Lipidomics - A Personal View

Aims and Definitions

It was perhaps inevitable that the sciences of genomics, proteomics, metabolomics, glycomics and so forth would lead to the ‘new’ science of lipidomics. The first mention of this that I could find in the literature was in a paper in 2001, which referred to the ‘lipidome’, i.e. the complete spectrum of lipids in a tissue, organelle or membrane. From 2002 onwards publications using the term lipidomics have appeared in increasing numbers. Lipids have been the Cinderellas of the biological world for far too long, and it is refreshing to see the sudden burst of interest in these fascinating molecules and in the techniques for their analysis.

General biochemistry textbooks of 20-30 years ago tended to devote one short chapter to lipids, stressing only their role as a source of fuel to maintain tissue functions or as components that determine the fluidity of cell membranes. There was little excuse for this as in 1929, George and Mildred Burr had demonstrated that linoleic acid was an essential dietary constituent. Then, in 1964, there was the discovery that the essential fatty acid arachidonate was the biosynthetic precursor of the prostaglandins with their effects on inflammation and other disease states. In 1979, a major milestone was the discovery of the first biologically active phospholipid, platelet-activating factor, and at about the same time, there arose an awareness of the distinctive functions of phosphatidylinositol and its metabolites.

In preparing the documents for the section of this web site dealing with lipid composition and biology, I have become aware that every single lipid class has a highly specific function that is independent of its role as a source of energy or as a building block of membranes. Lipids are relatively small hydrophobic molecules, which are good candidates for signalling purposes. The fatty acid constituents have well-defined structural features, such as cis-double bonds in particular positions, which can carry information by binding selectively to specific receptors. Many lipids can infiltrate membranes or be translocated across them to carry signals to other cells or organelles. Phospholipids have hydrophilic sites that can bind via hydrogen bonding to membrane proteins and influence their activities. Glycolipids carry complex carbohydrate moieties that have a part to play in the immune system, for example. Sphingolipids form microdomains in membranes, termed “rafts”, where specific proteins are located that control vital cellular functions. Every scientist should now be aware that lipids are just as interesting as all the other groups of organic compounds that make up living systems.

Through their various biological activities, lipids have been implicated in a number of human disease states both in detrimental and beneficial ways. In particular, disturbances in lipid metabolism play important roles in the pathogenesis of many common diseases, such as those associated with the metabolic syndrome, including obesity and insulin-resistant diabetes, as well as Alzheimer's disease, schizophrenia, cancer, atherosclerosis, and viral and bacterial infections. Many individual lipids operate either in concert with or in opposition to others, and the balance between them determines the health or otherwise of a tissue. Therefore, a comprehensive approach to compositional analysis is very important.

The aim of lipidomics is more than simply to analyse lipids in biological systems. It is to relate lipid compositions of tissues or membranes of animals, plants or microorganisms to their physical properties, enzymes and their biology in general. A brief definition is –

The analysis of lipids on the systems-level scale together with their interacting factors.

Alternatively, the more comprehensive definition [1] may be preferred –

The full characterization of lipid molecular species and of their biological roles with respect to expression of proteins involved in lipid metabolism and function, including gene regulation.

Both definitions contain two main elements. In relation to the analysis of lipids they require a full determination of all the lipids present and the nature of the molecular species of each lipid class in the biological sample being studied. Secondly, they suggest that such analytical data must be related to the biological function of lipids through knowledge of such enzymes, genes and other factors that may relevant. It is also implied that the metabolic and physical relationships of lipids to enzymes, receptors and nonlipid signalling molecules in their membrane environments must be considered. In many studies, the data obtained are intended to shed light on human disease states. A further aim is to eventually integrate all the various ‘omics’ into a single framework of cellular metabolism.

The topic has been subdivided, and I have seen the terms “phospholipidomics”, “sphingolipidomics”, “glycolipidomics”, “neurolipidomics”, “steroidomics” and even “endocannabinoidomics” in the literature. While a case can be made for what some call “focussed or targeted lipidomics”, there is a danger that lipidomics becomes simply the latest ‘buzz word’ to describe every study involving analyses of fatty acid compositions or of molecular species of a single lipid class in a tissue. I hope editors of journals will be sensitive to misuse of the term. For example, I have seen it used in the title of papers describing methods for fatty acid and for aldehyde analysis.

Is the subject really a new one? To answer this question, I have selected two papers from the 1960s that seem to me to fully satisfy the above definitions of lipidomics, at least in terms of the general knowledge of biochemistry of their time [2,3]. Randall Wood and his colleague isolated and quantified the main lipid classes in rat liver, a vital organ in terms of lipid metabolism in a widely used animal model. Then, the molecular species of each lipid were analysed, and the stereospecific positional distributions of fatty acids in all of the glycerolipids were determined with high precision. Finally, the results were interpreted fully in the light of existing knowledge of biochemical pathways. Of course, I could have chosen many other similar papers from this era. However, it is certainly true to say that the recognition of lipidomics as a discipline received an impetus with the emergence of new technologies at the start of this century.

