Lipids Take Charge | October 10, 2011 Issue - Vol. 89 Issue 41 | Chemical & Engineering News
Volume 89 Issue 41 | pp. 15-20
Issue Date: October 10, 2011

Cover Story

Lipids Take Charge

Mass spectrometry propels the field of lipidomics
Department: Science & Technology
News Channels: Analytical SCENE
Keywords: mass spectrometry, lipids, imaging, ion mobility, lipidomics
[+]Enlarge
Lipid Maps
Mass spectrometric imaging reveals unexpected lipid distributions in 12- by 13‑mm slices of mouse lung. A common palmitic acid-containing phospholipid (red) is distributed throughout the lung, but an arachidonic acid-containing phospholipid (green) accumulates at the edges of the airways.
Credit: Courtesy of Robert Murphy
Mass spec imaging reveals unexpected lipid distributions in slices of mouse lung. A common palmitic acid-containing phospholipid (red) is distributed throughout the lung, but an arachidonic acid-containing phospholipid (green) accumulates at the edges of the airways.
 
Lipid Maps
Mass spectrometric imaging reveals unexpected lipid distributions in 12- by 13‑mm slices of mouse lung. A common palmitic acid-containing phospholipid (red) is distributed throughout the lung, but an arachidonic acid-containing phospholipid (green) accumulates at the edges of the airways.
Credit: Courtesy of Robert Murphy
[+]Enlarge
Lipid interplay
Six of the eight classes of lipids are involved in macrophage lipid metabolism. Fatty acids, which are derived from acetyl coenzyme A, occupy a central place in the pathway. The colors represent the different lipid classes. Eicosanoids and sterol esters are subclasses of fatty acids and sterols, respectively.
Flowchart
 
Lipid interplay
Six of the eight classes of lipids are involved in macrophage lipid metabolism. Fatty acids, which are derived from acetyl coenzyme A, occupy a central place in the pathway. The colors represent the different lipid classes. Eicosanoids and sterol esters are subclasses of fatty acids and sterols, respectively.
[+]Enlarge
Class structure
The LIPID MAPS consortium organized lipids into eight broad classes. One example from each class is shown here, with the class name and the common name. Saccharolipids and polyketides are found only in plants and bacteria. The others are found in all organisms.
structures
 
Class structure
The LIPID MAPS consortium organized lipids into eight broad classes. One example from each class is shown here, with the class name and the common name. Saccharolipids and polyketides are found only in plants and bacteria. The others are found in all organisms.
[+]Enlarge
Picture Taker
SIMS instruments such as this one can image lipids.
Credit: Frederic Weber/Winograd Lab/Penn State
Secondary ion mass spectrometry instrument in the lab of Nicholas Winograd at Penn State.
 
Picture Taker
SIMS instruments such as this one can image lipids.
Credit: Frederic Weber/Winograd Lab/Penn State

In the “omics” world, lipids have long been in the shadows, while nucleic acids and proteins hogged the limelight. But now, this broad-ranging class of biomolecules is stepping into the spotlight as well. And mass spectrometry (MS) is the tool that is making it possible.

Like genomics and proteomics before it, lipidomics is about the profiling of a particular type of biomolecule in a cell or organism. But lipids are much more chemically diverse than either genomic DNA or proteins. By some estimates distinct lipid species number in the tens of thousands—or even more.

As a category of biomolecules, “lipids” is a bit of a catchall: Lipids are loosely defined as biological molecules that are generally hydrophobic in nature and in many cases soluble in organic solvents. They include the phospholipids found in cell membranes and the hormones that serve as signaling molecules.

Lipids “tend to be fairly small—not as big as biopolymers—but they’re very diverse,” says Robert C. Murphy, a mass spectrometrist and lipid biochemist at the School of Medicine at the University of Colorado, Denver. “One of the reasons mass spectrometry is so powerful is we can deal with all of these diverse structures.”

In the mass spectrometer, most lipids wind up singly charged, says Markus R. Wenk, head of the lipidomics program at the National University of Singapore. That means that the observed mass-to-charge ratios can be related directly to the molecular mass of particular lipids, making identification easy.

In addition, “mass spectrometry is nicely matched with the molecular-weight range of lipids,” Murphy says. Unlike proteins, which can be many kilodaltons, lipids are mostly clustered in the mass range below 1,200 daltons. When profiling lipids, he notes, “we never really challenge the mass-range capability of mass spectrometers.”

