| dc.description.abstract | Although obesity and prediabetes have been long-established risk factors for type 2 diabetes (T2D), excess fat deposition in visceral and ectopic organ sites (i.e. the “risky” fat depot) has been increasingly recognised as key conduits for T2D. However, quantification of these fat depots relies on expensive and time-consuming imaging techniques. There is a need for identification of biomarkers predictive of risky fat depot levels.
Metabolomics is a promising tool for discovering novel markers and generating mechanistic speculation. This PhD project aims to identify and understand plasma metabolite markers associated with metabolic risk factors including elevated fasting plasma glucose (FPG), fat deposition in visceral (VAT) and ectopic organ sites (liver and pancreas) in non-T2D human. To achieve this, a workflow to select the optimal injected concentration of different sample types for two complementary LC-MS untargeted analyses (polar metabolites and lipidomics) was established and applied to examine the value of measuring the plasma metabolome as a proxy for metabolite concentrations in various tissue sites including adipose tissue, muscle and the liver. Lastly, plasma metabolomic signatures for elevated FPG and VAT were characterised and metabolite markers predictive of VAT and ectopic fat deposition in the liver and pancreas were identified.
This PhD study highlighted the critical importance in optimising injected concentration for LC-MS analysis of different sample types to ensure the maximal number of linear features were obtained, and for the first time showed the plasma metabolomics profile was more reflective of the liver profile than muscle or adipose tissue. Subsequent metabolomics characterisation of clinical plasma samples reported profound associations between FPG or VAT with changes in several glycerolipid species independent of gender, ethnicity, age and body mass index (BMI). VAT was additionally associated with changes in phospholipid, ether-linked phospholipid and sphingolipid species independent of covariates. Liver fat deposition was predicted by a number of glycerolipid, phosphatidylethanolamine and dihydroceramide lipid species whose plasma concentrations were linearly correlated with the liver counterparts. A novel marker, sulfolithocholic acid, for the prediction of pancreatic fat independent of age, BMI and visceral adiposity was also identified. Finally, the study also reported an improved prediction of ectopic fat deposition by utilising a panel of metabolite markers compared to clinical measurements, and demonstrated the usefulness of the metabolomic signature to identify a subset of normoglycaemic individuals with a worse cardiometabolic profile. Findings from this PhD study highlighted the value of metabolomics as a promising tool to capture metabolic risk, and that candidate markers identified by metabolomics may offer opportunities for improved risk prediction and stratification, disease progression monitoring and to develop alternative means for the measurement of the effectiveness of dietary interventions. | en_US |
