Cells were cultured in 5% CO₂ at 37°C. TZM-bl cells were maintained in Dulbecco’s modified Eagle medium (DMEM, Gibco/ThermoFisher Scientific, Waltham, MA, USA) supplemented with 10% fetal bovine serum (FBS, Sigma/Merck, Darmstadt, Germany), 2 mM L-glutamine (Sigma), 0.1 mM MEM Non-Essential Amino Acids (Gibco), and 20 units/mL penicillin combined with 20 μg/mL streptomycin (Sigma). KHYG-1 cells (#ACC 725, DSMZ, Braunschweig, Germany) were maintained in Roswell Park Memorial Institute 1640 (RPMI, Sigma) medium supplemented with 10% FBS, 25 mM HEPES, 20 units/mL penicillin and 20 μg/mL streptomycin, and 100 units/mL of recombinant human interleukin-2 (IL-2, PeproTech, Cranbury, NJ, USA). J-Lat 10.6 cells were maintained in RPMI medium supplemented with 10% FBS, 25 mM HEPES, 20 units/mL penicillin and 20 μg/mL streptomycin. The J-Lat 1C10 cell line construction is described in (20). J-Lat 1C10 cells were cultured in cytokine-free media (RPMI 1640 medium, Hyclone/Cytiva, Marlborough, MA, USA), 10% FBS (Life Technologies), 1% Glutamax (Life Technologies), and 1% Penicillin-streptomycin (Life Technologies).

Pan-caspase inhibitor Z-VAD-FMK and individual caspase inhibitors (Caspase-Family Inhibitor Set IV) were purchased from Enzo Life Sciences (Farmingdale, NY, USA) and GrM inhibitor Ac-KVPL-CMK from Peptanova (Sandhausen, Germany). Pan-caspase inhibitors Q-VD-Oph and emricasan were purchased from Selleckchem (Houston, TX, USA). CellTrace™ Violet proliferation kit, LIVE/DEAD Fixable Violet Dead Cell Stain Kit and LIVE/DEAD Fixable Near IR Dead Cell Stain Kit were obtained from ThermoFisher Scientific. Flow cytometry antibodies CD3-PE (clone OKT3), CD16-PerCP-Cy5.5 (clone 3G8), and CD56-PE-Cy7 (clone HCD56) were obtained from Biolegend (San Diego, CA, USA). CCR5-BV421 (clone 2D7) and its isotype control (mouse C57BL/6 IgG2a, κ, BV421) were obtained from BD Biosciences (Franklin Lakes, NJ, USA).

Recombinant human IL-21, recombinant human IL-10, recombinant human MIP-1α (CCL3), recombinant human MIP-1β (CCL4), recombinant human RANTES (CCL5) were purchased from PeproTech. Prostratin, phorbol 12-myristate 13-acetate (PMA), and ionomycin were purchased from Sigma. JQ1 was purchased from Cayman Chemical (Ann Arbor, MI, USA) and romidepsin from Celgene (Brystol Meyers Squibb, Summit, NJ, USA).

Seeded KHYG-1 cells, enriched primary NK cells, or J-Lat cells were incubated at 37°C for the times indicated with 50 µM of the pan-caspase inhibitors (10 mM stock solutions in DMSO), 10 µM of individual caspase inhibitors (stock solution is 2 mM in DMSO) from the Caspase-Family Inhibitor Set IV that also contains the pan-caspase inhibitor control Z-VAD-FMK and cysteine cathepsin inhibitor Z-FA-FMK (negative control for FMK inhibitors), or only DMSO (0.5%) as a control. For collection of supernatants, cells were collected, centrifuged at 1500 rpm for 5 min and supernatant was collected. Fractionating of supernatants was performed using Amicon Ultracel Centrifugal Filter columns (3 kDa MWCO, 2 mL sample volume, Millipore/Sigma) according to manufacturer’s protocol with centrifugation at 4000 rpm for 60 min. For the unconcentrated >3 kDa fraction, 10% FBS was added, whereas the concentrated <3 kDa fraction was resuspended in serum-free medium to its original volume.

For the co-culture assay, J-Lat 10.6 cells were labelled with 2 µM CellTrace Violet dye on day one. On day two, 10x10⁶ J-Lat were collected and half was stimulated with 6 µM of prostratin for 4 h at 37°C, while the other half was left unstimulated as a control. Cells were washed and resuspended to a concentration of 0.5x10⁶ J-Lat cells/mL and incubated with z-VAD-FMK for 1 h at 37°C. After this incubation period, a cell suspension of 0.5x10⁶ KHYG-1 cells or only medium were added to the J-Lat cells in an effector:target ratio of 1:1. Z-VAD-FMK or DMSO were added to adjust the concentration and cells were incubated overnight in the same conditions as KHYG-1 are maintained, including 100 U/mL of rhIL-2.

