To identify a shared Ab signature among the five vaccinated rapid controllers from our previous vaccine trial, we assessed the vaccine-induced Ab responses developed in each animal at the time of the first challenge by systems serology (12). We had previously determined SIVmac239 neutralization titers and ADCC activity in these animals (6). Here, we analyzed Ab titers, FcγR affinity, and assessed three Fc-mediated effector functions: ADCP, NK cell degranulation, and ADCD (Figure 1). Because all Fc effector function assays were performed using human cells, rather than rhesus, we first confirmed that these circulating Abs bound human FcγRIIa, FcγRIIb, and FcγRIIIa (Supplementary Figure 1). In the group of eight Env-vaccinated Mamu-B^*17+ animals, we compared RMs that showed stringent control (n = 5) to those that did not (n = 3). Rapid controllers had slightly higher total IgG and IgG1 serum concentrations of Abs targeting SIV gp140, although this difference was not statistically significant. These Env-specific Abs also exhibited a higher affinity for rhesus FcγRIIa and FcγRIIIa relative to non-controllers (Figure 1A), which was also not statistically significant. Since Barouch et al. reported that protected Ad26-vaccinated RMs had developed highly polyfunctional Abs (3), we evaluated the capacity of our vaccine-induced Abs to mediate ADCP, ADCD and NK cell activation (Figure 1B). While we detected enhanced ADCD and NK cell degranulation activity (using staining for CD107a) mediated by the gp140-binding Abs present in the serum of the rapid controllers at day of challenge, these differences were not statistically significant.

Next, we explored the hypothesis that the stringent viremic control observed in our previous vaccine trial was mediated by vaccine-induced Abs. We evaluated the ability of these Abs to facilitate control of viral replication in vaccinated animals and whether depletion of these Abs in the stringent controller Mamu-B^*17+ RMs would result in viral rebound. While it is possible to isolate Ig from serum or plasma by collecting blood draws, the monthly blood available is limited by the weight of each animal. Thus, to overcome this restriction and isolate sufficient Ig to eventually transfer it to another set of vaccinated animals, we subjected r05007 and r09062, the two rapid controllers with the highest Env-specific Ab titers, to at least four rounds of therapeutic plasma exchange with a secondary plasma device using the Spectra Optia system (Terumo BCT) (13). This method has been previously used to remove circulating immunocomplexes and pathological auto-Abs in circulation (14–18). In contrast to plasmapheresis or standard therapeutic plasma exchange, where plasma is removed and replaced with fluids containing albumin and coagulation factors, we employed a continuous-flow centrifugation procedure, also known as immunoadsorption, to selectively remove Ig from plasma before reinfusing the Ig-depleted plasma back into the same animal. Immunoadsorption minimizes the side effects usually associated with plasmapheresis and allowed us to run one liter of plasma, corresponding to 2.5 times the animal's total blood volume, through four protein A or protein G columns. Using this technique, we isolated 15 g of total Ig from r05007 and r09062 over 3 months. After each round of immunoadsorption, IgG levels in both RMs were reduced by 53–76% compared to baseline (defined as the serum IgG concentration before each procedure) (Figures 2A,B). Additionally, we monitored the viral loads of these animals. At the time of the first procedure, both RMs were suppressing viral replication to <15 vRNA copies/ml (Figures 2A,B). Seven days after the first immunoadsorption, r05007 had a viral load of 650 vRNA copies/ml, and on day 14, viremia peaked at 940 vRNA copies/ml (Figure 2A). Viral loads in r05007 fluctuated between 15 and 940 vRNA copies/ml during all five rounds. r09062 maintained undetectable viral loads, <15 vRNA copies/ml, throughout all four procedures (Figure 2B). Thus, the temporary reduction of circulating Abs caused by immunoadsorption led to a transient rebound in SIV replication.

