Gone are the days when lipids were thought of as simple storage or structural molecules. Lately, lipids are getting more respect for their roles in myriad functions, from cell growth and metabolism to immunity and inflammation. Researchers are recognizing that these molecules can reveal even more about a cell’s phenotype than gene expression does. Unlike proteins, which are genetically encoded, lipids are synthesized by enzymes in response to environmental factors such as diet and exercise. They have been implicated in disorders including Alzheimer’s, cancer, and heart disease.
“We need to understand lipids to know what the cell is doing,” says Markus Wenk, principal investigator of LipidProfiles, a multicenter lipidomics initiative headquartered at the National University of Singapore. Scientists are turning to mass spectrometry (MS) for that task. “We will see a tremendous acceleration [of research] over the next 5 or 10 years,” Wenk says—similar to the recent rush of studies in genomics and proteomics.
Lipid molecules boast more chemical diversity than proteins. The full roster of lipids in a cell, known as the lipidome, contains tens of thousands of lipids that fall into eight categories based on how they’re synthesized. Yet despite the diversity of lipids, they share some properties that make them amenable to mass spectrometry. MS instruments identify and quantify molecules by ionizing them, separating them based on their mass, and then detecting the charge of the ions of different masses, which allows the machine to determine each ion’s mass-to-charge, or m/z, ratio. Many lipids contain phosphate and nitrogen groups, so they readily ionize. In addition, lipids tend to weigh less than 1,000 daltons, which is the ideal mass range for many mass spectrometers.
But grasping the breadth of the lipidome in a cell type or tissue is not always straightforward. Many highly abundant lipids, such as the phospholipids that make up cell membranes, are easily detectable by MS, but they can hinder the detection of scarcer lipids like the fatty acid–derived prostaglandins, which are involved in cell signaling.
The Scientist asked lipid-MS experts for their tips on how to take a productive trip through the lipidome. Here’s what they said.
What is the lipid profile of my sample?
Mass spectrometry can reveal a lot about the lipid content of biological samples and how that content changes under different conditions—as in yeast cells grown at a range of temperatures or human plasma cells from donors who have fasted overnight, for example.
The first step in an MS analysis is extracting the lipids from the cells or tissue of interest using a suitable organic solvent. Next, you’ll face a fork in the road. You can inject your extract directly into your MS instrument—an approach known as shotgun; alternatively, you can first separate the lipids based on their differences in polarity or size using a liquid chromatography (LC) column.
The choice depends on the type of sample you have and what you are trying to detect. “For a very low quantity of any molecule, you really have to use the power of [MS] coupled with chromatographic separation,” says Robert Murphy, professor of pharmacology at the University of Colorado, Denver. By separating out lipid molecules, LC enriches them, making it easier for the instrument to ionize and detect them.
One round of LC might suffice, but to detect an even broader range of lipids, Edward Dennis, a biochemist at the University of California, San Diego, and his colleagues devised a method for separating each lipid category from a sample and conducting LC on these more homogenous extracts separately. “We feel that, in the long run, our approach has the ability to see the many thousands of lipids that would be present in complex human tissue samples in very low amounts,” Dennis says.
You can up the ante for detection even further by analyzing your post-LC sample in an instrument, such as the popular triple quadrupole mass spectrometer, that can perform two rounds of MS. In so-called tandem MS, the first round identifies the m/z values of your sample’s lipid ions, as other instruments do, while the second round fragments the ions and detects those fragments. It is often easier for the machine to detect ion fragments than the parent ions, Dennis says.
However, LC-MS is not without drawbacks. Chromatography usually adds a lot of time to the experiment, Murphy says. If you’re only looking at the major lipids—phospholipids, for example—then you can save that time by using a shotgun approach.
Researchers in the shotgun camp have also devised methods to improve detection of a range of lipids in complex biological samples, including low-abundance lipids. Many of the tricks involve tweaking the solution in which lipids are dissolved to make their extraction from the biological sample more efficient or improve their detection in the MS instrument.
One method entails siphoning off a portion of your lipid extract, raising its pH, and injecting just that portion into the MS instrument. This allows the lipid molecules in your sample that are less charged to ionize more efficiently. “It’s kind of like chromatography, but you’re enriching certain lipid classes by selecting what the machine can ionize,” says Michael Kiebish, a senior scientist at Berg Diagnostics in Natick, Massachusetts, who uses this approach to study lipid changes in different diseases.
Shotgun also lends itself well to high-throughput studies once you have already worked out the conditions, such as the extraction solution, to improve detection of a particular lipid. “If you want to screen a thousand samples, boy, it doesn’t sound like you want to do chromatographic separation,” Murphy says.
How can I process my MS data?
