At A Glance
Name: Catherine Costello
Position: Director, Mass Spectrometry Resource, Boston University School of Medicine
Professor of biochemistry and biophysics
President of the American Society for Mass Spectrometry
Background: PhD studying organic reaction mechanisms by NMR, Georgetown University
Postdoc developing mass spectrometry techniques, MIT
Recently won a $12 million contract as part of a seven-year proteomics initiative of the National Heart, Lung, and Blood Institute.
What led you to proteomics?
I did my PhD at Georgetown University with Charles Hammer, studying NMR and organic reaction mechanisms. Midway through graduate school, in 1966, I went to one of those NATO summer conferences, in Glasgow, which was on mass spectrometry. I became aware at that point of what Klaus Biemann’s group was starting to do with mass spectrometry at MIT, and I applied to him for a postdoc position that started in 1970. I thought I was going there to do a normal-length postdoc, but as it happened, Klaus asked me if I would stay and become the Associate Director of his mass spectrometry resource, which I did, and I ended up staying there for 24 years.
From Klaus Biemann’s group came many of the present leaders of the mass spectrometry field. He was already studying proteins back in 1960, working on how to break the proteins into shorter peptides that could then be studied by GC/MS. In fact, he had a grant for studying proteins by mass spectro-metry which ran for 40 years; it ended in 1998.
During a large part of the time, I was working on the core research projects in the MS Resource, which meant developing new methods and new applications for instrumentation. We brought in field desorption, and then fast atom bombardment, and then matrix-assisted laser desorption/ionization. We had the first high-field four-sector mass spectrometer that was used for protein and carbohydrate sequencing by tandem mass spectrometry. It was installed in 1985.
Funding we obtained in 1989 to get an instrument to do MALDI prompted Marvin Vestal to start manufacturing that [instrument] within his company [Vestec], because at the time, there was no manufacturer. We submitted that grant proposal October 1, 1988, just at the time that MALDI of proteins had been described by Franz Hillenkamp and Michael Karas at a meeting in Bordeaux. We had followed their earlier papers, published during 1985-87, that first showed MALDI of peptides up to about 3000 Da, the mass range now used for most proteomics measurements. Our proposal was funded the next spring, and in the summer we placed the order. The instrument was delivered in December 1989.
Although I was involved in many projects with proteins, I was primarily developing — and I still have as a large priority — methods for carbohydrate and glycoconjugate analysis, which include glycoproteins.
When did you move to Boston University?
I moved over here to Boston University School of Medicine at the end of 1994, when Klaus was about to retire, and established the facility here. Part of the reason for coming here, to a medical school, was that we could work increasingly well with very small amounts of samples and complex mixtures. The immed- iate application to biological and medical problems was becoming more and more evident, and I thought it was better to have the Mass Spectrometry Resource in the context of a medical school where we could have very close communication with the people with the biological and medical problems and bring state-of-the-art mass spectrometry to where the problems were.
Has that promise borne out?
Yes, definitely. We have a number of collaborations that have to do with medical problems, and are developing a big program on proteoglycans. We are also part of an NHLBI program project grant for studying amyloid diseases. In fact, we have the dedication of that labora-tory tomorrow, and that receives a lot of financial support from patients and patient families. That kind of cooperation wouldn’t have happened from a distance.
What kind of amyloid diseases do you study?
We focus on two. One is an inherited disease involving a protein called transthyretin. It is a blood protein, and when there are mutations in the gene for that protein, they can lead to a deposition of the misfolded protein in different organs that causes organ failure and death. To date the only way to treat that successfully is a liver transplant, because most of the protein is produced in the liver. That certainly has helped many patients, but one wants to understand the disease better, so you could have a better treatment.
The other disease involves overproduction of immunoglobulin light chains that can deposit in organs, usually including the heart. Those patients have a very poor prognosis: In fact it used to be that no one survived. Now they are being treated with a combination of chemotherapy and autologous stem cell transplants, and many of them now survive longer and some may be cured.
We use mass spectrometry in both cases to identify the proteins, and find the mutations and the posttranslational modifications to try, first, to help with the diagnosis, and second, long-term, to understand the process of amyloid formation with the hope of developing new treatments.
You mentioned that, as part of your new NHLBI proteomics center, you want to develop a new FT-MS instrument. What is special about the instrument?
This instrument is different from available FT-MS instruments in that it will operate at very low temperatures. Large superconducting magnets that are used both for mass spectrometry and NMR, and also for imaging patients, are cooled with liquid helium and liquid nitrogen on the outside to keep them in a superconducting state. We want to use a magnet type that is used for NMR, which has a higher field and is more readily available than the ones that are made primarily for mass spectrometry. We want to remove the insulation that’s normally in the center and insert the mass spectrometer vacuum system and cell directly into the core of the magnet, so that the mass analyzer will be operating at liquid helium temperature. This will greatly improve the vacuum. We will also have the amplifiers and critical parts of the electronics at very low temperatures, which will improve their performance by reducing the noise. We should achieve a significant increase in sensitivity, and by using the higher field magnet, get higher sensitivity and higher resolution, while keeping the cost down. We would hope to go routinely to the femtomole levels, and perhaps below. We will also have a samples stage that will hold probably several hundred samples at once, so we can do high-throughput. We are working with a bioinformatics group to develop control software that will identify many peaks and target the unassigned peaks for MS/MS analysis. The instrument will particularly focus on posttranslational modifications, things like nitrosation in oxidative stress and maybe some we haven’t seen before.
Will the new instrument be commercially available?
That’s the intention. BU has filed for the provisional patent, and we have talked with some instrument companies, but it’s premature to predict the time course at this stage. We assume that this would be something that would be attractive, because it would give you a higher performance at a reasonable price.
What led you to study carbohydrates in your work?
We have a resource grant from the National Center for Research Resources, and a large part of the methods development within our resource grant is to create better methods for the analysis of carbohydrate structures because this is a much less developed field than protein or DNA sequencing. Yet carbohydrate modifications occur on probably half or more of the proteins, and there is a lot of reason to believe that there is biological significance to these. But the correlation between structure and activity has not been so well developed because it’s been so hard to study the structures. Now we are in a position where we can start to do these.
Where do you see proteomics headed in 5-10 years?
As we develop the capability to identify proteins and to compare cells and organisms, particularly humans, in different states of health and disease, we will be positioned to understand better how these complex biological systems work, and the pathways that are interacting. It’s very interesting that, so far, sets of interacting proteins that were identified on the basis of mass spectrometry do not show a lot of overlap compared to results people have obtained by using the yeast two-hybrid system. We are getting a different window to look on things, and to understand, if we want to develop treatments and interventions with diseases, what would be the points of entry that would be the most effective, that would do the most good, while having the fewest deleterious effects. It requires cooperating with the clinicians and the basic scientists and the computer scientists to get meaningful answers. I think our institution in particular is one in which people with these different skills are really well integrated.