PDAC.We further obtained a list of MTGs from the human cancer metastasis gene database.Their expression levels in PDAC tissues was analyzed to obtain DE-MTGs in PDAC We further screened the prognostic-related DE-MTGs based on survival analysis in the PDAC patient cohort of the TCGA-PAAD dataset.Key DE-MTGs were further selected based on LASSO regression and a gene signature was established.We use Xtile software to determine the best intermediate value of risk score to divide PDAC patients into high-risk groups and low-risk groups.The log-rank test in Kaplan-Meier survival analysis was used to test whether there are statistical differences in overall survival between different risk groups.The R language software package timeROC was used to construct a time-dependent receiver operating characteristic curve(ROC)curve,determine the area under the curve(AUC),and predict the 1,2,3-year overall survival rate of PDAC patients.And further calculate the consistency index C-index to evaluate the prediction effect of the model.Use the method described by Delong et al.to determine whether there are statistical differences between AUCs.Perform external verification in the external data set GSE62452 and PACA-AU to evaluate the prognostic prediction effect of the gene prognosis model in the external verification data set.We also used gene set enrichment analysis(GSEA)to further explore the potential molecular mechanisms of DE-MTGs gene model and poor prognosis of PDAC.Using ESTIMATE and CIBERSORT bioinformatics algorithms,we used transcriptome sequencing data to assess PDAC tissues.Immune cell infiltration,and Pearson correlation test was used to analyze the correlation between the expression level of DE-MTGs and immune cell infiltration.ResultsIn this part of the study,in order to obtain reliable and representative expression profile data in PDAC gene chip samples,we first downloaded and integrated 4 data sets of PDAC from the GEO database,GSE15471,GSE16515,GSE32676 And GSE22780,and used R language to process the data to identify differentially expressed genes in PDAC.We further obtained a list of metastasis-related genes from the human cancer metastasis gene database,analyzed their expression levels in PDAC tissues,and intersected the results of differential analysis to obtain differentially expressed metastasis-related genes(DE-MTGs)in PDAC.Bioinformatics Analysis was performed on the DE-MTGs.We further screened the prognostic-related DE-MTGs based on survival analysis in the pancreatic ductal adenocarcinoma patient cohort of the TCGA-PAAD dataset,and screened key genes based on LASSO regression,A gene signature was further established and the risk scores were calculated.We then used Xtile software to determine the best cutoff value of the risk score to divide PDAC patients into high-risk and low-risk groups.The log-rank test in Kaplan-Meier survival analysis was used to test whether there are statistical differences in overall survival between groups of different risk.The R language software package timeROC was used to construct time-dependent receiver operating characteristic curves(ROC),determine the area under the curve(AUC),and predict the 1,2,3-year overall survival of PDAC patients.Concordance index was further calculated to evaluate the prediction effect of the gene signature.The method described by Delong et al.was used to determine whether there are statistical differences between AUCs.External verification was performed in the external data sets GSE62452 and PACA-AU to evaluate the prognostic prediction effect of the gene signature.We also used gene set enrichment analysis(GSEA)to further explore the potential molecular mechanisms underlying the DE-MTGs related to the poor prognosis of PDAC.ESTIMATE and CIBERSORT bioinformatics algorithms were used to assess Immune cell infiltration of PDAC tissues based on transcriptome sequencing data.And Pearson correlation test was used to analyze the correlation between the expression of DE-MTGs and immune cell infiltration levels.ConclusionsIn this part of the study,the DE-MTGs we identified are closely related to the progression and prognosis of PDAC and are potential therapeutic targets.The genetic model and nomogram based on DE-MTGs have better prognostic prediction performance than traditional prognostic evaluation systems,and has great potential in the assistance of clinical decision-making.Part ?:The investigation of RNF138 on regulating pancreatic cancer malignant phenotype and associated molecular mechanismBackgroundPancreatic ductal adenocarcinoma(PDAC)is a highly malignant gastrointestinal tumor,characterized by poor prognosis and early invasion and metastasis.To date,surgical resection remains the only potentially curative treatment,yet only 15 to 20 percent of patients are candidates for pancreatectomy due to the late presentation of the disease.Chemotherapies lack significant benefit on the median overall survival of locally advanced or metastatic pancreatic cancer,which is less than one year.Recently,remodeling the immune system to eradicate tumour cells has been an area of extensive research.Despite clinical success of immunotherapy in other metastatic diseases such as lung cancer and melanoma,the application of immune checkpoint inhibitors(ICIs)in pancreatic cancer gave disappointing results.Ring finger protein 138(RNF138),a E3 ligase,is highly expressed in tumour compared to normal tissue.RNF138 possibly promotes tumorigenesis through regulation of cellular processes and exerted an immunosuppressive role on the tumor microenvironment(TME).However,the role of RNF138 in tumorigenesis and TME of pancreatic cancer has not been reported.This study suggests a strong association between high expression of RNF138 and PDAC.Therefore,investigating the pro-tumorigenic role and downstream molecular mechanism of RNF138 in PDAC may help elucidate the reasons behind high malignancy,provide new therapeutic targets,improve efficacy of ICIs and prolong overall survival.ObjectiveTo elucidate the expression of RNF138 in PDAC versus normal pancreatic tissue,and the prognostic value of RNF138 expression;to investigate