“So far, mass spectrometry is the method to go down to the lipid molecular level in terms of detection limits and sensitivity,” says Friedrich Spener, a biochemist at the University of Graz and cofounder of the Lipidomics Research Center Graz, in Austria.

“We don’t have those wonderful tools available in protein biochemistry to amplify and make large amounts” of lipids, Murphy explains. MS “deals with what the cell gives us.”

The nascent field of lipidomics owes much to a multi-institution initiative called LIPID MAPS(Lipid Metabolites & Pathways Strategy). The 10-year consortium, funded by the National Institute of General Medical Sciences at the National Institutes of Health, is now coming into the final stretch. Inspired by this drive, the four-year LipidomicNet project, funded by the European Commission, debuted in 2008 in Europe.

When LIPID MAPS got its start in 2003, “we had the notion of determining all the molecular species of lipids,” says Edward A. Dennis, a biochemist at the University of California, San Diego, who heads the project. But first, he and his fellow consortium members needed to come up with a way of classifying lipids. “The classification system for lipids that would be suitable for bioinformatics and data processing didn’t exist,” Dennis says.

So they created a classification system that divides lipids into eight categories—fatty acyls, glycerolipids, glycerophospholipids, sphingolipids, sterol lipids, prenol lipids, saccharolipids, and polyketides—according to their lipid backbones. The first six classes are found in all organisms, whereas the last two are found only in plants and bacteria. They also came up with a numbering system that gives every lipid its own 12-character identification code.

With that out of the way, they could turn to their real task—developing a method to quantitatively analyze individual lipid molecular species by MS. But they were stymied by the dearth of mass spectrometric lipid standards.

“Mass spectrometry as a technique relies on standards, because fragmentation patterns are different for every molecule,” Dennis says. Standards are typically stable-isotope-labeled versions of the compounds of interest. They help mass spectrometrists account for factors like recovery and ionization efficiency, which are essential for quantification. “Our first goal was to develop lipid standards for mass spectrometry. There were virtually none when we started,” Dennis recalls. “Today, there are more than 500.”

Ideally, a stable-isotope standard would be available for every lipid being quantified, but adding so many molecules to each sample isn’t practical. LIPID MAPS scientists spent a lot of time figuring out molecules that would be adequate internal standards for a broad range of molecules, says consortium member Alfred H. Merrill Jr. of Georgia Institute of Technology. For the sphingolipids, he says, they found that a single standard with an unusual chain length can suffice for all of its longer and shorter brethren.

The consortium members helped develop those standards, which Avanti Polar Lipids manufactures and sells. Now, everyone in the growing field of lipidomics can use them to quantify lipids.

Unless it’s quantitative, lipid analysis is useless, says Andrej Shevchenko, a researcher at the Max Planck Institute of Molecular Cell Biology & Genetics, in Dresden, who is not involved with LIPID MAPS.

Shevchenko explains why quantitation is the sine qua non for lipidomics, but not for proteomics: “The majority of proteomics work is not quantitative. If you identify a protein in such-and-such biological context, it usually means something. The protein has a character,” he says. “Well, a lipid doesn’t have a character. Lipids behave in a collective. You’re studying the collective behavior of certain similarly built molecules. That’s why lipid analysis is always 100% quantitative. We can speculate how accurate it is, but it’s absolutely ­quantitative.”

He goes on to liken proteins and lipids to knights and armies from the Middle Ages. “Knights are like proteins. Each knight has a pedigree, a story, an emblem, a flag. Lipids are like armies,” he says. “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.”

Classification system and standards in hand, LIPID MAPS members trained their eyes on the real prize: demonstrating that they could quantitatively profile lipids in a given kind of cell. As a test case, they chose mouse macrophages.

“Macrophages are a central lineage that you can isolate in blood, but they differentiate in all kinds of tissues,” Dennis says. Macrophages turn into alveolar cells in the lung, Kupffer cells in the liver, or microglial cells in the brain. They are precursors of the foam cells involved in atherosclerosis. In addition, mouse macrophages are easy to obtain and are available in a wide variety of genetic variants, making it possible to study the effects of genetic differences on lipid composition. LIPID MAPS has been able to quantitatively follow changes of some 500 lipids when macrophages are activated as in an infection.

