Abstract
Hundreds of molecular species make up the cellular lipidome. In this minireview, considerations for interpreting membrane and storage lipid profile changes that are often the focal point of lipidomic studies are discussed. In addition, insights how the most conserved molecular patterns are formed in eukaryotic systems and the consequences for the perturbation of lipid homeostasis are addressed. The implications of lipid identification specificity and experimental variability on modeling membrane structure and systemic responses are also discussed. The profile changes of membrane and storage lipids are bound to the kinetics of the metabolic system, and experimental design and functional interpretation in lipidomic research should be adapted accordingly.
Introduction
Lipids generally have three biological functions as follows: 1) storage of chemical energy, 2) essential structural components for biological membranes, and 3) signaling of molecular events (; ). In many lipidomics applications, membrane and storage lipids are measured and exploited as the basis for functional associations. For some structural lipid classes, such as lysolipids, ceramides, and diacylglycerols (and others), interpretations are often complicated as these lipids can also act as signaling molecules that might start a cascade of biosynthetic events. However, conventional omics data interpretation strategies, such as those established for expression data and proteomics, can be applied much more easily for signaling lipids because of the similar thresholds for fold-changes of two or more and its clearer functional association. This might be an oversimplification, but lipid signaling cascades, for instance, for small lipid mediators, phosphatidylinositides, and sphingosines (and others), can be equivalent to the activation of protein signaling cascades (Wymann and Schneiter, 2008; ). In both cases, a primary event, which is initiated by a small number of often low abundant molecular entities, leads to systemic changes on the cellular or organismal level.
From this perspective, lipid metabolic processes that change the make-up of the cell with respect to the membrane systems, and/or storage compartments follow slower kinetic and fold-changes and can be rather small but affect a large portion of the lipidome. The resulting consequences for experiment planning and data interpretation of lipid profiles will be discussed.
Lipid identification and quantitation are the basis for functional associations
Up-to-date mass spectrometry-based lipidomics procedures are not specific enough to identify lipids at the structure-defined level (; ). Specifically, the number of isomeric structures that are actually unresolved by chosen analytical approach should at least lead to the application of this recently introduced hierarchical nomenclature. In the case of complex lipids that are comprised of two or more aliphatic chains, the number of isomeric molecules can reach hundreds, which is the result of the limit of specificity at the molecular species identification level (; ). For the major and abundant lipid classes of eukaryotes that make up the membrane systems and the storage pools, the exact composition of the aliphatic chains are generally not defined with regards to the sn-position, double-bond position, and cis/trans isomerism (; ). However, the higher specificity for lipid identification in large-scale studies is on the horizon with the application of ion mobility in lipidomics (). Because changes in the gas phase collision cross section (CCS) can be minute (≤1%), higher resolution ion mobility is required, and the associated informatics approaches that properly detect these mass spectrometric features in three-dimensional datasets (rt, CCS, and m/z) are just now becoming publicly available ().
The structure of lipid molecules has a direct influence on the physical parameters of a biological membrane (Figure 1). The most obvious impact of structural lipid isomers is found in the packaging density, thickness, compressibility, and phase properties of a membrane (Figure 1A). The pleiotropic nature of the effects on membrane biology are a focal point of ongoing research (; ) [reviewed in ); )] and are mostly defined by the direct interaction of lipids with proteins and the modulation of lipid domains (). Another important effect of lipid profile changes concerning the aliphatic chains and lipid class composition is the maintenance of membrane asymmetry and morphology that is only possible within a certain compositional framework (). A well-established model is the application of lipid shapes that, based on the steric requirement of each lipid, allow possible membrane curvatures to be predicted [Figure 1A; ; ; ; )]. From this perspective, the need of preferable identification on a structural refined level and absolute quantitation of lipids (in molar amounts) is easy to perceive (). Large-scale lipid quantitation and its accuracy remain an open issue in lipidomics research and also reflect the wide range of analytical strategies based on LC-MSn and shotgun approaches (Zullig et al., 2020; ).
