Integrated analysis of ferroptosis and stemness based on single-cell and bulk RNA-sequencing data provide insights into the prognosis and treatment of esophageal carcinoma.

ABSTRACT

BACKGROUND: Cancer stem cells (CSCs) play a significant role in the recurrence and drug resistance of esophageal carcinoma (ESCA). Ferroptosis is a promising anticancer therapeutic strategy that effectively targets CSCs exhibiting high tumorigenicity and treatment resistance. However, there is a lack of research on the combined role of ferroptosis-related genes (FRGs) and stemness signature in the prognosis of ESCA. METHODS: The cellular compositions were characterized using single-cell RNA sequencing (scRNA-seq) data from 18 untreated ESCA samples. 50 ferroptosis-related stemness genes (FRSGs) were identified by integrating FRGs with stemness-related genes (SRGs), and then the cells were grouped by AUCell analysis. Next, functional enrichment, intercellular communication, and trajectory analyses were performed to characterize the different groups of cells. Subsequently, the stem-ferr-index was calculated using machine learning algorithms based on the expression profiles of the identified risk genes. Additionally, therapeutic drugs were predicted by analyzing the GDSC2 database. Finally, the expression and functional roles of the identified marker genes were validated through in vitro experiments. RESULTS: The analysis of scRNA-seq data demonstrates the diversity and cellular heterogeneity of ESCA. Then, we identified 50 FRSGs and classified cells into high or low ferroptosis score stemness cells accordingly. Functional enrichment analysis conducted on the differentially up-regulated genes between these groups revealed predominant enrichment in pathways associated with intercellular communication and cell differentiation. Subsequently, we identified 9 risk genes and developed a prognostic signature, termed stem_ferr_index, based on these identified risk genes. We found that the stem-ferr-index was correlated with the clinical characteristics of patients, and patients with high stem-ferr-index had poor prognosis. Furthermore, we identified four drugs (Navitoclax, Foretinib, Axitinib, and Talazoparib) with potential efficacy targeting patients with a high stem_ferr_index. Additionally, we delineated two marker genes (STMN1 and SLC2A1). Particularly noteworthy, SLC2A1 exhibited elevated expression levels in ESCA tissues and cells. We provided evidence suggesting that SLC2A1 could influence the migration, invasion, and stemness of ESCA cells, and it was associated with sensitivity to Foretinib. CONCLUSION: This study constructed a novel ferroptosis-related stemness signature, identified two marker genes for ESCA, and provided valuable insights for developing more effective therapeutic targets targeting ESCA CSCs in the future.