Drive to Discover:
An Interview with Peter Vogt
Professor Peter Vogt is head of the Division of Oncovirology in the Department of Molecular and Experimental Medicine at The Scripps Research Institute, where he has been conducting research since 1993. Prior to this time, he was the Hastings Distinguished Professor and Chairman of the Department of Microbiology at the University of Southern California School of Medicine. He also held faculty positions at the medical schools of the University of Washington and the University of Colorado.
Over his career, Vogt has focused on virology, genetics, cell biology, and cancer. Vogt made the fundamental discoveries that led to the identification of oncogenes, which control cell growth and division and lead to carcinogenesis—the process whereby benign cells in the body are transformed into cancerous ones. His work has been recognized by a number of prestigious awards, including election to the Institute of Medicine and the National Academies of Sciences.
Currently, Vogt and his laboratory focus on the molecular mechanisms of carcinogenesis. News&Views spoke with Vogt recently about topics including his entry into the field of cancer research, the evolution of this field, and aspects of cancer research today.
News&Views: When you first started in this field in the 1950s, how much of what we know today about carcinogenesis was known then? Did we know 10 percent or 50 percent? Did we understand the broad outlines?
Peter Vogt: It’s hard to quantify how much we really knew. I would say we knew little. And we knew little about molecular events in carcinogenesis. We had an idea of general changes in phenotypes and cell behavior, but we didn’t know the underlying mechanisms. It was an early stage of the field’s development, [and the field] was still in its infancy.
This was an exciting time in general for biology because of Watson and Crick’s discovery of the structure of DNA. How did this influence how scientists thought about viruses?
Watson and Crick had a tremendous impact. One of the early consequences... was that nucleic acids in general [came to be] regarded as the depositories or sources of genetic information. Maybe the most striking discovery in that respect was a virus-related finding. That’s the discovery that, if isolated in pure form, the genome of a virus—tobacco mosaic virus, which happens to be a single molecule of RNA—is infectious. It can make a completely new virus. And that was a wonderful discovery. It was made about the same time in Berkeley and in Tübingen. Tübingenwas where I got my Ph.D., and Berkeley was where I went for my post-doc. So, that discovery was at the beginning. It demonstrated very dramatically that nucleic acids are genetic material.
Did it influence your decision to go to Berkeley?
My decision to go to Berkeley was mainly prompted by the [pioneering] discovery of in vitro oncogenic transformation—virus-induced oncogenic transformation—at Caltech by Harry Rubin and Howard Temin. Harry Rubin then moved to take a faculty position at Berkeley. I had read Rubin’s papers, and I thought, “This is a beginning of a new field, the beginning of a new, molecular understanding of oncogenesis.” I wanted to work with him.
So moving into the '60s and '70s, with the advent of molecular biology, how rapidly did the understanding of cancer viruses progress?
I would say fairly rapidly. Howard Temin, on the basis of observations made with inhibitors of DNA synthesis, postulated that the viral genome becomes integrated into the cellular genome, and that this was the transforming event.
At the time, nobody believed him, and the evidence for the hypothesis was quite weak.
The whole question of how RNA tumor viruses, which are referred to as retroviruses, replicate remained open, and it was a big puzzle. Nobody knew how these viruses could generate progeny, why they are susceptible or sensitive to inhibitors of DNA synthesis. There was a big discussion in the field. [Some people] were for the pro-virus hypothesis and some people were against it. And that whole thing was only resolved with the discovery of reverse transcriptase.
By David Baltimore?
David Baltimore and Howard Temin. They discovered it at the same time. I had done some experiments, which sort of hinted in this direction, but didn’t go far enough. Together with [Peter] Duesberg, I showed that this requirement for DNA synthesis that was puzzling everybody was not a requirement for cellular DNA synthesis but for a DNA that was virus-specific, so it had to be viral DNA synthesis. I remember in our paper we speculated that if it’s truly viral, there is no enzyme in the cell that could make DNA from RNA.
We didn’t think far enough to say that that enzyme must be a viral enzyme.
How did that change our understanding?
That clearly showed that the viral genome was integrated.
And here you have a viral genome that becomes part of the cellular genome, so how does it transform the cell? That’s where oncogenes came into play. The existence of oncogenes was first suggested by viral mutants. We isolated the first temperature-sensitive mutants that later turned out to be mutants in this Src gene that made the virus temperature-sensitive—not in its ability to replicate, but in its ability to transform cells to induce oncogenic transformation. That showed clearly that it was viral genetic information that caused the tumor.
With the advent of molecular techniques that allowed analysis of large RNA molecules, I showed, in collaboration with Peter Dusberg, that you could actually take part of the viral-genome off, you could delete it, and still retain a virus that was perfectly viable. But that virus no longer caused tumors, no longer transformed.
So the suggestion was made that what we had deleted was actually tumor-inducing information. And that was clear evidence of the first oncogene in the cell. About 20 percent of the genome was deleted in these viruses. The virus was still viable, but it didn’t cause any tumors anymore.
Later, when the viral genetic map was worked out, it was found that this was an extra gene, not required for replication but required for transformation.
The next big discovery was by [Michael] Bishop and [Harold] Varmus—I participated in that as well—who showed that that particular piece of viral genome deleted in these non-transforming viruses is actually a cellular gene. It’s not a viral gene. It’s a cellular gene picked up by the virus.
In a way, that discovery took cancer research out of virology because it showed that all that viruses were needed for was to activate the gene and carry it from one cell to another. But, the gene itself and its function were both cellular. So looking at the cellular genes that could cause tumors doesn’t really require a virus. If you could identify other genes that upon activation were oncogenic, then you would have [identified] another oncogene. You didn’t need a virus.
