ORIGINAL RESEARCH published: 20 March 2020 doi: 10.3389/fimmu.2020.00418 In vivo Effects of Romidepsin on T-Cell Activation, Apoptosis and Function in the BCN02 HIV-1 Kick&Kill Clinical Trial Miriam Rosás-Umbert 1,2 , Marta Ruiz-Riol 1 , Marco A. Fernández 3 , Marta Marszalek 1 , Pep Coll 1 , Christian Manzardo 4 , Samandhy Cedeño 1 , José M. Miró 4 , Bonaventura Clotet 1,5,6,7 , Tomáš Hanke 8,9 , José Moltó 5,6 , Beatriz Mothe 1,6,7*, Christian Brander 1,7,10 and the BCN02 study group 1 IrsiCaixa AIDS Research Institute-HIVACAT, Badalona, Spain, 2 Department of Cellular Biology, Physiology and Immunology, Universitat Autònoma de Barcelona (UAB), Barcelona, Spain, 3 Flow Cytometry Facility, Health Sciences Research Institute Germans Trias i Pujol, Badalona, Spain, 4 Hospital Clinic- IDIBAPS, University of Barcelona, Barcelona, Spain, 5 Fundació Lluita contra la Sida, Hospital Universitari Germans Trias i Pujol, Badalona, Spain, 6 Department of Infectious Diseases, Hospital Germans Trias i Pujol, Badalona, Spain, 7 Centre for Health and Social Care Research (CESS), Faculty of Medicine, University of Vic – Central University of Catalonia (UVic – UCC), Vic, Spain, 8 The Jenner Institute, University of Oxford, Oxford, United Kingdom, 9 Joint Research Center for Human Retrovirus Infection, Kumamoto University, Kumamoto, Japan, 10 ICREA, Pg. Lluis Companys, Barcelona, Spain Edited by: Carolina Garrido, Romidepsin (RMD) is a well-characterized histone deacetylase inhibitor approved for the University of North Carolina at Chapel treatment of cutaneous T-cell lymphoma. in vitro and in vivo studies have demonstrated Hill, United States that it is able to induce HIV-1 gene expression in latently infected CD4+ T cells from Reviewed by: HIV-1+ individuals on suppressive antiretroviral therapy. However, in vitro experiments R. Brad Jones, Cornell University, United States suggested that RMD could also impair T-cell functionality, particularly of activated T Thomas Aagaard Rasmussen, cells. Thus, the usefulness of RMD in HIV-1 kick&kill strategies, that aim to enhance the Peter Doherty Institute for Infection and Immunity, Australia immune system elimination of infected cells after inducing HIV-1 viral reactivation, may *Correspondence: be limited. In order to address whether the in vitro observations are replicated in vivo, Beatriz Mothe we determined the effects of RMD on the total and HIV-1-specific T-cell populations bmothe@irsicaixa.es in longitudinal samples from the BCN02 kick&kill clinical trial (NCT02616874). BCN02 Specialty section: was a proof-of-concept study in 15 early treated HIV-1+ individuals that combined This article was submitted to MVA.HIVconsv vaccination with three weekly infusions of RMD given as a latency Vaccines and Molecular Therapeutics, reversing agent. Our results show that RMD induced a transient increase in the frequency a section of the journal Frontiers in Immunology of apoptotic T cells and an enhanced activation of vaccine-induced T cells. Although Received: 08 December 2019 RMD reduced the number of vaccine-elicited T cells secreting multiple cytokines, viral Accepted: 24 February 2020 suppressive capacity of CD8+ T cells was preserved over the RMD treatment. These Published: 20 March 2020 observations have important implications for the design of effective kick&kill strategies Citation: Rosás-Umbert M, Ruiz-Riol M, for the HIV-1 cure. Fernández MA, Marszalek M, Coll P, Keywords: romidepsin, HDAC inhibitor, kick&kill strategy, therapeutic vaccine, latency reversing agent (LRA) Manzardo C, Cedeño S, Miró JM, Clotet B, Hanke T, Moltó J, Mothe B, Brander C and the BCN02 study group (2020) In vivo Effects of INTRODUCTION Romidepsin on T-Cell Activation, Apoptosis and Function in the BCN02 Current antiretroviral therapy (ART) effectively suppresses HIV-1 replication in plasma, but it is HIV-1 Kick&Kill Clinical Trial. not able to completely eliminate the virus from infected individuals. Cessation of antiretroviral Front. Immunol. 11:418. treatment results in a rebound of plasma viremia within 3–4 weeks in most individuals (1). This doi: 10.3389/fimmu.2020.00418 rapid viral rebound after treatment interruption is due to the existence of a latent viral reservoir Frontiers in Immunology | www.frontiersin.org 1 March 2020 | Volume 11 | Article 418 Rosás-Umbert et al. Romidepsin Effects on Vaccine-Induced T-Cells and the inability of the immune system to effectively contain viral The proof-of-concept BCN02 trial evaluated a kick&kill replication. To date, numerous strategies have been pursued to strategy that combined the HIVconsv T-cell vaccines with achieve a functional cure or virus eradication, including early the HDACi RMD in a cohort of early-treated, HIV-1- ART initiation, ART intensification (2–6), passive administration infected individuals. Fifteen participants of BCN01 (12), of antibodies (7), therapeutic vaccination (8–13) and gene who previously received simian adenovirus-vectored vaccine therapy (14, 15), among others. ChAdV63.HIVconsv and MVA.HIVconsv, were invited 2–3 years HIV-1 kick&kill strategies are based on the use of latency later to receive two more dosing of the MVA.HIVconsv vaccine reversing agents (LRA) to induce production of HIV-1 proteins before (MVA1 ) and after (MVA2 ) three weekly-doses of RMD in latently infected cells and render these cells susceptible to (RMD1−2−3 ) followed by a monitored antiretroviral pause vaccine-induced virus-specific cytolytic T lymphocytes (CTL). (MAP) for a period of 32 weeks (NCT02616874). The combined The interest in LRA able to reactivate the latent provirus strategy was proven to be safe and vaccination was highly has increased over the past decade, with histone deacetylation immunogenic. RMD treatment resulted in marked increases inhibitors (HDACi) being some of the best characterized agents in histone acetylation and cell-associated HIV-1 RNA levels both in vitro and in vivo. The inhibitory effect of HDACi on compatible with induction of viral transcription. However, the histone deacetylation results in a higher degree of acetylated ultimate reduction of the viral reservoir in the BCN02 trial was histones, causing opening of chromosomes and increased gene overall minimal (27). Aside from effects on virus reactivation, for transcription. In ex vivo isolated cells from ART-suppressed a successful purge of the viral reservoir, it is critical that the LRA HIV-1-infected individuals, exposure to HDACi resulted in used in such strategies does not have any detrimental effects on reactivation of the integrated HIV-1 and led to viral protein the vaccine-induced immune cells (28, 29). Here, we assessed the expression by latently infected cells (16–18). Some HDACi, such in vivo impact of three weekly RMD doses on total and vaccine- as vorinostat (SAHA), panobinostat, and romidepsin (RMD) induced T cells in longitudinal samples from the BCN02 trial have also been tested for their in vivo potential to reverse (Figure 1). HIV-1 latency (19–22). RMD, a cyclic depsipeptide naturally produced by Chromobacterium violaceum, is a pan-HDACi MATERIALS AND METHODS that inhibits class I HDACs. RMD was clinically developed as an anti-cancer drug and is approved for the treatment of Study and Samples cutaneous T-cell lymphoma (23). Furthermore, RMD has been The BCN02 clinical trial (NCT02616874) was a phase I, open- shown to induce HIV-1 gene expression in latently infected label, single-arm, multicenter study in Spain (27). The study cells in vitro (18, 24) and in vivo, when administrated alone was approved by the institutional ethical review board of the (22) and in combination with therapeutic vaccine Vacc-4x (25) participating institutions (Reference Nr AC-15-108-R) and by in chronically-infected ART-suppressed individuals. Although it the Spanish Regulatory Authorities (EudraCT 2015-002300-84) was first thought that reactivation of the virus itself would lead to and was conducted in accordance with the principles of the robust immune activation and control of the rebounding virus, Helsinki Declaration and local personal data protection law it is generally accepted that prior stimulation of the immune (LOPD 15/1999). Fifteen participants were immunized with effector response by a therapeutic vaccine and/or an immune MVA.HIVconsv (MVA1 , 2 × 108 pfu intramuscularly), followed checkpoint inhibitor, may be needed in order to efficiently by three weekly-doses of romidepsin (RMD1−2−3 , 5 mg/m2 eliminate infected cells after LRA exposure (26). Therefore, body-surface area; BSA) and a second MVA.HIVconsv boost different immunotherapies are being investigated together with vaccination (MVA2, 2 × 108 pfu i.m.) before undergoing a LRAs to test their combined effect, especially combination monitored antiretroviral pause (MAP) 8 weeks later and for treatments that include T-cell vaccines. a maximum of 32 weeks. Cryopreserved peripheral blood FIGURE 1 | Study design. The BCN02 study was a single arm, open label, proof-of-concept study to address safety and effect on the viral reservoir of a kick&kill strategy combining MVA.HIVconsv vaccines with the HDACi RMD. Timepoints used for the analysis presented here are indicated for each assay by filled circles. Frontiers in Immunology | www.frontiersin.org 2 March 2020 | Volume 11 | Article 418 Rosás-Umbert et al. Romidepsin Effects on Vaccine-Induced T-Cells mononuclear cells (PBMC) were stored before, at the end and separation (MACS Milteny Biotec). CD8+ -depleted cells (CD4+ - after 8, 24 h (only for RMD1 ), and 3 and 7 days after all RMD enriched fraction) were stimulated with PHA (5 µg/ml) in doses for immunological and virological studies. RPMI plus 10% fetal bovine serum (R10) and antibiotics (penicillin 100 U/mL and streptavidin 100 µg/ml). After 3 Flow Cytometry days of stimulation, the CD4-enriched fraction was infected by Apoptosis Measurement spinoculation with HIV-1BaL and HIV-1IIIB laboratory-adapted PBMC viability was measured using a Pacific BlueTM Annexin strains at a multiplicity of infection (MOI) of 0.01 as reported V Apoptosis Detection Kit with 7-AAD (BioLegend). Lineage previously (12, 31). HIV-infected cells were cultured in duplicates surface markers (CD3, CD4, and CD8) and activation markers or triplicates in R10 medium with 20 U/ml of IL-2 in 96-well (HLA-DR, CD25, and CD69) were included in the staining. round-bottomed plates, alone or together with unstimulated Briefly, 1 × 106 of isolated PBMC were washed in PBS with CD8+ T cells obtained by positive magnetic bead separation the 1% FBS and resuspend in 100 µl of surface staining solution same day from an additional vial of frozen PBMC from screening (CD3, CD4, CD8, CD25, CD69, HLA-DR) and incubated for (week −4), 3 weeks after MVA1 (week 3, postMVA1 ), 1 week 20 min. After 2 washes with 300 µl of PBS with 1% FBS, cells after RMD3 (week 6, postRMD3 ), 4 weeks after MVA2 (week were resuspended in 100 µl of Annexin V Binding Buffer with 13, postMVA2 ) timepoints. Cultures at different CD8:CD4 ratios the corresponding Annexin V and 7-AAD. After 15 min of (E:T = 1:1 and 1:10) were harvested after 6 days. Cells were incubation, 250 µl of Binding Buffer was added to each tube and stained first with Aqua Live/Dead stained for surface markers acquired on a LSRII BD cytometer. The percentages of apoptotic (CD3 APC-H7, BD Biosciences, CD4 PerCP, BD Biosciences, and and live cells were analyzed using FlowJo software. The gating CD8 APC, BD Biosciences), then permeabilized (FIX & PERM R strategy is summarized in Supplementary Figure 1. Cell Permeabilization Kit, ThermoFisher). Cells were then fixed (FIX & PERM R Cell Permeabilization Kit, ThermoFisher) at T Cells and HIVconsv-Specific T-Cell Lineage, room temperature and finally stained with anti-Gag p24 antibody Activation and Cytokine Detection (KC-57-FITC; Beckman Coulter). CD8+ T-cell antiviral activity PBMCs were thawed and stimulated with anti-CD49d and anti- is expressed as % inhibition = [(fraction of p24+ cells in CD4+ CD28 antibodies (BD) in presence/absence of three peptides T cells cultured alone)–(fraction of p24+ cells in CD4+ T cells pools (containing 58, 54, and 54 peptides) covering the HIVconsv cultured with CD8+ T cells)]/(fraction of p24+ in CD4+ T cells immunogen protein in the presence of GolgiStop for 5 h. cultured alone) × 100. At least 100,000 cells were collected on Cultures were then stored overnight at 4◦ C until staining. a LSRII BD cytometer and analysis was performed using FlowJo Cells were stained first with a viability stain (Aqua Live/Dead 10 software. Fixable Dead Cell Stain kit, Invitrogen), followed by T cell lineage and maduration/activation markers (using anti-CD3- APC Cy7, anti-CD4 PECy5; anti-CD8 PerCP, anti-CCR7 B711, Statistical Analysis GraphPad Prism version 7 for Windows (San Diego, CA) was anti-CD45RA BV785, anti-HLA-DR BV650, anti-PD-1 BV605, used for statistical analysis. Mann-Whitney test and Wilcoxon anti-CD69 APC, and anti-CD25 PEDazzle594 chromogen- matched paired test were used for unpaired and paired conjugated monoclonal antiobodies; BioLegend) and dump comparisons, respectively. Significant values were considered channel (using anti-CD19-V450 for B-cells and anti-CD14-V450 for p <0.05. mAbs for monocytes; BioLegend) surface staining. Following the fixation and permeabilization step (Fix and Perm kit, Invitrogen), intracellular staining with conjugated antibodies RESULTS specific for cytokines (IFN-γ A700; Invitrogen, IL-2 PECy7, TNF-α FITC; BiolLegend and MIP1-β PE; RD Systems) was Increased Apoptosis After RMD Exposure performed. Approximately 105 cells were acquired on an Since RMD