Developing FTICR Mass Spectrometry Instruments with Unique Capabilities

Peter B. O'Connor, Professor of Analytical Chemistry, Warwick Centre for Analytical Science (WCAS), Department of Chemistry, University of Warwick, UK

Peter and his team work collaboratively with other research groups to demonstrate the effectiveness of higher specification fourier-transform ion cyclotron resonance (FTICR) mass spectrometry in specific applications. Here he discusses the diverse range of projects he is currently working on. He touches on the adaptability of mass spectrometry, enabling researchers to study cancer, neurological disease, petroleum, and biopharmaceuticals. 

Q: How did you come to work in this area of research?

A: Once I had completed my undergraduate degree I went on to do my PhD, although I originally planned to focus on computational modelling, once I got to university that didn’t really appeal to me. And so I wandered around attempting to find some other areas of interest, and eventually stumbled into the lab of Fred McLafferty (the American chemist known for his work in mass spectrometry) who was doing very similar research to what we are still doing now –although in a much more preliminary state at that point – I got to know about mass spectrometry, really fell in love with it and stuck with it all these years. It initially piqued my interest in the early ‘90s so roughly 25 years ago.

Q: Could you touch on your work within the area of cancer research?

A: I wouldn’t consider myself an expert in cancer research but my colleague, Peter Sadler, a co-author on one of our latest papers works on anti-cancer metallodrugs. The most famous of them is cisplatin which is probably the most prescribed anti-cancer drug of all.

Cisplatin acts as a sort of template for most of what Peter works on, except that he changes the metal and the ligands on the drug – this, of course, changes the chemistry of the molecule completely. The goal is to find anti-cancer drugs which are more efficacious, that have enhanced targetability, or, in this case, the team wanted to find an anti-cancer drug that was photo-activatable so that they could dose a person with an inert drug at fairly high concentrations and then just activate the molecule with blue light on specific target sites where the cancer is specifically located.

This strategy would mean that it would be possible to limit the damage to healthy cells through controlled positioning of the light, which is an interesting concept that Peter has brought forward. My personal interest was directed at the photo-activatable anti-cancer drugs, in figuring out what these drugs were doing when they became activated. What happens when you put this type of drug into a biological system, how do they actually kill cells, and what are they actually doing? It is a really interesting problem because it's a mass spectrometry problem. It is also a proteomics problem because most of the drug’s interactions are with proteins in the cells.

We started out with cisplatin and found ways to detect the modification sites of proteins in cisplatin, now we've extended this to quite a few other metallodrugs. We then performed liquid chromatography combined with mass spectrometry and tandem mass spectrometry – so fragmentation mass spectrometry. We used this approach to try to identify the proteins and the sites of modification. If you do this with a conventional proteomics method, you'll be able to find something but everything that's modified with a drug has a very strange isotope pattern because metals have very strange isotope patterns and there's a metal at the middle of the drug. So we had to find a way to automatically highlight all the peaks that had the metals on them. We were able to build a script that could examine the isotope patterns and identify only the peaks that would have the metals on them and we could then focus on those and perform more extensive mass spectrometry. It is a proteomics challenge, so you have a terrible mish mash of proteins and peptides in there, as usual. These are also human cell lines so they're quite complicated – they're not as simple as a yeast cell for example. Then we have the metals which make such a broad isotope distribution that it's rather challenging to find them and that tends to mess up a lot of the proteomics methodologies that are pretty well established now. So, while we had to develop the tools to do that, the high resolution of FTICR MS was very helpful, although probably not strictly necessary in all cases. Inevitably in biological systems you end up with overlapping isotope distributions and then that very high resolution becomes critical. So some of this could have been done with other proteomic set ups, but the FTICR is clearly going to be the best approach.

Q: What are the benefits of FTICR mass spectrometry and what applications is it particularly suited for?

