EARLY ANTIBODY THERAPY CAN INDUCE LONG LASTING IMMUNITY TO SHIV

Following an initial and massive burst of HIV replication¹⁵, viremia is incompletely controlled by the emergence of a virus-specific CD8⁺ T cell response, leading to a chronic phase during which the infection is never cleared^16,17. Virions that integrate their DNA into the host cell genome become a source of virus production, which persists despite anti-retroviral treatment over extended periods of time and gives rise to a recrudescent progressive infection when treatment is stopped¹⁸. This virus reservoir persists for the lifetime of the infected individual¹⁹.

Nonetheless, initiation of cART within 6 months of HIV-1 infection may limit damage to the immune system as well as the size of the reservoir, as measured by preservation of anti-virus T cell responses and lower levels of cell-associated viral DNA²⁰. In a rare subpopulation of HIV-1 infected persons, cART initiated during the early phases of infection resulted in sustained control of viremia to very low levels, following treatment interruption²¹. Durable control of plasma viremia has also been reported in some SIVsmE660 infected rhesus macaques after anti-retroviral treatment, started 24 or 72h post IV inoculation, was discontinued²².

Because bNAbs suppress viremia^3,23, accelerate the clearance of cell-free HIV-1 virions²⁴, enhance clearance of infected cells^6,25,26, and boost host humoral immunity²³ in humans, they have the potential to mitigate deleterious events occurring during the acute infection and possibly alter the long-term clinical course. Consistent with this hypothesis, experiments in humanized mice indicate that early intervention with bNAbs may be more effective in preventing establishment of the reservoir than does cART⁵. However, humanized mice do not have intact immune systems, and fail to sustain infection beyond 3–4 months.

We previously reported that initiation of bNAb monotherapy, 12 weeks after SHIV_AD8-EO inoculation, controls plasma viremia in macaques for 1 or 2 weeks, after which resistant viral variants emerge¹². Thus, it is problematic whether even early combination bNAb immunotherapy might durably control viremia. To address this question, combination 10-1074¹³ and 3BNC117¹⁴ immunotherapy, which targets non-overlapping epitopes on the envelope spike, was administered to rhesus monkeys at the earliest possible time (viz. 3 days post infection [PI]) when we were certain that a SHIV_AD8-EO infection had been established.

SHIV_AD8-EO infection resembles HIV-1 infection in a number of important ways. It is R5 tropic, generates sustained levels of plasma viremia in inoculated macaques, exhibits a Tier 2 neutralization sensitivity phenotype, produces neutralization resistant variants in bNAb and ART treated animals, and causes irreversible depletions of CD4⁺ T cells in infected monkeys^12,27–29 (and unpublished). Infection leads to symptomatic immunodeficiency associated with opportunistic infections and a fatal clinical outcome in untreated monkeys.

In the initial experiment to assess the potency of combination bNAbs during acute infection, 6 macaques were inoculated intrarectally with 1000 TCID₅₀ of SHIV_AD8-EO, which is sufficient to infect all animals challenged by this route¹¹. Beginning on day 3 after inoculation, each monkey received a single course of 3 weekly intravenous (IV) bNAb infusions (on days 3, 10, and 17 PI) of 10-1074¹³ plus 3BNC117¹⁴. Compared to untreated animals, extremely low levels of plasma viremia could be detected in 2 of the 6 bNAb recipients during the first 30 days of infection (Fig. 1a). Plasma viremia in the other 4 macaques was not measurable (<100 SHIV RNA copies/ml) using standard RT-PCR assays.

All 6 monkeys infused with the bNAbs experienced sustained periods of virus suppression lasting 56 to 177, days at which point rebound viremia occurred in 5 of the 6 treated macaques (Extended Data Fig. 1). Plasma viremia remained below levels of detection in the sixth macaque (DFIK) during the first 150 days of infection/treatment (Extended Data Fig. 1f). As expected, the time to virus rebound was directly related to the concentration of bNAb in the plasma. For example, the SHIV_AD8-EO rebound in macaque MVJ occurred on day 56 when the level of 3BNC117 decayed below 1 μg/ml in plasma, whereas monkeys DEPH and DEWP experienced prolonged suppression of viremia before rebound was observed (Extended Data Fig. 1a, d, and e). We conclude that a single 2 week course of combination antibody therapy with 3BNC117 and 10-1074, administered early after an IR infection, can control SHIV_AD8-EO viremia for up to 177 days.

