Department of Microbiology and Immunology, Stanford University
At A Glance
Name: Matthew Bogyo
Position: Assistant professor, Department of Microbiology and Immunology, Stanford University, since 2004.
Background: Adjunct faculty member, Department of Pharmaceutical Chemistry, University of California San Francisco, since 2002.
Scientific consultant, Axys Pharmaceuticals, 2000-2001.
Faculty fellow, UCSF, 1998-2001.
Postdoc, Harvard Medical School, 1997-1998.
PHD in biological chemistry, Massachusetts Institute of Technology, 1997.
Matt Bogyo gave a talk on proteomic profiling of proteases at this week's Association of Biomolecular Resource Facilities conference in Long Beach, Calif. ProteoMonitor spoke with Bogyo to find out more about his background and the technology he has developed to study proteases.
How did you get interested in studying proteases?
I trained initially in chemistry. I did my graduate work in the organic chemistry department at MIT, and then I realized that I was interested in more than just making molecules. So I got interested in how I could use chemistry to apply to biological systems.
I ended up working for an immunologist at MIT named Hidde Ploegh. He was studying MHC Class I and Class II antigen presentation pathways. Those two pathways both begin with production of antigenic peptides from larger proteins, and those both involve proteolytic pathways.
When I was in his lab, I started studying these enzymes called the proteosome, which is a multi-component protease complex that cleaves lots of different proteins inside the cell. It was a hard enzyme to study because it was made up of many different active sites that all are proteolytically active. So you really needed small molecule or chemical tools to dissect what each [subset] was doing.
I got interested in applying chemistry to make inhibitors that would covalently inactivate target active sites. What we realized really quickly was that if you could covalently attach something to the active site of a protease, you could then do that in the context of a complex proteome. So you add these small molecules if they were sufficiently selective, and they would only modify your target enzyme.
So we started building on that concept, and I eventually moved to studying other proteolytic enzymes — mainly lysosomal cysteine proteases.
I moved to UCSF after I finished my graduate work at the end of 1988 to start my own lab as a faculty fellow — a sort of independent, young investigator type of position. I was there for about three years. And then after that I went into industry for a little bit. At the time we were doing all the small molecule stuff, I was consulting for a small molecule protease inhibitor company called Axys Pharmaceuticals in South San Francisco.
What happened was Axys got purchased by Celera Genomics. Celera had just finished the human genome, and they were now moving on to the concept of doing proteomics. So they had made this big effort to buy a huge number of mass spectrometers, and to begin to do large-scale proteomic studies to basically look for biomarkers and also therapeutic targets in oncology.
Is that where you first got interested in proteomics?
No, I got interested in proteomics from the angle of mass spectrometry very early on. A lot of the stuff we were doing involved using a small molecule to modify a protein or set of proteins, and then isolating those proteins and identifying them by mass spec. So I got interested in mass spec back when the first peptide sequencing was being done.
We've always had a kind of split between chemistry and biological applications. So I've always had people in the lab that are doing real organic synthesis, and others that are doing more of applications of the small molecules to biological studies.
The reason why it's such a nice system from the standpoint of proteomics is that the major stumbling point, or Achilles' heel, of proteomic studies is there's this huge variance in dynamic range of expression of proteins inside a cell. It can range anywhere from 50 copies of a protein to 106 copies of protein. And that makes it virtually impossible to just break open a cell and separate all the proteins and analyze them by any sort of analytical method.
So we got interested in how small-molecule probes would allow you to enrich for specific targets — in this case proteases. If we wanted to study what proteases were doing, we wanted a way to just pull out the proteases that we were interested in, and not only just pull them out, but also tell how they were regulated at the level of activity, and not just protein expression. That's where we developed this concept of activity-based probes, which are basically small molecules that modify proteases in an activity-dependent way.
If a protease is inactive, or it's bound to some sort of inhibitor, you won't label it. You only label active forms. So it allows you to not only ask how much protein is present during, say, a disease progression, but it allows you say, 'As a tumor forms, which proteases get activated and when?'
Is your work similar to Benjamin Cravatt's work on activity-based proteomics?
Yes. Very similar. We basically started the field at the same time. Ben's a good friend of mine.
I think Ben's stuff has been much more directed towards the proteomics community. Our stuff is really focused more on proteases, and applications of those kinds of activity-based probes for looking at protease function in a number of specific biological systems.
We've been studying protease regulation in cancer, and protease regulation in various parasitic diseases.
Ben's done a lot more stuff focused on sort-of broad spectrum profiling — taking reactive species and looking at what kind of enzymatic targets you can modify.
How have you applied the technology that you've developed to disease research?
