Identification Of Common Key Gene For Cisplatin Acquired Resistance And Pulmonary Metastasis In Osteosarcoma By Integrated Bioinformatics Analysis And Experimental Study

Objective and significance:Osteosarcoma(OS)is the most common bone malignant tumor in children and adolescents.Chemotherapy resistance and pulmonary metastasis are important factors for poor prognosis in patients with OS.The current treatment dilemma for OS is that the 5-year survival rate is below 25%for patients between 5-26 years which with poor initial response to chemotherapy.Some serious side effects of chemotherapeutics and the primary or acquired resistance to treatment have brought great difficulties and challenges in clinical treatment.In addition,the prognosis of OS patients with metastatic tumors is significantly poorer than that of patients with primary tumors.The Overall survival rate of patients with metastatic or recurrent OS has remained almost unchanged over the past 30 years.Therefore,great importance should be attached to explore biomarkers and related regulatory pathways for the pathogenesis,metastasis,chemotherapy resistance and recurrence in OS.Methods:Chapter 1 Identification of key genes and construction of gene expression signature in osteosarcoma with acquired cisplatin resistanceThe Limma package of R language was used to screen the OS with acquired cisplatin resistance dataset GSE86053 obtained from the GEO database to screen differentially expressed genes.iDEP.90 was used for hierarchical clustering analysis,and DAVID was used for GO and KEGG analysis.Cytoscape visualized and screened the Hub gene,Followed by Kaplan-Meier plotter for survival analysis,and SurvExpress for gene expression signature model construction.Finally,key genes were verified by GEPIA and CCLE analysis.Chapter 2 Construction of weighted gene co-expression modules for osteosarcoma acquired cisplatin resistance,Gene set enrichment analysis and gene expression signature in osteosarcoma with acquired cisplatin resistanceThe R language WGCNA package was used to construct a weighted gene coexpression network for OS with acquired cisplatin resistance dataset in GSE86053,which obtained from the GEO database.and screened key module.STRING was used to construct a PPI network and Cytoscape was used to visualize and screen the hub gene.Then GSEA software was used for Gene set enrichment analysis,and key pathways and hub genes were screened.Then Kaplan-Meier plotter was used for survival analysis,and SurvExpress was used for gene expression signature model construction.Finally,key genes were verified by GEPIA and CCLE analysis.Chapter 3 Identification of key genes and construction of gene expression signature in osteosarcoma with pulmonary metastasisGEO2R was used to pre-process the datasets GSE85537 and GSE37552 both related with pulmonary metastasis in OS.We used Venn diagrams to take intersections to screen differentially expressed genes.At the same time,DAVID was used for GO and KEGG analysis.STRING was used to construct a PPI network and Cytoscape was used to visualize and screen the hub.Genes were subsequently analyzed by KaplanMeier plotter for survival analysis,and SurvExpress was used to construct gene expression signature models.Finally,key genes were verified by GEPIA and CCLE analysis.Chapter 4 Preliminary study of POSTN as a potential biomarker of acquired cisplatin resistance and pulmonary metastasis in osteosarcoma and study of POSTN overexpression stimulating cisplatin resistance and metastasis in osteosarcomaThe Hub gene related with pulmonary metastasis in OS dataset was used as a training set.The GSE86053 dataset related with acquired cisplatin resistance in OS was used as a test set for verification the training set.And then,multiple tumor database validation was performed on the screened common key gene,POSTN.Finally,to confirm the role of POSTN in OS,first POSTN in paired samples were assessed using a quantitative real-time PCR(qRT-PCR)assay from OS patients.Second,we applied lentivirus system to overexpression and knockdown of POSTN,respectively,in MG63 OS cell line.Results:Chapter 110 hub genes were screened from GSE86053 dataset related with acquired cisplatin resistance in OS by cluster analysis.Overexpression of APP,LAMB1 and KTN1 was associated with poor prognosis.The constructed gene expression signature model showed that 10 hub genes could stratify OS patients into two risk groups.Overexpression of LAMB1 and BMP4 is related to the poor prognosis of high-risk OS patients,and overexpression of APP is associated with low risk.Patients with OS are associated with a poor prognosis.Validated by the GEPIA database and CCLE database,KTN1 was overexpressed in both tumor tissue and OS cells.Chapter 2Four hub genes(VEGFA,H2AFX,SEMA3F,and ASNS)in black module was screened from GSE86053 data set related with acquired cisplatin resistance in OS through weighted gene co-expression network analysis.All of these genes are related to poor prognosis.Gene set enrichment analysis revealed that functional genes were mainly enriched in two pathways:ECM receptor interaction and Graft versus host disease.The four Hub genes LAMB1,COL 11 A,COL5A3 and HMMR are all related to the poor prognosis of patients with OS.The constructed gene expression signature model showed that the eight hub genes could stratify OS patients into two risk groups.Overexpression of LAMB1,COL5A3,and ASNS was significantly associated with the poor prognosis in the high-risk group.SEMA3F overexpression is associated with poor prognosis in patients at low risk.Validated by the GEPIA database and CCLE database,COL11A1 and COL5A3 were overexpressed in both tumor tissue and OS cells.Chapter 310 hub genes related with pulmonary metastasis in OS were screened from data sets GSE85537 and GSE37552,of which overexpression of MET,Postn,and COL1A2 was associated with poor prognosis in patients.With 10 hub genes were used to construct gene expression signature models,showing that OS patients can be stratified by high and low risk groups.The overexpression of DCN is significantly related to the poor prognosis of patients at high risk.The poor prognosis of patients in the low-risk group was significantly correlated.Validated by GEPIA database and CCLE database,COL1A2,POSTN,DCN,and VEGFC were overexpressed in tumor tissue and OS cells.Chapter 4The Hub gene selected from the dataset related with pulmonary metastasis in OS was used as a training set,and the dataset GSE86053 related with cisplatin acquired resistance in OS was used as a test set for verification.It was found that POSTN as a common key gene was involved in two processes,including cisplatin acquired resistance and pulmonary metastases.Multiple database revealed that POSTN was overexpressed in both OS tissue and cells line.Especially,the overexpression of POSTN mainly occurs in telangiectasia OS.And it is related to the poor prognosis of OS patients.Clinical samples found that POSTN was overexpression in human OS tissue,.Chemotherapy can increase POSTN expression in human OS tissue.In vitro experiments found that overexpression of POSTN increased resistance to cisplatin chemotherapy and promoted migration and invasion in vitro.Conclusions:KTN1,COL 11A1 and COL5A3 can be potential biomarkers of OS with acquired cisplatin resistance,and COL1A2,POSTN,DCN and VEGFC can be potential biomarkers of OS with pulmonary metastasis.Overexpression of POSTN promotes the invasion,metastasis and cisplatin resistance of osteosarcoma.POSTN can serve as a potential biomarker and therapeutic target for OS with acquired cisplatin resistance and pulmonary metastasis.In addition,POSTN can also be a potential biomarker for telangiectatic osteosarcoma,a rare type of OS.These results will help the research progress of OS.