Sustainable Antiviral Efficacy of Rejuvenated HIV-Specific Cytotoxic T Lymphocytes Generated from Induced Pluripotent Stem Cells

ABSTRACT

Persistence of HIV latently infected cells is a barrier to HIV cure. The “kick and kill” strategy for a cure includes clearance of the viral reservoir by HIV-specific cytotoxic T lymphocytes (CTLs). However, exhaustion and senescence of T cells accelerates during HIV infection, and does not fully recover, despite complete viral suppression under antiretroviral therapy. We previously established an induced pluripotent stem cell (iPSC) from a parental HIV-specific CTL clone and generated an iPSC-derived rejuvenated HIV-specific CTL clone (iPSC-CTL), which exhibited an early memory phenotype, high proliferation capacity and effector functions in vitro. Here, we assessed the antiviral efficacy of the HIV-specific iPSC-CTL by single- and multiple-round viral suppression assays (VSAs). The HIV-specific iPSC-CTL suppressed viral replication in an HLA-dependent manner with equivalent efficacy to the parental CTL clone in single-round VSA. In multiple-round VSA, however, the ability of the iPSC-CTL to suppress viral replication was longer than that of the parental CTL clone. These results indicate that HIV-specific iPSC-CTL can sustainably exert suppressive pressure on viral replication, suggesting a novel approach to facilitate clearance of the HIV reservoir via adoptive transfer of rejuvenated CTLs.

IMPORTANCE Elimination of latently HIV-infected cells is required for HIV cure. In the “kick and kill” strategy proposed for a cure to HIV, the host immune system, including HIV-specific cytotoxic T lymphocytes (CTLs), play a central role in eliminating HIV antigen-expressing cells following reactivation by latency-reversing agents (LRAs). However, CTL dysfunction due to exhaustion and senescence in chronic HIV infection can be an obstacle to this strategy. Adoptive transfer with effective HIV-specific CTLs may be a solution of this problem. We previously generated an induced pluripotent stem cell (iPSC)-derived rejuvenated HIV-specific CTL clone (iPSC-CTL) with high functional and proliferative capacity. The present study demonstrates that iPSC-CTL can survive and suppress HIV replication in vitro longer than the parental CTL clone, indicating the potential of iPSC-CTL to sustainably exert suppressive pressure on viral replication. Adoptive transfer with rejuvenated HIV-specific CTLs in combination with LRAs may be a new intervention strategy for HIV cure/remission.

INTRODUCTION

Despite effective antiretroviral therapy (ART), human immunodeficiency virus (HIV) remission and cure have not been achieved, with the exception of rare cases, due to long-lived latently HIV-infected cells evading viral clearance. A “kick and kill” strategy has been proposed to eradicate the HIV reservoir by reactivation with latency-reversing agents (LRAs) and induction of cell death by viral cytopathic effects and/or host immune responses (1). HIV-specific cytotoxic T lymphocytes (CTLs) could be a key effector for eradication of reactivated HIV-infected cells. However, phenotypic and functional impairment of T cells are frequently induced by persistent immune activation in chronic HIV infection (2–6). Immunosenescence is accelerated under these conditions, with accumulation of senescent CD28-CD57⁺CD8⁺ T cells in individuals living with HIV, which is not fully recovered by ART (7–10).

Combination therapies using various LRAs and therapeutic vaccines to induce HIV-specific CTLs have been assessed in clinical trials, but failed to show sufficient viral clearance in “kick and kill” strategies (11). Adoptive transfer of CTLs may be an alternative strategy to enhance effector function for elimination of the viral reservoir, but there are several critical hurdles. Allogeneic lymphocyte transfer to an HIV-infected progressor from an elite controller, resulted in only a transient reduction in viral load, followed by virus rebound after clearance of the donor cells from the recipient’s peripheral blood by day 8 (12). Allogeneic lymphocyte transfer in cynomolgus macaques has shown rapid clearance of donor cells from peripheral blood and lymphoid tissues, while adoptively transferred autologous lymphocytes stably persisted during the study period (13). These data indicate the limitation of allogeneic cells for adoptive transfer in “kick and kill” strategies. Additionally, adoptively transferred autologous simian immunodeficiency virus (SIV)-specific CD8⁺ T cell clones were rapidly lost from peripheral blood, possibly due to cell exhaustion from extensive in vitro expansion prior to infusion and there was no reduction in viral load (14). Thus, conventional technology is unlikely to provide HIV-specific CTLs with the potential to sustainably exert clinically relevant antiviral function after adoptive transfer.

