Page History

Andrew Rice: I'd like to thank the organizers for inviting me to present our work at this wonderful meeting. It's a great honor. Today, the mechanism of action of Tat is known in considerable detail. Towards the end of my talk, I'll give a slide on the most up to date model. In this presentation, my charge was to give a personal historical perspective. I'm going to cite some of the most influential [00:00:30] papers that led to our current understanding of Tat. I apologize to some authors in the audience and in the world who I don't cite their papers because of lack of time and my imperfect memory. Then I'm going to focus the talk on the work in my lab in the mid-1990s, that pointed the field into the direction that it's gone.

The story [00:01:00] begins in 1985 when Joe Sodroski, and Flossie [Wong-Staal], the [Robert] Gallo lab, and the [William] Haseltine lab, first identified an activity in HIV-infected cells that greatly activated HIV LTR directed gene expression. (1) This was a discovery of the Tat function. Shortly thereafter, a paper from Craig Rosen and the Haseltine lab, identified the cis-regulatory elements that responded to [00:01:30] Tat transactivation. (2) These experiments were done with plasmid transfections and batching the promoter.

In 1985, this is where we were. The 5'-LTR had binding sites for cellular transcription factors. It looked pretty ordinary. The TAR (trans-activation response) element was identified to be a downstream from the side of transcriptional initiation. At the time, this was really unprecedented. This could exist [00:02:00] as an RNA element, or it conceivably could be a DNA element. At this time, there was a paper that suggested that Tat activated through TAR through a post-transcriptional mechanism. I had just arrived, I'd done a postdoc at ICRF in London, with Ian Kerr studying interferon mechanisms of antiviral mechanisms, and I had a background in post-transcriptional mechanisms. [00:02:30] I thought that this would be an important topic to work on. I entered the field in about 1986.

Then, for the next probably 10 years, I would say, insight into the mechanism of action of Tat was a real mess. There were a whole number of publications that didn't stand the test of time. There were claims [00:03:00] that it was probably a translational control mechanism. It was an effect of an increase in initiation of transcription by RNA polymerase II, or an effect on elongation, and some papers claimed a combination of all these effects. But one of the most influential papers that got it right was from Matija Peterlin's lab at UCSF in 1987, were using a [00:03:30] horrible biochemical assay known as a nuclear run-on. He presented very strong evidence that Tat acted at the level of transcriptional elongation. (3)

This was followed the next year from a publication from Eric Holland, who was then at Stanford, using reporter plasmids in a genetic analysis, and he presented very compelling evidence that the TAR element, the element that responded to Tat, [00:04:00] was actually an RNA element. (4) Then, this was followed in 1989 from a publication from Jon Karn and colleagues at the MRC in Cambridge, that Tat bound directly to TAR RNA in vitro. (5)

So In 1989, this is where we were. It was established—evidence existed that Tat bound directly to the TAR [00:04:30] RNA element and that somehow activated transcriptional elongation. Of course, the question is, what is the mechanism of action? At this point, I had, in 1990, move my lab to Baylor in Houston. I was lucky to recruit a postdoc into my lab, Christine Herrmann. And if Carol Carter is in the audience, Christine, Chris is [00:05:00] a graduate of the Stony Brook, PhD program. She was a postdoc here at Cold Spring Harbor, where we met and got married. So, Cold Spring Harbor has a direct link to this. [laughter] Chris was promoted to the faculty over the years at Baylor. This is a picture where we were honored to receive a research award sponsored by Michael DeBakey at Baylor. [00:05:30] 

When Chris started her postdoc in 1992, after a maternity leave, and she was only working about 75% at that time. The lab had generated these reagents, these are recombinant Tat proteins expressed as GST fusions, so you could express them in E. coli very easily, purifying them by binding to [sepharose] beads, and get a nice set of purified [00:06:00] proteins. We had to characterize these in transactivation assays, so we had a variety of HIV-1 Tat, HIV-2 Tat proteins with and without functioning in vivo.

