The persistence of a latent form of HIV integrated into the genome of long-lived cells, particularly memory CD4 T cells, is considered the most significant barrier to curing the infection. A key element of cure research is the pursuit of strategies capable of inducing latent HIV to replicate, with the aim of causing the death of the latently infected cell either as a result of viral cytopathic effects or recognition and elimination by HIV-specific immune responses (or other targeting strategies). A number of candidate latency-reversing agents (LRAs) have emerged from studies, with the anticancer drugs HDAC inhibitors leading the pack. Two clinical trials of the HDAC inhibitor vorinostat have indicated that it is capable of inducing latent HIV to begin making viral RNA, but whether subsequent steps in the viral life cycle ensued—the manufacture of HIV proteins and their assembly into infectious virions—is not yet known. Two recently published studies set out to better understand the capabilities of the current crop of LRAs, including HDAC inhibitors, and report that they may not be up to the task, at least as single agents.
C. Korin Bullen and colleagues from the laboratory of Robert Siliciano at Johns Hopkins University tested the activity of a panel of LRAs including the HDAC inhibitors vorinostat, romidepsin, and panobinostat, along with three compounds with differing mechanisms: disulfiram (acetaldehyde dehydrogenase inhibitor), JQ1 (bromodomain inhibitor) and Bryostatin 1 (protein kinase C agonist). Results were compared to those obtained with PMA/ionomycin, which maximally induces replication of latent HIV by activating all CD4 T cells. A critical aspect of the study is that it was carried out using CD4 T cells sampled from HIV-positive individuals on ART; most prior research on LRAs has relied on laboratory models of HIV latency.
The researchers first employed an assay that measures viral outgrowth, and were able to demonstrate HIV production by CD4 T cells from 11 of 13 donors after stimulation with PMA/ionomycin. In contrast, no virus production was detected after exposure to any of the LRAs. Similarly, extracellular HIV messenger RNA could be detected in most cases after PMA/ionomycin stimulation, but in only one instance with a single LRA (Bryostatin 1). Lastly, HIV reactivation was assessed with an assay that quantifies intracellular viral messenger RNA; levels were detectable at baseline in 10 of 11 samples, and only increased significantly in response to Bryostatin 1 and PMA/ionomycin.
Since an increase in intracellular HIV messenger RNA has been reported in clinical trials of vorinostat, additional experiments were conducted to try and understand the basis for the discrepant results. The researchers note that HDAC inhibitors may activate human genes into which HIV has integrated, leading to the generation of some viral RNA but not the specific activation of the virus itself. Viral RNA generated by this mechanism is referred to as a “readthrough transcript,” because it reflects a reading of the HIV genetic code by the human gene (as opposed to HIV transcribing its own genes in order to assemble new virions). Using an assay designed to evaluate this possibility, the researchers found that the approximately twofold increase in levels of HIV gag RNA seen with vorinostat could be accounted for by an increase in readthrough transcripts, and did not appear to represent transcription of viral RNA initiated by HIV.
The overall conclusions drawn by the authors are fairly grim: “Our data demonstrate that none of the leading candidate non–T cell–activating LRAs tested significantly disrupted the latent reservoir ex vivo. The discordance between the effects of nonstimulating LRAs in in vitro models of HIV-1 latency and their effects ex vivo in resting CD4+ T cells from infected individuals on ART indicates that these models do not fully capture all mechanisms governing HIV-1 latency in vivo.” However, when second author Gregory Laird presented the findings recently at CROI (see webcast), he added the information that combinations of LRAs have since shown signs of promise; both romidepsin plus Bryostatin 1 and panobinostat plus Bryostatin 1 induced HIV RNA production to levels similar to PMA/ionomycin. Although it’s not clear if Bryostatin 1 is safe enough to enter human trials for HIV, several modified analogs are in development.
The second paper, published in PNAS by Anthony Cillo and colleagues from John Mellors laboratory, compared the activity of vorinostrat and CD4 T cell activation (using ant-CD3 and anti-CD28 stimulation), also using resting CD4 T cells isolated from individuals on ART. In this case, the researchers evaluated the proportion of integrated HIV that showed a response to each intervention. CD4 T cell activation led to HIV RNA production by 1.5% of the integrated viruses, whereas the proportion for vorinostat was 0.079% (not significantly higher than the background levels absent any intervention). While the relatively small proportion of latent HIV induced by CD4 T cell activation is partly explained by the presence of defective viruses, the authors note the results echo another recent study in suggesting the presence of a substantial HIV reservoir refractory to any currently used in vitro stimulus. Whether these viruses can be activated in vivo is not known. The paper concludes: “Identifying strategies to increase fractional proviral expression without inducing global T-cell activation is critical to achieving progress toward a cure of HIV-1 infection.”
