Study On The Intratumoral Microbiome And Its Interactions With The Tumor Microenvironment Based On RNA-seq Data Of Ovarian Cancer

BACKGROUD:ovarian carcinoma(OV)is the deadliest of the three major gynecologic malignancies.Although the survival rate has been improved over the past 40 years,the five-year survival rate is still less than 40%.Therefore,it is important to explore the factors associated with prognosis and treatment outcome of ovarian cancer.The tumor microenvironment(TME)is an important part of tumor that is closely related to the clinical outcome and therapeutic response.Immune cells including natural killer cells,lymphocytes and macrophages are important parts of the TME and have a significant impact on tumor development and the efficacy of antitumor therapy.the TME is conducive to microbial growth,and the presence of live microbes in tumors had been demonstrated more than a century ago.Recent studies have proved the presence of microbiota in most human tumors,including ovarian cancer.Microorganisms not only promote tumorigenesis as direct carcinogens,but also influence tumorigenesis and cancer treatment in several ways.The relationship among intratumoral microbes,TME and prognosis of ovarian cancer is unclear.In addition,the function of microorganisms associated with TME and prognosis of ovarian cancer is worth to explore.Based on transcriptome sequencing(RNA-seq)data obtained from The Cancer Genome Atlas(TCGA)-OV project,our study proposed to explore the interaction among intratumoral microbiome,host gene expression and prognosis of OV.The tumors were typed by Gene Set Variation Analysis(GSVA)and characterized by different subtypes.Non-human sequences were extracted from RNA-seq data and used for microbial annotation.Differences in microbial composition were compared between subtypes.Microbes associated with survival prognosis were screened to construct OV survival prognosis model.We further analyzed the association between microorganisms and TME within the tumor and verified the relationship between microorganisms and anti-tumor immunity using cellular experiments.METERIALS AND METHODS:A total of 373 primary tumor samples were included in our study.We downloaded the raw transcriptome sequencing data,gene expression matrices,clinical and survival information,TCR/BCR diversity,tumor purity and neoantigen load data.TME-related signature gene sets were collected from Bagaev et al.The intratumoral immune status of OV patients was analyzed using gene expression matrix and clustered to different subtypes.We further identified the characteristics of tumor immune microenvironment of different subtypes,including immune cells,tumor mutational burden(TMB),etc.Differentially expressed genes between groups were identified by DESeq2 and gene enrichment analysis were performed using Kyoto Encyclopedia of Genes and Genomes(KEGG)and Gene Ontology(GO)databases.For the raw transcriptome sequencing data,after sequence filtering,quality control,microbial annotation,contaminant identification and decontamination,a total of 2526 species and 855 genera were obtained.Differences in microbiome among different immune subtypes were explored by α-and β-diversity analysis and linear discriminant analysis effect size(LEfSe).CCLasso analysis were performed to identify microbial co-occurrence networks of the two subtypes using MetagenoNets platform.Univariate Cox proportional risk analysis was used to screen microbial markers related to survival prognosis.Further selection of prognostic microbial markers for ovarian cancer survival was performed using Lasso-penalized Cox analysis.The relationship between prognosis-related microbes and tumor immune microenvironment was explored based on Spearman’s rank correlation analysis.By co-culturing the Acinetobacter seifertii with ID8 mouse ovarian cancer cells and mouse peritoneal macrophages,six groups were designed:1)Acinetobacter seifertii group(As-10),2)Acinetobacter seifertii-treated ID8 cells group(ID8-As-10cell),3)Acinetobacter seifertii-treated ID8 cells conditioned medium(CM)group(ID8-As-10-CM),4)untreated ID8 cells group(ID8-cell),5)untreated ID8 cells in CM group(ID8-CM)and 6)DMEM medium group.The effect of Acinetobacter seifertii and ID8 cells on macrophage differentiation was first analyzed using quantitative reverse transcription polymerase chain reaction(RT-qPCR).Next,the effect of Acinetobacter seifertii and ID8 cells on the migration ability of macrophages was then analyzed using Transwell migration assay.RESULTS A total of 373 transcriptome sequencing data of primary tumor samples from untreated OV patients in TCGA and their corresponding clinical and survival information were included in this study.29 signature gene sets describing the tumor microenvironment were obtained from Bagaev et al.OV was classified into immuneenriched and-deficient subtypes based on GSVA and cluster analysis,with no significant differences in demographic and clinical indicators between the two subtypes.Compared to the immune-deficient type,the immune-enriched type had significantly longer overall survival(OS),progression-free interval(PFI),disease-specific survival(DSS),and disease-free interval(DFI).Immune-enriched type showed a hyperimmune state with significantly increased TMB and infiltration of immune cells such as ID8+T cells.Differentially expressed genes between groups were significantly enriched in pathways related to immune activation,regulation and immune-related diseases.A total of 3040 species and 871 genera were obtained after reads filtering,quality control,and microbial annotation.After contaminant identification and decontamination,2526 species and 855 genera were retained,and 17%of microorganisms were identified as contaminants which were removed from downstream analyses.β-diversity and CCLasso analyses revealed significant differences in the intratumoral microbiome between immune-enriched and immunedeficient types.58 species,mainly from Pseudomonas,were found to be significantly enriched in the immune-deficient type using LEfSe analysis,and 11 species were significantly enriched in the immune-enriched type.Functional analysis revealed that microbial functional genes in immune-deficient type were significantly up-regulated in phosphatidylinositol signaling system(map04070),PI3K-Akt signaling pathway(map04151)and taurine and hypotaurine metabolism(map00430).Univariate Cox proportional risk analysis identified 736 microorganisms significantly associated with OS,PFI,DSS,or DFI(P<0.05),of which 193 species were associated with OS.Further,32 prognostic species were screened through Lassopenalized Cox regression analysis.Correlation analysis showed that 5 immune cells were significantly correlated with 7 prognostic species(P<0.05).Among them,M1 cell was significantly correlated with 5 prognostic microorganisms(P<0.05):positively correlated with 2 protective factors,Achromobacter deleyi and Microcella alkiliile,and negatively correlated with 3 risk factors,Devosia sp.LEGU1,Ancylobacter pratisalsi and Acinetobacter seifertii.Survival analysis showed that M1 cell was significantly associated with longer OS and increased in tumors of the low-risk group.Thus,M1 cells were identified as key cells linking intratumoral microorganisms to ovarian cancer survival prognosis.RT-qPCR experiments showed that the M1 differentiation biomarkers was substantially upregulated in the macrophages treated with Acinetobacter seifertii compared to macrophages treated with untreated ID8 cells.the results of Transwell cell migration assay showed that the migration ability of macrophages was significantly decreased after Acinetobacter seifertii treatment,especially in the ID8-As-10-cell group,the migratory capacity of macrophages was significantly decreased(P<0.05).CONCLUSION In this study,OV was divided into immune-enriched anddeficient types.The immune-enriched type was significantly associated with longer survival time.Microbiome analysis showed that the intratumoral microbial composition between two subtypes was significantly different.We further screened 32 prognostic microbiota and found that intratumoral microbiota is significantly associated with the prognosis and of immune status OV.M1 cell was closely related to intratumoral microorganisms and negatively correlated with Acinetobacter seifertii,which could affect its intratumoral infiltration by inhibiting the migration ability of macrophages.