For the incubation of J-Lat cells with supernatants of KHYG-1 cells, a cell suspension of 1x10⁶ J-Lat cells was incubated with 6 µM of prostratin for 4 h at 37°C or left unstimulated as a control. J-Lat cells were then washed twice with PBS and resuspended to a concentration of 0.5x10⁶ J-Lat cells/mL using supernatant from KHYG-1 cells or only complete RPMI with DMSO (0.5%) or Z-VAD-FMK (50 µM) and incubated overnight at 37°C.

For the synergy experiments, 0.5x10⁶ J-Lat cells were resuspended in 0.5 mL complete RPMI or supernatants from treated KHYG-1 cells. In each condition, an additional LRA was supplemented to the culture; prostratin (2 µM), JQ1 (1 µM), romidepsin (17.5 nM), or no LRA. J-Lat cells were cultured under these conditions for 22 h at 37°C.

For 1C10 cells, cells were exposed to latency-reversal agents PMA (50 ng/ml) in combination with ionomycin (1 μM), or prostratin (6 μM), or supernatants of KHYG-1 for 16 h.

For the co-culture assay and the incubation with supernatant from NK cells experiment, J-Lat 10.6 cells were washed in ice-cold PBS and incubated with Near IR viability dye for 30 min at 4°C in the dark. The cells were then washed with ice-cold PBS and fixed with 2% paraformaldehyde for 20 min at 4°C in the dark. They were washed with ice-cold PBS one last time and finally resuspended in ice-cold PBS and stored at 4°C until they were analyzed. For 1C10 cells, cells were washed once with PBS, stained with Violet Dead Cell Stain and fixed in 2% formaldehyde for 15 min.

For the CCR5 expression experiment, J-Lat 10.6 cells were kept in five different conditions and TZM-bl cells were used as a positive control. On day one, a cell suspension of 1x10⁶ KHYG-1 cells/mL was stimulated with Z-VAD-FMK overnight at 37°C. On day two, J-Lat cells were resuspended to a concentration of 0.5x10⁶ cells/mL in (1): only RPMI 1640 medium (2), 100 U/mL IL-2, (3) Z-VAD-FMK, (4) 3 mL of supernatant of KHYG-1 cells or (5) 3 mL of supernatant of KHYG-1 cells that were stimulated with Z-VAD-FMK. All cells were incubated overnight at 37°C. On day three, cells were washed twice in FACS buffer (PBS with 1% BSA and 2 mM EDTA) and incubated with no antibody, BV421 isotype or CCR5-BV421 antibody diluted in FACS buffer with Near IR viability dye for 30 min at 4°C in the dark. The cells were then washed with FACS buffer and fixed with 2% paraformaldehyde for 20 min at 4°C in the dark. They were washed twice with ice-cold PBS and finally resuspended in ice-cold PBS and stored at 4°C until they were analyzed.

For staining of NK cell populations in PBMCs or enriched primary NK cell cultures, 0.5x10⁶ to 1.0x10⁶ cells were collected per staining. Cells were washed twice in ice-cold FACS buffer and incubated with CD3-PE, CD16-PerCP-Cy5.5, CD56-PE-Cy7 antibodies and Near IR viability dye for 30 min at 4°C in the dark. The cells were then washed with FACS buffer and fixed with 2% paraformaldehyde for 20 min at 4°C in the dark. They were washed twice with ice-cold PBS and finally resuspended in ice-cold PBS and stored at 4°C until they were analyzed.

All flow analysis was performed on BD FACSVerse (BD Biosciences), except for the analysis of 1C10 cells that was performed on a CytoFLEX S (Beckman Coulter, Pasadena, CA, USA).