FIGURE 1
However, the scientific community should not focus on favorite analytic strategies and the largest ID-number but instead use the scientific question and data quality as a first priority for the selection of a lipidomics approach (). With regard to the overall membrane composition, the molar composition is required to associate membrane phenotypes to lipid profiles that go beyond a pure description of changes but rather help reconstitute model membranes and perform functional studies. Furthermore, I would argue that a lack of standardization and proper normalization of lipid molarities to parameters like cell number, weight, protein content, and/or quantifiable phenotype limits the applicability of lipidomics. To overcome this essentially descriptive nature of lipidomics, a study plan and biological hypothesis should be formulated in an interdisciplinary discourse that includes the essential steps of the analytical method development concerning sampling and sample preparation. When considering the heterogenic nature of the biological sample matrix and its consequences for lipid identification and quantitation, the correlation between phenotypes and their putative link to the lipidome can be explored.
The regulation of lipids in response to perturbations of homeostasis
An intriguing illustration for analytical and biological challenges that one must face in lipidomics research was formulated by Andrej Shevchenko () in an analogy of knights and armies of the Middle Ages as follows: “Knights are like proteins. Each knight has a pedigree, a story, an emblem, a flag. Lipids are like armies. Knights are more visible. It’s easier to write poetry about them, but in the end it’s the army that wins, the collective of individuals. That’s the lipids.”
Indeed, a lipid metabolic perturbation that affects the state of a cell, tissue, organ, or organism often induces quantitative changes in many related molecules (Figure 1B). Such biosynthetic activities, which are the basis for cell membrane homeostasis, trafficking, and autophagy (to mention only the most obvious), require time by nature to be executed by the lipid metabolic system. This action of a biosynthetic system has an endogenous inertia to convert lipids. The turnover rates for membrane and storage lipids have only sparsely been determined if at all. These measurements should cover time scales in the range of seconds to days depending on the studied biological system to gain insights into the kinetics of a transition between different lipid metabolic states (Figure 1B). Furthermore, a possible occurrence of a pathology, where the biological system cannot be converted into another stable homeostatic state, can be characterized with much better resolution.
Model organisms such as D. melanogaster, Caenorhabditis elegans, and S. cerevisiae as well as cell culture systems are the ideal study objects to observe associations between the perturbation of lipid homeostasis and biological phenotypes. In a drosophila model of lysosomal storage disease, ) studied the emergence of lipid profile changes in association with the occurrence of neurodegenerative phenotypes. They showed that as an early phenotype, ceramide and sphingosine levels are altered before signs of failure of cell autophagy in the form of endomembranous structures and lipofuscin accumulation occurred. After 14 days of systemic lipid metabolism perturbation, significant behavioral alteration caused by neurodegeneration could be determined. A similar systemic breakdown of lipid homeostasis was observed for LET-767 RNAi, which is responsible for the elongation of fatty acids in C. elegans that only showed slight changes at first in TAG profiles but had in the second generation a fatal effect that led to developmental arrest and death (). A detailed lipid analysis revealed that the breakdown of sphingolipid biosynthesis was the phenotypic driver caused by a lack of branched chain fatty acids, which are an essential educt for sphingosine synthesis. Detailed lipidomics analysis during the polarization of MDCK cells showed that general reorganization of the membrane system and maintenance of apical and basolateral surfaces required molecular adjustments throughout the lipidome (). Within a time frame of approximately 7 days, a new lipid homeostatic state was reached signified by elongated sphingolipid acid chain lengths, increased hydroxylation, and less double bonds. At the same time, for glycerophospholipids, a higher degree of poly-unsaturated chains in combination with a tendency for longer chain length was observed.
