Like important inventions throughout history, mass spectrometry has taken a little while to realize its full potential. As sensitivity and throughput improve, researchers are finding it a more useful tool than its inventors could have imagined.

click the image to enlarge   IMS analysis of a 12 µm rat kidney section. (A) Partial survey protein profile in the m/z range from 9850 to 10050, obtained after homogeneous matrix deposition on the section, showing two strong signals at m/z 9943 and m/z 9979. (B) to (D), Ion density maps obtained for m/z 9943 and m/z 9979 with an imaging resolution of 250 µm. (Source: Richard Caprioli, PhD, and Pierre Chaurand, PhD, Vanderbilt University)

Some inventions are cool in and of themselves. The Slinky. The iPod. The massage chair. But for most, it's not the invention itself that is exciting. Take the wheel, for example. Pretty solid invention there, but by itself, the wheel is just a round thing with some spokes. Sliced bread, too. In and of itself, not so exciting. But take that sliced bread, plop a juicy Swiss-cheese hamburger on top, add a slab or two of hickory-smoked bacon, and now we've got something. Attach something cool to the wheels, something like a Ferrari F430 or one of those funky, impractical Segways, and now you've got a zippy invention.

The mass spectrometer is far from the coolest invention ever. It is a somewhat pedestrian device that measures the mass-to-charge ratio (m/z) of molecules, such as proteins and peptides, that have been electrically charged. But now consider what it can do, and then we've got something. Tasty burgers aside, mass spec is much more versatile than bread, and it may some day get as much mileage as the wheel. Some day soon.

Mass spectrometry provides valuable information to a wide range of professionals including chemists, pharmacists, physicians, engineers and astronomers. It's a cross-functional star that is just now realizing the upper limits of its potential. Advances in mass spectrometry instrumentation have added levels of sophistication to the technique and expanded the use of the devices. Tools such as high high-resolution mass spectrometers, electrospray ionization techniques, and hybrid instruments have led to increased sensitivity and higher throughput. Now, life science researchers are looking to push the outer limits of the techniques both in traditional areas such as biomarker discovery, and in new ways, like tissue imaging and biodefense.

Putting the air back into biomarker discovery
No area of study within proteomics has been more of a long-term tease than the search for and validation of protein biomarkers. Leigh Anderson, PhD, founder and CEO of the nonprofit Plasma Proteome Institute, Washington, D.C., is perhaps more familiar with this frustration than most. "The discovery and implementation of good biomarkers has not been as rapid as we would all like to see it be," says Anderson.

Most researchers shy away from working with plasma proteins because of the complexity of the research. Blood is rich in proteins and is probably equally rich in potential drug targets, if one can sort through the daunting number of high-abundance proteins to detect low-abundance proteins. The latest mass spectrometers allow researchers to do just that. Anderson is using mass spec to do what he says has been done in analytical chemistry for a long time. "Our objective is to employ the most capable mass spec systems for quantitation, rather than the typical interest in proteomics, which is to do discovery," he says. "We're primarily interested in emulating the approaches used in analytical chemistry to measure small molecules, like drug metabolites in clinical serum specimens."

Add a Little Salsa Researcher Daniel Liebler, PhD, professor of biochemistry and pharmacology, Vanderbilt University, Nashville, Tenn., co-developed software that specializes in searching for posttranslational modifications. Called SALSA, for scoring algorithm for spectral analysis, the tool enables researchers to identify protein targets for chemical modification and to map modifications at the level of amino acid sequence. Liebler describes SALSA as a "tool for discovering proteomic diversity [generated by] the various modified forms of proteins that result from damage by toxic chemicals and even endogenous reactive intermediates." Liebler says that advances in mass spectrometry, including new linear ion trap instruments, can "rapidly acquire [tandem mass spectrometry] data on complex peptide mixtures." He is using SALSA in combination with another tool he co-developed, called P-Mod, which uses peptide sequences to search tandem mass spec (MS-MS) data and assign correspondence between sequences and MS-MS spectra. "The newer instruments can acquire MS-MS spectra more rapidly than older instruments, which increases the likelihood that modified forms are detected." That's zesty indeed.