TZM-bl cells (2x10⁴ cells) were seeded in 96-wells plate one day prior to start of the experiment. Then, medium was removed, and supernatants of NK cells (200 µL) were added in triplicate. TZM-bl cells with only complete RPMI medium, RPMI with pan-caspase inhibitors (50 µM), RPMI with rhIL-2 (either 100 units/mL for experiments with KHYG-1 cells or 1000 units/mL for experiments with enriched primary NK cells), and RPMI with prostratin (6 µM) were used as controls. After one day of culture, supernatants were removed, washed once with PBS, and lysed in 50 µL Passive Lysis Buffer (Promega, Madison, WI, USA) for 30 min at 4°C. Finally, 20 µL lysates were transferred into a white 96-wells plate, 100 uL luciferase reagent (Promega) was added, and luminescence was measured on the Tecan Infinite^® 200 Pro microplate reader (Tecan Group Ltd, Männedorf, Switzerland). For each experiment fold change was calculated by dividing the relative luminescence units (RLU) of each condition by the RLU of TZM-bl cells cultured only with complete RPMI medium.

A cell suspension of 1x10⁶ KHYG-1 cells/mL was incubated with Z-VAD-FMK overnight at 37°C. The following day, 700 µL of supernatant was used to perform a Proteome Profiler™, Human Cytokine Array (#ARY005B, R&D Systems, Minneapolis, MN, USA) according to the instructions of the manufacturer. This array allows the determination of the relative expression levels of 36 human cytokines. Chemiluminescence was detected by ChemiDoc XRS+ (Bio-Rad, Hercules, CA, USA).

A cell suspension of 1x10⁶ KHYG-1 or primary NK cells/mL were stimulated with Z-VAD-FMK and incubated overnight at 37°C. Total RNA was extracted with Quick-RNA™ Miniprep Plus Kit (Zymo Research, Irvine, CA, USA) according to the instructions of the manufacturer. cDNA was synthesized using High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Waltham, MA, USA), with RNasin^® Ribonuclease Inhibitor (Promega). qPCR reactions were done using Premix Ex Taq™ (Probe qPCR) (Takara Bio Inc., Kusatsu, Japan) with the primer/probe sets (#4331182) for CCL3 (assay ID Hs00234142_m1), CCL4 (Hs99999148_m1), CCL5 (Hs99999048_m1) and GAPDH (Hs02786624_g1) from ThermoFisher Scientific. The qPCR was performed on an ABI Fast 7500 system (Applied Biosystems) according to manufacturer’s protocol. Gene expression analysis was performed using the ΔΔCT method. RNA samples were stored at -20°C and cDNA samples at 4°C.

PBMCs were isolated from buffy coats of anonymous blood donors using Ficoll density centrifugation (Ficoll^® Paque Plus, Sigma), aliquoted and stored in liquid nitrogen. A frozen aliquot of PBMCs was thawed and NK cells were negatively selected by incubating the sample in PBS containing 2% FBS with 1 mM EDTA and 50 µL/mL of Enrichment Cocktail (Stem Cell Technology, Vancouver, Canada) for 10 min at RT; followed by incubation with 100 µL/mL Magnetic Particles for 5 min at RT and the usage of a magnet to collect the supernatant containing the enriched NK cells. These cells were kept in AIM-V medium (Gibco) supplemented with 5% human AB serum (Sigma). The first day of culture, 1000 U/mL of rhIL-2 and 20 ng/mL of rhIL-21 was added. After the first day of culture, a daily dose of only 1000 U/mL of rhIL-2 was added. After 4 days of culture, enriched NK cells were treated with Z-VAD-FMK or DMSO for one day.

J-Lat 1C10 cells were washed with PBS and allowed to adhere to Polylysine coated slides (VWR, Radnor, PA, USA) marked with a hydrophobic barrier using A-PAP pen (Histolab, Gothenburg, Sweden). PLA was performed according to the manufacturer’s protocol (Sigma) with a few modifications: PLA plus and minus probes were diluted 1:20, amplification buffer (5×) was used at 10×. All washes were performed in PBS. Antibodies (1:1,000) were used against Tat (Abcam, Cambridge, United Kingdom), FLAG M2 (Sigma). Before DAPI staining and mounting with Duolink in Situ mounting media with DAPI (Sigma), FITC-conjugated anti-GFP (Abcam) was applied (1:500) for 1 h at ambient temperature protected from light. Slides were sealed with nail polish and stored at 4°C overnight before imaging. Slides were imaged using a Pannoramic Midi II slide scanner (3DHistech, Budapest, Hungary) and images were exported using the CaseViewer application. Images were analyzed with ImageJ (version 2.0.0-rc-69/1.52) and macros developed in house. Each RGB image was separated into separate channels. Based on the blue (DAPI) channel, nuclei were identified. In the red channel, spots (“maxima”) were detected with prominence thresholds 6–48.

