1.6 Harold Varmus — Animal Retroviruses and Cancer Research

Harold Varmus: [00:00:00] Well, it's a pleasure to begin the second half of what I call the Antiquities session. We're old, the information's old, the audience is partly old. I have a special word for John Coffin thanking him for setting me up, saving me some time. We negotiated ahead of time about how we were going to do this. I'm going to spend [00:00:30] most of my time talking about viral integration, why it's important, how it links to the ideas that Howard Temin (1934–1994) first espoused. I want to explain something about how it occurs and then talk about the relationship of the integration events to the oncogenic potential of retroviruses and conclude with some words about the role of integrated proviruses in HIV-infected individuals, with some recent work from a [00:01:00] postdoc of mine who left my lab 35 years ago. Another mark of antiquity, I suppose. 

Those of us who entered this field a long time ago, and indeed, I chose to work on what we then called RNA tumor viruses in 1969, before the discovery of reverse transcriptase and before we knew much about cancer genes, even cancer genes and viruses. Like many of my contemporaries, I was drawn to two things, the possibility of a [00:01:30] heretical life cycle, John used that word several times, and the capacity of these viruses to cause cancer. Indeed, my own predilection was to try to understand cancer by studying tumor viruses of some kind, either of the RNA fate or the DNA fate. These ideas, in many ways, are linked by an interest in the question of how [00:02:00] viral information is perpetuated in infected cells by, as you will see and as you all know, integration of a viral genome covalently into host chromosomes.

I'm going to tell you a lot about why, in thinking about the provirus hypothesis, it's correct to focus on reverse transcriptase. That is the big enchilada in the provirus world. John [Coffin] has beautifully told you about its discovery in many applications [00:02:30] and the importance of having a way to copy nucleic acid before we could do molecular cloning. That will play a very large role in what I had to tell you, but integration is also a vital aspect of what Howard conceived of as the provirus hypothesis. It explains the persistence of viral information as a provirus and infected cells, you all know that. It's relevant to the main topic of the discussion here this weekend, that the [00:03:00] HIV provirus is a major impediment to curing AIDS. And the integrase, as you will hear in a talk I believe tomorrow, is also a target for anti-HIV therapy. As I'll explain a little bit later, proviral integration does have a central role in oncogenesis by retroviruses, either as a result of capture of host proto-oncogenes or as a means [00:03:30] to activate endogenous proto-oncogenes by insertional mutagenesis.

Now, John has already introduced this concept of the two morphological forms of transformation by Rous sarcoma virus, I reintroduce it here because for Howard Temin, who had been exposed to prophage of bacteria, and should have been thinking, I would have thought, about the analogies between retroviruses and certain [00:04:00] bacteriophage. (1) He was more moved, as I understand it from his writings, by the idea that retrovirus like Rous sarcoma virus could come in different modes that would confer different but persistent phenotypes upon host cells, suggesting that it was the viral information itself and not some host program activated by a viral infection that conferred the fusiform or the round cell phenotype on [00:04:30] transformed chicken cells.

In thinking about how the viral information might persist, it's worth thinking in retrospect about the kinds of mechanisms that could have accounted for persistence of this information, but we all think now covalent integration into hosts chromosomes, but indeed there are other possibilities. For example, viral DNA could have been made and then replicated in a plasmid-like fashion by some [00:05:00] host DNA polymerases. That's certainly a possibility, it would be possible knowing more than was known in 1970, but it's known now that there are other retroviruses, hepadnaviruses, for example, like hepatitis B virus that can occasion the persistence of viral DNA by multiple rounds of reverse transcription as cells divide. Or you could imagine that there's integration, integration could be mediated by a virus-specific [00:05:30] mechanism, as we know is now true for all retroviruses. Or could have been mediated by some host machinery, efficient or inefficient. 

