SCIENCE
H/D exchange on ice cinnamic acid, acetonitrile, ethanol, and 0.1% TFA The "standard recipe" resulted Amide hydrogen/deuterium (H/D) exin a matrix with a pH slightly higher than change rates provide information about the desired pH 2.5. They adjusted the pH protein structure. Electrospray ionization with extra TFA, which did not adversely MS has been used successfully for such studies, but it can require a separation step affect ionization. The adjustment puts the pH of the matrix at the minimum for amide first. Moreover, ,he ESIIMS experiments hydrogen exchange. "The only thing you require parts of the instrument to be need to control is the temperature," says chilled, which requires modifications of Komives. "As long as you keep the tempercommercially available instruments. ature at [0 °C], you don't have to change There hasn't been much luck with [the components of] the matrix at all comMALDI MS studies of H/D exchange— until now. In this issue of Analytical Chem- pared to what people normally use." istry (p. 3987)) Elizabeth A. Komives snd One of the benefits of using MALDI for Jeffrey G. Mandell of the University of Cali- studies of the kinetics of amide H/D exfornia, San Diego, together with Arnold M. change is that a single mass spectrum proFalick of PerSeptive Biosystems describe a vides much of the needed data."If you want MALDI MS method for studying H/D exfive ttme pointss you analyze five samples," change. According to Komives, they wanted Komives says. "We can get a whole kinetic to develop a technique mat "anybody could run done easily in half a day." The time have access to and do very easily" and to points for the kinetic data are defined by eliminate as many "up front pieces of the the time at which the exchange reaction is puzzle as we could to get the sample into the quenched. In their analysis of the peptic machine as possible." fragments of cyclic AMP-dependent protein kinase (PKA), Komives and her co-workers According to Komives, others have were able to determine H/D exchange tried to measure amide H/D exchange in rates for 40 peptides, which covered 65% of proteins by MALDI MS, but to no avail. The key to her group's success turns out to the protein sequence. be simple. "All we did," Komives says with Komives believes this technique can be a laugh, "was put the MALDI target plate applied in any area H/D exchange would on a little ice pack." She adds, "We put the generally be used. "There's a lot of interest sample on a plate that's been chilled. Then in protein folding. People are starting to we take that plate and dry it more quickly [study] protein/ligand interactions and prothan usual by putting it in a vacuum desictein/protein interactions. Those are all areas cator for [one to two] minutes." Celia Henry we're exploring right now." Other than those minor changes, the experimental procedure resembles any other MALDI experiment and requires no changes to the instrumental set-up. Amide protons were allowed to exchange with D20 at pD 7.25; then the exchange rates were slowed considerably by quenching at pH 2.5,0 °C. The quenched samples were digested with pepsin prior to the MALDI analysis. Komives says that without chilling and Image summarizing the surface deuteration of PKA. The vacuum desiccation all the number of deuterons exchanged onto peptides was deuteriums back-exchange with the matrix before the pro- calculated by subtracting the centroid of the undeuterated tein ever makes it to the mass peptide mass envelope from the mass envelope for incubated peptides. Peptides with more than three spectrometer deuterons are colored red; peptides with 1.9-2.6 The researchers used a deuterons pink; peptides with 1.4-1.6 deuterons purple; and peptides with less than 1.1 deuterons blue. matrix of ce-cyano-4-hydroxy-
Transforming TOFMS, the Hadamard way Time-of-flight (TOY) MS ii intrinsically suited to pulsed methods of ionization. Unfortunately, many popular ionization techniques, such as electrospray, produce a continuous stream of ions. Pulsed orthogonal extraction provides a way to interface electrospray with TOFMS but analyzes a relatively small number of the ions generated by the continuous ionization source. To stop throwing away valuable ions generated by continuous ion sources, Angsar Brock, Nestor Rodriguez, and Richard N. Zare of Stanford University take a multiplexing technique that has already found a foothold in optical spectroscopy— the Hadamard transform (HT)—and apply it to MS. In doing so, the fraction of ions analyzed increases from 0.01% to 50%. They described their work in the September 15 issue of Analytical Chemistry (p. 3735)3 Just like its cousin, the Fourier transform, the HT involves modulating the signal. In the HT, the modulation is based on square waves, and the signal can be in one of only two states—on or off. If the sequence of on and off pulses is known, recovering the original signal by HT is simply a matter of addition and subtraction because the matrix consists of all l's and 0's. The fast HT is as much as lOx faster than a comparable fast Fourier transform, which requires multiplications. To modulate the ion beam, Zare and hii co-workers place a grid consisting of interspersed wire sets between the ion source and the TOF chamber. The ion source is positioned head-on to the flight tube, in contrast to orthogonal extraction TOF, in which the source and flight tube are at right angles. When the wire sets are held at the same potential, the ion beam passes undeflected. When opposing potentials are applied to the wires, the ion beam is split into two beams that miss the entrance slit to the TOF chamber, effectively turning the beam "off'. Because the on/off pulses are shorter than the drift time of the ion packets (nanoseconds versus microseconds), the faster ions from later pulses overtake the slower ions from earlier pulses. The resulting signal at the detector is the overlap of many TOF distributions, which look like noise. The TOF distributions can be recovered by
Analytical Chemistry News & Features, October 1, 1998 6 3 9 A
News
The gain expected from the HT—part of which is attributable to the collinear arrangement and part of which is generated by the multiplexing—is in the range of a factor of 10100. At the ACS meeting in Boston in August, Zare reported experimental Hadamard transform TOFMS experimental setup. evidence that the gain was just as theory had predicted applying a fast HT. "All the information is there, provided you know the sequence "The thing that makes [HT-TOFMS] potentially so appealing is that [the setup] with which you chopped [the ion beam]," is no different from [conventional] time of says Zare.