What Chris initially tried to do is, we've metabolically labeled cells with S35 [methionine], and then we were trying to bind these extracts, run gels and see proteins that specifically [00:06:30] associated with the wild type Tat, but not the mutant Tats. We could never see anything convincing, and Chris spent a considerable time trying to optimize labeling conditions, preparation of extracts and binding, dilution washing and we could find nothing above background.

At that point, I went back to my background in the interferon field, and I recalled that the interferon regulated [00:07:00] protein kinase, PKR, is autophosphorylated when you do in vitro reactions. I suggested to Chris that why don't we throw some P-32 labeled ATP into these reactions and see if we can detect a protein kinase activity. This was the first experiment that she did on April 1st, and we were hoping it wasn't a cosmic April Fool's joke. [00:07:30] We could detect a protein kinase activity that bound to the HIV Tat.

In this paper, we showed that that activity correlated with a functional Tat. (6) In fact, a 42-kilodalton polypeptide was autophosphorylated in the reaction, and we suspected that was the catalytic domain. We were fortunate [00:08:00] that the HIV-2 Tat protein is actually a substrate for this kinase, not the HIV-1 Tat. We don't have any real information of whether this is mechanistically important, but it was useful.

In our next publication, we show that this protein kinase, which Chris called TAK, for Tat associated kinase, was able to phosphorylate [00:08:30] the large subunit of RNA polymerase in its CTD (carboxylterminal domain) domain. (7) In the transcription field, it was known that RNA polymerase complexes that were engaged in active elongation had a hyper-phosphorylated CTD of the RNA polymerase. This correlated with the activation of transcriptional elongation.

This is one of my favorite experiments that Chris did in in vitro kinase reaction [00:09:00] using the HIV-2 Tat protein to pull out a kinase activity in nuclear extracts. This is a time course. You see this really robust phosphorylation of an exogenous CTD substrate and the so-called CTD-O form is one that's at hyperphosphorylated. I think it must have seven or more phosphorylation events and it really retards its migration in SDS gels. [00:09:30] 

We made use of our GST Tat collection and ran them through this in vitro kinase reaction. You see any recombinant Tat protein that would bind the kinase that could hyperphosphorylate the CTD was active in vitro and the mutants that had no function could not bind this kinase activity. This was a precise [00:10:00] correlation between binding this kinase activity and activation of elongation in vivo.

We presented this model in 1995, which has stood the test of time, although it's been made more immensely, more complicated. We proposed at that time that Tat bound to TAR RNA, and in this binding, it also required what was then termed a loop factor. Bryan Cullens' (b. 1951) lab had conducted genetic experiments that argued that there was a loop factor required for Tat to bind to TAR RNA, and then we proposed that in this complex was this Tat associated kinase, which would phosphorylate the RNA polymerase and activate elongation.

Here is an early more colorful rendition of this model, which was a Christmas gift from my brother and his wife, who are also virologists in Ashley Haase's department and it shows, I don't know if you can see, Tat binding to TAR, bringing with it TAK and hyperphosphorylating the RNA polymerase CTD. This is a unique model.

Shortly after our publication, David Price and colleagues, one of being Mike Mathews here at Cold Spring Harbor, showed that the kinase activity that we had identified was actually identical to a protein transcriptional elongation factor that David had termed P-TEFb, for positive transcription elongation factor B and in this paper, in Genes & Development, he showed that the catalytic subunit of P-TEFb or TAK was CDK9, which had formerly been known as PITALRE, but giving it's known identity, it was then termed CDK9.

