By John Russell
January 8, 2009 | Start with a person who should be sick but isn’t. Use DNA variation to hunt for broken genes that are keeping this person well. Set up a vigorous hypothesis generation and killing machine to mercilessly and quickly winnow through the options (no favored children allowed). Find and make drugs informed by the rare surviving hypothesis and race them through POC.
To a considerable extent, this is what Pfizer’s year-old Biotherapeutics and Bioinnovation Center is trying do with its Target Generation Unit led by distinguished human geneticist David Cox, some of whose accomplishments include co-founding Perlegen (see, Taking Data Storage to Infinity and Beyond), co-director of Stanford University Genome Center, and election to the Institute of Medicine (NAS). The BBC is based in San Francisco and much of the TGU is located at the BBC Research Technology Center (RTC), Cambridge, MA.
If drug finding still seems largely to be searching for needles in a haystack, Cox is trying to pick haystacks in which the right needles should glow and be inherently less prone to adverse events and he is using a genetics magnifying glass to find them. But this is not the whole formula. All of the BBC, which is led by world class developmental neurobiologist, biotech entrepreneur, and elected member of the National Academy of Sciences, Corey Goodman, seeks the nimbleness of biotech and a new, more effective approach to collaborating with academia.
The result, they and Pfizer hope, will be faster, less expensive, less problematic biotherapeutic discovery & development, and possibly the emergence a new paradigm that more effectively harnesses human genetics and is adopted throughout the Pfizer organization. Cox recently spoke with Predictive Biomedicine editor John Russell about the BBC and the TGU’s role.
JR: Can you start with a brief description of the Target Generation Unit?
Cox: The Target Generation Unit (TGU) is part of a new research division within Pfizer called the Biotherapeutics and Bioinnovation Center. Everything that we’re going to talk about is best put into that context. The BBC is a federation of small, independent biotech-like research units that bring together the best of biotech with the strengths of pharma to discover and develop high quality biotherapeutics through clinical proof-of-concept. One of those units is Rinat, a spinout from Genentech with cutting-edge antibody technology expertise; the second research unit is CovX, in San Diego, originally founded by scientists from The Scripps Institute, which is leading the development of an emerging biotherapeutics technology, uniting the therapeutic attractiveness of peptides with the beneficial clinical properties of antibodies.
The third unit is the RTC/Coley in Cambridge, Massachusetts and Düsseldorf, Germany, which is building Pfizer’s capability and competitive advantage in ribonucleic acid interference (RNAi) therapeutics. Each of these research units are focused on a novel and growing technology area, each unit harnessing its small size, unique culture and entrepreneurial drive to discover and develop breakthrough science in the most productive way
Also at each site are really good biologists carrying out human and animal cell based studies and animal in vivo studies. To complement their efforts, the TGU was created to put human genetics on the front end of that [process]. Human genetics has never been very successful by itself, so it needs human cell based models and it needs systems biology approaches to be really useful. I’d like to give you a sense for how we are integrating human genetics, cell biology and systems biology to provide an engine for the three units that make our biotherapeutics.
JR: So how are you going to generate good ideas for new targets?
Cox: The key is to start with a clinical question and outcome that you care about and use the genetics as a way of linking that clinical outcome into a specific biological pathway that’s meaningful. So what do I mean by that? Who are the kinds of scientists and what are the kinds of questions you might be interested in? For example, a key question today, with so many people suffering from diabetes, is why do these people have such a high risk of dying from heart attacks? There’s not a single medicine on the market today for type 2 diabetes that has been shown to reduce illness and death from heart attack. How could we figure out what part of the biology we need to focus on to deal with that major medical need?
Here’s how one might use human genetics. You start off with people that have all the risk factors for diabetes, but unlike almost everybody else with diabetes, they don’t develop cardiovascular disease. Those are unusual people. We’re not looking at people that have the disease that were trying to fix; rather, we’re looking at the people that should have gotten the disease but didn’t. Then what do we do with these people? By doing DNA sequencing quite globally on a large number of genes, we’re looking to see if there is evidence of rare DNA variations that may be the reason why these diabetics didn’t get cardiovascular disease.
JR: So these are genes that are working in healthy people generally, but broken in people who should have the disease but seem healthy?
Cox: We’re looking for the broken genes in the people who didn’t get sick. The likelihood that they are going to have side effects if you break those genes is much lower than for most genes. Now it doesn’t necessarily mean that most of these variants that you are going to see in these people will have completely knocked out the gene but in many cases the mutation will have reduced the affect of the gene somewhat but not completely. It doesn’t give you a home run, but it parachutes you in specifically to the biology you’d like to look at.
