Characterization and annotation of oxidized glycerophosphocholines for non-targeted metabolomics with LC-QTOF-MS data
Graphical abstract
Introduction
The status of an organism is governed by the activity of the cells building it, which balance biochemical reactions to maintain homeostasis. One of the crucial balances is ‘redox homeostasis’, which consists of the in vivo regulation of oxidative and reductive metabolism. Oxidation is one of the most commonly occurring reactions in a living system.
Among many endogenous oxidation processes, lipid peroxidation plays a vital role. Due to their widespread presence in all human cells, lipids are highly affected by oxidative stress. Oxidized lipids are involved in many important processes such as energy production through β-oxidation [1,2], signaling through eicosanoids [3,4] or uncontrolled oxidative degradation provoked by free radicals [5,6]. Therefore, not only intact lipids, but also their oxidized forms represent some of the most important features of mammalian biochemistry.
Currently, attention is being paid to the oxidation of glycerophospholipid (PLs) as intermediate products of oxidation. The work presented herein focuses on oxidized phosphocholines (oxPCs), therefore all subsequent statements and observations concern only to this particular class of oxidized phospholipids (oxPLs). Furthermore, among different types of oxPCs only two classes are covered (Fig. 1). They include: i) mildly oxygenated PC (class I oxPC), later called long chain-oxidized PC (LCh-oxPC), which are the products of the oxygen addition to the PC's unsaturated chain [7]. This includes oxidation via formation of hydroxyl -OH, dihydroxyl -(OH)2, peroxyl -OOH fatty acids as well as keto- and epoxy fatty acids; ii) oxidatively truncated PC (class II oxPC) later called short chain oxidized PC (SCh-oxPC) which occur as a result of fragmenting the oxidized chain of the PC after its previous oxidation [8]. These compounds generally present a semialdehyde (ω-CHO-SCh-oxPAPC) or dicarboxylic (ω-COOH-SCh-oxPAPC) chain in place of the unsaturated chain. Class III of cyclized oxPCs (cyc-oxPCs) and IV of oxidatively N-modified PCs [9] are not included in this publication.
Different classes of oxPC have different biological implications and therefore a proper identification and understanding of oxidation is crucial [[9], [10], [11], [12], [13]]. Considering the fact that the role of oxPC in health and disease is still not fully discovered, their analysis via non-targeted metabolomics seems to be fully justified.
OxPLs, especially oxPCs, have already been described in detail. However, the majority of the publications either refers to their biological properties and implication in health and disease states [7,11,13,14] or describes the mechanism of their formation [8,11]. Other publications have been devoted to the measurement of oxPC by means of LC-ESI-MS although only a few of them focused on the identification [[15], [16], [17]]. Current knowledge has been limited to low accuracy mass analyzers such as single quadrupole (Q), triple quadrupole (QQQ) or ion trap (IT).
This knowledge was significantly extended in 2015 with the work of Sala and colleagues [18], who analyzed oxPC using HILIC chromatography connected with a linear ion trap-Orbitrap mass spectrometer. They provided a large amount of information on MS level (accurate mass) and tandem mass spectrometry (MS/MS) level however the last was only as nominal mass. Very recently, significant advancements came from the work by Ni and colleagues [19], who proposed a software called LPPtiger for prediction and identification of oxPLs. It covers glycerophosphocholine (PC), glycerophosphoethanolamine (PE), glycerophosphoserine (PS), glycerophosphoglycerol (PG), and glycerophosphates (PA) and their lyso-forms, oxygen addition products (LCh-ox) and oxidative cleavage products (SCh-ox). The identification is performed based on the information from the negative ionization mode through five partial scores, based on data dependent analysis (DDA) fragmentation spectra.
Although oxidation has been described quite broadly, still the number of oxPLs in metabolomics databases is limited. Furthermore, the number of oxPLs found in databases exclusively devoted to lipids, such as LipidBlast [20] (none), LipidMaps [21] (26 lipids), LipidBank [22] (none) or LipidHome [23] (none) is minor (state for May 25th 2018). This point is crucial for global analysis, such as non-targeted metabolomics, where metabolites are measured anonymously.
