by Jeffrey M. Perkel
Today’s mass spectrometers are faster, more sophisticated and more sensitive than ever. Yet if the molecules that are fed into these machines aren’t charged, the machine is as blind as a bat.
That’s because mass spectrometers don’t actually measure a molecule’s mass, but rather its mass-to-charge ratio (m/z). Given a molecule’s m/z value and its charge state, researchers can compute the mass, but to do that the molecule first must be vaporized and ionized.
“If you don’t ionize or you can’t ionize, then you can’t do mass spectrometry,” says mass spec expert R. Graham Cooks, Professor of Chemistry at Purdue University. Naturally, then, ionization sources are critical components of any mass spec setup. So critical, in fact, that two of the most popular life-science ionization techniques, ESI (electrospray ionization) and MALDI (matrix-assisted laser-desorption ionization), won their inventors the Nobel Prize in Chemistry in 2002.
ESI and MALDI revolutionized biological mass spectrometry, and especially proteomics, because they are “soft”: They ionize biomolecules without smashing them to pieces in the process. Yet they are not the only ionization methods available to life-science researchers today. There are APCI and APPI, DESI and DART and LAESI, CI and EI and more. “Here’s the deal with ionization methods,” says Neil Kelleher, the Walter and Mary Elizabeth Glass Professor in the Life Sciences at Northwestern University, “it’s a bit of an alphabet soup—lots of acronyms for lots of new methods.”
In part, that’s because researchers still have not figured out how to overcome two fundamental issues: sample preparation and ionization efficiency. Sample ionization, and transmission of those ions into the mass spectrometer, are “the most inefficient parts of the [mass spec] process,” says Steve Smith, senior director of product management for mass spectrometry at Waters. Cooks notes that dozens of new ionization techniques have been developed in the past five years. “That wouldn’t be true if there weren’t still problems [with existing methods].” Fortunately, the alphabet soup of ionization approaches has sufficient diversity to handle just about any class of molecule.
MALDI and ESI
Mass spec ionization techniques can be subdivided into several categories. Most interface with chromatographic methods, including both liquid chromatography (e.g., ESI, nanoESI, APPI and APCI) and gas chromatography (e.g., EI and CI). Others handle surface-based samples that are either processed off-line (e.g., MALDI) or analyzed directly without preparation, that is, in situ, (e.g., DESI, DART and LAESI).
MALDI (available, for instance, in Bruker Daltonics’ microflex™, autoflex™ speed and ultrafleXtreme™ series of instruments) is most appropriate for use with a two-dimensional gel electrophoresis (2DGE) workflow, in which protein spots are excised from the gel, processed and spotted onto MALDI plates for MS analysis.
In MALDI, a sample is mixed with a UV-absorbing crystalline matrix material, such as 2,5-dihydroxybenzoic acid and alpha-cyano-4-hydroxycinnamic acid, and spotted onto a metal target plate. The plate is then inserted into the MS instrument, where it is placed in a vacuum (though an atmospheric-pressure variant, AP/MALDI, is also available from MassTech) and hit with a UV laser. The matrix absorbs the irradiation, heating and volatilizing the sample and ionizing it at the same time.
Generally, MALDI imparts a +1 charge to proteins (with an occasional +2 or +3, as well), which both simplifies and complicates downstream analysis. On one hand, mass calculation is trivial, as m/z = m for z of +1. But the +1 charge also makes it more difficult to analyze intact proteins, as their large size pushes their m/z values outside the “sweet spot” of most mass spectrometers.
The +1 ions also do not respond well to fragmentation, says Kelleher, making MALDI a tougher ionization source for use in tandem mass spec applications (such as post-translational modification analysis and peptide sequencing).
ESI, in contrast, produces a range of charged species for each molecule: +2, +3, +4 and so on. That kaleidoscope of ions complicates mass analysis but greatly enables tandem mass spec work. And, because it is a liquid-based method, ESI is compatible with the chromatographic separations so often used in biosample analyses, from ultra-fast UPLC to low-flow nanoLC.
