SMAC Mimetic Plus Triple-Combination Bispecific HIVxCD3 Retargeting Molecules in SHIV.C.CH505-Infected, Antiretroviral Therapy-Suppressed Rhesus Macaques

The most significant barrier to an HIV-1 cure is the existence of the latently infected viral reservoir that gives rise to rebound viremia upon cessation of ART. Here, we tested a novel combination approach of latency reversal with AZD5582 and clearance with bispecific HIVxCD3 DART molecules in SHIV.C.CH505-infected, ART-suppressed rhesus macaques. We demonstrate that the DART molecules were not capable of clearing infected cells in vivo, attributed to the lack of quantifiable latency reversal in this model with low levels of persistent SHIV DNA prior to intervention as well as DART molecule immunogenicity.

ABSTRACT

The “shock-and-kill” human immunodeficiency virus type 1 (HIV-1) cure strategy involves latency reversal followed by immune-mediated clearance of infected cells. We have previously shown that activation of the noncanonical NF-κB pathway using an inhibitor of apoptosis (IAP), AZD5582, reverses HIV/simian immunodeficiency virus (SIV) latency. Here, we combined AZD5582 with bispecific HIVxCD3 DART molecules to determine the impact of this approach on persistence. Rhesus macaques (RMs) (n = 13) were infected with simian/human immunodeficiency virus SHIV.C.CH505.375H.dCT, and triple antiretroviral therapy (ART) was initiated after 16 weeks. After 42 weeks of ART, 8 RMs received a cocktail of 3 HIVxCD3 DART molecules having human A32, 7B2, or PGT145 anti-HIV-1 envelope (Env) specificities paired with a human anti-CD3 specificity that is rhesus cross-reactive. The remaining 5 ART-suppressed RMs served as controls. For 10 weeks, a DART molecule cocktail was administered weekly (each molecule at 1 mg/kg of body weight), followed 2 days later by AZD5582 (0.1 mg/kg). DART molecule serum concentrations were well above those considered adequate for redirected killing activity against Env-expressing target cells but began to decline after 3 to 6 weekly doses, coincident with the development of antidrug antibodies (ADAs) against each of the DART molecules. The combination of AZD5582 and the DART molecule cocktail did not increase on-ART viremia or cell-associated SHIV RNA in CD4⁺ T cells and did not reduce the viral reservoir size in animals on ART. The lack of latency reversal in the model used in this study may be related to low pre-ART viral loads (median, <10⁵ copies/ml) and low preintervention reservoir sizes (median, <10² SHIV DNA copies/million blood CD4⁺ T cells). Future studies to assess the efficacy of Env-targeting DART molecules or other clearance agents to reduce viral reservoirs after latency reversal may be more suited to models that better minimize immunogenicity and have a greater viral burden.

IMPORTANCE The most significant barrier to an HIV-1 cure is the existence of the latently infected viral reservoir that gives rise to rebound viremia upon cessation of ART. Here, we tested a novel combination approach of latency reversal with AZD5582 and clearance with bispecific HIVxCD3 DART molecules in SHIV.C.CH505-infected, ART-suppressed rhesus macaques. We demonstrate that the DART molecules were not capable of clearing infected cells in vivo, attributed to the lack of quantifiable latency reversal in this model with low levels of persistent SHIV DNA prior to intervention as well as DART molecule immunogenicity.

INTRODUCTION

Since the beginning of the human immunodeficiency virus type 1 (HIV-1) epidemic, 74.9 million people have become infected with HIV, and 32 million people have died from AIDS-related illnesses (1). While antiretroviral therapy (ART) has drastically reduced morbidity and mortality, it does not eradicate HIV that persists in latently infected cells. Latently infected cells with integrated replication-competent provirus constitute the HIV reservoir. To target elimination of these latently infected cells, one strategy aims to induce HIV provirus expression using latency-reversing agents (LRAs) in order to enhance their elimination by the immune system in an approach called “shock and kill” (2). Most clinical trials or preclinical animal studies have shown only limited viral expression following treatment with candidate LRAs (3). However, we recently demonstrated that robust and persistent viral reactivation could be induced in vivo in ART-suppressed simian immunodeficiency virus (SIV)-infected rhesus macaques (RMs) and HIV-infected humanized mice by targeting the noncanonical NF-κB pathway with a mimetic of the second mitochondrial activator of caspases (or SMACm)/inhibitor of apoptosis (IAP) called AZD5582 (7). Treatment of these animals with AZD5582 induced SIV RNA expression in plasma and in resting CD4⁺ T cells from lymphoid tissues as well as increased HIV RNA levels in almost every tissue analyzed in humanized mice (7). A reduction in the viral reservoir size was not observed in these studies that lacked the “kill” component of the curative approach (3, 9–11), suggesting a lack of sufficient clearance of reactivated cells by either virus-induced cytopathic effect or immune effector cells. These results are in line with previous work suggesting that latency reversal alone may be insufficient to reduce the viral reservoir (12, 13).

Two important consequences of latency reversal may need to be addressed to effectively reduce the viral reservoir size: first, neutralizing the virions newly released from reactivated, latently infected cells and, second, clearing the infected cells themselves. Among the proposed “kill” approaches, antibodies (Abs) that bind the HIV envelope protein (Env), and particularly broadly neutralizing monoclonal antibodies (mAbs), are of great interest because of their ability to neutralize free virions in plasma and to potentially destroy reactivated cells by mediating antibody-dependent cell-mediated cytotoxicity (ADCC) (14, 15, 18, 19, 20, 21). Importantly, nonneutralizing Abs that bind to Env expressed on the surface of infected cells are also capable of mediating ADCC (22). A complementary strategy to Env-specific mAbs is represented by bispecific DART molecules, Ab-derived molecules with two unique antigen (Ag) binding sites that enable binding to two different targets simultaneously. For cancer immunotherapy, DART molecules with antitumor antigen and anti-CD3 specificities are designed to redirect cytolytic T cells to tumors (23–25). In the context of an HIV cure strategy, DART molecules with anti-HIV Env and anti-CD3 specificities are designed to redirect cytolytic T cells to HIV-infected cells that express the Env protein. The level of cell surface Env on latently infected cells may potentially be increased by LRAs and thus allow recognition by HIVxCD3 DART molecules, which have been shown to mediate the clearance of acutely HIV-infected CD4⁺ T cells in vitro and LRA-reactivated latently infected CD4⁺ T cells ex vivo (22, 26).