Sorting out these things, which ended up not being done at a systematic fashion, but by focusing on the most likely possibility, is actually important in retrospect to understand that the fact that retroviruses have a highly-specialized set of machinery for inserting viral DNA in a specific way [00:06:00] with specific orientations into host chromosomes is a very important aspect of how retroviruses operate.

Now the evidence for integration came along in different stages. One could argue, in a sense I just did, that Howard's belief that was articulated first very clearly in 1964 (1) that the Rous sarcoma virus genome would have been joined into host chromosomes, it wasn't really well supported [00:06:30] by the fragmentary evidence that was then present. A number of individuals especially, [Miroslav] Hill (b. 1929) and [Jana] Hillová, and Jan Svoboda (1934–2017, Czech virologist) carried out DNA transformation studies showing that some form of the viral genome, presumably DNA, was in cells, but they didn't rigorously prove that it was linked to host chromosomes. (2)

There was molecular hybridization experiments carried out with high molecular weight DNA. In my own lab, we did this with something called [00:07:00] the network test, which John as a master of retrovirus trivia probably knows but others don't, that other high molecular weight DNA prepared by centrifugation contain the viral genome suggesting it was probably linked to the host chromosome. Of course, what really cleared things up was doing molecular cloning that allowed [gene] mapping and sequencing of the viral ends and the joints to host chromosomes, showing that integration occurred at many different sites, [00:07:30] that the provirus always had long terminal repeats (LTR). It was always laid out in a manner that resembled the linear form of viral DNA, and linear DNA turns out to be the precursor to the provirus, not the circular forms that John showed in the slides in 1985. That made everything look very beautiful. You are all familiar with this kind of picture.

One point to make here that I won't have time to emphasize is that the provirus had a striking similarity to LTR-type retro—[00:08:00] transposons that are found in bacteria and mediate insertional events that cause mutations and allow transposition from one side to another. For those of us working on this problem at that time, that was a major way of getting us to think about retrovirus as [viral] vectors [for transduction] and number of other things.

When the joints were sequenced, it was apparent that these fine points were very precise, that the two base pairs were lost [00:08:30] from both ends of linear DNA when integration occurred, that insertion sites varied dramatically, although there is some very—retrovirus exhibits certain preference for certain kinds of sites, and that a host site is duplicated by a staggered cleavage at the target site, all indicative of the idea that there was some virus-specific, specialized mechanism for integration.

Now, further evidence for the idea [00:09:00] that there was a viral mechanism underlying integration, came, of course, the discovery of viral integrase, which was a discovery made in a variety of ways by looking for integration mutants as our Chairman Dr. [Stephen] Goff helped to work out and other labs including my own, and the discovery that the gag-pol polyprotein had a domain that was cleaved by protein, but that was generated by proteolytic cleavage, and then [00:09:30] a demonstration that either subviral particles or integrase itself could carry out the integration reaction and then the development of molecular structures for integration, which I won't have time to talk about in detail. 

But I think this one picture is emblematic of the progress that's occurred in this field over the recent years. This is a cryo-EM picture of the so-called intasome-nucleosome complex demonstrating, in this case, the integration of [prototype] foamy virus (PFV) [00:10:00] DNA into nucleosomes. (3)

I want at this point to switch to talk about the role of integration in oncogenesis. I'm going to begin by reminding you of the dramatic variety of retroviral genomes as we currently understand them, and point to the classification system that's most relevant to what I want to talk about, namely that we know that [00:10:30]some retroviruses have genes that are specifically devoted to transformation, not relevant to the replication of these viruses. The viruses that carry those genes about which I'll say a lot more in a moment have the ability to induce tumors rapidly, can transform cells in culture. Viruses on the left side don't have such genes.

