Stemness-induced chemoresistance and therapeutic improvements in HGSOC through multi-omics analysis

Chemotherapy resistance caused by cancer stem cells (CSCs) is a crucial reason of high grade serous ovarian cancer (HGSOC) recurrence. To characterize the complex interplay in the HGSOC tumor microenvironment (TME), we collected 47 samples approximately 180,000 cells to explore the changes of TME induced by chemotherapy at single cell resolution. By tracing the differentiation trajectory from CSCs to highly malignant cells, we identified key factors in tumor progression and revealed transcriptome variations induced by chemotherapy. Additionally, we observed complex crosstalk among TME components, especially involving the CD44 in immunity activation and chemotherapy resistance. Using the chemotherapy response score (CRS), we determined differentially expressed genes (CRS-DEGs) at the single-cell level and applied them to bulk RNA-seq data to develop a chemotherapy response prediction model. A novel cancer stemness signature (CSC.sig) was created based on CRS-DEGs, showing significant immunosuppressive characteristics. We further screened crucial genes associated with prognosis and integrated multiple machine learning algorithms to construct an overall survival assessment model. Our study provides comprehensive insights into the molecular mechanisms of chemoresistance due to CSCs, highlights potential therapeutic drugs, and develops models for prognosis assessment.