Identification of estrogen response-associated STRA6+ granulosa cells within high-grade serous ovarian carcinoma by single-cell sequencing

Abstract

Background: High-grade serous ovarian carcinoma (HGSOC) is a pathologic subtype of ovarian cancer (OC) with a more lethal prognosis. Extensive heterogeneity results in HGSOC being more susceptible to treatment resistance and adverse treatment effects. Revealing the heterogeneity involved is crucial.

Methods: We downloaded the single-cell RNA-seq (scRNA) data from GEO database and performed a scRNA analysis for cell landscape of HGSOC by using the Seurat package. The highly expressed genes were uploaded into the DAVID and KEGG database for enrichment analysis, and the AUCell package was used to calculate cancer-associated hallmark score. The SCENIC analysis was used for key regulons, the estrogen response enrichment scores in TCGA-OV RNA-seq dataset were calculated by using the GSVA package. Besides, the expression of STRA6 and IRF1 and the cell invasion and migration in si-STRA6 OC cells were detected by using the quantitative reverse transcription (qRT)-PCR method and Transwell assay respectively.

Results: We successfully constructed a single-cell atlas of HGSOC and delineated the heterogeneity of epithelial cells therein. There were five epithelial cell subpopulations, GLDC + Epithelial cells, PEG3+ leydig cells, STRA6+ granulosa cells, POLE2+ Epithelial cells, and AURKA + Epithelial cells. STRA6+ granulosa cells have the potential to promote tumor growth as well as the highest estrogen response early activity through the biological pathways analysis of highly expressed genes and estrogen response score of ssGSEA. We found that IRF1 and STRA6 expression was remarkably upregulated in the OC cancer cell line HEY. Silencing of STRA6 markedly decreased the invasion and migration ability of the OC cancer cell line HEY.

Conclusion: There is extreme heterogeneity of epithelial cells in HGSOC, and STRA6+ granulosa cells may be able to promote cancer progression. Our findings are benefit to the heterogeneity identification of HGSOC and develop targeted therapy strategy for HGSOC patients.

Keywords: GRNs; High-grade serous ovarian carcinoma; SCENIC; STRA6+ granulosa cells; Single-cell sequencing; TFs.

Fig. 1

Cell mapping in HGSOC. A: Cell mapping of 11 cell subpopulations in HGSOC. B–C: Expression levels of marker genes specifically expressed in 11 cell subpopulations. D–E: cell numbers and proportions of 11 cell subpopulations.

Fig. 2

Five heterogeneous subpopulations of Epithelial cells ya in HGSOC and the biological pathways they are involved in. A: Six heterogeneous subpopulations of Epithelial cells in HGSOC. B: Marker genes specifically expressed in five heterogeneous subpopulations of Epithelial cells in HGSOC. C: Biological pathways involved in GLDC + Epithelial cells. D: Biological pathways involved in PEG3+ leydig cells. E: Biological pathways involved in STRA6+ granulosa cells. F: Biological pathways involved in POLE2+ Epithelial cells. G: Biological pathways involved in AURKA + Epithelial cells.

Fig. 3

Enrichment analysis of the hallmark pathway and cell communication analysis of epithelial cell subpopulations. A: Heatmap of hallmark pathway enrichment scores for epithelial cell subpopulations. B–C: Network diagram of the number of ligand-receptor pairs interacting between epithelial cell subpopulations. (B) Cell-cell contact type. (C) Secreted signaling type.

Fig. 4

Communication density between STRA6+ granulosa cells and epithelial cell subpopulations. A: Cell-cell contact type. B: Secreted signaling type.

Fig. 5

STRA6+ granulosa cells activate estrogen response early activity. A: K-M curves of patients in the estrogen response early subgroup. B: estrogen response early enrichment score in a subpopulation of epithelial cells. C: Expression levels of estrogen pathway-related genes in a subpopulation of epithelial cells. , and D: Location of estrogen pathway-related genes highly expressed in STRA6+ granulosa cells within the KEGG pathway map.

Fig. 6

STRA6+ granulosa cells regulate estrogen response early via IRF1-dominant GRN. A: pearson correlation analysis of AUCell score of TFs with estrogen response early score of STRA6+ granulosa cells. B: Pearson correlation analysis of IRF1 with estrogen response early score in STRA6+ granulosa cells. C: IRF1-dominated biological pathways involved in GRN. Boxes represent TFs, circles represent target genes, and different colors represent different biological processes.

Fig. 7

Effects of IRF1 and STRA6 on OC invasion and migration. A: Expression levels of IRF1 in IOSE-80 and HEY. B: Expression levels of STRA6 in IOSE-80 and HEY. C–D: Effects of silencing STRA6 on OC cell invasion and migration. (*p < 0.05, **p < 0.01, ***p < 0.001).