To assemble the catalog of mouse macrophage lipids, the consortium developed a series of lipid-class-specific extraction protocols, to which they gave the moniker CLASS (for comprehensive lipidomics analysis by separation simplification). Consortium members took the view that, if they were going to construct a comprehensive lipid catalog, they needed to separate the lipids as much as possible before the molecules get to the mass spectrometer. They optimized extraction and separation conditions for each lipid class and then let the mass spectrometer separate any lipids still clustered together.

Using CLASS, Dennis’ laboratory has quantified more than 200 fatty acyls in mouse macrophages, including all of the major classes of eicosanoids, especially so-called resolvins and protectins. Derived from fish oil omega-3 fatty acids, these eicosanoids are being explored as novel anti-inflammatory drugs.

But CLASS is laborious and slow, and many scientists in the lipidomics field avoid it. They instead turn to “shotgun lipid­omics.” That is, they inject a lipid extract directly into the mass spectrometer with no prior separation.

“With some lipids and classes of lipids, you get almost identical results if you do shotgun” or separate first by liquid chromatography (LC), Wenk says. “With others, you don’t.”

“In principle, either method is okay, but you have to know the advantages and disadvantages,” Spener says. Spener’s lab uses LC methods similar to those advocated by the LIPID MAPS consortium. For shotgun methods, the main strength is speed. “Shotgun is a good method for quick discovery of differences” between two states, Merrill says. LC methods, on the other hand, are better able to capture low-abundance lipids, especially those involved in cell signaling.

Shevchenko advocates shotgun methods as being simpler and closer to the biological state. “It’s nice to be very comprehensive in getting low-abundance species, but it’s far more important to cover all the classes than all species in the class,” he says. “We might miss minor species in the major class, but we see all the lipid classes, and we see the major species within each lipid class.”

Xianlin Han, a researcher at Sanford-Burnham Medical Research Institute, in Orlando, combines shotgun methods with chemical derivatization to find low-abundance lipids. “Low-abundance lipid classes usually are the metabolites of abundant lipids,” he explains. “For example, eicosanoids are all oxidized either enzymatically or nonenzymatically from polyunsaturated fatty acids.”

All eicosanoids have a carboxylic acid that can be derivatized with an easily ionizable group. Other low-abundance lipid classes can be similarly derivatized at other functional groups. Derivatization makes low-abundance lipids more easily accessible by MS, even without separation methods.

Using the sample preparation tools and standards now available, researchers have been able to use MS to discover new lipids. Working with a collaborator at the U.S. Department of Agriculture, Merrill’s team identified a family of sphingolipids not previously seen in mammals. The sphingolipid backbone typically forms from the condensation of a fatty acyl-CoA and the amino acid serine. Merrill and coworkers found that the first enzyme in the pathway could also use alanine and glycine to form highly unusual ­sphingolipids. “Before we saw these with mass spec, there’d been no proof these molecules could be made by healthy humans,” Merrill says.

It’s not enough just to identify lipids. Researchers also want to know where those lipids are—which tissues and where in those tissues. For that information, they’re turning to mass spectrometric imaging, which converts spatial distributions of mass-to-charge ratios into pictures of the locations of different molecules.

With its second five-year grant, the LIPID MAPS consortium has added an imaging component. They are focusing on two types of mass spectrometric imaging—conventional matrix-assisted laser desorption/ionization (MALDI), headed by Murphy, and secondary ion mass spectrometry (SIMS), led by Nicholas Winograd at Pennsylvania State University.

Some lipid classes are better suited to MALDI imaging than others, Murphy says. For example, cholesterol, an abundant neutral molecule found in all plasma membranes, doesn’t ionize nearly as well as other lipids do. That makes it hard to detect with MALDI.

“We get ionization of neutral molecules, but the yield of those is much lower than the yield of secondary ions that are preionized in the physiological state,” Murphy says. “We see charged molecules much better.” A prime example is sulfatides in the brain. These molecules are sphingolipids with a sulfuric acid esterified to a galactosyl head group. Even though they are found at only low levels in the brain, “you can see them quite abundantly as negative ions,” Murphy notes. “It’s hard not to have them ionize with that sulfate.” Such easily ionized, low-abundance lipids can swamp neutral ones during ionization.

The spatial resolution of MALDI imaging is not yet good enough to see lipids inside cells. But even at such low resolution, imaging is revealing unforeseen things about lipids, Murphy says.

