The latency-reversing agent HODHBt synergizes with IL-15 to enhance cytotoxic function of HIV-specific T cells

Abstract

IL-15 is under clinical investigation toward the goal of curing HIV infection because of its abilities to reverse HIV latency and enhance immune effector function. However, increased potency through combination with other agents may be needed. 3-Hydroxy-1,2,3-benzotriazin-4(3H)-one (HODHBt) enhances IL-15-mediated latency reversal and NK cell function by increasing STAT5 activation. We hypothesized that HODHBt would also synergize with IL-15, via STAT5, to directly enhance HIV-specific cytotoxic T cell responses. We showed that ex vivo IL-15 + HODHBt treatment markedly enhanced HIV-specific granzyme B-releasing T cell responses in PBMCs from antiretroviral therapy-suppressed (ART-suppressed) donors. We also observed upregulation of antigen processing and presentation in CD4+ T cells and increased surface MHC-I. In ex vivo PBMCs, IL-15 + HODHBt was sufficient to reduce intact proviruses in 1 of 3 ART-suppressed donors. Our findings reveal the potential for second-generation IL-15 studies incorporating HODHBt-like therapeutics. Iterative studies layering on additional latency reversal or other agents are needed to achieve consistent ex vivo reservoir reductions.

Keywords: AIDS/HIV; Adaptive immunity; Cellular immune response; Immunology; Immunotherapy.

Figure 1. HODHBt synergizes with IL-15 to enhance HIV-specific cytotoxic CD8⁺ T cell responses ex vivo.

(A) Representative granzyme B ELISPOT results (from 1 of 14 donors). The indicated HIV and CMV antigens were represented by overlapping peptide pools. HIV-Gag, -Pol, and -Nef stimulations were each performed in triplicate. HIV-Env peptide, CMV-pp65, no peptide, and phytohemagglutinin (PHA) were performed in duplicate. (B and C) Combined ELISPOT results from the A5321 cohort of 14 ART-treated donors. (B) For each peptide stimulation condition (or control), results are presented as fold-change relative to Media. Shown are means ± SD. P values were calculated by Friedman’s ANOVA test (1 way), with all significant P values displayed. (C) Results are from the same data set as B, plotted to show pairing of IL-15 + DMSO and IL-15 + HODHBt conditions across participants. Shown are SFU per million PBMCs after subtracting the background from each corresponding no-peptide treatment condition. P values were calculated using 2-tailed, paired, nonparametric Wilcoxon’s tests.

Figure 2. Treatment with the combination of IL-15 and HODHBt enhances surface MHC-I and antigenicity of CD4⁺ T cells.

(A and B) scRNA-Seq results. (A) Heatmap of 40 genes that were differentially expressed (FDR < 0.05) between the treatments: DMSO (control), IL-15 + DMSO, and IL-15 + HODHBt; n = 3 donors. (B) GO terms identified by genes differentially expressed in T cells between IL-15 + DMSO and IL-15 + HODHBt conditions. P values were calculated by 1-sided Fisher’s exact test, and the size of each dot represents the ratio of input genes that are annotated in a term. (C) Flow cytometry results showing fold-changes in median fluorescence intensity (MFI) of MHC-I, relative to the DMSO control. Shown are individual data points for 6 donors, with mean ± SD. Donors with HIV are indicated with an x and donors without HIV as filled circles. P values were calculated by 1-way ANOVA with Dunnett’s multiple-comparison test. (D) PBMCs from ART-treated donor OM5220 were treated with DMSO, IL-15 (1 ng/mL), and HODHBt (50 μM), separately or in combination, for 4 days. CD4⁺ T cells were pulsed with a 15-mer peptide containing the RV9 epitope, then cultured or without with an autologous RV9-specific CD8⁺ T cell clone. Upper panel, surface MHC-I by treatment condition in no-clone conditions; data points indicate technical replicates. Lower panel, flow cytometry data from CD8⁺ T cell clone conditions showing the percentages of CD107a⁺ cells (degranulated). Shown are MFI ± coefficient of variation of CD107a from at least 18,000 viable CD8⁺ T cells. (E) CD8-depleted PBMCs from an HLA-B58⁺ donor without HIV were activated and infected with HIV_JRCSF or maintained as uninfected controls. After 60 hours, the antiretroviral agent T20 was added, and cells were treated with DMSO, IL-15 (20 μg/mL), and HODHBt (100 μM), separately or in combination. A CD8⁺ T cell clone specific for the HLA-B58–restricted epitope TW10 was then added to each culture across 3 replicates. Upper panel, flow cytometry data showing surface MHC-I levels. Data points indicate technical replicates and error bars represent SD. Lower panel, flow cytometry data showing the percentages of CD107a⁺CD8⁺ T cell clone. Shown are medians of technical replicates ± SD. DEG, differentially expressed gene; BP, biological process.

Figure 3. In ex vivo HIVE assays the combination of IL-15 and HODHBt induces T cell activation, STAT5 phosphorylation, and perforin/granzyme expression.

Each row depicts results from a different ART-treated individual with HIV: OM5334, OM5011, and OM5267. Columns from left to right display the following flow cytometry data: i) CD4⁺ T cell activation as %CD69⁺ at days 3 and day 9 of HIVE assay, ii) CD8⁺ T cell activation as %CD69⁺ at days 3 and day 9 of HIVE assay, iii) percentage perforin and granzyme double-positive cells within CD8⁺ T cells (Note that in OM5334 granzyme A was measured, while in OM5011 and OM5267 granzyme B was measured), and iv) MFI of phosphorylated STAT5 (pSTAT5; on total PBMCs). Samples were drawn from HIVE assays at 48-hour time points.

Figure 4. Virologic outcomes of HIVE assays.

Each row depicts results from different ART-treated PWH: OM5334, OM5011, and OM5267. The leftmost columns display ultrasensitive p24 ELISA results, measured at day 3. As described in Methods, each condition was plated across multiple wells of a 96-well plate. Here, each data point corresponds to an ELISA measurement from a single well. The horizontal dashed line indicates the limit of detection (LOD). P values were calculated by Kruskal-Wallis test with Dunnett’s test (comparing with DMSO [no NK] condition). The remaining columns display HIV proviral DNA measurements from the intact proviral DNA assay (IPDA), from left to right: intact HIV proviruses, 3′-defective HIV proviruses (3′ deleted or hypermutated), and 5′-deleted HIV proviruses. All proviral DNA measures are presented as mean ± SD (8 replicates) copies of HIV/10⁶ CD4⁺ T cells. Kruskal-Wallis tests were performed, and resulting P values are shown with each graph. Where these were significant, post hoc Dunnett’s tests were performed (compared with DMSO only), and all significant P values are shown. A total of 8 replicates were performed for each. Means with SDs are shown.