On June 8, 2006, the Food and Drug Administration approved the vaccine Gardasil, the first-ever cancer vaccine that blocks four types of human papilloma virus (HPV). This includes the two that give rise to nearly 75 percent of cervical cancer cases and two additional types that cause about 50 percent of genital warts. For roughly two decades, Richard Schlegel, M.D., Ph.D.—professor and chair of the Department of Pathology at Georgetown University’s Lombardi Comprehensive Cancer Center—conducted collaborative research that has resulted in this major medical breakthrough.
The vaccine technology was generated by a team of Georgetown University researchers—including A. Bennett Jenson, M.D., and Shin-je Ghim, Ph.D., now both at the James Graham Brown Cancer Center in Louisville, Ky.—in the early 1990s and licensed for commercial development.
In August, Schlegel sat down with Frank Reider, editor of Georgetown Medicine, and Bill Cessato, the magazine’s senior writer, to discuss the discovery and its future implications.
Georgetown Medicine: Thank you, Dr. Schlegel, for talking to us today. Tell us a little bit about what your lab is doing.
Richard Schlegel: My laboratory studies the human papilloma viruses (HPV) and their role in cancer. We have been working on HPV for almost 20 years actually—trying to study the complex activities of its oncogenes and how it immortalizes and transforms cells. Additionally, we are interested in studying how to combat these viruses and prevent them from infecting humans.
GM: Why is this important?
RS: Actually, when we first began our studies in 1982, we didn’t think that our experiments would be important clinically. At that point, we were studying a cow papilloma virus. I was intrigued by this cow virus because it gave us a simple system to study cell transformation, specifically how viruses can alter normal cellular behavior and induce tumors. The bovine virus provided an interesting model system because it could transform cells in the laboratory and the viral genome had been cloned and sequenced. This enabled us to express the genes and study their effects on host cells. While we were studying the cow virus, studies from Germany—specifically in Harold zur Hausen’s laboratory— showed that there were newly discovered human papilloma viruses present in cervical cancer. This was the first hint that the virus might actually have an etiological role in cervical cancer.
We therefore shifted our attention away from studying the bovine virus and began investigations into the role of HPV in cervical cancer. The transition from the bovine virus to the human virus was fairly straightforward because their genomic organizations were similar and the structure of the two virus particles was similar when examined under the electron microscope.
GM: Why is the work that your lab is doing now significant and important?
RS: I think the greatest impact of our research right now is the clinical translation of our laboratory findings. There are about a half a million women every year who develop cervical cancer, and about a quarter of a million women die from the disease every year.
We’re lucky in this country to have an affordable, effective, and widely used screen for early detection of cervical cancer—the PAP smear. Early detection allows clinicians in the United States to intervene early, before the disease progresses to late-stage cancer. However, in developing countries, the lack of sufficient health care access results in many women progressing to late-stage carcinoma.
The work that we’ve done in the last 20 years has led to the development of a vaccine that allows us to prevent women from being infected with HPV. Initial vaccine trials have proven extremely efficacious and safe, and we believe that the widespread use of the vaccine will have a positive impact on the health of women globally.
GM: You are also working on the next generations of this vaccine, correct?
RS: Yes. The first-generation vaccine against HPV is based upon technology developed at Georgetown in about 1992. We discovered that we could express the capsid protein of an HPV in the laboratory in an antigenic form that mimicked the intact virus. When over-expressed in either yeast or insect cells, the capsid protein was later observed to form virus-like particles, or VLPs, which look very similar to the “real” virus. This is the basis for Merck’s vaccine, Gardasil, as well as GlaxoSmithKline’s vaccine, Cervarix.
There are two types of HPV which cause approximately 70-75 percent of cervical cancers, types 16 and 18. The commercial vaccines are type-specific, and in order to protect against these two types, the vaccines contain a mixture of VLPs. These vaccines are both very successful, but they can be quite expensive (about $360) and appear to require refrigeration, which could limit their distribution to developing countries.
In order to develop a cheaper and more stable form of the vaccine, we have been collaborating with Bob Garcea at the University of Colorado. Bob had already shown that papilloma virus capsid protein could be expressed in bacteria, using a molecular method to “fuse” the capsid protein to GST. This fusion protein remained soluble in the cell and made purification much easier.
