 |
|
|
| |
IAS Towards an HIV Cure Symposium, 29th-30th June 2013, Kuala Lumpur
|
| |
| |
Written for NATAP by David Margolis MD, UNC Chapel Hill and the Collaboratory of AIDS Researchers for Eradication (CARE)
Dr. Francoise Barre-Sinoussi has made "TOWARDS AN HIV CURE" the signature initiative of her tenure as president of the IAS. The most recent fruit of this effort was a 2-day pre-conference symposium held immediately prior to the 7th IAS Conference on HIV Pathogenesis, Treatment and Prevention in Kuala Lumpur on 29-30 June 2013. The symposium provided a platform to present state-of-the-art basic science and clinical research on HIV cure research.
Attacking persistent HIV infection
Two presentations, one in the humanized mouse model of HIV infection and the other in the real thing --- people --- headlined potential advances in the attempt to eradicate HIV infection. Martin Tolstrup reported the efforts of the group at Aarhus University in Denmark to test the effect of repeated cyclic dosing of the HDAC inhibitor Panobinostat in HIV+ patients on ART.
Panobinostat (PAN) is the 2nd selective and potent HDAC inhibitor tested in HIV+ patients, a drug developed by Novartis and currently under study to treat multiple myeloma and acute leukemia. Like vorinostat, it targets the Class I HDACs -1, -2, and -3 thought to be central to the maintenance of HIV latency, but is more potent than VOR on a molar basis.
To avoid low platelet counts seen when the drug is used in oncology, the Danish group elected to give 12 doses of PAN 20 mg three times a week every other week, 4 cycles over 8 weeks. Fifteen patients, all male, with medians of age 47 years, HIV diagnosis 82 months, HIV therapy 43 months, CD4 nadir 350, and CD4 at baseline 935, were studied. 2 patients had initiated ART during acute infection (infection duration undefined).
PAN appeared, like VOR, to be well tolerated. Only mild Grade I events were reported, CD4 counts were stable, and platelet counts trended down over longer times of exposure but never met a gradable toxicity criteria. This appears very similar to the experience with VOR.
HDAC inhibitors increase the amount of protein within a cell that is modified by the addition of an acetyl group (acetylation). Therefore total cellular acetylation can be used as a marker of biological effect to prove that the HDAC drug got inside a cell and did its job. The group is to be applauded for its careful study of total cellular acetylation in PBMCs as a biomarker of PAN effect within cells, as was done with the single-dose study of vorinostat. Acetylation of histone H3 was seen to increase 2-3-fold within 2 hours of dosing, and appeared to generally persist at 2 hours, and likely out past 52 hours. In fact the level of global acetylation within cells did not appear to have totally returned to baseline prior to the next round of cyclic dosing. Unfortunately the group did not have or could not yet present data collected on single copy plasma HIV RNA, or cell-associated HIV RNA. Presumably we will hear this at the next meeting.
However, tantalizingly, the group reported some preliminary evidence that PAN might have perturbed latent HIV infection using a transcription-mediated amplification (TMA) assay. TMA is a technology different than PCR that detects DNA or RNA. It can be very sensitive, but is not as precise as PCR. However, it is a good tool for discerning "presence" or "absence" of a target, such as HIV RNA. The TMA assays used here were reported to be 50% sensitive for the detection of 3.8 copies of HIV RNA/ml, and 95% sensitive for the detection of 12 c/ml. During the study, at various points after PAN had been given, 60% of participants and positive TMA tests for HIV RNA in their blood, compared with only 28% before panobinostat. Only one of the 15 participants showed no positive TMA tests throughout the study period. So if a negative TMA test is mostly less than 3 copies and a positive test is mostly more than 3 copies or as much >12 copies, we are talking about pretty tiny amounts of HIV RNA and tiny amounts of change. The results of the single copy plasma HIV RNA, or cell-associated HIV RNA assay could really clarify things. It would be a great interest to really verify that a 2nd HDAC inhibitor could ("kick") perturb latent HIV infection.
