1.DNA Copy Number Variations As Biomarkers For Prognosis Prediction Of Bladder Cancer 2.The Molecular Functions Of Virus-encoded MiRNAs In EBV-associated Epithelial Cancers

This PhD project consists of two parts of relatively independent research work.Chapter 1 DNA copy number variations as biomarkers for prognosis prediction of bladder cancerBladder cancer(BC)has been a debilitating and costly disease for patients and their families.The high rate of recurrence for non-muscle-invasive bladder cancer(NMIBC)and the high risk of distant metastases for muscle-invasive bladder cancer(MIBC)made it urgent to find effective signatures for prognosis.Various biomarkers have been recently developed for detection,but there has not been any for prognosis evaluation.A number of genomic alterations occur in the carcinogenesis of bladder cancer,yet none of them has been referred in the current risk group stratifications.In this study,we aim to find significant copy number variations(CNVs)for prognosis of bladder cancer by exploring from tumor tissue to urine sediment.In this study,the array-based comparative genomic hybridization on 65 bladder caner samples were previously performed for discovering significant copy number variations in bladder cancer tumor tissue.CNVs of 5 candidates(CEP63,FOSL2,GHR,PAQR6,ZFAND3)and 1 reference gene(TBP)were validated by real-time PCR.In this study,we analyzed the correlations between identified CNVs and clinicopathological parameters of 219 patients.Prognostic value of identified CNVs had been evaluated in predicting disease-free survival(DFS)of NMIBC and MIBC patients using Kaplan-Meier analysis,Cox proportional hazards regression and Logistic regression.The identified CNVs were further tested in BC urine sediment to explore their potential for clinical application.Real-time PCR(n=123)and 3D digital PCR(n=40)were applied to detect identified CNVs in urine for assessing their correlations with clinicopathological parameters and evaluating their prognostic value for DFS.We found that copy numbers of CEP63,FOSL2 and PAQR6 were significantly gained(all p<0.01)in 219 bladder cancer tumor samples.CNVs of CEP63 showed significant correlations with tumor stage(p<0.01),high grade(p<0.01)and lymph node metastasis(p<0.01).CNVs of FOSL2 were associated with tumor stage(p<0.01)and grade(p=0.047).The median follow-up time for all enrolled patients was 61 months(range,3 to 188 months).Overall disease-free survival at 5 and 10 years for the entire cohort was 48.3%and 22.7%,respectively.Multivariate Cox regression analysis indicated that copy number variations of CEP63(p=0.002),FOSL2(p=0.004)in NMIBC patients and FOSL2(p=0.022),PAQR6(p=0.024)in MIBC patients were independent prognostic factors for DFS.The prognostic indexes(PIs)for recurrence hazards were calculated by the prediction models for NMIBC and MIBC,respectively(NMIBC:PI=1.5095× CEP63+1.47123×FOSL2;MIBC:PI=2.0440×FOSL2+2.2079×PAQR6).NMIBC and MIBC patients with higher PI tended to have poorer DFS(p=0.0003 and p=0.0001,respectively).NMIBC Patients with high PIs tended to have poor DFS(p=0.00056),indicating that prognostic index can help to identify those with worse outcomes among high risk NMIBC patients.For exploring clinical application value,we tested CNVs of CEP63,FOSL2 and PAQR6 in urine sediment of 123 BC patients.Results of real-time PCR showed that copy numbers of CEP63 and FOSL2 significantly gained in BC patients' urine samples(p=0.036 and p<0.01,respectively).The innovative high-sensitivity 3D digital PCR further revealed the correlations between CNVs of CEP63,FOSL2 and tumor stage in urine sediment.Copy numbers of CEP63 and FOSL2 in urine were significantly higher in MIBC patients than those in NMIBC patients(p<0.01 and p=0.046,respectively).Copy number gains of CEP63 and FOSL2 in urine sediment were found related to poorer DFS(p<0.01 and p=0.018,respectively)for NMIBC patients.A model based on CNVs of CEP63 and FOSL2 in NMIBC urine was constructed to predict the hazards of recurrence within 1 year.Patients with high prognostic index according to this predition model were under recurrence risk 3.594 times(p=0.002)higher than patients with low prognostic index.To compare the efficiency on detecting 2 identified CNVs by real-time PCR and 3D digital PCR,ROC analysis was applied on urine samples tested by both methods.The sensitivity(0.750 and 0.650,respectively),negative predictive value(0.786 and 0.708,respectively)and AUCs(0.848 and 0.803,respectively)of 3D digital PCR on detecting CNVs of CEP63 and FOSL2 were all higher than the sensitivity(0.475 and 0.425,respectively),negative predictive value(0.644 and 0.646,respectively)and AUCs(0.601 and 0.685,respectively)of real-time PCR.A diagnostic model based on CNVs of CEP63 and FOSL2 detected by 3D digital PCR was constructed with AUC of 0.858,sensitivity of 85.0%and negative predictive value of 83.8%.In conclusion,our findings indicated that CNVs of CEP63,FOSL2 and PAQR6 had considerable clinical value in prognosis of BC.Long term follow-up revealed that patients with copy number gains of 3 identified genes in tumor had worse prognosis.Prognostic value of CEP63 and FOSL2 CNVs had been further validated