Search

Enhanced BDNF expression has been implicated in the mechanisms of action of antidepressant treatment. Previous findings have shown that chronic antidepressant drug treatment produces a ’bi-phasic effect’ on total BDNF gene expression, i.e. down-regulation at 4h and up-regulation at 24h. The earlier effect is common to acute administration, while the latter is unique to chronic treatment.

Despite recent advances, the prognosis of glioblastoma (GBM), the most common brain cancer, remains poor. Evidence suggests that the epidermal growth factor receptor (EGFR)/matrix metalloproteinase-9 (MMP9) axis is implicated in gliomagenesis. Recent studies have shed light on EGFR interaction with the highly-glycosylated transmembrane mucin 4 (MUC4) with implication in carcinogenesis. MUC4 expression in glioma grades and its link with EGFR/MMP9-driven glioma progression have not been studied so far. Changes in the glycosylation pattern through alterations of glycoproteins, such as mucins, in brain cancer represent potential biomarkers in glioma diagnosis and monitoring. In this thesis, EGFR, MUC4, and MMP9 tissue expression was investigated to assess their utility as a panel of biomarkers. Secondly, this thesis aimed to define the molecular mechanisms underlying their role in GBM. Lastly, Raman spectroscopy (RS) with combined bioinformatic analyses were used to monitor changes in glycosylation for glioma diagnosis and monitoring. Immuno-staining techniques and ELISA were used in fixed tissue samples and blood serum samples. MUC4 was overexpressed in glioma grades, and especially in GBM, with MMP9. This expression was found in neoplastic cells and in the microvasculature. The two proteins combined had prognostic value. In-vitro assays using cellular models was employed to define the molecular mechanisms of the pathway and suggested that MUC4 might be involved in cancer cell dissemination and may interact with EGFR in glioma cells. Gene expression analysis supported the involvement of the three proteins in a common pathway. The utility of Raman spectroscopy combined with classification learning was lastly demonstrated on cellular models. This methodology showed a promising high sensitivity potential for the detection of subtle glycoprotein patterns in cell lines. Importantly, this approach was found useful in discriminating glioma grades with high accuracy in tissue and serum samples. Discrimination was possible owing to a significant change in glycosylation.