A team led by University of Wisconsin-Madison researcher Josh Coon has performed the first large-scale proteomic experiment using negative electron-transfer dissociation mass spectrometry.
The effort, which was detailed in a paper published last month in the online edition of Analytical Chemistry, demonstrates the feasibility of using NETD on a proteome-wide scale, and provides a new approach for mass spec analysis of proteins and post-translational modifications like phosphorylation and glycosylation, Coon told ProteoMonitor.
LC-MS/MS-based proteomics typically uses acidic chromatography conditions followed by positive electrospray ionization – a workflow well suited to ionizing highly basic proteins and peptides. Acidic proteins and peptides, on the other hand, ionize poorly under such conditions, meaning they are often missed in a conventional LC-MS/MS workflow.
To correct for this, ESI can be done in the negative mode to generate peptide anions. However, collision-activated dissociation of these anions produces spectra that are much more difficult to interpret than those generated by the more conventional application of CAD to peptide cations. On the other hand, ETD – which uses free-radical-driven chemistry to fragment peptide ions for mass spec analysis – results in much more informative peptide anion spectra.
ETD has grown in popularity in recent years due to its ability to retain many post-translational modifications and provide even fragmentation along the entire length of large peptides, and today most mass spec vendors offer instruments capable of this form of fragmentation (PM 2/25/2011). Prior to the UW-Madison effort, however, it had not been used for high-throughput analysis of peptide anions, Coon said.
"There have been a few publications on using NETD, but they were just taking a standard, putting it in [the mass spec] by infusion, and showing that you could [fragment it]," he said. "This, I would say, is the largest number of peptides from a complex mixture that anyone has been able to achieve via analysis in the negative mode."
In the Analytical Chemistry study, the researchers analyzed 2,004 unique peptides in Saccharomyces cerevisiae via NETD on a Thermo Scientific Orbitrap Velos instrument. Using standard ETD and CAD they identified, respectively, 4,403 and 6,693 unique peptides.
That result, Coon said, demonstrates that, though still in need of refinement, NETD could be a viable proteomic technology.
"Before this, you couldn't do it at all, so we thought being able to do shotgun proteomics [using NETD] at even a third of the capacity you can get with state-of-the-art positive mode [workflows] was a huge step," he said.
The technique, Coon added, could prove particularly useful for studying acidic post-translational modifications — especially histidine phosphorylation, which is highly labile under the acidic chromatography used in standard LC-MS/MS workflows.
"Histidine phosphorylation is very hard to detect in the positive mode because if you go under acidic conditions it's hydrolyzed and lost so you don't see it," he said. "If you keep pH high you can retain it, and so with our negative mode analysis you should be able to detect it."
Since completing the initial NETD work, the researchers have begun studying histidine phosphorylation and are currently working on a separate paper detailing these findings, Coon said.
"It's a nice example of an area where there's really no information currently because the main tool used to get at PTMs – mass spec in the positive mode – loses them, so there's not much there," he said.
The technique could also prove handy for studying phosphorylation and glycosylation more generally, Coon said. Both of these modifications are of keen interest to proteomics researchers with protein kinases becoming prominent drug development targets of late and glycoproteins comprising roughly 80 percent of the protein biomarkers currently in clinical use.
"Some glycans would be much more readily ionized to anions than cations, so [glycoproteomics] is a potential application," he said. "And if we really get going well I would think phosphorylation in general [would be a good use], because phosphopeptides should deprotonate much better than they protonate."
Modification of the Orbitrap Velos to run NETD experiments required adjusting the chemical ionization source conditions; optimizing instrument voltages for transmitting reagent cations; and changing the instrument's firmware. However, adjusting the mass spec instrument itself was "the trivial part," Coon said, noting that "any system that can do ETD should be able to do NETD," as well.
More complicated, he said, are the upfront separation and back-end informatics steps involved. Because mass spec-based proteomics is typically done using low pH separations, the Coon team had to build its own high-pH separations workflow for the NETD analysis. Much of this work, Coon noted, was done by Jason Russell, a graduate research assistant in his lab and an author on the Analytical Chemistry paper.
Construction of the separations system ran into a variety of technical challenges, including chromatographic reproducibility, spray stability, and general robustness, but, Coon said, "the recipe is in the paper" for researchers who would like to do NETD work, and, were demand for the technique to grow, commercial vendors could likely produce such systems "without too much problem."
For the informatics, Coon recruited National Institutes of Health researcher Lewis Geer, developer of the Open Mass Spectrometry Search Algorithm and also an author on the paper, to build an algorithm that could identify peptides using anion fragments.
"We made [the algorithm] work well enough that we could get thousands of peptide IDs," Coon said. "But there are a lot of subtleties about how to do that search the best way, and I would say we certainly haven't explored those."
"I would say there are a number of areas where we could certainly make [the workflow] a little bit better, but if we can get on our first real experiment a third of the positive IDs you can get in [standard] shotgun proteomics, that's pretty good," he said. "How much further we can go I'm not sure, but it gives us hope that we can go after some of these things that right now we don't see, [like histidine phosphorylation], and start sequencing them."
John Syka, a Thermo Fisher scientist and author on the paper, told ProteoMonitor that incorporating the technology into the company's machines would primarily be a matter of modifying the instruments' firmware.
Whether the company chose to incorporate it, he noted, would most likely depend on researcher demand.
"If people read this paper and decide that they want [to do NETD] and they talk to their [Thermo Fisher] representative and ask for it, that tends to get more resources allocated to get that sort of thing done," he said. "It's a very new technique in the sense of being directed at any semblance of an analytical application. So we'll just have to see."
Have topics you'd like to see covered in ProteoMonitor? Contact the editor at abonislawski [at] genomeweb [.] com.