We can also think, as Peter Vogt (b. 1932) taught me to think, about viruses as having the genetic [00:11:00] capacity to replicate independent of a helper virus, R+, or to transform cells in culture, T+. Only one virus (Rous sarcoma virus) on this slide is R+ T+, and that was the virus that our own group used for many of the study I'll now talk about. It still to my knowledge remains solitary as a virus completely competent for both replication [00:11:30] and transformation, and hence particularly convenient as a virus to work on for trying to understand the genetic and biochemical properties of retroviruses. We can also divide these viruses into the simple and complex. That's a story you'll hear a lot more about later in this meeting. 

Now, going back to one of the simpler viruses, the Rous sarcoma virus. The genetics of that virus as displayed from around 1975 are shown here: [00:12:00] three genes known to be involved as requirements for replication, gag, pol and envelope (env). One gene, src, defined both by conditional mutants and by deletion mutants, that I think have already been mentioned, were isolated both by Peter Vogt and Hidesaburo Hanafusa (花房 秀三郎, 1929–2009, Japanese virologist). Those deletions occur near the 3’ end of the viral genome removed the ability of those viruses to carry out transformation. The conditional [00:12:30] mutations, perhaps the most important in which was identified by Steve Martin when he was a postdoc in [UC] Berkeley in 1970, is illustrated here. Temperature-sensitive mutant that displays the number of important features, can't transform cells at the higher temperature. They're comparing 41° to 35, but strikingly this virus grows perfectly well at both the high and the low temperatures.

That tells you [00:13:00] in combination with the fact that that transformed cells when raised to a higher temperature revert to a normal phenotype; and in the reverse normal-looking cells become transformed, suggest that that an active gene is required, not just to initiate but to maintain, the cancer-like state; and that the transforming gene is not required for the replication of the virus. Very different from what one sees trying to do analogous studies with DNA tumor viruses, for example, where the oncogenic [00:13:30] properties of the virus are also implicated in the replication of the virus.

It raises the question of why Rous sarcoma virus should even bother to have such a gene, it's just extra baggage. [This] led us in San Francisco, Mike Bishop and I and our colleagues to ask about the nature of that gene. We were helped in trying to analyze the gene by the fact that there were deletion mutants that spanned most or all of the gene and that we had reverse transcriptase [00:14:00] in the pre-cloning era to make a radioactive probe with reverse transcriptase that was specific for the region of the Rous sarcoma virus that contain most or all of the src gene and was deleted in the generation of transformation defective mutants. It was possible to take radioactive, small fragments, products of reverse transcriptase of the Rous sarcoma virus genome, select them [00:14:30] against the transformation detected deletion mutant, and gather up a pot of radioactive DNA that was then specific for the src deletion and presumably specific, we now know that was essentially true, for the src gene.

By hybridizing that so-called src probe to DNA from various birds, it was possible to show that there was DNA in normal cells [00:15:00] from chickens and from many other avian species. We now know from essentially all metazoa, you can find homologous sequences that will hybridize or at least show by DNA sequencing, as we've subsequently demonstrated, that the src gene is present in normal cells in a slightly different form from the way it appears in the virus, and that the gene is highly conserved during evolution. Some of you will notice that radioactive DNA from the replication genes [00:15:30] also hybridized the chicken DNA. That's because of what Howard Temin had barely demonstrated many years before, the background in his hybridization assay, and attributable to endogenous retroviruses, to which there's no similarity in the other avian species examined for src DNA.

By various ways, we showed that the normal cell had the capacity [00:16:00] to make a protein similar to the protein product of the viral src gene. I don't have time to go into the difference between viral and cellular src genes, except to say that they differ in very significant ways and that the proteins made by these two genes differ dramatically in protein kinase activity with the v-src (viral src) gene being unremittingly activated and causing cell transformation, and the c-src (cellular src) [00:16:30] protein having a biochemically-regulated transforming activity that fails to transform cells.

We bring that all together by saying that there is a normal gene function of which still remains largely unknown in normal cells called the c-src gene, a proto-oncogene that somehow was captured by a virus that has replication functions, but [00:17:00] not transforming functions to generate the Rous sarcoma virus genome. I'll come back near the end to a compilation of how we think that happens.