was previously described to have a toxic effect in LSRFortessa BD instrument, and analysis was performed using vitro (28, 29), especially on activated T cells, we first measured FlowJo 10 software. The gating strategy is summarized in the effect of RMD on cell viability both in total CD8+ and Supplementary Figure 2. CD4+ T cells and in activated, HLA-DR+ expressing cells in Intracellular cytokine staining analyses were done applying PBMC from the BCN02 participants. Viability in total T cells boolean gates in FlowJo 10, subtracting unstimulated signals was assessed by flow cytometry by Annexin V/7AAD staining using Pestle v1.7 program and represented using SPICE v5.35 before and after the three RMD doses. Increases in the number of software (provided by the National Institute of Health, Mario apoptotic CD8+ T cells (Annexin V+ ) were detected at 24 h after Roeder, ImmunoTechnology Section, Vaccine Research Center, RMD1 (Wilcoxon signed-rank, p = 0.0151) and 3 days after each NIAID, NIH, Bethesda) (30). RMD dose (Wilcoxon signed-rank, RMD1 p = 0.0413, RMD2 p = 0.0181, and RMD3 p = 0.0833, respectively, Figure 2A). A Viral Inhibition Assay similar pattern was observed in CD4+ T cells, with apoptotic CD8+ T-cell mediated viral inhibition capacity was measured at cells being significantly increased 3 days after RMD2 and RMD3 1:1 and 1:10 CD8 effector to CD4 target ratios. Cryopreserved (Wilcoxon signed-rank, RMD2 p = 0.0067 and RMD3 p = 0.0413, PBMCs were obtained from timepoints before the BCN02 respectively, Figure 2B). One week after RMD3 the levels of intervention and CD8+ cells were depleted by magnetic bead apoptotic cells in both CD8+ T cells and CD4+ were not restored Frontiers in Immunology | www.frontiersin.org 3 March 2020 | Volume 11 | Article 418 Rosás-Umbert et al. Romidepsin Effects on Vaccine-Induced T-Cells FIGURE 2 | Effect of RMD on the viability on T cells. Apoptotic cells (Annexin V+ /7AAD+ and Annexin V+ /7AAD− ) percentages are shown for CD8+ (A) and CD4+ T cells (B). Changes relative to baseline in the percentage of apoptosis in HLA-DR+ (in red) and HLA-DR− cells (in purple) are shown for CD8+ (C) and CD4+ (D) T cells. RMD administration cycles are indicated by gray bars. Sampling time points at 4 h, 8 h, 3 days, and 7 days after each RMD administration are indicated. Median with interquartile range is shown. P-values (*p < 0.05, **p < 0.01, ***p <0.001) are indicative for the corresponding timepoint compared to pre-RMD1 . to pre-RMD levels (p = 0.0054 for CD8+ and p = 0.0181 for induction of apoptosis due to RMD exposure as the activated CD4+ ), suggestive of an accumulative toxic effect of RMD. HLA-DR+ cells. As activated T cell have been found to be particularly sensitive to RMD exposure (28), levels of apoptosis in CD4+ and CD8+ CD8+ and CD4+ Naïve Populations T cell with high or low expression of the activation marker HLA-DR were assessed. Changes in the percentage of apoptotic Increase After MVA1 and Up to 24 h After cells relative to baseline during RMD exposure was comparable 1st Romidepsin Exposure Followed by a between HLA-DR+ and HLA-DR− CD8 T cells (Figures 2C,D). Shift to Memory Phenotypes The same results were observed in CD4+ T cells. Therefore, As activation status did not impact susceptibility of T cells to non- activated HLA-DR− cells had the same susceptibility to RMD-induced apoptosis, we next assessed whether RMD could Frontiers in Immunology | www.frontiersin.org 4 March 2020 | Volume 11 | Article 418 Rosás-Umbert et al. Romidepsin Effects on Vaccine-Induced T-Cells FIGURE 3 | Evolution of T-cell subsets over the course of the intervention. Averages of relative frequency of CD8+ (A) and CD4+ (B) T-cell subsets based on CCR7 and CD45RA expression are shown. TEMRA: highly differentiated effector cells expressing CD45RA (CCR7− /CD45RA+ ), EM: effector memory cells (CCR7− /CD45RA− ), CM: central memory (CCR7+ /CD45RA− ), and naïve (CCR7+ /CD45RA+ ). P-values (*p < 0.05, **p < 0.01, ***p < 0.001) are indicative for the corresponding timepoint compared with values from baseline samples prior to MVA1 . affect the distribution of different T-cell differentiation subsets in highly differentiated effector cells expressing CD45RA (TEMRA) the peripheral blood. Flow cytometry was used to measure the (Supplementary Figure 3). frequency of CD4+ and CD8+ differentiation subsets defined by Importantly, upon in vitro stimulation with peptides derived CCR7 and CD45RA expression at each timepoint. Three weeks from the vaccine immunogen, changes in the HLA-DR+ in after MVA1 , the frequency of naïve CD8+ T cells increased expression were already detected after MVA1 . Both, HIVconsv- from median frequency of 30.5–35.2% and up to 49.6% 24 h specific CD8+ and CD4+ T cells, defined upon in vitro antigen- after RMD1 (Wilcoxon signed-rank p = 0.0122 and p = 0.0005, specific stimulation and detection based on their ability to respectively, Figure 3A). Similar results were observed for naïve produce cytokines (IL-2, MIP1-β, TNF-α, IFN-γ) in response to CD4+ T cells, which increased from 33% at baseline to 35.2% antigen specific stimulation, showed detectable increases of HLA- 3 weeks after MVA1 and up to 41.7% at 24 h after RMD1 DR+ expression earlier and up to higher levels compared to the (Wilcoxon signed-rank p = 0.0227 and p = 0.0110, respectively, changes seen in total CD8+ and CD4+ T cells (Figure 4A). The Figure 3B). The frequency of CD8+ and CD4+ naïve cells were percentage of HLA-DR+ among HIVconsv-specific CD8+ cells not further increased after RMD2 or RMD3 , as their frequency increased from median of 5.1% at baseline to 10.11% 3 weeks actually decreased after RMD2 and RMD3 . At the same time, the after MVA1 , and up to 21% 3 days after RMD1 (Wilcoxon signed- median frequency of CD8+ effector memory and CD4+ central rank p = 0.0245 and p = 0.0002, Figure 4B) and percentage memory T cells progressively increased over RMD1−2−3 . T-cell of HLA-DR+ , HIVconsv-specific CD4+ cells increased from differentiation subsets were not further changed 4 weeks after median of 4.2% at baseline to 8.3% 3 weeks after MVA1 , and MVA2 compared to subset population observed after RMD doses. up to 21.8% 3 days after RMD1 (Wilcoxon signed-rank p = 0.5614 and p = 0.0023, Figure 4C). High levels of activation were maintained during all RMD doses (Figures 4C,D) and reached RMD Treatment Increases T-Cell Activation the peak 3 days after RMD3 suggestive of an additive effect of Especially of the HIVconsv-Specific T Cells RMD on the activation of vaccine-stimulated CD4+ and CD8+ In order to assess the effect of RMD on T-cell activation and T cells. exhaustion markers, we evaluated the frequency of CD4+ and As changes in in vivo histone acetylation and induction of CD8+ T cells expressing HLA-DR, CD69, CD25, and PD-1. A viral