Update 2/10/14: An hour or so after this post was completed, PLoS Pathogens published a study by Datsen George Wei and colleagues reporting that the HDAC inhibitor romidepsin is capable of inducing virion production by latently infected CD4 T cells sampled from individuals on ART (abstract and link to full text added below, see figure 3 and table S2). Results are also reported for vorinostat, but those seem compatible with the Bullen and Cillo papers described above. I'm not sure what the explanation might be for the seeming conflict with the results reported by Bullen et al for romidepsin, although perhaps they might relate to differences in duration of in vitro exposure to the drug as the most consistent effects reported in the PLoS Pathogens paper appear to occur after six days whereas Bullen et al used 18 hours. The proportion of latent HIV induced to produce virus by romidepsin is not reported. Two clinical trials are assessing the activity of romidepsin in HIV-positive individuals (see TAG's clinical trials table) so hopefully clearer answers about its activity will be forthcoming.
Nat Med. 2014 Mar 23. doi: 10.1038/nm.3489. [Epub ahead of print]
Bullen CK, Laird GM, Durand CM, Siliciano JD, Siliciano RF.
Abstract
HIV-1 persists in a latent reservoir despite antiretroviral therapy (ART). This reservoir is the major barrier to HIV-1 eradication. Current approaches to purging the latent reservoir involve pharmacologic induction of HIV-1 transcription and subsequent killing of infected cells by cytolytic T lymphocytes (CTLs) or viral cytopathic effects. Agents that reverse latency without activating T cells have been identified using in vitro models of latency. However, their effects on latently infected cells from infected individuals remain largely unknown. Using a new ex vivo assay, we demonstrate that none of the latency-reversing agents (LRAs) tested induced outgrowth of HIV-1 from the latent reservoir of patients on ART. Using a quantitative reverse transcription PCR assay specific for all HIV-1 mRNAs, we demonstrate that LRAs that do not cause T cell activation do not induce substantial increases in intracellular HIV-1 mRNA in patient cells; only the protein kinase C agonist bryostatin-1 caused significant increases. These findings demonstrate that current in vitro models do not fully recapitulate mechanisms governing HIV-1 latency in vivo. Further, our data indicate that non-activating LRAs are unlikely to drive the elimination of the latent reservoir in vivo when administered individually.
Published online before print March 31, 2014, doi: 10.1073/pnas.1402873111
PNAS March 31, 2014
Anthony R. Cillo, Michele D. Sobolewski, Ronald J. Bosch, Elizabeth Fyne, Michael Piatak, Jr., John M. Coffin, and John W. Mellors
Contributed by John M. Coffin, February 19, 2014 (sent for review September 26, 2013)
Abstract
Reversal of proviral latency is being pursued as a curative strategy for HIV-1 infection. Recent clinical studies of in vivo administration of the histone deacetylase inhibitor suberoylanilide hydroxamic acid (SAHA; vorinostat) show increases in unspliced cellular HIV-1 RNA levels in resting CD4+ T cells. A critical unknown, however, is the proportion of latent proviruses that can be transcriptionally reactivated by SAHA or T-cell activation. In this study, we quantified the fraction of HIV-1 proviruses in resting CD4+ T cells from patients on suppressive antiretroviral therapy that were reactivated ex vivo with SAHA or antibodies to CD3/CD28. At concentrations of SAHA achieved clinically, only 0.079% of proviruses in resting CD4+ T cells were reactivated to produce virions, compared with 1.5% of proviruses in cells treated with anti-CD3/CD28 antibodies after correcting for spontaneous virion production in the medium control. A significant positive correlation (ρ = 0.67, P < 0.001) was found between levels of virions in the supernatant and unspliced cellular HIV-1 RNA following anti-CD3/CD28 treatment, but not following SAHA treatment (ρ = 0.21, P = 0.99). These results reveal that the majority of HIV-1 proviruses are not reactivated by current therapeutic approaches and that more effective means of reversing proviral latency will likely be required to deplete HIV-1 reservoirs.