We then tested the purified fusion protein in our animal model and found it to be as effective as the VLP formulations. In addition, the purified vaccine can be dried down into a powder and reconstituted with water without loss of activity. We also discovered that the fusion protein was not forming capsids but, instead, capsomeres. A capsomere is comprised of five L1 molecules, and 72 capsomers make up the capsid. What we have produced in bacteria, therefore, are capsomeres that are more stable and retain all the immunogenicity of the entire capsid. These are the critical aspects of the second-generation vaccine.
GM: How about the third generation?
RS: The third-generation vaccine that we are working on is in collaboration with Bob Garcea; Lutz Gissmann in Heidelberg, Germany (who helped discover the association of papilloma viruses and cervical cancer); and Luisa Villa in Sao Paulo, Brazil, where there is a facility to produce and test the candidate proteins. The third-generation vaccine is a combined prophylactic-therapeutic vaccine.
GM: Thinking of the development of your vaccine as a timeline, what are the major milestones that you can identify for us?
RS: In about 1988, while I was at the NIH, I had initiated some collaborative studies with Bennett Jenson, who at that time was in the Department of Pathology at Georgetown. We were considering methods we could use to investigate the immunogenicity of the capsid proteins and were specifically interested in defining the types of neutralizing epitopes, linear or conformational, that were present on the virus.
These initial interactions initiated our collaborations to develop a vaccine against HPV. Our research efforts really hit full steam when I accepted a position in the Department of Pathology at Georgetown University in 1990. Bennett Jenson and I had adjacent laboratories and began sharing ideas on a daily basis as to where we wanted to go with the project. Over the course of several months of research, we came to the realization that a conformational epitope, the shape of the capsid protein, was the critical key and the correct avenue to pursue in making an effective vaccine. The recently approved Gardasil vaccine blocks four types of HPV, including two that give rise to nearly 75 percent of cervical cancer cases and two other types that cause about 50 percent of genital warts. Over the course of several months of research, we came to the realization that conformational epitopes, the shape of the capsid protein, was the critical key and the correct avenue to pursue in making an effective vaccine.
GM: How did you stumble onto that idea?
RS: The idea actually derived from early studies in Bennett’s laboratory. Our first hypothesis was that linear epitopes might be protective. Linear epitopes are short, contiguous sequences of amino acids in the protein. After several unsuccessful attempts with this model, we realized that shape was important. This redirected our thinking and led us to investigate methods that would allow us to make this protein in the conformationally correct form. That is when we began to think about different ways of expressing the protein correctly.
I would say another critical turning point came in about 1992. That was probably one of the “eureka” moments when we were able to show, in eukaryotic cells, that we could produce the protein and it was in the correct shape. Bennett and Shin-je can probably recall this experiment as well as I can. The three of us looked at cells overexpressing the L1 protein under the microscope. Shin-je had stained the cells with an antibody that recognizes conformationally correct L1, and we could visualize brightly immunofluorescent nuclei of cells stained with this antibody. We knew at that point that it was possible to produce the L1 protein with the correct shape, and this indicated to us that L1 could serve as a prophylactic vaccine in humans.
I would say the next step came in late 1995, when we published a paper in the Proceedings of the National Academy of Sciences. The publication was the culmination of several experiments in which we showed that immunizing dogs with the L1 protein protected them against subsequent challenge from the virus. That was the first time anyone showed that we could take the vaccine, immunize animals, and then we could challenge them with high-dose virus and they were completely protected against a mucosal papilloma virus (like those in cervical cancer). And the protection was 100 percent. It wasn’t 70 percent or 80 percent. This proof of principle, showing that the vaccine worked in vivo, was a major hurdle.
The next major step came when we contacted and turned the technology over to Medimmune, a biotechnology company in Gaithersburg, Md. Scientists at Medimmune utilized our expression vectors for HPV L1 and their huge fermentation capability to generate enough protein to initiate clinical trials.
The next step after that, I think, would be the sublicensing of the technology from Medimmune to Glaxo- SmithKline (GSK). GSK had the interest, expertise, and financial resources to carry out the very expensive phase II and phase III clinical trials. As you know, phase II and phase III clinical trails can take years and present major hurdles in the development of a vaccine or therapeutic drug. Any toxicity or adverse event has the potential to end the program. Fortunately, there weren’t any unexpected hurdles along the way.
GM: Help us to understand what it takes to make a vaccine, including the path that you took and some of the scientific hurdles you had to overcome.