These examples underline the requirements for systematic lipidomic studies to establish well-defined sampling times and consider the kinetic aspects of the biological process to be studied. In addition, for the study objectives where a lipid metabolic switch is investigated and not a single “golden molecule,” this should denote a systemic alteration. Furthermore, one can postulate that there always should be a panel of lipids connected by a metabolic network that responds to the systemic perturbation of lipid homeostasis. From my experience and reported studies, aliphatic chains of structural lipids are one of the building blocks that are usually modified due to environmental stress, genetic defects, and cell biological processes. This has two practical consequences as follows: first, lipid class profiles are defined by the underlying fatty acid composition, and second, the detection of many systemic lipidome changes are dependent on the enzyme kinetics of fatty acid transformation, transport, and exchange. From these findings, it can be argued that major abundant structural lipids should always reflect the overall fatty acid composition. For instance, this is the case in most eukaryotic system palmitic acids (FA 16:0), oleic acid (FA 18:1), and stearic acid (FA 18:0). The resulting profiles of lipid classes— phosphatidylcholine (PC), phosphatidylethanolamine (PE), and phosphatidylserine (PS)—are consequently built of 16:0_16:0, 16:0_18:0, 16:0_18:1, 18:1_18:1, and 18:0_18:1 aliphatic backbones that represent a large molar percentage of the eukaryotic lipidome.
Regarding the aforementioned examples of systemic lipidome alterations, the publication of apparently structurally and/or synthetically unrelated (quite distanced in the metabolic pathway with many enzymatic steps in between) lipid biomarkers only based on a descriptive study should be read with caution. It has been reported that a proper experimental group size, well-described experimental procedures, a set of quality controls, and a critical assessment on the possible errors and statistics should be subsequently considered and outlined (). Furthermore, tracer and flux analyses should become an integral part of functional lipidomic studies whenever possible (; ; ).
The experimental error propagation defines thresholds for the detection of lipid profile changes
When lipid molecular species are preferentially transformed within a metabolic network leading to profile changes, several sources of experimental variability and their impact should be critically discussed. For functional studies, it is evident that there is interest in detecting the primary lipid metabolic effect. However, biomedical studies are driven by a phenotypic characterization based upon imaging, fitness, survival, and other morphological metrics that introduce a bias toward late sampling time points that will already be dominated by secondary metabolic events. Driven by the need to have a phenotype characterized, when it is noticeable without a doubt, one can easily understand that kinetic aspects of systemic metabolic changes are often underrepresented in the literature. I would encourage to increase the number of experimental groups from only one pair consisting of “Control” and its counterpart like “Diseased,” “Mutant,” and/or “KnockDown,” to a reasonable number of phenotypically and kinetically defined experimental groups (). In the context of the adaptive power of lipid metabolic systems, i.e. the molecular snapshot that is usually taken in the form of lipid profiles and interpreted, the reversibility of an observed state/phenotype, its stability, and the number of involved regulative processes should be considered (Figure 1B).
A further decisive choice for experimental success in pinpointing primary lipid metabolic alterations is the limit of spatial resolution to cover tissue heterogeneity and cell biological compartmentalization, which to some extent can be studied with state-of-the-art lipid MS-imaging (; ), coupling of laser microdissection (; ), and cell fractionation techniques () () with lipidomics. These complex methodical approaches require further considerations concerning quantitation, possible contamination, and limits of specificity, which are discussed elsewhere in detail (; ). If such enrichment strategies for the specimen of interest is not possible, a critical assessment of the averaging effects of mixing with other molecular pools in the isolated sample is necessary.
Regarding these concerns, one should examine whether the fold-change for a lipid quantity is realistically measurable after completing the analytical workflow. In my experience, fold-changes of 0.25 for a lipid are detectable in well-planned and executed lipidomic studies. This lower limit of detectable lipid metabolic alteration integrates the experimental error propagation that is occurring during homogenization and volumetric processing, and due to batch effects of internal standards as well as instrument instability. Additionally, I would suggest that functional lipid metabolic studies be performed with at least six biological replicates to ensure statistical validity. In the context of large-scale experiments and an increasing number of variables in biomedical applications, an appropriate power calculation for experimental group sizes should be performed. However, in my experience, high foldchanges of minor lipid species in the context of systemic metabolic switches are often overvalued in comparison to smaller changes of major abundant components of a lipidome, even when the affected molar percentage is magnitudes higher.