Data was analyzed and plotted using GraphPad Prism v8.4.2 or v9.0.0 (San Diego, CA, USA). Statistical analysis using ANOVA was performed using GraphPad Prism.

Since we were interested in whether NK cells could inhibit HIV-1 replication in a GrM-dependent manner, we performed a co-culture experiment using the HIV-1 latently infected cell line J-Lat 10.6 as target cells and the NK cell lymphoma cell line KHYG-1, which has high expression levels of GrM but not granzyme B, as effector cells. J-Lat cells contain an integrated copy of HIV-1 with GFP in place of Nef, an accessory gene. Thus, HIV-1 reactivation and replication can be monitored by GFP expression. We also labelled J-Lat cells with CellTrace Violet to discriminate them from KHYG-1 cells in flow cytometry analysis ( Figure 1A ). We briefly reactivated J-Lat cells with the LRA prostratin and then co-cultured them with KHYG-1 cells in an 1:1 effector:target (E:T) ratio for 1 day. GrM inhibitor Ac-KVPL-CMK was added to specifically inhibit the proteolytic activity of GrM. We included the pan-caspase inhibitor Z-VAD-FMK in the co-cultures to limit bias of target cell death. The GrM inhibitor had no effect on the reduction in percentage of GFP-positive J-Lat cells when co-cultured with KHYG-1 cells ( Figure 1B ). The presence of Z-VAD-FMK prevented cell death of J-Lat cells induced by KHYG-1 cells ( Figure 1C ) and maintained the E:T ratio of 1:1 throughout the experiment ( Figure 1D ). The percentage of GFP-positive J-Lat cells was increased when co-cultured with KHYG-1 and in the presence of Z-VAD-FMK ( Figure 1B ). This effect only occurred in co-culture with KHYG-1 cells, but not when J-Lat cells were cultured alone with Z-VAD-FMK.

Next, we assessed whether cell-to-cell contact was required for Z-VAD-FMK-treated KHYG-1 cells to induce HIV-1 reactivation. We incubated KHYG-1 cells with Z-VAD-FMK for one day, collected the supernatant, and cultured either unstimulated or prostratin-stimulated J-Lat cells with the supernatants. Supernatants of Z-VAD-FMK-treated KHYG-1 cells induced and increased the percentage of GFP-positive J-Lat cells ( Figure 1E ). Thus, the pan-caspase inhibitor Z-VAD-FMK appeared to affect the KHYG-1 cells to secrete a biological product that subsequently induces or increases HIV-1 reactivation within J-Lat cells.

Since the supernatant of Z-VAD-FMK-treated KHYG-1 cells only weakly induced HIV-1 reactivation in J-Lat cells, we tested whether there could be synergy in HIV-1 reactivation with some other LRAs. For this, we cultured J-Lat cells in normal medium or in collected KHYG-1 supernatants. In each of these conditions we either added prostratin (protein kinase C agonist), JQ1 (BET bromodomain inhibitor), romidepsin (histone deacetylase inhibitor), or no additional LRA and measured the percentage of GFP-positive cells by flow cytometry after one day of culture. Again, supernatants of Z-VAD-FMK-treated KHYG-1 weakly induced HIV-1 reactivation with approximately 3.9% GFP-positive J-Lat cells without any additional LRA ( Figure 2A ). We observed an increase in prostratin-induced reactivation from 69.2% to 86.6% and JQ1-induced reactivation from 4.7% to 18.2% when J-Lat cells were cultured in supernatants of Z-VAD-FMK-treated KHYG-1 cells ( Figures 2B, C ). These increases in GFP-positivity were clearly higher than the added percentages of GFP-positivity induced by the supernatant of Z-VAD-FMK-treated KHYG-1 or the individual LRA alone, indicating a synergistic effect. For romidepsin, there was no significant increase in the percentage of GFP-positive J-Lat when cultured in supernatants of Z-VAD-FMK-treated KHYG-1 cells (57.2%) compared to romidepsin-treated J-Lat cells (50.7%) ( Figure 2D ), although it was significantly increased compared to DMSO-treated supernatants (36.7%). Except for romidepsin treatment, viability was equally high in all conditions and thus did not affect the percentages of GFP-positivity cells ( Figure 2E ). Therefore, treatment of KHYG-1 cells with Z-VAD-FMK can work in synergy with some other LRAs to induce a stronger HIV-1 reactivation.