Induced pluripotent stem cell (iPSC) technology allows reprogramming of a somatic cell into a pluripotent stem cell with the potential to differentiate into any cell type. Several cell types have been successfully generated from iPSC and application of this technology for therapeutic interventions against a variety of diseases is in progress (15). Although there are challenges that need to be addressed for the clinical use of iPSC technology (15), new therapeutic strategies using iPSC technology should be considered for diseases without effective therapies. Induced pluripotent stem cells can be established from T-cell clones and re-differentiated into functional T cells, with antigen-specificity maintained (16, 17). CD8⁺ T lymphocytes derived from iPSC with anti-tumor activities are a promising new cell source for cancer immunotherapy (18–21). We previously established an iPSC from an HIV-specific CTL clone obtained from an individual living with HIV and successfully generated a rejuvenated HIV-specific CTL clone derived from the iPSC (iPSC-CTL) (17). Rejuvenated iPSC-CTL exhibited an early memory phenotype with CD45RA and CD27 expression after re-differentiation into CD8 single-positive T cells, with the potential to become highly proliferating effector cells (22). In the present study, we examined the potential of HIV-specific iPSC-CTL to sustainably exert suppressive pressure on HIV replication.

RESULTS

Characterization of HIV-specific iPSC-CTL.

H254SeV3, an iPSC-CTL specific for the HLA-A*24:02-restricted CTL epitope, Nef134-8 (RYPLTFGW), was previously generated from a CTL clone of the same specificity, H25-4, by re-differentiation into CD8-single positive T cells by our newly established technique (22). The iPSC-CTL used in this study was stimulated and expanded 6 to 7 times with PHA and irradiated peripheral blood mononuclear cells (PBMCs), resulting in greater than 10²⁰-fold expansion (22). The phenotypic profile of iPSC-CTL was compared with the parental Nef134-8-specific CTL clone, H25-4, by flow cytometry (Fig. 1A). The levels of CD3, CD8α, and CD8β expression in iPSC-CTL were similar to H25-4. Both stained with HLA-A*24:02/Nef134-8 tetramer at an equivalent level, indicating that expression of the epitope-specific TCR is maintained in iPSC-CTL. Similar to the parental clone, iPSC-CTL exhibited an effector memory (CD45RA-/CCR7-/CD27-/CD28-) phenotype. Following single-round expansion, iPSC-CTL showed an early memory phenotype (22), suggesting that these cells can differentiate during multiple in vitro expansion cycles. Expression of the activation markers HLA-DR and CD38 was also comparable to the parental clone, without expression of the exhaustion and senescence markers, PD-1 and CD57, respectively.

FIG 1.

Characterization of HIV-specific iPSC-derived CTL. (A) Expression of surface markers on iPSC-CTL and the parental clone, H25-4. Isotype controls are shown in gray. (B) Functional avidity of iPSC-CTL and the parental clone. CTLs (1 × 10³) were stimulated with A24-GXR or CEM-GXR cells pulsed with serial dilutions of Nef134-8 peptide. IFNγ production was measured by the ELISpot assay. Percentage of the maximal IFNγ response of each CTL is shown. The mean and standard deviation (SD) from five independent experiments are shown.

TCR functional avidity was measured by IFNγ ELISpot assay with serially diluted Nef134-8 peptides. The iPSC-CTL and parental clone showed similar levels of responses (Fig. 1B), indicating that the functional avidity of the iPSC-CTL is equivalent to the parental clone.

Single-round in vitro viral suppression assay.

