Hany Farwanah and Thomas Kolter, LiMES—Program Unit Membrane Biology and Lipid Biochemistry, Bonn, Germany
Lipids are hydrophobic or amphiphilic low-molecular-weight substances with a low solubility in water. Glycerophospholipids, sphingolipids, and cholesterol are the building blocks of cellular membranes, and triacylglycerols are the major molecular storage forms of metabolic energy. Various lipids serve as signaling substances either as first or second messengers in signal transduction; as autocrinic, paracrinic, or endocrinic regulators;or are covalently bound to cellular proteins. Several diseases are caused by, or at least associated with, alterations in lipid metabolism, such as inherited disorders of lipid metabolism, atherosclerosis, diabetes, obesity, or Alzheimer's disease.
Lipidomics is the determination of a lipid profile of a given source under given conditions with the aim to understand lipid function in a biologic system. This understanding might be achieved by correlating changes in this composition with physiologic or pathophysiologic alterations of the system. Current research focuses on the determination of lipid profiles, which is attempted predominantly by the combination of extraction and separation techniques coupled with mass spectrometry. A major challenge is the variety of molecular lipid species found in a given cell. This article provides an overview on current methodologic approaches that find application in lipidomics. Because the function of a lipid can be associated critically with its distinct subcellular or even suborganellar localization, some limitations of this approach are discussed also.
Lipidomics (1) is the approach to determine and to understand the lipid profile of a given biologic source in terms of systems biology (2). This “understanding” requires information about the interacting partners. Systems biology requires a comprehensive set of quantitative data that can be interpreted by bioinformatic approaches to obtain insight into structure and dynamics of a given system. Current attempts to treat the hydrophobic part of the metabolome in this respect face technologic problems that must be overcome to determine lipid profiles comprehensively and accurately. With the advent of modern mass spectrometric techniques, metabolic snapshots of parts of the lipidome have become possible. They can be compared with corresponding snapshots of other genotypes and with those obtained under pathologic conditions or after drug treatment. Such differences can be visualized in the form of a network model to correlate compositional differences with functional aspects. However, the information on the levels of all lipid classes in a relevant system, such as an organ, tissue, or cell type, is often incomplete or even entirely missing. This incomplete system is, in part, caused by the large heterogeneity found among cellular lipids, and also by the structural complexity especially of glycolipids, the unavailability of appropriate standard substances, and difficulties with the extraction and ionization of certain lipid classes such as the phosphatidylinositolphosphates. Moreover, decades of lipid research taught us that not only is the concentration of a lipid critical, but also its subcellular distribution can be critical for its function (3). This article provides a brief overview on current techniques used within lipidomics.
Lipids are a structurally heterogeneous group of low molecular compounds with common solubility properties. The majority of lipids is only slightly soluble in water, but they can be extracted from biologic sources with organic solvents. Lipids can be defined as hydrophobic or amphiphathic small molecules that originate entirely or in part by carbanion-based condensation of thioesters such as fatty acids, polyketides, and their derivatives, or by carbeniumion-based condensation of isoprene units such as the terpenes, which include the sterols. A recently suggested nomenclature and assignment system for lipids will facilitate the data exchange and processing required in lipidomics (4). It consists of a 12-digit code for each molecular lipid species and classifies lipids into eight categories: fatty acyls, glycerolipids, glycerophospholipids, sphingolipids, sterol lipids, prenol lipids, saccharolipids, and polyketides (4). The first two letters of the alphanumeric Lipid-ID are a fixed database designation, the next two letters specify the category, and the numbers specify class, subclass, and the individual lipid. In this way, also more hydrophilic lipid derivatives can be considered within the field of lipidomics. The structures of selected lipids are shown in Fig. 1.
Figure 1. Structures of selected lipids.
Lipids differ not only in their chemical structures but also in the types of aggregates that they form in the aqueous environment of a living cell. Highly hydrophobic lipids are fats and oils, which are used as storage forms of metabolic energy. In an aqueous environment, these substances aggregate to lipid globules. Within cells, these lipids are deposited as insoluble droplets of up to 200 pm in size especially in the cytoplasm of adipocytes. Closer examination of these aggregates found in vivo reveals a highly metabolically active and organized complex-layered structure of an own organelle, in which triacylglycerols, diacylglycerols, cholesterol, and cholesterol esters are surrounded by a phospholipid/lysophospholipid monolayer, together with accessory proteins (5). In addition, amphiphilic lipids contain structural elements of higher polarity. Dependent on their molecular shape, these lipids form different supramolecular aggregates in an aqueous environment. The most important feature of amphiphilic lipids, with a cylindrical shape like many glycerophosphatides and sphingolipids, is their ability to form lamellar phases spontaneously, the structural basis of biologic membranes. Cone-shaped lipids with small hydrophilic headgroups such as phosphatidylethanolamine (PE), phosphatidic acid (PA), and also cardiolipin (1,3-bis-(3-sn-phosphatidyl)-glycerol) tend to form hexagonal phases in pure form. Incorporation of “nonbilayer” lipids such as PE into lamellar phases can induce a concave phase frontier and a more dense packing of the lipid tails. PA, and lysophosphatidic acid (LPA), which can be inter-converted enzymatically, favor opposite curvatures that occur in budding vesicles in vivo. Inverted cone-lipids such as detergents and lysolipids—which lack one fatty acid residue—form micelles (6). Already binary systems of bilayer-forming lipids can show the coexistence of different lipid phases at a given temperature. This differential miscibility can lead to lipid microdomains that might also be found in biologic membranes under physiologic conditions. A crucial role for membrane properties is played by the sterols, cholesterol in animals, sitosterol in plants, and ergosterol in lower eukaryotes. Within evolution, they became available after the occurrence of higher oxygen concentrations in the ancient atmosphere [see Mouritsen (7)] and enabled eukaryotic cells to form the so-called liquid-ordered membrane phases. This liquid is essential for membrane stability, impermeability, and for the function of many membrane proteins. Within lipidomics, the analysis of membrane-forming lipids, or “structural” lipids, such as phospholipids and cholesterol, can be distinguished from “mediator-lipidomics” (8). In the latter, highly bioactive lipids are determined, which are present usually in very low concentrations but are found in greater amounts in response to extracellular stimuli. Lipid mediators are substances such as the eicosanoids, platelet activating factor, LPA, sphingosine-1-phosphate, and ceramide-1-phosphate. Because “structural” lipids determine the function of membranes and can be altered under pathophysiologic conditions, both lipid classes have to be determined in a comprehensive approach.