Anderson believes that there are two ways to approach biomarker validation. The first is the traditional approach with immunoassays, which he acknowledges is an effective method. "The alternative is to think about using mass spectrometry as the detection step," he says. "That's the approach we're taking. And it looks most productive as a near-term solution to the problem of rapidly making a lot of different assays that can be tested. The jury is still out, and will be for a long time, with respect to what's the best way to implement clinical assays."

Many researchers have turned to antibody arrays (See "Antibody Arrays Are a Powerful Tool for Proteomics," page 26). "That's a great way to go," says Anderson, "but it's expensive and time-consuming to make these antibodies. A rigorously optimized commercial immunoassay costs approximately $2 million to $4 million to develop. That's what motivated us to look at a platform to identify candidate markers using mass spec. If, for example, we set up assays that required very high sensitivity, we could use one antibody, together with the mass spec as the second antibody. The advantage of doing that is that mass spectrometers, particularly triple quadrupoles, are very quantitative. And they give highly reproducible measurements. Secondly, because of the resolution and sensitivity of mass spectrometers, you can get absolute analyte specificity."

Using a triple quadrupole mass spectrometer, researchers can select a particular peptide mass, fragment it into very specific pieces, and select one piece to look at more closely. Advances in mass spec have allowed for resolution high enough that researchers can be certain they are measuring the peptide they intend to measure, thus bypassing a lot of the difficulty of generating immunoassays.

Anderson describes a cancer biomarker study the Plasma Proteome Institute is conducting, in which researchers found 1,400 candidate biomarkers which might be related to cancer. "If you take that as a starting point, then a useful alternative strategy to coming up with validated panels of high specificity in a disease state is to try to prioritize those lists and make specific assays using mass spec, and then measure them very precisely," he says. "The advantages of doing it this way are that using specific assays instead of a general discovery approach, you can potentially get more sensitivity, and you can get much higher quantitative accuracy. Those are two features that will be critical in the optimization of collections of proteins you'd like to evaluate together to get a panel result.

"There are many good reasons to believe that a collection of proteins can give a much more specific and a much more sensitive indicator than a single protein will," says Anderson. "We need to be looking for these panels. Evaluation of these candidate markers, the ones that exist and the ones in the process of being discovered, is really the rate-limiting step in getting more markers all the way through to the clinic. We've been concentrating on developing the technology for that, and exploring how it will be possible to fund this evaluation effort."

One logjam has been—surprise!—lack of funding. Anderson is blunt in his view that the National Institutes of Health (NIH), Bethesda, Md., hasn't played its part in funding biomarker research. He says sardonically that, "Historically, the NIH has not funded anything as practical as serious concerted marker validation. The diagnostics industry hasn't done it either, mainly because it's a much smaller industry than the pharmaceutical industry and it isn't really funded to do basic research."

Rolling the DIGE
The successful exploitation of underground petroleum in Titusville, Pa., in 1859 marked the beginning of the oil drilling industry in the United States. At the time, petroleum

click the image to enlarge   DIGE technology enables quantification with statistical confidence for 2D gel experiments. It can be used to analyze relative stochiometry and/or posttranslational modifications of proteins in defined complexes. (Source: David Friedman)

wasn't good for much more than lighting lanterns, until the internal combustion engine was developed for use in the automobiles of the late 1800s. It wasn't long before striking oil meant a little more than being able to read Edgar Allan Poe during the evenings.

Mass spectrometry isn't likely to change the face of transportation as we know it, but in combination with two-dimensional (2D) gel electrophoresis, it may shed enough light to reveal differences in protein expression between normal and cancerous cells, among other things. David Friedman, PhD, is working with 2D difference gel electrophoresis (DIGE) technology from GE Healthcare, Chalfont St. Giles, UK, to find subtle protein abundance changes in tumor tissue.

