Metabolic profiling by Nuclear Magnetic Resonance (NMR) spectroscopy is a key technique in the field of stratified medicine and has been used to examine changes in concentrations of metabolites present in biofluid samples from different clinical or population groups. In addition, NMR provides a unique opportunity to measure metabolite mobility, e.g. by measuring diffusion properties using Diffusion-Ordered NMR Spectroscopy (DOSY) in a complex biological mixture. The aim of this work was to evaluate the potential of the DOSY experiment as an addition to the currently used metabolic profiling toolbox.
A standard DOSY pulse sequence suitable for analysing simple mixtures of metabolites was optimised for the analysis of human urine. For the analysis of human blood plasma, a relaxation-edited DOSY pulse sequence was optimised to select small metabolite signals, while a diffusion-edited pulse sequence DOSY was developed to select macromolecule signals. These experiments have been tested under various conditions and optimised for high-throughput metabolic profiling on human biofluids.
To characterise the effects of sample matrix on the diffusion coefficient profile, titrations of protein and glycerol were performed into a model solution of blood metabolites. It was possible to determine which metabolites were interacting with the protein based on the changes in their diffusion coefficient. The value of DOSY in metabolite identification was also demonstrated, and the experiment applied to distinguish between similar metabolites in urine fractions.
Further experiments were carried out using a set of blood serum samples from a sepsis trial, with the aim of using DOSY to differentiate between clinical groups based on metabolite diffusion coefficients. DOSY spectra were acquired and analysed to determine diffusion coefficient markers differentiating survivors from non-survivors of sepsis. This thesis demonstrates the value of DOSY experiments in metabolic profiling and the potential of diffusion coefficients as markers for clinical research.