Stratification of ovarian cancer borderline from high-grade serous carcinoma patients by quantitative serum NMR spectroscopy of metabolites, lipoproteins, and inflammatory markers

Background: Traditional diagnosis is based on histology or clinical-stage classification which provides no information on tumor metabolism and inflammation, which, however, are both hallmarks of cancer and are directly associated with prognosis and severity. This project was an exploratory approach to profile metabolites, lipoproteins, and inflammation parameters (glycoprotein A and glycoprotein B) of borderline ovarian tumor (BOT) and high-grade serous ovarian cancer (HGSOC) for identifying additional useful serum markers and stratifying ovarian cancer patients in the future. Methods: This project included 201 serum samples of which 50 were received from BOT and 151 from high-grade serous ovarian cancer (HGSOC), respectively. All the serum samples were validated and phenotyped by ¹H-NMR-based metabolomics with in vitro diagnostics research (IVDr) standard operating procedures generating quantitative data on 38 metabolites, 112 lipoprotein parameters, and 5 inflammation markers. Uni- and multivariate statistics were applied to identify NMR-based alterations. Moreover, biomarker analysis was carried out with all NMR parameters and CA-125. Results: Ketone bodies, glutamate, 2-hydroxybutyrate, glucose, glycerol, and phenylalanine levels were significantly higher in HGSOC, while the same tumors showed significantly lower levels of alanine and histidine. Furthermore, alanine and histidine and formic acid decreased and increased, respectively, over the clinical stages. Inflammatory markers glycoproteins A and B (GlycA and GlycB) increased significantly over the clinical stages and were higher in HGSOC, alongside significant changes in lipoproteins. Lipoprotein subfractions of VLDLs, IDLs, and LDLs increased significantly in HGSOC and over the clinical stages, while total plasma apolipoprotein A1 and A2 and a subfraction of HDLs decreased significantly over the clinical stages. Additionally, LDL triglycerides significantly increased in advanced ovarian cancer. In biomarker analysis, glycoprotein inflammation biomarkers behaved in the same way as the established clinical biomarker CA-125. Moreover, CA-125/GlycA, CA-125/GlycB, and CA-125/Glycs are potential biomarkers for diagnosis, prognosis, and treatment response of epithelial ovarian cancer (EOC). Last, the quantitative inflammatory parameters clearly displayed unique patterns of metabolites, lipoproteins, and CA-125 in BOT and HGSOC with clinical stages I-IV. Conclusion: ¹H-NMR-based metabolomics with commercial IVDr assays could detect and identify altered metabolites and lipoproteins relevant to EOC development and progression and show that inflammation (based on glycoproteins) increased along with malignancy. As inflammation is a hallmark of cancer, glycoproteins, thereof, are promising future serum biomarkers for the diagnosis, prognosis, and treatment response of EOC. This was supported by the definition and stratification of three different inflammatory serum classes which characterize specific alternations in metabolites, lipoproteins, and CA-125, implicating that future diagnosis could be refined not only by diagnosed histology and/or clinical stages but also by glycoprotein classes.

Mitochondrial Dysfunction Pathway Alterations Offer Potential Biomarkers and Therapeutic Targets for Ovarian Cancer

The mitochondrion is a very versatile organelle that participates in some important cancer-associated biological processes, including energy metabolism, oxidative stress, mitochondrial DNA (mtDNA) mutation, cell apoptosis, mitochondria-nuclear communication, dynamics, autophagy, calcium overload, immunity, and drug resistance in ovarian cancer. Multiomics studies have found that mitochondrial dysfunction, oxidative stress, and apoptosis signaling pathways act in human ovarian cancer, which demonstrates that mitochondria play critical roles in ovarian cancer. Many molecular targeted drugs have been developed against mitochondrial dysfunction pathways in ovarian cancer, including olive leaf extract, nilotinib, salinomycin, Sambucus nigra agglutinin, tigecycline, and eupatilin. This review article focuses on the underlying biological roles of mitochondrial dysfunction in ovarian cancer progression based on omics data, potential molecular relationship between mitochondrial dysfunction and oxidative stress, and future perspectives of promising biomarkers and therapeutic targets based on the mitochondrial dysfunction pathway for ovarian cancer.

MASS SPECTROMETRY-BASED MITOCHONDRIAL PROTEOMICS IN HUMAN OVARIAN CANCERS

The prominent characteristics of mitochondria are highly dynamic and regulatory, which have crucial roles in cell metabolism, biosynthetic, senescence, apoptosis, and signaling pathways. Mitochondrial dysfunction might lead to multiple serious diseases, including cancer. Therefore, identification of mitochondrial proteins in cancer could provide a global view of tumorigenesis and progression. Mass spectrometry-based quantitative mitochondrial proteomics fulfils this task by enabling systems-wide, accurate, and quantitative analysis of mitochondrial protein abundance, and mitochondrial protein posttranslational modifications (PTMs). Multiple quantitative proteomics techniques, including isotope-coded affinity tag, stable isotope labeling with amino acids in cell culture, isobaric tags for relative and absolute quantification, tandem mass tags, and label-free quantification, in combination with different PTM-peptide enrichment methods such as TiO₂ enrichment of tryptic phosphopeptides and antibody enrichment of other PTM-peptides, increase flexibility for researchers to study mitochondrial proteomes. This article reviews isolation and purification of mitochondria, quantitative mitochondrial proteomics, quantitative mitochondrial phosphoproteomics, mitochondrial protein-involved signaling pathway networks, mitochondrial phosphoprotein-involved signaling pathway networks, integration of mitochondrial proteomic and phosphoproteomic data with whole tissue proteomic and transcriptomic data and clinical information in ovarian cancers (OC) to in-depth understand its molecular mechanisms, and discover effective mitochondrial biomarkers and therapeutic targets for predictive, preventive, and personalized treatment of OC. This proof-of-principle model about OC mitochondrial proteomics is easily implementable to other cancer types. © 2020 John Wiley & Sons Ltd. Mass Spec Rev.