Biologic membranes serve several vital functions such as compartmentalization, maintenance of gradients, signal transduction, and many others (9). Dependent on their source, membranes differ in their total lipid composition, but they also show lateral heterogeneity as well as differences between their two leaflets. Although glycosphingolipids (with glucosyl-ceramide as an exception) are found exclusively on the extracytoplasmic leaflet, phosphatidylcholine (PC) and sphingomyelin (ceramide-1-phosphorylcholine; SM) are found predominantly in the outer leaflet, and the aminophospholipids phosphatidylserine (PS), the PE, and also the phosphatidylinositols and their phosphates are localized predominantly on the cytoplasmic leaflet (6). This asymmetric distribution is maintained actively by flippases and floppases, but it can also be destroyed actively by scramblases in response to signals. For example, surface exposure of PS is a sign of pathologically altered cells (10). Lateral differences in lipid compositions cause microdomains to develop that account for differences in bilayer thickness and, subsequently, protein transport. Although it is still a matter of debate how these lipids are organized laterally, on which time scale the resulting microdomains (“rafts”) exist and with which functional relevance (11), lipid compositions differ not only between such short-living microdomains, but also between the apical and basolateral membranes of a polarized cell. Although information on the spatial distribution of membrane lipids can be important (3), most mass-spectrometric approaches used in lipidomics focus currently on the determination of total lipid composition. An advanced mass-spectrometric surface analysis technique, however, secondary ion mass spectrometry (SIMS) determines lipid compositions within different areas of freeze-fractured cells (12). This technique and related imaging techniques will become important tools in lipidomics.
Different local concentrations of phosphoinositides are one of the many examples of the involvement of lipids in signal transduction processes (13), Lipid mediators can act not only as membrane components, but also as autocrinic and paracrinic mediators (14); when bound to proteins, lipids fulfill crucial functions. Starting with the pioneering work of Rudolf Schoenheimer on cholesterol, it has been recognized that lipid metabolism and transport are tightly controlled processes. More recently, the molecular mechanisms that underlie lipid homeostasis have been recognized (15). Much of the interest in lipidomics results from new possibilities for the understanding, diagnosis, and treatment of human diseases caused by disturbances of this homeostasis. Therefore, determination and interpretation of lipid profiles is the aim of lipidomics. A key discipline within lipidomics is lipid analysis. Mass-spectrometric methods, however, allow not only the determination of lipid classes that share a common head group, such as PC or SM, but also the distribution of molecular species that develop by combination of different acyl chains, which can be determined with high sensitivity.
Lipid analysis starts with sample preparation, followed by extraction of the lipids from their biologic matrices. The resulting crude lipid extracts can be subjected directly to extensive and global mass spectrometric (MS) analysis, which is an approach sometimes called “shotgun lipidomics.” Alternatively, lipids or lipid classes are separated prior to analysis, either by gas chromatography (GC), liquid chromatography (LC), thin layer chromatography (TLC), or—less frequently—capillary electrophoresis (CE). These methods are coupled online or offline to mass spectrometry or tandem mass spectrometry (MS/MS; e.g., GC/MS, LC/MS, LC/MS/MS, CE/MS). In both cases, structural heterogeneity and complexity of lipids provide a huge amount of data. Accordingly, bioinformatic treatment of the data is used increasingly to facilitate data handling and evaluation.
Sample preparation and lipid extraction
In many cases, sample preparation is mandatory prior to lipid extraction. This crucial process might include tissue homogenization, determination of dry weights, cell numbers, protein content, or DNA content for normalization purposes. The addition of internal standards is not only used to control extraction efficiency, but also is required for lipid quantification by mass spectrometric analysis, in which an ion current is translated into a lipid concentration. Many standard lipids that can be distinguished from endogenously occurring lipids by using rare chain length combinations or the incorporation of isotopes are available commercially. In the field of complex lipids, especially of glycosphingolipids, standards must be chemically prepared in a laborious manner. To ensure reproducible results, sample preparation should be standardized and might rely on advanced methods that can be conducted in parallel, such as immunomagnetic cell separation. Also, the chemical derivatization of lipids is used when extraction and chromatographic properties of a certain lipid or lipid class modified in this way are improved, or when other lipid classes can be separated more easily from the modified ones by different solubilities in a given solvent.
