Browsing by Issue Date, starting with "2019-07-08"
Now showing 1 - 1 of 1
Results Per Page
Sort Options
- Non-invasive, innovative and promising strategy for breast cancer diagnosis based on metabolomic profile of urine, cancer cell lines and tissuePublication . Silva, Catarina Grace Sousa Luís; Câmara, José de Sousa; Tomás, Helena Maria Pires GasparThe work presented in this thesis aimed to establish the metabolomic profile of urine and breast cancer (BC) tissue from BC patients (samples cordially provided by Funchal Hospital), in addition to BC cell lines (MCF-7, MDA-MB-231, T-47D) as a powerful strategy to identify metabolites as potential BC biomarkers, helping on the development of non-invasive approaches for BC diagnosis and management. To achieve the main goal and obtain a deeper and comprehensive knowledge on BC metabolome, different analytical platforms, namely headspace solid-phase microextraction (HSSPME) combined with gas chromatography-quadrupole mass spectrometry (GC-qMS) and nuclear magnetic ressonance (1H NMR) spectroscopy were used. The application of multivariate statistical methods - principal component analysis (PCA) and orthogonal partial least square – discriminant analysis (OPLS-DA), to data matrix obtained from the different target samples allowed to find a set of highly sensitive and specific metabolites metabolites, namely, 4-heptanone, acetic acid and glutamine, able to be used as potential biomarkers in BC diagnosis. Significant group separation was observed in OPLS-DA score plot between BC and CTL indicating intrinsic metabolic alterations in each group. To attest the robustness of the model, a random permutation test with 1000 permutations was performed with OPLS-DA. The permutation test yielded R2 (represents goodness of fit) and Q2 values (represents predictive ability) with values higher than 0.717 and 0.691, respectively. Several metabolic pathways were dysregulated in BC considering the analytical approaches used. The main pathways included pyruvate, glutamine and sulfur metabolisms, indicating that there might be an association between the metabolites arising from the type of biological sample of the same donor used to perform the investigation. The integration of data obtained from different analytical platforms (GC-qMS and 1H NMR) for urinary and tissue samples revealed that five metabolites (e.g., acetone, 3-hexanone, 4-heptanone, 2methyl-5-(methylthio)-furan and acetate), were found significant using a dual analytical approach.