| Background"World Cancer Report 2014" reported that newly diagnosed cases and mortality for esophageal cancer(EC) ranks first in the world. Histologically, esophageal squamous cell carcinoma(ESCC) accounts for 90% of EC in China. In the past two decades, the overall five-year survival rate of EC has maintained constant and is in the range of 15%-25%. One of the main reasons for such poor prognosis is that most EC patients are at moderate or advanced clinical stages when first present in clinic, and at least 50% of patients have distant metastases. Obviously, it is of utmost importance to clarify the risk factors and molecular mechanism of ESCC, to identify ESCC biomarkers with high sensitivity and specificity and to explore new therapeutic targets and treatment modalities, with a purpose to reduce morbidity and mortality for ESCC.As the ultimate executioner of life activities, proteins have important biological functions in the formation and development of diseases. Proteins participate the maintenance of biological activities through protein-protein interaction network. Under pathological conditions, protein-protein interaction networks will dynamically change. Therefore, it is impossible for an individual protein to reflect physiological or pathological conditions of the body, resulting in limited sensitivity and specificity in terms of disease diagnosis. The application of mass spectrometry, protein chips and bioinformatics in life science enable us to understand the nature of life and disease from the perspective of system biology. In addition, mass spectrometry based proteomics has limits such as poor reproducibility, sample dependent data quality and low protein sequence coverage. Therefore, it is necessary to find alternative approach to address these deficiencies and simultaneously increase the probablity of identification of low-abundance proteins. Protein-protein interaction network analysis not only provides insights in molecular function, biological pathways and processes but also compensates the limits of proteomics to some extent. Aims1. To identify the differential expressed proteins by iTRAQ labeling and mass spectrometry between ESCC and adjacent normal esophageal epithelium.2. The seed proteins were determined by Gene Ontology enrichment of the differentially expressed proteins followed by mapping protein-protein interaction network. The key hub/bottleneck proteins were identified by topological analysis of protein-protein interaction network followed by KEGG pathway enrichment analysis.3. To explore the clinical relevance of ESCC biomarker candidates including Fibronectin, Transgelin, Integrinβ1, 14-3-3ζ identified by the present study. Materials and Methods:1. Proteins were extracted from 10 cases of ESCC and adjacent tissues(collected from LinZhou Tumor Hospital) and protein concentration was measured by the Bradford method. iTRAQ labeling was performed on each pool of ESCC and adjacent non-ESCC comprising 10 individual ESCC and its counterpart, respectively followed by MS/MS analysis. The criteria of protein identification included different ratio ≥ 2 and unique peptides ≥ 2.2. One hundred seven seed proteins was selected in accordance with ten characteristics of tumor biology after GO functional enrichment analysis performed on BinGo plugin in Cytoscape. The seed proteins were searched STRING(http://www.string-db.org/) to build protein- protein interaction networks. By virtue of HUBBA website(http://hub.iis.sinica.edu.tw/Hubba/index.php), hub and bottleneck nodes were identified by "degree(DEGREE)" and "bottleneck(BOTTLE NECK)" algorithm followed by KEGG analysis by CuleGo plugin. According to co-expression and syngergy of cancer hot spots, the ESCC biomarker candidates were identified.3. Western blot analysis was used to verify the expressions of biomarker candidates for ESCC. Separation of equal amount proteins was carried out by SDS-PAGE and transferred to PVDF membrane. After incubation of PVDF membrane with first and secondary antibodies, ECL was used to detect the target protein expression followed by semi-quantitative analysis.4. Software SPSS 17.0 for Windows was used to evaluate the statistical significance with P < 0.05 considered statistically significant. Results1. The present study identified a total of 243 proteins with different expression, which included 119 up-regulated proteins and 125 down-regulated proteins.2. GO enrichment analysis of 243 differentially expressed proteins identified 123 GO terms with the "hypergeometric test" analysis and "FDR" correction for multiple testing(P <0.0001). Among them, 17 GO terms strongly associated with tumor characteristics was selected as seeds for subsequent protein-protein interaction network establishment by searching STRING. The network included a total of 117 proteins and 255 interactions.3. After identification of 36 key hub and bottlenecks proteins by network topology analysis, KEGG pathway analysis showed that the following pathways were enriched including focal adhesion, proteasome, extracellular matrix receptor interaction, hypertrophic cardiomyopathy, dilated cardiomyopathy, pathogenic E.coli infections, amoebiasis, leukocyte transendothelial migration, myocardial contractility, arrhythmogenic right ventricular cardiomyopathy, complement and coagulation cascade. The nearest neighbours to evidence-proved cancer proteins including Fibronectin, Transgelin, Integrin β1, 14-3-3ζ was regarded as ESCC biomarker candidates.4. Western blot results showed that Fibronectin and Integrin β1 were up-regulated whereas Transgelin was down-regulated remarkable in ESCC compared with adjacent non-tumor tissue. The expression of 14-3-3ζ showed a tendency of down-regulation in ESCC without statistical significance. Conclusions 1 The ESCC candidate biomarkers identified in the present study were correlated with ESCC but its clinical significance and mechanisms implicated in the evolution of ESCC needs further investigation to elucidate. |
PDF Full Text Request