PLoS Pathog 10(4): e1004071. doi:10.1371/journal.ppat.1004071
Datsen George Wei, Vicki Chiang, Elizabeth Fyne, Mini Balakrishnan, Tiffany Barnes, Michael Graupe, Joseph Hesselgesser, Alivelu Irrinki, Jeffrey P. Murry, George Stepan, Kirsten M. Stray, Angela Tsai, Helen Yu, Jonathan Spindler, Mary Kearney, Celsa A. Spina, Deborah McMahon, Jacob Lalezari, Derek Sloan, John Mellors, Romas Geleziunas, Tomas Cihlar
Published: April 10, 2014DOI: 10.1371/journal.ppat.1004071
Abstract
Persistent latent reservoir of replication-competent proviruses in memory CD4 T cells is a major obstacle to curing HIV infection. Pharmacological activation of HIV expression in latently infected cells is being explored as one of the strategies to deplete the latent HIV reservoir. In this study, we characterized the ability of romidepsin (RMD), a histone deacetylase inhibitor approved for the treatment of T-cell lymphomas, to activate the expression of latent HIV. In an in vitro T-cell model of HIV latency, RMD was the most potent inducer of HIV (EC50 = 4.5 nM) compared with vorinostat (VOR; EC50 = 3,950 nM) and other histone deacetylase (HDAC) inhibitors in clinical development including panobinostat (PNB; EC50 = 10 nM). The HIV induction potencies of RMD, VOR, and PNB paralleled their inhibitory activities against multiple human HDAC isoenzymes. In both resting and memory CD4 T cells isolated from HIV-infected patients on suppressive combination antiretroviral therapy (cART), a 4-hour exposure to 40 nM RMD induced a mean 6-fold increase in intracellular HIV RNA levels, whereas a 24-hour treatment with 1 µM VOR resulted in 2- to 3-fold increases. RMD-induced intracellular HIV RNA expression persisted for 48 hours and correlated with sustained inhibition of cell-associated HDAC activity. By comparison, the induction of HIV RNA by VOR and PNB was transient and diminished after 24 hours. RMD also increased levels of extracellular HIV RNA and virions from both memory and resting CD4 T-cell cultures. The activation of HIV expression was observed at RMD concentrations below the drug plasma levels achieved by doses used in patients treated for T-cell lymphomas. In conclusion, RMD induces HIV expression ex vivo at concentrations that can be achieved clinically, indicating that the drug may reactivate latent HIV in patients on suppressive cART.
Very nice summary of these important reports, Richard.
I think one of the main differences in the two approaches was that in the Bullen paper they developed a PCR assay that discriminates between host-initiated and HIV LTR-initiated transcription. I think the rationale for this assay is that if viral transcripts are initiated by upstream host gene promotors, the viral transcript will not be complete and so can not lead to functional virions (hope I have understood that correctly?). Most viral PCR assays only measure part of the gag gene or gag and LTR (which are proximal), but are not designed to detect host gene fragments upstream of initiation. Therefore viral transcripts detected by conventional HIV PCR assays, may well include partial fragments that have been initiated by the activation of host genes. These will not reflect genuine full viral transcripts. The assays used in the other papers reporting an effect of HDAC inhibitors, presumably measure HIV using conventional PCR assays for gag and LTR sequences, so they would fall foul of this.
Why does it matter? Because the mechanism of action of HDAC inhibitors is to "release" genes from a "silent" state to enable their expression. This will effect host genes as well as HIV. Therefore HDAC-inhibitors are liable to initiate HIV transcription by activating upstream host genes. As such this means that the detection of HIV gag sequences by PCR in HDAC treated cells may in fact be a "false" signal.
This all leads me to think that the Bullen paper is more robust, whereas the Wei paper may be on a false lead. What flies in the face of this is the fact that Wei claim to have pelleted HIV virions from supernatant of HDAC-inhibitor treated CD4 cells. They could identify the HIV sequences in these virion pellets as they could in the host cells. This would suggest that mature HIV virions have been formed as a result of latent viral activation by HDAC inhibitors.
However it is not impossible that this is still wrong. I'm not going to speculate about how likely this is, but extracellular forms of RNA can be found that derive from host cells, as part of normal cell biology, contained in extracellular vesicles or exosomes. These are sadly about the same size and density as HIV virions and so very hard to distinguish by centrifugation. So I would argue that, sadly, the presence of "virions" in the supernatant of HDAC inhibitor-treated CD4 cells in the Wei paper, does not 100% prove that HDAC inhibitors are inducing viral activation.
However the fact that Gregory Laird (from the Bullen group) subsequently presented data claiming an effect from combined LRAs, presumably with the same host-viral PCR assay, is perhaps cause for more optimism.
Posted by: Gareth Hardy | May 18, 2014 at 10:52 AM