Because of the heterogeneous nature of lipids, no single extraction method exists that extracts all lipids efficiently. Lipid extraction can be achieved by liquid-liquid extraction, Soxhlet extraction, or solid-phase extraction. In 1957, Folch developed a popular method by which many lipids are extracted using a mixture of chloroform and methanol in a volume ratio of 2:1. The extraction is followed by a wash step with water and is frequently applied for total lipid determination. The occasional formation of emulsions is a disadvantage, and more hydrophilic lipids like gangliosides are separated into the upper phase. To date, most extractions are performed by a method introduced by Bligh and Dyer in 1959 or by modifications of this method. In this case, the extraction is carried out by a monophasic mixture of chloroform, methanol, and water in a volume ratio of 1:2:0.8. Changing the ratio to 2:2:1.8 (V/V/V) leads to a phase separation, in which most lipids can be recovered from the lower phase. Proteins precipitate mostly between the two phases. It must be considered that quantitative lipid extraction by either of these methods might require repeated re-extraction, and that lipid exposure to air or enzyme sources might lead to lipid alteration or decomposition (16).
Extraction parameters such as solvent type, mixture ratios, metal ion concentration, pH of the aqueous phase, extraction time, and temperature influence the recovery of extracted lipids and must be validated to ensure reliable results. For example, the recovery of the acidic lipids PA and phosphatidylglycerol (PG) can be less than 30% in classic Folch and Bligh Dyer extraction, where these lipids can become bound to proteins tightly (17). Lipids bound to proteins covalently are only released under appropriate conditions, which depend on the type of lipid-protein linkage. For example, ceramides bound to protein of the cornified envelop in the human skin (18) can be extracted after mild alkaline hydrolysis of the ester linkage between lipid and protein. Special conditions are required for extraction of more polar lipids such as gangliosides, lysophospholipids and lysosphingolipids, or phosphatidylinositol-phosphates.
The dynamic development of mass spectrometry has had a huge impact on lipid analysis. Currently, a variety of suitable mass spectrometers is available. In principal, a mass spectrometer consists of an ion source, a mass analyzer, and an ion detector. The typical features of each instrument (Fig. 2) result mostly from the types of ion source and mass analyzer. To date, the ionization techniques applied to lipid analysis include Electrospray Ionization (ESI or nano-ESI), Atmospheric Pressure Chemical Ionization (APCI), Matrix-Assisted Laser Desorption/Ionization (MALDI), and, more recently, Atmospheric Pressure Photo Ionization (APPI) and Desorption Electrospray Ionization (DESI). For the majority of analytical tasks in lipidomics, ESI and nano-ESI are the most common choices.
Figure 2. Schematic representation of a triple quadrupole mass spectrometer. MRM is based on monitoring multiple specific precursor/product ion pairs.
ESI uses a high cone voltage to produce single- or multiple-charged ions from an analyte-containing solution by creating a fine spray of highly charged droplets. In the case of higher flow rates, this process can be assisted pneumatically by nitrogen as drying gas. ESI is a soft ionization method that can be run without fragmentation of the analyte molecules. It is suitable for membrane lipids (19). Generally, neutral lipids show poor ionization efficiencies in classic ESI, so that additional measures like adduct formation, derivatization, or application of nano-ESI are required. Typical drawbacks of ESI include signal suppression in complex matrices and varying adduct patterns.
Nano-ESI is a miniaturization of ESI with enhanced sensitivity and reduced signal suppression. It can be carried out either as classic offline approach analysis or as online nano-flow LC/MS, which are both attractive for lipidomics. Performance differences regarding the mass spectrometric response may result from different conductive coatings of the used sample capillaries and the used ionization mode. Many lipid classes, which include glycerophospholipids (20-22); sphingolipids (23, 24); triacyglycerols, cholesterol, and ergosterol (25); and cholesterol esters (26), have been analyzed using ESI and nano-ESI (20). This approach has already been extended for automated (21) and high-throughput investigations (26).
APCI is complementary to ESI for compounds of limited polarity (27). In APCI, the analyte solution is subjected to a heater of 400-500° C. The resulting plasma is ionized at a corona discharge needle with the help of nitrogen. The nitrogen molecules transfer the charge onto the analyte molecules in an indirect fashion. Because of the heat, the ionization conditions are less soft than ESI, and sensitive molecules may show fragmentation. The solvent plays a less dominant role for ionization than in ESI. This result can be beneficial in case of normal phase LC/MS approaches with organic solvents such as chloroform or hexane. Adduct formation can be reduced when suitable conditions are used. APCI has been used to analyze ceramides (28), sphingomyelin (29), triacylglycerols, as well as cholesterol and its oxidation products (30).