RS: There are several hurdles or milestones. One of them, I think, was made early on when we discovered you could express a single protein, in this case L1 (there are two capsid proteins in the virus, L1 and L2), and generate antibodies against that protein that were capable of recognizing the intact virus and neutralizing it. This discovery was made in about 1992, and was a tremendous break-through. Since there are two proteins which comprise the capsid, L1 accounting for 95 percent of the protein in the capsid, we were uncertain if we would have to express both proteins. If that was the case, it would increase the complexity of the studies dramatically. Based on abundance, we went after L1, thinking, “Well, if that is most of it, maybe that is the most promising target.” As it turned out, L1 was the protein that was the real determinant of immunogenicity. After that, it was really trying to figure out how to make the protein most effectively in the right shape, or conformation.
In the beginning, we were trying to make the protein in bacteria. We thought we would be quick about it. However, it just didn’t work and we thought, “This is a mess. Perhaps we need to make the vaccine in mammalian cells or, at least, a eukaryotic cell.”
We switched to mammalian cells that have nuclei (unlike the bacteria). We thought the protein might be processed differently. We did something else differently, too. We didn’t make a fusion protein; we made L1 alone. This proved successful and was the genesis of developing the HPV vaccine. Both Merck and GlaxoSmithKline used eukaryotic cells. One takes advantage of expression in yeast, and the other uses insect cells that they infect with a baculovirus carrying the L1 gene. Initially, we thought the presence of the nucleus must be the trick in getting expression of a protein capable of being recognized by conformationally correct antibodies.
But then Bob Garcea, who was at Harvard at the time, discovered he could make a small amount of L1 protein in bacteria. He mentioned, “It was a real pain to purify the protein, but it was possible.” Even though it was difficult, it did appear that we would be able to use bacteria to produce L1 and purify it enough where it assembled into a capsid structure. This advancement was after our initial attempts in bacteria and when technologies had improved with regards to expression and purification. Because of this, we returned to experiments with bacteria to synthesize the second-generation vaccine. We have essentially come full circle, starting with failed attempts at expression in bacteria, switching to the eukaryotic system, and then coming to realize that we could actually use bacteria again, in the end. The second- and third-generation vaccines are based on expression in bacteria.
GM: Explain what the vaccine does.
RS: How does the vaccine generate protective immunity? There is an interesting story in the answer. Papilloma viruses involved in cervical cancer infect the mucosal cervical epithelium. They are termed mucosal papilloma virus infections, as opposed to skin infections, or epidermal infections. For many years, people thought that protection against infection by mucosal viruses was mediated by a specific type of antibody called IgA. You could find the IgA antibody in saliva, vaginal secretions, and the lungs. The circulating antibodies in your blood are IgG. What we found very early on are two important things: immunity was mediated by antibodies, and secondly, it wasn’t by IgA, but protection was provided by IgG. IgG was the main determinant, and if you wanted to protect an animal, that was the antibody that you had to induce.
I think that the critical experiment in the animal studies was to immunize one set of dogs, allow them to generate antibodies in their circulating blood, and then isolate those antibodies from the serum, purify them, and inject them into dogs that never saw the vaccine. They now had circulating antibodies against the virus. We challenged the second set of dogs on their mucosa, and it turns out that they were completely protected.
I should mention that there appear to be two different mechanisms for antibodies to “inactivate” a virus. One way is that the antibody prevents the virus from binding to the cell. It sort of wraps up the shell of the virus, and it can’t interact anymore with the host cell. The amazing thing is you can also protect cells even after they have bound to the virus. You can add the antibody at a later time, and they are still protected from infection, even though the virus has attached to the cell surface. We don’t fully understand the mechanism, but we think that it prevents some later step—entry of the virus into the cell. Therefore, the antibodies have two discrete steps at which they neutralize the virus.
GM: What technology emerged along the way that was important in advancing your work?
RS: One of the major hurdles for us came in 1992, when we were trying to make the L1 protein in the right shape and in sufficient quantity to work with. We had previously been working for years studying a papilloma virus oncoprotein called E5. That was a real bear to work with. It was very difficult to express in mammalian cells. We had developed a technique to express the E5 protein utilizing an SV40 virus vector and promoter. The technology is not only based on a strong promoter, but it is also based on the fact that the DNA is amplified hundreds or thousands of times when transfected into the mammalian cells. The cells contain a protein which amplifies the plasmid DNA, increasing the copy number of genes you introduce into the cell, such as E5. This system allowed us to obtain high levels of expression of E5 in eukaryotic cells, and we had a lot of experience with the system. After coming to Georgetown, Bennett and I were having one of our many discussions and we asked, “How can we express this protein, and how can we get high levels expressed in eukaryotic cells?”