transcription occur rapidly after RMD exposure (22, 27), peak in the frequency of CD8+ HLA-DR+ T cells was observed we assessed the expression of early activation markers CD69 and 3 days after RMD1 (Wilcoxon signed-rank, p < 0.0001) and CD25 in CD4+ and CD8+ T cells. Increases in the expression of increased levels were maintained over the course of the 3 RMD CD69 in CD4+ T cells were observed upon each RMD dosing, doses, as their median frequency more than doubled (4.1% at with significant increases seen after RMD2 and RMD3 in the baseline to 9.6% 1 week after RMD3 , Wilcoxon signed-rank, CM subset (from baseline levels of 2.5 to 7.1% (RMD2 p = p = 0.0002, Figure 4A). The same results were observed in 0.0107) and 6.7% (RMD3 p = 0.0479) and in the TEMRA subset CD4+ T cells, with levels of HLA-DR expression increasing (baseline levels of 2.7 to 9% (RMD2 p = 0.0203) and 6.6% 3 days after RMD1 (Wilcoxon signed-rank, p < 0.0001) and (RMD3 p = 0.0315), respectively (Supplementary Figure 4B). with twice as many HLA-DR expressing cells at day 7 after As with HLA-DR surface staining, expression of CD69 returned RMD3 (median 3.1% at baseline to 6.3%, Wilcoxon signed- to baseline levels 4 weeks after last vaccination. No significant rank, p = 0.0017, Figure 4A) compared to baseline. The largest changes were observed in CD69 expression on CD8+ T cells increase in percentage of cells expressing HLA-DR in both CD4+ (Supplementary Figure 4A) nor in the expression of CD25 on and CD8+ T cells was observed in effector memory (EM) and CD8+ or CD4+ T cells over the course of the intervention. Frontiers in Immunology | www.frontiersin.org 5 March 2020 | Volume 11 | Article 418 Rosás-Umbert et al. Romidepsin Effects on Vaccine-Induced T-Cells FIGURE 4 | CD4+ and CD8+ T cell activation status over the course of the intervention. Percentage of HLA-DR expression is shown for CD8+ in orange and CD4+ T cells in purple (A). Levels of HLA-DR are shown for CD8+ (B) and CD4+ T cells (C) in HIVconsv specific and total T cells. HIVconsv-specific T cells were defined upon in vitro antigen-specific stimulation and detection based on their ability to produce cytokines (IL-2, MIP1-β, TNF-α, IFN-γ) in response to stimulation. Percentage of PD-1 is shown for CD8+ in orange and CD4+ T cells in purple (D). Median with interquartile range is shown. P-values (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001) are indicative for the corresponding timepoint compared with baseline. The expression of PD-1+ increased rapidly in CD8+ T cells PD-1 expression in both CD8+ and CD4+ T cells were transient, from median baseline levels of 4.6–6.4, 7.2, and 6.1% at 4 h after and the levels of PD-1 expression were consistently restored to the end of RMD1−2−3 treatment (Wilcoxon singed-rank, p = baseline levels within 3 days after each RMD dose. 0.0215, p = 0.0238, and p = 0.0574, respectively, Figure 4D). The same kinetics was observed in CD4+ T cells where the frequency of PD-1 expressing cells raised from median baseline Polyfunctionality of HIVconsv-Specific levels of 8.5–11.9, 12.35, and 10.7% at 4 h after RMD1−2−3 Responses Increases After MVA1 and (Wilcoxon singed-rank, p = 0.0004, p = 0.0002, and p = 0.0046, Decreases Over RMD Treatment respectively, Figure 4D). However, in contrast to the additive Considering the effect of RMD on the expression of markers effect of RMD1−2−3 on T-cell activation markers, changes in of cell death, exhaustion and activation, we assessed the Frontiers in Immunology | www.frontiersin.org 6 March 2020 | Volume 11 | Article 418 Rosás-Umbert et al. Romidepsin Effects on Vaccine-Induced T-Cells FIGURE 5 | Longitudinal assessment of the T-cell cytokine production over the course of the intervention. Polyfunctionality of CD8+ and CD4+ T cells was analyzed by Boolean gating. Pie charts illustrate relative proportion of each of the different subsets (cells producing 1, 2, 3, or 4 cytokines, respectively). effect of RMD on the functionality of vaccine-elicited T cells. MVA1 nor RMD1−2−3 altered the inhibition activity, which Polyfunctionality of HIVconsv-specific T cells was measured was median of 64, 52, and 53,5% at the screening, postMVA1 by stimulation of PBMC with the HIVconsv peptides and and postRMD3 time points at the E:T ratio 1:1, respectively, enumeration of the cells producing IFN-γ, IL-2, MIP1-β, against HIV-1BaL and median 46, 44, and 43% at the screening, and/or TNF-α by flow cytometry. As shown in Figure 5, postMVA1 and postRMD3 time points, respectively, against HIV- MVA1 vaccination increased relative polyfunctionality in both 1IIIB . In fact, there was a weak, but statistically significant increase CD4+ and CD8+ HIVconsv-specific T cells. Three weeks after in the inhibitory capacity against HIV-1BaL after the second MVA MVA1 (postMVA1 ), the highest increase in polyfunctionality vaccination given 4 weeks after RMD3 from median of 64% at was observed in vaccine immunogen-specific CD4+ T cells baseline to 69% at postMVA2 at the E:T ratio 1:1 and from median that produced 2, 3, and 4 cytokines (post MVA1 , Figure 5), of 11% at baseline to 21% at postMVA2 , for the E:T ratio1:10 with a particular increase in INF-γ secreting cells. On the (Wilcoxon signed-rank, p = 0.0200) (Figure 6A) and against other hand, vaccine-specific CD8+ T cells produced INF-γ, HIV-1IIIB from median of 12% at baseline to 16% at postMVA2 , IL-2, and TNF-α and were mostly polyfunctional, producing for at the 1:10 E:T ratio (Wilcoxon signed-rank, p = 0.0156) 2 or 3 cytokines. However, during the treatment with RMD, (Figure 6B). These data indicate that antiviral activity, measured the polyfunctionality was reduced to baseline levels in both by VIA with laboratory-adapted viral strains, was not negatively HIVconsv-specific CD4+ and CD8+ T cells and the final MVA2 impacted during RMD treatment. vaccination was not able to re-boost functionality profiles (post MVA2, Figure 5 and Supplementary Figure 5). These results suggest that RMD treatment might have impaired functionality of vaccine-induced responses and prevented their subsequent DISCUSSION booster vaccination effect. The BCN02 study was an HIV-1 kick&kill proof-of-concept trial combining MVA.HIVconsv vaccines with the HDACi RMD given CD8+ Cells Maintain Antiviral Activity of as an LRA to the previous ChAdV63.HIVconsv-MVA.HIVconsv After the RMD Treatment regimen recipients from the parental BCN01 trial (12). The To further study the functionality of vaccine-induced T cells immune analyses in BCN02 showed a highly significant shift during the intervention, we measured the in vitro antiviral of HIV-1-specific T-cell responses toward conserved regions of capacity of CD8+ T cells. The in vitro replication inhibition HIV-1 covered by the immunogen and marked increases in capacity of PBMC-derived CD8+ T cells was