Signaling pathway network alterations in human ovarian cancers identified with quantitative mitochondrial proteomics

Abstract

Relevance: Molecular network changes are the hallmark of the pathogenesis of ovarian cancers (OCs). Network-based biomarkers benefit for the effective treatment of OC.

Purpose: This study sought to identify key pathway-network alterations and network-based biomarkers for clarification of molecular mechanisms and treatment of OCs.

Methods: Ingenuity Pathway Analysis (IPA) platform was used to mine signaling pathway networks with 1198 human tissue mitochondrial differentially expressed proteins (mtDEPs) and compared those pathway network changes between OCs and controls. The mtDEPs in important cancer-related pathway systems were further validated with qRT-PCR and Western blot in OC cell models. Moreover, integrative analysis of mtDEPs and Cancer Genome Atlas (TCGA) data from 419 patients was used to identify hub molecules with molecular complex detection method. Hub molecule-based survival analysis and multiple multivariate regression analysis were used to identify survival-related hub molecules and hub molecule signature model.

Results: Pathway network analysis revealed 25 statistically significant networks, 192 canonical pathways, and 5 significant molecular/cellular function models. A total of 52 canonical pathways were activated or inhibited in cancer pathogenesis, including antigen presentation, mitochondrial dysfunction, GP6 signaling, EIF2 signaling, and glutathione-mediated detoxification. Of them, mtDEPs (TPM1, CALR, GSTP1, LYN, AKAP12, and CPT2) in those canonical pathway and molecular/cellular models were validated in OC cell models at the mRNA and protein levels. Moreover, 102 hub molecules were identified, and they were regulated by post-translational modifications and functioned in multiple biological processes. Of them, 62 hub molecules were individually significantly related to OC survival risk. Furthermore, multivariate regression analysis of 102 hub molecules identified significant seven hub molecule signature models (HIST1H2BK, ALB, RRAS2, HIBCH, EIF3E, RPS20, and RPL23A) to assess OC survival risks.

Conclusion: These findings provided the overall signaling pathway network profiling of human OCs; offered scientific data to discover pathway network-based cancer biomarkers for diagnosis, prognosis, and treatment of OCs; and clarify accurate molecular mechanisms and therapeutic targets. These findings benefit for the discovery of effective and reliable biomarkers based on pathway networks for OC predictive and personalized medicine.

Keywords: Biomarkers; Hub molecules; Mitochondrial proteomics; Ovarian cancer; Pathway networks; Predictive preventive personalized medicine.

Fig. 1

Pathway overlapping analysis and validation of cancer associated canonical pathways. a The interactions among those identified canonical pathways. b qRT-PCR quantitative analysis of the expression levels of TPM1, LYN, CALR, GSTP1, CPT2, and AKAP12 in OC cells TOV21G and SKOV3 compared to control cells IOSE80. c Protein expression levels of TPM1, LYN, CALR, GSTP1, CPT2, and AKAP12 in OC cells TOV21G and SKOV3 compared to control cells IOSE80. d Western blot quantitative analysis of TPM1, LYN, CALR, GSTP1, CPT2, and AKAP12 in OC cells TOV21G and SKOV3 compared to control cells IOSE80. n = 3. *p < 0.05; **p < 0.01; ***p < 0.001. NS no significance

Fig. 2

The hub molecules were identified within OC tissues. a The protein–protein interactions (PPIs) with combined scores greater than 0.4 were selected to construct PPI network. b–g The entire PPI network was analyzed using MCODE, and six modules (module 1 score = 24.357, module 2 score = 15, module 3 score = 10.435, module 4 score = 10.4, module 5 score = 8.769, and module 6 score = 6.609) were obtained

Fig. 3

Biological processes, co-expression and co-localization analysis of hub molecules. a The hub molecules were classified according to the biological process. b The results of co-expression at the mRNA level between hub molecules. c The results of chromosome location and co-expression at the protein level between hub molecules. d The cell location of those hub molecules

Fig. 4

Kaplan–Meier (KM) survival curve of hub molecules ALDH1L1, COL1A1, COL6A2, CXCL12, GSTM3, HSPA5, HSPE1, LUM, P4HB, PRKACB, RPS19, and XRCC6 in OCs