A metabolomics study of selected perturbations of normal human metabolism

Abstract

Metabolism is an integrated network of biochemical pathways involving a series of enzymecatalysed anabolic or catabolic reactions in cells. Metabolites are chemical compounds that are involved in or are products of metabolic pathways, and the metabolome is defined as the total complement of all the low molecular weight metabolites present in a cell at any given time. Metabolomics is a relatively new research technology utilised for the global investigation, identification and quantification of the metabolome. Three aims were defined for the metabolomics study presented here: • The use of metabolomics technology to generate new biological information; • Application of the metabolomics technology to gain information on the three natural perturbations, namely the menstrual cycle, pregnancy and aging; and • Reflection on metabolomic studies as a hypothesis-generating approach. I obtained three sets of urine samples from women during their menstrual cycle, samples from sixteen pregnant and eleven non-pregnant women for the second natural perturbation, and data sets from previous investigations on infant and child groups, as well as thirty-two urine samples from adults for the study of the metabolomic profiles due to age. These urine samples were analysed to determine the organic acid metabolite profiles. The metabolites were identified by means of AMDIS and were manually quantified. Data matrixes were compiled, which underwent certain data reduction steps, prior to statistical analysis. Different statistical approaches were used to generate information on these three natural perturbations due to the clear differences between the three experimental groups used. The investigation of the menstrual cycle did not show a distinct difference between the three phases involved in the cycle, whereas the pregnancy perturbation showed a difference between pregnant groups and non-pregnant groups. The most pronounced difference in metabolite profiles were found when the different age groups were compared to one another. Finally a hypothesis on the effect of age on metabolism was defined and an experimental approach was proposed to evaluate this hypothesis. In conclusion three proposals were formulated from this investigation: 1. If it appears that an insufficient number of participants can be generated for a metabolomics study, such a study should be discarded in the interest of a more feasible investigation. 2. It is advisable that a number of appropriate analytical validation parameters should be incorporated in the early stages of a metabolomics study, specifically linked to the context of the perturbation chosen for the investigation. 3. The control and experimental groups should be homogenous that is to say as comparable as possible with regard to age, ethnicity, diet, and gender, lifestyle habits and other possible confounding influences, except for the specific perturbation being studied. In a perfect world this would be possible, specifically when hypothesis formulation, testing and finally the expansion of scientific knowledge is a desired outcome of the investigation.