Fred Hutchinson Cancer Research Center
At A Glance
Name: Barry Stoddard
Position: Faculty and principal investigator, Fred Hutchinson Cancer Research Center, since 1993. Co-director, graduate program in molecular and cellular biology, University of Washington School of Medicine and FHCRC, since 2000.
Background: Department of Biochemistry, University of California, Berkeley, 1990-1992.
PhD in biophysical chemistry, Massachusetts Institute of Technology, 1990.
Barry Stoddard is scheduled to give a talk on engineering new proteins for research and medicine at next month's Canadian Proteomics Initiative conference in Edmonton, Alberta, Canada. ProteoMonitor spoke with Stoddard to find out about his background and work in enzyme engineering.
How did you get into structural biology and protein engineering?
I met a professor in graduate school who was an x-ray crystallographer who just absolutely sold me on the things in that field. I worked for Greg Petsko at the Massachusetts Institute of Technology. He was just the most enthusiastic and interesting researcher I had ever met, and he basically completely changed my ideas about what I might be interested in. So in my case, I didn't intend to go to structural biology. It was a case of hooking up with the right mentor who really showed me how interesting the field is.
Greg was very interested in the question of what makes enzymes such efficient catalysts of chemical reactions. He was really interested in the structure and function mechanism of enzymes. So I worked with him on a couple of different systems. This was back in the mid- to late-1980s. And I still work on the same types of questions today. I'm very interested in eyzymatic catalysts and their properties.
So I did my PhD with Greg, and I got that in the 1990, and then I actually took a little sabbatical from crystallography and structural biology. I did my postdoc with Dan Koshland, who for quite some time was the editor of Science magazine. And I went to work with him because I wanted to learn a little more about protein biochemistry and molecular biology. I worked for him for two years, then I came to the Fred Hutchinson center here, and I've been here ever since.
What kind of questions did you work on when you got to Fred Hutchinson?
It took a few years to start something new and original. I came here really continuing to work on the same sorts of enzymes and questions that I did as a graduate student and postdoc. Originally I was working in an area called time-resolved crystallography. It was a method for trying to solve structures of intermediates in a reaction pathway — trying to take snapshots during a reaction, and to piece together images along a pathway of a reaction using a variety of methods.
Subsequent to that, I really developed an interest in the area of enzymes that are used for targeted genetic applications. And a large part of that is trying to engineer those enzymes to have novel properties and functions that are of use. So the common thread in all of that is I've always been interested in proteins and enzymes as reagents and as molecules that you can play with and do things to — you can alter their properties, and treat them as tools, and work with them in that way.
Back when I was doing time-resolved crystallography, I was very interested in manipulating enzymes to allow me to take snapshots of them in action. And then after I became interested in enzymes that are interesting for targeted genetic applications, I became very interested in engineering them to have novel properties of substrate recognition.
Is the work that you do now primarily having to do with enzyme engineering?
I should say upfront that we're engaged in both computational experiments and genetic selection experiments, and in both cases, there's a collaboration with another laboratory that really provides us with the bulk of technology to allow us to carry those things out.
The protein engineering work that we do is all really driven by David Baker's lab at the University of Washington. It's his laboratory that has really done amazing things developing computational algorithms to engineer enzymes and proteins. What my lab has contributed has been the underlying structure/function studies of both the wild type and the re-engineered variants of those enzymes, as well as the applications of interest. We've been very interested in enzymes involved in genetically targeted applications, so we have some really good model systems to apply engineering tools to.
But all of the actual computational code and algorithms for the engineering work we've been involved with has been developed by David Baker's lab.
Do you first engineer the enzymes computationally, then go about testing the engineered enzymes experimentally?
Yes. Typically, there's a design cycle, which the two labs are intimately involved with. The design cycle usually starts with understanding the structure and function and mechanism of a wild-type protein, and then subsequent to that, applying computational tools for some element of redesign, or alteration of properties, followed by going back to structural and biochemical analyses to validate and verify the results of the redesign.
The design cycle concept is a really import one. You start with the structure/function information on the naturally occurring protein, and then you apply tools to redesign that protein, and then you go back and further study structure and function mechanisms of the redesigned proteins.
So the two labs work very synergistically together. Between the two labs, we have all the tools we need for all elements of the design cycle.
What kind of tools do you use on the experimental side?
On the experimental side, structure determination is the biggest, using x-ray crystallography, because at the heart of everything the two labs are doing together is trying to redesign the structure of a protein in some way, and determining experimentally the structure is very important to see how accurate the predictive methods are, and to validate the redesign.
