At A Glance
Name: Laurence Brill
Position: Principal investigator, Protein Profiling, Mass Spectrometry, the Genomics Institute of the Novartis Research Foundation, San Diego, since 2004
Background: Postdoctoral Fellow, GNF, 2002-2004
Postdoctoral Research, the Scripps Research Institute, 1997-2002
PhD in molecular, cell, and developmental biology, University of California Los Angeles, 1997
MS in plant pathology, University of Illinois at Urbana-Champaign, 1990
BS in agriculture, Ohio State University, 1985
How did you first get involved with proteomics?
I became interested in proteomics several years before I got involved with it. I went to talks, especially those of John Yates, and it occurred to me that this is a great way to analyze lots of proteins simultaneously and produce what’s found in biological systems. And then that should result in increased understanding of how the biological systems are actually functioning, all the components together, or as many as we can detect and analyze.
What were you doing before you got into proteomics?
I was doing more traditional protein biochemistry looking at one protein in great detail. At that time, I was across the road at the Scripps Research Institute. [In particular], I was working on biochemical and biophysical analysis of a viral protein termed the tobacco mosaic virus movement protein. The data was used to determine some structural information, and a topological model of this putative membrane protein was proposed and refined. Our longer term goal was solving the crystal structure of the protein, and that work is [still] ongoing.
Tell me a little about what you are working on now.
In general, we are working on phosphoproteomics; that’s one of the major efforts of our group here at GNF. We focus the most on tyrosine phosphorylation, largely because it is such a key to cell signaling, to the proteins that control what cells are going to do: if they are going to proliferate, if they are going to differentiate, if they are going to die, et cetera. ERBB2 is a receptor tyrosine kinase that is frequently overexpressed in cases of very aggressive cancer, most notably breast, ovarian, and some gastric [cancers] as well. We sought to identify a lot of the components of the signaling network of ERBB2, which has been poorly understood, in spite of extensive efforts at the single protein level. The future goals would be to take a closer look at some of those that seem to be key in metastatic cancer behavior, for example.
What techniques do you use to do that?
Our technology platform is based on immobilized metal affinity chromatography, or IMAC, LC tandem mass spectrometry. We developed the technology platform to the point where it’s highly selective for phosphopeptides from complex mixtures. Then we subject those enriched phosphopeptides to an MS/MS analysis in order to assign the tyrosine phosphorylation sites. And we also [analyze] the serine threonine sites that come along as well. One key to the technology is using anti-phosphotyrosine immunoprecipitation, because tyrosine phosphorylation is at such low levels.
How many of the components of the ERBB2 signaling network have you identified?
We’ve identified 88 proteins that are involved with the ERBB2 signaling network. That was on an older ion trap instrument [an LCQ]. With some [of the] present work, we are getting larger numbers of phosphorylation sites, largely because [of] a faster, more sensitive ion trap that we are now using [an LTQ].
You have been using this information in conjunction with drug treatment?
Yes, indeed. In the ERBB2 research project that we have had going, we use a monoclonal antibody, which is actually an approved therapeutic called Herceptin, [which] specifically inhibits the activity of the ERBB2 receptor. That was a very focused way to see which of various [phosphorylation] sites were involved in ERBB2 signaling. It turns out most of them that we detected in untreated cells probably are, because most of them disappeared, [or] at least went down to an undetectable level, following Herceptin treatment.
What is the next step?
Several things come to mind. From [a] technical standpoint, we want to get more and more sites that are biologically relevant. [What] it is going to involve I think is increased sensitivity from the LC/MS standpoint, and we have some ideas in mind for trying to get higher sensitivity. It has to do with especially highly sensitive LC setups. The ion trap is pretty well set as far as its sensitivity. We are trying to improve the actual chromatography itself, which is directly coupled to the electrospray ionization.
What else are you working on in proteomics?
[What] we are doing is best suited to looking at cancer biology systems, that’s what we are pretty much focusing on. We have looked at the BCR-ABL chronic myelogenous leukemia system [published in PNAS in 2003]. It’s a cellular system where the leukemia cells in the vast majority of cases have a constitutively active fusion tyrosine kinase called the BCR/ABL kinase. Inhibition of that, specifically, can lead to effective therapy for CML patients.
Do you collaborate with other groups?
We do have some internal collaborations here at GNF, as well as a few external collaborations. The oncology group here at GNF is one of our main internal collaborations, and we have a collaboration with Toshiaki Kawakami at the La Jolla Institute of Allergy and Immunology, and one with Bert Vogelstein at Johns Hopkins University. We also have an agreement that was just signed for a collaboration with a company, and we hope that collaboration matures into one that will be useful for the cell signaling/protein phosphorylation community.
When you think about the field in general, what are some of the big obstacles or difficulties that still need to be overcome?
The first thing that comes to mind is being able to really narrow it down a little bit more effectively to the proteins that are especially important in the biological processes, really being able to sift through huge numbers of different proteins and more effectively find the ones that are key. And then, in a related vein, I think that we need to be able to get ever increasing numbers of proteins. It seems as though the field is a little bit stuck at circa 1,500 peptides per MudPIT analysis. I think that improved chromatography is going to be helpful for getting things better separated. And also [what I think needs to occur more] is focusing on a more selected sets of proteins, for example analyzing the ribosomes, nuclear pore complexes, specific organelles, or subcellular assemblies.
Is there a need for other technological improvements?
I think that sensitivity of instrumentation is pretty key, being able to pick up extremely low level peptides with reliable signal-to-noise ratios. I think that there has been a tremendous amount of progress in that area, but that continued progress is really going to help. I think that bioinformatics will be very, very important to be able to more effectively sift out and identify the real, reliable peptide hits from those that are a bunch of noise.
Do you think there is still value to making these lists of proteins?
I think there is substantial value to that but that’s not enough. Those lists of very exciting and important players have to be followed up in more detail and narrowed down to pursue a more traditional biology/cell biology type approach.
What does the future hold for proteomics?
People need to realize that there are no overnight miracles, that there is a great deal of promise. We are starting to get closer to realizing some of the promise of proteomics, but a lot of work still needs to be done. In such a young field, we just need to chug along and keep making all the improvements we can and learn what we can on the way, do smarter and smarter experiments, and get more focus.
Do you think expectations have been set too high, or too fast?
I think so, definitely. In the biotech area in general I think that’s been true ever since it started in the mid-70s. People had silly ideas of what was possible. As a little bit of an exaggeration, they thought that we grow pork chops on soybean plants, which obviously isn’t going to happen. I think that expectations need to be realistic, yet optimistic.
What would be the proteomics equivalent of that expectation of growing pork chops on soybean plants?
I would say instant drug targets [or the idea that we] are going to suddenly cure all cases of cancer. Biology is just so complicated. I think that proteomics is perhaps the strongest tool we have for analyzing and understanding, but it’s going to be a tremendous challenge for many decades to get a better handle on biological systems.