We had heard a rumor that David was claiming the catalytic activity was a cyclin-dependent kinase, and actually, I heard that at a Cold Spring Harbor meeting. I was aware of the CDK literature so I went to the library here and within 15 minutes had guessed through the autophosphorylation or kinase being a 42 kilodalton subunit that TAK was indeed CDK9. So we published this in 1997, and in this paper, we also showed that P-TEFb activity or TAK activity is downregulated in resting CD4 T cells. (8) I'll get to that point in a few later slides. With the identity of the catalytic subunit of TAK, a graduate student in the lab, Moses Gold, conducted this experiment in a reporter plasmid generated in the Cullen lab, they had replaced the TAR element with the Rev response element (RRE) and then one could direct a rev fusion to this artificial promoter to see if it would activate LTR directed gene expression. Moses had actually thought of this before we knew the identity of CDK9, and he had shown that one could activate this reporter plasmid with a fusion of CDK8 but it's unrelated to our story today. Moses performed this experiment and using the CDK9 or PITALRE-Rev fusion, he got robust activation of this reporter plasmid and we reported this in 1998. (9)

At this time, Kathy Jones at the Salk, using a more modern methodology, doing mass spec analysis of proteins that are bound to Tat, identified cyclin T1 is the regulatory subunit of CDK9. (10) Now we had the core enzyme activity, the regulatory subunit is Tat, the catalytic subunit is CDK9.

Going back to this old model in 1995, this is where we were in the mid 1990s and now, if we fast forward 21 years, this is a simplified model of the mechanism of action of Tat and it's come through the work of a number of labs, there are too many to cite, but what's believed now is that core P-TEFb, CDK9, cyclin T1 in most cells are exist in an RNP complex containing the non-coding RNA 7SK with a number of other proteins.

This is work originally done from [unintelligible 00:15:29] lab at UC Berkeley and Olivier Bensaude in France. Tat somehow extracts  [unintelligible 00:15:36] P-TEFb from this RNP complex. Then other proteins assemble in a complex known as a super elongation complex, and then the catalytic subunit of CDK9 phosphorylates subunits of negative elongation factors that normally associated with RNA polymerase and restrict elongation. CDK9 and also phosphorylates residues in the carboxy-terminal domain of RNA polymerase. There are yet a number of mechanistic questions that still exist in the field and it's still a credibly active and productive area.

Going back to 1998, we now knew the components of this Tat associated kinase P-TEFb, we had reagents. In this paper, with Kathy's lab, we looked at CD4 T cells in monocytes, macrophages, at the regulation of this kinase and we found that it is downregulated in CD4 lymphocytes in resting cells (11). This is now relevant to the current HIV cure efforts if one believes reactivation of latent virus as any component of that strategy. And shown here now is a demonstration of how this kinase is downregulated in resting CD4 T cells if you isolate them from healthy blood donors. (12, 13) Shown here are two blood donors. In resting cells, the level of cyclin T1 is very low. We know the messenger RNA levels of cyclin T1 are actually quite high here. We've identified in subsequent papers that there are micro RNAs that repress the expression of cyclin T1 and we also know that proteostome-mediated proteolysis contributes to the low level of cyclin T1 and it's strongly upregulated when you activate these cells.

If we look at CDK9, the story is slightly different. Generally, the basal level of CDK9 in a resting T cell is quite high and upon activation, it does increase. But if you look at this phosphorylation in the so called CDK9 T loop, this is a residue that has to be phosphorylated to give the enzyme catalytic activity, and you see in a resting cell, the T loop is phosphorylated at a very low level, and then upon activation, its phosphorylation increases and we know there's some identified protein phosphatases and kinase that regulate this issue.

In more recent experiments in the field, a number of molecules have been identified in primary CD4 T-cell models of latency. They all indicate that P-TEFb or what we used to call TAK, has to be upregulated to reactivate latent virus. In fact, there are a number of compounds or methods that can modestly or minimally activate T cell and up-regulate this kinase. We showed in this old paper that prostratin can up-regulate cyclin T1 without increasing activation markers on resting T cells. More recent results with more physiologically—or compounds with more potential for humans, such as ingenol derivatives have shown that these compounds can up-regulate this kinase without activating T cells and reactivating latent HIV in in vitro systems. Some of HDAC inhibitors, romidepsin, vorinostat, which has been shown to reactivate latent virus in people, we've shown recently that these compounds can upregulate this protein kinase.

