Maldi

A BRIEF HISTORY

In MALDI, a sample is mixed with an excess of matrix material and dried onto a target plate. The matrix absorbs energy at the wavelength of a laser, releasing it into the sample as heat. This causes the sample to vaporize and form ions, which can then be mass analyzed, typically in a time-of-flight mass analyzer (which determines an ion's mass based on the time it takes to reach the detector).

Prior to MALDI (and ESI), scientists usually used fast atom bombardment (FAB) or chemical derivatization of peptides in order to analyze them spectrometrically. But these methods were time-consuming and, in the case of FAB, expensive. "It's not that people weren't doing proteomics before MALDI and electrospray became available. But it really took off once these technologies became available," says Phil Andrews, director of the Michigan Proteome Consortium, whose lab has used MALDI to map protein masses directly from 2-D electrophoresis gels.

The technique's evolution began nearly four decades ago, with a device called LAMMA, or laser micropulse mass analyzer. LAMMA could report the mass of inorganic ions with submicron resolution via laser desorption of the analytes from an organic resin-based matrix; Hillenkamp was one of its developers.

"During the course of that development, I realized two things," says Hillenkamp. "First of all, that there was always a background of signals in our spectra, which were obviously coming from the organic matrix.... And, I realized that if [we used] different resins, we would see different fingerprint mass spectra." He concluded that it should be possible to use the technique to provide fingerprint spectra for organic materials and even macromolecules like proteins or DNA.

With Karas, Hillenkamp systematically probed every amino acid with a 266-nm neodymium-doped yttrium-aluminum-garnet (Nd:YAG) laser, observing that aromatic amino acids like tryptophan and tyrosine could be desorbed with less energy than aliphatic ones, but that if two different types of amino acids were mixed, both were desorbed simultaneously. Thus, Hillenkamp concluded, "if you have one absorbent component in your mixture, then more difficult to absorb molecules will fly along with it." The pair published their observations in 1985, marking MALDI's official debut.¹

Initially the technique was limited to small organic molecules such as amino acids. Hillenkamp and Karas extended it to the mass range of larger peptides, but it was Koichi Tanaka of the Shimadzu Corp., in Kyoto, who first made the technique work for large proteins. Tanaka, who would go on to win the 2002 Nobel Prize in chemistry for his work, dissolved the proteins in a suspension of ultra-fine metal powder and glycerol so that intact macromolecules such as proteins could be ionized. In 1988, he and Hillenkamp/Karas simultaneously demonstrated the technique's ability to generate mass spectra of proteins with masses over 20 kDa and 100 kDa, respectively.

KEY DEVELOPMENTS

MALDI has undergone several key modifications since 1988, including the matrix used. Hillenkamp and Karas' first efforts employed nicotinic acid, but that matrix didn't work very well. "The spectra they got of proteins were extremely broad peaks, and it was hard to get accurate masses from [them]," recalls Brian Chait of Rockefeller University in New York.

So Chait, then-postdoc Beavis (now CEO of Beavis Informatics in Winnipeg, Canada), and Hillenkamp, set out to find better formulations. Their results – derivatives of cinnamic acid, along with 2,5-dihydroxybenzoic acid – are still the materials most widely used today. "All of the good matrices were found empirically in the first years," notes Karas.

The laser was also modified, from Nd:YAG to a less expensive, yet more efficient, 337-nm nitrogen laser, as was the system geometry. Early MALDI-TOF systems used an axial configuration, in which the laser is in line with the target and flight tube. Orthogonal technology (introduced commercially in the late 1990s), in which the MALDI source is at a right angle to the detector, eliminates sample discrepancies so that they are not amplified at the end of the run, giving these instruments up to 3 ppm mass accuracy with external calibration, explains Armin Holle, director of life science research and development at Bremen, Germany-based Bruker Daltonics, a mass spectrometer manufacturer.

"With an axial MALDI-TOF, you'll never be able to do that with an external calibration, because slight height differences in the sample make slight differences in flight time, and again make the mass assignment a bit wrong," he says. Axial instruments are still the best choice for experiments that require high sensitivity, however, as they can analyze samples down to a few attomoles.

One of MALDI's most important improvements was the introduction of delayed ion extraction. Developed in the late 1950s, this technique was first employed in commercial MALDI-TOF instruments by Marvin Vestal of PerSeptive Biosystems (now Applied Biosystems) in 1995 to compensate for differences in the initial velocities of ions.