In this study, we sought to evaluate if the viral reservoir could be reduced by a combination of AZD5582 and HIVxCD3 DART molecule treatments in ART-suppressed RMs infected with simian/human immunodeficiency virus SHIV.C.CH505.375H.dCT. The virus SHIV.C.CH505 is derived from SIVmac766 but encodes a clade C transmitted/founder Env and thus can be targeted by HIV-specific Abs (27). Three DART molecules were selected based on their ability to bind HIV-1 CH505-infected cells in vitro and mediate killing of CD4⁺ T cells isolated from acutely SHIV.C.CH505-infected RMs ex vivo (28). A cocktail of DART molecules with A32, 7B2, and PGT145 anti-HIV-1 Env specificities was administered together with AZD5582, and both latency reversal and the impact on the replication-competent SHIV reservoir were assessed.

RESULTS

HIVxCD3 DART molecule binding and redirected killing of infected CD4⁺ T cells in vitro.

The rhesus Fc-bearing HIVxCD3 DART molecules were composed of a rhesus cross-reactive anti-CD3 arm (based on hXR32, a humanized anti-CD3ε mAb) combined with an anti-HIV-1 Env arm based on the A32, 7B2, or PGT145 mAb (Fig. 1A). A32 and 7B2 are nonneutralizing human mAbs that recognize gp120 C1, cluster A, and gp41 cluster I, respectively. PGT145 is a broadly neutralizing human mAb that recognizes gp120 V2, a quaternary epitope at the V2 apex of HIV-1 (29–31).

FIG 1.

HIVxCD3 DART molecule design, binding, and redirected killing activity. (A) Schematic representation of HIVxCD3 DART molecules used in this study. Each molecule has 2 antigen binding arms linked to a rhesus IgG1 Fc domain with mutations to inactivate Fcγ receptor binding but retain neonatal Fc receptor (FcRn) binding to prolong the serum half-life (green). The first arm is an anti-human CD3ε domain, which cross-reacts with the rhesus homolog with comparable affinity (blue). The second arm is an anti-HIV-1 envelope domain, which consisted of A32 (anti-gp120 CD4i C1, cluster A), 7B2 (anti-gp41 cluster I), and PGT145 (anti-gp120 V2 glycan), pictured in red, purple, and yellow, respectively (https://www.hiv.lanl.gov/content/immunology/ab_search.html). (B) Gating strategy for flow cytometry analysis of DART molecule binding to mock-infected cells (top) and HIV CH505-infected cells (bottom). For HIV CH505-infected cells, p24⁺ cells were analyzed for binding by staining with secondary Ab (2ary) after each DART molecule. DART molecules were stained individually in separate wells. FSC, forward scatter; SSC, side scatter. (C) Surface binding of HIVxRSV DART molecules to HIV-1 CH505 IMC-infected cells. To specifically assess binding to cell surface HIV-1 envelope, DART molecules with HIV arms paired with RSV arms instead of CD3 arms were utilized for this analysis. Primary human CD4⁺ T cells were activated with anti-CD3 and anti-CD28 Abs in vitro and then infected for 72 h with HIV-1 CH505 IMC or mock infected. The percentage of infected cells (p24⁺) with bound DART molecules (at 10 μg/ml) was measured by flow cytometry after 24 h. A secondary Ab with only the Fc portion was also used as a negative control (Fc 2ary alone). Error bars represent standard deviations (SD). Results were obtained using two donors. (D) Redirected killing of primary human HIV-1 CH505 IMC-infected CD4⁺ T cells by individual HIVxCD3 DART molecules. Titration curves represent HIVxCD3 DART-mediated specific lysis of primary human CD4⁺ T cells activated with anti-CD3 and anti-CD28 Abs in vitro and then infected for 72 h, with HIV-1 CH505 IMC as targets (T) and autologous CD8⁺ T cells as effectors (E) at E:T ratios of 33:1 and 0:1. Percent specific lysis was measured 24 h after incubation of T+E+DART molecules. Titration curves represent infected cell lysis mediated by PGT145, A32, or 7B2 DART molecules. RSVxCD3 was used as a negative control. Results show average specific lysis mediated by human HIVxCD3 DART molecules using two donors.

To specifically interrogate binding to Env on the surface of CH505 infectious molecular clone (IMC)-infected primary human CD4⁺ T cells, we generated variant DART molecules with anti-HIV arms paired with anti-respiratory syncytial virus (anti-RSV) arms instead of anti-CD3 arms. All 3 HIVxRSV DART molecules exhibited binding to CH505-infected cells but not mock-infected cells, indicating that the different Env epitopes were accessible and recognizable (Fig. 1B and C). All 3 fully functional HIVxCD3 DART molecules redirected CD8⁺ T cells to kill HIV-1 CH505 IMC-infected primary human CD4⁺ T cells at an effector-to-target cell (E:T) ratio of 33:1 with 50% effective concentrations (EC₅₀) of 13.5, 7.6, and 1.5 ng/ml for the PGT145, 7B2, and A32 DART molecules, respectively (Fig. 1D). No redirected T cell killing was mediated by RSVxCD3 control molecules or by HIVxCD3 DART molecules in the absence of CD8⁺ effector cells (E:T ratio = 0:1).

Nonhuman primate study design.