For the moment, the really critical thing was to think about other transforming viruses of the retrovirus class that had the transforming functions, T, but were deficient in replication. All those cases, it's been found that there is a progenitor gene, [00:17:30] a proto-oncogene that is the source of a modified form of the gene found in various viruses.

Many of our colleagues from those days are shown here. I don't think anybody else is here. David Baltimore is coming near the end of this meeting but otherwise some of these folks have been mentioned but they're not in the audience. Happily, they're alive.

Audience: Can you name them all?

[laughter]

Harold Varmus: I can name them all but I don't even have time [00:18:00] to do so. I can do that for you later if you like. (4) In any event, highlighted here are some genes that are probably familiar to you, ErbB is the EGFR (epidermal growth factor) receptor, Myc, ABL (tyrosine-protein kinase ABL1), Ras, and some others are frequently mutated in human cancers, and the conduit from retroviruses to an understanding of human cancer generated by what I used to call the “retroviral genome project” that brought to light many of the important genes in human cancer. [00:18:30]

This leaves out a consideration of those viruses that I told you about before that lack a viral oncogene, a gene derived from a host gene and yet those viruses are also able to cause cancers—the cancers arise slowly, the viruses don't transform cells in culture. On the left are a few examples: the avian leukosis virus (ALV), murine leukemia virus (MLV), mouse mammary tumor virus (MMTV). We now know through studies of viral integrations [00:19:00] that, in general, those viruses cause cancers by activating proto-oncogenes by proviral integration.

In some cases, they affect known proto-oncogenes like the c-myc gene which is a precursor to an avian gene I'll mention in a moment. Sometimes they allow us to discover new oncogenes by causing insertion mutations, I'll show you an example of that very briefly. The insertions [00:19:30] inevitably activate expression of the proto-oncogenes and they can do that as a result of enhancer elements found in the proviruses, proviruses being upstream or downstream in either orientation. And this notion of a DNA rearrangement inducing the activation of an oncogene, sets a precedent for understanding certain viral-independent mutations found in human cancers [00:20:00] that create oncogenes.

I'll give you one example with perhaps one of most famous genes on that list, the c-myc locus shown here. That locus can be activated by capture of the proto-oncogene to create a virus, in this case, an avian leukemia virus, that's R-T+, the [avian] myelocytomatosis virus (AMV). It can be activated by infection of chickens with an avian leukosis virus (ALV) [00:20:30], it's R+T- that causes bursal lymphoma, but activates the c-myc gene, and as a consequence of a proviral integration, or it's found activated in a variety of human tumors and mouse tumors as well, by a rearrangement, for example, in this case, a translocation that activates expression. That's one set of genes that activate known proto-oncogenes.

But there are also viruses like [00:21:00] the mouse mammary tumor virus that tend to cause tumors in this case, breast cancers by activating proto-oncogenes that were not previously known. And the way of going about doing that was to use an approach based on transposon tagging, looking at the location of a provirus and asking what's nearby, and when that was done with the mammary tumor virus of, well, over 30 years ago, Roel Nusse (b. 1950) and I discovered a gene called, [00:21:30] initially, the "integration site 1" gene int1, which proved through studies carried out later by Roel Nusse to be the mammalian homolog of the Drosophila wingless gene (Wg), the gene we now call Wnt1. (5)

That was the way in which the Wnt signaling pathway was discovered. A lot of studies on Wnt gene evolution, Wnt signaling, the world of Wnt and development, and regulation of [00:22:00] stem cell replication had still did not reveal any role for Wnt itself in tumorigenesis, but by looking at the signaling pathway, that Wnt1, a secretory protein activates, it was possible to show starting in the mid-'90s, that many other components of the signaling pathway, for example, the APC gene, β-catenin (beta-catenin), and on the AXIN gene and others, do play important roles in a variety of human cancers, [00:22:30] especially colon cancer, but also liver and other cancers.

