Development of a nanoLC-MS system for global lipidomics

Abstract

In the context of this thesis, an advanced nano-flow liquid chromatography (nLC) – nanoelectrospray (NSI) - mass spectrometry (MS) system was developed for a global (i.e. not lipid class- or species-restricted) lipidomics approach. In order to grant broad range analyses of different lipid classes and species, the loading- and sample preparation procedures were particularly optimized. A direct sample injection with the resuspension mixture of BuOH:IPA:H2O 8:23:69 + 5 mM phosphoric acid (8Bu+) thereby proved to be suitable for global lipidomics experiments. In this regard, the addition of phosphoric acid to the samples strongly enhanced the general detectability of lipids with terminal phosphate groups, which were found to be hardly analyzable under standard reversed-phase LC conditions. The robustness of the nLC-MS method was then evaluated based on repetitive analyses of a yeast extract that was spiked with standards. Over a measuring period of two days, high retention time stabilities were achieved for all analyzed lipids. The bulk of phospholipids could be stably detected with relative standard deviations below 10 %, while the MS-response was less stable for late eluting very hydrophobic lipids. Using normalization approaches, these deviations could nevertheless be reduced with structurally- or retention time-related standards. A direct sensitivity comparison to a corresponding narrow-bore HPLC platform revealed an increase of 2-3 orders of magnitude for the MS response by application of the nLC system that conformed with an increase of up to two orders of magnitude for the linear dynamic range. Sub-femtomol traces of individual lipids (< nM concentrations) could thus be detected with the nLC-MS platform in both positive and negative mode analyses. In order to permit a thorough comparison of the lipidome coverage, a semi-autonomous workflow was developed that allowed an accurate identification of individual (molecular) lipid species. The presumably simple yeast (Saccharomyces cerevisiae) lipidome was then selected as a model for the in-deep evaluation of the lipidome coverage. The direct comparison of the yeast phospholipidome to the corresponding narrow-bore HPLC revealed a more than three-fold increase in lipid identifications on molecular species level when identical sample concentrations were comparatively analyzed with both LC systems. All in all, 935 lipid species, with each comprising multiple molecular species isomers, were identified across 30 analyzed lipid classes from 4 lipid categories in yeast. Even though this thesis uncovered more endogenous yeast lipids than any reference study, a detailed analyses of isomeric signals further indicates that also novel lipid classes, such as headgroup-methylated PS variants, and oxidized lipids can be found in natural yeast lipid extracts. Overall, this thesis shows that the insights into the complexity of a lipidome can be strongly deepened by using an nLC-MS system for a global lipidomics experiments.