So based on those gene products you can do cell based models, which we’ll do in the TGU, to say if you knock out that gene completely and you’re looking at cell-based phenotypes, what kind of end points do you impact? What do you change? Here’s where the systems biology comes in, you perturb the system with that very focused initial gene and now you say, based on the output of that perturbation, are there better things within that aspect of biology that you should be looking at then the initial gene that you used. You know there’s two very different ways of thinking about systems biology. There’s this way, which is very focused. It’s basically starting with one part of the universe and expanding out. The other approach is collecting all the data and reverse engineering how everything works. They’re both absolutely fine intellectual constructs and they both undoubtedly contribute really useful information, but one is much more focused than the other.
JR: it’s really a top down versus bottom up approach.
Cox: Exactly right.
JR: OK, how will you do that? Is there a group of scientists for assay and another group for modeling?
Cox: One of the ways we’re doing this is to be completely integrated and not have separate groups. We want the people that are doing the genetics, the cell-based models and the systems approaches to be completely integrated with each other and we want to maximize that synergy.
JR: Do you build your groups around projects?
Cox: Yes, In order to improve efficiencies in research and development, Pfizer Global Research and Development has identified six high priority areas with the highest probability for success and address an unmet medical need. These include Oncology, Alzheimer’s disease, Schizophrenia, Pain, Rheumatoid Arthritis, and Diabetes. The BBC is no different. Right now, we are mostly focused on the areas of diabetes and oncology.
JR: How do you get from the vision to the reality? What needs to happen inside BBC and the TGU to develop this team approach to developing biotherapeutics more quickly and winnowing through lots of ideas to find the right targets?
Cox: The key is to have an intellectual framework of where each of the different components fits. You know the hypothesis that you’re testing, and when you do your cell based models you actually have a hypothesis that your are trying to refute by doing an experiment. You can have many ideas come from either the biology or the genetics but then within the BBC or within the TGU - these experiments are going on in both places - you collect the data and say is the hypothesis that you were trying to test still viable or did this experiment make it go away. Did it just de-prioritize it because another experiment or another hypothesis was much more promising?
JR: In pursuing this approach will you focus mostly on human cell assays and models versus animal models?
Cox: Well, we’re really trying as much as possible to do patient studies and the genetics is what allows us to do that.
Another critical component to this approach is external collaborations. For quite some time, Pfizer and others in the pharmaceutical industry have talked about how important it is to have external academic collaborations
The historic problem is that pharma has given money to academics but often hasn’t give any real detailed knowledge about what they are trying to accomplish. And the academics are happy to take the money but they aren’t really interested in the questions the pharma industry is interested in and they use it to do whatever they want. Here is an example of our new approach. Let’s say within the TGU or in the BBC or in Pfizer we don’t have the expertise on a particular area of human genetics. Are we going to set this up ourselves? No, we are going to go out and find the scientist that does have that expertise. But rather than just say do some cool stuff and get back to us if you find anything, we’re going to say, “We’re really interested in your models for this particular purpose and that we’d like to work with you to get this particular answer.” Now that’s a really different model because in the past I think that in general pharmaceutical companies have been quite reticent to share with people what they’re interested in and the approaches they are taking.
There will be information flowing out to the academic collaborators as well as information flowing back to the TGU and the BBC. Obviously not everything is going to be open to the public, but it will be a much freer flow of information that will allow there to be real scientific collaboration with external experts.
JR: Could people publish on their work even in those cases?
Cox: Absolutely. It’s a very different type of collaboration than is the standard. So now you say, ‘Is this pie in the sky?’ There’s lots of things that have to be worked out on this but we already have some collaborations and other ones that will be announced shortly. I think that leveraging what’s out in the biomedical world in an effective way is really, really important.
JR: Is there a good example of that you can cite now?
Cox: One is our recently announced relationship with UCSF. In a novel experiment to advance new drug discovery and development, as well as stimulate basic research with potential biomedical application, Pfizer entered a three-year collaboration with the multi-campus California Institute for Quantitative Biosciences (QB3), headquartered at UCSF. There we have a scientific relationship where UCSF scientists can put forward an idea on something that they are working on and are interested in. Then, if there’s a Pfizer scientist who feels that idea could be interesting for something they’re working on, those two scientists get together and write a proposal, which is submitted to a team assembled to help identify promising areas of mutual interest and facilitate project management for those projects that Pfizer agrees to fund. The new collaboration intimately links the scientific talent of UCSF with the extraordinary expertise of Pfizer. This truly is an exciting collaboration as it allows these researchers to come up with ideas and projects of mutual interest.
JR: How are you going to judge BBC progress?
Cox: Very simple. What matters to us in the BBC - our progress - is how many things that we put into humans in clinical trials show positive proof of concept in patients. Period. There’s no intermediate success.
JR: When do you think you will know how well you are succeeding? Biotherapeutics can be moved more quickly. Will it be in a year of six months?
