An integrative approach based on GC–qMS and NMR metabolomics data as a comprehensive strategy to search potential breast cancer biomarkers

• Autor(a) principal: Silva, Catarina Luís
• Data de Publicação: 2021
• Outros Autores: Perestrelo, Rosa, Capelinha, Filipa, Tomás, Helena, Câmara, José S.
• Tipo de documento: Artigo
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10400.13/4703
• topic: Breast cancer; Tissue; Urine; NMR; MS; Metabolomics; Chemometric tools; .; Faculdade de Ciências Exatas e da Engenharia; Centro de Química da Madeira

Introduction Globally, breast cancer (BC) is leading at the top of women's diseases and, as a multifactorial disease, there is the need for the development of new approaches to aid clinicians on monitoring BC treatments. In this sense, metabo lomic studies have become an essential tool allowing the establishment of interdependency among metabolites in biological samples. Objective The combination of nuclear magnetic resonance (NMR) and gas chromatography–quadrupole mass spectrometry (GC–qMS) based metabolomic analyses of urine and breast tissue samples from BC patients and cancer-free individuals was used. Methods Multivariate statistical tools were used in order to obtain a panel of metabolites that could discriminate malignant from healthy status assisting in the diagnostic feld. Urine samples (n=30), cancer tissues (n=30) were collected from BC patients, cancer-free tissues were resected outside the tumor margin from the same donors (n=30) while cancer-free urine samples (n=40) where obtained from healthy subjects and analysed by NMR and GC–qMS methodologies. Results The orthogonal partial least square discriminant analysis model showed a clear separation between BC patients and cancer-free subjects for both classes of samples. Specifcally, for urine samples, the goodness of ft (R2 Y) and predictive ability (Q2 ) was 0.946 and 0.910, respectively, whereas for tissue was 0.888 and 0.813, revealing a good predictable accuracy. The discrimination efciency and accuracy of tissue and urine metabolites was ascertained by receiver operating charac teristic curve analysis that allowed the identifcation of metabolites with high sensitivity and specifcity. The metabolomic pathway analysis identifed several dysregulated pathways in BC, including those related with lactate, valine, aspartate and glutamine metabolism. Additionally, correlations between urine and tissue metabolites were investigated and fve metabo lites (e.g. acetone, 3-hexanone, 4-heptanone, 2-methyl-5-(methylthio)-furan and acetate) were found to be signifcant using a dual platform approach. Conclusion Overall, this study suggests that an improved metabolic profle combining NMR and GC–qMS may be useful to achieve more insights regarding the mechanisms underlying cancer.