measured by histone acetylation and induction of viral transcription in vivo by standard viral inhibition assay (VIA) (12, 31) using autologous RMD. However, the reduction of the viral reservoir after these CD4+ T cells infected with two laboratory-adapted HIV-1 interventions was overall minimal (27). Already low baseline strains BaL (R5 tropic virus) and IIIB (X4 tropic virus). Virus levels of viral reservoir in the early-treated participants enrolled replication was measured by flow cytometry as the percentage in the BCN02 study likely limited latency reversal activity of of HIV-1 Gag p24-positive CD4+ cells. CD8+ T-cell inhibitory RMD and/or a potential toxicity of RMD (28, 32) on vaccine- capacity was determined at screening, 3 weeks after MVA1 induced T cells might have precluded the capability to observe (postMVA1 ), 1 week after RMD3 (postRMD3 ), and 4 weeks a more pronounced reduction in the size of the viral reservoir. after MVA2 (postMVA2 ). In contrast to the changes in cytokine To address the latter possibility, we here evaluated the effect production/polyfunctionality detected by flow cytometry, neither that the HDACi RMD had on the T-cell viability, activation and Frontiers in Immunology | www.frontiersin.org 7 March 2020 | Volume 11 | Article 418 Rosás-Umbert et al. Romidepsin Effects on Vaccine-Induced T-Cells FIGURE 6 | In vitro viral replication inhibition capacity. Levels of CD8+ viral inhibitory capacity are shown for HIV-1BaL (A) and HIV-1IIIB (B) at two E/T ratios (effector/target ratio) 1:1 and E/T 1:10 in individuals that underwent the intervention (n = 14 for HIV-1BaL and n = 15 for HIV-1IIIB ). P-values (*p < 0.05) are indicative for the corresponding timepoint compared with respective bsl. functionality throughout the intervention that included RMD These changes are in line with the increase in the magnitude given in a three weekly 5 mg/m2 BSA regimen. of the immunogen-specific responses upon vaccination (27). Our data document a in vivo toxic effect of RMD on T However, the magnitudes of vaccine-specific responses were not cells 3 days after each RMD exposure. Although these effects increased further during RMD administration, suggesting that were transient and viability of T cells was partially recovered viral reactivation did not contribute in a major way to the 7 days after RMD1 and RMD2 , there was an additive increase increased levels of cell activation during RMD. In contrast, RMD in apoptosis over the full RMD regimen, which was not fully treatment changed levels of CD69 and PD-1 more rapidly albeit restored by 7 days after RMD3 . Jones et al. and Zhao et al. have transiently, peaking at 4 h after the end of each RMD dose. In described increases in cell death upon long term (18- and 72- h line with our observation on CD69 longitudinal expression, an period) in vitro exposure to RMD, but no cytotoxic effect were increase in CD69+ expression was previously described upon observed when T cells were exposed to RMD for 4 h only, which RMD, panobinostat and vorinostat treatments in vitro (18, 24) is more similar to the in vivo regimen where the terminal half-life and also in vivo given alone (22) or in combination with a vaccine (t 1/2 ) of RMD is estimated to be ∼3 h (28, 32). Indeed, Clutton (34). This enhanced expression of PD-1 and CD69 could be the et al. exposed ex vivo PBMC to RMD for 3 h, washed them result of a generalized increased expression of the host genes and cultivated them further in the absence of RMD for 3 days induced by the action of RMD. So far, little is known about in order to mimic in vivo exposure, and reported that the cell the effect of RMD and other LRA on the host genome gene viability was not changed after 3 or 24 h, but was significantly expression profiles. There is an urgent need to better understand reduced at 48 and 72 h (33). These results concur with our these effects and, especially, to clarify how they can affect the observations showing a peak of apoptotic T cells 3 days after antiviral immunity against HIV-1 (35) or, in fact, fueling it. each RMD infusion and could explain the transient decreases in Although it was previously described that RMD can impair peripheral CD4+ T-cell counts observed in BCN02 (27). Also, the T-cell and NK function in vitro (28, 29, 36, 37), Søgaard et al. cumulative effect on viability observed after 3 RMD doses is in presented in vivo data from the REDUC trial, indicating that line with the delayed effects of RMD suggested by Clutton. RMD did not alter the proportion of HIV-1-specific T cells As RMD may induce T-cell activation resulting from reservoir nor inhibited T-cell cytokine production (22), at least in the reactivation and viral antigen presentation, which could also peripheral blood. Nonetheless, there appears to be a trend drive increase in exhaustion levels prior to the induction of toward decreased HIV-specific T cell responses after RMD apoptosis, we evaluated the frequency of activated CD4+ and treatment, which may have not reached statistical significance CD8+ T cells. Indeed, the additive increase in HLA-DR+ due to the limited group size. On another note, vaccine- activation markers in both CD8+ and CD4+ T cells over RMD induced responses in the REDUC trial were weaker compared treatment, especially in HIVconsv-specific T cells, was already to the responses induced in the BCN02 trial (12, 27), possibly observed after the first MVA.HIVconsv vaccination. Moreover, as due to weaker vaccine vectors/immunogens employed in the the increased expression of HLA-DR was mostly observed in T REDUC trial. In addition, REDUC did not include early-treated effector memory cells and the fact that HIVconsv-specific T cells individuals with relatively intact immunity, and the detection of showed higher levels of HLA-DR expression, the data suggest reduced polyfunctionality of vaccine-induced T-cells could have that there could be a different memory distribution among been limited. In BCN02, the first MVA.HIVconsv vaccination HIVconsv-specific T cells compared to the total T cell population. enhanced polyfunctionality, especially in CD4+ T cells, in line Frontiers in Immunology | www.frontiersin.org 8 March 2020 | Volume 11 | Article 418 Rosás-Umbert et al. Romidepsin Effects on Vaccine-Induced T-Cells with previous reports showing that MVA-vectored vaccination RMD has a transient effect on T-cell viability, exhaustion and can improve the polyfunctionality and T effector memory increased cell activation in an additive way over three weekly phenotype in CD4+ T cells more so than in CD8+ T cells (38). doses. Although this can result in a decrease of polyfunctionality This increased effector function profile induced by MVA1 was of vaccine-induced HIVconsv-specific responses, the in vitro reduced upon RMD treatment but did not lead to a