A team led by researchers at the University of Washington has developed a multiplexed approach to data-independent acquisition mass spec that could offer significant gains in precursor selectivity.
Described in a paper published this week in Nature Methods, the technique uses the simultaneous measurement of multiple narrow m/z isolation windows to reduce precursor interference issues associated with DIA analysis while retaining the method's breadth and speed.
According to Michael MacCoss, a UW researcher and author on the paper, currently the technique is only possible using Thermo Fisher Scientific's Q Exactive instrument. Several Thermo Fisher researchers were co-authors on the study.
Traditionally, shotgun proteomics experiments have used data-dependent acquisition wherein the mass spectrometer performs an initial scan of precursor ions entering the instrument and selects a sampling of those ions for fragmentation and generation of MS/MS spectra. Because instruments can't scan quickly enough to acquire all the precursors entering at a given moment, however, many ions – particularly low-abundance ions – are never selected for MS/MS fragmentation and so are not detected.
In DIA, on the other hand, the mass spec selects broad m/z windows and fragments all precursors in that window, allowing the machine to collect MS/MS spectra on all ions in a sample.
DIA was initially proposed almost a decade ago by Scripps Institute researcher John Yates III. Waters was the first vendor to introduce a DIA product, launching in 2006 its MSE method, an approach based on the company's ion mobility technology. And, last year, AB Sciex introduced its Swath DIA method – developed in collaboration with Swiss Federal Institute of Technology Zurich researcher Ruedi Aebersold – for use on its TripleTOF 5600+ machine (PM 6/8/2012). Via its open source proteomics software package Skyline, MacCoss' lab offers support for DIA methods on instruments from a variety of vendors including Agilent, Bruker, and Thermo Fisher.
As mass spec hardware and software has in recent years improved sufficiently to take advantage of DIA analysis, the technique has seen a surge in popularity. In particular, the method has proven useful for combined qualitative/quantitative analysis of samples, allowing researchers to do shotgun-style discovery and then interrogate the same data set for targeted quantitative information on analytes of interest.
One issue with DIA analysis, however, is that the wide m/z fragmentation windows typically used in the approach create significant opportunity for precursor interference, resulting in what MacCoss and his co-authors characterized as a "typical five- to 10-fold reduction in precursor selectivity compared to data obtained with data-dependent acquisition or selected reaction monitoring."
"If you consider something like [AB Sciex's] Swath method, which has a 25 m/z-wide window, there are a lot of [precursors] in that isolation window," MacCoss told ProteoMonitor. And this abundance of precursors, he noted, leads to a loss of selectivity.
An obvious solution to this selectivity problem is narrowing the fragmentation window used. This, however, would reduce the m/z range researchers could measure with the technique. As the study authors noted, at current mass spec acquisition speeds of 10 Hz, a 20 m/z-wide window is required to sample a 400 m/z range every two seconds. Narrowing the fragmentation window would result in a proportional narrowing of the overall m/z range captured by the analysis.
To get around these limitations, the UW team turned to a multiplexed approach, simultaneously analyzing five different 4 m/z windows per spectra. This allowed them to cover in one spectrum the equivalent of a 20 m/z-wide window but with the improved selectivity inherent in the smaller 4 m/z windows. Using the technique, they achieved a five-fold increase in precursor selectivity.
The method was enabled, MacCoss said, by the configuration of the Q Exactive, which, he noted, has a quadrupole for precursor selection, a collision cell for fragmentation, and a C trap where the fragments of one precursor window can be stored while the next is selected and fragmented.
"You need to be able to isolate a mass, fragment that mass, and then store the fragments, and then isolate another mass, fragment it, and store those fragments without losing any of the existing fragments," he said, adding that while "you could imagine other" potential mass spec configurations capable of the technique, the Q Exactive is currently the only instrument with the necessary combination of components.
MacCoss noted, as well, that the technique was best suited to instruments that, like Orbitraps, offer fast isolation and fragmentation relative to the speed of mass analysis. In the case of the Q Exactive, "you can do the isolation and fills in about 20 to 50 millisecond time frames," while the mass analysis takes on the order of "60 to 100 milliseconds," he said. "So there is an advantage where you can collect and store multiple precursors and measure those multiple precursors in one spectrum without losing time."
By the same token, MacCoss said, implementing the technique on a time-of-flight machine wouldn't make sense because "a time-of-flight is so quick that you gain nothing by measuring multiple isolation windows in parallel, because you could measure them all individually faster than you could isolate and fragment them."
The researchers tested the technique by using it for quantitation of a six-bovine protein digest spiked into yeast lysate. They quantified five of the proteins based on measurements of 36 peptides, with an average lower limit of detection for those peptides of 8.66 femtomoles.
MacCoss and his colleagues also used the technique for distinguishing between forms of a peptide with and without an oxidative modification – something, he said, that would be impossible with DIA methods that rely on wider isolation windows.
Because many of the fragment ions of these peptides are identical, the only way to distinguish between them is by using their precursor masses, MacCoss noted. However, being doubly charged, the two forms of the peptide are 8 m/z apart, meaning that even were the researchers to use a 10 m/z window they would still be unable to distinguish between the two forms.
Using the multiplexing technique, on the other hand, the researchers were able to cover a 20 m/z range while also achieving precursor selectivity down to a 4 m/z window, well below the 8 m/z separating the peptide forms.
"So it does clean up the data a fair amount," MacCoss said.
While the UW team settled on multiplexing groups of five 4 m/z windows, a variety of other combinations could also work, MacCoss said, noting that the optimal window size would depend in large part on the particular experiment.
"There is a balance between transmission and selectivity," he noted. "The wider the window, the more signal you are going to get, and the narrower the window, the better the selectivity. So it will depend on the sample and on the analyte you are measuring. If you have an issue where you don't have enough ions to see something, then you are going to want to open the windows wider. If you have a situation where you can see lots of ions but you have lots of interference, then you'll want to narrow them."