For example, working with tissue slices from mouse lungs, Murphy’s team found unexpected accumulation of arachidonic acid in the lung. “We expected it to be fairly diffuse in the various cells of the lung,” he says. Instead, cells lining the airways in the lung contained large amounts of arachidonic acid-containing phospholipids. Some of the mediators of arachidonic acid metabolism are involved in asthma, in which airways get restricted. One of the questions that Murphy is still trying to answer is whether the cells lining the airways are a reservoir of arachidonic acid.

SIMS imaging provides better spatial resolution than MALDI imaging, but in some circumstances, it can be a much harsher technique. In SIMS, a “primary” ion beam bombards a surface and dislodges secondary ions, which are analyzed by a time-of-flight mass spectrometer.

MaldI and SIMS imaging work at some of the same length scales, Winograd says, which allows researchers to check that both methods yield similar results. Then SIMS has the added benefit of better spatial resolution and three-dimensional capabilities. The 3-D aspect comes from acquiring images of underlying layers as a function of depth.

Imaging the mouse macrophages used in LIPID MAPS, Winograd’s team showed that SIMS with a C60 ion beam could detect a wide range of lipids. They imaged 10 species of sphingomyelins and nearly 50 glycerophospholipids, including 22 glycerophosphocholines, 12 glycerophosphoinositols, nine glycerophosphoglycerols, and five glycerophosphates, Winograd says.

Ion mobility combined with MS is another new technique researchers are applying to lipid analysis. Whereas MS separates species by mass-to-charge ratio, a proxy for size, ion mobility separates molecules by their collisional cross section, which is related to the shape of the molecule.

The additional separation dimension comes in particularly handy with lipids because so many species share the same nominal mass. It can also help prevent easily ionized lipids from suppressing the detection of other lipids.

Ion mobility easily separates major lipid classes such as the sphingolipids and glycerophospholipids, says John A. McLean, a chemistry professor at Vanderbilt University. With those two groups “the only real difference is in how the fatty acyl tails are connected to the phosphate and head group,” he observes. Sphingolipids use sphingosine, whereas glycerophospholipids use glycerol. “It’s an innocuous difference but it turns out to have a very strong effect on the prevailing structures the lipids adopt,” McLean says. “We’re able to discern them structurally using ion mobility.” After McLean’s team uses the mobility to separate the lipid classes, they depend on the resolving power of the mass spectrometer itself to separate individual lipids within those classes.

Michal Kliman, a postdoc in McLean’s lab, applied ion mobility to fruit fly brains in studying a Drosophila model of epilepsy. With this so-called bang-sensitive phenotype, fruit flies go into a comatose state and fall out of the air when their jar is banged on a countertop. Using a combination of MS and ion mobility, Kliman narrowed a field of approximately 1,200 lipids to the 38 lipids whose levels in the normal and bang-sensitive fruit flies differ significantly.

“All of a sudden, we could develop biochemical models to interpret a possibility of what was happening in the epilepsy model,” McLean says. “It was very important that we isolated only those signals corresponding to lipids.”

Ion mobility might also help with one of the challenges to identifying lipids—distinguishing between structural isomers. Simply moving the position of a double bond in a fatty acyl tail—for example, omega-3 versus omega-6 fatty acids—changes the lipid and its properties.

Using ion mobility, “we believe that you can discern the number of double bonds in the fatty acyl tails, but it still remains challenging to determine specifically where they occur,” McLean says.

Australians Todd Mitchell and Stephen Blanksby of the University of Wollongong are working on other ways to pinpoint the position of double bonds. Their first approach was to react lipids with ozone inside a mass spectrometer. The ozone reacts with double bonds in the chain and causes the lipid to fragment in specific locations during MS/MS experiments.

“Probably a lot of studies are really underreporting what’s actually there, because we’re not seeing these double-bond isomers,” Mitchell says. “Sometimes small changes in these isomers won’t actually show up in mass spec because they cancel each other out.”

Despite the many advances, MS can’t yet reveal certain key things about lipids. “We would like to know specifically where lipids are in a cell,” Murphy says. “That’s a fundamental piece of information we don’t have any techniques to get into right now.” Perhaps future techniques will allow scientists like Murphy to take MS into the cell without ripping it apart first. “There’s a lot of promise there,” he says, “but also lots of hard work to be done.”

With two years left before LIPID MAPS winds down, the consortium has accomplished the things it set out to do. “LIPID MAPS provided the catalyst to develop this whole new field of lipidomics,” Dennis says. “We see a tremendous expanding future.”

 
Chemical & Engineering News
ISSN 0009-2347
Copyright © American Chemical Society

Leave A Comment

*Required to comment