Conclusion
In this minireview, I have provided some personal viewpoints on the experimental limits and interpretation guidelines for the application of lipidomics to study lipid homeostasis. This is to raise awareness of some principles in studying systemic lipid metabolic changes concerning experimental planning and general data interpretation. The reader is invited to further the discussion on lipid profiles as a snapshot of membrane compositions in the context of biophysics and spatial and temporal resolution. The interest in lipids in biomedical research is rightfully increasing, and the lipidomic community should pursue the development of appropriate analytical methods and the provision of high quality data with a determination to integrate lipidomics into interdisciplinary research.
Statements
Author contributions
DS conceptualized and wrote the manuscript.
Funding
DS is grateful for the funding of his research from the German Center for Infection Research (DZIF) of the research topic “Thematic Translational Unit Tuberculosis” (TTU 709, 813) and the German Network for Bioinformatics Infrastructure (de.NBI, project LIFS2 FKZ 031L0108B).
Conflict of interest
The author declares that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors, and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
References
1
ArnaudC. H. (2011). Lipids take charge. Washington: Chemical and Engineering News, 89. Available at: https://pubsapp.acs.org/cen/support/contact.html.
2
BednarikA.PrysiazhnyiV.BezdekovaD.SoltwischJ.DreisewerdK.PreislerJ. (2022). Mass spectrometry imaging techniques enabling visualization of lipid isomers in biological tissues. Anal. Chem.94, 4889–4900. 10.1021/acs.analchem.1c05108
3
BrandenburgJ.HeyckendorfJ.MarwitzF.ZehethoferN.LinnemannL.GischN.et al (2022). Tuberculostearic acid-containing phosphatidylinositols as markers of bacterial burden in tuberculosis. ACS Infect. Dis.8, 1303–1315. 10.1021/acsinfecdis.2c00075
4
CockcroftS. (2021). Mammalian lipids: Structure, synthesis and function. Essays Biochem.65, 813–845. 10.1042/EBC20200067
5
CullisP. R.de KruijffB. (1979). Lipid polymorphism and the functional roles of lipids in biological membranes. Biochim. Biophys. Acta559, 399–420. 10.1016/0304-4157(79)90012-1
6
EiersbrockF. B.OrthenJ. M.SoltwischJ. (2020). Validation of MALDI-MS imaging data of selected membrane lipids in murine brain with and without laser postionization by quantitative nano-HPLC-MS using laser microdissection. Anal. Bioanal. Chem.412, 6875–6886. 10.1007/s00216-020-02818-y
7
EllisS. R.BrownS. H.Het PanhuisM.BlanksbyS. J.MitchellT. W. (2013). Surface analysis of lipids by mass spectrometry: More than just imaging. Prog. Lipid Res.52, 329–353. 10.1016/j.plipres.2013.04.005
8
EntchevE. V.SchwudkeD.ZagoriyV.MatyashV.BogdanovaA.HabermannB.et al (2008). LET-767 is required for the production of branched chain and long chain fatty acids in Caenorhabditis elegans. J. Biol. Chem.283, 17550–17560. 10.1074/jbc.M800965200
9
FuT.KnittelfelderO.GeffardO.ClementY.TestetE.ElieN.et al (2021). Shotgun lipidomics and mass spectrometry imaging unveil diversity and dynamics in Gammarus fossarum lipid composition. iScience24, 102115. 10.1016/j.isci.2021.102115
10
HanX.GrossR. W. (2022). The foundations and development of lipidomics. J. Lipid Res.63, 100164. 10.1016/j.jlr.2021.100164
