At A Glance:
Name: Scott Ficarro
Age: 27
Position: Postdoc, protein profiling group, protein sciences department, Genomics Institute of the Novartis Research Foundation, San Diego, since 2001.
Background: PhD, University of Virginia, working with Don Hunt, 1997-2001. Studied peptide phosphorylation, androgen receptor phosphorylation, analysis of the phosphoproteome.
Recently published a paper online ahead of print in PNAS entitled “Profiling of tyrosine phosphorylation pathways in human cells using mass spectrometry.”
How did you get into proteomics?
I first got into mass spectro metry while I was an undergrad at Rider University in Lawrenceville, N.J. I was working in the mass spectrometry lab at American Cyanamid, and doing immunology research for Dr. Riggs at Rider. As I was finishing up, I talked to my boss at Cyanamid about what I should do for grad school, and he said 'there's this guy, Don Hunt, down at the University of Virginia who uses mass spectrometry to solve biologically problems.' So I started looking at the research, found it interesting, applied, was accepted, and started research [at U.Va]. My first project dealt with phosphorylation and IMAC enrichment. I found the field of phosphorylation very interesting, because, when phosphorylation is deregulated, it often leads to cancer.
How did you get to the Genomics Institute of the Novartis Research Foundation?
I had heard of Pete Schultz’s research, and met Eric Peters and Ansgar Brock [of the institute's protein profiling group] at an ASMS conference. As I finished my PhD, I found they were interested in doing phosphorylation work.
What's it like working in a private foundation like GNF? Does it feel any different from academia?
I like the situation I am in now because GNF is kind of a hybrid between academia and industry. We have more resources than you typically have access to in an academic setting and also the freedom to pursue whatever our research interests are. I really enjoy developing the methods and publishing them to help other people.
In your recent PNAS paper, you used a new approach to find tyrosine phosphorylation sites across a large number of proteins. What distinguishes this method from others?
We're finding a large number of tyrosine phosphorylation sites in a single analysis. [With] other methods you're typically looking at phosphorylation on a protein- by-protein basis. For example, some approaches utilize gel electrophoresis and then mass spectrometric analysis of each spot. This is very labor-intensive.
You mention in the paper that you initially performed immunoprecipitation of undigested proteins, rather than tryptic peptides, and went straight to HPLC/MS. Why didn't this approach work?
In a typical immunoprecipitation, you're dealing with on the order of more than a thousand proteins. When you digest that mixture, most of the peptides are not going to be phosphorylated, and these non-phosphorylated peptides get in the way. We did see peptides from a few signaling proteins, but most peptides were from cytoskel-etal proteins, like actin and myosin.
You then added methyl esterification of the tryptic peptides and IMAC, before performing nanoflow reverse-phase HPLC/uESI/MS. How did you come up with this additional step, and what does it add?
I first used methyl esterification to increase the selectivity of the IMAC column toward phosphopeptides while at the University of Virginia, but this was applied to total protein digests, where we targeted serine and threonine phosphorylation. The idea came while I was showing results from IMAC of a complex mixture at a group meeting. Don [Hunt] was asking ‘you did IMAC, so why are you not seeing much phosphorylation?’ I explained that when you take a complex mixture and put it over the IMAC column, it is enriched for acidic peptides, and these obscure the detection of most phosphopeptides. Don suggested blocking the carboxyls by making methyl esters. I went ahead and demonstrated that if you did the methyl esterification on digests of complex protein mixtures, and then performed IMAC on the resulting peptides, you greatly increased the select ivity of the column for phosphopeptides.
When I [began working] at GNF I saw a lot of interest in tyrosine phosphorylation. Since it represents a very small fraction of total phosphorylation in the cell, it’s difficult to visualize. We worked out a method whereby we could enrich for this particular type of phosphorylation.
You were able to find a number of additional tyrosine phosphorylation sites, including over 30 that were previously not identified. What is the impact of this finding?
Tyrosine phosphorylation is very important in regulating protein function. The sites we found are obvious candidates to study biologically. We’re trying to use tools like Scansite that, from the motif of phosphorylation, can give you some insight into potential binding partners and phosphorylating kinases. For example, we detected a novel phosphorylation site on Cas-L, which is a docking protein involved in T-cell signaling. [Cas-L] is known to interact with the Lck kinase, and actually the site we find has a consensus for binding to the Lck SH2 domain. The binding for these two proteins could be through the site we found. So, the bottom line is that the procedure yields a large amount of information from a single MS analysis; this information can place proteins in signaling pathways and serve as a starting point for biological investigation.
Is that what you are doing now?
Right now, we are looking at different systems for novel members of tyrosine kinase cascades that could serve as drug targets.
How effective is Scansite in doing its job?
There's only a limited number of kinases in the database so it doesn't reveal every interaction. But in certain cases it helped us make some connections, like with the Cas-L protein.
You also looked at the impact on tyrosine phosphorylation of treating cells with STI571, or Gleevec. What is the significance of your findings for understanding this drug?
Our results show phosphorylation patterns of cells expressing the BCR/Abl tyrosine kinase before and after treatment with STI571. In the drug-treated cells, we see inhibition of BCR/Abl phosphorylation and induction of phosphorylations that indicate the cells are different iating. Thus, our method can be used to test the efficacy of drugs in treating CML. This profiling could in principle be applied to any system, for any drug. Although many novel phosphorylation sites were discovered in our experiments, further work will be necessary to demonstrate their relevance to CML.
Right now we are looking at different cancer cell lines to try to find new members of signaling pathways, using the same method.
What else are you working on?
We are also focused on developing the method from two different fronts, the first being sample preparation, the second being instrumentation. We'd like to automate the procedure, because right now, the sample handling and column preparation requirements are quite extensive. We would like to automate the IMAC and LC/MS steps. We would like to have the derivatization procedures, especially the methyl esterification, automated as well, but this has been difficult. We'd also like to improve our ability to quantify the changes in phosphorylation by differential methyl modification. I did some initial work on this at U.Va. by looking at differences in phosphorylation that happened during sperm matur- ation. Also, we'd like to see improvements from the instrumentation side. The ion trap is very powerful but it does have some limitations in terms of mass accuracy and resolution. We'd like to transform this technology to [one] using quadrupole time-of-flight mass spec, or even MALDI-FTMS. Our group has just published some articles on the MALDI-MS platform — in American Pharmaceutical Reviews and LCGC Europe.
We have our MALDI FTMS platform working. It's fully automated and we get about 2 ppm mass accuracy with about a resolution of a hundred thousand. We can analyze a 1,536-well plate in four hours. We are applying this technology to study several different systems. In the future, I will work IMAC enrichment into the platform so we can leverage the advantages of FTMS to phosphopeptide analysis.
In your paper, you mention that you did not detect some known phosphorylation sites. Are you developing new methods to try to find these sites?
We definitely would like to get at these other sites by trying different detergents that solubilize these other proteins or utilizing different proteases, so the peptide fragments are amenable to mass spectrometric analysis. We're working on that now.
In general, these analyses give a lot of information, but it will take a long time to decipher all the interactions. We're seeing a snapshot of the phosphorylation in the cell, but its kind of a blurry picture; we're not seeking everything. But it can give you a glimpse of the major players.