"There are several different technology platforms to solve problems in proteomics," Friedman says, "such as gel-based, LC/MS-based, and more targeted approaches. We try our best to cover all those bases, because we believe all of these technology platforms are very complementary. You don't just do gels, or just a shotgun approach. Each technology has its own strengths and its own unfortunate weaknesses, but they complement each other very nicely."

Friedman uses mass spectrometry for protein identification purposes and for posttranslational modification binding. "The MALDI-TOF-TOF [matrix-assisted laser desorption/ionization time-of-flight] instruments have greatly aided our ability to identify low-abundant proteins," he says. "We could take the samples to an ion-trap-type instrument, but that's a very time-consuming experiment. For a low-abundant protein, you can barely see it in the gel, you have an abundance issue. You also have [the number of] trips and cleavage sites there are and how they are distributed."

Combining mass spec with DIGE has allowed Friedman to do global differential display analyses with multiple variables, with repetition built into every experiment. "We use DIGE for any gel-based approach we take. At the quantification stage, it gives you very high levels of confidence in even subtle protein changes. We can measure the repetition and look at multiple drug doses over a time course, multiple genotypes, all at the same time. If you look at a lot of the proteomics being done out there, people do these fantastic experiments, but don't repeat them. It's an n of 1. The data are very fascinating, but is it biologically significant? With DIGE, we can build that into the experiment, run these fantastic comparisons, and have high statistical confidence."

Searching for biological relevance has been a staple of Friedman's research. He admits that it drives him nuts when he sees research that lists measurement of proteins but with no bearing on the biological and statistical significance of those measurements. "It could be that the technician sneezed in the culture when they grew up the control," he says. " 'Oh, wow, we see a change!' And when you go in and identify it, it's human mucosal material."

"DIGE technology is still young, but I do see it growing. It's not the end to the experimental design, but it's a means to an end. You do all this DIGE work, and you get proteins that are changing, but the researcher still has to go back and validate these in some way, show some other biological significance."

A current focus of Friedman's research is protein resolution and quantification. Using a gel-based approach as opposed to the shotgun approach, he and his colleagues are studying the upregulation and downregulation of isoforms. "We can actually see these different charged isoforms on the 2D gels, and in some cases, one isoform is up-regulated while another isoform is down-regulated. You can catch that on the gels, and with DIGE, you can quantify that with statistical confidence. But if you [used] the shotgun approach, depending on how upregulated or downregulated the isoforms were, you'd never even see that that change occurred, unless you had mass spec information on those peptides that contained the modification."

Nice gadget, but what can it do?
Early cellular telephones were often more hassle than they were worth. Models such as the Motorola DynaTAC 8000X were roughly the size and weight of a cinder block and had minimal battery power. They were impractical enough that for years they took a back seat (bad pun intended) to carphones. Incremental improvements in size, power, and functionality have made the cell phone as ubiquitous as Paris Hilton.

Early mass spectrometry instruments were far from ideal for work in tissue imaging, and thus were rarely if ever used for the task. A problem arose in vibrational decomposition of molecules, says Richard Caprioli, a professor of biochemistry and director of the Mass Spectrometry Research Center at Vanderbilt University, Nashville, Tenn. "You could see parts of some lipids, but you never even got the whole lipid," he says. "When MALDI was invented a number of years ago, it occurred to me that it might get around the problem, and in fact it did."

It turns out, says Caprioli, that huge proteins can be imaged. "We just [imaged] something of a molecular weight of 247,000. Of course, there are all kinds of problems along the way. When proteins get above 100,000 molecular weight, they become harder and harder to analyze. Even though we've done things at 200,000 to 300,000, the average person whose lab is not dedicated to this would find it hard to do," he says. "Certainly the major advantage of our approach is that it allows proteins and very large molecules to be imaged, as well as the small ones."