We performed a single-round viral suppression assay (VSA) to examine the ability of the iPSC-CTL to inhibit virus replication in vitro using a green fluorescent protein (GFP)-reporter cell line, CEM-GXR, stably transfected with a Tat-inducible LTR-GFP (Fig. 2A) (23). When HIV-infected A24-GXR and CEM-GXR cells were cultured for 1 week without effectors, the frequency of GFP-positive cells increased to 50% to 70% (Fig. 2B). However, when HIV-infected cells were cocultured with either the iPSC-CTL or the parental clone, the frequency of GFP-expression was less than 10% in A24-GXR cells. Reduction in HIV replication (GFP positive cell frequency) by iPSC-CTL on day 6 was more than 90% (GFP-expressing cell frequencies were 6.24% and 68.4% with and without iPSC-CTL, respectively). In contrast, in CEM-GXR cells showing equivalent or rather less efficient HIV replication without effectors compared with A24-GXR, reduction in GFP-expressing cell frequencies was not observed even in the presence of the iPSC-CTL or the parental clone. These observations were confirmed in a second independent experiments. These results indicate that HIV replication was suppressed by the iPSC-CTL through a TCR-dependent antigen-specific manner. Despite downregulation of HLA class I expression in A24-GXR cells after NL4-3 infection (Fig. 2C), in vitro viral replication was completely suppressed, suggesting that iPSC-CTL can efficiently recognize and eliminate HIV-infected cells even with lowered HLA class I expression.

FIG 2.

Single-round viral suppression assay. (A) Representative gating strategy used to define the frequency of GFP-expressing cells in total target cells. Numbers in the GFP vs FSC-A plots indicate the percentages of GFP-expressing cells in the violet-negative subset. (B) Frequency of GFP-expressing cells in the violet-negative subset. (C) Surface expression of HLA class I on A24-GXR cells infected with HIV NL4-3. PE-labeled anti-HLA class I antibody was used for the staining. Numbers in the plots indicate the percentages of GFP-expressing cells. A representative result in two independent experiments is shown.

Antigen-dependent proliferation capacity of HIV-specific iPSC-CTL.

Adoptively transferred CTLs for HIV cure interventions are required to have not only effector functions, including cytokine production and killing activity, but also proliferative capacity for sustained viral control. To evaluate proliferation following antigenic stimulation, iPSC-CTL and the parental clone labeled with violet dye were cocultured with reactivated latently HIV-infected cells: ACH-2 or HLA-A*24:02-expressing ACH-2 (A24-ACH). Reactivation with TNF-α induced HIV p24 expression in more than 80% of both latently-infected cells (Fig. 3A). Co-culture of iPSC-CTL with reactivated A24-ACH, but not ACH-2, resulted in proliferation on day 3, which continued through day 6 (Fig. 3B). Both iPSC-CTL and the parental clone, showed modest proliferation when cocultured with non-reactivated A24-ACH cells, for which 8.92% of target cells expressed p24. The decline in median fluorescent intensity (MFI) indicated approximately four division cycles of iPSC-CTL over 6 days (Fig. 3C). The antigen-specific proliferation capacity of iPSC-CTL was equivalent to the parental clone. These data indicate that iPSC-CTL efficiently proliferate following stimulation by reactivated latently infected cells.

FIG 3.

Proliferation of HIV-specific iPSC-CTL after stimulation by reactivated latently HIV-infected cells. (A) Expression of HIV-p24 48 h after activation by TNF-α in A24-ACH and ACH-2 cells. ACH-2 cells were used for nonspecific controls while the reason of the SSC change in ACH-2 after stimulation is unclear. (B) HIV-specific iPSC-CTL (red) or the parental clone, H25-4 (blue) were cocultured with ACH-2 or A24-ACH with (bold line) or without (filled) 48-h TNF-α treatment before coculture. Violet histograms of gated CD3⁺CD8⁺ cells are shown. Histograms of CTLs before coculture (day 0) are shown by a light gray line. (C) Decay of violet in the iPSC-CTL and the parental clone cocultured with A24-ACH cells or ACH cells with or without TNF-α treatment. Relative median MFI of violet are shown. A representative result in two independent experiments is shown.

Multiple-round in vitro viral suppression assay.

We established a multiple-round VSA to investigate the longevity of iPSC-CTL for virus suppression in vitro. Co-culture of violet-labeled CTLs and HIV-infected target cells were maintained for 24 days with the addition of HIV-infected target cells every 3 days, and GFP expression was monitored (Fig. 4A). During coculture of iPSC-CTL with HIV-infected A24-GXR cells, viral replication was suppressed by approximately 80% through day 12, compared with the culture without CTL (3.48% versus 20.4% on day 12, respectively), and although the level of suppression gradually declined, it was sustained until day 21 (12.1% [coculture] versus 23.2% [no iPSC-CTL]) (Fig. 4B and C). In contrast, GFP-expressing target cell frequencies in the coculture with the parental clone, H25-4, increased from day 15, and remained elevated at levels similar to the culture without CTLs. These results indicate that the potency of iPSC-CTL to suppress viral replication is maintained longer than the parental clone. There was no difference in GFP-expressing cell frequencies between the samples, irrespective of the presence of either iPSC-CTL or H25-4, when CEM-GXR cells without A24 expression were used as target cells (Fig. 4D), confirming that the extended viral suppression by iPSC-CTL is HLA-dependent. These results were confirmed in an additional independent experiment.