APPI is derived from APCI, but instead of the corona discharge needle, the ionization takes place after irradiation with a krypton lamp that emits photons of 10.0-10.6 electron volt. Different methods with and without dopant have been recommended. APPI, in combination with LC, has been used to analyze glycosphingolipids (31). In addition, APPI has been reported to be more sensitive and efficient than APCI and ESI for the analysis of fatty acid esters and acylglycerols (32, 33). Because the method has not found widespread use to date, it is too early to estimate its potential.
MALDI is another soft ionization method, but in contrast to ESI and APCI, it is carried out in solid state. The analyte is placed onto a metallic plate together with an UV-absorbing matrix. Usually, the matrix consists of aromatic acids, which are cocrystallized with the sample. Irradiation of suitable positions on the target with a laser of appropriate wavelength causes analyte desorption and ionization. MALDI is used widely in the analysis of proteins and polymers. Recently, it has also been applied to the analysis of lipids and in particular glycolipids (34). MALDI is a qualitative method that normally provides no quantitative results. The matrix used, the heterogeneities within the matrix, analyte concentrations, and preparation techniques influence the results greatly. MALDI is mostly used offline. LC/MALDI coupling can be implemented via fraction collection and automated target preparation. Compared with ESI, MALDI is more tolerant to salts and other disturbing components.
DESI is performed by directing electrosprayed-charged droplets onto a surface for analysis under atmospheric conditions. The collision of the charged droplets with the surface leads to the ionization and desorption of the analyte (35). Then, the ions produced in the gas phase are sampled by an atmospheric interfaced mass analyzer. DESI has been used to create two-dimensional images related to the distribution of lipid species in human tissues (36).
Mass analyzers differ in sensitivity, accuracy, and resolution. Among others, they include quadrupole, triple quadrupole (TQ), ion trap, linear ion trap, time of flight, Fourier transform ion cyclotron resonance, and hybrid combinations such as quadrupole-time of flight (Q-TOF) analyzers.
The deliberate generation of fragments is a valuable tool in structure elucidation and quantification. It is mainly carried out as a second mass analysis after determination of the mass per charge (m/z) ratio of the molecular ions. This approach is called tandem mass spectrometry and it can be carried out either as tandem-in-space using a triple quadrupole (or combinations of
The principle of frequently used scan modi is indicated. Please note that a quadrupole and another mass analyzer), or as tandem-in-time using an ion trap. The latter also allows multiple-stage mass spectrometry (MS"). On the other hand, a triple quadrupole allows different MS/MS scan modes, namely product ion scan, precursor or parent ion scan, and neutral loss scan.
TQ mass spectrometers (Fig. 2) have been used frequently for lipid analysis. They consist of three quadrupoles, Q1, Q2, and Q3, which are located in a row, with Q2 as the fragmentation cell filled with an inert collision gas (e.g., argon). A product ion scan starts with selecting a lipid species of a certain m/z-value in Q1, fragmenting it through collisions with the collision gas (collision-induced dissociation; CID) in Q2, and determining the m/z values of the fragments in Q3. Usually, this scan type is used to ascertain characteristic fragments and fragmentation conditions such as the suitable CID energy. Parent (precursor) ion scan (PIS) is the conversion of the product scan. Here, Q3 is fixed to m/z values of desired fragments, whereas the corresponding parent molecular ions are scanned in Q1. PIS is used to identify lipid classes and individual lipid species according to characteristic charged fragments of their functional head groups or their backbones. Neutral loss scan (NL), finally, represents the loss of a neutral fragment from a charged parent molecule. In this case, both Q1 and Q3 are used to scan, but with a constant mass offset. NL has been frequently used to analyze a variety of lipids. For example, many phospholipids have been identified recently according to the NL of their fatty acids in the positive ionization mode. A structural identification of lactosylceramide using NL of two hexoses is illustrated in Fig. 3.
Figure 3. Tandem electrospray mass spectrum (positive mode) of lactosylceramide, which shows the neutral loss of two hexose residues.
These scan modes or combinations of them have been used not only for identification, but also for quantification of lipids. In this context, multiple parent ion scan (MPIS) of head group and backbone (fatty acids) fragments has been shown recently to be suitable for the determination of individual phospholipid amounts relative to an internal standard. This analysis has been performed directly from crude lipid extracts by using a modified Q-TOF mass spectrometer (21). On the other hand, multiple reaction monitoring (MRM) is applied frequently for lipid quantification. MRM is a technique based on monitoring compound-specific transitions of precursor ions to product ions. In addition to specifity, sensitivity and selectivity are key advantages of MRM. This advantage is particularly valuable by the quantification of small compound amounts and/or by the occurrence of increased background levels. An interesting application of MRM is the quantification of many endogenous sphingolipids carried out by Zheng et al. (37). In a single LC run, multiple combinations of specific precursor-ion/product-ion transitions can be obtained, monitored, and applied to achieve sensitive quantification of individual lipid species.
As mentioned above, two different mass spectrometric approaches to lipid analysis exist: The first one is performed directly from lipid extracts without prior chromatographic separation and is referred to as “shotgun” lipidomics. Here, lipid classes are separated in the ion source according to their intrinsic electrical properties (24). Detailed and unambiguous structural and quantitative analysis of individual species is obtained by means of multiplexed mass spectrometry using NL, PIS, and combinations of them. A remarkable contribution to this field is the analysis of ceramides, cerebrosides (β-glycosylceramides), and phospholipids by Han. Refinements of this approach using advanced mass spectrometers have led to additional improvements.