FcRn has been described as the main factor responsible for prolonging IgG half-life by protecting it from the lysosomal degradative pathway (19–22). IgG recycling is possible because the FcRn binds to the Fc portion of IgG with high affinity at low pH, but not at neutral pH (23, 24). The pH dependence of the FcRn-IgG interaction facilitates the translocation of IgG from acidic endosomes to the cell surface, then its subsequent release into the near-neutral pH of circulation. Hence, it is not surprising that in vivo FcRn dysfunction has been linked to low plasma IgG concentrations (25). Recently, Smith et al. used Rozanolixizumab, an anti-FcRn monoclonal Ab, to safely and efficiently block FcRn and deplete circulating IgG in cynomolgus macaques (26). With this in mind, we reasoned that a prolonged reduction in IgG levels would result in loss of viremic control if the non-neutralizing Abs circulating in the Mamu-B^*17+ rapid controllers had a role in suppressing SIVmac239 replication. We administered a simianized Rozanolixizumab chimera (anti-FcRn rhIgG1 [RozR1LALA]) to the two rapid controllers. Monkey r05007 received an intravenous dose of 30 mg/kg followed by a subcutaneous daily dose of 5 mg/kg for 1 week, while r09062 received a single intravenous dose of 300 mg/kg. We then quantified circulating IgG and SIVmac239 viral loads. r05007 received the anti-FcRn Ab infusion two days after the last round of immunoadsorption. At the time of the first anti-FcRn Ab administration, the IgG concentration in this animal was already reduced by 45.8% relative to baseline, and viremia had increased from 500 to 790 vRNA copies/ml. Three days after commencing Ab infusion, viral loads peaked at 1,700 vRNA copies/ml. Maximal IgG depletion, an 80.3% decrease relative to baseline, was observed 10 days following the first infusion (or three days after the final subcutaneous dose) (Figure 2A). Unlike r05007, r09062 controlled viral replication throughout all four rounds of immunoadsorption. Although these two Mamu-B^*17+ animals had been infected for over five years (Supplementary Figure 2), r09062 had maintained viral loads below detection limits for over 4.5 years (Supplementary Figure 2B). Two weeks after the last immunoadsorption, r09062 received a single infusion of anti-FcRn Ab and its IgG levels started to decrease. One week later, IgG levels were reduced 73.9% relative to baseline (Figure 2B). Viral loads remained <15 vRNA copies/ml for 27 days after the anti-FcRn Ab administration. We then started to observe plasma viral “blips” ranging from 15 to 25 vRNA copies/ml over the following 100 days. It should be noted that both animals developed moderate titers of anti-drug Abs after treatment with anti-FcRn Ab, independent of the infusion route and dose (Supplementary Figure 3).

To further investigate the virologic control observed in the Mamu-B^*17+ RMs of our previous vaccine trial, six RMs expressing Mamu-A^*01+ and/or Mamu-A^*02+ MHC-I alleles (Table 1) were vaccinated with gag, nef, tat, rev, vif, vpr, and vpx. SIV genes were delivered using a recombinant vesicular stomatitis virus (rVSV) prime followed by a recombinant adenovirus type 5 (rAd5) boost. Both vaccine vectors utilized in this study included gag, tat and nef, or tat fused with other genes (i.e., vif, vpr, vpx, and rev) but did not include env. As a reference, our Mamu-B^*17+ cohort was vaccinated with env, vif, nef, tat, and rev, which were delivered using a recombinant yellow fever virus 17D prime followed by three boosts with recombinant DNA plasmids delivered by electroporation (6). These animals were then administered rAd5, rVSV and rhesus monkey rhadinovirus (rRRV) constructs, and were challenged almost 4 months after the last rRRV boost. At the time of the first SIV challenge, we detected SIV-specific T cell responses by intracellular cytokine staining (ICS) in the peripheral blood mononuclear cells (PBMCs). These animals had developed vaccine-induced CD8⁺ T cells mainly directed against vif and nef, and CD4⁺ T cells against env. Conversely, we were unable to detect SIV-specific T cell responses, by ICS, in the PBMCs of our Group 1 vaccinees at the time of the first intrarectal SIV challenge (data not shown), which was 2 years after the rAd5 boost. One day before the first SIV challenge, all vaccinated RMs (Group 1) received 200 mg/kg of the purified Ig derived from the two Mamu-B^*17+ vaccinated animals. A large proportion of the infused Ig targeted SIV Env (Figure 3). At the time of SIV challenge, detectable levels of Env-binding Abs were present in all recipient RMs. Ab levels were comparable among all Group 1 animals and ranged from 221 to 321 μg/ml of gp140-specific IgG in circulation (Table 2). No neutralization activity against SIVmac239 was detected in the purified total Ig (data not shown). We evaluated the contribution of these Abs to protection against acquisition and suppression of viremia post-infection by subjecting all animals to six repeated intrarectal challenges with a marginal infectious dose of SIVmac239, as previously described (27), starting one day following the Ig passive transfer. Six additional unvaccinated, non-Ig-infused RMs expressing Mamu-A^*01+ and/or Mamu-A^*02+ MHC-I alleles (Group 2) served as controls for this challenge experiment (Table 1). In Group 1, Env-specific Ab titers decreased by 35% during the first week after infusion and continued to decline until the sixth challenge. Eleven weeks later, we detected 2.25 μg/ml of gp140-specific Abs in r15010, the only RM from Group 1 that remained uninfected after six SIVmac239 challenges (Table 2). Of note, r15010 was the only RM in Group 1 or 2 that also expressed the elite control-associated Mamu-B^*17+ allele (Table 1). We did not observe any correlation between the level of Env-binding Abs in the sera of the Group 1 vaccinees and the rate of SIVmac239 acquisition. Although vaccinees and control animals acquired SIV at similar rates (Figure 4A), Group 1 exhibited significantly lower peak (P = 0.0079) and chronic phase viral loads (P = 0.0317) (Figures 4B–D). The difference in geometric mean viremia between both groups was also statistically significant (P = 0.0003) (Figure 4E). Collectively, these data suggest that non-neutralizing Abs may contribute to suppression of SIVmac239 replication in vivo.