Identification starts with the annotation of signals querying databases through the experimental masses; thus the power of identification is among other parameters a function of mass accuracy [24]. The confirmation of the annotations can be achieved by the analysis of authentic standards. Nevertheless, due to their often limited availability and/or high price, this strategy may be challenging [25]. As an alternative, MS/MS can be used. MS/MS spectra of PLs are relatively easy to interpret since they follow known fragmentation patterns that have already been described in detail. A range of independent studies has been performed using different mass analyzers that help in defining a list of product ions and neutral losses (NLes) undoubtedly confirming the presence of a particular PL [[26], [27], [28]].
In general, each spectrum can be divided into three characteristic regions [26] including: i) low mass region with product ions of head group; ii) mid-mass region with fatty acids and fatty acids-related signals and iii) high-mass region of NLes indicating the ionization (adduct formation) (see Fig. 1S panels A and B, supplementary material). However, in the case of many spectra, product ions corresponding to the fatty acids are not explicable. Furthermore, some NLes cannot be explained by the presence of adducts such as sodium or potassium (Fig. 1S panels C and D, supplementary material) [26].
To explain these unknown signals, a profound study of many spectra from biological samples, including samples of patients with newly diagnosed type 2 diabetes mellitus (T2DM), was performed. Diabetes was chosen since it is well established that high hyperglycemia causes strong oxidative stress leading to the formation (among other oxidation products) of oxPCs [29,30]. The aim of this work is to perform global characterization of oxPCs for LC-MS analysis and their recognition in MS/MS spectra. This is particularly important for data independent analysis (DIA), since most of the existing solutions correspond to the DDA [19].
Section snippets
Chemical and reagents
Ultrapure water, used to prepare all the aqueous solutions, was obtained “in-house” from a Milli-Direct16 system (Millipore, Billerica, MA, USA). LC-MS grade acetonitrile was purchased from Honeywell (Sigma-Aldrich Chemie GmbH, Steinheim, Germany) and Fisher (Fisher Scientific, Loughborough, UK). Analytical grade formic acid was purchased from Fluka Analytical (Sigma-Aldrich Chemie GmbH, Steinheim, Germany) and the analytical standard was a mixture of oxidized PAPCs (oxPAPCs) purchased from
Results and discussion
The origin of this work was structural elucidation of MS/MS spectra of PCs from human plasma samples. Though spectra were correctly assigned representing PCs as a class, identification of the exact lipids with particular compositions of fatty acids remained ambiguous due to additional unexplained product ions. This leads to the hypothesis that some of the observed PCs may have undergone modifications affecting their structure and leading to the formation of additional ions. Since peroxidation
Conclusions
A general non-targeted metabolomics method has been employed to analyze a complex standard mix of oxidation products of PAPC. CID has been employed to study the fragmentation pattern of the analytes. General diagnostic characteristics from MS and MS/MS levels have been defined. Specific signals (both product ions and NLes) for the presence of oxidation in the PC structure were determined. These signals do not necessary lead to the identification of a particular oxidized PAPC. However, a fast
Conflicts of interest
All authors declare neither financial nor commercial conflicts of interest.
Acknowledgements
This work was supported by grants from the Spanish Ministerio de Economía y Competitividad (grant CTQ2014-55279-R) and USPCEU PCON10/2016. AGF acknowledges Fundación Universitaria San Pablo CEU for his PhD fellowship. JS acknowledges European Social Fund Postdoc Project No. CZ.1.07/2.3.00/30.0004; The Ministry of Education, Youth and Sports of the Czech Republic (NPU – LO1204) and Nadace Český literární fond for travelling support. The authors are indebted to Mr. Dan Cherry for his careful
References (36)
- et al.
A lipidomic perspective on inflammatory macrophage eicosanoid signaling
Adv Biol Regul
(2014) - et al.
Oxidized phospholipids: from molecular properties to disease
Biochim. Biophys. Acta
(2007) - et al.
Chemistry of phospholipid oxidation
Biochim. Biophys. Acta
(2012) - et al.
Lipid peroxidation generates biologically active phospholipids including oxidatively N-modified phospholipids
Chem. Phys. Lipids
(2014) - et al.