If two ions of the same mass start off with different velocities and are accelerated down the flight tube, they will hit the detector at different times, leading to peak broadening and low resolution. Delayed ion extraction varies the electrical field each ion experiences based on its initial velocity, so that "in the end, it's all made such that [all ions] arrive at the same time on the detector," says Holle.

ENTERING THE PROTEOMICS REALM

Despite their many advantages, early MALDI-TOF instruments had a major limitation: They reported an analyte's molecular weight, but not its structure. The instruments thus could not be used to determine a protein's sequence or post-translational modifications – data that is critical to protein chemists.

To overcome this, post-source decay (PSD), in which ions that fragment in the electrical field-free region of the flight tube are separated with a reflector, was introduced in the early 1990s. "PSD was kind of a compromise; it allowed you to fragment the molecule so that you could get ... some sequence information, so that helps you in identifying peptides and then the proteins from which they come," says Michael Gross of Washington University in St. Louis.

Another way to obtain sequence information for peptide identification is to use tandem mass spectrometry (MS/MS). "In a tandem mass spectrometer, what you get is a fingerprint of the amino acid sequence for a peptide, and that's a much more specific piece of information to use in a database search," says John Yates of The Scripps Research Institute in La Jolla, Calif.

The first tandem MALDI-TOF/TOF instrument was developed by Vestal in the late 1990s and commercialized as the 4700 MALDI-TOF/TOF by Applied Biosystems of Foster City, Calif., the current market leader in MALDI instruments. "Bringing this advance of MS/MS on the platform, and adding things like autoloaders, we brought the whole concept of high-throughput definitive proteomics into a whole new realm, and invented a whole new class of instrument," says Dave Hicks, Applied Biosystems' senior director for proteomics.

THE FUTURE OF MALDI

Because MALDI is so versatile, current-generation systems are not limited to TOF analyzers; MALDI can be coupled with ion trap and quadrupole instruments, for example. "It's now being adapted and marketed and configured as an ionization option, or as an ionization source on a variety of different mass analyzers. So it's not just TOF or TOF/TOF any longer, it can be used in a variety of ways," Hicks points out.

And MALDI ion sources can provide even greater resolution and mass accuracy when coupled to highly sophisticated Fourier transform (FT)-MS instruments. "The future ... may very well be MALDI-FT instruments, which would have resolving powers 100 times better than the current best TOF instruments," says Gross, who adds that this type of instrument has not yet taken off, for a variety of reasons. "One of the beauties of the MALDI-TOF instrument is that the TOF is quite simple. And the FT instrument is a little bit more complicated, and more expensive. It requires a superconducting magnet like NMR [nuclear magnetic resonance] does," he explains.

Indeed, it is MALDI-TOF's simplicity that makes it so popular. Cornell University's Lee notes that MALDI-TOF instruments are more accessible to the nonexpert than other MS instruments, including ESI. "You no longer have to be trained as an analytical chemist to get high-quality data on characterizing your biomolecules of interest – the technology's evolved now far enough, and it's high enough performing, that many more investigators can begin to access that tool," says Lee. And not just in proteomics labs. MALDI-TOF instruments are also being used for genotyping analysis, biomarker identification, pathogen detection, and even imaging (see Sidebar).

ESI-based instruments will likely remain complementary, and required components of the proteomicist's toolbox for some time to come. For one thing, ESI and MALDI often detect significantly different protein sets. "There may be 30 or 40 percent of the peptides that are identified by MALDI that are not identified by electrospray, and vice versa," says Andrews of the Michigan Proteome Consortium.

And, unlike MALDI, electrospray sources can be coupled online to liquid chromatography systems, reducing liquid handling steps and allowing the analysis to be done "on the fly." According to Gross, "If you want real-time information, then MALDI is not appropriate."

Offline analysis has its advantages, too, however. In LC-ESI, the sample is analyzed as soon as it comes off the device. But, if the separated proteins are spotted onto the MALDI target after the LC run, they can be archived. "[If] you store the chromatogram, you can reinspect [the sample. You have as much time as you want for analysis," Karas says.

Having watched the development of MALDI technology, Franz Hillenkamp is no longer surprised by its popularity, but he is pleased by it. "If you have done something in your life that has found so large scale and widespread acceptance by the colleagues who find it very useful, that is certainly something very satisfying. Not surprising, but very satisfying."

The Evolution of MALDI