Obviously, the first thing that rang in my mind was to use the expression system we had developed. We decided to clone the L1 gene into the SV40 vector and transfect it into mammalian cells to determine if we could achieve high levels of expression. It worked on the first shot.
That was, like I said, a momentous occasion for us because we realized we would be able to make the L1 protein.
GM: What led you to study HPV?
RS: I was doing a combined pathology residency and post-doctorate fellowship at Harvard. In the Pathology Department, I worked with Tom Benjamin, who led a very exciting research program and a lab that was filled with energetic, interesting scientists. Tom’s research interest was in the mechanism of cell transformation by a mouse transforming virus called polyoma virus. This viral system was ideal because it allowed investigators to study viral-mediated transformation in the laboratory. Although we didn’t know it at the time, polyoma virus was in the same family of viruses as the papilloma viruses. It is a little smaller, but it is the same shape, has the same genomic organization, and has a double-stranded genome.
I decided at that point that I wanted to work on polyoma virus and did so for the next two to three years. I then came to NIH and worked in the lab of Ira Pastan, who helped guide me in studies of virus binding and uptake. Peter Howley, an NIH colleague of Ira’s, came to me one day and said, “You know, there is a position available in the Pathology Department in NCI. Would you be interested in transferring over and working on papilloma viruses?” My first thought was: “Papilloma viruses—what do they do?”
I later realized they were very much like polyoma viruses. They had a very similar genomic organization and they could transform cells in vitro. I saw enormous potential for studying this at the molecular level, and I transferred to Peter’s lab in 1982. I was working in Peter’s lab on the cow papilloma virus when, as I mentioned earlier, zur Hausen isolated human papilloma virus from cervical cancer specimens. That is when my focus turned to studying HPV.
GM: Has it been in any way advantageous for you and your work to be here at Georgetown?
RS: I wouldn’t be involved with the vaccine if I hadn’t moved down to Georgetown in 1990. Georgetown actually had a couple of well-established papilloma virus laboratories. Wayne Lancaster’s laboratory and Bennett Jenson’s laboratory were working on papilloma viruses before I came here. There was a rich history of studying papilloma viruses, and I thought the opportunity for collaboration was immeasurable.
Unfortunately, Wayne’s lab left before I came to Georgetown. If it really weren’t for Bennett being at Georgetown, and sort of pulling me—screaming and kicking—into the field of immunology, I wouldn’t have been doing it on my own. At that time, I was very interested in transforming genes and cervical cancer development. Bennett was interested in host-immune responses to the virus. I think it was the fusion of Bennett’s lab with my lab that generated the ideas and concepts that were critical for the development of the vaccine.
GM: Do you think that there are things that you stumbled onto or developed that have an influence and opened some scientific doors elsewhere?
RS: The idea of being able to express a single protein and have it assemble into this virus-like particle was a big step for us. I think a lot of people are using this information to modify that structure— similar to what we are doing for the therapeutic vaccine. There are several laboratories trying to adapt these virus-like particles in order to provide therapeutic protection as well as prophylactic protection.
Another concept, not developed by us, is the ability to use these virus-like particles as a means to transfer DNA into cells. This concept is not at the translational level yet, but in terms of molecular biology, I think that we are going to learn a lot in the next three to five years in being able to transfer viral and cellular genes using these virus-like particles.
I also think that the canine model for papilloma virus infection has given us an extraordinary system for comparing the in vitro and in vivo activities of the viral genes. In addition, the dog model should enhance our ability to develop improved vaccines. There are over 100 papilloma viruses that can infect humans. Unfortunately, each type is unique, even though they may have the same pathology. If you make antibodies against type 16 and type 18—the most important ones for cervical cancer— you don’t protect against type 1 or type 2, which induce common warts. There’s little or no cross-reactivity or crossprotectivity.
What we would like to do is to be able to generate an agent which is crossprotective. The dog model allows us to perform these experiments because there are a couple of different types of canine papilloma viruses. In addition, the dog model has allowed us to study a therapeutic drug we have identified, the Chinese herb called artemisinin. This compound was used by the Chinese 2,000 years ago to treat certain intestinal diseases. Currently, artemisinin is the standard course of treatment for malaria infections which have developed resistance to chloroquine.