And then also, there are a large number of biophysical and biochemical characterizations of underlying functions and stability. Things we might do is study the folding and refolding behavior of the redesigned proteins — we do a lot of that. We study substrate recognition and binding affinity and specificity, and we measure catalytical efficiency — actual turnover rate.
How do you pick which proteins to work with?
In my lab, I work in two different fields that both are associated with genetic targeting, or gene-specific applications. We work on proteins that in some way can be applied in a genetically targeted manner toward a disease state.
One of the types of proteins we work with are called homing endonucleases. Those have the property of recognizing specific sequences, or a site, in an entire genome. They have the innate ability to recognize and bind to one and only one DNA target site in a complex genome, and then to do some sort of chemistry there. Usually it's a gene conversion event — it's driving a change in the DNA sequence. And that's potentially useful for gene therapy, and for population engineering, and for all sorts of genetically targeted applications.
Have you actually redesigned a homing endonuclease?
So what we are doing is solving the structures of the naturally occurring versions of these homing endonucleases, and then we're engaged in selection and redesign towards novel sites. And what we want is to change their specificity eventually towards a site that's found in, say, a gene therapy target — perhaps a sequence in the cystic fibrosis gene, or the factor VII gene, which is mutated in hemophilia A.
[W]e actually have a paper coming out in Nature in a few weeks showing that we could use computational redesign to change specificity of these proteins towards novel DNA sites. We haven't yet redesigned a naturally occurring homing endonuclease to recognize a physiological target site. That's the next step. But we've been able to change specificity towards a novel site. So we basically alter the structure of the protein to recognize a novel DNA target.
Part of that is you want to maintain specificity. It's easy to take any enzyme, including a homing endonuclease, and make it less specific, so it recognizes a broader collection of substrates. But that's not what you want to do. What you really want to do is you want to shift the specificity, instead of just broadening specificity. So you want to have both positive and negative design. That's something that we're working on very hard in our design cycle — to incorporate elements of both positive and negative selection, or redesign. Because we want the novel reagents that we make to have specificities that are equivalent to their naturally occurring counterparts. And that's probably the holy grail of protein design and engineering in general. Because biology is really all about specificity at the molecular level, and that's a very hard thing to recapitulate with a computer program.
So the next step would be to actually synthesize the redesigned protein?
So once you've done a redesign in silico, then, yes, the next step is to actually create the necessary construct to make the corresponding protein, and then to test it. And we test the redesigned protein both in vitro, by purifying the protein and testing it against purified DNA, and also in vivo, by putting these redesigned proteins into cells, and asking them to actually carry out a biological process.
We're just starting with in vivo testing of some of things that we have redesigned.
So homing endonucleases are one type of protein that we've been working very hard on doing reengineering on. The other family of enzymes that we work on are enzymes called cytosine deaminase. They are used for an anti-cancer therapy, called pro-drug gene therapy.
We've done redesign to introduce novel properties into those enzymes, and we're just starting to put them into in vivo tests — mostly cell culture, as well as animal models.
Are these redesigned proteins something that you would patent, or commercialize?
Yes, in principle. We have not done that, but in principle, a reengineered protein that has a novel property is a new material or reagent that could be patented.
So with these redesigned proteins, you're currently at the stage of testing them in vitro and in vivo?
Yes, that's right. My hope is that within the next three to four years, we might be able to generate engineered variants of these types of enzymes that could actually be used for pre-clinical trials against various types of disease states.
From what I've seen of structure-based engineering, I think the field has matured over the past five years to the point where people can actually start thinking of applications of various sorts. It's a very exciting time.
Are you involved in developing technology to improve this area of research?
That's really David Baker. The two labs together have had a lot of success, and it's primarily due to the fact that between the two labs, we can apply every step of the design and validation cycle to very specific applications. So we have the combination of some very powerful models systems, the development of technology, and the subsequent validation, all of which is important, encapsulated between the two labs.
Where do you get your funding from?
Mostly from the NIH, and we also have support from the Gates foundation.
What other projects do you have planned for the future?
The other big project which I hope to eventually bring to the point of engineering, is we've been involved for a while in looking at structure and function of a protein called factor VIII, and another protein called Von Willebrand factor, both of which are clotting factors, or blood coagulation factors. When these two proteins are mutated, people have either hemophilia A, or Von Willebrand's disease, which are two of the most common types of bleeding disorders.
And the treatment for those disorders is a protein-based treatment. You actually treat people with recombinant factor VIII usually. So we've solved the structure of full-length factor VIII, and I hope eventually to be able to apply tools of protein engineering to improve stabilized versions of factor VIII. That would eventually help to treat hemophilia.
Seattle has a history of really excellent research in blood coagulation, and there were laboratories here that had constructs and opportunities for us to work in this area.