Assistant professor, Departments of Chemistry and Biomolecular Chemistry
University of Wisconsin-Madison
At A Glance
Name: Josh Coon
Position: Assistant professor, Departments of Chemistry and Biomolecular Chemistry, University of Wisconsin-Madison, since 2005.
Background: in Don Hunt's lab, University of Virginia, 2003-2005.
PhD in chemistry, University of Florida, 2002.
Joshua Coon gave a talk on using electron transfer dissociation for protein sequence analysis at this week's Pittcon conference in Orlando, Fla. ProteoMonitor spoke with Coon before his talk to find out more about ETD technology, and about Coon's work.
How did you get into proteomics and using electron transfer dissociation to study proteins?
I got started in proteomics when I was in graduate school at the University of Florida. I was doing laser desorption ionization, and in the last few years I was there got excited about proteomics work. This was around 2000, and there was beginning to be a large [amount] of interest in the field. That was when I began my interest in protein work. I was doing instrumental mass spectrometry, but then started getting interested in methods for developing proteomics tools.
In graduate school I did some laser desorption at atmosphere with subsequent chemical ionization for looking at peptides out of gels. That was interesting, but after I finished my graduate work, I decided that if I wanted to do more work with instrumentation in the proteomics field, I would need a postdoc.
The postdoc that I chose was at Don Hunt's lab in the University of Virginia. Don had a good reputation for being a leader in protein sequencing and also in instrumentation, so I thought it was a good fit.
When I went to Don's lab in 2002, he had a variety of applied work and instrumental work, and I was drawn to the instrumental work. In Don's lab I met John Syka and Jarrod Marto. Don and John and Jarrod had talked about electron capture dissociation — Fred McLafferty's group had developed that, and it had been around for a while.
We were interested in analyzing modifications, primarily phosphorylation. But we were basically using ion trap methods, and the electron capture method wasn't applicable to those instruments.
That group had discussed the possibility of using the reactions of anions with peptide cations in an ion trap to sort of get a similar type of fragmentation, and then when I came along in 2002, it was a project that was described. I said, 'That sounds pretty good,' and so John Syka and I spent about a year to build a linear ion trap mass spectrometer to do ion-ion reactions.
Up to that time, the ion-ion reaction work had been pioneered by Scott McLuckey. He had been using three-dimensional ion traps, and we thought that the technology of the future was linear ion trap, and that we could probably do the work best in a linear ion trap. It took us about a year to adapt a Thermo Finnigan LTQ instrument to do ion-ion reactions so that we could try to figure out if we could find anions that would do electron transfer dissociation.
That took us until about 2003, and then when we first tried out our instrument, we were fortunate to find that if you used the right type of anion, and you reacted it with a peptide cation, you could get an electron to transfer to the peptide. And it turns out that if an electron transfers to the peptide, you get a dissociation that is pretty much identical to what you get in electron capture dissociation, except that you can do this in an ion trap mass spectrometer, which is much less expensive. And we found that we could do it pretty quickly, so we could do those reactions on a 20 to 60 millisecond time scale, which is completely compatible with chromatography.
Can you explain the difference between electron capture and electron transfer dissociation?
[Fred] McLafferty's group [at Cornell University] back in the late 90s [was] using ion cyclotron resonance instruments, or FTICR. Those instruments use a magnetic field to trap peptide ions, and to measure their mass. In that trapping device you can contain low-energy electrons in an overlapping space with the peptide cations that you wish to dissociate. If you do that, a peptide cation — some of them — will react with these low-energy electrons, and they will fall apart in a very specific way.
The problem is that on an ion trap mass spectrometer, or any mass spectrometer that uses radio frequency fields to contain peptides, rather than magnetic fields, the electron won't be confined in that space. You can't keep a low-energy electron around in those type of devices.
In electron transfer, dissociation basically works in that an ion trap device that uses a radio frequency field can contain cations of peptides and anions in the same space at the same time. And it turns out that if you pick the negative ion correctly — if you choose the right composition, the right type of anion — you can put them in the same space with the peptide cation in the ion trap and you can get the electron to jump off the anion onto the peptide.
So you use this anion as a way to deliver an electron. The idea is if we can't put low-energy electrons in our device, we can just use an anion of a small molecule, and use that as a vehicle to deliver an electron to a peptide. And once you've delivered an electron, the process works the same as for electron capture dissociation for the most part.
So the main reason for doing the electron transfer as opposed to the electron capture is so that you can work with an ion trap mass spec?
Yes. So you can't do ECD very well on any type of instrument besides an ICR. So to get that sort of fragmentation on ion traps or any device that uses radio frequency fields, you can use electron transfer, and basically it's a way to do that type of dissociation on the most common mass spectrometer.
It turns out it works surprisingly well, and that's good for us. We can do the technique with chromatographic separations, with data-dependent analysis, just like you would use collisional activation in a typical proteomic analysis.
What did you study after you developed the electron transfer technique?
We spent 18 months doing that, and the first thing we wanted to study before I left my postdoc was post-translational modifications. Phosphorylation was the biggest driver.
It turns out if you use collisional activation to fragment phosphorylated peptides, they don't always fall apart very well, and more often than not you can't get enough backbone breakages of the peptide to determine the sequence.