Amazingly, the anti-malarial artemisinin drug is fantastically effective at killing cervical cancer cells but not normal cervical cells. We actually have a new NIH grant using the dog model to develop treatment protocols with that Chinese herb. You might say, “Why do you need a therapy if you are developing a vaccine?”
The bottom line is that the current vaccines—either from Merck or GlaxoSmithKline—are targeting only two oncogenic viruses, 16 and 18, which cause maybe 70 or 75 percent of human cancers (cervical cancers). What happens to the other 30 percent of the people with early cervical cancer? Do you just say, “Well, sorry, we cannot help you”? There are certainly existing surgical methods to treat these early cervical lesions, but our work with artemisinin suggests that there might be a simple, safe method to rapidly and cheaply treat these HPV infections which could have a great impact for developing countries. Actually, experimental work with artemisinin has been very fruitful and Georgetown has already “spun off ” a biotechnology company called Franz Viral Therapeutics to further develop the clinical potential of this compound.
GM: How could an interested philanthropist help you with your work?
RS: I think we have to continue the crusade to develop cheaper, more stable, and more cross-protective vaccines. There are 12 to 15 viruses that cause cervical cancer in humans, and we are targeting only two of them. Realistically, pharmaceutical companies won’t be able to make all 15 types and mix them together. We would be paying $3,000, rather than $300, for vaccination procedures. In addition, we need to determine if we can make the papilloma virus vaccines therapeutic as well as prophylactic. These are long-range studies that require very significant investments. However, the impact globally is enormous.
GM: What would a gift provide? Would it give you more people? More infrastructure?
RS: Dog studies are not cheap. Each dog is approximately $500, and dog trials can last for several months and cost in excess of $50,000. We try to adopt out all the dogs that we work with. Currently, we only have space for 30 animals. This limits our ability to have multiple studies running at the same time. With the shrinking pool of NIH dollars, competition for the limited resources is tough. Gifts to the university would allow our group to expand and speed up our development of new vaccines and therapeutics.
GM: Does a lot go into this in the way of training and education, as well as the actual work?
RS: We spend significant effort training graduate students, as well as undergraduate students, in our laboratory. They play an important, integral role in moving our research programs forward, and during the process, they learn to become independent, creative scientists. I would not be here if it were not for someone mentoring me as a graduate student. I see it as an obligation and a joy to teach the undergraduate and graduate students at Georgetown. We are paving the way for the next series of medical breakthroughs.
As a sidelight, I believe that Georgetown has a unique infrastructure that is not present at most other medical schools—the presence of an adjacent undergraduate campus. This facilitates undergraduate training in medical sciences, as well as interactions between basic sciences and medical sciences. It is a real highlight of our research program.
GM: You often hear the expression “from bench to bedside.” What is it like to be someone who has actually seen a project progress from the lab to a vaccine?
RS: When I began working on papilloma viruses, I was interested in the bovine virus and cell transformation. I really wasn’t thinking about translational research at that time. It wasn’t until Harald zur Hausen had discovered HPV in cervical cancer that we began thinking, “Gee, if we could make a vaccine against this virus, we could prevent a worldwide health problem.”
In the beginning, it was all very cerebral and hypothetical. Every medical scientist has wishful thoughts like, “Wouldn’t it be great if I could do something that had a major impact on human health?” We realize, however, that this isn’t the usual outcome. It took 10 years of research to get us to the point in 1992, when Shin-je, Bennett, and I realized that this really had the possibility to work. The entire journey had moments of great excitement and great doubt. We could get expression in cells, but would it work in animals? We had positive results in the animals, but would it work in humans? Then when you see it actually work in human clinical trials with no toxicity and 100 percent protection—it was a dream realized.
About Georgetown University Medical Center
Georgetown University Medical Center is an internationally recognized academic medical center with a three-part mission of research, teaching and patient care (through our partnership with MedStar Health). Our mission is carried out with a strong emphasis on public service and a dedication to the Catholic, Jesuit principle of cura personalis -- or "care of the whole person." The Medical Center includes the School of Medicine and the School of Nursing and Health Studies, both nationally ranked, the world-renowned Lombardi Comprehensive Cancer Center and the Biomedical Graduate Research Organization (BGRO).