Mass Spectrometry – The Advantages

If the subject is not new, the methodology that is now being applied to the analysis of lipids in the name of lipidomics certainly is novel and extremely powerful. Modern mass spectrometry methods involving ionization by electrospray (ESI), fast atom bombardment (FAB), atmospheric pressure chemical-ionization (APCI), atmospheric pressure photo-ionization (APPI), and matrix-assisted laser desorption (MALDI) techniques are highly sensitive and can produce excellent quantitative data. The newest generation of mass spectrometers include ion trap, mass quadrupole:linear ion trap (Q:LIT), quadrupole:time of flight (QToF), linear ion trap:`Orbitrap' (LIT:Orbitrap), and Fourier transform ion cyclotron resonance (FT-ICR) and enable highly detailed analyses, though often at substantially greater capital cost. All of these methods give a wealth of structural information about analytes.

Of the various ionization methods, electrospray ionization appears to be the most popular, as it offers both high sensitivity and good quantification with less need for substantial structure-dependent response factors than the other ionization methods. It gives excellent results with phospholipids, sphingolipids and simple lipids such as triacylglycerols. There may be no need for extensive sample preparation or for a chromatographic interface with ESI instruments, since samples can be injected directly into the ion source. In addition, tandem MS methods greatly enhance the information obtainable. While the instruments are costly, they are becoming more affordable.

ESI-MS is a mild ionization technique that generates intact molecular ions with relatively little fragmentation and remarkable sensitivity (femtomole range). In consequence ESI-MS provides a direct measurement of intact molecular species of lipid classes, without an absolute need for isolation of lipid classes and subsequent chemical or enzymatic degradation for structural analysis. Much of the dubiety that might arise from the absence of a preliminary fractionation step can be removed by the application of diagnostic tandem MS/MS scans. In this technique, individual mass ions from one MS scan are selected in a collision cell where fragment ions are generated by a combination of argon gas and applied potential before separation in a second quadrupole and transmission to the detector. In this way, much more structural information is obtained. The term “shotgun lipidomics” is increasingly being used for this approach [4]. As a single cell type can contain well over 1000 distinct molecular species of lipids, ESI-MS appears to be the only method for obtaining a comprehensive analysis in a reasonably short time.

Mass Spectrometry – The Drawbacks

All methods for the analysis of complex mixtures of lipid classes have some limitations. While it could be argued that direct-inlet MS has fewer than most, there are drawbacks to even this methodology of which we should be aware? I retired from active research before the new MS techniques became widely available, so my knowledge of the subject is derived from my reading of the literature not from personal experience. Please consider the comments that follow in this light.

The main drawback of MS is the absence of a defined relationship of the intensity of an ion with the concentration of the analyte corresponding to this ion. Therefore, quantification can only be accurately performed through a measurement of a ratio of the intensities between the ions of an analyte and a stable isotopically labelled analogue of this analyte under identical experimental conditions. This is especially a problem with liquid chromatography interfaced to MS, as slight changes in elution times can cause major changes in ion suppression. Reproducibility of LC-MS conditions is seen as an important problem in metabolomics in general.

Information is available on positional distributions of fatty acids in glycerophospholipids by modern MS methods, but not with the precision of classical methods. The fatty acids in the primary and secondary positions in triacyl-sn-glycerols can be distinguished by mass spectrometry, for example, but full stereospecific analysis is not possible. This requires the application of complicated methodologies for which the services of skilled analysts are necessary. Indeed, all chiral lipids will require a chromatographic step for definitive characterization. Similarly, information can be obtained on double bond positions in fatty acid components of intact lipids by ESI-MS, but not in as comprehensive or as simple a manner as with gas chromatography linked to MS.

Then, there is the problem of lipids that have the same molecular weights, and a recent paper highlights the difficulty of differentiating phosphatidylglucose and phosphatidylinositol by mass spectrometry, suggesting the former might easily be missed in samples (and probably has been in many analyses) [5]. The same problem arises with phosphatidylglycerol and lysobisphosphatidic acid. There must be many more such examples. While tandem MS techniques may solve many of these difficulties, they must be applied with care by knowledgeable analysts.

A major challenge for shotgun lipidomics is that many lipids in solution tend to form aggregates, which cannot be ionized efficiently. This depends on a number of factors, including the concentration and polarity of a given lipid, the chain length and degree of unsaturation of the fatty acid constituents, and the solvent employed. Lipid samples must therefore be in a sufficiently dilute solution to prevent aggregation, otherwise quantification is impossible.

Mass spectrometry of intact lipids is a highly specialised technique. A vast amount of data can be produced over a relatively short period, so that time that otherwise would have been spent on derivatization and chromatography is now spent analysing this information on a computer. Until recently, newcomers to the field were obliged to develop their own software to record, interpret and apply appropriate statistical analyses to these data. Happily, software tools are now being made available free of charge that enable high-throughput analysis of extensive data sets from lipid extracts [6]. It seems probable that a high degree of automation of data handling will soon be possible. The science of bioinformatics is certainly assuming increasing importance.

