| Background and objectiveOsteosarcoma(OS)is the most common malignancy of the motor system,occurring primarily in children and adolescents,accounting for approximately 60% of all childhood malignancies and the second leading cause of cancer-related death in adolescents.Micrometastases are present in more than 50% of patients with osteosarcoma at diagnosis,which is one of the key factors leading to treatment failure.With the extensive use of neoadjuvant chemotherapy,the prognosis of patients with localized osteosarcoma has been significantly improved,with the 5-year survival rate reaching 60%-70%.However,the prognosis for patients with distant metastases remains poor,with 5-year survival rates hovering around 15%-30%.Elucidate the cellular and molecular mechanisms of early micrometastases of osteosarcoma,which will help to provide a new direction for intervention and treatment of osteosarcoma from a systematic perspective.In recent years,tumor heterogeneity has been proved to be an important factor affecting tumor progression and treatment effect,and has been widely studied in a variety of solid tumors such as liver cancer,breast cancer,colorectal cancer and kidney cancer.As a highly malignant tumor,osteosarcoma also has significant heterogeneity.However,studies on the effect of tumor heterogeneity on micrometastases of osteosarcoma are still limited.In this study,we aimed to parse the complex tumor microenvironment of osteosarcoma tissue,identify cell subsets that are closely associated with early micrometastases of osteosarcoma,and explore the application of this information in the prognosis assessment of patients with osteosarcoma.Materials and methods1.Analysis of microenvironment and intercellular communication in osteosarcoma(1)Using single-cell sequencing data from osteosarcoma patients,the complex cell composition of osteosarcoma tissue was analyzed through rigorous quality control screening,batch effect elimination,linear and nonlinear combined dimension reduction,and database and literature were used for identification of cell types.(2)Tumor tissue samples from osteosarcoma patients were taken for paraffin sections,and immunohistochemical staining was performed according to the marker genes of different cell types,to verify the results of single-cell sequencing of osteosarcoma.(3)Use i TALK ligand receptor database to compare and analyze the dominant expressed genes of each cell population,so as to discover the important communication relationship between cell populations.2.identifying the different invasive osteosarcoma subclusters(1)Extracted a group of cells announced as osteosarcoma cells from the Seurat object containing all cells to construct a new Seurat object.Dimensionality reduction and unbiased clustering was performed again to further subdivide the osteosarcoma cells.(2)The invasive ability of different osteosarcoma subgroups was evaluated according to the expression level of characteristic genes,activation degree of key pathways and important biological processes.(3)The transcriptional dynamics of each osteosarcoma subcluster were analyzed by the pseudo-time algorithm,and the cell differentiation trajectory in the process of tumor development and early micrometastasis was simulated,in an attempt to analyze the internal mechanism of different invasive osteosarcoma subclusters.3.Establishing a prognostic model of osteosarcoma patients based on the characteristics of key cell subsets(1)Transcriptome sequencing data and corresponding clinical survival information of osteosarcoma patients were collected from TARGET and GEO databases.Patients were divided into three groups: training group,internal validation group,and external validation group.Transcriptome sequencing data and corresponding clinical information of patients with HCC and colorectal cancer were collected from the TCGA database to clarify the specificity of the prognostic model.(2)By comparing the gene expression patterns of different invasive cell subsets of osteosarcoma,the gene set related to early micrometastasis of osteosarcoma was constructed after strict screening.(3)The prognostic model of osteosarcoma was established by univariate COX analysis and lasso regression analysis.The effects of the model were verified by drawing K-M survival curve and ROC curve of the training group,internal validation group and external validation group.The model was applied to other tumor patients to verify its specificity for osteosarcoma patients.(4)Combined with relevant literature,and wound healing and transwell experiments were conducted in vitro to clarify the influence of modeling genes on the biological characteristics of osteosarcoma,trying to analyze the molecular mechanism of the predictive ability of the model.Results1.Analysis of microenvironment and intercellular communication in osteosarcoma(1)Osteosarcoma is mainly composed of tumor cells,macrophages,mesenchymal stem cells,T cells,B cells,endothelial cells,pericytes,and myoblasts.The proportion of tumor cells was the highest,followed by macrophages.(2)There is a compact and complex cell-cell communication network in osteosarcoma.These cell-cell interactions regulate important biological processes,including tumor inflammatory response,immune escape,solid tumor growth,and new blood vessel formation.2.Identification of different invasive osteosarcoma cells(1)Osteosarcoma cells were clustered into nine subclusters(named OS-0 to OS-8).(2)Certain genes,which have been shown to inhibit OS metastasis,were highly expressed by the cells in OS-6 and OS-7.In addition,OS-8 cells showed high expression levels of some genes have been confirmed to promote metastasis of osteosarcoma.PI3K-AKT pathway,one of the key pathways promoting the metastasis in osteosarcoma,was activated in OS-8,but was inhibited in OS-6 and OS-7.(3)Pseudo-time analysis showed that there was obviously difference in the transcription dynamics between low metastasis potential subclusters(OS-6 and OS-7)and high metastasis potential subcluster(OS-8).3.Establishment of the Risk Regression Model based on the characteristics of key cell subsets(1)A total of 813 key genes associated with osteosarcoma metastasis were collected by comparing the expression patterns of different cell subsets with different metastasis potentials.(2)Establish a risk regression model for patients with osteosarcoma,which is composed of 5 genes.Among the training group,internal validation group and external validation group,the prognosis was significantly better in the low-risk group than the high-risk group,with significant differences in 5-year survival and median survival time.(3)This model is not suitable for prognosis prediction in patients with liver cancer and colorectal cancer,indicating that this model has significant specificity for patients with osteosarcoma.(4)Among the five genes that constitute the model,the influence of three genes on the proliferation,apoptosis,migration,invasion and other biological characteristics of osteosarcoma has been thoroughly studied.Two other modeling genes(EFEMP2 and GALNT14),the wound healing and the transwell assay showed that silencing their expression significantly inhibited the migration and invasion ability of osteosarcoma cells.ConclusionIn this study,starting from single-cell sequencing data from osteosarcoma,we elucidate the detailed cell composition of the tumor tissue and analyze the complex intercellular communication in the tumor microenvironment.Different osteosarcoma subsets have different metastasizing potential.Changes in transcriptional dynamics may be the underlying mechanism of different biological functions.Based on the above results,the risk regression model was established which could accurately and specifically predict the prognosis of patients with osteosarcoma.In conclusion,this study revealed the mechanism of early micrometastasis of osteosarcoma at the cellular and molecular levels,established a risk regression model,providing new ideas for the treatment and prognosis assessment of osteosarcoma patients. |
PDF Full Text Request