To determine whether low-level viral replication prior to virus rebound persisted despite apparent control of viremia, we performed ultrasensitive nested RT PCR (qRT-PCR) on plasma samples from animals with viral loads below levels of detection by standard methods (Extended Data Table 1). For example, in macaque DEPH, which did not generate a detectable virus rebound until day 177 (Extended Data Fig. 2d), the ultrasensitive assay measured less than 2 viral RNA copies/ml in plasma, between days 34 and 106 PI, and 10 RNA copies/ml on day 141 PI, a month prior to rebound. We conclude that low levels of virus are continuously being produced systemically and released into the circulation in the infected and bNAb treated macaques, even when plasma viremia is not detected by standard assays.

Two patterns of “post-rebound” plasma viremia were observed in the 6 bNAb treated monkeys. In the first, (DFIK, MVJ, and DEWP; Controller [Fig 2a–c]), the rebound SHIV_AD8-EO RNA loads in plasma returned to undetectable levels after a maximum of 20 weeks and remained so for an additional 20 to 30 weeks of observation. The second pattern occurred in monkeys DEPH, DFKX and DFFX (Fig. 2d–f; Non-Controller). These three macaques never fully controlled their viremia after rebound and displayed viral loads of 1 × 10^2–10³ RNA copies/ml between 110 and 130 weeks after virus inoculation.

Sustained control of plasma viremia by early bNAb therapy following a mucosal SHIV_AD8-EO challenge in 3 of 6 monkeys suggested that the same intervention might also be effective against challenge by the IV route, for which a smaller virus inoculum is typically required to establish an infection. Macaques were therefore inoculated with either 100 or 1000 TCID₅₀ of SHIV_AD8-EO IV and then administered 3 weekly bNAb infusions starting on day 3 after infection. As shown in Fig. 1b, the 4 recipients of 1000 TCID₅₀ of virus showed peak plasma virus loads ranging from 4.2 × 10³ to 1.1 × 10⁴ RNA copies/ml during the first 2 weeks of bNAb administration. Two of the 3 recipients (DEMR and DEHW) inoculated with 100 TCID₅₀ of virus IV had lower levels of plasma viremia during the early treatment period, and in the third animal (DEBA), viral loads remained below the level of detection throughout this time (Fig. 1c). Viremia declined to undetectable levels by 30 days following bNAb administration and remained suppressed for 48 to 110 days PI in all 7 combination bNAb treated monkeys inoculated by the IV route (Extended Data, Fig. 2). As previously observed in the IR challenged animals, virus rebound in the IV inoculated monkeys was directly associated with the decline of circulating bNAb levels.

To determine whether virus rebound was due to declining levels of bNAbs in the plasma or to the emergence of mutation(s) conferring resistance, we sequenced plasma viral RNA from the four monkeys (DELV, DF06, DEWL, and MAF), generating detectable viremia at a time when the 3BNC117 mAb was not detectable, but 10-1074 mAb concentrations remained above threshold levels (see Extended Data Fig 2). When samples collected at the peak of virus rebound from these 4 animals were analyzed, no changes were observed in amplicons from macaques DELV and DEWL, but 9 of 18 amplified sequences from macaque DF06 carried the N332S change and 12 of 24 amplicons from monkey MAF had the S334N substitution, both of which eliminated the glycan at position 332 of gp120 (Extended Data Fig. 3), the epitope targeted by the 10-1074 mAb. This result indicated that resistance to the 10-1074 mAb likely occurred in these 2 animals during.

Similar to the macaques challenged by the IR route, passive immunotherapy after IV infection also led to sustained virus control following the resolution of rebound viremia in some of the challenged monkeys. In 3 macaques (DEWL, MAF, and DEMR; Controllers), rebound viremia remained detectable for 42 to 90 weeks, and was then followed by long periods (25 to 56 additional weeks) during which virus replication was stably suppressed below the limits of detection (Fig. 3a–c). CD4⁺ T cell counts were maintained at a level of 800 cells/μl or higher in these 3 controller animals. In the second group of IV challenged animals (monkeys DF06, DELV, DEHW, and DEBA; Non-Controllers), virus replication was never completely suppressed following rebound (Fig. 3d–g). Failure to control viremia was associated with CD4⁺ T cell loss and progression to AIDS in 1 (macaque DF06) of these 4 monkeys.

Humoral and cellular immune responses were assessed before and after SHIV_AD8-EO rebound to determine whether anti-viral immunity was associated with controller status. No measurable correlation was observed between virus control and antibody levels. The controllers generated very low levels of anti-gp120 binding antibodies that were barely detectable by ELISA. Anti-viral CD8⁺ T cell responses were measured at 4 different times during the bNAb treatment, virus rebound, and post-rebound phases of their infections in both controllers and non-controllers. Although anti-SIVmac239 Gag CD8⁺ T cell responses were higher in animals challenged by the IV route, no major differences were observed between controllers and non-controllers (Extended Data, Fig. 4a). Furthermore, the polyfunctionality of the CD8⁺ T cell responses and distribution of CD8⁺ T cell memory subsets in IV inoculated controllers and non-controllers at different times following infection, treatment, and rebound were comparable (Extended Data Fig. 4b, c).