Since HIV-1 reactivation starts with the activation of the HIV-1 long terminal repeat (LTR) promoter, we examined whether the supernatants of Z-VAD-FMK-treated KHYG-1 cells can induce HIV-1 promoter activation. For this we used TZM-bl cells, which have an integrated copy of the luciferase gene under the HIV-1 LTR promoter. First, KHYG-1 cells were incubated with or without pan-caspase inhibitor Z-VAD-FMK for 1 day. Then, we collected the supernatants and cultured the TZM-bl cells in these supernatants for 1 day before we performed a luciferase assay. Prostratin was used as a positive control and showed increased HIV-1 promoter activity ( Figure 3A ), as expected. Supernatant of Z-VAD-FMK-treated KHYG-1 induced significantly higher HIV-1 promoter activity, compared to DMSO-treated supernatants ( Figure 3A ). HIV-1 promoter activity decreased when we diluted the supernatants. Altogether, this indicates that a secreted factor by KHYG-1 cells after treatment with the pan-caspase inhibitor Z-VAD-FMK can induce activation of the HIV-1 LTR promoter.

Since TZM-bl cells do not express Tat, we used J-Lat 1C10 cells as another cell model to examine whether the reactivation was Tat-driven or Tat-independent (20). This cell line allows detection of transactivation response element (TAR)-bound Tat in proximity to an LTR-bound zinc finger protein through a proximity ligation assay. Although the parental cell line 5A8 is known to be less responsive to some common LRAs (21), we did observe a significant, albeit low, increase in GFP positivity of the 1C10 cells when cultured with supernatants of Z-VAD-FMK-treated KHYG-1 cells compared to supernatants of control KHYG-1 cells ( Supplementary Figure 1A ). However, no increase in LTR-bound Tat was detected in the 1C10 cells that were incubated with supernatants of Z-VAD-FMK-treated KHYG-1 cells ( Supplementary Figure 1B ). This indicates that the observed reactivation occurs before Tat binding to the TAR region of HIV-1 RNA.

Since Z-VAD-FMK inhibits a wide range of caspases, we tested more specific caspase inhibitors to figure out which inhibited caspase within the KHYG-1 cells was responsible for the NK-cell mediated induction of HIV-1 promoter activity. We incubated the KHYG-1 cells with ten specific caspase inhibitors as well as with a positive control inhibitor (VAD) that is identical to the Z-VAD-FMK used in all our other experiments and a negative control inhibitor (FA). All inhibitors contain the N-terminal benzyloxycarbonyl (Z-) group and C-terminal fluoromethylketone (FMK) inhibitor group. After 24 h, supernatants were collected and added to TZM-bl cells. However, none of the specific caspase inhibitors alone could increase HIV-1 promoter activity ( Figure 3B ).

Further, we examined whether two other pan-caspase inhibitors, Q-VD-Oph and emricasan, could cause KHYG-1 cells to activate the HIV-1 promoter. Q-VD-Oph did not increase the ability of KHYG-1 supernatants to induce HIV-1 promoter activity, whereas emricasan activated the HIV-1 promoter similar to Z-VAD-FMK ( Figure 3C ). Heat-inactivation of the supernatants at 60°C before adding them to the TZM-bl cells, did not alter the capacity of supernatants of Z-VAD-FMK-treated KHYG-1 cells to induce HIV-1 promoter activity ( Figure 3D ). Heat-inactivation at 99°C abolished HIV-1 promoter activation by the supernatants.

Next, we wanted to assess the kinetics of Z-VAD-FMK induced HIV-1 promoter activation. In a time course experiment, we observed an increase in HIV-1 promoter activity with supernatants of KHYG-1 treated with Z-VAD-FMK for 8 h, followed by a plateau ( Figure 4A ). We also separated the supernatant of Z-VAD-FMK-treated KHYG-1 cells into two fractions based on size with a cut-off of 3 kDa. The fraction below 3 kDa showed no activity, whereas the fraction above 3 kDa showed the same activity as unfractionated supernatants ( Figure 4B ). Finally, we observed that Z-VAD-FMK-treated KHYG-1 cells were forming clumps ( Figure 4C ), indicating a proliferating or activated phenotype. However, since there was no increase in proliferation ( Figure 1D , and data not shown), it likely represented activation of the cells. Altogether, these data suggest that Z-VAD-FMK-treated KHYG-1 cells induce HIV-1 promoter activity in target cells most likely through a secreted protein.

