• target therapy
• proteomics
• Glioblastoma
• cancer

Background: Glioblastoma multiforme (GBM) is the most aggressive, highly invasive and neurologically destructive among malignant brain tumors. None of the treatments currently used is effective, making GBM lethal within 12 months from diagnosis. Despite major evolution in the understanding of the molecular mechanisms involved during GBM development and progression, patients are yet in need of a successful treatment. Nowadays, one of the most promising approaches aimed to cure cancer consists on the antibody-based therapy. The systemic delivery of specific antibodies, coupled with highly cytotoxic drugs, directed towards the tumor site is considered as a promising way of treatment. Unfortunately, the bottleneck of such approach consists in the identification of antigens accessible by the blood stream.
Aim of the work: This work aims to identify and validate in-vivo new biomarkers produced either by the tumor itself or by the surrounding stroma in response to tumoral demand, which are reachable by a systemic administered compound and could be used as therapeutic targets in human GBM. Material and methods: Human GBM specimens, two different human glioblastoma cell lines (U373 and T98G) grafted in nude mice and U87-derived tumors grown in egg choriallantoic membrane (CAM) were biotinilated ex-vivo. Labeled proteins were successively isolated using affinity chromatography based on streptavidin beads. Proteins were eluted and digested with trypsin and the resulting peptides were analyzed using the 2D nano-LC MS/MS technique. Protein identification was carried out using the Mascot® search engine (Matrix Sciences, Boston, MA, USA) and the Swisprot® protein database (Swiss Institute for Bioinformatics, Basel, Switzerland). For the purpose of validation U87-derived tumors were grown on CAM. Intravenous administration of specific monoclonal antibodies into the CAM vessels was carried out in order to validate in-vivo the accessibility of the target.
Results and conclusion: expression profiles for each experimental model were determined using the Mud-PIT technique. About 30 to 35% of the total proteome were found to be accessible (membrane associated, extracellular and secreted proteins). Among the proteins identified, the study highlighted the hyaluronan receptor CD44 and tenascin-C (onco-fetal Tenascin) already known to be overexpressed in human GBM along with other new potential targets such as sparc-like 1, prosaposin and collagen 6 α1. The validation phase carried out using immunohistochemical analysis confirmed the overexpression of the proteins in high-grade gliomas. Additionally, in-vivo experiment with the systemic administration of the monoclonal anti-human CD44 and COL6α1 antibodies into CAM vessels resulted in a site-specific tumor accumulation of the antibodies suggesting these proteins as readily accessible target for the treatment of GBM. Taken together, these results demonstrate the potential of the biotinilation technique in searching for potential accessible biomarkers. The study pointed at the usefulness of the CAM system as an alternative model of biomarker validation in comparison to the more cost and labor intense mouse model. Further investigations focusing on the development of antibody-based treatment of tumor bearing animals are the next step to envision before proposing these targets for clinical trials on humans.