1H-NMR-based metabolomic applications in biomedical research

NMR spectroscopy represents a powerful, versatile and reproducible technique for the analysis of complex biological matrices. In fact, virtually, all biologically relevant molecules are characterized by at least one NMR signal with a specific intensity, frequency (or chemical shift) and magnetic relaxation properties, all reflecting the chemical environment surrounding the detected nucleus. In a high-throughput vision of metabolomic analysis, the very high reproducibility, the minimal sample preparation required, and the possibility to simultaneously detect all metabolites presenting NMR active nuclei, make NMR spectroscopy one of the most suitable techniques for the analysis of any type of biological matrix, enabling the rapid and global evaluation of an NMR spectrum in its entirety or the determination of the concentrations of all metabolic features that are above the μM detection limit. The NMR versatility allowed a wide variety of metabolomic applications in life science research, especially for both human and veterinary biomedicine. As metabolites indicate intermediate and end-points of gene expression and cell activity, under the combined influence of external stimuli, metabolomics can provide a holistic approach to understand the phenotype of a certain biological system, holding promises for both clinical and precision medicine. In this context, the presented thesis aims at demonstrating the potential of untargeted NMR-based metabolomic approach in biomedical research, addressing different topics, mainly regarding the use of untargeted NMR-based metabolomic on body fluids to disentangle characteristic fingerprints and/or metabolic markers for different types of both human and animal diseases or healthy conditions; also with the aim of paving the way to personalized individual’s healthcare. Considering the occasional misunderstandings present in the literature about the different aims of “fingerprinting” and “profiling” approaches of the untargeted analysis, and the different tools to achieve them, this thesis also proposes a study where we demonstrate that the criticism on the main drawbacks of the commonly used bucketing procedure of NMR spectra are not valid when an untargeted metabolomic analysis is planned via a fingerprinting approach. In conclusion, the results presented in this thesis contribute to the demonstration that untargeted NMR-based metabolomics, coupled with biochemistry, analytical chemistry, bioinformatic tools and statistical analysis, can be considered as a comprehensive analytical technique with reasonable and actual prospects of being implemented in biomedical research.