When mass spectrometry is interfaced with a high-performance liquid chromatography (HPLC) system, some of the potential confusion over the identities of lipids is avoided. The value of this approach has been highlighted in a recent review of methodology for the analysis of sphingolipids, for example [7]. The authors state, “The resulting LC–MS/MS analyses are one of the most analytically rigorous technologies that can provide the necessary sensitivity, structural specificity, and quantitative precision with high-throughput for 'sphingolipidomic' analyses in small sample quantities”. For example, they permit separation of isometric and isobaric species such as glucosyl- and galactosylceramides. I am sure that this view is correct and that chromatographic separations will continue to have great importance for the foreseeable future. The use of narrow-bore columns (“Ultra-HPLC”) reduces analysis times and improves the sensitivity.

Conclusions

In short, modern mass spectrometric methods are revolutionizing the approach to the analysis of lipids, and they are encouraging biologists to consider the importance of individual molecular species of specific lipid classes in living systems. While I have concentrated on ESI-MS here, other MS techniques are producing invaluable data. On the other hand, while mass spectrometry alone can give useful information on fatty acid compositions, the precision and robustness of the flame ionization detector will ensure that gas chromatographic analysis will continue to be the method of choice for detailed fatty acid analysis, though MS will certainly remain an invaluable aid to identification. The enzymic methodology for determination of positional distributions of fatty acids in glycerolipids, which has now been available for nearly fifty years, will still be required for accurate analysis, especially with triacyl-sn-glycerols. Nuclear magnetic resonance spectroscopy will continue to have its place, as will the various optical spectroscopy methods. In short, regardless of the power of the newer MS methods, I am encouraged to believe that the discipline of ‘lipidomics’ is going to continue to need skilled analysts and chromatographers for some considerable time to come.

I can recommend recent reviews on the aims of lipidomics [8,9] and a book I jointly authored with Xianlin Han on the methodology [10]. The January issue (2009) of the European Journal of Lipid Science and Technology is devoted to lipidomics.

References

1. Spener, F., Lagarde, M., Géloën, A. and Record, M. What is lipidomics? Eur. J. Lipid Sci. Technol., 105, 481-482 (2003).
2. Wood, R. and Harlow, R.D. Structural studies of neutral glycerides and phosphoglycerides of rat liver. Arch. Biochem. Biophys., 131, 495-501 (1969) (DOI: 10.1016/0003-9861(69)90421-4).
3. Wood, R. and Harlow, R.D. Structural analyses of rat liver phosphoglycerides. Arch. Biochem. Biophys., 135, 272-281 (1969) (DOI: 10.1016/0003-9861(69)90540-2).
4. Han, X.L. and Gross, R.W. Shotgun lipidomics: Electrospray ionization mass spectrometric analysis and quantitation of cellular lipidomes directly from crude extracts of biological samples. Mass Spectrom. Rev., 24, 367-412 (2005) (DOI: 10.1002/mas.20023).
5. Haimi, P., Uphoff, A., Hermansson, M. and Somerharju, P. Software tools for analysis of mass spectrometric lipidome data. Anal. Chem., 78, 8324-8331 (2006) (DOI: 10.1021/ac061390w).
6. Nagatsuka, Y., Tojo, H. and Hirabayashi, Y. Identification and analysis of novel glycolipids in vertebrate brains by HPLC/mass spectrometry. Methods Enzymol., 417, 155-167 (2006) (DOI: 10.1016/S0076-6879(06)17012-3).
7. Haynes, C.A., Allegood, J.C., Park, H. and Sullards, M.C. Sphingolipidomics: Methods for the comprehensive analysis of sphingolipids. J. Chromatogr. B, 877, 2696-2708 (2009) (DOI: 10.1016/j.jchromb.2008.12.057).
8. Oresic, M., Hanninen, V.A. and Vidal-Puig, A. Lipidomics: a new window to biomedical frontiers. Trends Biotechnol., 26, 647-652 (2008) (DOI: 10.1016/j.tibtech.2008.09.001).
9. Khalil, M.B., Hou, W.M., Zhou, H., Elisma, F., Swayne, L.A., Blanchard, A.P., Yao, Z.M., Bennett, S.A.L. and Figeys, D. Lipidomics era: accomplishments and challenges. Mass Spectrom. Rev., 29, 877-929 (2010) (DOI: 10.1002/mas.20294).
10. Christie, W.W. and Han, X. Lipid Analysis - Isolation, Separation, Identification and Lipidomic Analysis (4th edition), 446 pages (Oily Press, Bridgwater, U.K.) (2010) - www.pjbarnes.co.uk/op/la4.htm.

Additional references on this topic can be found in the ‘Literature Survey’ section of the Lipid Library. Subsequent to publication here, a version of this web page was published elsewhere (Christie, W.W. Lipidomics - A personal view. Lipid Technology, 21, 58-60 (2009) - DOI: 10.1002/lite.200900009.