net reduction replication inhibition capacity of CD8+ T cells was not impaired of the latent reservoir compared to baseline. and should not preclude effective killing upon RMD-induced Our data also indicate that despite fluctuations in activation, viral reactivation. As RMD increased higher levels of HIV-1 maturation phenotypes and polyfunctionality, RMD treatment transcription on the RMD2 and RMD3 dose, it is tempting did not impair the in vitro antiviral capacity of CD8+ T cells. to speculate that increasing the number the RMD doses could While some in vitro studies have shown a diminished inhibitory result in further increased in levels of reservoir reactivation capacity when CD8+ T cells were exposed to RMD, but not without overly inhibiting the antiviral capacity of CD8+ T to other HDACi (39, 40), such an effect was not observed in cells. Regardless whether this would occur in vivo with longer vivo in clinical trials using RMD. In the REDUC trial, the viral term RMD administration, the present data indicate that inhibition assay showed a trend toward increased inhibitory timing and order of LRA and T-cell immunotherapy regimens activity post-immunization that was lost after RMD exposure, are critical in order to achieve the clearance of reactivated but overall, antiviral capacity did not significantly change over latently HIV-1-infected cells. Larger controlled clinical trials are time. Similarly, we did not see a reduction in the antiviral capacity needed to further investigate combinations of LRA and immune of T cells over the course of RMD1−2−3 , suggesting that the intervention in order to find the best strategy to achieve a preservation of the antiviral capacity may depend on the balance functional cure of HIV-1. between the deleterious effects of RMD and the potency of the employed vaccination strategy. DATA AVAILABILITY STATEMENT Of note, our results highlight that there is no direct comparability between assays used to characterize CTL The datasets generated for this study are available on request to functionality, including multiparametric flow cytometry and the corresponding author. in vitro inhibition assays. In particular, in our study, the polyfunctionality of CD8+ T cells decreased slightly during RMD treatment, while in vitro VIA activity was maintained. ETHICS STATEMENT In addition, the MVA2 boost vaccination did not augment This study was carried out in accordance with the the proportion of cell secreting multiple cytokines, while a recommendations of and approval by the Ethics Committee moderately increased in vitro suppressive capacity was observed. of the Hospital Universitari Germans Trias i Pujol (Badalona, These data are consistent with other studies, which showed a Spain). All subjects provided their written informed consent to disconnection between cytokine secretion and antiviral capacity participate. The study was conducted according to the principles (41, 42), indicating that polyfunctionality of HIV-1-specific expressed in the Declaration of Helsinki. CD8+ T cells (at least as measured in standard assays) is not directly associated with viral suppression capacity (31, 43, 44). This lack of consistency between the two techniques could AUTHOR CONTRIBUTIONS be due to the cytokines measured in standard protocols for intracellular cytokine assays: whilst IFN-γ, IL-2, MIP1-β, and MR-U, MR-R, BM, and CB conceived and designed the study and TNF-α are used to measure functionality of HIV-1-specific T drafted the manuscript. MF, MM, PC, CM, SC, JMM, BC, TH, cells, evaluating secretion of granzyme B and perforin could be and JM contributed to the study design. MR-U, MR-R, and MF more accurate when assessing T-cell killing activity and may performed the experiments. All authors revised the manuscript provide better concordance with VIA activity (45). However, critically for important intellectual content and approved the whether antiviral capacity measured by standard in vitro VIA final version of the manuscript. assay using laboratory-adapted viral strains will translate into effective in vivo killing or reactivated, HIV-1 infected cells FUNDING remain to be determined as well. In addition, further work using autologous virus might be more representative of the The BCN02 trial was funded by the HIVACAT program, ISCIII physiological conditions in kick&kill strategies. Finally, viral (PI15/01188) grant and Fundació Glòria Soler. Sub-analyses were suppression capacity was measured in total bulk CD8+ T cells partly funded by the European Union’s Horizon 2020 research while cytokine secretion was measured in vaccine specific and innovation program under grant agreement 681137- cells, which could have been more susceptible to in vivo effects EAVI2020, by the Spanish Ministry of Economy, Industry by RMD. and Competitiveness (AEI/MINECO/FEDER), grant SAF2017- This present study has a number of limitations, which include 89726-R and by NIH grant P01-AI131568. The vaccine GMP a small sample size and the lack of control arms, both placebo manufacture was jointly funded by the UK Medical Research and single intervention arms to discern the effects exerted by Council and the UK Department for International Development RMD or the vaccine alone. Thus, the presented results need to (DFID) under the MRC/DFID Concordat agreements (MRC be interpreted with caution. Still, the present study shows that G0701669). JMM received a personal 80:20 research grant from Frontiers in Immunology | www.frontiersin.org 9 March 2020 | Volume 11 | Article 418 Rosás-Umbert et al. Romidepsin Effects on Vaccine-Induced T-Cells the Institut d’Investigacions Biomèdiques August Pi i Sunyer Germans Trias i Pujol Research Institute, Badalona, (IDIBAPS), Barcelona, Spain, during 2017–19. CM received a Spain: Ana Maria Barriocanal, Magi Farré, Cristina personal post-doctoral research grant (Pla Estratègic de Recerca Perez-Reche. i Innovació en Salut -PERIS- 2016/2020) from the ‘Departament Pharmacokinetic/pharmacodynamic modeling and de Salut de la Generalitat de Catalunya’, Barcelona, Spain. simultation, Institut de Recerca de l’Hospital de la Santa Creu i Sant Pau-IIB Sant Pau, Barcelona, Spain: Marta ACKNOWLEDGMENTS Valle. Hospital Clinic-HIVACAT, IDIBAPS, University of Barcelona, Spain: Christian Manzardo (currently working at Special thanks to all volunteers participating in this study for their Hospital Universitari Arnau de Vilanova, Lleida, Spain), Juan perseverance and dedication, without whom the phase I clinical Ambrosioni, Irene Ruiz, Cristina Rovira, Carmen Ligero and trial would not have been possible. Jose M. Miro. Projecte dels noms, BCN Checkpoint, Barcelona, Members of the BCN 02 Study Group Spain: Antonio Carrillo, Michael Meulbroek, Ferran Pujol IrsiCaixa AIDS Research Institute-HIVACAT, Hospital and Jorge Saz. Universitari