I'll end here. This was the size of my lab in the mid-1980s at Baylor. This is Chris who gets the lion's share of the work, credit. Graduate students, Hyangshuk Rhim, Moses Gold, and Xinzhen Yang, and finally, the acknowledgments. This all started at Cold Spring Harbor and I'd like to acknowledge Mike Mathews and Jim Watson for their support during that time, and then at Baylor, of course, Christine gets the bulk of the credit for showing that Tat targeted [00:20:30] protein kinase and all of these students and postdocs contributed, and of course, this couldn't have been done without funding. Thank you.

[applause] 

Paul Bieniasz (Moderator): Excellent. We have time for a couple of questions.

Michael Emerman: Andy, thanks for that background. I want to convey to somebody who was not working in this field but read up the papers, how incredibly confusing the models were. [00:21:00] You could read two papers and they're all contradictory to each other.

Andrew: I remember there would be two Cell papers back to back that contradicted each other in those days.

Paul: To this day, there are some people in the field who think that this is what Tat does but it also does other things too.

Andrew: Right, my view is Tat certainly does this, but there is no divine law [00:21:30] that this viral protein can acquire other functions because it has lots of surfaces to make contact.

Paul: In terms of transcriptional activation, is this is the whole story or is there—?

Andrew: Probably not. I think this is a story that's that solid and the field agrees upon, but there are minor arguments that there are effects on transcriptional initiation and maybe even post-transcriptional event.

John Coffin: Of other genes as well?

Andrew: [00:22:00] Other HIV genes?

John Coffin: Other cellular genes.

Andrew: This elongation factor is absolutely essential for the elongation of almost most, if not all, cellular [RNA polymerase II] transcripts.

John Coffin: The question is that, is there any evidence, or even much speculation about effects of Tat itself on the expression of other genes? There's a lot of that in tax in HTLV, this is why I asked the question.

Andrew: There [00:22:30] is a lot of transcriptional profiling of cells expressing Tat and Tat-induced genes have been identified, but to my knowledge, there's no evidence that Tat directly activates those genes. The mechanism whereby they're induced is unclear.

Paul: Ron.

Ron Desrosiers: Maybe you said this and I missed it, but if one were to change the TAR sequence to something more innocuous [00:23:00] or not folded, would you predict that such a virus would be Tat-independent? Has anyone done that experiment and if not, wouldn't that be an interesting experiment?

Andrew: You can inactivate the TAR element in a provirus and it doesn't replicate. So it's—

Ron: What do you mean inactivate? Change the nucleotides?

Andrew: Change nucleotides so Tat can't recruit P-TEFb, and that virus does not replicate. [00:23:30]

Ron: How do you interpret that if the polymerase is not held up there, if you've changed it to a sequence where there's good readthrough?

Andrew: The signals for RNA polymerase to stall are built into the RNA polymerase complex when it binds. And there are negative factors, two protein complexes that enforce this stall on RNA polymerase and that has to be overcome by [00:24:00] recruiting P-TEFb.

Paul: Last question.

John Mellors: Very nice, Andy. two questions. How do you think the Susana Valente compounds, the Cortistatin A is working either acutely or post-exposure and—?

Andrew: I'm very concerned that when those compounds are used over such a long term, [00:24:30] the cells that were reactivated die in the culture.

John Mellors: Second, you've studied this for 25 years, what about a small molecule to persuade Tat to—?

Andrew: That's certainly feasible, but I think—

John Mellors: Where would you target? In this complex, what you would target?

Andrew: Tat interacts [00:25:00] with TAR RNA and that's essential. That's a interaction between two viral components.

John Mellors: You have to get P-TEFb there also or will Tat escort it?

Andrew: Tat has to escort it. If you could prevent that interaction, you would prevent P-TEFb and that would probably inhibit viral replication, but for the pharmaceutical industries to want to do that, they would have to pump in a billion dollars, and the classes of antivirals that exist now are so [00:25:30] wonderful—that it's not realistic to think that we want another class of drugs.

John Mellors: I'm thinking of activating transcription. They kick through, mediated through Tat. Can't we leverage its inherent activity to—?