The ionization process in MALDI is much gentler than earlier mass spec iterations, which is easier on molecules. "Any other way of trying to put energy into the molecule, or even the laser directly, without the matrix, does not work well," Caprioli says. He compares the output from the mass spectrometer to an image from a confocal microscope. "But the advantage is that you get that image in only a single molecular weight," he says. "It's a molecular-weight-specific image. It's like looking at a piece of liver, and saying, I want to take a picture, but all I want to see is the protein of 27,327. The way you'd do that today would be to raise an antibody to that protein. Then you'd have to put a fluorescent tag on it, and you'd see that protein light up. All we do is put it in the mass spectrometer, hit a button, and we can see the protein light up."

MALDI mass spectrometry has applications in biology and medicine, anywhere data on the spatial distribution of molecules is important, says Caprioli. For this reason, pathologists are beginning to use mass spectrometry to determine which molecular signatures are present in diseases, with the goal of better classifying disease. In the molecular age, pathologists are desperate to move up and do molecular diagnosis. MALDI techniques offer that. Still, there's a long way to go before mass spectrometry gains wide acceptance and use among pathologists. "We're only at the beginning of the technology cycle," Caprioli says.

Future advances will have to improve sensitivity and mass range. But anything that can be sliced and sectioned can be put into a mass spectrometer. Caprioli cautions, however, that sample preparation is not a trivial matter. "It has to be done carefully, but someone who has lab experience but no tissue experience can do this fairly quickly."

Much of Anderson's work is being done with triple quadrupole mass spectrometers from Applied Biosystems (ABI), Foster City, Calif. ABI has developed both advanced mass spectrometry instruments and corresponding software for analysis. "We have some software tools that allow our users to ultimately take images of tissues," says Joseph Anacleto, PhD, senior director of the company's small- molecule division. "It is a fairly straightforward application, despite the delicate sample preparation. You're taking a tissue, freezing it, slicing it very thin, placing it onto a MALDI plate, putting some matrix on it for it to form ions, and once you get to that stage, you put it into a mass spec instrument."

Much of Caprioli's work is directed toward singling out diagnostic and therapeutic markers and better understanding protein-protein interactions. Anacleto says there are additional applications in drug discovery. "[Our customers] use it to look for distribution of drugs and drug metabolites in various tissues. Is the drug going where you expect it to go? What happens to the drug in that tissue; is it being metabolized?"

Anacleto says that drug discovery work is predicated on speed, sensitivity, and throughput. "The mantra in pharma is always 'fail fast, fail cheap'. You want to screen more things, get better quality target lead compounds that you're going to further develop. These are efforts that push the limits of mass spec from a performance and throughput aspect."

Many of Applied Biosystems' customers are also interested in using mass spectrometery for biodefense. "[The Centers for Disease Control and Prevention] uses one of our systems for a variety of applications," says Anacleto. "They're looking for chemical agents and metabolites of chemical agents. If you suspect an exposure, you can take a urine sample for certain metabolites and look for protein signatures. A lot more work will be done in this area in the future."

Next up for mass spec
The next generation of mass spectrometers will be faster and have higher sensitivity, providing even greater utility for life science applications. "The greatest opportunity is in the clinical area," says Vanderbilt's Caprioli, "where this molecular diagnosis of disease will certainly be very important. I'm collaborating with many clinicians now, and I see that only increasing in the future."

Asked what the ultimate therapeutic benefit of advances in mass spectrometry will be, Vanderbilt's Friedman jokes, "Well, do you want me to cure cancer?" All kidding aside, he says he expects simply "more sensitive and more accurate protein identifications. It all depends on the biology. The best you can hope for the technology is to give you better, more accurate pictures of what you're analyzing. But win or lose, it all depends on the samples that you bring to the instrumentation."

The great inventor Thomas Edison once said he never perfected an invention that he did not think about in terms of the service it might give others. Researchers using mass spectrometry in innovative ways have clearly taken those words to heart.

This article was published in G & P magazine: Vol. 5, No. 4, May, 2005, pp. 12-16.