In a poster presentation, Wightman et al. (Monash University & Burnet Institute, Australia) suggested that a third HDAC inhibitor, Entinostat, might also be useful. Entinostat also selective for class 1 HDACs, and in a model of latency developed using primary cells, not a cell line, the drug induced the expression of latent virus with no changes in markers of T-cell activation or expression of the HIV co-receptors CXCR4 or CCR5. The authors stated that this compound was an attractive novel option for future clinical trials.
In a very important study Victor Garcia of UNC demonstrated the ability of a toxin linked to an HIV envelope antibody to clear residual cells producing HIV in a humanized mouse model. In full disclosure, this writer's laboratory assisted with some assays used in this study. In a complex study led by Paul Denton, humanized BLT mice were constructed, HIV infected, and treated with ART, or left untreated. In this model, an immunodeficient mouse, unable to reject transplanted cells, if reconstituted with a full human immune system derived from stem cells and a thymic organoid. In this system, HIV infection, ART, and latent infection during ART has already been validated in various publications from Garcia's lab and others. Garcia sent samples to Angela Kashuba's laboratory, were it was demonstrated carefully that ART (raltegravir, tenofovir, FTC) fully penetrated all the relevant tissues studied within the mouse.
As with humans, after several weeks of ART, HIV RNA is undetectable in the plasma, but HIV RNA+ cells are still abundant in tissues (blood, marrow, thymus, liver/spleen, lungs, GALT). This situation mimics the so-called "residual active reservoir" seen in humans on ART, wherein rare cells are expressing HIV RNA, although new rounds of infection and viral replication is blocked by ART. Garcia's group sought to test the efficacy of a novel targeted cell killing approach to deplete productively infected cells in vivo, using the immunotoxin 3B3-PE38 targeting HIV Env (by 3B3 antibody) that carries the potent toxin produced by Pseudomonas bacteria (PE38). Immunotoxin or vehicle control was administered in addition to ART every other day for 14 days to animals. Animals then sacrificed and levels of residual RNA in tissues were compared between HIV-infected but untreated, ART treated, and ART/immunotoxin treated mice. While an immunotoxin approach in humans will be challenged by the vast diversity of viral envelopes presented by an infecting HIV swarm, in this model system the humanized mouse is infected with a single clonal viral inoculum, and so targeting by 3B3 is possible.
ART alone strongly suppressed HIV RNA levels in the plasma and tissue, but
HIV cell-associated RNA expression in tissues persisted. The immunotoxin further reduced RNA levels measured by PCR up to 3.2 logs in individual tissues. This result was confirmed by in situ hybridization that directly visualizes HIV RNA molecules within infected cells.
This finding sets the stage for several important experiments in the future. It demonstrates the ability of a reagent to clear persistently infected cells in a measurable way in this tractable small animal model. And further, while the 2nd type of persistent infection --- the latent, quiescent viral reservoir as opposed ot the residual active reservoir --- was not studied in this experiment, it sets the stage for the testing of latency induction strategies, followed by strategies to clear residual active expression.
The Boston Cure patients
The important stories of the Berlin patient, followed by the Visconti cohort, and the Mississippi baby have taught us much, stretched our imagination, and raised hopes for the research towards HIV curative strategies. Tim Henrich of Brigham and Women's Hospital and Harvard Medical School added another chapter of discoveries from the clinic this with studies of HIV-1 reservoirs and other phenomenon in unique HIV+ patients treated in Boston for malignancies with allogeneic stem cell transplantation.
Earlier, Henrich previously reported very significant depletion of HIV DNA within peripheral blood cells in 3 individuals following reduced-intensity conditioning allogeneic hematopoetic stem cell transplantation (RIC-alloHSCT) from donors that had "normal" (wild-type) CCR5 receptor genes, and whose cells were susceptible to HIV infection, unlike the cells transplanted in the Berlin patient case.
At 4.3 years after transplantation, one patient was completely transplanted, with only donor cells detected, and no HIV DNA in two pools of 25 million PBMCs that were tested, a reduction of HIV DNA of at least 1500-fold compared to the level of DNA prior to transplant.
At 2.6 years after transplantation the second patient was completely transplanted, with only donor cells detected, and no HIV DNA in one pool of 50 million PBMCs that were tested, and no HIV recovered after the culture of 150 million PBMCs. A third patient died of underlying cancer 188 days after transplant, with declining levels of HIV DNA.