11
HebbarS.KhandelwalA.JayashreeR.HindleS. J.ChiangY. N.YewJ. Y.et al (2017). Lipid metabolic perturbation is an early-onset phenotype in adult spinster mutants: A Drosophila model for lysosomal storage disorders. Mol. Biol. Cell28, 3728–3740. 10.1091/mbc.E16-09-0674
12
HebbarS.SchulzW. D.SauerU.SchwudkeD. (2014). Laser capture microdissection coupled with on-column extraction LC-MS(n) enables lipidomics of fluorescently labeled Drosophila neurons. Anal. Chem.86, 5345–5352. 10.1021/ac500276r
13
IsraelachviliJ. N.MitchellD. J.NinhamB. W. (1977). Theory of self-assembly of lipid bilayers and vesicles. Biochim. Biophys. Acta470, 185–201. 10.1016/0005-2736(77)90099-2
14
JoardarA.PattnaikG. P.ChakrabortyH. (2022). Mechanism of membrane fusion: Interplay of lipid and peptide. J. Membr. Biol.255, 211–224. 10.1007/s00232-022-00233-1
15
KimI. Y.ParkS.JangJ.WolfeR. R. (2020). Quantifications of lipid kinetics in vivo using stable isotope tracer methodology. J. Lipid Atheroscler.9, 110–123. 10.12997/jla.2020.9.1.110
16
KirkwoodK. I.PrattB. S.ShulmanN.TamuraK.MaccossM. J.MacleanB. X.et al (2022). Utilizing Skyline to analyze lipidomics data containing liquid chromatography, ion mobility spectrometry and mass spectrometry dimensions. Nat. Protoc.17, 2415–2430. 10.1038/s41596-022-00714-6
17
KnittelfelderO.TraikovS.VvedenskayaO.SchuhmannA.SegeletzS.ShevchenkoA.et al (2018). Shotgun lipidomics combined with laser capture microdissection: A tool to analyze histological zones in cryosections of tissues. Anal. Chem.90, 9868–9878. 10.1021/acs.analchem.8b02004
18
KofelerH. C.EichmannT. O.AhrendsR.BowdenJ. A.Danne-RascheN.DennisE. A.et al (2021). Quality control requirements for the correct annotation of lipidomics data. Nat. Commun.12, 4771. 10.1038/s41467-021-24984-y
19
LiebischG.FahyE.AokiJ.DennisE. A.DurandT.EjsingC. S.et al (2020). Update on LIPID MAPS classification, nomenclature, and shorthand notation for MS-derived lipid structures. J. Lipid Res.61, 1539–1555. 10.1194/jlr.S120001025
20
LiebischG.VizcainoJ. A.KofelerH.TrotzmullerM.GriffithsW. J.SchmitzG.et al (2013). Shorthand notation for lipid structures derived from mass spectrometry. J. Lipid Res.54, 1523–1530. 10.1194/jlr.M033506
21
LondonE. (2022). Ordered domain (raft) formation in asymmetric vesicles and its induction upon loss of lipid asymmetry in artificial and natural membranes. Membr. (Basel)12, 870. 10.3390/membranes12090870
22
LorentJ. H.LeventalK. R.GanesanL.Rivera-LongsworthG.SezginE.DoktorovaM.et al (2020). Plasma membranes are asymmetric in lipid unsaturation, packing and protein shape. Nat. Chem. Biol.16, 644–652. 10.1038/s41589-020-0529-6
23
McdonaldJ. G.EjsingC. S.KopczynskiD.HolcapekM.AokiJ.AritaM.et al (2022). Introducing the lipidomics minimal reporting checklist. Nat. Metab.4, 1086–1088. 10.1038/s42255-022-00628-3
24
Moran-GarridoM.Camunas-AlbercaS. M.Gil-de-La FuenteA.MariscalA.GradillasA.BarbasC.et al (2022). Recent developments in data acquisition, treatment and analysis with ion mobility-mass spectrometry for lipidomics. Proteomics22, e2100328. 10.1002/pmic.202100328
25
Palusinska-SzyszM.JurakM.GischN.WaldowF.ZehethoferN.NehlsC.et al (2022). The human LL-37 peptide exerts antimicrobial activity against Legionella micdadei interacting with membrane phospholipids. Biochim. Biophys. Acta Mol. Cell Biol. Lipids1867, 159138. 10.1016/j.bbalip.2022.159138