Andrew: We have some data that we're writing up that in primary cells infected with HIV, [00:26:00] if you do a flow cytometry analysis, the cells with the highest levels of p24 (capsid protein) have the highest levels of cyclin T1, and vice versa. The data is suggesting maybe, that if you just tickle the latent virus enough to make some Tat, that will auto induce positive feedback induce P-TEFb, and then you get a boost of viral replication. That's if [00:26:30]you think of the shock and kill strategy, that is a way that you could maybe selectively reactivate latent cells without activating the non-infected cells.

Paul: Thank you, Andy. Let's move on.

Citations

1. Sodroski, Joseph G., Craig A. Rosen, Flossie Wong-Staal, Syed Zaki Salahuddin, Mikulas Popovic, Suresh K. Arya, and Robert C. Gallo. “Trans-Acting Transcriptional Regulation of Human T-Cell Leukemia Virus Type III Long Terminal Repeat.” Science 227, no. 4683 (January 11, 1985): 171–73. doi:10.1126/science.2981427.
2. Rosen, Craig A., Joseph G. Sodroski, and William A. Haseltine. “The Location of Cis-Acting Regulatory Sequences in the Human T Cell Lymphotropic Virus Type III (HTLV-III/LAV) Long Terminal Repeat.” Cell 41, no. 3 (July 1985): 813–23. doi:10.1016/S0092-8674(85)80062-3.
3. Kao, Shaw-Yi, Andrew F. Calman, Paul A. Luciw, and B. Matija Peterlin. “Anti-Termination of Transcription within the Long Terminal Repeat of HIV-1 by tat Gene Product.” Nature 330, no. 6147 (December 3, 1987): 489–93. doi:10.1038/330489a0.
4. Feng, Sandy, and Eric C. Holland. “HIV-1 Tat Trans- Activation Requires the Loop Sequence within Tar.” Nature 334, no. 6178 (July 1988): 165–67. doi:10.1038/334165a0.
5. Dingwall, Colin, Ingemar Ernberg, Michael J. Gait, Sheila M. Green, Shaun Heaphy, Jonathan Karn, Anthony D. Lowe, Mohinder Singh, Michael A. Skinner, and Robert Valerio. “Human Immunodeficiency Virus 1 tat Protein Binds Trans-Activation-Responsive Region (TAR) RNA in Vitro.” Proceedings of the National Academy of Sciences 86, no. 18 (September 1, 1989): 6925–29. doi:10.1073/pnas.86.18.6925.
6. Herrmann, Christine H., and Andrew P. Rice. “Specific Interaction of the Human Immunodeficiency Virus Tat Proteins with a Cellular Protein Kinase.” Virology 197, no. 2 (December 1, 1993): 601–8. doi:10.1006/viro.1993.1634.
7. Zhu, Yuerong, Tsafrira Pe’ery, Junmin Peng, Yegnanarayana Ramanathan, Nick Marshall, Tricia Marshall, Brad Amendt, Michael B. Mathews, and David H. Price. “Transcription Elongation Factor P-TEFb Is Required for HIV-1 Tat Transactivation in Vitro.” Genes & Development 11, no. 20 (October 15, 1997): 2622–32. doi:10.1101/gad.11.20.2622.
8. Yang, Xinzhen, Moses O. Gold, Derek Ng Tang, Dorothy E. Lewis, Estuardo Aguilar-Cordova, Andrew P. Rice, and Christine H. Herrmann. “TAK, an HIV Tat-Associated Kinase, Is a Member of the Cyclin-Dependent Family of Protein Kinases and Is Induced by Activation of Peripheral Blood Lymphocytes and Differentiation of Promonocytic Cell Lines.” PNAS  94, no. 23 (November 11, 1997): 12331–36. doi:10.1073/pnas.94.23.12331.
9. Gold, Moses O., Xinzhen Yang, Christine H. Herrmann, and Andrew P. Rice. “PITALRE, the Catalytic Subunit of TAK, Is Required for Human Immunodeficiency Virus Tat Transactivation In Vivo.” Journal of Virology 72, no. 5 (May 1, 1998): 4448–53. doi:10.1128/JVI.72.5.4448-4453.1998.
10. Wei, Ping, Mitchell E Garber, Shi-Min Fang, Wolfgang H Fischer, and Katherine A Jones. “A Novel CDK9-Associated C-Type Cyclin Interacts Directly with HIV-1 Tat and Mediates Its High-Affinity, Loop-Specific Binding to TAR RNA.” Cell 92, no. 4 (February 20, 1998): 451–62. doi:10.1016/S0092-8674(00)80939-3.
11. Herrmann, Christine H., Richard G. Carroll, Ping Wei, Katherine A. Jones, and Andrew P. Rice. “Tat-Associated Kinase, TAK, Activity Is Regulated by Distinct Mechanisms in Peripheral Blood Lymphocytes and Promonocytic Cell Lines.” Journal of Virology 72, no. 12 (December 1, 1998): 9881–88. doi:10.1128/JVI.72.12.9881-9888.1998.
12. Sung, Tzu-Ling, and Andrew P. Rice. “Effects of Prostratin on Cyclin T1/P-TEFb Function and the Gene Expression Profile in Primary Resting CD4+T Cells.” Retrovirology 3, no. 1 (October 2, 2006): 66. doi:10.1186/1742-4690-3-66.
13. Ramakrishnan, Rajesh, Eugene C. Dow, and Andrew P. Rice. “Characterization of Cdk9 T-Loop Phosphorylation in Resting and Activated CD4+ T Lymphocytes.” Journal of Leukocyte Biology 86, no. 6 (2009): 1345–50. doi:https://doi.org/10.1189/jlb.0509309.