Germans Trias i Pujol, Badalona, Spain: Susana The Jenner Institute, The Nuffield Department of Medicine, Benet, Christian Brander, Samandhy Cedeño, Bonaventura University of Oxford, UK: Nicola Borthwick, Alison Crook, Clotet, Pep Coll, Anuksa Llano, Javier Martinez-Picado, Marta Edmund G. Wee and Tomáš Hanke. Marszalek, Sara Morón-López, Beatriz Mothe, Roger Paredes, Maria C. Puertas, Miriam Rosás-Umbert, Marta Ruiz-Riol. SUPPLEMENTARY MATERIAL Fundació Lluita contra la Sida, Hospital Universitari Germans Trias i Pujol, Badalona, Spain: Roser Escrig, Silvia Gel, Miriam The Supplementary Material for this article can be found López, Cristina Miranda, José Moltó, Jose Muñoz, Nuria Perez- online at: https://www.frontiersin.org/articles/10.3389/fimmu. Alvarez, Jordi Puig, Boris Revollo, Jessica Toro. 2020.00418/full#supplementary-material REFERENCES with control of HIV-1 replication. Sci Transl Med. (2013) 5:166ra2. doi: 10.1126/scitranslmed.3004682 1. El-Sadr WM, Abrams D, Losso M, General De Agudos H-T, Ramos Mejia JM, 10. Autran B, Murphy RL, Costagliola D, Tubiana R, Clotet B, Gatell J, et al. Rappoport C. CD4+ count-guided interruption of antiretroviral treatment. N Greater viral rebound and reduced time to resume antiretroviral therapy after Engl J Med. (2006) 355:2283–96. doi: 10.1056/NEJMoa062360 therapeutic immunization with the ALVAC-HIV vaccine (vCP1452). AIDS. 2. Buzón JM, Massanella M, Llibre JM, Esteve A, Dahl V, Puertas MC, (2008) 22:1313–22. doi: 10.1097/QAD.0b013e3282fdce94 et al. HIV-1 replication and immune dynamics are affected by raltegravir 11. Schooley RT, Spritzler J, Wang H, Lederman MM, Havlir D, Kuritzkes intensification of HAART-suppressed subjects. Nat Med. (2010) 16:460–5. DR, et al. AIDS clinical trials group 5197: a placebo-controlled trial doi: 10.1038/nm.2111 of immunization of HIV-1-infected persons with a replication-deficient 3. Llibre JM, Buzón MJ, Massanella M, Esteve A, Dahl V, Puertas MC, et al. adenovirus type 5 vaccine expressing the HIV-1 core protein. J Infect Dis. Treatment intensification with raltegravir in subjects with sustained HIV- (2010) 202:705–16. doi: 10.1086/655468 1 viraemia suppression: a randomized 48-week study. Antivir Ther. (2011) 12. Mothe B, Manzardo C, Sanchez-Bernabeu A, Coll P, Morón-López S, 17:355–64. doi: 10.3851/IMP1917 Puertas MC, et al. Therapeutic vaccination refocuses T-cell responses towards 4. Gutiérrez C, Díaz L, Vallejo A, Hernández-Novoa B, Abad M, Madrid N, et al. conserved regions of HIV-1 in early treated individuals (BCN 01 study). Intensification of antiretroviral therapy with a CCR5 antagonist in patients EClinicalMedicine. (2019) 11:65–80. doi: 10.1016/j.eclinm.2019.05.009 with chronic HIV-1 infection: effect on t cells latently infected. PLoS ONE. 13. Hancock G, Morón-López S, Kopycinski J, Puertas MC, Giannoulatou (2011) 6:e27864. doi: 10.1371/journal.pone.0027864 E, Rose A, et al. Evaluation of the immunogenicity and impact on the 5. Vallejo A, Gutierrez C, Hernandez-Novoa B, Diaz L, Madrid N, Abad- latent HIV-1 reservoir of a conserved region vaccine, MVA.HIVconsv, in Fernandez M, et al. The effect of intensification with raltegravir on the HIV- antiretroviral therapy-treated subjects. J Int AIDS Soc. (2017) 20:21171. 1 reservoir of latently infected memory CD4 T cells in suppressed patients. doi: 10.7448/IAS.20.1.21171 AIDS. (2012) 26:1885–94. doi: 10.1097/QAD.0b013e3283584521 14. Tebas P, Stein D, Tang WW, Frank I, Wang SQ, Lee G, et al. Gene editing of 6. Gandhi RT, Coombs RW, Chan ES, Bosch RJ, Zheng L, Margolis DM, et al. No CCR5 in autologous CD4 T cells of persons infected with HIV. N Engl J Med. effect of raltegravir intensification on viral replication markers in the blood (2014) 370:901–10. doi: 10.1056/NEJMoa1300662 of HIV-1-infected patients receiving antiretroviral therapy. J Acquir Immune 15. Wang CX, Cannon PM. The clinical applications of genome editing in HIV. Defic Syndr. (2012) 59:229–35. doi: 10.1097/QAI.0b013e31823fd1f2 Blood. (2016) 127:2546–52. doi: 10.1182/blood-2016-01-678144 7. Lynch RM, Boritz E, Coates EE, DeZure A, Madden P, Costner P, et al. 16. Ververis K, Karagiannis TC. Potential non-oncological applications of histone Virologic effects of broadly neutralizing antibody VRC01 administration deacetylase inhibitors. Am J Transl Res. (2011) 3:454–67. during chronic HIV-1 infection. Sci Transl Med. (2015) 7:319ra206. 17. Archin NM, Liberty AL, Kashuba AD, Choudhary SK, Kuruc JD, Crooks doi: 10.1126/scitranslmed.aad5752 AM, et al. Administration of vorinostat disrupts HIV-1 latency in patients on 8. Sneller MC, Justement JS, Gittens KR, Petrone ME, Clarridge KE, antiretroviral therapy. Nature. (2012) 487:482–5. doi: 10.1038/nature11286 Proschan MA, et al. A randomized controlled safety/efficacy trial 18. Wei DG, Chiang V, Fyne E, Balakrishnan M, Barnes T, Graupe of therapeutic vaccination in HIV-infected individuals who initiated M, et al. Histone deacetylase inhibitor romidepsin induces antiretroviral therapy early in infection. Sci Transl Med. (2017) 9:eaan8848. HIV expression in CD4 T cells from patients on suppressive doi: 10.1126/scitranslmed.aan8848 antiretroviral therapy at concentrations achieved by clinical dosing. 9. García F, Climent N, Guardo AC, Gil C, León A, Autran B, et al. PLoS Pathog. (2014) 10:e1004071. doi: 10.1371/journal.ppat.10 A dendritic cell-based vaccine elicits T cell responses associated 04071 Frontiers in Immunology | www.frontiersin.org 10 March 2020 | Volume 11 | Article 418 Rosás-Umbert et al. Romidepsin Effects on Vaccine-Induced T-Cells 19. Rasmussen TA, Tolstrup M, Brinkmann CR, Olesen R, Erikstrup C, Solomon 35. Ruiz-Riol M, Brander C. Can we just kick-and-kill HIV: possible challenges A, et al. Panobinostat, a histone deacetylase inhibitor, for latent-virus posed by the epigenetically controlled interplay between HIV and host reactivation in HIV-infected patients on suppressive antiretroviral therapy: immunity. Immunotherapy. (2019) 11:931–5. doi: 10.2217/imt-2019-0092 a phase 1/2, single group, clinical trial. Lancet HIV. (2014) 1:e13–21. 