Principal component analysis on LC-MS/MS and 2DE-MALDI-TOF in glioblastoma cell lines reveals that mitochondria act as organelle sensors of the metabolic state in glioblastoma

Glioblastoma is a difficult disease to diagnose. Proteomic techniques are commonly applied in biomedical research, and can be useful for early detection, making an accurate diagnosis and reducing mortality. The relevance of mitochondria in brain development and function is well known; therefore, mitochondria may influence the development of glioblastoma. The T98G (with oxidative metabolism) and U87MG (with glycolytic metabolism) cell lines are considered to be useful glioblastoma models for studying these tumors and the role of mitochondria in key aspects of this disease, such as prognosis, metastasis and apoptosis. In the present study, principal component analysis of protein abundance data identified by liquid chromatography coupled to tandem mass spectrometry (LC-MS/MS) and matrix-assisted laser desorption/ionization-time of flight mass spectrometry (MALDI-TOF) from 2D gels indicated that representative mitochondrial proteins were associated with glioblastoma. The selected proteins were organized into T98G- and U87MG-specific protein-protein interaction networks to demonstrate the representativeness of both proteomic techniques. Gene Ontology overrepresentation analysis based on the relevant proteins revealed that mitochondrial processes were associated with metabolic changes, invasion and metastasis in glioblastoma, along with other non-mitochondrial processes, such as DNA translation, chaperone responses and autophagy. Despite the lower resolution of 2D electrophoresis, principal component analysis yielded information of comparable quality to that of LC-MS/MS. The present analysis pipeline described a specific and more complete metabolic status for each cell line, defined a clear mitochondrial performance for distinct glioblastoma tumors, and introduced a useful strategy to understand the heterogeneity of glioblastoma.