To assess this cocktail of 3 HIVxCD3 DART molecules as part of a novel “shock-and-kill” regimen in vivo, we next combined them with the SMAC mimetic AZD5582, recently shown to induce SIV or HIV latency reversal in ART-suppressed animal models (7). To this end, we selected 13 female RMs negative for MamuB*08 and MamuB*17, major histocompatibility complex (MHC) haplotypes associated with an increased frequency of spontaneous lentiviral control (33–35). All RMs were infected intravenously (i.v.) with SHIV.C.CH505 at a dose of 10 ng of p27 antigen. SHIV.C.CH505 is a chimeric simian/human immunodeficiency virus with a SIVmac766 backbone and the clinically relevant HIV-1 transmitted/founder clade C envelope CH505. A virus expressing HIV-1 Env is needed to evaluate the activity of the HIVxCD3 DART molecule cocktail. All RMs were treated with an ART regimen consisting of tenofovir (TDF) (5.1 mg/kg of body weight), emtricitabine (FTC) (40 mg/kg), and dolutegravir (DTG) (2.5 mg/kg) administered subcutaneously daily from week 16 postinfection (p.i.) to the end of the study. The control group included 5 RMs maintained on ART until the end of the study, at 65 weeks p.i. The experimental group included 8 RMs who additionally received AZD5582 and a cocktail of 3 HIVxCD3 DART molecules at 59 weeks p.i., after sustained plasma viral load measurements of <60 copies/ml for over 9 months. The experimental treatment was administered weekly for 10 consecutive weeks and consisted of a dose of the DART cocktail (1 mg/kg each molecule) followed by a dose of AZD5582 (0.1 mg/kg) 2 days later (Fig. 2A).

FIG 2.

Experimental design. (A) Thirteen Indian rhesus macaques were infected intravenously (i.v.) with SHIV.C.CH505.375H.dCT (10 ng of p27 Ag). Antiretroviral therapy (ART) was initiated at 16 weeks postinfection (p.i.) and maintained for the duration of the study. At week 63 p.i., 8 animals received the experimental treatment consisting of an i.v. infusion of a cocktail of the 3 DART molecules at 1 mg/kg each followed by an i.v. infusion of AZD5582 at 0.1 mg/kg 2 days later. This treatment was given weekly for 10 consecutive weeks. The remaining 5 RMs were maintained on ART only and served as controls. All animals were euthanized at the end of the study for tissue collection. (B) Longitudinal analysis of plasma SHIV RNA before intervention in the experimental (left) and control (right) groups. The shaded gray area represents the period of ART administration.

All animals experienced an exponential increase in plasma viral loads following SHIV.C.CH505 challenge, with a peak of between 10⁶ and 10⁷ SHIV RNA copies/ml of plasma observed at 2 weeks p.i. (Fig. 2B). At 16 weeks p.i. (pre-ART initiation), plasma viral loads ranged from <60 to 10⁴ SHIV RNA copies/ml. Three of 13 RMs demonstrated transient control of viremia before ART initiation, with at least one measurement of <60 copies/ml. These animals were allocated to both the control (1 RM) and experimental (2 RMs) groups. The initiation of ART induced a rapid decline in plasma viral loads to <60 SHIV RNA copies/ml of plasma in all animals after 2 weeks of treatment. All animals maintained a plasma viral load of <60 copies/ml on ART for the 40-week period before experimental interventions.

SHIV RNA levels remained stable in plasma and cells during treatment with AZD5582 plus the DART molecule cocktail.

To assess the latency reversal activity of AZD5582 in ART-suppressed SHIV.C.CH505-infected RMs, the plasma viral load was first monitored three times per week during the AZD5582 plus DART molecule cocktail (AZD5582+DART molecule cocktail) treatment period using a standard assay with a limit of detection of 60 copies/ml of plasma. In this model, and in contrast to what we previously observed in SIV-infected, ART-suppressed RMs (7), no increases in plasma viral loads above 60 copies/ml were observed following AZD5582 treatment in any of the 8 experimental macaques (Fig. 3A). As expected, the control animals also maintained plasma viral loads of <60 copies/ml at equivalent time points.

FIG 3.

Plasma viral loads during AZD5582+DART molecule cocktail administration. (A) Longitudinal assessment of plasma SHIV RNA levels using a standard assay with a limit of detection of 60 copies/ml during treatment with AZD5582 and the DART molecule cocktail (left) and at an equivalent time period for the controls (right). (B) Longitudinal assessment of plasma SHIV RNA levels using an ultrasensitive assay with a limit of detection 5 copies/ml during treatment with AZD5582 and the DART molecule cocktail. The shaded gray area represents the period of ART treatment. Black and green vertical lines represent AZD5582 and DART molecule cocktail infusion time points, respectively. Horizontal dotted lines represent the limits of detection for the standard (upper line) and ultrasensitive (lower line) assays.

Second, the plasma viral load was measured using an ultrasensitive assay with a limit of detection of 5 copies of SHIV RNA/ml of plasma in the experimental group. To this end, we selected plasma samples from 3 time points prior to treatment with AZD5582+DART molecules as baseline values and 10 time points during and after treatment. Two RMs showed single occurrences of plasma viral loads above the limit of detection of the ultrasensitive assay, one occurrence before the experimental treatment and one occurrence after 2 doses of AZD5582+DART molecule cocktail (Fig. 3B). All SHIV RNA values in the remainder of the RMs were below 5 copies/ml plasma.

To further evaluate the impact of AZD5582+DART molecule cocktail treatment on viral transcription in ART-suppressed SHIV.C.CH505-infected RMs, we next quantified cell-associated RNA in CD4⁺ T cells enriched from various tissues. Levels of SHIV RNA in CD4⁺ T cells isolated from peripheral blood, lymph nodes, and bone marrow were not statistically different before and after AZD5582+DART treatment (Fig. 4A, top row). We also cross-sectionally compared SHIV RNA levels in CD4⁺ T cells isolated from the gastrointestinal tract and spleen of the experimental group versus controls at the end of the study. No significant differences were observed between groups (Fig. 4B, top row). Thus, we did not observe latency reversal, as measured by increased levels of viral RNA in CD4⁺ T cells or plasma, following AZD5582 treatment in this model of ART-suppressed SHIV.C.CH505-infected RMs.

FIG 4.

Viral reservoir assessment. (A) Comparison of cell-associated SHIV RNA (top) and SHIV DNA (bottom) levels in total CD4⁺ T cells isolated from peripheral blood, lymph nodes, and bone marrow before (Pre) and after (Post) AZD5582+DART molecule treatment. (B) Cell-associated SHIV RNA (top) and SHIV DNA (bottom) levels in CD4⁺ T cells isolated from the gastrointestinal tract and spleen of experimental RMs after AZD5582+DART molecule treatment or control RMs at equivalent time points. (C) Assessment of frequencies of CD4⁺ T cells carrying replication-competent SHIV by a quantitative viral outgrowth assay in lymph nodes and spleen of experimental RMs after AZD5582+DART molecule treatment or control RMs at equivalent time points. Statistical significance was determined using Wilcoxon matched-pairs signed-rank tests in panel A and Mann-Whitney tests in panels B and C. n.s., not significant.