On the other hand, the complexity of the lipid mixtures that compose a variety of isobaric compounds as well as the signal suppression effects during MS ionization frequently require a chromatographic separation. Therefore, LC/MS and LC/MS/MS are valuable tools. Zheng et al. (37) have developed a normal phase LC-ESI-MS/MS method that allows the separation of sphingolipid classes and the subsequent quantification of molecular sphingolipid species, which includes ceramides, glucosylceramides, lactosylceramides, and sphingomyelin in one run. Others have shown that a combination of MALDI-MS and ESI-MS/MS with thin layer chromatography (TLC) enables a successful analysis of gangliosides (38) and of low abundant phosphoinositides (39), respectively.
The amount of data obtained by mass spectrometry is enormous, particularly when many lipid classes or several lipid species are investigated simultaneously. As such, manual management of the different data sets is impractical. For this reason, various attempts have been initiated to create software that is capable of evaluating and handling the generated data sets in a qualitative, quantitative, and comparative manner. One example is the software called Lipid Profiler, which has been used to achieve automated identification, deconvolution, and absolute quantification of glycerophospholipid species analyzed by MPIS (21). Another model is given by a collection of computer algorithms referred to as the computational lipid analysis program, which has been developed to monitor time- or treatment-dependent changes in cellular phospholipids (40-42). This program consists of data handling routines and a package of multistep statistical analysis and is based on normalization of the raw signal intensities observed at a particular m/z value. Such normalization allows the comparison of dynamic changes within different phospholipid profiles. As in other fields of systems biology, the interest on developing management systems for the amount of generated mass spectrometric data is increasing.
The mentioned data management systems are only first steps toward comprehensive lipid databases and global lipid networks. An example for such a data bank is LIPIDMAPS (http://www.lipidmaps.org), which covers structures and annotations of biologically relevant lipids (43). The structures originate from the core laboratories of the LIPIDMAPS consortium and their partners. In this database, users can search the LIPID MAPS proteome database using either text-based or structure-based search options. In addition to LIPIDMAPS, other databases in Europe (http://www.lipidomics.net) and Japan (http://www.lipidbank.jp) have been initiated.
After maturation of the technology used for lipid analysis, bioinformatics will be needed to correlate quantitative data on lipid concentrations with those of the enzymes and other proteins involved in lipid metabolism on the mRNA and proteome level. These data might be generated from experiments that use RT-PCR, DNA-microarrays, cellular fluorescence imaging, or flow cytometry (44). To this end, an important role in lipidomics will be exchange and integration of data from genomics and proteomics. Thus, the entire lipidomics approach must be understood as a part of system biology, including the above-mentioned fields.
Applications and Challenges
Lipid metabolism is a tightly regulated process. Alterations in lipid metabolism and aberrant lipid levels have been associated with frequently occurring diseases, for example, atherosclerosis, diabetes, cancer, schizophrenia, neurodegenerative and respiratory diseases, and Parkinson or Alzheimer disease. Therefore, lipidomics found entrance into diagnosis and drug development, such as in the discovery and optimization of biomarkers, therapeutic targets, and lead compounds (1). In addition to this field, lipid levels can be altered in rare inherited disorders of lipid metabolism, or in response to drugs such as the large group of cationic amphiphilic drugs that are used as, for example, antidepressants, neuroleptics, β-Adrenoceptor Antagonists, or Antiarrythmics. A growing group of inherited disorders is known to affect lipid metabolism. Examples are defects in proteins required for cellular lipid trafficking (44), other defects impair sphingolipid metabolism (45), and even others might be unrecognized causes of even apparently unrelated diseases such as mental disorders or epilepsy (45). Although most of the genetic defects underlying these diseases and the identities of the major storage substances are known for more than a decade, the detailed changes in lipid composition and the pathogenic mechanisms that lead to the different phenotypes are far from clear. An improved understanding of the pathogenesis of these and other diseases can be expected from the application of lipidomics to this field.
Because of the inherent difficulties in lipid analysis, a comprehensive treatment of the lipidome in terms of systems biology has not been achieved to date, but it will be attempted for systems of reduced complexity. For example, alterations of all lipid classes in response to genetic and pharmacologic changes will be analyzed in murine macrophages by the multicentered LIPIDMAPS consortium. Recently, the protein and lipid composition of synaptic vesicles isolated from rat brain have been determined by mass spectrometry (46). With the exception of cholesterol, whose content of about 40% is very high in the vesicles, the lipid analysis confirmed data obtained by classic lipid analysis. But because ESI-MS allows determination of the chain length distribution in phospholipids, these data have been collected also for this (undisturbed) system. From a certain cell type via a certain vesicle type, let us mention a certain lipid. Cardiolipin is a bacterial lipid, and in eukaryotes it is the only phospholipid synthesized in the mitochondria. The chain length distribution of molecular cardiolipin species of rat liver and bovine heart is known for fifteen years. Changes in cardiolipin content, acyl chain composition, and cardiolipin peroxidation affects mitochondrial function and can contribute to human diseases like ischemia, hypothyroidism, aging, heart failure, and cardioskeletal myopathy (47).