FIG 4.

Multiple-round viral suppression assay. (A) Experimental scheme of multiple-round VSA. (B) Gating strategy used to define target cells and effectors. The percentage of GFP expressing cells in the violet-negative subset in the target gate (black) and the percentage of violet-positive subsets in the CTL gate (red) are shown. The plots on days 12 are shown as representative data. (C, D) Frequency of GFP expressing cells before (circles) and after (triangles) addition of HIV-infected target cell every 3 days in A24-GXR (C) and CEM-GXR (D) target cells. A representative result in two independent experiments is shown.

To understand the mechanism of extended viral suppression by iPSC-CTL, we monitored the frequency of violet-positive cells (CTLs) and the MFI of violet dye (Fig. 5). A profound decrease in violet-positive cells was reproducibly observed after day 12 in the parental clone when cocultured with A24-GXR cells, compared with iPSC-CTL (Fig. 5A and B). In contrast, no significant difference in MFI decay kinetics was observed between iPSC-CTL and the parental clone (Fig. 5C), indicating equivalent proliferation in both of these CTL clones. These results suggest that the extended viral suppression by iPSC-CTL is due to its longer survival but not higher proliferation.

FIG 5.

Survival of HIV-specific iPSC-CTL and the parental clone in a multiple-round VSA. (A) The pseudo-color plots of violet vs FSC-A in the CTL gate are shown. Numbers in the plots indicate the percentage of violet-positive subset (the same as Fig. 4B). Frequency (B) and relative violet MFI (C) of violet-positive subset in the CTL gate are shown. Data for the parental clone cocultured with A24-GXR cells are missing after day 21 due to their absence from the gate. A representative result in two independent experiments is shown.

Susceptibility to activation-induced cell death of HIV-specific iPSC-CTL.

Repeated encounter with antigen induces activation-induced cell death (AICD) to CD8⁺ T cells (24). In order to evaluate the sensitivity to AICD of the iPSC-CTL and parental clone, the frequencies of Annexin V⁺/propidium iodide (PI)- (early apoptotic cells) and Annexin V⁺ cells (total apoptotic cells) after TCR stimulation were analyzed. The frequency of apoptotic cells was higher in the parental clone than the iPSC-CTL (Fig. 6A and B), implying that the iPSC-CTL is less susceptible to AICD than the parental clone. The expression of functional molecule, perforin, that is crucial for CTL cytotoxicity, after antigen stimulation was also examined. The expression level was equivalent in both CTLs (Fig. 6C). These results suggest that the extended viral suppression by the iPSC-CTL is attributed to longer survival of the iPSC-CTL due to its less susceptibility to AICD compared to the parental clone.

FIG 6.

Susceptibility to antigen-induced cell death of HIV-specific iPSC-CTL and the parental clone. (A) The pseudo-color plots of PI vs Annexin V of unstimulated (lower panels) and stimulated (upper panels) HIV-specific iPSC-CTL (left panels) and the parental clone (right panels) 0 h (just after stimulation) and 6 h after stimulation by HLA-A*24:02/Nef134-8 tetramer. (B) Frequency of early (Annexin V⁺/PI-, left panels) and total (Annexin V⁺, right panels) apoptotic cells in unstimulated (lower panels) or stimulated (upper panels) CTLs after 0, 2, 6, and 24 h. (C) Expression of functional molecules after antigen stimulation. HIV-specific iPSC-CTL (left panels) and the parental clone (right panels) were cocultured with an HLA-A*24:02-expressing B-LCL pulsed with (upper panels) or without (lower panels) peptide for 6 h. Perforin and IFNγ expression in the CD3⁺CD8⁺ cells are shown. A representative result in two independent experiments is shown.