The 14 Indian RMs (Macaca mulatta) used in this study were housed at the Wisconsin National Primate Research Center (WNPRC). All animals were cared for in accordance with Weatherall report guidelines and the recommendations included in the National Research Council Guide for the Care and Use of Laboratory Animals. All procedures were approved by the University of Wisconsin Graduate School Animal Care and Use Committee (animal welfare assurance number 16-00239 [A3368-01]; protocol number G005248). RMs were separated as follows: Group 1 (Ig-infused vaccinated RMs, n = 6), Group 2 (controls, n = 6), and Group 3 (Mamu-B^*17 rapid controllers, n = 2). Purified Ig was prepared in saline bags and administered intravenously into each animal. The rhesus IgG1 chimeric version of Rozanolixizumab, Anti-FcRn Ab [RozR1LALA], was engineered and produced by the nonhuman primate reagent resource (NIH Nonhuman Primate Reagent Resource Cat# PR-00-1, RRID: AB_2888630). Anti-FcRn Ab, which was infused in Group 3 animals, was initially administered intravenously, followed by a subcutaneous daily dose for a week in one of the Group 3 RMs. We collected plasma and serum at the indicated time points to measure viremia, total Ig and Env-specific Ab levels.

To determine the relative concentrations of total IgG, IgG1, IgG2, IgG3, IgG4, and IgM in the sera of the Mamu-B^*17+ RMs at the time of the first SIV intrarectal challenge, we first coupled SIVmac239 gp140 to magnetic beads (Luminex Corporation) using NHS-ester linkages with Sulfo-NHS and EDC (Thermo Fisher). Next, we incubated SIVmac239 gp140-conjugated magnetic beads with diluted serum overnight at 4°C on a shaker. The following day, plates were spun down and washed before adding these reagents to the appropriate wells: mouse anti-monkey IgG PE (SouthernBiotech, #4700-09), to determine total IgG, mouse anti-rhesus Ig (rhIg) G1 (NHPRR, clone 7H11), rhIgG2 (NHPRR, clone 3C10), rhIgG3 (NHPRR, clone 2G11), rhIgG4 (NHPRR, clone 7A8), and rhIgM (Life Diagnostics, #2C11-1-5), to define the relative concentration of each isotype, or PE-conjugated FcγRIIa, FcγRIIb, and FcγRIIIa (R&D Systems, #1330-CD-050, #1875-CD-050, #4325-FC-050, respectively). The plates were incubated for 1 h at room temperature on a shaker. Next, we added goat anti-mouse IgG Fc PE (Invitrogen, #31861) to the corresponding “isotype” wells and incubated the plates for 1 h at room temperature on a shaker. After washing all plates, fluorescence was quantified using an IQue 3 flow cytometer (Intellicyt) and analysis was performed using IntelliCyt ForeCyt (v8.1).

A THP-1 phagocytosis assay was performed using SIVmac239 gp140-coated yellow/green fluorescent-neutravidin beads (Molecular Probes Inc, #F8776), as previously described (51). First, the fluorescent beads were incubated with diluted plasma for 2 h at 37°C. Following this incubation, THP-1 cells were added to the plates and incubated at 37°C overnight to allow phagocytosis. The next day, cells were fixed and acquired using a LSRII flow cytometer (BD Biosciences). Pooled plasma samples were first titrated and then individually tested at a final dilution of 1:500. We calculated the phagocytic scores as the geometric mean fluorescent intensity (MFI) of the FITC-fluorescent beads multiplied by the bead uptake percentage.