Since NK cells can exert their functions through the secretion of a variety of cytokines, we examined whether there was a difference in the cytokine protein profile of Z-VAD-FMK-treated KHYG-1 cells that could inform how HIV-1 reactivation occurred in the target cells. We collected supernatants of control or Z-VAD-FMK-treated KHYG-1 cells and subjected the supernatants to a cytokine array that can detect protein levels of 36 different cytokines ( Figure 5A ; Supplementary Figure 2 ). Since KHYG-1 requires recombinant human IL-2 for growth in culture, we detected equal amounts of IL-2 in both supernatants. MIF and SerpinE1 were also identified in equal levels in the supernatants of control and Z-VAD-FMK-treated KHYG-1. However, protein levels of the cytokines MIP1α (CCL3), MIP1β (CCL4) and RANTES (CCL5) increased in the supernatants of Z-VAD-FMK-treated KHYG-1 cells. These three cytokines are ligands for the HIV-1 entry co-receptor CCR5 and act as HIV-1 suppressors by blocking entry of HIV-1 into the target cells (22–24). Then, we examined the effect of Z-VAD-FMK on the mRNA levels of CCL3, CCL4, and CCL5 within KHYG-1 cells by qPCR. Both CCL3 and CCL4 mRNA levels were increased after Z-VAD-FMK treatment, whereas CCL5 mRNA levels were not affected by the treatment ( Figure 5B ).

To assess whether the cytokines CCL3, CCL4, and CCL5 were responsible for the HIV-1 reactivation, we incubated TZM-bl cells with different physiological concentrations of each of the cytokines. Neither individually nor in combinations did these cytokines induce HIV-1 promoter activity ( Figure 5C ). Also, at a non-physiologically high dose we could not detect HIV-1 promoter activation by these chemokines ( Supplementary Figure 3A ). Although the J-Lat cells used in this study did not express CCR5 as measured by flow cytometry ( Supplementary Figure 3B ), we did also examine whether these cytokines could induce HIV-1 reactivation in these cells in a CCR5-independent manner. No increase in GFP expression was observed after J-Lat cells were incubated with any or combinations of the three cytokines ( Supplementary Figure 3C ). Altogether, Z-VAD-FMK induced secretion of CCL3, CCL4, and CCL5 by KHYG-1 cells, but these HIV-1 suppressor cytokines were not responsible for the reactivation of HIV-1.

To study the effect of pan-caspase inhibitors under more physiological conditions, we assessed whether Z-VAD-FMK can modulate primary NK cells to secrete HIV-1 latency reversal factors. We purified primary NK cells from primary PBMCs and after 4 days of ex vivo culture, we incubated the cells with Z-VAD-FMK, or DMSO as control. Although we obtained highly purified primary NK cell cultures following enrichment protocols, the culture showed low viability and proliferation in the following days as well as inconsistent results on HIV-1 promoter activation (data not shown). Since primary NK cells are highly dependent on cytokines and stimulatory signals from other (feeder) cells (25), we performed a semi-enrichment that yielded less pure primary NK cell cultures but could be consistently cultured ex vivo for several days. Following NK cell dynamics of CD3^-CD56^dimCD16⁺ and CD3^-CD56^brightCD16^- NK cells together in cultures from four donors showed high variation in the percentage of NK cells within PBMCs at the start of the semi-enrichment (3%-16%) and during the experiment ( Figure 6A ; Supplementary Figures 4–6 ). Since CD3⁺ T cells proliferate at a higher rate in response to IL-2 in the first days of ex vivo culture compared to NK cells, there is a minor drop in NK cell frequency. The one-day treatment of Z-VAD-FMK decreased the percentage of NK cells within the culture population ( Figure 6B ). This is most likely due to increased survival of mainly other cells in the culture where cell death was inhibited by Z-VAD-FMK. Analysis of semi-enriched ex vivo NK cell cultures from eight anonymous donors showed that supernatants of Z-VAD-FMK-treated cultures increased activation of the HIV-1 promoter compared to the control cultures ( Figure 6C ). This activation appears to be mediated by NK cells, since supernatant of Z-VAD-FMK-treated whole PBMCs show very low increased HIV-1 promoter activation ( Figure 6D ). Finally, we extracted mRNA from all cells in the Z-VAD-FMK-treated or control ex vivo semi-enriched primary NK cell cultures. Similar to the KHYG-1 cell line, we observed an increase in CCL3 and CCL4 mRNA expression in the Z-VAD-FMK-treated cells ( Figure 6E ). This data indicates that treatment of primary NK cells with a pan-caspase inhibitor induces the upregulation of CCL3 and CCL4 mRNA levels as well as secretion of a latency reversal factor.