Oxidized phospholipids in control of inflammation and endothelial barrier
Transl. Res.
(2009) - et al.
Structural characterization of oxidized phospholipid products derived from arachidonate-containing plasmenyl glycerophosphocholine
J. Lipid Res.
(2000) - et al.
Identification of oxidized phospholipids by electrospray ionization mass spectrometry and LC-MS using a QQLIT instrument
Free Radic. Biol. Med.
(2011) Update of the LIPID MAPS comprehensive classification system for lipids
J. Lipid Res.
(2009)Shorthand notation for lipid structures derived from mass spectrometry
J. Lipid Res.
(2013)Knowledge-based metabolite annotation tool: CEU Mass Mediator
J. Pharmaceut. Biomed. Anal.
(2018)
Characterization of chain-shortened oxidized glycerophosphocholine lipids using fast atom bombardment and tandem mass spectrometry
Anal. Biochem.
The biochemistry and physiology of mitochondrial fatty acid β-oxidation and its genetic disorders
Annu. Rev. Physiol.
Ceramides and mitochondrial fatty acid oxidation in obesity
Faseb. J.
Eicosanoid storm in infection and inflammation
Nat. Rev. Immunol.
Oxidative stress and free radicals in COPD--implications and relevance for treatment
Int. J. Chronic Obstr. Pulm. Dis.
Oxidative stress and autophagy: the clash between damage and metabolic needs
Cell Death Differ.
Studies of phospholipid oxidation by electrospray mass spectrometry: from analysis in cells to biological effects
Biofactors
Generation and biological activities of oxidized phospholipids
Antioxidants Redox Signal.
Cited by (14)
Analytical approaches for studying oxygenated lipids in the search of potential biomarkers by LC-MS
2021, TrAC - Trends in Analytical ChemistryOxidized lipids in the metabolic profiling of neuroendocrine tumors – Analytical challenges and biological implications
2020, Journal of Chromatography ACitation Excerpt :After the exclusion of any possible coelutions, the presence of a possible oxidation was considered. For this reason, the possible oxidations were calculated with the difference in mass between the non-oxidized and oxidized forms: 15.9949 Da for –OH, 31.9898 Da for –(OH)2 or –OOH [33]. The recalculated masses of non-oxidized forms for –(OH)2/–OOH and –OH lead to obtain, after a CMM search, LPC(O-16:1) and LPC(16:0) as possible candidates by [M]+ both of them, respectively.
Computational solutions in redox lipidomics – Current strategies and future perspectives
2019, Free Radical Biology and MedicineCitation Excerpt :On the other hand, this approach allows to pick potential oxidized species in the minutes scale. CEU mass mediator, a web base tool for metabolites annotation by quiring different databases, recently provided a semi-automated solution for identification of oxPLs [66]. Based on the LC-MS/MS experiments using oxPAPC standards, the authors defined and validated the fragmentation rules of oxidized PLs including modification type specific neutral losses and fragment ions, some of which were implemented in CEU mass mediator module for oxidized lipids.
Oxidized glycerophosphatidylcholines in diabetes through non-targeted metabolomics: Their annotation and biological meaning
2019, Journal of Chromatography B: Analytical Technologies in the Biomedical and Life SciencesCitation Excerpt :We observe that oxidation increases the number of isomers leading to the formation of different oxidized forms according to the position of the double bond and its stereochemistry. The identity of each peak was confirmed based on the MS/MS spectrum and known fragmentation pattern [20,39]. This was performed using a semi-automated tool for oxGPCs annotation within the CeuMassMediator platform [37,40].
Cultivar differentiation of Citri Reticulatae Pericarpium by a combination of hierarchical three-step filtering metabolomics analysis, DNA barcoding and electronic nose
2019, Analytica Chimica ActaCitation Excerpt :Plant secondary metabolites are an enormous group of small-molecule natural products with high structural diversity [5]. Metabolomics is a rapidly emerging field of postgenomic research that aims to comprehensively analyze the total population of metabolites in samples [6,7]. Recently, metabolomics analysis has been successfully applied to discriminate and differentiate plant phenotypes.
- 1
These authors contributed equally to this work.