At this point, we're able to say something about how retroviruses like the Rous sarcoma virus or mouse sarcoma viruses acquired their oncogenes, putting together a whole bunch of attributes of retroviruses that came from efforts to understand how retroviruses work: efficient quasi-random, viral integration that is mediated by a specialized, [00:23:00] virus-specific mechanism. The ability of proviruses to activate genes nearby the insertion site, to expand the clone to make a tumor that has an activated oncogene, transcriptional read through of the long terminal repeat to enter the domains of the oncogene, or there are other ways to that happen as well, [00:23:30] I don't have time to talk about that. Packaging of two subunits of viral RNAs, and we haven't talked about here yet, but making viral heterodimers, making use of something that John talked to you about, namely, the ability of the reverse transcriptase to move from one template to another to basically jump from one template to the next to create recombinants.

When you put all that together, integration near cellular oncogene, co-packaging of RNAs, recombination during reverse transcriptase, [00:24:00] you end up in the next round of infection with a provirus that has a modified proto-oncogene, activated viral oncogene incorporated into the viral genome.

My last three minutes, I'm going to talk a little bit about HIV. Just very briefly. To go back to this question of HIV persistence in patients and its relationship to integration and integration sites. We know that [00:24:30] the persistence of HIV, which is obviously a critical issue for this audience to think about over the next couple of days, it's not due simply to continued replication of the virus. We know that patients whose virus replication is suppressed by antiretroviral therapy, patients who are natural controllers of replication, still show persistence of the genome.

Presumably, that's because HIV proviruses—many are defective, but some are infectious—are integrated at many different [00:25:00] sites and host genomes. When you analyze the integration sites of collections of HIV-infected cells, the results show that some infected cells expand disproportionately in a clonal fashion. Some clones presumably are eliminated for other reasons, but in work done by [Stephen H.] Steve Hughes, the postdoc who left my lab 35 years ago, has been at Frederick [National Laboratory for Cancer Research] most recently, [00:25:30] after a time, Cold Spring Harbor and the and the Frederick-branch of the intramural program or the NCI. Steve has shown that fairly large percentage of infected cells are in such preferentially expanded clones, and those clones can exist for a very long time. That's obviously an impediment to cures.

Now, why do those clones expand? Well, there could be either of two reasons for that come to his mind and to mine as well. One is [00:26:00] that an integrated provirus might act a little bit like the proviruses we've been talking about in tumors that are induced by retroviruses without their own oncogenes, and might be able to affect the expression of some gene involved in growth regulation. Two have been identified in his studies MKL2 and BACH2, these are both encode the factors that are involved in transcriptional control.

What you might expect to have happen then is that there would be an [00:26:30] overtime marked increase in clonal abundance, something which is sometimes but not very commonly observed. John Coffin and Steve [Hughes] have recently written a review (6) and that shows that they first infected peripheral blood mononuclear cells with virus and showed that there were many integrations in this MKL2 gene in different orientations, but when they looked at at least one patient who had many clones with insertions in this gene, [00:27:00] all the insertions were in a fairly restricted area, or these most of them were, and they were all on the same orientation. There were a number of transcripts, abundant transcripts associated with those insertions.

On the other hand, to try to explain the other expanded clones, it might be the case that infected immune cells are responding to antigens or to cytokines and other homeostatic signals in preferential ways that would increase [00:27:30] the abundance of those clones, and that shouldn't affect the relative abundance over time—hard thing to test, but that seems to be largely what's happening. Most findings are consistent with that model. But here's the problem which needs to be better understood and represents one of the things for the future.

As you've heard throughout this session, retro biology has a long history now over a century. Some of the heroes and critical events are shown here and if there's anything [00:28:00] that I've said that you really need to take home and think about is that we do have a lot more learned about how retroviruses operate, especially those that are causing a devastating disease in the human population. I hope that some of you will solve some of the remaining problems. Thanks very much.

