The Effects of Nutritional Intervention on Metabolic Phenotype

Abstract

This research aimed to examine the relationship between dietary interactions, health and disease and the effects of nutritional intervention on metabolic phenotype. High throughput proton nuclear magnetic resonance spectroscopy and liquid chromatography mass spectrometry analytical platforms coupled with multivariate statistical methods were utilised to identify metabolic changes in biological fluids related to nutritional status and intervention. These metabonomic approaches were applied initially to an animal study and subsequent human studies. The aims of this research were tackled by examination of a preliminary animal study to assess the effects of methyl eugenol, a natural dietary flavour component that is genotoxic, on rat metabolic phenotype by analysis of urinary metabolic profiles. Methyl eugenol alterations of urinary metabotype at relatively low doses (1-50mg/kg bw/day) could not be discriminated, however discrimination based upon age, sex and administration method (orogastric gavage vs. dietary matrix) could be seen. A second study examined the effects in humans of dietary fat and carbohydrate composition on the development of metabolic syndrome by analysis of urinary and faecal water metabolic profiles. Dietary intervention did not produce statistically significant alterations of human urinary and faecal water metabotype. Furthermore, regression analyses of bacterial population numbers to 1H-NMR spectroscopic data of both urine and faecal water also did not produce statistically significant modelling of the datasets. This study highlighted the need to limit inter-subject variability particularly in human nutritional intervention studies. The third metabonomic study investigated the effects of cruciferous vegetable, meat meal and caffeine dietary intervention on the human urinary metabolic phenotype by 1H-NMR spectroscopy and UPLC-MS analysis. Chemometric analysis of the 1H-NMR spectral dataset, use of two-dimensional NMR experimentation and synthetic standard urine spiking experiments identified S-methyl-L-cysteine-sulphoxide (SMCSO) and N-acetylated-SMCSO as urinary biomarkers of cruciferous vegetable consumption in man. Multivariate statistical analysis of the 1H-NMR spectral dataset identified post-prandial biomarkers of meat meal consumption and bacterial co-metabolite excretion in human urine. Targeted interrogation and multivariate statistical modelling of the UPLC-MS database identified 36 tentative biomarkers of cruciferous vegetable, meat and caffeine dietary interventions. This research demonstrated the successful application of metabonomics to the study of nutritional intervention and the identification of biomarkers of dietary exposure. Also highlighted were the challenges of biomarker identification particularly in UPLC-MS datasets and the need to limit both experimental and inter-subject variations in metabonomic studies of human biofluids.