To determine whether CD8⁺ T cells might, in fact, be mediating the sustained suppression of virus replication, we initially administered the CD8 T cell depleting mAb MT807R1, which is specific for the CD8α chain, to all 6 controller macaques (MVJ [at day 434 PI], DEWP [at day 434 PI], DFIK [at day 336 PI], DEMR [at day 609 PI], MAF [at day 917 PI], and DEWL [at day 917 PI]) (Fig. 4a–f). All animals responded with a burst of plasma viremia, reaching levels between 10⁵ and 10⁷ viral RNA copies/ml, and which subsequently declined to baseline in all of the monkeys, except for DFIK.

Prior to the administration of the anti-CD8α mAb, quantitative virus outgrowth assays were performed on samples collected from the 6 controllers to measure the frequency of circulating CD4⁺ T cells releasing replication-competent virions. As shown in Extended Data Table 2, less than 1 in 10⁶ CD4⁺ T cells carrying infectious virus was detected in all of the controller monkeys immediately before infusion of the depleting anti-CD8α mAb. Interestingly, in 5 of the 6 anti-CD8α mAb recipients in which plasma viremia had returned to baseline levels following mAb infusion, the frequency of cells carrying replication-competent virus, determined by virus outgrowth assays, was also less than 1 in 10⁶ CD4⁺ T cells (Extended data Table 2).

A major deficiency of using the anti-CD8α mAb to deplete CD8⁺ T cells is that NK, NKT and γδ T cells are also targeted for depletion, as documented by FACS analyses for monkeys MVJ, DEWL, and DEMR (Extended Data Fig 5a,b). We therefore administered the CD8b255R1 anti-CD8β mAb, which specifically targets macaque CD8⁺ T cells, to deplete CD8⁺ T cells in these same 3 controller monkeys. As shown in Fig. 4a–c (red arrows), infusion of the anti-CD8β mAb caused an immediate increase in plasma virus loads, a decline in levels of CD3⁺ CD8⁺ T cells (Extended Data Fig. 5c), and no changes in circulating CD3^–CD8⁺ cells (Extended Data Fig. 5d). We conclude that CD8⁺ T cells are responsible for control of virus replication in controller macaques and depletion of this subset leads to recrudescence of viremia.

Although the non-controller monkeys failed to suppress plasma viremia to undetectable levels, 4 (DFKX, DFFX, DEHW, and DEBA) of 7 of these macaques have maintained low virus loads (105 to 385 RNA copies/ml) and did not experience significant changes in levels of circulating CD4⁺ T cell for 2 to 3 years after SHIV_AD8-EO infection (Figs. 2 and 3). Taken together, this indicates that 10 (6 controllers and 4 non-controllers) of the 13 bNAb treated monkeys benefited from early immunotherapy.

As a control for bNAb immunotherapy during the acute SHIV_AD8-EO infection, 3 macaques were inoculated by the IR route and treated daily with cART for 15 weeks, starting on day 3 after infection. This course of cART corresponded to the mean time period during which 2 weeks of bNAb therapy suppressed virus replication prior to rebound. cART therapy controlled and maintained undetectable levels of viremia for the entire 15 week treatment period (Extended Data Fig. 6). However, all 3 animals developed high sustained levels of plasma viremia following cessation of cART and none became controllers.

We speculate that the continuous production of low levels of progeny virions, as measured by ultrasensitive RT PCR (Extended Table 1) during the 50–140 day period before virus rebound in the antibody treated macaques, could drive the formation of immune complexes. Antigen-presenting dendritic cells expressing activating Fc receptors can bind to these immune complexes, leading to their activation and efficient antigen processing for presentation and cross-presentation to CD4⁺ and CD8⁺ T cells³⁰. In contrast, the near complete inhibition of virus replication by the cART regimen employed may limit the amount of viral antigen available to induce immunity.

As proof of concept, our results demonstrate that a SHIV infection, established during the acute phase, can, in fact, be controlled. They also suggest that a delicate balance may exist between: 1) preservation of helper CD4⁺ T cells; 2) the size and stability of the virus reservoir; and 3) the continuous production of sufficient quantities of antigen to generate a potent and sustained CD8⁺ T cell response. Although rhesus macaque infections with SHIV differs from HIV-1 infections in a number of important ways, immunotherapy should be explored as a way to control systemic dissemination of virus, contain damage to the CD4⁺ T cell lineage, and mobilize a robust immune response that may be capable of controlling the infection in humans.