The SHIV reservoir was not reduced by AZD5582+DART molecule cocktail treatment.

A potential explanation for the observed lack of an increase in SHIV RNA following experimental treatment could be clearance of the cells reexpressing SHIV and/or clearance of the produced virus by the DART molecules. To evaluate this hypothesis, we next investigated whether AZD5582+DART molecule cocktail treatment had an impact on the levels of SHIV DNA in ART-suppressed SHIV.C.CH505-infected RMs. To this end, SHIV cell-associated DNA was measured in CD4⁺ T cells purified from peripheral blood, lymph nodes, and bone marrow before and after AZD5582+DART molecule cocktail treatment in the experimental animals. However, we did not observe significant differences in SHIV DNA levels with this longitudinal sampling (Fig. 4A, bottom row). We also compared SHIV DNA levels in CD4⁺ T cells isolated from the gastrointestinal tract and spleen of the experimental group versus controls at the end of the study. These data indicated no differences between the groups (Fig. 4B, bottom row).

To further assess the impact of AZD5582+DART molecule cocktail treatment on the viral reservoir in ART-suppressed SHIV.C.CH505-infected RMs, we estimated the size of the replication-competent reservoir by a quantitative viral outgrowth assay (QVOA) in the experimental animals after they received 10 doses of AZD5582+DART molecule cocktail and compared it to that of the control animals at an equivalent time point. Virus outgrowth was detected in CD4⁺ T cells from lymph nodes for all RMs and in CD4⁺ T cells from the spleen in all but 2 RMs. Although the size of the replication-competent SHIV reservoir in the lymph nodes was smaller in the experimental group than in the control animals (median of 0.21 infectious units per million CD4⁺ T cells [IUPM] [n = 8] versus 0.46 IUPM [n = 5]), this difference was not statistically significant (Fig. 4C). Furthermore, the sizes of the replication-competent reservoir in the spleen were similar in the AZD5582+DART molecule cocktail and control groups (median of 0.31 IUPM [n = 8] versus 0.36 IUPM [n = 5]) (Fig. 4C). Altogether, these results suggest that the SHIV reservoir was not significantly reduced by AZD5582+DART molecule cocktail treatment.

DART molecule pharmacokinetics and immunogenicity.

We evaluated the pharmacokinetics (PK) and antidrug antibody (ADA) responses of the weekly administered DART molecule cocktail. Peak and trough serum concentrations (collected immediately after and at 1 week postinfusion, respectively) of total HIVxCD3 DART molecules (all 3 combined) are shown in Fig. 5, with individual plots for each experimental animal. Mean peak and trough serum concentrations after the first DART molecule cocktail dose were 1.05 × 10⁵ ng/ml (range, 0.44 × 10⁵ to 1.37 × 10⁵ ng/ml) and 7.32 × 10³ ng/ml (range, 4.23 × 10³ to 10.88 × 10³ ng/ml), respectively. These serum concentrations greatly exceed the EC₅₀ values (1.5 to 13.5 ng/ml) for redirected killing activity in vitro (Fig. 1C), indicating ample DART molecule exposure to mediate activity against reactivated, Env-expressing, latently infected cells, if induced by AZD5582 administration. Noticeable declines in peak serum concentrations became evident after 3 to 6 infusions, and trough concentrations declined to undetectable levels after 3 to 8 infusions, consistent with accelerated DART molecule elimination by RMs in response to repeat DART molecule cocktail dosing. Measurement of ADAs in serum demonstrated that all animals developed ADAs against each of the DART molecules in a time frame that coincided with the declines in total DART molecule serum concentrations. Thus, repeat-dose studies in RMs are limited by the immunogenicity of HIVxCD3 DART molecules, even despite modification of the DART molecules to include rhesus Fc sequences.

FIG 5.

HIVxCD3 DART molecule pharmacokinetics and immunogenicity. Individual representations of total (A32, 7B2, and PGT145) DART molecule serum concentrations (left axis) and antidrug antibody (ADA) responses against each HIVxCD3 DART molecule (right axis) in serum of the 8 RMs of the experimental group are shown. S/N, signal-to-noise ratio. Dashed lines indicate one-half of the lower limit of quantification of DART molecules in rhesus serum. For animal 14-01, the increase in the DART concentration following infusion on day 42 was not detected because the postinfusion serum sample was not collected. For animal 14-19, the last DART infusion (approximately one-half of the total dose) was administered on day 42.

Safety of combined AZD5582 and DART molecule cocktail administration.

Clinical and laboratory parameters were assessed longitudinally over the study time period to determine the safety and tolerability of AZD5582+DART molecule cocktail administration as well as general health during this study. Longitudinal complete blood counts showed normal values, with the exception of transiently decreased hemoglobin in 4 animals (14-01, 4U3, 14-91, and 14-55) in the experimental group and 1 animal (14-77) in the control group at several time points during the period of intensive blood collections and increased white blood cell counts above the upper normal limit at a single time point in 2 experimental animals (14-19 and 6U7) and 1 control animal (X85) during the intervention phase (Fig. 6A). Platelet counts varied widely across both groups of animals. Longitudinal assessment of serum chemistry in 12/13 animals showed values within normal ranges, suggesting that hepatic and renal functions were not adversely impacted by AZD5582+DART molecule cocktail treatment (Fig. 6B). One control animal (1R4) had elevated alanine transaminase (ALT) levels throughout the study, but no clinical concerns were reported by the veterinarians. Body weight, used as an indicator of overall health, showed some fluctuations but stayed within the expected ranges (Fig. 6C). One animal (14-19) had a reaction characterized by tachypnea, tachycardia, and pallor during the 6th DART molecule cocktail infusion. The 6th AZD5582 dose for this animal was withheld; however, during the 7th administration of the DART molecule cocktail, the animal developed tachypnea, tachycardia, and pallor again, and the infusion was halted, with no further doses of AZD5582 or the DART molecule cocktail given. A similar reaction was noted in another RM (6U7) after the 9th DART molecule cocktail infusion but resolved shortly thereafter. Unfortunately, during the 10th DART molecule cocktail infusion, this animal developed urticaria with bradycardia and bradypnea and died. The adverse events in these 2 RMs are consistent with acute infusion reactions secondary to hypersensitivity reactions triggered by the immunogenicity of the DART molecules, which are structured with rhesus Fc but human antibody variable domain protein sequences. ADAs to recombinant human proteins and hypersensitivity reactions are detected relatively often in monkeys (36–38) but are not predictive of immunological reactions in humans (39). As in our prior study (7), AZD5582 appeared safe in ART-suppressed RMs.