Usually, lipidomics is understood as a way to determine the global lipid composition of a sample like tissues or cells. On the subcellular level, however, lipid content of certain organelles, the function of the lipid patterns, as well as the identity of the interacting partners is an actual area of research. For example, the lipidome of the nucleus, where a remodeling of phosphatidylcholine to molecular species with saturated acyl chains takes place, is analyzed currently not only with “static” but also with “dynamic” lipid profiling with the aid of stable isotope labeling (48). Indirect methods like antibody staining or photoaffinity labeling were necessary to determine lipid sorting associated with the maturation of intralysosomal membranes: Membranes are prepared for their lysosomal degradation by reduction of their cholesterol content and enrichment of a lipid characteristic for intralysosomal membranes, bis(monoacylglycero)-phosphate (45).
Novel lipid species have been discovered by mass spectrometric analysis. Knowledge of their structure and occurrence is required for the understanding of their function for the organism in health and disease, which includes a potential use as biomarkers. Only three examples for the growing number of such discoveries are mentioned briefly here: Fatty acid amides such as N-arachidonoylethanolamine and N-oleoylethanolamine are signaling substances found in the brain. Within “targeted lipidomics,” the structures of low abundant acylamides like N-arachidonoyl-dopamine and N-arachidonoyl-glycine have been determined by a combination of HPLC and ESI-MS (49). These lipids act in part as endogenously occurring cannabinoids and can regulate pain, immune function, reproduction, and appetite. A second example is the discovery of 3-ketodihydroceramides. This metabolite is usually not found because the precursor lipid, 3-ketosphinganine, is reduced rapidly. But it occurs in cultured cells that overexpress serine palmitoyltransferase and that are supplied simultaneously with sufficient amounts of the metabolic precursors serine and palmitic acid (37). Under these conditions, the capacity for ketone reduction seems to be exhausted, and excess ketosphinganine is acylated by one or more of the cellular acyltransferases. A final example is the discovery of a protein-bound glycosphingolipid. Knockout animals as source offer the advantage of greater amounts of otherwise not detectable metabolites. In mice deficient in the degradation of glucosylceramide, a novel posttranslational modification has been identified. In the animals, glucosylceramides M-hydroxylated in the acyl residue are bound covalently to proteins of the cornified envelope of the skin (50). Covalently bound lipids are essential not only for organ development, but also for the function of the human skin (18).
As mentioned before, knowledge of total lipid composition is not necessarily sufficient to understand lipid function. For example, the total cholesterol content of neurons from mice with a deficiency of the Niemann-Pick disease Cl-protein is not different from that of control animals. Closer examination of this animal model of the human Niemann-Pick disease, type C1, however, revealed that the cholesterol level in cell bodies is increased drastically at the expense of that in distal axons (51). A determination of members of all lipid classes present in a tissue is still difficult, and analysis of lipids within different areas of a cell [e.g., by SIMS (see above)] requires sample preparation by freeze fracture. For the analysis of some problems, techniques such as metabolic labeling with radioactive biosynthetic precursors might be appropriate. Results obtained in experiments with artificial (e.g., fluorescence-labeled) lipids must be interpreted with care because of entirely different properties compared with that of their native counterpart. After a period of development of methods for analysis, data processing, exchange, and visualization, a significant impact of this technology on our understanding of lipid metabolism and function can be expected.
1. Wenk MR. The emerging field of lipidomics. Nat. Rev. Drug. Discov. 2005; 4:594-610.
2. Kitano H. Systems biology: a brief overview. Science 2002; 295:1662-1664.
3. van Meer G. Cellular lipidomics. EMBO J. 2005; 24:3159-3165.
4. Fahy E, Subramaniam S, Brown HA, Glass CK, Merrill AH Jr, Murphy RC, Raetz CR, Russell DW, Seyama Y, Shaw W, Shimizu T, Spener F, van Meer G, VanNieuwenhze MS, White SH, Witztum JL, Dennis EA. A comprehensive classification system for lipids. J. Lipid Res. 2005; 46:839-862.
5. Martin S, Parton RG. Lipid droplets: a unified view of a dynamic organelle. Nat. Rev. Mol. Cell Biol. 2006; 7:373-378.
6. Janmey PA, Kinnunen PKJ. Biophysical properties of lipids and dynamic membranes. Trends Cell Biol. 2006; 16:538-546.
7. Mouritsen OG. Life-As a Matter of Fat. The Emerging Science of Lipidomics. 2005; Springer, Berlin, Germany.
8. Serhan CN, Hong S, Lu Y. Lipid mediator informatics-lipidomics: novel pathways in mapping resolution. AAPS J. 2006; 8:E284-E297.
9. Edidin M. Lipids on the frontier: a century of cell-membrane bilayers. Nat. Rev. Mol. Cell Biol. 2003; 4:414-418.
10. Zwaal RF, Comfurius P, Bevers EM. Surface exposure of phos- phatidylserine in pathological cells. Cell. Mol. Life Sci. 2005; 62:971-988.
