Pijun Gong, Jia Li, Yinbin Zhang and Shuqun Zhang* Pages 1 - 19 ( 19 )
Background: Ovarian cancer (OV) is one of the deadliest gynecologic cancers, and approximately 75% of serous ovarian cancer (SOC) patients are diagnosed at advanced stages due to the lack of effective biomarkers.
Objective: Immunogenic cell death (ICD) has been investigated in many comprehensive studies, and the role of ICD in ovarian cancer and its impact on immunotherapy is not yet known.
Methods: The NMF clustering analysis was employed to categorize OV samples into different subgroups. Survival, mutation, and CNV analyses were performed in these clusters. ESTIMATE, CIBERSORT, TIDE, and drug sensitivity analyses (based on GDSC) were also performed on the subtypes. Then, differentially expressed immunogenic cell death genes (DE-ICDGs) in OV were obtained by crossing the DEGs between cluster3 vs. cluster1, DEGs from the TCGA-GTEx dataset, and DEGs from the GSE40595 dataset. Functional enrichment analysis of DE-ICDGs was then performed. The signature genes related to the prognosis of OV in three OV datasets were excavated by drawing Kaplan-Meier curves. Finally, quantitative real-time PCR (qRT-PCR) was performed to verify the expression trends of the signature genes.
Results: The NMF clustering analysis categorized OV samples into three distinct groups according to the expression levels of ICDGs, with differential analysis indicating that Cluster3 represented the subgroup with high ICD expression. Mutation and CNV analysis did not differ significantly between clusters, but Amp and Del's numbers did. Immuno- infiltration analysis revealed that cluster3 showed significant differences from cluster1 and cluster2. Immunotherapy and drug sensitivity analysis showed differences in immunotherapy and chemotherapy sensitivity between the clusters. The DEGs in cluster3 vs. cluster1, TCGA-GTEx dataset and GSE40595 dataset were intersected to obtain a total of 71 DE-ICDGs, and functional enrichment result suggested that the DE-ICDGs were significantly correlated with inflammatory response, complement system and positive regulation of cytokine production. 2 DE-ICDGs (FN1 and LUM) were identified that were associated with OV prognosis and were validated significantly down-regulated in the SOC group with PCR.
Conclusion: We identified ICD-associated subtypes of OV and mined 2 OV prognostic genes (FN1 and LUM) associated with ICD, which may have important implications for OV prognosis and therapy.
Immunogenic cell death, immune infiltration, ovarian cancer, drug sensitivity, immunotherapy, chemotherapy.