FIG 6.

Safety data. Shown are data for longitudinal assessments of complete blood counts (WBC, white blood cell; HGB, hemoglobin) (A), serum chemistry (AST, aspartate transaminase; ALT, alanine transaminase; GGT, gamma-glutamyl transaminase; BUN, blood urea nitrogen) (B), and body weight (C). The shaded gray area represents the period of ART treatment, and horizontal lines indicate the normal range. Black and green vertical lines represent DART molecule cocktail and AZD5582 infusions, respectively.

Correlation between virologic parameters and LRA sensitivity.

To better understand the lack of observed SHIV reactivation and reservoir reduction, we investigated potential virologic parameters that may have contributed to the lack of latency reversal in this model. In a previous study in a group of 12 ART-suppressed SIV-infected RMs, we showed that 8 animals experienced increased plasma viral loads during AZD5582 treatment, while 4 maintained stably suppressed plasma SIV RNA levels in the presence of continued ART (7). In this prior work, the RMs with stably suppressed viral loads during AZD5582 treatment had lower levels of plasma SIV RNA immediately before ART initiation and lower levels of SIV DNA in peripheral CD4⁺ T cells before AZD5582 treatment, suggesting a relationship between reservoir size and potential for latency reversal. We thus compared these virologic parameters in the 8 SHIV.C.CH505-infected RMs treated with AZD5582 in the current study with those in the previously described SIV-infected RMs. As shown in Fig. 7, levels of plasma viral RNA before ART initiation were significantly lower in the SHIV.C.CH505-infected RMs than in both groups of SIV-infected animals (by a median of 1 to 3 logs). Similarly, levels of viral DNA (vDNA) in peripheral blood CD4⁺ T cells before AZD5582 treatment were lower in this group of SHIV.C.CH505-infected RMs than in SIV-infected RMs who experienced increased plasma viral loads (by a median of ∼1 log) and were similar to those of the group of SIV-infected RMs with stably suppressed plasma viral loads during AZD5582 treatment. In line with our previous work, these results suggest a correlation between the latency-reversing activity of AZD5582 and the viral reservoir size before intervention and provide a hypothesis for the lack of latency reversal in the SHIV model.

FIG 7.

Comparison of virologic parameters in RMs infected with SIV or SHIV. Comparative analyses were performed among three groups of RMs: SIV-infected RMs who experienced increased on-ART plasma viral loads (PVL) following AZD5582 treatment, SIV-infected RMs who maintained stably suppressed on-ART plasma viral loads following AZD5582 treatment, and SHIV-infected RMs who maintained stably suppressed on-ART plasma viral loads following AZD5582 treatment. Plasma viral loads prior to ART initiation (A) and SHIV/SIV DNA levels in peripheral CD4⁺ T cells before AZD5582 administration (B) are shown. PB, peripheral blood.

DISCUSSION

HIV persists in millions of people around the world. Despite the reduction of plasma viremia to below the detection limits of clinical assays, ART cannot clear the virus, and thus, lifetime ART adherence is needed to prevent the viral rebound that occurs in the majority of patients after treatment interruption (41–43). As such, the most significant barrier to an HIV cure is the existence of infected cells with integrated replication-competent provirus that give rise to rebound viremia upon cessation of ART (i.e., the latent reservoir). Under ART, this latent reservoir is invisible to the immune system, and it is therefore essential to understand how to best target reservoir cells to promote immune-mediated recognition and clearance. Approaches that focus on the reactivation of virus from latency in reservoir cells followed by immune-based clearance of infected cells, termed shock and kill, are a cornerstone of current research efforts toward an HIV cure. In the current study, we used an SHIV-infected RM model and suppressive ART to evaluate the reservoir-reducing potential of the SMAC mimetic AZD5582 as a latency reversal agent in combination with a cocktail of 3 HIVxCD3 DART molecules with different anti-HIV Env specificities as clearance agents.

Infection of RMs with SHIV.C.CH505 (27, 44) allowed us to utilize HIV-1 Env-targeting agents as the clearance strategy. Many studies have shown the utility of RM infection using SHIV chimeric viruses to evaluate transmission (14, 45, 46). SHIV.C.CH505 has been demonstrated to maintain the properties of primary HIV-1 strains and to mimic viral replication dynamics in a prior study (27). In the current work, 13 female RMs were successfully infected with SHIV.C.CH505, achieving peak plasma viral loads as high as those seen in SIV/HIV-1 infection (10⁷ to 10⁸ copies/ml). However, different from the prior study (27), several of the RMs that we investigated had evidence of plasma SHIV RNA spontaneously decreasing to low levels pre-ART, and three RMs had plasma viral loads of <60 copies/ml for at least one time point prior to ART initiation. This result is notable as animals were selected for MHC alleles that are not typically associated with natural control of SIV infection (33–35). The viral replication curves found here in 13 RMs are similar to those observed in a group of 6 adult macaques also infected with SHIV.C.CH505, in which ART was initiated 12 weeks after infection (47).