in BC urine sediment.We have done a pioneer research by applying the novel high sensitive 3D digital PCR for detecting BC urinary DNA CNVs,which was in prospect of providing a new method for urine-based molecular detection.Chapter 2 The molecular functions of virus-encoded miRNAs in EBV-associated epithelial cancersIt is more than 50 years since Epstein-Barr virus(EBV),the first human oncogenic virus,was discovered.EBV,with the most common and persistent infection in humans,and roughly 95%of the world's populations sustaining an asymptomatic life-long infection,mainly infects lymphocytes and oropharyngeal epithelial cells.EBV has subsequently been found to be associated with a diverse range of neoplasms,such as Burkitt lymphoma(BL),Hodgkin lymphoma(HL),lymphoproliferative disorders(PTLD),AIDs-related non-Hodgkin lymphoma,diffuse large B cell lymphoma(DLBCL),NK/T cell lymphoma,nasopharyngeal carcinoma(NPC)and EBV associated gastric carcinoma(EBVaGC).Several studies have investigated the roles played by EBV in carcinogenesis and have shown that EBV gene transcripts,including EBV nuclear antigen 2(EBNA2)and latent membrane protein 1(LMP1)can disturb signal transduction in host cells,causing immortalization or transformation.Although these studies have shed light on the oncogenic properties of EBV,many unsolved problems still remained.miRNAs play important roles in many complex biological processes,including cellular development and differentiation.An increasing number of studies have found that dysregulation of certain miRNAs induces carcinogenesis in various organs.EBV was the first virus found to encode miRNAs,which expressed in infected B cells and some EBV-associated neoplasms.Several studies have profiled EBV-encoded miRNA expression in EB V-associated malignancies and have revealed that specific viral miRNAs play roles in carcinogenesis.Although several EBV encoded miRNAs had been studied for their functions in tumorigenesis,the majority of viral miRNAs were still being unknown.In this study,we comprehensively profile the expression of 44 known EBV-encoded miRNAs in clinical samples from 8 gastric carcinoma patients and 33 nasopharyngeal carcinoma patients by miRNA microarray and next-generation sequencing.EBV associated gastric carcinoma only occupied 6%of gastric cancer in Asian patients.Since EBER-ISH was not the rutine test for gastric patients,the amount of EBV associated gastric carcinoma samples was limited.Up till now,there has not been any research based on fresh-frozen EBV associated gastric carcinoma tissue samples reported in China.In this research,we have found 4 EBV-encoded miRNAs highly expressed in EBV associated gastric carcinoma tumor tissue and intrestingly significantly correlated with 4 host miRNA,with correlation coefficient higher than 0.9.Surprisingly,2 host miRNAs shared the exact same 5' seed sequences with 2 EBV encoded miRNAs.The host target genes of this 4 EBV encoded miRNAs were predicted and were enriched in the cancer related biological process,such as cell migration,cell cyle regulation and DNA replication.Since there were so many predicted target genes of host miRNAs and EBV encoded miRNAs,we aimed to find the common target genes which were differently expressed in EBV associated gastric carcinoma tumor tissue.Therefor,we sequently analyze the differently expressed host genes not only in 8 samples of our study but also in 26 EBV associated gastric carcinomas in The Cancer Genome Atlas and 12 EBV associated gastric carcinomas in Gene Expression Omnibus.We have found 994 differently espressed host genes in the all 46 EBV associated gastric carcinomas.By intersect the the target genes of host and EBVmiRNAs and the differently expressed genes in EBV associated gastric carcinoma,we have got the common host target genes of host and EBV miRNAs which were also differently expressed in EBV associated gastric carcinoma.The KEGG signal pathway analysis of those host target genes had revealed the possible cell signal pathways in which both host miRNA and related EBV miRNAs were involved.The common cell signal pathway of the host miRNAs and related EBV miRNAs were“Signaling pathways regulating pluripotency of stem cells" and "Proteoglycans in cancer",both wich were related with cell suvival.The in vitro research futher confirmed the biological functions of miR-BART1-3p and miR-BART7-3p.The results of cell counting Kit-8(CCK8)analysis,transwell migration assays demonstrated that both EBV encoded miRNAs could inhibit the proliferation and migration of gastric cancer cells.The flow cytometric analysis showed that the two EBV miRNAs could induce cell cycle arrest in G0G1phases.We have also investigated the biological functions of miR-BART7-3p which expressed higher than miR-BART1-3p in NPC.It is found that miR-BART7-3p can also inhibit proliferation and migration in NPC cells.In conclusion,we have found that miR-BART1-3p and miR-BART7-3p could inhibit cell growth in EBV associated gastric carcinoma.The possible mechanism of negative regulation by the EBV-encoded miRNAs was illustrated in this study and may helped to further understand the virus' role of oncogenesis in EBV associated epithelial carcinoma.