DISCUSSION

In this study, we show the antiviral efficacy of iPSC-derived HIV-specific CTL in vitro. The iPSC-CTL was shown to have the potential to proliferate following antigenic stimulation with reactivated latently infected cells and suppress HIV replication during multiple rounds of productive HIV infection. This study is the first, to our knowledge, to describe the successful establishment of a human virus-specific CTL with effective antiviral activity using induced pluripotent stem cell technology.

We established a flow cytometry-based VSA using fluorescence-labeled CTLs and CEM-GXR cells, in which HIV-infected cells could be monitored, while simultaneously allowing assessment of the proliferation kinetics of the effector cells without additional staining. Multiple-round VSA showed that HIV replication was suppressed by iPSC-CTL for a longer time relative to the parental clone, which showed rapid loss of CTL when cocultured with A24-GXR. Frequency of apoptotic cells after antigen stimulation was lower in the iPSC-CTL compared to the parental clone, suggesting that longer survival of the iPSC-CTL could be attributed to resistance to AICD due to rejuvenation.

Several classes of LRAs with distinct mechanisms to reactivate latently HIV-infected cells have been proposed for HIV cure and some have been evaluated in clinical trials. Although various LRAs showed potent virus reactivation in vitro (25–30), no clinical trials to date have shown a successful reduction of the HIV reservoir size by the LRA alone (31–36), suggesting that latency disruption will need to be paired with other therapeutic approaches to eliminate HIV-infected cells.

Although CTLs are intrinsic in eliminating virus-infected cells, exhaustion and senescence of T cells are observed in HIV-infected individuals, which is not fully reversed after prolonged successful ART (8, 37). Furthermore, the number of aging people living with HIV is increasing. Normal aging promotes functional immunological changes, such as increased inflammatory cytokines, low naive/memory ratio, increased T cell activation, and decreased vaccine responsiveness (38). Therapeutic vaccination to enhance HIV-specific CTLs might not work as efficiently in the elderly or individuals for whom immunonsenescence may be elevated prior to ART initiation.

Clinical trials of therapeutic vaccination to enhance T cell responses against HIV conserved regions, combined with LRA (vorinostat and romidepsin), in ART-suppressed individuals during acute/recent infection, have been performed to evaluate the “kick and kill” strategy (11, 39). Although the vaccines elicited robust and broad HIV-specific T cells, and the combined intervention showed a reduction in HIV total DNA levels and extended viremic control after ART interruption in some patients, there was no significant benefit on HIV reservoir reduction in most cases. Combination therapy with LRAs and vaccines to enhance HIV-specific T cells may not achieve HIV cure in individuals living with HIV.

In vitro expansion with antigenic stimulation is a necessary procedure for adoptive transfer using HIV-specific CTLs to provide sufficient numbers for effective viral control. However, this often results in the effectors being terminally differentiated and susceptible to apoptosis. Induced pluripotent stem cells have the potential for infinite proliferation, and generation of iPSC-derived CD8⁺ T cells with early-memory phenotype has been achieved (22). Rejuvenated iPSC-CTL could persist with antiviral activity for a longer time following adoptive transfer in vivo and help reservoir elimination during intermittent reactivation of latently HIV-infected cells by LRAs. The feasibility of cure by the kick and kill strategy is still under discussion, and functional cure may be a realistic goal to be aimed for. The iPSC-CTL could contribute to sustainable viral control, possibly leading to functional cure. It is difficult to use primary target cells for multiple-round VSA, and all the experiments have been performed using established cell lines to show antiviral activity of HIV-specific iPSC-CTL in the present study. Further analyses using primary cells and animal models may be the next issues.

A major concern regarding CTL-based immunotherapy against HIV/AIDS is the pre-existence of CTL escape mutants. Adoptive transfer of a single epitope-specific CTL clone led to selection of escape variants in a clinical trial in the pre-ART era (40). It is critical for successful adoptive transfer to select CTL targeting peptides with susceptible form to CTL recognition. A recent report noted that functional CTL targeting of peptides that contain topologically important viral residues, where it is strongly related to low viral sequence entropy, contributes to immune control in HIV infection (41). Especially for functional cure, it would be beneficial to identify CTL targets intolerant to amino acid change by applying structure-based network analysis. Adoptive transfer with a mixture of HIV-specific iPSC-CTLs specific for multiple epitopes may also be effective to suppress HIV escape variants.

