At A Glance
Name: Philip Felgner
Position: Professor, Center for Virus Research, University of California, Irvine, since 2002
Chief scientific officer and founder, Gene Therapy Systems, since 1998.
Background: Chief scientist and various other positions, Vical, 1988-98.
Staff scientist, Syntex Research, 1982-88.
Post-doc in biophysics, University of Virginia, 1978-82.
PhD, biochemistry and neuroscience, Michigan State University, 1978.
BS, biochemistry, Michigan State, 1972.
How did you first get involved with proteomics?
I developed a DNA vaccine technology at abiotech company in San Diego called Vical. We discovered that you can inject naked DNA into mouse muscle and the gene product would be expressed there. And when the gene product gets expressed, if it’s a gene encoded by a microorganism, then the animal gets an immune response. This is an interesting alternative way for mounting an immune response. After we made that discovery and started thinking about other organisms that we might want to produce an immune response against, we had another problem: Some of these organisms had many genes. For instance, I was working with a group from the Navy on the malaria vaccine. They were using DNA vaccination in humans, and they found that they could produce immune responses similar to the ones we were seeing in animals, but they would try these vaccines out with single antigens from malaria, and they wouldn’t get sufficient protection to protect against the actual challenge. So we concluded that what’s needed is a cocktail of antigens — probably 10 or 20. But the organism encodes 5000 potential antigens, so the question is, how do you go about deciding among 5,000 antigens which ones to choose?
My colleague Denise Doolan at the Navy had access to patient populations in Africa — very well-characterized groups of individuals who had been previously infected with malaria. One group gets exposed to malaria once in their lifetime, gets the infection and gets the disease, and then they recover, and then following that, they never come down with a new infection. So they’ve been immunized. But then there’s another group that consistently comes down repeatedly with a new infection. And so what we were interested in then was in comparing the immune responses of these two groups. We thought that if we could comprehensively monitor, antigen by antigen, the immune response against each antigen in this organism, and compare those responses — the group that’s -protected after infection with the group that’s not protected after it’s infected — then we find out what’s present in those people who are protected that’s lacking in the people who are not protected. Those antigens are the ones that we would want to use in this DNA vaccine.
So that’s when you came up with the idea of using the array?
Right. That was the conceptual thing, and then it became a technical problem: How does one go about making all of the individual antigens from an organism that has 5,000 antigens? So the first thing we did was develop a technology called transcriptionally active PCR. That’s a patented approach for making PCR fragments transcriptionally active that we published in the Journal of Biological Chemistry. It’s not hard to make thousands of different PCR — you can buy thousands of primers, and put them in a PCR machine. One of the standard PCR machines that a lot of people buy is called Tetrad, by MJ Research, and that thing has space for four 96-well plates. So that gives you about 400 reactions in one run in the morning, and then you can do another run in the afternoon, which would give you 800, and a few days of that and you’d have 5000. So that makes it very feasible if you have a way of using the PCR machine to make transcriptionally-active PCR fragments.
How do you make PCR fragments transcriptionally active?
You run the PCR reaction in the presence of a promoter and a terminator fragment, and if you set it up properly so everything overlaps in the right way, you end up with a gene that has a promoter and a terminator appended to it. So that was the first thing we developed, around 2000. Then, as we kept playing around with these things, we came up with a way of doing recombination cloning. So actually we have two ways of doing it: We can make a PCR fragment that will insert itself into a plasmid, and you can do that in a real high-throughput way, in groups of hundreds at a time. Normally people do cloning just one or a few genes at a time. You make different PCR fragments encoding each gene, then you mix it up with a fragment of a plasmid, and then the bacteria do the recombination and put the gene into the plasmid. It’s called in vivo recombination cloning. That allows us to get plasmids almost as fast as we get these PCR fragments, and actually we prefer the plasmid approach now. Because what we can do is get these plasmids and sequence them, and make sure we have the right gene in there. Then we can inventory the plasmids.
After that, you have to do one more step, express the protein in a cell-free in vitro transcription/translation system.
So how long would it take to produce all the proteins in malaria in this way?