[applause] 

Julie Overbaugh (Moderator): We have three or four minutes for questions [00:28:30]

Harold: I never expect much in the way of questions at historical talks, but happy to welcome them.

Beatrice Hahn: Hi, Harold, I have a question. A virus like Rous sarcoma virus, has that ever been found in like outbred birds, not the hens in the henhouse, but wild species, is such a thing--

Harold: My own view is that the selective pressure that allows us to isolate viruses [00:29:00] that have oncogene is: scientists. Scientists find tumors, and Peter Vogt has been perhaps the most important selective pressure I know because he spent many weeks at poultry factories, taking birds with tumors, stashing them in the freezer. That's how the JUN oncogene was discovered. Just go to a tumor, and especially when the tumor arises in an animal that we know is being exposed [00:29:30] to viruses that lack oncogenes, you will find tumors that are particularly useful. Peter [Vogt] has found many and [Hidesaburo] Hanafusa found several, Robin [Weiss] made—Robin has actually looked at wild birds infected in jungle, you want to say something about that?

Robin Weiss: They had the leukosis viruses, but I spent my last six weeks in Peter’s lab in 1972 at this chicken slaughterhouse in LA. I didn't eat chicken for about a year afterwards. Peter came along in the first day with— [00:30:00] he refused to wear hard hats, and which the vet in charge insisted on so he was banned. I left him with this freezer full of tumors…

Harold: So it was you! Okay, you're the selective force.

Robin: He got further ones. I can't think of a single mammalian or avian oncogene-carrying retrovirus that is naturally transmitted.

Harold: That's probably true.

Robin: As you they’re pathologists’ rarities, [00:30:30] one-off sports.

Harold: There's a difference, it is obviously—

Robin: [crosstalk]

Harold: —it's a disadvantage for the virus to be carrying that gene. It's not it's just extra baggage and indeed, if you grow Rous sarcoma virus to sarc+, very quickly the culture becomes dominated by helper virus, with the deletion mutations occur, the high frequency and the deletion mutants grow to higher titre.

John Coffin: Well, the deletion mutants also grow because the transformed cells tend to fall out of your culture.

Robin: There's something odd about the [00:31:00] R+T+, the replication competent sarcoma viruses that Schmidt-Ruppin and Prague and Bratislava [strains of RSV]— They'd all been passed through mammals, right? Forced in at high titre, and what came out was probably a recombinant.

Harold: I don't know any evidence for their being recombinant.

Robin: There isn't hard evidence that--

John Coffin: [crosstalk] between the two virus genomes [00:31:30].

Harold: Okay, sorry, professor.

John Mellors: Thanks for that, Harold. The genome of the CD4 cells [are] bombarded, right, with retroviral insertions and this clonal expansion suggests that some of the insertions are influencing cell biology. Do you have concerns that after a long latency period, some of these clones may little further up and transform? One thing that's very obvious, so [00:32:00] far, is we can't find a CD4⁺ T cell lymphoma. They're extremely rare in HIV infection. What do you think's going on there? Have we not waited long enough? Or is there something inherent about a CD4 cell that is unlikely to result in transformation?

Harold: Yes, and someone else probably answered this question. I've never studied this, but of course everyone knows here that [00:32:30] HIV is a risk factor for a variety of tumors. Then, to my knowledge, no one's ever found a cancer that seems to be driven by an HIV insertion, but whether there's some selection against CD4⁺ lymphomas in AIDS patients, I really don't know the answer to that question, Bob [Gallo], or others?.

John Mellors: There is one tumor currently that's been reported with a BACH2 insertion [00:33:00].

Bob Gallo: You might as well say none.

Harold: If it's only one, he probably says there are none.

Bob: Going back to what Beatrice said, I always thought of it that if you had an oncogene-containing virus that was replication-competent, the species would almost go out of existence if it was highly replicating.