11. Hancock JF. Lipid rafts: contentious only from simplistic standpoints. Nat. Rev. Mol. Cell Biol. 2006; 7:456-462.
12. Ostrowski SG, Kurczy ME, Roddy TP, Winograd N, Ewing AG. Secondary ion MS imaging to relatively quantify cholesterol in the membranes of individual cells from differentially treated populations. Anal. Chem. 2007; 79:3554-60.
13. Carlton JG, Cullen PJ. Coincidence detection in phosphoinositide signaling. Trends Cell Biol. 2005; 15:540-547.
14. Hla T. Genomic insights into mediator lipidomics. Prostaglandins Other Lipid Mediat. 2005; 77:197-209.
15. Kalaany NY, Mangelsdorf DJ. LXRS AND FXR: the Yin and Yang of Cholesterol and Fat Metabolism. Annu. Rev. Physiol. 2006; 68:159-191.
16. Manirakiza P, Covaci A, Schepens PJ. Comparison of different extraction methods for lipid content determination in various matrices. Food Comp. Anal. 2001; 14:93-100.
17. Nollet LML, ed. Food analysis by HPLC, 2nd edition 2000. Marcel Dekker, New York, p. 254.
18. Holleran WM, Takagi Y, Uchida Y. Epidermal sphingolipids: metabolism, function, and roles in skin disorders. FEBS Lett. 2006; 580:5456-5466.
19. Briigger B, Erben G, Sandhoff R, Wieland FT, Lehmann WD. Quantitative analysis of biological membrane lipids at the low picomole level by nano electrospray ionization tandem mass spectrometry. Proc. Natl. Acad. Sci. U.S.A. 1997; 94:2339-2344.
20. Ekroos K, Chernushevich IV, Simons K, Shevchenko A. Quantitative profiling of phospholipids by multiple precursor ion scanning on a hybrid quadrupole time-of-flight mass spectrometer. Anal Chem. 2002; 74:941-949.
21. Ejsing CS, Duchoslav E, Sampaio J, Simons K, Bonner R, Thiele C, Ekroos K, Shevchenko A. Automated identification and quantification of glycerophospholipid molecular species by multiple precursor ion scanning. Anal. Chem. 2006; 78:6202-6214.
22. Schwudke D, Oegema J, Burton L, Entchev E, Hannich JT, Ejsing CS, Kurzchalia T, Shevchenko A. Lipid profiling by multiple precursor and neutral loss scanning driven by the data-dependent acquisition. Anal. Chem. 2006; 78:585-595.
23. Han X. Characterization and direct quantitation of ceramide molecular species from lipid extracts of biological samples by electrospray ionization tandem mass spectrometry. Anal. Biochem. 2002; 302:199-212.
24. Han X, Cheng H. Characterization and direct quantitation of cerebroside molecular species from lipid extracts by shotgun lipidomics. J. Lipid Res. 2005; 46:163-175.
25. Sandhoff R, Briigger B, Jeckel D, Lehmann WD, Wieland FT. Determination of cholesterol at the low picomole level by nano-electrospray ionization tandem mass spectrometry. J. Lipid Res. 1999; 40:126-132.
26. Liebisch G, Binder M, Schifferer R, Langmann T, Schulz B, Schmitz G. High throughput quantification of cholesterol and cholesteryl ester by electrospray ionization tandem mass spectrometry (ESI-MS/MS). Biochim. Biophys. Acta 2006; 1761:121-128.
27. Byrdwell WC. Atmospheric pressure chemical ionization mass spectrometry for analysis of lipids. Lipids 2001; 36:327-346.
28. Farwanah H, Nuhn P, Neubert R, Raith K. Normal phase LC separation of stratum corneum ceramides with detection by evaporative light scattering and APCI massspectrometry. Anal. Chim. Acta 2003; 492:233-239.
29. Byrdwell WC, Perry RH. Liquid chromatography with dual 46. parallel mass spectrometry and (31)P nuclear magnetic resonance spectroscopy for analysis of sphingomyelin and dihydrosphingomyelin. I. Bovine brain and chicken egg yolk. J. Chromatogr. 47. A 2006; 1133:149-171.
30. Raith K, Brenner C, Farwanah H, Muller G, Eder K, Neubert RH. A new LC/APCI-MS method for the determination of cholesterol 48. oxidation products in food. J. Chromatogr. A 2005; 1067:207-211.
31. Roy S, Delobel A, Gaudin K, Touboul D, Germain DP, Baillet A, 49. Prognon P, Laprevote O, Chaminade P. Liquid chromatography on porous graphitic carbon with atmospheric pressure photoionization mass spectrometry and tandem mass spectrometry for the analysis 50. of glycosphingolipids. J. Chromatogr. A 2006; 1117:154-162.
32. Cai SS, Syage JA. Comparison of atmospheric pressure photoionization, atmospheric pressure chemical ionization, and electrospray 51. ionization mass spectrometry for analysis of lipids. Anal. Chem. 2006;78:1191-1199.
33. Cai SS, Syage JA. Atmospheric pressure photoionization mass spectrometry for analysis of fatty acid and acylglycerol lipids. J. Chromatogr. A 2006; 1110:15-26.
34. Levery SB. Glycosphingolipid structural analysis and glycosphin- golipidomics. Methods Enzymol. 2005; 405:300-369.
35. Takats Z, Wiseman JM, Gologan B, Cooks RG. Mass spectrometry sampling under ambient conditions with desorption electrospray ionization. Science 2004; 306:471-473.