Kelleher recently published a study in which he used top-down proteomics (an approach in which proteins are analyzed intact, rather than as peptides, to study their post-translational modifications) to identify and characterize some 3,000 protein species produced from 1,043 human genes.  The study was driven by a pair of high-end ESI-powered mass spectrometers coupled to a “nanocapillary reversed-phase liquid chromatography” system, including a Thermo Scientific 12-Tesla LTQ FT Ultra Fourier-transform ion cyclotron resonance MS and a Thermo Scientific Orbitrap Elite.
John Yates, a mass spectrometrist at the Scripps Research Institute in La Jolla, Calif., also favors ESI-based mass specs, mostly Orbitraps. His lab, though, tends to build its own low-flow sources rather than using off-the-shelf solutions. “Since we make our own columns, it’s easier to use our own source than to try to fit to commercial ones,” he says.
Yates’ lab recently acquired a new Thermo Scientific Q Exactive, a hybrid quadrupole-Orbitrap instrument. Featuring high mass resolution and accuracy and fast scan speeds, the instrument, he says, “is smoking.”
A class issue
According to Smith, the key variable to consider in choosing an ionization source is the kind of molecule a researcher is looking for. “Each method is good for molecules of a certain polarity,” he says.
ESI most effectively ionizes relatively polar molecules, especially proteins and peptides, says Keith Waddell, director of LC-MS marketing at Agilent Technologies. (The company offers a custom form of ESI, called Agilent JetStream, which uses heated nitrogen gas to concentrate and direct the electrospray into the MS inlet, increasing sensitivity five- to 10-fold, Waddell says.)
For less polar metabolites and other compounds, such as steroids, researchers can consider ESI variants such as APCI (atmospheric pressure chemical ionization) and APPI (atmospheric pressure photoionization). APCI uses an ion molecule reaction to impart charge, whereas APPI uses light energy to do the same thing.
Waters offers as a standard component of its mass specs a dual-mode ionization source called ESCI, which rapidly alternates between ESI and APCI modes.
For highly nonpolar molecules, Waddell says, researchers should consider gas chromatography (GC)-based approaches instead. GC most commonly employs one of two ionization methods: EI (electron impact) or CI (chemical ionization). In EI, molecules are ionized by collision with electrons produced by a heated filament, generally by “chipping off an electron” and producing a positive charge, says Waddell. CI does the same thing, but in the presence of a gas such that the gas becomes charged; collision with the gas ions ionizes the sample instead.
Surface-based ambient ionization
According to Cooks, one of the key issues in biological mass spectrometry today is sample preparation. “Mass spectrometrists are not spending their time with the samples in the instrument, they spend most of their time on chromatography,” he says.
This is especially true if researchers are interested in radically different classes of molecules, as an extraction method that works for, say, sugars, will not capture lipids. Many researchers, Cooks says, would prefer to analyze samples directly—that is, to read their molecular profiles without all the sample preparation and chromatographic acrobatics. “Put simply, the desire people have is the desire to do mass spectrometry in situ,” he says.
Enter ambient ionization methods, which enable researchers to analyze the chemistry of biological samples with little or no sample preparation by essentially blasting ions off their surfaces.
According to Cooks, some 40 surface-based ambient ionization approaches have been described in the literature in the past five years or so, and four have been commercialized. One, developed by Cooks and coworkers, is DESI (desorption electrospray ionization).
Commercialized by Prosolia for instruments from AB Sciex, Agilent Technologies, Bruker Daltonics, Leco, Thermo Scientific and Waters, DESI essentially directs a beam of solvent at a tissue sample on a slide. As the solvent pools, it extracts some of the molecular components of the sample, which then splash up into the MS inlet upon subsequent solvent impacts, yielding an electrospray.
According to Cooks, ambient interface methods can be used in one of two ways: “point-and-shoot” and MS imaging. In the former case, the goal is to take a snapshot of the sample’s molecular profile, for instance, to test a food sample for contaminants; in the latter, researchers collect spectra at various points across a sample (that is, pixel by pixel).