We recently reported an efficient method to generate functional antigen-specific CD8⁺ T cells from iPSCs transduced with a TCR for clinical application (42). Applying these new technologies could allow the development of HIV-specific iPSC-CTLs for immunotherapy, in which quantity (the number of iPSC-CTL) and quality (antigen specificity, differentiation status, and function of each iPSC-CTL) could be personalized.

In summary, the present study showed sustainable antiviral activity of iPSC-derived rejuvenated HIV-specific CTLs in vitro. HIV-specific CTLs generated with iPSC technology may contribute to the development of a new strategic intervention toward HIV cure.

MATERIALS AND METHODS

iPSC-CTL and the parental CTL clone.

An HIV-specific iPSC-derived CD8⁺ T cell clone (iPSC-CTL), H254SeV3, was established from an HLA-A*24:02-restricted HIV-1 Nef134-8-specific CTL clone, H25-4, as previously described (17, 22). T cell expansion was performed by stimulating iPSC-CTL and the parental CTL clone (H25-4) with 5 μg/mL PHA-P (Wako, Osaka, Japan) and coculture with irradiated PBMCs in α-MEM (Thermo Fisher Scientific) supplemented with 20% heat-inactivated fetal calf serum (FCS), ITS (Insulin, Transferrin, Selenium Solution) (Thermo Fisher Scientific), 50 μg/mL L-ascorbic acids (Sigma-Aldrich), 100 U/mL of penicillin, 100 ng/mL of streptomycin, and 2 mM l-glutamine (Sigma-Aldrich) (M20), in the presence of 10 ng/mL IL-7 and IL-15 (Miltenyi Biotec). CTLs were used for all assays following 2 weeks of stimulation.

Cell lines and virus.

CEM-GXR, a GFP reporter CD4⁺ T cell line which encodes GFP driven by HIV-1 LTR, kindly gifted from Mark Brockman (23), and ACH-2, a latently-HIV-infected CD4⁺ T cell line (43, 44), were cultured in RPMI 1640 (Sigma-Aldrich) supplemented with 10% heat-inactivated FCS, 100 U/mL of penicillin, and 100 U/mL of streptomycin (R10). HLA-A*24:02 was transduced into CEM-GXR and ACH-2 cells using a lentivirus-vector, pLenti6.3/V5-DEST (Invitrogen), to establish HLA-A*24:02-expressing CEM-GXR (A24-GXR) and ACH-2 (A24-ACH) cells. Expression of HLA-A*24:02 on A24-GXR and A24-ACH cells was confirmed by flow cytometry using biotinylated anti-HLA-A9 serotype antibody (One Lambda) and streptavidin-phycoerythrin (PE) (Molecular Probes) (Fig. 7). HLA-A*24:02 is included in the HLA-A9 serological family (45).

FIG 7.

Expression of HLA-A*24:02 in A24-GXR and A24-ACH cells. HLA-A*24:02 expression on CEM-GXR and ACH-2 cells transduced with the HLA-A*24:02 gene was assessed using anti-HLA-A*24:02 antibody. Histograms of each cell line with (red) or without (black) HLA-A*24:02 are shown. Histograms of the isotype control are also shown (filled light gray).

HIV-1 NL4-3 strain was propagated by transfection of human 293T cells with plasmid pNL4-3 using Lipofectamine 2000 (Invitrogen) and expanded on PM1 cells.

Flow cytometry.

The following antibodies and reagent were used for flow cytometric analysis: anti-CD57–fluorescein isothiocyanate (FITC), anti-CD27-FITC, anti-CD8α-FITC, anti-HLA-class I-PE, anti-CD3-PE, anti-Perforin-PE, anti-CD8α-peridinin chlorophyll protein (PerCP), anti-CD3-PerCP, anti-CD38-PerCP, anti-CCR7-PE-Cy7, anti-PD-1-allophycocyanin (APC)-Cy7, anti-CD45RA-APC-Cy7, anti-HLA-DR-APC-Cy7, anti-IFNγ-APC-Cy7, anti-CD3-Pacific Blue (PB), anti-CD28-PB, anti-CD8-PB (BioLegend), anti-CD8β-PE (Beckman Coulter), and PE-conjugated HLA-A24/Nef-134-8 tetramer. LIVEDEAD^TM Fixable Aqua Dead Cell Stain kit (Thermo Fisher Scientific) was used to monitor cell viability. Anti-HIV-1 p24 antibody, Nu24, labeled with Alexa Fluor 647 (46) was used for intracellular staining of HIV-infected cells. All flow data were acquired using a FACS Canto II (Becton Dickinson) and the analysis was performed using FlowJo software ver.9.6.6 or ver. 10.8.0 (Tree Star).

