Abstract
BACKGROUND: Understanding the tumor microenvironment (TME) is essential for the advancement of immunotherapy for ovarian cancer (OC). Nonetheless, predicting transcription factor (TF) regulation from the TME using single-cell RNA sequencing (scRNA-seq) data is challenging. METHODS: The OC scRNA-seq data were analyzed with a specialized scRNA-seq transcriptome analysis application. The OC TME was utilized to instruct the SCENIC procedure for TF regulation. We built a risk model using Lasso regression and identified immunological subgroups using ConsensusClusterPlus. To analyze the percentage of invading immune cells, the algorithms CIBERSORT, ESTIMATE, and xCell were used. We computed the stromal score, immunological score, estimate score, and tumor purity to evaluate the risk model's capacity to predict the tumor immune microenvironment. Additionally, the expression of immunological checkpoints was examined, and for pertinent evaluation, the imvigor 210 dataset of the immunotherapy cohort was used. pRophetic predicted the sensitivity of 138 GDSC database drugs. In addition, we examined the expression of unproven risk model genes using qPCR and immunohistochemistry (JCHAIN, UBD, and RARRES1). Cell proliferation was assessed by colony formation assays. Transwell experiments were used to examine the invasion and migration ability of OC cells. RESULTS: Six immunologically malignant cell subpopulations have been identified within the cancer immune microenvironment (referred to as TC0-6). Unique in its immunological profile, TC0 demonstrates the most intimate interactions with immune cells. Following a meta gene screen in the TC0 subpopulation using the top 30 targets of 14 transcription factor (TF) factors, two distinct immunological molecular subtypes-the C1 and C2 subtypes-with notable survival differences were discovered. On the basis of nine genes whose expression differs between the C1 and C2 subtypes, a risk model was constructed. The risk model is an accurate method for forecasting the effectiveness of immunotherapies, clinicopathological characteristics, and survival. JCHAIN and UBD expression in OC tissues was found to be low according to qPCR and IHC analyses, whereas RARRES1 expression was found to be high. The functional experiment results indicated that downregulation of JCHAIN and UBD and overexpression of RARRES1 could suppress the proliferation, migration, and invasion of OC cells in vitro. CONCLUSION: Based on TF regulatory networks in the tumor microenvironment, this study developed a 9-gene risk model for the prognosis of ovarian cancer. This model may aid in the future promotion of personalized OC immunotherapy. In addition, JCHAIN, UBD, and RARRES1 were identified as three novel immune-related biomarkers for OC.