Murdoch University Research Repository

Welcome to the Murdoch University Research Repository

The Murdoch University Research Repository is an open access digital collection of research
created by Murdoch University staff, researchers and postgraduate students.

Learn more

Unravelling the Lipidome of Idiopathic Pulmonary Fibrosis and its Spatial Distribution using High Resolution Mass Spectrometry

Nambiar, Shabarinath (2018) Unravelling the Lipidome of Idiopathic Pulmonary Fibrosis and its Spatial Distribution using High Resolution Mass Spectrometry. PhD thesis, Murdoch University.

[img]
Preview
PDF - Whole Thesis
Download (5MB) | Preview

Abstract

Chronic lung diseases are complex, progressive disorders with increasing incidence and mortality. Chronic obstructive pulmonary disease (COPD), asthma and pulmonary fibrosis are examples of chronic lung conditions that can significantly impact the quality of life. Minimally-invasive diagnostic methods that eliminate bronchoscopic and surgical biopsy from patients are ideal; metabolomics therefore holds considerable promise for the discovery of biomarkers that can aid diagnosis and treatment with greater sensitivity, specificity and precision.

The main aim of this project was to employ ultra-performance liquid chromatography-quadrupole time-of-flight (UPLC-QTOF) high resolution mass spectrometry (HRMS) and matrix-assisted laser desorption ionisation (MALDI) mass spectral imaging (MSI) together with multivariate statistics-based metabolomics to identify and characterize potential lipid biomarkers of idiopathic pulmonary fibrosis (IPF). This dissertation consists of the following studies: (1) literature review of metabolomics in chronic lung diseases; (2) application of HRMS for untargeted metabolic profiling of chronic lung disease including COPD and IPF; (3) investigation of a novel data-independent acquisition (DIA) approach to augment untargeted approaches for lipid biomarker identification; (4) development of a novel matrix application technique to improve MALDI-MSI acquisitions of tissue sections whilst maintaining spatial localisation of endogenous metabolites; and (5) exploiting potassium adduct formation to resolve the spatial distribution of lipids in fibrotic tissues.

A total of 65 clinical plasma samples (from 20 healthy control subjects, 21 COPD and 24 IPF patients) were profiled using UHPLC-QTOF-MS. A fundamental challenge in using HRMS for untargeted profiling of complex, chronic lung diseases is the heterogeneity of the human samples. Various contaminations present in fibrotic tissues or adjacent non-fibrotic constituents can confound characterization and encumber the discovery of reliable biomarkers. The results of this study revealed significant correlation between COPD and IPF clinical phenotypes and plasma metabolite profiles. The unbiased metabolomics workflow and deconvolution pipeline provided end-to-end analysis from peak picking and annotation through to metabolite identification.

Subsequently, the ability of the UPLC-QTOF-MS method to discriminate between lipid species was enhanced by the application of a DIA method to distinguish between “stable versus progressor” IPF patients. This DIA method is known as SONAR and uses a wide, continuously sliding precursor window for fragmentation, thereby allowing correlation of precursor and fragment ions. SONAR lipid data were processed using Progenesis QI and searched against LIPID MAPS for structural elucidation and metabolite confirmation. The lipids identified were found to be intermediates of key metabolic pathways such as the glycolytic/TCA cycle, mitochondrial-beta oxidation and lipid metabolism and hold considerable promise as biomarkers of disease.

The matrix deposition step in MALDI-MSI is crucial for simultaneous extraction of metabolites from tissue sections as well as maintaining the spatial dimensionality of the endogenous metabolites. A novel, efficient and cost-effective preparative method referred to as the “freeze-spot” method was developed using wheat seed sections to demonstrate extraction efficiency and reliability, whilst maintaining the spatial resolution of the acquired MALDI-MSI images. The technique was also found to be simple and robust, forming fine matrix crystals that enabled efficient ionisation of surface metabolites, further eliminating the need for sophisticated matrix application approaches.

In the final study, 10 healthy and 10 fibrotic tissues were compared using MALDI-MSI. The MSI technique developed uses potassium adduct formation to improve spatial resolution and dimensionality of lipid species such as triglycerides (TG), ceramides, sphingolipids and glycerophospholipids. The results of this study showed changes in lipid composition of IPF tissues compared to healthy controls. This study identified lysophosphatidylcholine (LysoPC), phosphatidylcholine (PC) and phosphatidylethanolamine (PE) as potential lipid biomarkers of the disease and requires further study as targets of intervention and treatment. Both SONAR and MSI successfully identified similar classes of lipids (TG, PE, LysoPC and PC) which may play a role in the pathophysiology of the IPF lipidome.

This project highlighted the complementarity of HRMS and MSI based metabolomics for the characterization of unique lipid features in fibrotic tissue and plasma samples. The study also demonstrated the discriminative power of the unbiased DIA approach for the identification of lipids via fragment ion patterns that were indicative of specific lipid classes. In addition, the application of chemometric principal component analysis (PCA) and orthogonal partial least-squares to latent structures-direct analysis (OPLS-DA) proved useful for the identification of statistically significant lipids. This statistical approach allowed for the assessment of covariance and correlation between lipids and the modelled lung diseases, and further illustrated lipid compositional changes in chronic lung diseases.

Taken together, the experimental work presented in this thesis show the large potential for mass spectrometry-based metabolomics as a tool for discovery. The specificity of the novel methods outlined will be highly beneficial for compound identification and further confirmation of disease biomarkers.

Item Type: Thesis (PhD)
Murdoch Affiliation: School of Veterinary and Life Sciences
Supervisor(s): Trengove, Robert, Sze How, Bong, Gummer, Joel and Moodley, Yuben
URI: http://researchrepository.murdoch.edu.au/id/eprint/43984
Item Control Page Item Control Page

Downloads

Downloads per month over past year