Use this URL to cite or link to this record in EThOS:
Title: Metabolomics applied to biomarker discovery in liver related diseases
Author: Xu, Jin
Awarding Body: King's College London
Current Institution: King's College London (University of London)
Date of Award: 2019
Availability of Full Text:
Access from EThOS:
Access from Institution:
The liver is the only human internal organ capable of regeneration. It is indispensable to most metabolic functions, such as bile production, metabolism of lipids, carbohydrates, proteins and amino acids, and breakdown of hormones. Since almost all other organs in the body rely on the liver for their smooth functioning, it is not surprising that the liver is closely related to a plethora of diseases. Liver failure, cholestasis and sepsis are common liver related disorders. A major drawback related to current biomarkers for liver disorders, such as aspartate aminotransferase (AST), gamma-glutamyl transferase (GGT) and interleukins, is that they have limited diagnostic accuracy and specificity. Therefore, there is an urgent need to investigate metabolites that associate with different liver conditions, which can then potentially be utilised as biomarkers and therapeutic targets. The focus of my thesis was to apply metabolomics on human and mouse liver tissues to further understand its biochemistry under stress and facilitate the discovery of potential endogenous biomarkers under those conditions. Chapter two involved an in-depth analysis of metabolic variations in two main types of liver donor tissues. Lipid fingerprints in 112 matched pre- and post-transplant liver biopsies from donation after circulatory death (DCD, n=36) and donation after brain death (DBD, n=76) were investigated using Reversed Phase-Mass Spectrometry (RP-MS). Two lysophosphatidylcholines, LysoPC (16:0) and LysoPC (18:0), showed higher levels in DCD, before transplantation (q < 0.01). These findings imply that the two LysoPCs play a role in perpetrating tissue damage induced by warm ischaemia. Interestingly, elevated amounts of these two lipids were also observed in recipients undergoing early allograft dysfunction (EAD) (p < 0.05). After comparing DBD and DCD groups and identifying two potential lipid biomarkers of EAD, we decided to target lipids that are known to induce inflammation since ischaemia reperfusion injury (IRI) is associated with the release of pro-inflammatory mediators. In Chapter 3, a targeted analysis of five ceramides was carried out in 46 (10 DBD, 13 DCD) paired biopsies at both transplantation stages. It was found that C18, C20, C24 ceramides showed significant differences in DBD (p<0.05) pre- and post-transplantation, while C22 ceramide (p<0.05) showed more pronounced changes in the DCD group. A strong correlation between levels of C18 ceramide and bilirubin and creatinine, two common markers of liver dysfunction, suggested the involvement of C18 ceramide with IRI in DCD. In Chapter 4, we aimed to extend the lipidomics results and investigated more polar molecules. Hydrophilic interaction chromatography-MS (HILIC-MS) analysis was performed on 35 DBD and 12 DCD paired liver samples (n=94). Orthogonal Projections to Latent Structures-Discriminant Analysis (OPLS-DA) modelling suggested the involvement of five metabolites from the adenosine monophosphate (AMP) pathway. Elevated levels of AMP and adenine were indeed found in the DBD group, after univariate analysis (p<0.01). These changes were also seen in recipients experiencing immediate graft function (IGF) (p<0.05), suggesting that they may contribute to healthy allograft function and improve graft survival rate. In chapter 5, an animal model of cholestasis was studied. Untargeted lipidomic analysis and semi-targeted bile acids analysis were applied to mouse livers (7 controls, 6 cholestatic) and plasma (8 controls, 10 cholestatic) from a cholestatic model. A panel of 5 lipids, including 2 phosphatidylcholines (PC (36:3), PC (34:2)) and 3 cholesterol esters (CE (16:0), CE (16:1), CE (18:1)) showed significant differences between cholestatic and control mice (p<0.05). In addition, Receiver Operating Characteristic (ROC) curve revealed a high potential of these five lipids in discriminating cholestatic mice from controls, with area under the curve (AUC) of 100%, when compared to plasma enzymes (91.2%) and bile acids (77.5%). Lastly, my focus moved to sepsis in order to understand inflammation in the liver and if this can affect the brain via the liver-brain axis. Liver and brain samples (4 controls, 6 septic) from a septic mouse model were analysed with both RP-MS and HILIC-MS untargeted approaches. A panel of 9 annotated metabolites, including previously discussed LysoPC (16:0), CE (16:0) and CE (18:1), were uncovered through a meta-analysis which employed three machine learning tools. Subsequently, a correlation analysis was performed between this panel of metabolites and general indicators of sepsis. Based on all the statistical tools applied, glycerylphosphorylethanolamine; a lipid depleted in both liver and brain from septic mice (p<0.05); was selected as the most promising candidate biomarker for endothelial dysfunction during sepsis. Overall, this work has revealed a panel of key metabolites that are all specifically associated with a range of liver related diseases. Through comparing two types of liver donation, DBD and DCD, LysoPC (16:0) and LysoPC (18:0) were found to be potential biomarkers of liver tissue damage. In the meanwhile, adenine and AMP were discovered to be putative biomarkers of allograft function and survival. Furthermore, inflammatory lipids, ceramides, were also studied in DBD and DCD groups, diverse distributions of five chosen ceramides and correlations with liver dysfunction markers were observed. Moving from human studies to mouse experiments, in a cholestasis model, a panel of 5 lipids, including phosphatidylcholines and cholesterol esters, were proved to have higher predictive ability compared with traditional enzymes. In addition, glycerylphosphorylethanolamine was chosen as a promising biomarker of endothelial dysfunction in a mouse sepsis study involving liver-brain axis. Future work entails validating these findings in independent cohorts and/or clinical studies.
Supervisor: Legido Quigley, Cristina ; Ma, Yun Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available