Despite the powerful SIV latency reversal activity of AZD5582 in vivo (7), here, we did not find evidence of virus reactivation, as measured by increases in on-ART viremia or increases in levels of SHIV RNA in infected CD4⁺ T cells. These results are in contrast to our previous work in which AZD5582 treatment resulted in increased on-ART plasma viral loads and viral RNA levels in CD4⁺ T cells from tissues in HIV-infected ART-treated humanized mice and SIV-infected macaques with long-term viral suppression on ART (7). In this prior SIV study, latency reversal occurred in the group of 8 RMs that had higher levels of pre-ART viremia and pre-AZD5582 CD4⁺ T cell-associated SIV DNA, suggesting that a larger reservoir is more prone to virus reactivation with this approach. Using these two parameters as predictors of the capacity to reverse latency with AZD5582, we demonstrate that the group of SHIV.C.CH505-infected RMs treated with AZD5582 aligns with the four SIV-infected RMs without virus reactivation in the prior study. With all these data now in context, we propose that the SHIV.C.CH505 model utilized here is likely not one in which latency reversal can be easily demonstrated. We suggest that the utility of SHIV models, in general, for “reservoir/cure” studies, despite their attractiveness for the evaluation of HIV-1 Env-targeted approaches, may be limited without scrupulous attention to pre-ART virologic parameters indicative of spontaneous control of virus replication and restricted reservoir size. A long period of untreated virus replication (e.g., ≥6 months) to allow for sufficient evaluation of these parameters before ART initiation, with censoring of controllers, seems warranted for future studies. Importantly, HIV-infected patients who initiated ART during chronic infection maintain CD4⁺ T cell-associated HIV DNA at levels (49) comparable to those of SIV-infected macaques who demonstrated increased on-ART viremia in response to AZD5582, suggesting that these patients may also experience similar latency reversal if given a therapy that induces noncanonical NF-κB activation.

The ultimate goal of the shock-and-kill strategy is to achieve clearance of newly reactivated infected cells, thereby reducing the size of the persistent viral reservoir. As we did not find evidence of SHIV reactivation with AZD5582+DART molecule cocktail treatment, the efficacy of the HIVxCD3 DART molecules on clearance in the animals was not fully evaluable. Based on the stable levels of total vDNA in CD4⁺ T cells before and after AZD5582+DART molecule cocktail treatment and similar replication-competent reservoirs in the experimental and control groups, as well as ample DART molecule exposure in most of the study, the HIVxCD3 DART molecules were not capable of clearing infected cells in vivo in the absence of latency reversal, which is needed to induce Env expression and allow recognition by DART molecules. Our results also demonstrate that repeat-dose studies of HIVxCD3 DART molecules with human recombinant protein components in SHIV-infected, ART-suppressed RMs are limited by DART molecule immunogenicity. Future studies utilizing infection with SIV or more robustly replicating SHIVs, rather than SHIV.C.CH505, and fully rhesusized DART molecules or other clearance agents may better enable evaluation of strategies to reduce the viral reservoir after effective latency reversal. Our data also suggest that clinical trials using LRAs in combination with clearance agents like HIVxCD3 DART molecules in persons living with HIV-1 may more readily demonstrate a measurable effect in those who initiated ART during chronic infection with a larger persistent reservoir.

MATERIALS AND METHODS

Animals and infection.

Nonhuman primate studies utilized 13 RMs (Macaca mulatta). All animals were born at Yerkes National Primate Research Center (all female and 4 to 5 years of age at the start of the study), and all were negative for Mamu-B*08 and Mamu-B*17 alleles that are known to be associated with natural SIV control. Animals were treated in accordance with guidelines of the Institutional Animal Care and Use Committee (IACUC) of Emory University and Yerkes National Primate Research Center (PROTO201700286). Animal care facilities are accredited by the U.S. Department of Agriculture and the Association for Assessment and Accreditation of Laboratory Animal Care International. Prior to assignment to the study, animals were confirmed negative for SIV, simian T cell lymphotropic virus, simian retrovirus type D, and herpes B virus. All RMs were challenged intravenously (i.v.) with SHIV.CH505.375H.dCT (10 ng p27 Ag in 1 ml). Infection was confirmed by plasma viral load 1 week later. Animals were divided into experimental (n = 8) and control (n = 5) groups. Additionally, data obtained from a group of 12 RMs infected i.v. with 3 × 10³ tissue culture infective doses (TCID₅₀) of SIVmac239 were used for comparative analyses (7).

Antiretroviral therapy.

ART was initiated at 16 weeks (112 days) postinfection. The ART cocktail contained two reverse transcriptase inhibitors, tenofovir (TDF) (20 mg/kg) and emtricitabine (FTC) (40 mg/kg), plus an integrase inhibitor, dolutegravir (DTG) (2.5 mg/kg) (50). This ART cocktail was administered subcutaneously daily at 1 ml/kg of body weight.

Sample collection and processing.

All animals were anesthetized prior to any collection procedure. EDTA-anticoagulated blood samples were collected regularly for a complete blood count, chemical analysis, and immunostaining. During the 10-week intervention phase for the experimental group and a corresponding time period for controls, blood was collected three times weekly to monitor viral loads (weeks 59 to 68 for the experimental group and weeks 53 to 62 for the control group). Plasma was separated from blood by centrifugation within 1 h of phlebotomy and preserved at −80°C for further investigation. Mononuclear cells were isolated from blood and bone marrow by density gradient separation and cryopreserved until use. Bone marrow and one lymph node from each animal were collected before AZD5582+DART molecule cocktail intervention. All other tissue samples, including lymph nodes, spleen, and gut, were collected postmortem. Lymph node biopsy specimens were washed with RPMI 1640 twice, and fat tissues were separated. Cleaned lymph nodes were cut and ground over a 70-μm strainer for cell separation. The obtained cells were counted and cryopreserved. Connective and fat tissues of gut samples were removed, and cleaned gut samples were cut into small pieces. Gut cells were isolated by digestion with collagenase and DNase I for 2 h at 37°C and then passed through a 70-μm cell strainer. The cell suspensions obtained were washed and cryopreserved.

DART molecules and AZD5582 administration.