Robin: No, I don't think.

John Coffin: It's got to have a probability being transmitted [00:33:30] before it kills.

Bob: Yes, but I do want to point out that the HTLV Tax protein can produce leukemia by itself, but it has no cellular homolog. So in a sense, it's an oncogene. In a sense, it's not because it's not what we define as an oncogene.

Harold: Well, no, I think you could call it a viral oncogene if the tax gene has some attributes of a cancer gene. You don't have to have a cellular homolog to be oncogenic.

Bob: Well, therefore, then there is one. It's horizontally transmitted, but it's not very infectious.

Harold: It's weak, it's weak.

Bob: Weak [00:34:00] yes, but it's not weakly transforming, it's very efficient transforming maybe more than almost any naturally occurring tumor virus that I know of. 3% to 4% leukemia per lifetime is not weak, but it's weak to transmit, is extremely weak to transmit, it's transmitted only in a cell. I think that's why we don't get a lot.

Robin Weiss: [unintelligible 00:34:21]

John Coffin: To John Mellors' point—

Harold: Or EBV.

John Coffin: —I would argue about CD4⁺ T cell tumors and AIDS patients. I would argue that malignant CD4 cells or premalignant cells are probably excellent targets for the HIV that's already present in that patient and are very likely to be rapidly killed off if they arise I think. Just like [unintelligible 00:34:46] things like that.

Participant: [unintelligible 00:34:49]

John Mellors: John, that would be the case if they were viremic, but if they're on suppressive (antiretroviral therapy) ART, [00:35:00] those cells won't be infected.

John Coffin: They would have to have been infected and transformed over a very narrow window before you started the therapy.

John Mellors: Why?

John Coffin: Why couldn't the transformation occur after?

John Mellors: Because the virus has been around--

Harold: Give him the microphone.

[laughter]

Julie: I just have the feeling that John and John have argued this point before, that's all.

[laughter]

Just guessing, just a feeling it's been discussed.

Robin: It may be trying to move on to feline leukemia viruses.

Julie: [laughs] Yes, exactly. [00:35:30] Okay, thanks, Harold.

[applause]

[00:35:36] [END OF AUDIO]

Citations

1. This is either:

• Temin, Howard M. “Homology Between RNA from Rous Sarcoma Virus and DNA from Rous Sarcoma Virus-Infected Cells.” Proceedings of the National Academy of Sciences 52, no. 2 (August 1, 1964): 323–29. doi:10.1073/pnas.52.2.323.
• Temin, Howard M. “The Participation of DNA in Rous Sarcoma Virus Production.” Virology 23, no. 4 (August 1, 1964): 486–94. doi:10.1016/0042-6822(64)90232-6.

 2. According to Robin A. Weiss, “A Perspective on the Early Days of RAS Research,” Cancer and Metastasis Reviews, July 29, 2020, doi:10.1007/s10555-020-09919-1, these articles include:

• Hlozánek, Ivo, and Jan Svoboda. “Characterization of Viruses Obtained after Cell Fusion or Transfection of Chicken Cells with DNA from Virogenic Mammalian Rous Sarcoma Cells.” The Journal of General Virology 17, no. 1 (October 1972): 55–59. doi:10.1099/0022-1317-17-1-55.
• Hill, Miroslav, and Jana Hillová. “Virus Recovery in Chicken Cells Tested with Rous Sarcoma Cell DNA.” Nature New Biology 237, no. 71 (May 10, 1972): 35–39. doi:10.1038/newbio237035a0.
  

3. Maskell, Daniel P., Ludovic Renault, Erik Serrao, Paul Lesbats, Rishi Matadeen, Stephen Hare, Dirk Lindemann, Alan N. Engelman, Alessandro Costa, and Peter Cherepanov. “Structural Basis for Retroviral Integration into Nucleosomes.” Nature 523, no. 7560 (July 16, 2015): 366–69. doi:10.1038/nature14495.