Cox: You’ll have some idea in a year or two, but to show proof of concept in humans, I think it’s unrealistic to think that we will have the answer in less than three. But it better be a lot faster than ten years. That’s the big problem. You have to change the system where there’s accountability, where you’re going to have your ass on the line for your ideas. If you know that you’re going to be dead before your ideas will ever be shown to be good or bad, or you’re not going to be working at the company anymore, that’s really not a good incentive for having people showing urgency.
JR: What attracted you to this opportunity?
Cox: The ability to plug human genetics into a situation that can impact medicine. It’s remarkable what’s happened in human genetics over the past five or ten years. But the actual impact on health outcomes, to date, has been pretty pathetic. Part of the reason is that human genetics, by itself, can’t do it. It can play a role, perhaps, in picking existing treatments and treatment options and use them more effectively, but the way our society is set up, it’s really hard to have that make commercial and regulatory sense. What I realized is that it’s going to be really hard doing it in a diagnostics way so why not apply human genetics to make new medicines for unmet medical needs. That is what attracted me to this opportunity.
JR: Will Pfizer’s huge size be a problem to operating in this nimble fashion?
Cox: Not at all. We have a tremendous opportunity to show how this can work. In fact, the BBC represents an innovative operating model for Pfizer and the pharmaceutical industry, as it brings the best features of biotech and pharmaceutical scale together in one place for the first time. BBC is partnering closely with PGRD to maximize platform line, therapeutic area and technology expertise to drive the development of high-value compounds across both organizations. And to be able to leverage these resources to be more effective, then you can really feel proud of what you’ve done in life.
JR: Will there be a role for the BBC with the rest of Pfizer?
Cox: Absolutely. There will be lots of interactions. The thing though is what you want to do is figure out which unit or which set of people are ones that have the best expertise and the best shot at carrying something through. The BBC will take compounds through proof of concept. We have our own clinical unit. But for all of the phase three trials and all of what needs to get leveraged from big pharma, that’ll be part of the Pfizer organization. So we leverage the strengths of the big organization but we stay separate enough that we can use a biotech approach for the first parts of research and development.
JR: How important will in silico approaches and expertise be inside the TGU and the BBC broadly?
Cox: I believe, personally, that they’ll be quite important in this focused way that I’ve described. This isn’t my area of expertise. But we have some pretty smart scientists in the group who are really quite skilled at this and I think the key thing here is planting your flag in this one place in biology and then expanding out from that in a hypothesis based way. That’s how I’m hoping we’re going to be able to do this.
JR: You’ve talked of using human genetics; will you also use other ‘omics such as protein-protein interactions in the TGU?
Cox: Sort of. But the problem with those is that they aren’t tied to the fundamental clinical outcomes that we care about. The reason for looking at the DNA variation is because you can associate that in a casual way with the clinical outcomes that you care about. The problem you have is that you don’t have any idea how those variations, from a biological point of view, are leading to the clinical outcomes. So then we use that as the starting point and then we use all the tools of biology we can use to get a better understanding of that.
JR: So you’ll start with a genetic roadmap and turn that into a mechanistic view of what’s going on.
Cox: Absolutely. We believe without understanding the mechanism the chances that things are going to work in humans as a drug are very unlikely. We use that just at the starting point and then we keep pushing at the mechanism until we are confident that we have enough understanding that when we put that in humans, it’s actually going to have the effect we want.
JR: One of the things these small systems biology-oriented companies have found is that reaching no go decisions - basically killing projects - that the reward for that is not very high. People want to find out something that will work.
Cox: Yes, but the motive to kill projects is very high for us because we expect most of our ideas are going to be not very good. It’s not like we’re going to pick ideas we think are stupid; but that we believe no matter how good our ideas are, if we do the right science, most of them are going to go away. That’s just the way science works.
JR: So that suggests two things. You need to work more efficiently so you can go through the ideas more quickly and then when you do come up with an idea, that idea should have a much lower attrition rate.
Cox: That’s exactly right, and as you point out, in silico or system biology can take an idea that seems really good and based on kinetic parameters or just mass action it can tell you shouldn’t even bother to do a single experiment because it’s dead on arrival. I think that’s where the systems biology can be very, very helpful. Lowering clinical attrition is the key. If we can put biotherapeutics into the clinic that have a higher probability of making it to the market to become human drugs, that will be a win for Pfizer, and a big win for human health.
JR: Do you expect the TGU and BBC to grow in staff or stay stable?
Cox: It’s stable and the reason is that we don’t want to get too much bigger. We want to preserve the unique cultures of these sites and maintain their entrepreneurial spirit. The notion is to have groups that you can get everybody together in a small auditorium and talk to everybody without a microphone - they’re all focused on the technology, and it becomes really exciting.
JR: Thank you for your time.
This article first appeared in Bio-IT World’s Predictive Biomedicine newsletter. Click here for a free subscription.