Understanding Glioblastoma Biology and Function: A Proteomic Approach Open Access
Downloadable ContentDownload PDF
Glioblastoma multiforme is among the most common and deadly types of brain tumors affecting adults. The incurability of malignant glioblastoma is mainly attributed to its highly invasive nature coupled with resistance to chemo- and radiation therapy. Tumor cell invasiveness is partially dictated by secreted proteins that can degrade the surrounding basement membrane and stroma and act as chemoattractants to promote migration. The tumor cell secretome is thus a key source of insight into the molecular mechanisms of tumor cell invasion and for defining potential novel therapeutic targets and associated biomarkers. We used stable isotope labeling by amino acids in cell culture to characterize the secretomes of four human glioblastoma cell lines exhibiting different degrees of invasiveness. The most invasive glioblastoma cell line, U87, had a unique secretome signature which set it apart from the less invasive glioblastoma cell lines T98, LN18 and U118. Using Pavlidis template matching we identified several proteins whose expression levels correlated with U87 cell invasiveness. The most striking of these was chitinase-3-like protein 1 (CH3L1), with CH3L1 representing more than 3.6% of total U87 secreted proteins. Antibody neutralization of secreted CH3L1 reduced U87 cell invasiveness by 30% indicating a role for this protein in glioblastoma invasiveness. CH3L1 was secreted as different glycoform species with acidic and basic isoelectric points. These glycovariants were modified by either tri-antennary and tetra-antennary N-linked glycans with varying numbers of terminal sialic acid residues, or high mannose structures. Finally, a comprehensive study comparing glioblastoma to astrocyte secretomes enabled the identification of biomarkers with potential diagnostic and prognostic value. These include collagen alpha-3(VI), galectin-1 and insulin-like growth factor-binding protein 5. This study has provided informative data suggesting potential roles for a number of glioblastoma secreted proteins in the tumor cell invasion process, particularly CH3L1. In addition, by mining the unique signature of the glioblastoma secretome, we were able to identify potentially useful clinical markers. Such continuing advancements in the understanding of this disease will hopefully provide us with novel diagnostic tools and therapeutic targets to improve patient outcomes in the future.