A: The biggest advantage of FTICR by far, is that it has much higher resolution than any other type of mass spectrometer. It is about 10 times better than the nearest competitor which is the orbitrap. High resolution is very useful, particularly whenever you have particularly complex samples. One of the biggest areas that researchers work on in this area, although I don't work on this myself, is petroleum. Petroleum is a horrendously complex mixture, hundreds of thousands of components, all low molecular weight and so they tend to give you just huge signal and you might have hundreds of individual peaks at every one dalton spacing in the spectrum, so it gets really complicated. Another area that we're working on quite a bit is synthetic polymers, particularly the block copolymers where combinatorial confusion in the spectrum is common because you have distributions of polymers that you tag together in a block copolymer and so they get rather heterogeneous.  A more advanced mass spectrometer is necessary to make sense of the spectra.

Both petroleum and polymers can be either not very amenable to chromatographic separation or they can be too complicated for chromatographic separation so you need to be able to handle many different components at the same time in the spectrum, so you absolutely need the resolution. Proteomics is very similar in that sense. In this case, like in many cases, we start by digesting the proteins in the cells down to peptides and work from there. It's what we call a bottom-up approach to proteomics but that means that you're turning maybe tens of thousands of proteins into millions of peptides or at least hundreds of thousands of peptides. Then you have to separate out that mixture and process it. Now peptides are very amenable to chromatography, so they usually do separate out pretty well. So you don't have terribly complex samples but if you have a pretty good chromatography set up you might be able to separate out your mixture into say 1,000 or 2,000 fractions and if you've got 100,000 or a million infractions that means that you've got quite a few components coming out at any particular time and you still need to have high resolution to be able to deal with the complexity of the sample. For proteomics, the approach is to try to run the experiment and take advantage of the high performance of the mass spectrometer for selection and tandem mass spec. They have to run the mass spectrometer really, really fast and try to select as many peaks as possible in every time slice in order to fragment them and get sequence information. This approach does work, and we do that too, but we have the added advantage of having very high resolution meaning we can separate out the mixtures at the same time. Which helps us process this data. Then there's all the other elements such as glycomics and genomics that are also amenable to this high-performance mass spectrometry in different contexts.

Q: Your team's mission is to "develop new FTICR mass spectrometry instruments with unique capabilities."  Could you elaborate on this?

A: Development is most of what we do. We work on developing advanced FTICR mass spectrometry, more specifically, we develop new tools using the high-resolution capabilities of the FTICR.

I'm currently working on building a web app that I intend to use to teach people how to calculate isotope distribution. Isotope distributions are not actually very complicated, but it really is important to get them right if you're going to be comparing them to real data. A lot of web applications that you find make some pretty simple assumptions that are often fine – but can sometimes cause problems. We have to calculate isotopic distributions for all metals which is not always easy. We also work on different ways of fragmenting molecules. In the mass spectrometer you will typically measure your mass spectrum and get a number of peaks for peptides. You'll then select one peak using some of the voltages and wave forms that we have in the mass spectrometer, smash it using a tandem mass spectrometry fragmentation method and then measure the masses that you've created – which will presumably allow you to get some structural information for peptides. This method usually enables the partial (or sometimes full) sequencing of the peptides. So there's a lot of tricks to sequencing using tandem mass spectrometry.

How do you break the molecules? Most of the time, we collide them with background gas and break them solely by collisions. We are also working on doing photo dissociation which involves breaking molecules with UV and IR light, and we also do reactions with electrons. There are half a dozen different types of reactions you can perform with electrons that causes the molecules to break in different ways. Each reaction tends to result in slightly different information, so when you combine several of these reactions together you can get a very complete picture of sequence information which is very useful. We’ve actually developed a new fragmentation technique, called two dimensional mass spectrometry, which is very exciting. We've got a few papers out on it now and we're about to submit three or four more.

