Application of liquid chromatographymass spectrometry and chemometrics in the automated characterization of molecular lipid species
Not peer reviewed
MetadataShow full item record
Lipidomics is an important field that has attracted extensive interest worldwide, due to the increasing awareness of crucial lipid functions in biological systems. Lipidomics aims at detecting, characterizing and quantifying lipid species comprehensively. In the work for the present thesis, analytical strategies based on liquid chromatography-mass spectrometry (LCMS) and chemometrics were developed for characterization of molecular species of major lipid classes, i.e. triacylglycerols (TAG) and glycerophospholipids (GPL) from marine oils and biological systems.
The applicability of liquid chromatography electrospray tandem mass spectrometry (LC-ESIMS2) for the structural characterization of naturally occurring TAG in cod liver oil was investigated. A computational algorithm was developed to automatically interpret mass spectra and elucidate TAG structures, and the results of the algorithm were compared against the lipase benchmark method. It was proved that LC-ESI-MS2 provides a suitable and powerful strategy for the structural characterization of TAG in cod liver oil.
The thesis also evaluates different strategies for differentiating marine oils by means of principal component analysis (PCA). The TAG composition and four different types of data, including total ion current (TIC) and total mass spectral (TMS) profiles derived from LC-ESIMS and LC-ESI-MS2, were used as the datasets for PCA. The results show that using the tandem TMS profiles from LC-ESI-MS2 experiments was the most rapid and convenient approach for the differentiation of the various marine and plant oils investigated, and for the representation of the characteristic TAG patterns.
The thesis proposes a least square spectral resolution (LSSR) approach for the automated characterization and deconvolution of the main GPL species, i.e., phosphatidylcholine (PC) and phosphatidylethanolamine (PE) in biological extracts. Class-specific scanning methods, such as precursor ion scanning and neutral loss scanning, in LC-MS were applied to acquire the lipidomic dataset. The methodology is based on least squares resolution of spectra and chromatograms from theoretically calculated mass spectra with the isotope distribution. The described algorithm was able to resolve PC and PE species of reference mixtures, porcine brain sphingomyeline, cod and mouse brain lipid extracts.
Recent advances in high-resolution mass spectrometry have revolutionized the lipidomics field by providing high-resolution data. The LSSR methodology was further extended to be compatible with this type of data for an accurate identification and quantification of lipid species. The methodology has been expanded to cover the analysis of other major lipid classes such as GPL, sphingolipids, glycerolipids. Examples for the analysis of natural lipids extracts from egg, porcine brain and bovine liver are presented. The flexibility of the methodology allows supporting more lipid classes and more data interpretation functions, which in turn makes LSSR a promising tool for lipidomic data analysis.
LSSR methodology was applied on LC-MS data to evaluate the effects of methylmercury (MeHg) and EPA on intact PC and PE species in mouse brain. The effects of EPA and MeHg on PC and PE composition in brain were evaluated by PCA and ANOVA. The results demonstrate that EPA reduces the levels of arachidonic acid (AA) containing PC and PE species in brain, while MeHg tends to elevate the levels of AA containing PC and PE species. EPA also significantly increases the levels of n-3 polyunsaturated fatty acids (PUFA) containing PC and PE species in brain. The results indicate that EPA may counteract the alterations of the PC and PE pattern induced by MeHg, and thus alleviate MeHg neurotoxicity in mouse brain through the inhibition of AA-derived pro-inflammatory factors.
The LSSR methodology was further applied to evaluate the effects of MeHg and EPA on the PC and PE composition in mouse liver and plasma by PCA and ANOVA in conjunction with biological and toxicological analyses. Similar to results from brain, EPA significantly elevates the levels of PC and PE species that contain n-3 PUFA and reduces the levels of PC and PE species that contains AA. MeHg increases the levels of PC and PE species with AA to a lower extent. MeHg induces more prostaglandin E2 and less prostaglandin E3, thus increasing proinflammatory factors, while EPA displays the ability to decrease the AA-derived inflammatory factors. The histological analysis of cell damage and necrosis and the measurements of biochemical indexes also indicate that MeHg induced chronic inflammatory symptoms in mice, and that EPA can alleviate the MeHg-induced hepatic toxicity. Collectively, EPA may have protective effects against MeHg-induced toxicity in mice due to the favourable modification of membrane phospholipid composition and the inhibition of inflammatory factors release.
In summary, the described strategies and algorithms represent promising tools for the analysis of TAG and GPL species in oils, fats and biological systems. The application of these methodologies on different objects can provide insights into various research areas, such as food and nutrition, health, pharmacology and toxicology.