ADCD was assessed by monitoring C3b complement deposition using SIVmac239 gp140-coated red fluorescent neutravidin beads (Molecular Probes Inc, #F8775). Beads were first incubated with diluted plasma for 2 h at 37°C. After washing the bead/plasma mixture, we added guinea pig complement (Cedarlane Laboratories, #CL4051), diluted in gelatin veronal buffer with calcium and magnesium (Boston Bioproducts), and incubated the plates for 20 min at 37°C. Following this incubation, all wells were washed with cold 15 mM EDTA, then incubated with FITC-conjugated goat anti-guinea pig C3 Ab (MP Biomedicals, #855385) for 15 min at room temperature. Last, the plates were washed, and beads were resuspended in PBS. Samples were acquired using a LSRII flow cytometer (BD Biosciences). Percentage of complement deposition was determined by the geometric MFI multiplied by the percentage of beads positive for C3b deposition.

Plates were coated with SIVmac239 gp140 and incubated overnight at 4°C. All coated wells were washed and blocked with 5% BSA in PBS for 1 h at 37°C. Plates were then washed and diluted plasma samples were added into the corresponding wells for 1 h at 37°C. Simultaneously, CD107a PE Cy5 (BD Biosciences, #555802), brefeldin A (Sigma, #B7651), and GolgiStop (BD Biosciences, #554724) were incubated with human NK cell culture for 2 h. After both incubations were completed, we added the stimulated NK cells to the plates (50,000 cells per well), and incubated them for 5 h at 37°C. Next, NK cells were transferred to V-bottom plates containing CD56 PE Cy7 (BD Biosciences, # 557747), CD16 APC Cy7 (BD Biosciences, #55775), and CD3 Alexa Fluor 700 (BD Biosciences, #557943) and incubated for 15 min at room temperature. Plates were then washed and resuspended in PBS. Samples were acquired using a LSRII (BD Biosciences). Human NK cells were isolated using a RossetteSep NK cell enrichment kit (StemCell Technologies, #15065). NK cells were defined as CD3⁻/CD16⁺ and/or CD56⁺. Ab-dependent NK degranulation was calculated as the percentage of NK cells positive for CD107a.

The characterization and in vivo half-life of the purified Ig isolated from the Mamu-B^*17+ RMs were determined by ELISA. ELISA plates were coated with either purified mouse anti-human Ig (BD Biosciences, #555784) or purified SIVmac239 gp140 (Immune Technologies, #IT-001-140p) at 1 μg/ml and incubated overnight at 4°C. The next day, all plates were washed with 1 × PBS-Tween20 and blocked with 5% powdered milk in PBS at 37°C for 1 h. Following this incubation, the plates were washed and 100 μl of serially diluted serum samples or purified Mamu-B^*17-derived total Ig was added to the corresponding wells for 1 h at 37°C. After washing the plates, 100 μl of diluted goat anti-human IgG-HRP Ab (SouthernBiotech, #2045-05) was added to all wells and incubated for 1 h at 37°C. The plates were washed before being developed with 3,3′,5,5′-Tetramethylbenzidine (EMD Millipore, #613544-100ML). Shortly after, the reaction was stopped with TMB stop solution (SouthernBiotech, #0412-01) and all plates were read at 450 nm on an Epoch microplate reader (Biotek).

We assessed the SIVmac239 neutralization activity of the purified Ig and sera of the Ig-infused RMs using a previously described assay (52). Replication-competent SIVmac239 was produced by transfecting HEK293T cells with the plasmid using JetPrime (Polyplus transfection, #114-01). Seventy-two hours later, supernatant was harvested and stored at −80°C. Purified Ig and RM sera were incubated with SIVmac239 for 1 h at 37°C before being transferred onto TZM-bl cells (NIH AIDS Reagent program, #8129). After 48 h, neutralization activity was calculated based on the luminescence measurements of each duplicate well using the Bright-Glo luciferase assay (Promega, #E2620).