Two-dimensional mass spectrometry technology is well established. The concepts and methodologies were worked back in the late 80s, but it required too much computational power for the time, we needed modern computers to be able to start to successfully process the data. When it comes to two-dimensional mass spectrometry, you may have one spectrum that equates to 200 Gb of data – your average computer can't process that, it requires cluster computers, housed here in the university. This enhanced computing power allows us to get it down into a format that we can use. The technology finally enabled us to use this method even though it's been sitting there for close to 30 years in the literature! It allows us to fragment the molecules.

In two-dimensional mass spectrometry you don’t have to isolate the molecules. What you do instead is you modulate all of the molecules in and out of a fragmentation zone and you basically code each pre-cursor ion with its own modulation frequency and because that is dragging it back and forth to the fragmentation zone the fragments end up carrying the same modulation frequency as the pre-cursor. This means that you can fragment all the molecules at the same time without doing any isolations and it greatly speeds up the whole process because you don't have to do them one by one, you just do them all together. It speeds it up and it also gives you much more detail in the information because what inevitably happens with complex samples is you can fragment the first 5 or 10 and then you run out of time to be able to do the next bit so it's basically a much faster way of doing the whole experiment. It will end up giving us a lot more information about structures and molecules in complex mixtures. It's an ongoing development project though so stay tuned on that one I guess!

Q: What projects are you currently working on?

A: Two-dimensional mass spectrometry; that's a big development project for us going forward and we're working on it in pretty much every avenue that we can think to apply it. We have recently had two scientific papers accepted that outline the technology and its use for the analysis of whole proteins rather than the analysis of smaller peptides that you would typically create by digesting the proteins –  top-down proteomics. Of course, we're using it on proteomics and polymers and we are also making good progress on the use of two-dimensional mass spectrometry for glycomics. There are a few application areas that are particularly interesting.

The first project to mention is in collaboration with a medical school in Beijing. The researchers have scorpion venom and scorpion venom extracts that they are using that have particular anti-coagulant activities. The idea is to try to figure out what is it in the structures of these scorpion venom extracts that is responsible for the anti-coagulant activity. It turns out that most of the molecules in the extract are proteins, not very big ones, pretty small, under 10,000 daltons, but they're tied up in really tight balls with many disulphide bonds, presumably so that they'll survive when they're out in the environment and being in this conformation prevents them from immediately being chopped up by proteases and allows them to maintain their stability. But, first of all, you need to try to figure out what these molecules are. The scorpions aren't sequenced so we don't have a database to search against so we have to sequence them ‘de novo’, using just the mass spectrometry data - which is much harder. We also must try to figure out exactly where those disulfide bonds are, you tend to just scramble where the disulfide bonds are (if you even keep that information) so it's a challenge to try to figure out exactly what these venoms are and how they work. That's a fun project which is making some nice progress at the moment.

We are also working on analysis of amyloid plaques in Alzheimer’s disease patients, specifically trying to figure out what impact metallic particles have on Alzheimer’s disease. There has been a lot of discussion about aluminium and iron in your diet or that you pick up from environmental pollution and how this might impact Alzheimer’s disease. I have a collaborator who works on that and our part of the project is to try to figure out what modifications we see in the Alzheimer’s disease proteins, the amyloid beta protein, as a function of these metallic nanoparticles. They also then go off to the synchrotrons and try to figure out exactly what the chemical state of the metal is in that particular nanoparticle. Is it actually a metal or is it a metal that is bound to something else? It is an interesting problem from a lot of different directions, technologically as well as in terms of medicine.

We also have a big project on antibodies. This is a collaboration project with a biopharmaceutical company, who manufacture and use a lot of different antibodies. They're extremely interested in all the post-translational modifications that feature on them. If these antibodies are being produced for pharmaceutical purposes they want to ensure that the antibodies remain stable over time and that they aren’t creating some kind of side reaction that then causes problems with the activity of the drug. Therefore, there is a lot of work currently focused on post-translational modifications of antibodies. 

Peter B. O'Connor was speaking to Laura Elizabeth Lansdowne, Science Writer for Technology Networks.