At A Glance
Name: Fred McLafferty
Position: Professor Emeritus, Cornell University, since 1992
Prior Experience: Professor of Chemistry, Cornell University, 1968-92
Professor of Chemistry, Purdue University, 1964-68
Dow Chemical, 1950-64 (from 1956: Director, Eastern Research Laboratory for basic research, Framingham, Mass.)
Postdoctoral fellow, University of Iowa, 1949-50
PhD in Organic Chemistry, Cornell University, 1950
BS in Chemistry and Mathematics, University of Nebraska, 1943
Received the 2003 American Society for Mass Spectrometry Award for a Distinguished Contribution in Mass Spectrometry this week.
Most of your earlier work was focused on the analysis of small organic molecules by mass spectrometry. How did you become interested in studying proteins?
Proteins and peptides are linear molecules, [which] are perfect for mass spectrometry. It’s just like having a string of pocket beads of different colors, or different masses. If you can break [the bonds] between every one of the beads and get the masses of each of those fragments, then you know the whole thing. There are 20 amino acids, and except for leucine and isoleucine they have different masses. So when we put a protein into the gas phase, and we break it up, then from those masses we can get the sequence. I wanted to apply our techniques to other molecules [besides small organic molecules], and it was pretty obvious that proteins were important things.
I tried to analyze proteins from 1980 [on], but it wasn’t until John Fenn invented electrospray ionization that we were able to put proteins into the mass spectrometer. Our first publication on this was in 1989.
What’s your approach for analyzing proteins by mass spectrometry?
We do top-down proteomics [by Fourier transform mass spectrometry], [so] we start with the whole protein and break it up inside the mass spectrometer. We now have automatic programs that take the information from the mass spectrometer and put it all together into sequences.
Probably the biggest thing about this top-down approach isn’t the identification of the protein you are looking at. The bottom-up approach [can also do this and] is so tremendously automated; there are labs with 50 mass spectrometers doing it 24 hours a day. The place where we really shine is [in] showing where the posttranslational modifications are, because [people] have a terrible time [getting those with the bottom- up approach].
If we run a molecular weight of a protein, and it’s supposed to be 40,000, and we find it’s 40,016, there is something [modified]. When we break it up into two pieces, one of them will have the correct mass, and the other one will be off by 16, so we know that the modification is in that [of the molecule]. Carbonic anhydrase, [for example,] we broke up in 189 places, and all you have to do is look and see which one of these places is off. In that case, we found that the sequence that had been in the protein database for all these years was wrong. By starting at the top, not only can you identify proteins very quickly, but also, when you see that the molecular weight is off, you know that something’s happened to the protein.
How did you develop electron capture dissociation? Is it true that you discovered it by accident?
It was an accident, I must admit We had been looking for a way to break different bonds in the protein, because all of these wonderful ways would still essentially break the weakest bond. We [used an] excimer laser that we were trying to break different bonds with. An excimer laser has [a wavelength of] 193 nanometers and 6.4 electron volts of energy. Bond dissociation is only 3 or 4 electron volts, so we thought ‘maybe this will break a specific bond before the energy gets randomized.’ One day a postdoc found these crazy different peaks. What he had done with this $30,000 laser is [to] not aim it very well. We store our ions inside the magnet in the mass spectrometer, and the laser is supposed to go right on through and just hit the ions that are sitting there in the central part of the magnetic field. He aimed it so it hit the side, hit metal, and knocked out electrons, and it was these very low-energy electrons that then got captured.
Finally Roman Zubarov, who was a new postdoc, put an extra pair of trapping plates on the cell that he could make negative. With his extra pair of trapping plates, he could keep electrons and positive ions in there. It turned out that 10 cents worth of wire [to create the electrons] gives you as much as a $30,000 laser does. That’s how we finally got that method going. Our first publication on that was 1997 or 1998.
What about the top-down approach can still be improved?
One of the problems is to break all of these bonds. The first electron that goes in can break almost any bond, but the products are still there. You keep putting in more electrons to break more of the molecular ions. To get all of these cleavages, you have to have big product ions as well as little ones. One of the problems was to do this fast enough and efficiently enough. This spring we published a plasma method (Sze et al., Analytical Chemistry, Apr 1;75(7):1599-603) that gets around this — not completely, but in one spectrum we had 183 cleavages out of 250 possible ones.
Another thing that’s terribly important, of course, is the data handling. The best research for automating this whole process has been done by Neil Kelleher at the University of Illinois, a former student of mine. He has automated the way you get the samples in there, and he has automated the programs for interpreting the spectra. For example, [for a protein that has] two disulfide bonds, it turns out that a very simple sort of bashing of the molecule that only gives 30 or 40 cleavages is enough to tell these disulfide bonds. His procedure is to break things up partly and have the computer analyze it and see if the problem is solved. He is trying to do the complete proteomics of a small organism and to do it in this automated fashion and evolve the techniques at the same time.
FTMS has a reputation for being difficult. Do you think it will become more mainstream in proteomics?
It’s certainly important that the instrument manufacturers make it user-friendly. Our own mass spectrometer — parts of it we got in 1984, we put on a 6-Tesla magnet in 1990, and most of it is either homemade or more than 10 years old — runs almost all the time. But we do not have any instrument manufacturer help us fix it, the grad students and postdocs do everything.
Where do you see mass spectrometry for protein analysis going in the next few years?
One of the great advantages of mass spectrometry is sensitivity. As soon as you do proteolysis or anything that you have to do outside the mass spectrometer, you use up a lot of sample. We published in Science in 1996 that we were able to do MS/MS on less than an attomole of sample, and we were able to identify carbonic anhydrase on 10 attomoles of sample. To show you what that means, a red blood cell contains a femtomole, or 10-15 moles, of protein. Carbonic anhydrase is a one percent component of a red blood cell, so that’s 10-17 moles. We put in a protein mixture that had 10-17 moles of carbonic anhydrase, and not only got its accurate molecular weight but broke it up into pieces, and those pieces then easily chose carbonic anhydrase out of the protein database. That’s where all the future of FTMS is. The only trouble is, one flake of dandruff is of course one million times or so that amount. Getting that kind of sample into the mass spectrometer is [a problem.] My former postdoc Gary Valaskovic is the guy who did that, and he now has a company called New Objective. We have got the sensitivity, and there are these techniques for getting samples out of single cells, and one of these days, [people] are going to be doing that with this kind of top-down FTMS mass spectrometry. They are going to be identifying things at this attomolar level, and that’s going [to show] how individual cells differ from each other.
Most biologists know you for your FTMS work. But what are you most famous for?
I have five wonderful children [laughs].
Chemists all know me because of the McLafferty rearrangement [the six-membered ring hydrogen rearrangement in the gas-phase ion chemistry of carbonyl compounds]. As soon as they take organic chemistry, they have to learn that [and] I take the blame for that. Lots of other people saw [this rearrangement], too, but luckily my name got stuck on it.