the role of RNF138 and its downstream pathway in mediating the highly metastatic,proliferating,and immunosuppressive properties of PDAC.MethodsExpression of RNF138 was evaluated by immunohistochemistry(IHC)staining in tissue samples from PDAC,matched normal adjacent tissue,and normal pancreas.The correlations between RNF138 expression and clinicopathological factors,immune cell infiltration,and overall survival were evaluated to elucidate the prognostic value of RNF138.Protein expression of RNF138 was assessed in five pancreatic cancer cell lines and one normal pancreas cell line through western blotting and quantitative real-time polymerase chain reaction(qRT-PCR).The effect of RNF138 on cell proliferation was preliminarily evaluated with small interfering RNA(siRNA)knockdown of RNF138 in the PANC-1 cell line.RNF138 knockdown PANC-1 and BXPC-3 cell lines were established with lentivirus mediated short hairpin RNA transfection.Cell motility,invasion and metastasis,colony formation,proliferation,apoptosis,and cell cycle were evaluated through scratch assay,transwell assay,colony formation assay,cell counting kit-8(CCK8),and flow cytometry(FACS)cell cycle and apoptosis analysis,respectively.RNF138KD PANC-1 was subcutaneously implanted into adaptive immune deficient nude mice to investigate tumorigenesis properties in vivo.RNA-sequencing and ingenuity pathway analysis(IPA)and Gene Ontology(GO)enrichment analysis compared transcriptome of wildtype and RNF138KD PANC-1 and suggested key downstream signaling pathways through which RNF138 exerts its tumorigenic effects.Change in protein expression of STAT1 and activation of IFNy related pathway in RNF138KD PANC-1 was confirmed through western blotting and qRT-PCR.Re-expression of wildtype RNF138 in RNF138 KD PANC-1 through plasmid transfection rescued effects of RNF138KD.Co-immunoprecipitation(Co-IP)of RNF138-FLAG and immunofluorescence(IF)co-localization confirmed molecular interaction between RNF138 and STAT1.Six plasmids containing loss of function mutation in particular functional domain were constructed and transfected into RNF138KD PANC-1,to investigate the molecular mechanism of the RNF138-STAT1 interaction.Statistical methods included the Cox proportional hazards regression analysis,Mann-Whitney U test,Kaplan-Meier survival analysis,Pearson chi-square test and Student's t test.P values of less than 0.05 were considered statistically significant.Results1.The prognostic value of RNF138 in PDAC.IHC demonstrated that RNF138 expression is significantly higher in PD AC tissues compared with matched normal adjacent tissue and normal pancreas tissue(P<0.005).The expression level of RNF138 is positively correlated with TNM staging and grading(P<0.05).Univariate survival analysis suggested correlation between RNF138 expression level and prognosis(P<0.05),and overall survival of the RNF138 high subgroup is significantly shorter than RNF138 low subgroup(P<0.05).Multivariate survival analysis confirmed that high RNF138 expression is an independent risk factor of worse outcome(P=0.039,HR=1.855).High expression of RNF138 is also negatively correlated with CD8+T cell infiltration and PD-L1 expression(P<0.05).2.RNF138 regulation of malignant phenotype of PD AC.RNF138 expression is high in pancreatic cell lines but low in normal pancreatic cells.Transient knockdown of RNF138 arrested cell cycle in S phase,resulting in increased G1 and S but decreased G2/M(P<0.05).In PANC-1 and BXPC-3 cell lines,stable knockdown of RNF138 significantly inhibited proliferation and colony formation,arrested cell cycle,and promoted apoptosis.Cell motility,invasion and metastasis were not affected(P>0.05).Interestingly,lowered expression of RNF138 PANC-1 potently inhibits in vivo growth,with no tumor formation upon subcutaneous injection into T cell deficient nude Balb/c mice.3.The molecular mechanism of RNF138 in regulating phenotypes of PDAC.RNA-sequencing and ingenuity pathway analysis(IPA)compared the transcriptomes of wildtype and RNF138KD PANC-1 cells.Knockdown of RNF138 significantly increased expression of STAT1 and genes that were mapped onto activation of IFNy pathway(P<0.05).In PANC-1 and BXPC-3 cell lines,knockdown of RNF138 led to increased STAT1 and IRP1 protein expression,increased activating phosphorylation of STAT1 at 70 1y,and increased mRNA expression of a of genes downstream of IFNy signaling,including ISGF3,IRF1,CXCL9,CXCL10,HLADRA,IDO1,and IFNg.Re-expression of wildtype RNF138 in RNF138KD PANC-1 cells lowered expression of STAT1 and interferon-stimulated genes.Co-immunoprecipitation and immunofluorescence double staining demonstrated colocalization and interaction of RNF138 and STAT1.The above results suggested that ubiquitination and proteasomal degradation of STAT1 and inactivation of the downstream IFN? pathway mediate the malignant and immunosuppressive phenotype of RNF138 high PDAC.Further re-expressing domain-specific mutated RNF138 in RNF138KD PANC-1 cells revealed two functional domains,UIM and RING,are crucial for interaction with STAT1 and its ubiquitination and degradation.Conclusion1.The expression of RNF138 is significantly higher in PDAC tissue compared with matched normal adjacent tissue.High expression of RNF138 in PDAC is an independent risk factor of prognosis,negatively correlated with immune cell infiltration and overall survival.2.High expression of RNF138 promotes proliferation and inhibits apoptosis in PD AC cell lines.Expression of RNF138 is critical for tumorigenesis in vivo.3.Knockdown of RNF138 expression results in increased STAT1 protein expression and activated IFN downstream signaling pathway.C-myc induces RNF138 mRNA transcription and protein expression.RNF138 can bind to p53 and mediate the inhibition of STAT1 and IFN downstream signaling pathway.RNF138 relies on its ubiquitinated ligase domain to regulate STAT1 and IFN signaling pathways.In the PDX pancreatic cancer model with humanized immune system,RNF138 siRNA combined with PD-1 monoclonal antibody can increase the infiltration of CD8+cells and other immune cells in the tumor microenvironment,and effectively inhibit tumor growth.