Running a lipid sample through an MS instrument will generate an array of m/z values and signal intensities for each. Software packages make predictions from these raw data about the identity and quantity of the lipid molecule that corresponds to each m/z value.
Separate software exists for LC-MS versus shotgun data. LC-MS programs rely on an additional set of raw data measuring the time it takes for a molecule to come off the chromatography column. These values, together with the m/z values and intensities, generate a three-dimensional chromatogram which helps identify the lipid.
XCMS (metlin.scripps.edu/xcms) is one popular free software package for analyzing LC-MS data, but it requires users to learn a programming language, says Gary Siuzdak, senior director of the Center for Metabolomics and Mass Spectrometry at the Scripps Research Institute in La Jolla, California, whose lab developed the program. An online version (xcmsonline.scripps.edu) launched last year does not require knowledge of programming. Users upload raw data, select from preset parameters such as the type of MS instrument used, and receive their results usually the same day. The program identifies lipids by comparing their m/z values to a repository of MS metabolite data developed at the Scripps center. XCMS Online presents the data in different graphs, such as one that compares lipid levels between two samples.
For more control over the analysis of your LC-MS data, MZMine 2 (mzmine.sourceforge.net/index.shtm) is a free, open-source program that lets you write your own algorithms or adapt them from other programs such as XCMS. MZMine has a graphical interface and does not involve programming language knowledge.
For shotgun data, many researchers develop their own software because fewer good commercial options exist, Berg Diagnostics’s Kiebish says. One option, however, is, LipidXplorer (sourceforge.net/projects/lipidxplorer/file), a free downloadable package that lets you type in commands, such as asking the program to identify certain lipids.
How can I visualize my lipid of interest?
MS has proven its mettle for being able to reveal the types and amounts of lipids in a sample, but the technique is also emerging as an indispensable way to visualize the distribution of lipids in tissues. Protein biologists can track down their targets with the help of antibodies or fluorescent tags, but lipid biologists have few available lipid-specific antibodies or dyes at their disposal. Mass spectrometry-based imaging, or MSI, “is a tool where you can get information about the location of specific lipids in tissues that would be difficult to do by other techniques,” Murphy says.
Matrix-assisted laser desorption/ionization (MALDI) imaging is probably the most widely used approach for mapping lipids, though protein biologists can use it too. The technique involves mounting a thin slice of tissue on a slide, then dousing it in an acidic matrix solution that absorbs ultraviolet light. The slide is then placed inside a mass spectrometer equipped with a MALDI source that directs a UV laser beam at the sample. The laser evaporates the lipids from the sample and ionizes them, releasing the ions into the mass analyzer. The slide passes under the beam on a stage at set points in an X-Y grid to create a mass spectrum for each coordinate.
The narrower the laser beam, the better the spatial resolution of molecules, says Richard Caprioli, director of the Mass Spectrometry Research Center at Vanderbilt University Medical Center in Nashville, Tennessee, who developed some of the early techniques in MS imaging. “You can routinely get down to 5 to 10 microns with MALDI” by upgrading to a laser and lenses that can generate this size laser spot, he says. The result is close to subcellular-level spatial resolution, and Caprioli’s group is working on improving it further still.
A technique called secondary ion mass spectrometry (SIMS) imaging is already achieving 1-micron or finer resolution of lipids, and has the most promise for determining the subcellular localization of lipids, Murphy says. SIMS is analogous to MALDI, except that it does not require treatment with an acidic matrix as MALDI does. Another difference is that the SIMS technique uses high-energy ions to ionize lipids. The ions hit a smaller area of the sample than the UV laser beam in MALDI and make it possible to focus on a smaller area for each coordinate on the X-Y grid.
But the higher spatial resolution of SIMS comes with a trade-off: as the ion beam focuses on a smaller area, some signal is lost because there are fewer molecules present to detect. “Once you get inside the cell it becomes less certain what’s going on,” says Nick Winograd, a professor of chemistry at Penn State University. Although adding a matrix step can enhance ionization and improve detection, Winograd’s group generally focuses on structural lipids, which are abundant enough to visualize by SIMS.
A newer imaging technique known as desorption electrospray ionization (DESI) ionizes lipids outside the MS instrument, making it possible to analyze a range of materials, including the surface of living tissue, such as a fingertip. Instead of a beam of UV light or ions, DESI hits the sample with charged droplets that dissolve the lipids, splash them off the sample, and ionize them. Although the spatial resolution is comparable to MALDI, DESI is faster because it does not require adding a matrix. It is also more affordable: a DESI ion source costs about $70,000 compared with more than $100,000 for the MALDI equipment, says R. Graham Cooks, professor of analytical chemistry at Purdue University in Indiana, whose lab developed the approach.