The Group 1 RMs were vaccinated with genes encoding SIVmac239 gag, nef, tat, rev, vif, vpr, and vpx using a rVSV prime followed by a rAd5 boost. The rVSV prime consisted of three plasmids containing the full-length Gag polyprotein (VSV-Gag₁-G_nj6), Nef protein (VSV-Nef₁-G_nj6), or Tat-Rev fusion protein (VSV-Tat-Rev₁-G_nj6). Profectus Bioscience manufactured these vectors. Each Group 1 animal received an intramuscular injection containing 1 × 10⁸ PFU of each VSV vector. Six weeks later, the Group 1 RMs were infused with three rAd5 vectors produced by ViraQuest. These vectors encoded gag (rAd5-Gag), nef (rAd5-Nef), or a fusion of vif-vpx-vpr-tat-rev genes. Each animal received an intramuscular injection containing 1x10¹¹ PFU of each Ad5 vector.

PBMCs from Group 1 RMs at the time of the first SIV challenge were stimulated with Gag, Nef and Tat peptides in R10 (RPMI, 10% FBS, 1 × Pen/Strep/Amphotericin B) media containing anti-CD28 and CD49d costimulatory Abs (BD Biosciences), as well as a PE-conjugated anti-CD107a Ab (BioLegend), for 1 h at 37°C in a 5% CO₂ incubator. Following this incubation, brefeldin A (Sigma, #B7651), and GolgiStop (BD Biosciences, #554724) were added to inhibit protein transport. All tubes were incubated for eight additional hours at 37°C. The antigenic stimuli used for ICS consisted of pools of overlapping 15-mer peptides covering the entire open reading frames (ORFs) of SIVmac239 Gag, Tat and Nef at a final concentration of 1 μM. The following day, PBMCs were washed prior to adding the surface staining Ab cocktail. This cocktail included ARD Aqua (live/dead viability dye) and Abs against: CD4, CD8, CD14, CD16, CD20. Next, the cells were fixed and permeabilized using the Cytofix/cytoperm fixation/permeabilization solution kit (BD Biosciences, #554714). After permeabilization, PBMCs were incubated with Abs against CD3, IFN-γ, TNF-α, and CD69 for 1 h at room temperature. Last, the cells were washed and acquired using a FACSAria II (BD Biosciences). We analyzed the data using FlowJo 10.6 by gating on live CD14⁻/CD16⁻/CD20⁻/CD3⁺ lymphocytes and on cells expressing either CD4 or CD8, but not both markers. Next, we gated cells positive for IFN-γ, TNF-α, or CD107a to determine if they were co-expressed with CD69, a marker of recent lymphocyte activation. We utilized the Boolean gate to analyze all possible combinations. Leukocyte activation cocktail-stimulated PBMCs were stained and used as a positive control.

The challenge stock used in this study was produced by the Virology Services Unit of the WNPRC using SIVmac239 hemi-genome plasmids obtained from the NIH AIDS Research and Reference Reagents Program. Plasmids were transfected into 293T cells. Supernatant containing SIVmac239 virions was harvested and propagated on mitogen-activated rhesus PBMCs. Several healthy SIV negative rhesus macaques were used as donors. The titer of the stock utilized was 90,000 50% tissue culture infective doses (TCID₅₀) per ml. All Mamu-A^*01+ and Mamu-A^*02+ animals were intrarectally challenged with 200 TCID₅₀ (4.8 × 10⁵ vRNA copies) of SIVmac239, as previously described (27). Each challenge occurred every 2 weeks. Plasma samples were taken at 7 and 10 days after exposure to determine if an animal had become infected. Once an animal was SIV positive, it was no longer challenged.

SIV RNA viral load measurements were based on a previously published protocol (53). Total RNA was extracted from 0.5 ml of EDTA-anticoagulated Indian RM plasma using QIAgen DSP virus/pathogen Midi kits on a QIASymphony SP instrument. A two-step RT-PCR reaction using random hexamers was completed for each of the six replicates per sample, followed by 45 PCR cycles. The limit of detection in 0.5 ml of plasma was 15 vRNA copies/ml. The primers and probe used for this procedure were: forward primer, SGAG21, 5′ GTCTGCGTCAT(dP)TGGTGCATTC 3′; reverse primer SGAG22, 5′ CACTAG(dK)TGTCTCTGCACTAT(dP)TGTTTTG 3′; probe, PSGAG23, 5′ FAM-CTTC(dP)TCAGT(dK)TGTTTCACTTTCTCTTCTGCG-BHQ1 3′.

Differences between the Mamu-B^*17+ rapid controllers and non-controllers at day of challenge, as well as viral loads of the Group 1 RMs, were determined using Mann-Whitney U test. Kaplan-Meier survival analysis was used to determine if the SIVmac239 acquisition rate differed between both groups. A 0.05 significance threshold was utilized for each test.