Index

• 1.1 James D. Watson — Welcome
• 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
• 4.1 Ronald Desrosiers — The Origin of SIVmac: Non-human Primate Models for HIV
• 5.1 Flossie Wong-Staal — Discovery of Human Retroviral Transactivators
• 5.2 Joseph Sodroski — Primate Host-Specific Selection of Immunodeficiency Virus Gag and Env Proteins
• 5.6 Michael Emerman — Host-virus Co-evolution
• 8.1 John Mellors — MACS and Beyond: Epidemiology, Viremia and Pathogenesis
• analytical separation
• Baylor College of Medicine
• blood — banks, donors, plasma, screening, transfusions, clotting factors (factor VIII), PBMCs
• capsid, capsid protein (p24)
• Carter, Carol A.
• Cell (journal)
• Cold Spring Harbor Laboratory (CSHL)
• credit, priority
• Cullen, Bryan R. (b. 1951)
• cure vs. remission of HIV/AIDS
• cyclin T1
• E. coli
• education and early career
• funding and grants
• Genes & Development (journal)
• Gold, Moses O.
• GST (glutathione S-transferase), GST pull-down assay
• Haseltine, William A. (b. 1944)
• Herrmann, Christine H.
• Holland, Eric C.
• HTLV (human T-lymphotropic virus)
• Imperial Cancer Research Fund (ICRF)
• in vitro vs. in vivo
• interferons
• Jones, Katherine A.
• Karn, Jonathan
• Kerr, Ian
• kinases
• LTR (long terminal repeat)
• macrophage
• mass spectrometry
• Mathews, Michael B.
• mechanism
• Medical Research Council (MRC)
• models (model systems, model organisms, modeling)
• nuclear run-on assay
• pharmaceutical industry
• Price, David H.
• provirus
• radionuclide, radiolabeling, radioactive tracer
• Rev response element (RRE)
• Rosen, Craig A.
• Salk Institute for Biological Studies
• Session 5: Molecular Biology of the Extraordinary Virus
• Session 8: Pathogenesis and Prospects
• Stanford University, Stanford University School of Medicine, Stanford University Medical Center
• Stony Brook University, Renaissance School of Medicine
• tat
• tax
• trans-activation response element (TAR)
• transactivation
• transfection, transduction, viral vector
• UCSF (University of California San Francisco)
• University of Cambridge
• Valente, Susana
• viral reservoir, viral latency, disease reservoir