Both patients had no detectable HIV-specific cellular immune function by ELISpot IFN-gamma screenings of total PBMCs involving comprehensive HLA-specific peptide panels. But they did have significant chronic graft-vs-host disease (GVHD), that is the patients engrafted immune systems were exhibiting a chronic active immune response against the patient's native cells. One interpretation of these events is that it is the GVHD that is seeing all HIV+ patient cells and killing them, resulting in cure as a side effect of chronic GVHD.
As anticipated since the last presentation of this experience, Henrich then reported that the patients had both elected to undergo an observed, IRB-approved interruption of antiretroviral therapy. At the time of the meeting, both patients were still without detectable HIV RNA on DNA in PBMCs upon weekly sampling for 7 and 14 weeks respectively. Speaking with Dr. Henrich about a month later, this is still the case. Further follow-up and study is ongoing.
ART and Cure
Several studies highlighted a role of early and/or intensive ART in eradication strategies, or at least in shrinking the target viral reservoir. The work of Cheret, Rouzioux , and others illustrated the impact of 12 months of ART on cell-associated HIV-DNA in acute primary HIV-1 infection in the OPTIPRIM ANRS 147 trial. 90 patients were treated a median of 35 days from the estimated date of infection, 43% had HIV1 Western-Blot with ≤1 antibody; 92% were male and 96% had symptoms. At baseline median values for CD4, HIV-RNA and HIV-DNA were 472 cells/mm3 [IQR: 368-640], 5.4 log copies/ml [IQR: 4.9-5.8], and 3.65 log copies/million PBMC [IQR: 3.35-4.02], respectively. Patients received raltegravir, maraviroc, darunavir/r, emtricitabine/tenofovir (Arm 1) or darunavir/r, emtricitabine/tenofovir (Arm 2). The combined results in both arms at 12 months showed 239 cells/μl rise in CD4s, plasma HIV-RNA was < 50 cp/ml in 91%, and a large decrease in HIV DNA in PBMCs. Comparative data will be analyzed at 24 months.
In another presentation from Wolf in Munich, 5-drug ART during primary HIV infection led to a reduction of proviral DNA levels greater than that typically seen after treatment in chronic infection. Similarly, ART included 2 NRTIs +1 PI +Maraviroc (MVC) +Raltegravir (RAL). In a parallel cohort of chronically infected patients ART was intensified with MVC+RAL. At month 24, median proviral DNA levels were significantly lower in PHI pts (2.1 vs. 2.6 log cp/106 PBMC, p=0.001). The SPARTAC trial also showed that HIV-1 DNA levels after antiretroviral therapy in primary infection predict disease progression. The HIV-1 DNA reservoir level after 48-weeks of treatment strongly predicted disease progression, with
total HIV-1 DNA levels being more predictive than integrated HIV-1 DNA levels.
Laurent Hocqueloux reported the findings of an ARNS study that showed that chronically infected patients who started ART above 500 CD4/mm3 achieved better higher CD4 counts and lower HIV DNA levels. 309 patients were included after a median of 4 years of ART. Patients on ART before CD4s < 500 cells/μl had significantly lower HIV-DNA level (median=2.50 log).
The benefits of early therapy seem clear and universal, when it can be successfully implemented. Such patients would be prime candidates for future curative efforts.
Measurements of HIV infection
Sarah Palmer, from the Westmead Millenium Institute for Medical Research and University of Sydney, Australia reported an analysis of HIV infection frequencies and genetic make-up of intracellular HIV-1 in the tissue compartments of patients treated after acute and chronic infection. Using single-genome and single-proviral sequencing techniques, intracellular HIV-1 genomes from defined naïve, central-, transitional-, and effector-memory T cells in peripheral blood, GALT, and lymph node tissue were sequenced. After 4-12 years of suppressive therapy samples were collected 6 months apart from 5 subjects who initiated therapy during acute infection and 3 who initiated therapy during chronic infection. The major findings were:
· Frequencies of infection in each population did not change much over time in most patients
· Infection frequencies of all T-cell subsets were higher in subjects treated during chronic infection than acute infection, with central memory carrying most of the HIV DNA. In GALT effector T cells carried the most HIV DNA
· There was little evidence of sequence changes (evolution) of HIV DNA in blood, LN, GALT in central, effector, naïve, and transitional memory cell populations
· The increase of clonal HIV-1 sequences from effector-memory T-cells in 2 subjects, especially in one patient with a replication-incompetent HIV species carrying a large deletion in protease, indicates an expansion of cells with integrated proviral DNA rather than active viral replication.