26
PosorY.JangW.HauckeV. (2022). Phosphoinositides as membrane organizers. Nat. Rev. Mol. Cell Biol.23, 797–816. 10.1038/s41580-022-00490-x
27
SahaS.DasA.PatraC.AnilkumarA. A.SilP.MayorS.et al (2022). Active emulsions in living cell membranes driven by contractile stresses and transbilayer coupling. Proc. Natl. Acad. Sci. U. S. A.119, e2123056119. 10.1073/pnas.2123056119
28
SampaioJ. L.GerlM. J.KloseC.EjsingC. S.BeugH.SimonsK.et al (2011). Membrane lipidome of an epithelial cell line. Proc. Natl. Acad. Sci. U. S. A.108, 1903–1907. 10.1073/pnas.1019267108
29
SchmittD.BozkurtS.Henning-DomresP.HuesmannH.EimerS.BindilaL.et al (2022). Lipid and protein content profiling of isolated native autophagic vesicles. EMBO Rep.23, e53065. 10.15252/embr.202153065
30
SchwudkeD.SchuhmannK.HerzogR.BornsteinS. R.ShevchenkoA. (2011). Shotgun lipidomics on high resolution mass spectrometers. Cold Spring Harb. Perspect. Biol.3, a004614. 10.1101/cshperspect.a004614
31
SezginE.LeventalI.MayorS.EggelingC. (2017). The mystery of membrane organization: Composition, regulation and roles of lipid rafts. Nat. Rev. Mol. Cell Biol.18, 361–374. 10.1038/nrm.2017.16
32
ShevchenkoA.SimonsK. (2010). Lipidomics: Coming to grips with lipid diversity. Nat. Rev. Mol. Cell Biol.11, 593–598. 10.1038/nrm2934
33
SkotlandT.EkroosK.KavaliauskieneS.BerganJ.KauhanenD.LintonenT.et al (2016). Determining the turnover of glycosphingolipid species by stable-isotope tracer lipidomics. J. Mol. Biol.428, 4856–4866. 10.1016/j.jmb.2016.06.013
34
StrandbergE.TiltakD.EhniS.WadhwaniP.UlrichA. S. (2012). Lipid shape is a key factor for membrane interactions of amphipathic helical peptides. Biochim. Biophys. Acta1818, 1764–1776. 10.1016/j.bbamem.2012.02.027
35
SychT.LeventalK. R.SezginE. (2022). Lipid-protein interactions in plasma membrane organization and function. Annu. Rev. Biophys.51, 135–156. 10.1146/annurev-biophys-090721-072718
36
TharkeshwarA. K.TrekkerJ.VermeireW.PauwelsJ.SannerudR.PriestmanD. A.et al (2017). A novel approach to analyze lysosomal dysfunctions through subcellular proteomics and lipidomics: The case of NPC1 deficiency. Sci. Rep.7, 41408. 10.1038/srep41408
37
van MeerG. (2005). Cellular lipidomics. EMBO J.24, 3159–3165. 10.1038/sj.emboj.7600798
38
WymannM. P.SchneiterR. (2008). Lipid signalling in disease. Nat. Rev. Mol. Cell Biol.9, 162–176. 10.1038/nrm2335
39
ZulligT.TrotzmullerM.KofelerH. C. (2020). Lipidomics from sample preparation to data analysis: A primer. Anal. Bioanal. Chem.412, 2191–2209. 10.1007/s00216-019-02241-y
Summary
Keywords
lipidomics, biophysics, lipid homeostasis, membrane, analytical precision
Citation
Schwudke D (2023) What information is contained in experimentally determined lipid profiles?. Front. Anal. Sci. 3:1157582. doi: 10.3389/frans.2023.1157582
Received
02 February 2023
Accepted
06 March 2023
Published
23 March 2023
Volume
3 - 2023
Edited by
Jared B. Shaw, e-MSion, Inc., United States
Reviewed by
Carmen Bedia, Institute of Environmental Assessment and Water Research (CSIC), Spain
Updates
Copyright
© 2023 Schwudke.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Dominik Schwudke, dschwudke@fz-borstel.de
This article was submitted to Omics, a section of the journal Frontiers in Analytical Science
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.