36. Wiseman JM, Ifa DR, Song Q, Cooks RG. Tissue imaging at atmospheric pressure using desorption electrospray ionization (DESI) mass spectrometry. Angew. Chem. Int. Ed. Engl. 2006; 45:7188-7192.
37. Zheng W, Kollmeyer J, Symolon H, Momin A, Munter E, Wang E, Kelly S, Allegood JC, Liu Y, Peng Q, Ramaraju H, Sullards MC, Cabot M, Merrill AH Jr. Ceramides and other bioactive sphingolipid backbones in health and disease: lipidomic analysis, metabolism and roles in membrane structure, dynamics, signaling and autophagy. Biochim. Biophys. Acta 2006; 1758:1864-1884.
38. Dreisewerd K, Muthing J, Rohlfing A, Meisen I, Vukelic Z, Peter-Katalinic J, Hillenkamp F, Berkenkamp S. Analysis of gangliosides directly from thin-layer chromatography plates by infrared matrix-assisted laser desorption/ionization orthogonal time-of-flight mass spectrometry with a glycerol matrix. Anal. Chem. 2005; 77:4098-4107.
39. Wenk MR, Lucast L, Di Paolo G, Romanelli AJ, Suchy SF, Nussbaum RL, Cline GW, Shulman GI, McMurray W, De Camilli P. Phosphoinositide profiling in complex lipid mixtures using electrospray ionization mass spectrometry. Nat. Biotechnol. 2003; 21:813-817.
40. Milne S, Ivanova P, Forrester J, Brown HA. Lipidomics: an analysis of cellular lipids by ESI-MS. Methods 2006; 39:92-103.
41. Forrester JS, Milne SB, Ivanova PT, Brown HA. Computational lipidomics: a multiplexed analysis of dynamic changes in membrane lipid composition during signal transduction. Mol. Pharmacol. 2004; 65:813-821.
42. Ivanova PT, Milne SB, Forrester JS, Brown HA. LIPID arrays: new tools in the understanding of membrane dynamics and lipid signaling. Mol. Interv. 2004; 4:86-96.
43. Sud M, Fahy E, Cotter D, Brown A, Dennis EA, Glass CK, et al. LMSD: LIPID MAPS structure database. Nucleic Acids Res. 2007; 35:D527-D532.
44. Schmitz G, Liebisch G, Langmann T. Lipidomic strategies to study structural and functional defects of ABC-transporters in cellular lipid trafficking. FEBS Lett. 2006; 580:5597-5610.
45. Kolter T, Sandhoff K. Sphingolipid metabolism diseases. Biochim. Biophys. Acta 2006; 1758:2057-2079.
46. Takamori S, Holt M, Stenius K, Lemke EA, Grpnborg M, Riedel D, et al. Molecular anatomy of a trafficking organelle. Cell 2006; 127:831-846.
47. Chicco AJ, Sparagna GC. Role of cardiolipin alterations in mitochondrial dysfunction and disease. Am. J. Physiol. Cell Physiol. 2006; 292:C33-C44.
48. Hunt AN. Dynamic lipidomics of the nucleus J. Cell Biochem. 2006; 97:244-251.
49. Tan B, Bradshaw HB, Rimmerman N, Srinivasan H, Yu YW, Krey JF, et al. Targeted lipidomics: discovery of new fatty acyl amides. AAPS J. 2006; 8:E461-E465.
50. Doering T, Proia RL, Sandhoff K. Accumulation of protein-bound epidermal glucosylceramides in β-glucocerebrosidase deficient type 2 Gaucher mice. FEBS Lett. 1999; 447:167-170.
51. Karten B, Vance DE, Campenot RB, Vance JE. Cholesterol accumulates in cell bodies, but is decreased in distal axons, of Niemann-Pick C1-deficient neurons. J. Neurochem. 2002; 83:1154-1163.
Berliner JA, Zimman A. Future of toxicology-lipidomics, an important emerging area for toxicologists: focus on lipid oxidation products. Chem. Res. Toxicol. 2007; 20:849-53.
German JB, Gillies LA, Smilowitz JT, Zivkovic AM, Watkins SM. Lipidomics and lipid profiling in metabolomics. Curr Opin Lipidol. 2007; 18:66-71.
Pulfer M, Murphy RC. Electrospray mass spectrometry of phospholipids. Mass Spectrom. Rev. 2003; 22:332-364.
Watson AD. Lipidomics: a global approach to lipid analysis in biological systems. J. Lipid Res. 2006; 47:2101-2111.
Yetukuri L, Katajamaa M, Medina-Gomez G, Seppanen-Laakso T, Vidal-Puig A, Oresic M. Bioinformatics strategies for lipidomics analysis: characterization of obesity related hepatic steatosis. BMC Syst. Biol. 2007; 1:12-26. http://www.lipidlibrary.co.uk/lipids.html
Lipid Domains, Chemistry of
Lipids from Whole Cells and Tissues, Extraction of
Lipids, GC-MS of
Mass Spectrometry: Overview of Applications in Chemical Biology
Lipid Bilayers, Properties of