flight," says Zare. "All you need is a grid and a computer. The grid can potentially be made very cheaply, and you need a computer anyway." "I'm really taken with [the Hadamard transform]," Zare says. "Once you understand it, it's so amazingly simple. My hope is that lots of people will use this in the future." He hopes to interest manufacturers of TOF mass spectrometers in the technique. He imagines being able to multiplex sources through a single grid and, thus, run many samples simultaneously. "In TOF mass spectrometry, you pay for the vacuum system. Why use this big vacuum system for just one beam? Why not use 10 beams or 100 beams and get the same information out? You'd get 10 or 100 ttmes [higher] throughput." Celia Henry
NEWS FROM THE ACS NATIONAL MEETING Alan Newman reports from Boston.
Fast, hot, and on the move
ture changes arise from the system's low thermal mass. "You don't want to heat an oven, just the column," he reports. Two Over the past few years, fast GC has grown "off-the-shelf', 100-um i.d., 1-m long GC in sophistication, and now Ed Overton and columns are wound around each other like his research group at Louisana State Unia double helix. The combination—along versity have moved the technique one step with a heater and sensor wires—is loaded further by creating a small yet versatile into a fiberglass sheath measuring less instrument. Informally dubbed the "shoethan 1-mm in diameter. This arrangement box GC", Overton's system weighs just 15 isolates the heating and provides the rapid lbs, measures 8 x 11 x 14 in., includee temperature changes without cold spots three electronic pressure controllers, and that could broaden peaks. carries two GC columns and detectors in Samples are fed into a conventional parallel. It can be configured as a portable heated inlet and then trapped onto a solid GC. The key to the system, says Overton, sorbent such as Tenax. Rapid thermal desis obtaining precise and reproducible fast orption injects the analytes onto the two temperature programming at rates as high columns. According to Overton, with programmable temperature changes of 5 °C/ 20 °C/s. According to Overton, the fast tempera- s, volatile hydrocarbons up to C10 need less than 20 s to elute, and semivolatile compounds up to C2o take less than 1 min. Coelutions are no problem because of the dual-column design; peaks that do not separate o n oTif* column C3.n be identified from the chromatogram of the second. Or, the two chromatograms can be used ss fingerprints for compeex materials. Cycle times are under 5 min., and Overton has run up to 80 samples, with The shoebox GC next to a standard Hewlett-Packard blanks, in one day. Moreinstrument. 640 A
Analytical Chemistry News & Features, October 1, 1998
over, the system handles various techniques such as purge and trap, solid-phase microextraction, and pyrolysis. The instrument design also allows back-flow cleaning of the entire analytical train, and Overton reports that as a result columns have long lifetimes. The current design accommodates temperatures up to 250° C, but Overton is looking at increasing that value to 350° C. Hydrogen gas consumption with the dual flame ionization detectors is typically around 30 mL/min, says Overton, which should allow days of field work with a typical gas cylinder. The instrument, now called microFast GC2, has just been commercialized under the auspices of start-up company Chromalytics.
Problem-based learning A recent report arising from NSF-sponsored workshops advocates that analytical chemists introduce problem-based learning into their undergraduate courses (Anal. Chem. 1998, 70,176 A-77 A)) Although the concept has been discussed in education circles for years, to most analytical chemists this is a new idea. At the ACS meettng several practitioners of problem-based instruction shared their experiences. Several more examples will appear as feature articles in Analytical Chemistry.
