Abstract

Development of Methods for Global and Targeted Metabolomics Analysis of Biological Fluids Dmitri Sitnikov, Ph.D.
Concordia University, 2021
Protein precipitation with organic solvents is the most common extraction method used for global metabolomics of human plasma. This standard method has broad selectivity but does not provide good coverage of low abundance metabolites. Increasing the metabolite coverage by parallel analysis of plasma samples prepared using different organic solvents is inefficient due to the low orthogonality of their selectivity. Moreover, there is an increasing demand for more selective sample preparation methods suitable for targeted and untargeted metabolomics analysis of biofluids that can also provide better coverage of metabolites missed by current standard workflows. To achieve these goals, a novel sequence of orthogonal but complementary extraction methods was designed and evaluated after performing a quantitative systematic side-by-side comparison of seven extraction protocols with different selectivity to establish the most orthogonal combinations with good analytical performance. A novel sequential solid-phase extraction protocol was introduced to fractionate the metabolome into anion, cation, neutral, and zwitterion fractions, followed by a rigorous analytical assessment. The improvement in metabolome coverage in metabolomics study could be achieved also by analyses of alternative samples such as oral fluid. Therefore, the final goals were to evaluate oral fluid as a sample source for targeted metabolomics analysis via the development and validation of a liquid chromatography with tandem mass spectrometry assay and accumulate sufficient background knowledge to handle this sample type in future developments.
The execution of the first objective, i.e., the side-by-side comparison of solvent precipitation methods confirmed their low orthogonality and was accompanied by severe matrix effects in comparison to more selective methods. The comparison of solid-phase extraction (SPE) and liquid-liquid extraction methods revealed their orthogonal metabolite coverage to other methods. However, further experiments showed that methyl tert-butyl ether incompletely removed lipids, which resulted in the significant and undesirable splitting of lipids between aqueous and organic layers. Thus, the final SPE sequential fractionation protocol combined sample deproteinization by methanol followed by the sequential SPE executed on mixed-mode strong anion-exchange and mixed-mode strong cation-exchange polymeric sorbents. This produced four fractions enriched in metabolites with distinct ionic properties providing flexibility to analyze the fraction individually or in strategic combinations tailored to various mass spectrometry methods. The final protocol resulted in a 1.6-fold increase in total metabolite coverage compared to methanol precipitation with a 2-fold increase in the analysis time. The method demonstrated excellent signal repeatability for both targeted (relative standard deviation < 13%) and global (relative standard deviation < 30% for 75% metabolites) metabolomics. Moreover, excellent separation of anion and cation metabolites (< 4% overlap) and small (< 27%) overlap between other pairs of metabolite classes allow supplementary assignment of ionic properties to unknown metabolites that can aid metabolite identification. The utility of the above sequential solid-phase extraction method for fractionation of the polar metabolome can, in future, extend beyond plasma to other biofluid types, such as oral fluid. Oral fluid allows portable sample acquisition and reflects the composition blood at least for several metabolites. Moreover, the similar first step (methanol precipitation) between two methods promised easier and faster transfer of the solid-phase extraction protocol from plasma to oral fluid. Two model analytes cortisone and cortisol were selected for a preliminary study in oral fluid to serve as standard metabolites in the future implementation of solid-phase extraction protocol. Their selection was driven by their biological and clinical role and a commercial availability of stable isotope labelled standards. Methanol precipitation and a short reversed-phase chromatography run were combined with tandem mass spectrometry analysis. The successful development and validation of the method revealed an accurate, precise (< 15% variability), and sensitive (low limit of quantitation = 0.31 ng/mL) high-throughput assay, whose selectivity outcompeted an immunoaffinity method, while requiring only 30 µL of oral fluid sample.
In summary, this thesis establishes the first steps towards the adaptation of the sequential solid-phase extraction method for simultaneous global and targeted metabolomics analyses of plasma and oral fluid. Such studies will benefit from the systemic, more in-depth characterization of metabolite status, enriched pathway coverage, and faster identification and transition of putative biomarkers from the discovery to the targeted stage. Therefore, my research provided the necessary groundwork for developing and validating universal preparation and analysis workflows, which will permit the execution of pluripotent metabolomics studies.
.