We've done a collaboration with Doug Hanahan to go into transgenic mouse models of cancer. Basically, the mice get cancer over a period of about 10 to 12 weeks, and the particular model we studied was the pancreatic cancer model. And basically what we did was use our active site probes in vivo. We introduced them into the animal and asked, 'Which proteases got active during progression of the tumor formation, and when, and what happens if we inhibit those targets — how does that affect the tumor?'
So it allowed us to do proteomic profiling in vivo and ask which proteases might be important for the disease.
We've also done some imaging applications of that, where we can take fluorescent versions of our small-molecule probe, and then go in and ask not only, 'Which proteases get activated?' but 'Where are they localized? What cells are expressing those proteases?'
It's a very powerful tool. It allows you to focus from the whole proteome down to a small subset of proteins that you're interested in.
Have you found certain proteases that would be good targets for disease therapy?
Yes. We published a paper in May 2004 in Cancer Cell showing that several of the lysosomal cysteine proteases known as cathepsins seem to be key players, because if you block those enzymes you have dramatic effects on both angiogenic switching — initiation of formation of blood vessels — and also effects on tumor growth, and the invasiveness of the tumors.
We think that those are definitely interesting targets. There are currently no anti-cathepsin inhibitors in the clinics for oncology. Most of the cathepsin inhibitors that are in the clinic are for areas like osteoporosis, because they're involved in bone remodeling. Also atherosclerosis is another area, because they seem to play a role in the formation of atherosclerotic plaques.
Those are the main indications at the moment, and part of the reason may be that protease activity that we see is coming from the microenvironment of the tumor, and not from the tumor itself so much. A lot of the models that pharma uses to look for effective inhibitors, or drug candidates, are simple xenograph models, which probably aren't very good.
I think in the near future we're going to see a lot more of proteases in the role of cancer, and potentially more compounds going into the clinic directed towards these targets.
How long were you at Celera?
Two years. We were doing research and development stuff. We were designing and building new probes and applying them in proteomic applications.
Why did you decide to leave?
I just decided that the academic environment was better for what I wanted to be doing. I basically liked the freedom of academics, and I also liked the collaborative aspects of academics, which kind of get lost in industry.
What are you working on now at Stanford?
We're doing technology development. We're continuing to develop activity-based probes. We're trying to expand into other protease families, particularly serine proteases.
We're trying to apply all these tools to specifically look at proteolytic pathways in a couple of human parasites, as well as in the cancer system.
We study the malaria parasite in my lab. The parasite has a number of protease systems that it uses to survive inside the host, and we're basically looking at which proteases are important, when they're activated in the life cycle of the parasite, and whether they're good targets for therapeutics.
For cancer we're looking at the lysosomal cysteine proteases.
How do you decide which proteases to target?
In this case it was a matter of having access to active site probes that we could use to look at specific targets, and also knowing from the literature that several of these types of proteases have been upregulated in cancer, though it was hard to tell what they were doing. So it's sort of having the tools, and knowing that they were important through past literatures, is the way we selected them.
Do you think the work with active site probes can be applied in vivo to humans?
Absolutely. We're doing studies now with fluorescent probes in mice, and we're just starting collaborations to move into radio-labeled probes. If those show efficacy in animal models and show low toxicity, those can move towards humans fairly quickly because currently those types of tracers are being used in humans — not for proteases, per se, but for other things.
If you're trying to figure out where your tumors are, or where the margin between tumor and normal is, you want something that will light up the tumor specifically. Since proteases seem to play a role in this remodeling, where a tumor is growing and cleaving up its local environment, we tend to see a lot of activity right around the edge of the tumor, so this is a useful thing if you're trying to figure out where the tumor starts and ends.
Is any of this technology patented?
Yes. We filed some provisional patents on some of the imaging agents, and we filed some disclosures and patents on some of our inhibitors based on the small molecule scaffolds we work with. One of the angles is therapeutics — can we come up with good inhibitors of these enzymes and use them as drugs, and the other angle is can we use them to image where the tumors are, and we've filed on both of those.
Is there commercial interest in your technology?
Yes, definitely. Certainly if we can get imaging tools that can be used for things that can actually go into humans, I think then they would have very large commercial interest if they have benefits, and they can be used for early stage diagnostics, or looking at efficacy of treatment of patients that are under a certain chemotherapeutic regime.
With the small molecules, there are several companies in the area that are interested in licensing the actual compounds as inhibitors to go forward with as drug candidates in oncology. We'll see how far that goes.
Do people request reagents from you so that they can do their own protease studies?
Absolutely. We send out reagents weekly. We get tons of requests all the time.
We received funding as part of a national technology center grant through the NIH Roadmaps Initiative. We put together, together with the Burnham Institute in San Diego, a protease center grant that's focused on proteolytic pathways.
Using that grant we have a chemistry core, which is funded to do just that — produce reagents that can then be made available to the community.