The experimental group (n = 8) received weekly 30-min i.v. infusions of a cocktail of 3 HIVxCD3 DART molecules at 1 mg/kg each for 10 consecutive weeks starting at 59 weeks postinfection. Each DART molecule was composed of (i) an anti-human CD3ε arm derived from rhesus cross-reactive humanized antibody hXR32, (ii) an anti-HIV-1 Env arm derived from human antibody A32 (gp120 CD4i C1, cluster A), 7B2 (gp41, cluster I), or PGT145 (gp120 V2), and (iii) a rhesus IgG1 Fc domain, which retains neonatal Fc receptor (FcRn) binding for prolongation of serum half-life, but is inactivated by point mutagenesis for the Fc gamma receptor and complement binding. Two days after each DART molecule cocktail dose, all animals in the experimental group also received AZD5582 at 0.1 mg/kg via i.v. infusion over 30 min using an in-line filter. AZD5582 infusions were prepared from AZD5582 (Chemietek) powder and dissolved into a 10% Captisol solution to 0.4 mg/ml.

Serum collection and processing.

Before and after each DART infusion, 0.5 to 1 ml of blood was collected. Serum was separated, and virus was inactivated by mixing 10 parts of serum with 1 part of a mixture of Tween 80 (polyethylene glycol sorbitan monooleate, polysorbate 80) (10%) and TNBP [tri(n-butyl) phosphate] (3%) in phosphate-buffered saline (PBS) (51) and then aliquoted into duplicates for subsequent pharmacokinetic and antidrug antibody (ADA) analyses.

PK assay.

The total concentration of the combined HIVxCD3 DART molecules in serum was measured by a sandwich immunoassay using an immobilized goat polyclonal antibody that contains anti-hXR32 antibody (recognizes the CD3 arm of DART molecules) for capture and biotinylated anti-EK antibody (recognizes the linker region integral to DART molecules) with streptavidin-horseradish peroxidase (SA-HRP) for detection. The lower limit of quantification was 6.1 ng/ml.

ADA assay.

ADAs to individual HIVxCD3 DART molecules were measured by affinity capture elution (ACE). Serum samples were acidified and added to enzyme-linked immunosorbent assay (ELISA) plates with immobilized DART molecules to capture ADA. Bound ADA was eluted, immobilized to new plates, and then detected with a biotin-labeled DART molecule and a streptavidin SulfoTag using an MSD Quickplex SQ 120 plate reader.

Plasma viral load and ultrasensitive plasma viral load assays.

Standard SHIV.C.CH505 plasma viral load quantification was performed regularly throughout the study and three times per week during the AZD5582+DART molecule treatment period at the Translational Virology Core Laboratory of the Emory Center for AIDS Research using a standard quantitative PCR (qPCR) assay (limit of detection of 60 copies per ml of plasma) as previously described (52). Ultrasensitive SHIV.C.CH505 plasma viral load quantification (limit of detection of 5 copies per ml of plasma, as performed for this study) was performed for 3 time points before AZD5582+DART molecule treatment and 10 time points during AZD5582+DART molecule treatment as previously described (8, 53).

CD4⁺ T cell isolation for cell-associated SHIV DNA/RNA quantification.

CD4⁺ T cells were enriched from cryopreserved peripheral blood and bone marrow mononuclear as well as cryopreserved lymph node and spleen cell suspensions using a MACS CD4⁺ T cell nonhuman primate isolation kit (Miltenyi Biotech) for all samples. Enriched cells were then aliquoted into 1 million or 5 million CD4⁺ T cells and lysed in 350 μl Buffer RLT Plus RNeasy Plus lysing buffer (Qiagen) with 1 mM 2-mercaptoethanol (Sigma-Aldrich Chemistry).

Cell-associated SHIV DNA/RNA quantification.

Cell-associated SHIV RNA and DNA levels were measured simultaneously in total CD4⁺ T cells isolated from peripheral blood mononuclear cells (PBMCs), bone marrow mononuclear cells, lymph nodes, and gut cell suspensions (1,000,000 to 5,000,000 cells) lysed in Buffer RLT Plus (Qiagen) plus 2-mercaptoethanol. Both DNA and RNA were extracted using the Allprep DNA/RNA minikit (Qiagen).

Quantification of SHIV gag DNA was performed on the extracted DNA by quantitative PCR using the 5′ nuclease (TaqMan) assay with an ABI7500 system (PerkinElmer Life Sciences). The sequence of the forward primer for SHIV gag was 5′-GCAGAGGAGGAAATTACCCAGTAC-3′, the reverse primer sequence was 5′-CAATTTTACCCAGGCATTTAATGTT-3′, and the probe sequence was 5′-FAM (6-carboxyfluorescein)-TGTCCACCTGCCATTAAGCCCGA-TAMRA (6-carboxytetramethylrhodamine)-3′. For cell number quantification, quantitative PCR was performed simultaneously for monkey albumin gene copy numbers. The sequence of the forward primer for albumin was 5′-TGCATGAGAAAACGCCAGTAA-3′, the reverse primer sequence was 5′-ATGGTCGCCTGTTCACCAA-3′, and the probe sequence was 5′-AGAAAGTCACCAAATGCTGCACGGAATC-3′. RNA was reverse transcribed using a high-capacity cDNA reverse transcription (RT) kit (Thermo Scientific) and random hexamers. SHIV gag and the rhesus macaque CD4 gene were quantified by qPCR of the resultant cDNA using TaqMan universal master mix II (Thermo Scientific). The CD4 primers and probe were Rh-CD4-F (5′-ACATCGTGGTGCTAGCTTTCCAGA-3′), Rh-CD4-R (5′-AAGTGTAAAGGCGAGTGGGAAGGA-3′), and Rh-CD4-probe (5′-AGGCCTCCAGCACAGTCTATAAGAAAGAGG-3′).

SHIV quantitative viral outgrowth assay.