Interferon gamma (IFNγ) ELISpot assay.

The ELISpot assay was performed as previously described (47). One-thousand CTL were cultured for 18 h with 1 × 10⁵ A24-GXR or CEM-GXR cells pulsed with 5X-serial dilutions of synthetic Nef134-8 peptide (Sigma-Genosys). Spots were counted using an ImmunoSpot analyzer (Cellular Technology Limited). Results are expressed as the percentage of the maximal responses.

Proliferation analysis.

The iPSC-CTL and parental clone were labeled with 5 μM violet (CellTrace, Thermo Fisher Scientific) and cocultured with A24-ACH or ACH-2 cells, reactivated with 10 ng/mL TNF-α (Miltenyi Biotec) for 48 h prior to coculture in M20, without any cytokines. The cells were stained with anti-CD3-PB and anti-CD8-PerCP on days 3 and 6 and analyzed using FACS Canto II. The MFI of violet dye (violet) in CD3⁺CD8⁺ cells was monitored. Results are shown as the decline of violet MFI relative to that on day 0. The experiments were performed twice independently.

Viral suppression assay (VSA).

Target cells for VSA were prepared by infecting 1 × 10⁶ A24-GXR and CEM-GXR cells with 1,000 and 1,500 TCID₅₀ HIV-1 NL4-3, respectively, for 2 h in the presence of 10 μM Diethylaminoethyl cellulose (DEAE) and cultured in R10. HIV infection was monitored by GFP expression. HIV-infected A24-GXR or CEM-GXR cells were cocultured with violet-labeled iPSC-CTL or the parental clone at an effector: target (E:T) ratio of 1:3 in M20 supplemented with 1 ng/mL each of IL-7 and IL-15, 24 h after infection. The frequency of GFP-expressing cells in the violet-negative subset was monitored from day 3 following coculture, every 24 h through day 9. The level of HLA class I expression on the cell surface was measured 4 days after HIV infection in an independent experiment.

A multiple-round VSA was established to evaluate long-term viral suppression by iPSC-CTL. First, 1 × 10⁶ A24-GXR and CEM-GXR cells were infected with 100 and 300 TCID₅₀ NL4-3, respectively. Infected cells were used as targets after 72 h of culture in R10, at which time, the percentage of GFP-positive cells was less than 2%. violet-labeled iPSC-CTL or the parental clone (3.3 × 10⁵) were cocultured with HIV-infected target cells (1 × 10⁶) in M20 supplemented with 1 ng/mL each of IL-7 and IL-15 at an E:T ratio of 1:3, and GFP expression in the target cells (violet-negative subset) was monitored on day 3. Half of the cultured cells were then used for the second-round of the assay, by adding 1 × 10⁶ A24-GXR or CEM-GXR cells infected with NL4-3 for 3 days (as performed for the first round), to provide additional HIV-infected target cells. The subculture was repeated 7 times over 24 days. In the FSC-A versus SSC-A plot, target cells and CTLs were gated, respectively. The frequency of GFP-expressing cells in the violet-negative target cells and the frequency of the violet-positive CTLs were monitored. VSAs were performed twice independently.

Analysis of activation-induced cell death (AICD).

The iPSC-CTL and the parental clone were cultured in the presence or absence of HLA-A*24:02/Nef134-8 tetramer (1.25 μg/mL of pHLA) for 0, 2, 6, and 24 h. Cell death was quantified using Annexin V apoptosis detection kit (Thermo Fisher Scientific) according to the manufacturer’s instructions. Analyses were performed twice independently.

Intracellular cytokine staining.

The iPSC-CTL and parental clone were stimulated with an HLA-A*24:02-expressing B lymphoid cell line (B-LCL) pulsed with 100 nM Nef134-8 peptide in the presence of anti-CD28, anti-CD-49d and monensin (GoldiStop) (BD Bioscience). Following incubation for 6 h, intracellular cytokine staining was performed as previously described (48).