Overall, these findings again reinforced the idea that successful suppressive ART fully blocks HIV replication, as defined by full rounds of the viral lifecycle necessary to generate genetic diversity and evolution. This clear data has not yet ended the debate between the camp (including the author) of those that feel that ART has reached its maximum virological effectiveness (and that more or better therapy will not cure HIV infection), and those that view part of the problem of persistent HIV infection as one of "persistent replication."
Richard Koup presented the work of his colleague Joe Casazza in the Vaccine Research Center, studying the details of CD4+ T cells isolated from peripheral blood mononuclear cells (PBMCs) that are HIV RNA +, and therefore likely to be actively transcribing HIV RNA. HIV RNAs come in two major flavors: spliced HIV RNAs that represent the early stages of viral expression, and unspliced, long HIV RNAs that represent the later stages of viral replication that can result in virion production, and expression of HIV antigens on the infected cell surface. Using specially designed PCR primers, these two general species of RNA can be distinguished, identifying which cells are brewing HIV but might be less visible to the immune system, and which cells may be producing virions and might be cleared by an immune response.
Live CD8-negative PBMCs from 6 HIV-infected individuals not on ART were sorted into CD4+ bright, CD4+ dim, and CD4-negative populations. This was an important feature, as it is known that CD4+ cells can lose the expression of CD4 on their surface in the productive phases of HIV infection, due to the action of the viral protein Nef. Surface markers of T cell activation and surface expression of HIV antigens were used to further classify cells expressing HIV. The frequency of T cells containing HIV RNA, and the quantity of HIV RNA in each population was determined using limiting dilution RT qPCR. HIV RNA copy number per cell was determined using values from wells near the endpoint of dilution that were likely to containing a single HIV RNA+ cell. Measurable spliced RNA was used as evidence of active transcription from proviral DNA.
The median frequency of cells containing unspliced RNA in the CD4 bright population was about 1:2500 significantly greater than the frequency of spliced HIV RNA in the same population, about 1:5000. In the CD4 dim population there was routine co-expression of spliced and unspliced RNAs with a median frequency of about 1 in 60 cells. Median spliced and unspliced RNA copy number increased with decreasing surface CD4 T cell expression, as might be expected. When cells expressing markers of T cell activation were studied, up to 1 in 30 (3.3%) were actively transcribing HIV.
However, it was said that it was very difficult using broadly neutralizing env-specific monoclonal antibodies to identify HIV RNA + cells directly ex vivo. The message seemed to be that although it might be possible to detect HIV RNA-positive cells, it was much more difficult to detect HIV antigen-positive cells using current antibody techniques. This may pose a challenge to the detection of infected cells in ART-treated patients during attempts to pursue viral purging/eradication strategies.
Una O'Doherty of the University of Pennsylvania reported better success in detection of HIV Gag antigen-positive reservoir cells. She stained phenotypically resting CD4+ T cells (HLA-DR-, CD25-, CD69-) from HIV infected individuals on ART with an antibody against HIV Gag and sorted for Gag+ and Gag- cells. She then measured HIV DNA in the Gag+ and Gag- fraction to determine if HIV DNA was enriched in the Gag+ population. Higher levels of HIV DNA were measured in Gag+ cells in all 5 patients in this preliminary study. Gag+ cells were calculated to represent up to 2% of the resting CD4+ cells that were also HIV DNA+. It would be important to learn, if possible, how much of the 98% of cells that were Gag-negative could become Gag+ upon stimulation, and what proportion of these cells (both Gag+ without activation, and Gag+ only after activation) produced replication-competent HIV.