The advantage of MS imaging is that instead of simply quantifying the bulk abundance of various compounds in a sample, researchers can produce a kind of spatially resolved map, or image, of the sample’s molecular composition. Vanderbilt University mass spec expert Richard Caprioli, who perfected the use of MALDI as an imaging ionization method, has compared the resulting images to the red/green/blue channels of a digital image, except that in this case, each “channel” corresponds to a single metabolite. 
One disadvantage of ambient ionization methods, says Cooks, is sensitivity: Because there is no sample preparation and enrichment, molecules of interest may tend to be lost amidst the signal produced by far more abundant, but irrelevant, compounds. Furthermore, ion suppression, in which one molecule inhibits the ionization of another (thereby limiting sensitivity) can be particularly problematic in such experiments, he says.
MALDI imaging is not actually an ambient approach, as it occurs in a vacuum and requires sample preparation. Nevertheless, it is an alternative, complementary imaging approach. Kelleher recently has begun using that approach (in which MALDI matrix is sprayed on top of a tissue slice on a MALDI target plate) in his top-down work, getting a first-pass surface analysis of potential biomarker molecules using MALDI imaging, followed by a more detailed, “grind-and-find” ESI MS-based analysis to actually characterize those proteins in detail.
“There’s a good link between top-down proteomics and MALDI-based imaging,” Kelleher notes. “When people do MALDI imaging they do often direct analysis, top-down—they don’t digest.”
Another recently introduced imaging approach is LAESI (laser ablation electrospray ionization), commercialized by Protea Biosciences as the LAESI DP-1000 Ionization System. Developed by Akos Vertes at George Washington University, LAESI uses a mid-IR laser (2,940 nm) to excite the O-H bonds in water, acting as a kind of endogenous matrix. Gas-phase particles are created from the ablation of the sample and then ionized through interactions with an electrospray ionization plume, all at ambient pressure.
According to Alessandro Baldi, vice president and general manager at Protea Biosciences, the advantage of this approach compared with MALDI imaging is that it requires no matrix, simplifying the experiment and producing cleaner spectra, especially in the low m/z range. It also produces multiply charged ions, as does ESI. But perhaps most importantly, he says, LAESI allows researchers to take three-dimensional molecular profiles by repeatedly digging into a sample and reanalyzing it.
“LAESI can scratch the surface and then go deeper,” Baldi says. “So it’s a difference between seeing what is on the surface versus what is inside the biological entity. And this can be performed directly on living cells.”
The technique produces pixels approximately 200 microns in diameter and about 50 microns deep, Baldi says. The LAESI DP-1000, which was named one of the top 10 innovations of 2011 by The Scientist magazine , costs about $250,000.
Before you buy
By all accounts, it is relatively simple to install a new ionization source—a matter of removing and installing a few bolts, plus electrical and gas connections.
“These are very interchangeable,” says Iain Mylchreest, vice president of research and development for chromatography and mass spectrometry products at Thermo Scientific. “Typically they are small ‘source housings’ or probes that bolt on to the front of the instrument. None requires any kind of service intervention. It can take minutes to switch, for instance, from ESI to APCI.”
Expect to pay less than $30,000 or so for most standard sources (e.g., APCI and APPI), Waddell says. Third-party sources may cost more, but the bigger issue is physical compatibility: Instrument architecture varies among vendors and across instrument classes (e.g., ion trap, time-of-flight (TOF), quadrupole), so you’ll need to make sure your desired ion source can interface with the instrument in your lab.
And don’t be surprised if you need to purchase another ionization source in the years ahead. As research priorities change, it’s always possible that you’ll find yourself studying a new class of molecules that just doesn’t respond well to the technologies you have on hand.
“There’s no universal ionization technology out there,” Mylchreest says.
 Tran, JC, et al., “Mapping intact protein isoforms in discovery mode using top-down proteomics,” Nature, 480:254-8, 2011.
 Perkel, JM, “Mass spectacle,” The Scientist, 23(3):61, March 2009.
 The Scientist Staff, “Top 10 Innovations, 2011,” The Scientist, January 2012.
The image at the top of this page is Thermo Fisher's Q Exactive.
Editor's Note: The article, “ESI Versus MALDI For Protein Characterization By Mass Spec,” (May 8, 2007) has been updated with the current article.