Objective: Ovarian cancer has the highest mortality rate among gynecological malignant tumors, and it preferentially metastasizes to omental tissue, leading to intestinal obstruction and death. scRNA-seq is a powerful technique to reveal tumor heterogeneity. Analyzing omentum metastasis of ovarian cancer at the single-cell level may be more conducive to exploring and understanding omentum metastasis and prognosis of ovarian cancer at the cellular function and genetic levels.
Methods: The omentum metastasis site scRNA-seq data of GSE147082 were acquired from the GEO (Gene Expression Omnibus) database, and single cells were clustered by the Seruat package and annotated by the SingleR package. Cell differentiation trajectories were reconstructed through the monocle package. The ovarian cancer microarray data of GSE132342 were downloaded from GEO and were clustered by using the ConsensusClusterPlus package into omentum metastasis-associated clusters according to the marker genes gained from single-cell differentiation trajectory analysis. The tumor microenvironment (TME) and immune infiltration differences between clusters were analyzed by the estimate and CIBERSORT packages. The expression matrix of genes used to cluster GSE132342 patients was extracted from bulk RNA-seq data of TCGA-OV (The Cancer Genome Atlas ovarian cancer), and least absolute shrinkage and selection operator (LASSO) and multivariate Cox regression were performed to establish an omentum metastasis-associated gene (OMAG) signature. The signature was then tested by GSE132342 data. Finally, the clinicopathological characteristics of TCGA-OV were screened by univariate and multivariate Cox regression analysis to draw the nomogram.
Results: A total of 9885 cells from 6 patients were clustered into 18 cell clusters and annotated into 14 cell types. Reconstruction of differentiation trajectories divided the cells into 5 branches, and a total of 781 cell trajectory-related characteristic genes were obtained. A total of 3769 patients in GSE132342 were subtyped into 3 clusters by 74 cell trajectory-related characteristic genes. Kaplan-Meier (K-M) survival analysis showed that the prognosis of cluster 2 was the worst, P < 0.001. The TME analysis showed that the ESTIMATE score and stromal score in cluster 2 were significantly higher than those in the other two clusters, P < 0.001. The immune infiltration analysis showed differences in the fraction of 8 immune cells among the 3 clusters, P < 0.05. The expression data of 74 genes used for GEO clustering were extracted from 379 patients in TCGA-OV, and combined with survival information, 10 candidates for OMAGs were filtered by LASSO. By using multivariate Cox regression, the 6-OMAGs signature was established as RiskScore = 0.307*TIMP3 + 3.516*FBN1-0.109*IGKC + 0.209*RPL21 + 0.870*UCHL1 + 0.365*RARRES1. Taking TCGA-OV as the training set and GSE132342 as the test set, receiver operating characteristic (ROC) curves were drawn to verify the prognostic value of 6-OMAGs. Screened by univariate and multivariate Cox regression analysis, 3 (age, cancer status, primary therapy outcome) of 5 clinicopathological characteristics were used to construct the nomogram combined with risk score.
Conclusion: We constructed an ovarian cancer prognostic model related to omentum metastasis composed of 6-OMAGs and 3 clinicopathological features and analyzed the potential mechanism of these 6-OMAGs in ovarian cancer omental metastasis.
Keywords: 6-OMAGs; Omentum metastasis; Ovarian cancer; Prognosis; Tumor microenvironment; scRNA-seq.
© 2022. The Author(s).