A unique four-hub protein cluster associates to glioblastoma progression

• Published in: PloS one Vol. 9; no. 7; p. e103030
• Main Authors: Simeone, Pasquale, Trerotola, Marco, Urbanella, Andrea, Lattanzio, Rossano, Ciavardelli, Domenico, Di Giuseppe, Fabrizio, Eleuterio, Enrica, Sulpizio, Marilisa, Eusebi, Vincenzo, Pession, Annalisa, Piantelli, Mauro, Alberti, Saverio
• Format: Journal Article
• Language: English
• Published: United States Public Library of Science 22.07.2014; Public Library of Science (PLoS)
• Subjects: Aging; Biology and Life Sciences; Biomarkers, Tumor - analysis; Biomarkers, Tumor - metabolism; Brain; Brain - metabolism; Brain - pathology; Brain cancer; Brain Neoplasms - metabolism; Brain Neoplasms - pathology; Brain research; Brain tumors; c-Myc protein; Cell cycle; Classification; Cluster Analysis; Clustering; Crosstalk; Diagnostic systems; Discriminant analysis; Electrophoresis, Gel, Two-Dimensional; Epidermal growth factor receptors; Evaluation; Gene expression; Glioblastoma; Glioblastomas; Glioma; Glioma - metabolism; Glioma - pathology; Heterogeneity; Histopathology; Humans; Huntingtin; Identification; Immunohistochemistry; Mass Spectrometry; Mass spectroscopy; Medical treatment; Medicine; Medicine and Health Sciences; Multivariate Analysis; Mutation; Myc protein; Nervous system; Network hubs; p53 Protein; Pathogenesis; Pathology; Principal components analysis; Protein Interaction Maps; Proteins; Proteome - analysis; Proteome - metabolism; Proteomics; Proteomics - methods; Science; Signal Transduction; Signaling; Tumor proteins; Tumors; Two dimensional analysis; Western blotting; United States > US; Italy
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0103030

Gliomas are the most frequent brain tumors. Among them, glioblastomas are malignant and largely resistant to available treatments. Histopathology is the gold standard for classification and grading of brain tumors. However, brain tumor heterogeneity is remarkable and histopathology procedures for glioma classification remain unsatisfactory for predicting disease course as well as response to treatment. Proteins that tightly associate with cancer differentiation and progression, can bear important prognostic information. Here, we describe the identification of protein clusters differentially expressed in high-grade versus low-grade gliomas. Tissue samples from 25 high-grade tumors, 10 low-grade tumors and 5 normal brain cortices were analyzed by 2D-PAGE and proteomic profiling by mass spectrometry. This led to identify 48 differentially expressed protein markers between tumors and normal samples. Protein clustering by multivariate analyses (PCA and PLS-DA) provided discrimination between pathological samples to an unprecedented extent, and revealed a unique network of deranged proteins. We discovered a novel glioblastoma control module centered on four major network hubs: Huntingtin, HNF4α, c-Myc and 14-3-3ζ. Immunohistochemistry, western blotting and unbiased proteome-wide meta-analysis revealed altered expression of this glioblastoma control module in human glioma samples as compared with normal controls. Moreover, the four-hub network was found to cross-talk with both p53 and EGFR pathways. In summary, the findings of this study indicate the existence of a unifying signaling module controlling glioblastoma pathogenesis and malignant progression, and suggest novel targets for development of diagnostic and therapeutic procedures.