O'Doherty also presented evidence that cells that are Gag+ can be cleared by HIV-specific cytotoxic T cells (CTLs). Finally, in a cell culture model system, the efficiency of CTL clearance of Gag+ cells vitro model correlated strongly with the frequency of integrated HIV DNA in that patient's cells in vivo. She concluded that a fraction of the reservoir in resting CD4+ T cells expresses HIV Gag proteins in patients on ART, that such cells may be cleared by CTL, and that patients with more efficient CTL should have a smaller reservoir size.
Mechanistic studies
Several interesting mechanistic studies illustrated potential future targets for anti-latency therapy, pointing to cellular mechanisms that may enforce latent infection and maintain reservoirs of dormant HIV infection despite ART. Targeting such mechanisms may force latent virus out of hiding. Mathias Lichterfeld of the Massachusetts General Hospital proposed that latent HIV-1 infection was driven by the host factor beta-catenin in a relatively new type of CD4 T memory cell called a memory stem cell. Further, he doubled down on this by proposal in suggesting that drug inhibitors of beta-catenin could purge such latent reservoirs.
CD4+ T memory stem cells (Tscm) represent a recently discovered subpopulation of T cells that appear to be long-lived. These cells are thought to be maintained by a cell signaling pathway that is involves a cascade of cell factors known as the Wnt/beta-catenin pathway. This is a complex pathway built of many human cellular factors, and one that regulates many human genes.
Beta-catenin with its coactivator CBP (the binding protein of the cAMP response element-binding protein CREB); together with other transcription factors binds to places in the human genome that contain a DNA sequence called the Wnt-responsive element, and activates transcription of a wide range of target genes. The Wnt/beta-catenin signaling pathway regulates cell morphology, motility, and proliferation; aberrant regulation of this pathway leads to neoplastic proliferation (ie. cancer). Under study to treat cancers thought to be caused by abberant Wnt signaling, the experimental inhibitor PRI-724 blocks this pathway in cancer stem cells and appears to have antineoplastic activity.
Lichterfeld tested the idea that this inhibitor could disrupt HIV latency, and used PCR assays to determine the amount of cell-associated HIV-1 DNA in sorted CD4 T cell subsets from ART-treated HIV-1+ patients. He found high levels of HIV DNA in Tscm in patients, and the presence of replication competent virus within the CD4 T cell subsets using viral reactivation assays. The recovery of HIV from cells exposed to PRI-724 suggested that pharmaceutical beta-catenin inhibitors might have an adjunct or additive role for reducing long-term viral persistence in CD4 Tscm.
Examining a more typical population of cells, the canonical central memory CD4+ T cell, Tokameh Mahmoudi of the Erasmus Medical Centre, Netherlands also showed evidence that activation of the Wnt pathway by natural ligands or small molecule inhibitors could induce the expression of latent HIV. Her work was done in classical laboratory cell line models of latency, the J-Lat 1.1 and S-Lat30 cell lines. While these lines are sometimes criticized as representing laboratory artifacts rather than the true character of the latent reservoir in patients, many studies in these simple models have been later borne out in more complex ones or in cells from patients. Mahmoudi found that treatment of latently HIV infected cells with activators of the Wnt pathway resulted in recruitment of TCF/LEF and beta-catenin, the molecular effectors of Wnt signaling, to the latent HIV LTR. This event was measured using chromatin immunoprecipitation, a gold-standard laboratory method for showing the occupancy of a specific cellular factor at a specific gene regulatory region, or promoter. Wnt-mediated activation of the latent HIV LTR was synergistically enhanced in the presence of histone deacetylase inhibitors, already known the ability to induce latent HIV. She also confirmed these studies in a model of latency developed by infecting primary CD4 cells in the laboratory, as described by Bosque and Planelles. Mahmoudi studies reinforced those of Lichterfeld, However, the ultimate value of this approach remains to be fully validated in such studies using highly specific small-molecule inhibitors and molecular tools in primary cells, preferably from HIV+ patients, rather than the non-specific reagents used to induce Wnt signaling (lithium) and HDAC inhibition (valproic acid).
Overall, the IAS Cure Symposium in KL had some excellent science and was well worth the trip. It is exciting to see the field continue to expand and move steadily forward.
|
| |
|
|
|
|
|