Replication-competent SHIV reservoirs were measured by the Viral Reservoir Core Laboratory of the Emory Center for AIDS Research. Latently infected cells were quantified using a limiting dilution culture assay in which CD4⁺ T cells enriched from lymph node or spleen cells using magnetic beads and column purification (Miltenyi Biotec) were cocultured with CEMx174 cells in 5-fold serial dilutions ranging from 5 × 10⁶ cells per well to 4 × 10⁵ cells per well. The cells were cultured in RPMI 1640 containing 10% fetal bovine serum (FBS) and 100 U ml⁻¹ interleukin-2 (IL-2) (Sigma). The ratio of target cells added was 4:1 for the two highest dilutions. A constant number of 1 × 10⁶ CEMx174 cells was added to all other wells. The cultures were split every 7 days, and fresh medium was added. After 21 days, the growth of virus was detected by RT-qPCR. SIV RNA was isolated from 400 μl of the culture supernatant using the Zymo viral RNA isolation kit (Zymo Research). DNase treatment was performed using an RQ1 RNase-free DNase kit (Promega). One-step RT-qPCR targeting SIV gag was performed using an Applied Biosystems 7500 real-time PCR system and TaqMan fast virus 1-step master mix (Thermo Scientific) for RT-qPCR with the following primers and probe: SIVgagFwd (5′-GCAGAGGAGGAAATTACCCAGTAC-3′), SIVgagRev (5′-CAATTTTACCCAGGCATTTAATGTT-3′), and SIVgag probe (5′-FAM-TGTCCACCTGCCATTAAGCCCGA-3IBFQ-3′). The frequencies of infected cells were determined by the maximum likelihood method and were expressed as infectious units per million CD4⁺ T cells.

Infectious molecular clone.

The HIV-1 infectious molecular clone (IMC) for subtype C transmitted/founder isolate CH505 (GenBank accession number KC247577) was generated with the backbone derived from the NL4-3 isolate as previously described (54, 55). CH505 IMC expressed the Renilla luciferase reporter gene and preserved all 9 viral open reading frames. The Renilla luciferase reporter gene was expressed under the control of the HIV-1 Tat gene. Upon HIV-1 infection of CD4⁺ T cells, the expression of Tat during HIV-1 replication will induce luciferase expression, which allows the quantitation of infected cells by measuring relative luminescence units (RLU).

Infection of primary CD4⁺ T cells with HIV-1 CH505 IMC.

Cryopreserved PBMCs, obtained from healthy individuals by leukapheresis procedures (56, 57), were thawed and stimulated in R20 medium (RPMI 1640 medium [Invitrogen] with 20% fetal bovine serum [Gemini Bioproducts], 2 mM l-glutamine [Invitrogen], 50 U/ml penicillin [Invitrogen], and 50 μg/ml gentamicin [Invitrogen]) supplemented with IL-2 (30 U/ml) (Proleukin), anti-CD3 (25 ng/ml clone OKT-3; Invitrogen), and anti-CD28 (25 ng/ml; BD Biosciences) antibodies for 72 h at 37°C in 5% CO₂. Two donors were used in separate experiments to test human HIVxCD3 DART molecules. Primary CD4⁺ T cells were isolated by CD8 depletion (CD8 Microbead; Miltenyi Biotech). A total of 1.5 × 10⁶ cells were infected using 1 ml of the virus supernatant by spinoculation (1,125 × g) for 2 h at 20°C. After spinoculation, 2 ml of R20 medium supplemented with IL-2 was added to each infection mixture, and infection mixtures were left for 72 h. Cells were plated in a 12-well plate at a concentration of 0.5 × 10⁶ cells/ml in R20 medium plus 30 U/ml of IL-2.

Flow cytometric analysis of HIVxRSV DART binding to HIV-1-infected CD4⁺ cells.

To eliminate CD3 binding, these studies were conducted with HIVxRSV DART molecules in which anti-HIV arms were paired with anti-RSV arms instead of anti-CD3 arms. HIV-1-infected or mock-infected primary CD4⁺ T cells were obtained as described above. Cells incubated in the absence of virus (mock infected) were used as a negative control. Following infection, cells were washed in PBS, dispensed into 96-well V-bottom plates at 2 × 10⁵ cells/well, and incubated with 10 μg/ml HIVxRSV DART molecules for 2 h at 37°C. Subsequently, cells were washed twice with 250 μl/well of wash buffer (WB) (1% FBS–PBS), stained with vital dye (Live/Dead fixable aqua dead-cell stain; Invitrogen), and washed with WB. Cells were then resuspended in 100 μl of 1 μg/ml biotin-conjugated mouse anti-EK antibody (recognizes the E/K heterodimerization region of DART proteins; MacroGenics), mixed with 1:500-diluted streptavidin-phycoerythrin (PE) (BD Biosciences), and incubated in the dark for 45 min at room temperature. After washing, cells were resuspended in 100 μl/well Cytofix/Cytoperm (BD Biosciences), incubated in the dark for 20 min at 4°C, washed in 1× Cytoperm wash solution (BD Biosciences), stained with anti-p24 antibody (clone KC57-RD1; Beckman Coulter) at a final dilution of 1:100, and incubated in the dark for 25 min at 4°C. Cells were washed three times with Cytoperm wash solution and resuspended in 125 μl PBS–1% paraformaldehyde. The samples were acquired within 24 h using a BD Fortessa cytometer. A minimum of 50,000 total events was acquired for each analysis. Gates were set to exclude doublets and dead cells. The appropriate compensation beads were used to compensate the spillover signal. Data analysis was performed using FlowJo 9.6.6 software (TreeStar). Assays were repeated twice, and the averages of the results are shown with standard deviations (SD).

Redirected T cell cytotoxicity assay against HIV-1 CH505 IMC-infected CD4⁺ T cells.

Primary CD4⁺ T cells infected with the HIV-1 IMC as described above were used as target cells using a previously described assay (22). Briefly, resting CD8⁺ effector T cells were isolated by negative selection from autologous PBMCs using a CD8⁺ T cell isolation kit (Miltenyi Biosciences) at 33:1 and 0:1 E:T ratios in the absence or presence of HIVxCD3 DART molecules for 24 h at concentrations ranging from 1,000 to 0.0001 ng/ml. Uninfected and infected target cells alone were included as additional controls. Each condition was tested in duplicate. After incubation, the ViviRen live-cell substrate (Promega) was added, and RLU were measured on a luminometer; the percent specific lysis (%SL) of target cells was calculated from luminescence counts (RLU) as cytotoxicity (%) = 100 × (1 − [RLU of sample/RLU of control]), as described previously (58). Data were fit to a sigmoidal dose-response function to obtain the 50% effective concentration (EC₅₀).

Statistical analyses.

Statistical analyses were performed using GraphPad Prism software (version 8). Analyses across groups were performed using Mann-Whitney U tests. Analyses within groups were performed using Wilcoxon matched-pair rank sum tests. For all statistical analyses, a P value of <0.05 was considered significant.