With ETD, it turns out you get a very rich sequence. You get a very good fragmentation, regardless of the presence of a modification. We think that ETD for the area of post-translational modification is going to be a benefit. It's going to allow for you to see things that otherwise are challenging to see on bench-top instruments.
Now in my new position here at the University of Wisconsin, my research has moved towards fragmenting whole proteins. It turns out that with ETD, you can take very large peptides, or even whole proteins, and if you do this reaction to them, they fall apart very randomly across the backbone, without regard to size, the amino acid sequence, or the presence of modifications.
One of the areas we're very excited about is that you don't have to use trypsin to make small peptides any more. One of the major reasons people like to use trypsin in their mass spectrometry is you get a small peptide that doesn't contain any internal basic residues, so it falls apart pretty well with collisions. But it turns out with these types of electron-based dissociation methods like ETD or ECD, you can break apart bigger things pretty effectively. So now you have this whole area of large molecule mass spec that you can start to approach with ion trap instrumentation.
We're very interested in looking at large peptides, say 3 to 10 kilodaltons in size. Electron transfer dissociation will make them fall apart very effectively, but it tends to leave a very complicated spectrum. As things get bigger, you have more fragments, the number of charges are higher. So one of the things we're very excited about is using ETD in series with other ion-ion reactions.
For ion trap work — Scott McLuckey pioneered this whole field — if you choose a different type of anion, instead of delivering an electron, you can move a proton. So one of the things we're pushing here at Wisconsin is that if you use ETD to dissociate a large peptide, you'll generate peptides that are in lots of different charge states. It's a very complicated spectra. But then if you do a second reaction with a proton transfer-type anion, you can remove charge from those fragments, and then you can simplify your spectra so you can basically read off the N and C terminus of a whole protein.
That's an area that we're very interested in. Sequential ion-ion reactions is how I describe it. You can have reactions that do different functions and process ions in different ways and by doing that you can start to pretty effectively attack large peptides and whole proteins pretty quickly.
We're interested in moving towards looking at bigger things. I think the reasons you would want to do that are some of the ones Neil Kelleher has been talking about for the last five years. That is, you can see alternative splicing, you can see patterns of modifications across, for example, a whole N-terminal sequence of something. So that's important, and also if you don't chop the protein mixture into as many pieces, you don't have to sort out as many later.
That's where we're headed, and I think there are other ion-ion reactions that are not developed yet that could be useful. One of our main motivations is to develop this ion-ion chemistry to be automated and to routinely handle large molecules effectively.
Do you have any plans to commercialize the technology that you're developing?
Well it turns out that at present, to my knowledge, there's no commercial system that you can buy today that will do ETD or any other ion-ion reactions. But I understand that several instrument manufacturers are currently working on commercial versions of instrumentation that can do electron transfer dissociation. What other reactions those instruments will accommodate is not clear.
I think the early applications of ETD, and what people are excited about right now, are the ability to randomly to cleave peptides of typical size and get more information per peptide, and post-translational modification analysis is a really important application that people recognize. I think that as more people start to get these instruments and start to do the chemistries, this large molecule trend will develop later.
I think right now the most interest in ETD is generated because you can see modifications pretty well. Certainly there [are] instrument manufacturers who are going to, I would guess, within the next year start selling [ETD].
My group is collaborating with Thermo Electron on some of the instrumental aspects.
Aside from working on development of techniques, are you also doing applied work?
Yes. About a third of our work is on instrumentation, about a third is applied work — phosphorylation site quantitation and different cell states, identifying splice variants of proteins that occur in cancer cell lines, and also Wisconsin has a really good strength in stem cell research, so we're getting involved in some cell signaling and phosphorylation work in stem cell research. And the other third of our interest is in bioinformatics, and how do we handle this data.
As we talked about area, you get more information, on average, from ETD. We're very excited about the new types of bioinformatics that will go with this technology. So if you get more information per spectrum, then your ability to do things like generate different search modes, and maybe even developing de novo sequencing algorithms, becomes much more practical because there's more information that you can get out of the spectra. That's an interest that we have, and we're starting to figure out how we can take advantage of this unique type of data that we generate.
I expect that other labs will also develop bioinformatics for ETD, and other people will start generating new types of search modes and algorithms to process the data.
What are your main goals for the future?
There [are] several main goals. One is to really characterize ETD and to get the most out of it that we can. That means sort-of fundamental issues of making this reaction the most efficient and fastest that it can be, pushing the envelope in terms of trying to analyze bigger things more effectively, sequencing ion-ion reactions, developing new bioinformatics platforms like de novo sequencing capabilities.
And then the last area that we would like to push that we haven't talked about yet is that all this ion-ion chemistry is being performed, at least in our lab, most effectively on ion trap instrumentation. And linear ion traps are common in hybrid mass spectrometers, like LTQ-FT and LTQ-Orbitrap. So it's feasible that you could adapt this type of technology to instrumentation that has high mass accuracy and high mass resolution. So if you can couple all these processing abilities that ion-ion chemistry brings, ETD included, with high mass resolution and high mass accuracy instrumentation, you're much better off because you can handle bigger things more effectively. So I think that's another area where we would like to drive in the next several years — sort of moving this platform into hybrid instrumentation where you get more information per analysis.