Mapping single-cell transcriptomes of endometrium reveals potential biomarkers in cancer

Abstract

Abstract Background Deconvolution of immune microenvironment that drive transcriptional programs throughout the menstrual cycle is key to understanding regulatory biology of endometrium. Methods We comprehensively analyzed single cell transcriptome of 59,397 cells across ten human endometrium samples. Cell specific expression of genes were revealed and transcription factors that potentially regulated these genes were identified by SCENIC. CellChat was used to analyze the cell-cell communications. The RNA-based molecular subtypes of human endometrial cancers were revealed by nonnegative matrix factorization analysis. Results Single cell transcriptome analyses revealed the dynamic cellular heterogeneity throughout the menstrual cycle. In particular, we identified two perivascular cell subtypes, four epithelial subtypes and four fibroblast cell types in endometrium. Moreover, we inferred the cell type-specific transcription factor (TF) activities and linked critical TFs to transcriptional output of diverse immune cell types, highlighting the importance of transcriptional regulation in endometrium. Dynamic interactions between various types of cells in endometrium contribute to a range of biological pathways regulating differentiation of secretory. Integration of the molecular biomarkers identified in endometrium and bulk transcriptome of 535 endometrial cancers (EC), we revealed five RNA-based molecular subtypes of EC with highly intratumoral heterogeneity and different clinical manifestations. Mechanism analysis uncovered clinically relevant pathways for pathogenesis of EC. Conclusions In summary, dynamic immune microenvironment analyses provide novel insights into future development of RNA-based treatments for endometriosis and endometrial carcinoma.