Characterization of Differentially Expressed Proteins for Ovarian Cancer

Abstract

Ovarian cancer (OC) is the most lethal gynaecological malignancy worldwide and is the fifth leading death cause from cancer in women. Nonspecific and vague early symptoms often result in late diagnosis of OC with around five year survival rate (˂ 30%). Early diagnosis is the key factor for successful treatment and improved patient survival. Advanced diagnostics and modern treatment strategies even though caused substantial drop in mortality associated with OC in developed states, progress is still needed for improved patient survival in developing countries. In the country like Pakistan, where although the prevalence of OC is comparatively low, late diagnosis of the disease results in higher mortality. Existing approaches being used for ovarian tumor detection are either invasive or devoid of enough sensitivity and specificity. Also, there are no FDA recommended screening tools for early stage OC detection in the general population. Thus development of non-invasive, sensitive and specific diagnostic/screening procedures is required for improved early diagnosis and surveillance in OC. Proteomic methodologies have been employed for discovery of molecular markers associated with numerous malignancies including colorectal, chronic myeloid leukemia, pancreatic, prostrate, oral, lung and renal cell carcinoma with noteworthy outcomes. In the present work, a methodical approach combining two dimensional- gel electrophoresis (2D-GE) and matrix-assisted laser desorption/ionisation mass spectrometry (MALDI-MS) is used for proteomic profiling of human ovarian cancer and benign disease control ovarian tissue samples with an objective of finding potential markers of the disease. Equal protein content was resolved by 2D-GE from OC and benign control tissue lysate samples followed by 2D gel staining with colloidal coomassie. Resultantly the 2DE map of OC and control tissue lysates iv produced overall 376 ± 121 and 427±115 spots, respectively, when comparison was performed for all of the three stages studied. Selection of differentially expressed statistically significant 2D gel spots was performed by quantitation of individual spots by Dymension v.3.0.1.2 and ImageMaster 2D Platinum 7.0 software. 2D gel spots having at least one-fold intensity difference (117 ±29) were selected in the initial screening followed by FDR analysis with value set at ≤ 0.2; indicating 80% accuracy in findings. Only these selected spots were analyzed by MALDI-TOF-MS analysis for protein identification. Reliability of the peptide mass fingerprinting identification data was assessed by calculating PMF score for each identified protein with ≥ 79 threshold value set for positive hits. MS analysis resulted in identification of 31 distinct proteins with the 11 proteins recognized with better FDR ≤ 0.05 (95% accuracy confidence) and PMF score ˃ 79. In silico characterization of the MS identified dataset proteins was also performed. Gene Ontology (GO) based categorization classified present data proteins according to their molecular functions, biological processes and cellular components. To have an overview of the functional interactions among data proteins, these were subjected to analysis using the software tool STRING v10. The analysis revealed that four of the proteins in the present data; CYHR1, COL6A2, KRT2 and KRT9 were functionally separated from the rest of dataset differentially expressed proteins. Hence, individual interacting partners of these four proteins and another two proteins of interest in present study, i.e. ANXA6 and TBB4B were analyzed. ANXA6 appears to be involved in interaction with proteins that are directly involved in apoptotic growth of the cells, tumourigenesis, inflammation, immune response, signaling, membrane damage repair and stress responses; while TBB4B is found to be involved in interaction with one of the differentially expressed proteins in the dataset (TBA1C) v and other tubulin family proteins that are involved in tumor development and survival of tumor cells. Additionally, characterization of the differentially regulated proteins achieved by Ingenuity Pathway Analysis (IPA) online core analysis software through annotating downstream effect analysis customized for inquiry of data proteins association with diseases and disorders, cellular and molecular functions, associated networks analysis, canonical pathways and biomarker filter analysis. According to IPA, three of the top diseases linked with the 31 MS identified proteins were cancer, reproductive and endocrine system disorders. Networks involved in cancer, organismal injury and abnormalities, reproductive system disease were found dysfunctional in OC. Important canonical pathways in which differentially abundant present data proteins were classified included; remodeling of epithelial adherens junctions, gap junction signaling, 14-3-3 mediated signaling and Sertoli cell-Sertoli cell junction signaling. IPA biomarker filter analysis selected four dataset proteins as potential markers of OC i.e. ANXA6, ALB, LDHA and APOA1. All of the above mentioned computational analyses explain the likelihood of involvement of differentially regulated proteins in the health status of an individual. After MS analysis, many of the identified proteins were represented by multiple spots with slight change in MW or pI suggesting post translational modifications. To address this difference is due to PTMs, data proteins were subjected to PTM analysis using NetNGlyc 1.0 (http://www.cbs.dtu.dk/services/NetNGlyc/) and Protter (http://wlab.ethz.ch/protter/) server. Analysis revealed single or multiple glycosylation sites in many of the identified protein sequences namely ANXA6, VIM, TBB4B, TBB5, ACTG, COL6A2 etc., with significantly high scores (threshold ≥ 0.5) along vi with other PTMs viz. acetylation, phosphorylation, S-nitrosylation and Ubiquitination of potential amino acid sequences. Of the 31 candidate proteins identified in the present work, eleven proteins that met the four tier criteria (fold change ≥ 1.5, p-value ˂ 0.05, FDR ≤ 0.05 and PMF score ˃ 79) were selected for further discussion/analysis. Among these eleven proteins, nine proteins (ANXA2, ANXA4, ANXA5, ANXA6, VIM, TBB4B, TBA1C, ACTG and COL6A2) exhibited enhanced expression in OC compared to the benign controls while two (LDHA and APOA1) showed reduced staining. Several proteins that established differential regulation in the present work have been previously described to be associated with other malignancies and could signify premalignant alterations. Two of the marker proteins identified in this study, both having up regulated expression in OC, have not been found previously associated with OC. These included ANXA6 and TBB4B and were only selected for verificational analysis. Validation by western blot (WB) confirmed their enhanced expression in OC tissues compared to benign controls. Quantitative estimation of ANXA6 and TBB4B by ELISA in the plasma samples of subjects also confirmed their up regulation in OC. Furthermore, the differentiating expressions of both of the candidate proteins were closely associated with the clinical manifestation of the disease and were further elevated in the advanced stage (III and IV) OC than the less advanced tumors (stage II). Hence, ANXA6 and TBB4B up regulation in OC reported here for the first time, is novel. It may provide insight into the mechanism of OC metastatic progression, potentially leading to the design of novel diagnostic and therapeutic strategies and serve as biomarkers of ovarian cancer. Furthermore, assessment of these potential markers likely association with OC, in a larger population, besides developing clinically useful biomarkers also improve our understanding about disease progressive mechanism in future