If that was the goal, I’d say we could do 500 a week, so we’d get all 5,000 done in 10 weeks. But we’re not doing it that way. When we get a few hundred of them, we get the proteins and then we start putting them down on chips, and then we start looking at the immune responses. Because there’s other technical things that need to be handled — we need to be able to get those proteins and then use them for monitoring immune responses. And so those are some of the things we’re doing now.
So you’re basically putting the proteins on chips and running serum over them, and then taking out the ones that have a strong response?
So what stage are you at now?
We’re printing chips now. We’re working out the methods for using those chips to monitor antibody titers in mice and in humans.
We’ve put a lot of emphasis on smallpox. On campus here, there’s a center called the Center for Virus Research. They have a lot of experience with vaccinia virus, which is the smallpox vaccine. So we can do all the things we need here with vaccinia. We can infect animals with vaccinia virus, and we can infect animals with the live virus, and then look and see what the immune responses are that they have against the virus — which proteins do the mice react against. We can use our system to characterize that. And then we also have a lot of human samples. We published a paper in the Journal of Immunology, and we showed that you can take blood from people who have been vaccinated up to 60 years ago, and get this time course. We have a 60-year time course following the immune response of individuals who have been vaccinated with vaccinia smallpox vaccine. Then we have people who have been vaccinated recently, and we have naïve people who have never been vaccinated. Because some people who work in the lab need to get this vaccine before they start working. So we collect their blood before and after they’ve been vaccin-ated, and we get the early time points. This system allows us to do all the things we would need to do in animals, including doing challenge studies and protection studies in mice, and it also allows us to look at human serum from people who have been vaccinated with the vaccine that everybody agrees is actually effective against smallpox.
So for people who have been vaccinated a long time ago, how does their antigen response change?
What we’ve found is that the immune response is high a few weeks or months after people have been immunized. That immune response declines to about 10 percent the level that it was at the start, and stays constant for the whole 60-year period of the study. So you can measure these anti-vaccinia virus immune responses 60 years later. When we look at individual proteins from the virus, what we see is there is an anamnestic response. That means the immune response is weak against the individual proteins, but if somebody gets a boost of this vaccine and you look three weeks later, the antibody response is way up against the antigens. If you do the same thing with a person who is naïve and you measure their immune response 3 weeks after they’ve been immunized, their immune response is really low. It takes a naïve person a lot longer to build up immune response in response to the virus than somebody who had been previously vaccinated, even if that person had been vaccinated 40 years ago. You can prime the immune response, and the immune response remembers that prime.
So the individual proteins all retain a low level of response over 60 years?
Yeah, that’s right. So we’re looking at all the individual responses against each individual protein, and what we want to do now is decide which of the 200 or so proteins that are made by the vaccinia virus we want to use in a subunit vaccine against smallpox.
So the ultimate idea is to take the most reactive antigens and make a safer vaccine out of those?
Those would be harmless, the proteins. Whereas the current, old-fashioned vaccine is really very toxic. We see those people, and those reactions that they get, many are really nasty and painful. It was something people were used to, but nowadays, our health system wouldn’t allow new products out with that kind of a toxicity profile. So anything that would be developed today would have to be much more benign.
Tell me a little about what you’re working on with your partners in England.
We’re working with them on Francisella tularensis, in collaboration with Rick Titball’s group at the Defense Science and Technology Laboratories in Proton Down England.
We have a grant now to do what I’ve been describing to you — not on malaria, but for Francisella tularensis. It’s a bacteria with 2,000 genes. So we go through this process I described, make all the genes and proteins, and use the proteins to monitor immune responses from infected animals, and also from infected people. The animal studies require a special facility because [F. tularensis] is so infectious. In Porton Down, they have one of the most advanced centers for being able to do animal work on really dangerous organisms. Things we can do here we’ll do, but things that require work with live organisms, go on there.
Is your work being funded as part of an RCE? (see PM 8-1-03, 9-12-03)
It’s not from an RCE. But it is under the NIH biodefense program. We have several other projects that are still pending with the NIH — a smallpox project, and we have a plague project with Porton Down. Then there’s another organism called Burkholderia pseudomallei — it’s like F. tularensis actually — it’s an intracellular gram negative bacteria. A lot of these dangerous ones seem to be intracellular organisms.