MSACL 2019 EU Abstract

Enhanced Mass Spectrometric Profiling of the Human Blood Exposome Using an Optimised Dispersive Solid Phase Extraction Protocol

Mark David (1), Ada H.Y. Yuen (2), Stephane Camuzeaux (3), Caroline J. Sands (4), Toby J. Athersuch (5), Zoltan Takats (6), Matthew R. Lewis (7)
(1) MRC-NIHR National Phenome Centre, Department of Surgery and Cancer, Imperial College London, IRDB Building, Du Cane Road, London W12 0NN, UK, (2) Biomolecular Medicine, Department of Surgery and Cancer, Faculty of Medicine, Imperial College London, Sir Alexander Fleming Building, South Kensington, London, SW7 2AZ, UK

Warning: Undefined variable $headshot in /var/www/html/view_abstract/view_abstract_in_program.php on line 704

Mark David (Presenter) University of New South Wales >> POSTER (PDF)

Presenter Bio: Mark David joined the National phenome centre in 2013 as a technician and is responsible for generating urinary, plasma or serum metabolic signatures using Liquid Chromatography-Mass Spectrometry platforms. He’s interests in the work led him to undertake a PhD at Imperial college London and is now in he’s final year. Mark’s main research interests lie with xenometabolome profiling approaches for improved human population phenotyping –using MS based platforms to study aspects of human health and disease. This area of research stemmed from 8 years working as a Forensic toxicologist in Australia. There he developed targeted methods on UPLC-MS for drugs of abuse. The effects of these drugs on certain metabolites led him to the field of metabolomics and he’s studies, and interests are a part of ongoing efforts to develop tools and explore the xenometabolome.

Relevant Financial Disclosures (within past 24 months) Salary MRC-NIHR National Phenome Centre Conflict mitigation NOT REQUIRED.

INTRODUCTION: The untargeted nature of metabolomics allows measurement of biofluid chemistry related to both endogenous metabolism and host-environment exposures (i.e. the exposome). Comprehensive coverage of chemically diverse metabolites present in human blood products benefits from the use of multiple methods, each oriented toward a small molecule subset generally segregated by polarity and hydrophobicity. Whilst recent developments in LC-MS profiling methodologies have delivered numerous solutions for the analysis of polar molecules (e.g. via HILIC-MS) and complex lipids, the analysis of moderately hydrophobic and amphipathic molecules in blood products (including much of the exposome) by RPC methodology, is complicated by the suppressive effects of lipids on the ionisation of low molecular weight (LMW) metabolites. Efficient and inexpensive solutions are required for the separation of small molecules from the remaining sample matrix fit for large scale and high throughput applications.
OBJECTIVES: To develop and optimise a blood preparation protocol enabling the use of a coordinated suite of analytical assays, specifically removing lipids and protein efficiently and inexpensively, with minimal effect on other LMW metabolites to enable RPC-UPLC-TOFMS profiling.
METHODS: A lipid removal sample preparation technique was developed using a novel dispersive solid phase extraction (DSPE) technique. Factors optimised to facilitate lipid removal with minimal loss of other analytes included design of experiment(DOE) protocols for aspects of the DSPE sorbent specification, and the solvent composition used for extraction. The feasibility and robustness of the extraction methodology explored on an exemplar epidemiological plasma dataset (n=285) using well-characterised RP-UPLC-TOFMS phenotyping for high resolution detection of chemical species and data processing pipelines.
RESULTS: A rapid sample preparation method using DSPE for the removal of lipids was optimised for human blood products. DSPE provided a straightforward reproducible approach which enabled the use of uncompromised RPC-UPLC-TOFMS to complement the coverage provided by HILIC and lipid analyses. UPLC-TOFMS data on the exemplar study was processed using XCMS software resulting in 3093 detected metabolite features. Repeated observation of specific reference features from pooled QC samples throughout an analytical batch demonstrated mean retention time RSD <0.3% and mean peak area RSD <5% with no post normalisation. Finally, to demonstrate the methods suitability to explore the exposome, both endogenous metabolites and known xenobiotics were annotated, and their population prevalence reported.
CONCLUSION: The optimised lipid removal method described enables improved characterisation of the human exposome using high-throughput metabolic phenotyping platform. Additional advantages include reduced cost and increased robustness when compared to conventional solid-phase sample clean-up protocols.