36. Pace M, Williams J, Kurioka A, Gerry AB, Jakobsen B, Klenerman P, et al. doi: 10.1016/S2352-3018(14)70014-1 Histone deacetylase inhibitors enhance CD4 T Cell susceptibility to NK 20. Elliott JH, Wightman F, Solomon A, Ghneim K, Ahlers J, Cameron MJ, cell killing but reduce NK Cell function. PLoS Pathog. (2016) 12:e1005782. et al. Activation of HIV transcription with short-course vorinostat in HIV- doi: 10.1371/journal.ppat.1005782 infected patients on suppressive antiretroviral therapy. PLoS Pathog. (2014) 37. Garrido C, Spivak AM, Soriano-Sarabia N, Checkley MA, Barker E, Karn J, 10:e1004473. doi: 10.1371/journal.ppat.1004473 et al. HIV latency-reversing agents have diverse effects on natural killer cell 21. Archin NM, Bateson R, Tripathy MK, Crooks AM, Yang K-H, Dahl NP, function. Front Immunol. (2016) 7:356. doi: 10.3389/fimmu.2016.00356 et al. HIV-1 expression within resting CD4+ T cells after multiple doses of 38. Gómez CE, Perdiguero B, García-Arriaza J, Cepeda V, Sánchez-Sorzano vorinostat. J Infect Dis. (2014) 210:728–35. doi: 10.1093/infdis/jiu155 CÓ, Mothe B, et al. A phase I randomized therapeutic MVA-B vaccination 22. Søgaard OS, Graversen ME, Leth S, Olesen R, Brinkmann CR, Nissen SK, et al. improves the magnitude and quality of the T cell immune responses The depsipeptide romidepsin reverses HIV-1 latency in vivo. PLoS Pathog. in HIV-1-infected subjects on HAART. PLoS ONE. (2015) 10:e0141456. (2015) 11:1–22. doi: 10.1371/journal.ppat.1005142 doi: 10.1371/journal.pone.0141456 23. Kim YH, Demierre M-F, Kim EJ, Lerner A, Rook AH, Duvic M, et al. 39. Sung JA, Lam S, Garrido C, Archin N, Rooney CM, Bollard CM, et al. Clinically meaningful reduction in pruritus in patients with cutaneous T- Expanded cytotoxic T-cell lymphocytes target the latent HIV reservoir. J Infect cell lymphoma treated with romidepsin. Leuk Lymphoma. (2013) 54:284–289. Dis. (2015) 212:258–63. doi: 10.1093/infdis/jiv022 doi: 10.3109/10428194.2012.711829 40. Walker-Sperling VE, Pohlmeyer CW, Veenhuis RT, May M, Luna KA, 24. Rasmussen TA, Søgaard OS, Brinkmann C, Wightman F, Lewin SR, Kirkpatrick AR, et al. Factors associated with the control of viral replication Melchjorsen J, et al. Comparison of HDAC inhibitors in clinical development: and virologic breakthrough in a recently infected HIV-1 controller. Effect on HIV production in latently infected cells and T-cell activation. Hum EBioMedicine. (2017) 16:141–9. doi: 10.1016/j.ebiom.2017.01.034 Vaccines Immunother. (2013) 9:993–1001. doi: 10.4161/hv.23800 41. Payne RP, Kløverpris H, Sacha JB, Brumme Z, Brumme C, Buus S, 25. Leth S, Schleimann MH, Nissen SK, Højen JF, Olesen R, Graversen ME, et al. et al. Efficacious early antiviral activity of HIV Gag- and Pol-specific Combined effect of Vacc-4x, recombinant human granulocyte macrophage HLA-B∗ 2705-restricted CD8 + T Cells. J Virol. (2010) 84:10543–57. colony-stimulating factor vaccination, and romidepsin on the HIV-1 reservoir doi: 10.1128/JVI.00793-10 (REDUC): a single-arm, phase 1B/2A trial. Lancet HIV. (2016) 3:e463–72. 42. Salido J, Ruiz MJ, Trifone C, Figueroa MI, Caruso MP, Gherardi MM, doi: 10.1016/S2352-3018(16)30055-8 et al. Phenotype, polyfunctionality, and antiviral activity of in vitro 26. Shan L, Deng K, Shroff NS, Durand CM, Rabi SA, Yang HC, et al. stimulated CD8+ T-cells from HIV+ subjects who initiated cART at Stimulation of HIV-1-Specific cytolytic T lymphocytes facilitates elimination different time-points after acute infection. Front Immunol. (2018) 9:2443. of latent viral reservoir after virus reactivation. Immunity. (2012) 36:491–501. doi: 10.3389/fimmu.2018.02443 doi: 10.1016/j.immuni.2012.01.014 43. Ruiz-Riol M, Llano A, Ibarrondo J, Zamarreño J, Yusim K, Bach V, et al. 27. Mothe B. Therapeutic vaccines after early ART: impact on reservoir. In: Oral Alternative effector-function profiling identifies broad HIV-specific T-cell session on Therapeutic Vaccines and Cure Strategies at the Joint Keystone responses in highly HIV-exposed individuals who remain uninfected. J Infect Symposia on HIV Vaccines and Functional Cures and the Eradication of HIV, Dis. (2015) 211:936–46. doi: 10.1093/infdis/jiu534 2019 Mar 24–28. Whistler, BC (2019). 44. Rosas-Umbert M, Llano A, Olvera A, Ruiz-riol M, Rocafort M, Cobarsi P, 28. Jones RB, Connor RO, Mueller S, Foley M, Szeto GL, Karel D, et al. Mechanisms of abrupt loss of virus control in a cohort of previous HIV et al. Histone deacetylase inhibitors impair the elimination of HIV- controllers. J Virol. (2019) 93:e01436-18. doi: 10.1128/JVI.01436-18 infected cells by cytotoxic T-Lymphocytes. (2014) 10:e1004287. 45. Migueles SA, Osborne CM, Royce C, Compton AA, Joshi RP, Weeks KA, doi: 10.1371/journal.ppat.1004287 et al. Lytic granule loading of CD8+ T cells is required for HIV-infected cell 29. Jones RB, Mueller S, Connor RO, Rimpel K, Sloan D, Karel D, et al. elimination associated with immune control. Immunity. (2008) 29:1009–1021. A subset of latency-reversing agents expose HIV-infected resting CD4 doi: 10.1016/j.immuni.2008.10.010 + T-cells to recognition by cytotoxic T-lymphocytes. (2016) 12:1–25. doi: 10.1371/journal.ppat.1005545 Conflict of Interest: JMM reports grants and personal fees from Abbvie, Angelini, 30. Roederer M, Nozzi JL, Nason MC. SPICE: exploration and analysis of post- Contrafect, Genentech, Gilead, Jansen, Medtronic, MSD, Pfizer, ViiV Healthcare, cytometric complex multivariate datasets. Cytometry A. (2011) 79:167–74. outside the submitted work. TH reports grants from Medical Research Council doi: 10.1002/cyto.a.21015 UK, during the conduct of the study, and has a patent US 7981430B2 issued. CB is 31. Yang H, Wu H, Hancock G, Clutton G, Sande N, Xu X, et al. Antiviral founder, CSO and shareholder of AELIX THERAPEUTIC, S.L. BM is a consultant inhibitory capacity of CD8+ T cells predicts the rate of CD4+ T-cell decline for AELIX THERAPEUTICS, S.L., outside the submitted work. in HIV-1 infection. J Infect Dis. (2012) 206:552–61. doi: 10.1093/infdis/jis379 32. Zhao M, De Crignis E, Rokx C, Verbon A, van Gelder T, Mahmoudi T, The remaining authors declare that the research was conducted in the absence of et al. T cell toxicity of HIV latency reversing agents. Pharmacol Res. (2019) any commercial or financial relationships that could be construed as a potential 139:524–34. doi: 10.1016/j.phrs.2018.10.023 conflict of interest. 33. Clutton G, Xu Y, Baldoni PL, Mollan KR, Kirchherr J, Newhard W, et al. The differential short- and long-term effects of HIV-1 latency- Copyright © 2020 Rosás-Umbert, Ruiz-Riol, Fernández, Marszalek, Coll, Manzardo, reversing agents on T cell function. Sci Rep. (2016) 6:30749. doi: 10.1038/ Cedeño, Miró, Clotet, Hanke, Moltó, Mothe, Brander and the BCN02 study group. srep30749 This is an open-access article distributed under the terms of the Creative Commons 34. Tapia G, Højen JF, Ökvist M, Olesen R, Leth S, Nissen SK, et al. Attribution License (CC BY). The use, distribution or reproduction in other forums Sequential Vacc-4x and romidepsin during combination antiretroviral therapy is permitted, provided the original author(s) and the copyright owner(s) are credited (cART): Immune responses to Vacc-4x regions on p24 and changes and that the original publication in this journal is cited, in accordance with accepted in HIV reservoirs. J Infect. (2017) 75:555–71. doi: 10.1016/j.jinf.2017. academic practice. No use, distribution or reproduction is permitted which does not 09.004 comply with these terms. Frontiers in Immunology | www.frontiersin.org 11 March 2020 | Volume 11 | Article 418