At A Glance
Name: Kenneth Standing
Age: 78
Title: Professor Emeritus, department of physics, University of Manitoba
Co-director (with Werner Ens), Time-of-Flight Mass Spectrometry Laboratory, University of Manitoba, since 1979
Prior Experience: Professor, department of physics, University of Manitoba, since 1953
Director, cyclotron laboratory 1959-67 and 1968-74
PhD, Princeton University, 1955
BS, University of Manitoba, 1948
What led you into mass spectrometry?
I am really a nuclear physicist by trade, that’s what I did in my thesis [on] at Princeton. When I left Princeton, I came back to Manitoba, which was my undergraduate institution, expecting to stay for a year or two, and it has stretched out for another 40-odd years. When I came back here, I was in charge of constructing a cyclotron, and then I was the director of that lab for a number of years. Among other things, we were asked whether we had some nuclear method of analyzing grain for protein content. From my work at Princeton, I knew there was a particular isotope of oxygen that could serve as a signature of the protein content.
When we were doing this, I came to realize that just finding the protein content was by no means enough, you really wanted to know what the balance of amino acids was. I asked a colleague who was doing mass spectrometry if there was any way of doing this with mass spec, and he said I should go down and see Ron Macfarlane at Texas A&M. He had developed the so-called PDMS (plasma desorption mass spectrometry) technique for looking at large biomolecules by mass spectrometry, and he was one of the first to be able to see masses on the order of 1,000 Da, which was very large at that time, back in the 1970s.
Brian Chait, who is now at Rockefeller, was working with me at the time on this grain business. After I had been down to see Ron, we decided that we would stick our necks out and build a time-of-flight mass spectro- meter. We got a grant for about $4,000 from our graduate studies faculty, and we were foolhardy enough to go ahead. Our first result was the spectrum of potassium iodide, not exactly a large biomolecule, which we reported at the ASMS meeting in Seattle in 1979. It didn’t cause too many waves, but shortly afterwards I ran out of money, and we found a job for Brian at Rockefeller.
At almost the same time, there was a student coming around looking for a place to lay his head; this was Werner Ens. He started working with me, and shortly afterwards another very good student, Ron Beavis, came and started working with me, and in the next few years we made a little bit of an impact on the field. We have been active in it ever since, particularly developing methods for analyzing large biomolecules by time-of-flight mass spectrometry, and at the same time trying to do enough applications to show that these techniques really had some value for biologists.
How did your collaboration with Sciex come about that resulted in the QStar?
The people at Sciex had been dipping their toe into the time-of-flight waters, but they had never decided to go ahead and do anything along that line. The catalyst for this development was Bob Boyd, who brought us together. I came up with the quadrupole time-of-flight hybrid, because I thought we knew something about time-of-flight, and of course Sciex was already an expert in quadrupole machines, so it was kind of a natural fit. We got together for dinner at a meeting [and] hashed that over with the Sciex people, and it was decided to go ahead with developing a quadrupole time-of-flight hybrid.
The first thing we were going to do is to take one of their quadrupoles and fit it onto a time-of-flight machine that we already had. An excellent Russian postdoc named Igor Chernushevich was available for this job in our lab, and the Sciex people sent a team of theirs to settle the nitty-gritty — what sort of bolt holes you need and things like that. Within a few months, by the end of that year, 1996, we had a hybrid machine working. There was quite a long gap — a couple of years or so — between the time we developed that and the time that the commercial instrument was put on the market.
With the assistance of another excellent Russian postdoc, we had already developed a method for putting electrospray onto the time-of-flight machine, under a cooperative grant between Sciex and ourselves, from the Canadian government. When we proposed that, we had suggested that it might also be possible to couple MALDI to this sort of machine. In the meantime, we had been developing the method of cooling ions by putting them through a collision cell, which had not previously been applied to a time-of-flight machine. That turned out to be a very essential part of the idea of coupling MALDI to this hybrid machine.
So your group also developed the orthogonal MALDI source for the instrument?
That was developed in our group more or less by ourselves, and Sciex decided to license it. When we originally proposed that, [it] was thought to be a pretty far out idea, so it was something that was very approp- riate to work on in a university rather than in an outfit like Sciex that has to worry about the bottom line.
[That technology is also used in PerkinElmer’s new MALDI-TOF]. The person who was most instrumental in developing that was Alexandre Loboda, another Russian postdoc in our lab who went on to work at Sciex. It’s an orthogonal injection time-of-flight [instrument] that has rather high resolution and mass accuracy.
Were you and Sciex the first to develop a hybrid quadrupole time-of-flight mass spectrometer?
No, the people who did it first were people at Micromass and Howard Morris at Imperial College; they were slightly ahead of us. We were working along the same lines at the same time, but they published before we did.
How do the ABI/Sciex QStar and the Waters/Micromass Q-TOF differ mainly?
Forgetting about the MALDI aspect of it, they are basically the same machines: I don’t think there is any real difference in principle between them. Micromass uses an octopole instead of a quadrupole for one of their components — a couple of details like that.
Your group did extremely well in this year’s ABRF proteomics research group study — you were the only ones to determine two phosphorylation sites correctly. What was your secret to success?
We used our QqTOF machine and were careful about mass accuracy, and did it in a very straightforward way. There are fancy things you can do to try to upgrade the amount of phosphorylated protein, and you can do some fancy HPLC. We didn’t do that, and it turned out that the people who tried had limited success.
One very important component of our success was that we had separated the sample by offline HPLC and put it down as a whole series of spots on a target. If you do this sort of thing online, you are limited in the amount of time you can look at things. You have a fraction coming out of the HPLC, and then a minute later another fraction comes out. In our case, we look at it with MALDI, and we can sit around and get some results and go off for lunch and come back and look at it some more. These time constraints that are always prevalent in these online couplings are just not present anymore in the offline way of doing things. [Also, using] the HPLC cleaned up the spectrum, it improved the signal-to-noise by more than an order of magnitude in the peaks that we were interested in looking at, and when we did the MS/MS measurement on them, it improved things by another order of magnitude.
What kind of performance do you usually get?
In general, we tend to measure things to 3 decimals. We get mass accuracies that tend to be better than 10 mDa (millidaltons). That’s fairly definitive. In some cases, it’s really the average of a number of peaks that you are measuring, and we get better than a few mDa. There are a number of cases where there is an ambiguity, that is, there are several possibilities at a given mass number, and if you are going to distinguish these, you have to get down to the mDa range. I don’t know that everybody realizes just how well these machines can work. The people who do FTICR measurements always talk about their high resolution and high mass accuracy, and certainly they can get higher resolution than anybody else, and they can get high mass accuracy if they are very careful. But for most purposes, the mass accuracy that we can get seems to be sufficient to distinguish these possibilities if you are doing sequencing.
What do you think will be the most significant advances in mass spectrometry for protein analysis in the next four years?
I personally think the more significant developments are going to be in being able to characterize posttranslational modifications more effectively. Most of the things we have looked at have been individual proteins, and in just about every case, there are various complications that take a long time to sort out. I think the development is going to be to be able to do these things much more efficiently and on a larger scale. Bioinformatics tools have to be improved a lot. I think the machines that are being built are gradually improving, and I suspect that in a few years, a few mDa mass accuracy is going to be pretty well routine in the state-of-the-art machines.
What can still be improved?
You need to worry about things like temperature variations, all sorts of little things that are relatively minor perturbations, but they add up. These things are pretty well known at the moment, but it’s just a question of incorporating them into the instruments that are on the market, and doing it in a way that you can send the instrument out into a lab and have it perform as well as it does in the factory.