4. On the slide, from left to right are: Harold Varmus, J. Michael Bishop, Charles Sherr, David Baltimore, Edward Scolnick, Robert Weinberg, Stuart Aaronson, Inder Verma, and George Vande Woude. The cartoon was made by Jamie Simon, created for a CSHL meeting in 1983. See Robin Scheffler, A Contagious Cause: The American Hunt for Cancer Viruses and the Rise of Molecular Medicine (Chicago: University of Chicago Press, 2019), page 218.

5. Nusse, Roel, and Harold E. Varmus. “Many Tumors Induced by the Mouse Mammary Tumor Virus Contain a Provirus Integrated in the Same Region of the Host Genome.” Cell 31, no. 1 (November 1, 1982): 99–109. doi:10.1016/0092-8674(82)90409-3.

6. Hughes, Stephen H., and John M. Coffin. “What Integration Sites Tell Us about HIV Persistence.” Cell Host & Microbe 19, no. 5 (May 11, 2016): 588–98. doi:10.1016/j.chom.2016.04.010.

Index 

• 1.4 Robin Weiss — Retrovirus History and Early Searches for Human Retroviruses
• 1.5 John Coffin — The Origin of Molecular Retrovirology
• 2.4 Robert Gallo — Discoveries of Human Retrovirus, Their Linkage to Disease as Causative Agents & Preparation for the Future
• 4.3 Beatrice Hahn — Apes to Humans: The Origin of HIV
• 8.1 John Mellors — MACS and Beyond: Epidemiology, Viremia and Pathogenesis
• 8.6 David Baltimore — Bringing it to an End (And Where Are We Going?)
• ABL
• analogy
• antiretroviral therapy (ART)
• APC
• avian leukosis virus (ALV)
• avian myeloblastosis virus (AMV)
• AXIN
• BACH2
• bacteriophage
• Bishop, J. Michael (b. 1936)
• chicken
• Cold Spring Harbor Laboratory (CSHL)
• colon cancer
• cytokines
• Drosophila
• education and early career
• EGFR
• env
• Epstein-Barr virus (EBV)
• ErbB
• evolutionary developmental biology (evo-devo)
• feline leukemia virus (FeLV)
• Frederick National Laboratory for Cancer Research (FNLCR, NCI-Frederick)
• freezer
• gag
• gene mapping
• Hanafusa, Hidesaburo (花房 秀三郎, 1929–2009)
• Hill, Miroslav (b. 1929)
• Hillová, Jana
• HTLV (human T-lymphotropic virus)
• iconoclasm in science
• information, genetic
• integrase
• JUN
• kinases
• leukemia and lymphoma
• Los Angeles
• LTR (long terminal repeat)
• Martin, G. Steven
• mechanism
• microscope — electron and optical
• MKL2
• molecular cloning
• mouse mammary tumor virus (MMTV)
• murine leukemia virus (MLV)
• Myc
• National Cancer Institute (NCI)
• National Institutes of Health (NIH)
• natural selection, evolutionary selection, evolutionary fitness
• nucleic acid hybridization
• Nusse, Roeland (b. 1950)
• observation
• oncovirus
• pathology (medical discipline)
• pol
• prototype foamy virus (PFV)
• radionuclide, radiolabeling, radioactive tracer
• Ras
• reverse transcriptase
• Rous sarcoma virus (RSV)
• sequencing
• Session 1: The Story of Animal Retroviruses
• src
• structural biology
• Svoboda, Jan (1934–2017, Czech virologist)
• tax
• Temin, Howard M. (1934–1994)
• transfection, transduction, viral vector
• transformation (animal)
• transposon
• UC Berkeley (University of California, Berkeley)
• veterinary medicine
• Vogt, Peter K. (b. 1932)
• Wnt gene family (int1, Wingless, etc.) 
• β-catenin