It’s true that for a while, viruses were still used to discover new oncogenes. Whenever you isolated a new retrovirus, you had a chance of finding a new oncogene. But eventually that approach became exhausted, and more oncogenes were found by directly analyzing tumors and cellular genes.
At what point does a cell become a cancer cell? In other words, cancers in general are an up regulation of the things that cause it to grow or the down regulation of things that slow the growth. But, there are many different types of cancer cells in different tissues and so forth. So, is there a sense in the field that you can have a cell that is almost a cancer cell?
Well, there is a sense in the field of a number of criteria that need to be fulfilled for a cell to be called a cancer cell. There are a couple of classic papers in the field that define what a cancer cell is, but the general consensus is that a cancer cell does not develop as a result of a single genetic change. You need several genetic changes that interact, accumulate, and make the cancer cell.
Cancer is a process that goes through a number of consecutive genetic changes that either occur in a timeline or occur all at once and are selected for. But, you need more than one change. A single change [may] give you some growth advantage and take you out of certain regulatory circuits, but it doesn’t make a cancer cell.
Malignant neoplasms are currently the second leading cause of death in the United States, so obviously cancer is still a very big problem, and we have a long way to go in terms of treatment. But how much further do you think we have to go in the understanding of cancer and genes that cause cancer?
If you look at statistics, in the past few years there has been a slight reduction in mortality from cancer. But we have a LONG way to go. And we also have a long way to go in understanding. We know about oncogenes. We know about tumor-suppressor genes. We still don’t know how they work together. There are still aspects of cancer that we have almost no understanding of as yet.
For example, the possible role of small and non-cording RNAs in the control of gene expression and in the control of cancer. So microRNAs are a big surprise as players in oncogenicity, and they remain almost completely unexplored.
They are non-coding RNAs. They are regulatory RNAs, very small RNAs. They were first discovered as regulators of gene expression in C. elegans, but now have been found in mammals, and there is some evidence that they are involved in tumor genesis. There are papers that suggest they actually define high breakpoint regions in the genome that are important in chromosomal rearrangements in leukemia. So, there is still a lot of work to be done to show that these microRNAs are really crucial, but there are strong suggestions they might be important. It is a completely new area.
Given that, is it possible that there are other whole areas of control we are not aware of?
Non-coding RNAs is an upcoming area of study, as is glycobiology. We know far too little of the importance of carbohydrates as surface molecules and surface markers, both as antigen markers and in mediating interactions between cells.
And presumably, post-translational modification in general.
Post-translational modification in general. Phosphorylation has been studied for quite awhile now, but glycosylation is very incompletely understood. And there are lots of other post-translational modifications coming up: methylation, acetylation. All of these could be important.
Can you explain in general terms how we go from understanding or identifying a gene to some sort of therapy?
One of the benefits that could derive from studying oncogenes and tumor suppressors is to find normal targets for small molecule interactors—new drug targets. And these targets are not only the oncogenes, but also the entire signaling chains and signaling cascades connected to these oncogenes.
Target verification is an important aspect of cancer research today. The most simple and straightforward approach to finding interactors is to look for inhibitors of enzymatic activity. Traditional drug design follows that strategy. A more daring strategy that still needs broad confirmation is to look at protein-protein interactions and try to affect those with small molecules. There have been some success stories, but also suggestions it can’t be done. But if we successfully developed an approach to modify protein-protein interactions, we could increase the number of specific targets many fold.
The other exciting development in that area originated here at Scripps [in the laboratory of K. Barry Sharpless], and that’s the principle of using bivalent ligands put together with “Click Chemistry.” These promise to have a high degree of specificity and increased potency.
What do you think are the most promising new tools in molecular biology? Genome-wide scans, microarrays...
Well, genome-wide scans are starting to become useful. The problem there is to design scans that are linked to function. If that is successful, and in a few cases it has been, then you get really interesting information. In a simple fashion, you can ask for genome-wide surveys of proteins or clones that have a certain phenotype—either transformation or some changes in morphology or changes in the localization of a protein. Those tend to be very instructive. So, if you can link function to genome-wide survey, you usually get some very interesting information. That’s where the field is moving.
Microarrays, just as plain microarrays initially were exciting, but they tend not to give you enough information on function. So you end up with lists of genes up or down regulated, and then you can rationalize these lists by saying, “Ok, these are all related functions.” But you have to verify function, and that is decidedly not that high-throughput activity. It takes a lot of manpower to do that. So this is still a big technological challenge to do functional genomics in a high-throughput fashion.
The other things that have had a tremendous effect on cancer research are knock-out technology, RNAi technology, and the ability to develop mouse strains where you have conditional knockouts of one or several genes. You can show the interaction of these genes by crossing the lines and putting knockouts together. That has been incredibly informative and spectacularly successful. It has shown us how oncogenes and tumor-suppressor genes interact. It has been the prime tool for designing model systems, murine models of human tumors that then can be used to look for specific anti-cancer drugs.
All right. If you look towards the future, what do you see on the horizon?
It’s difficult for me to say anything about this. You know, it’s the old [Yogi Berra] saying, “It’s difficult to make predictions, especially of the future.”
I would expect targeted therapy to be very successful, both for individuals and also to [find] specific new targets derived from our knowledge of oncogenes and tumor-suppressor genes.
Anything else that you want to say about your research?
No... except that I love doing it.
I wanted to end by asking you about something else you love—painting… Do you see painting as a refuge from your daily work?
For me, my own painting is a response to visual stimuli I enjoy. I want to transform them into something durable [that others may find] enjoyable.
It is not a refuge; it’s a different way of discovering. The basic drive is to discover.
Send comments to: firstname.lastname@example.org