HIV-1 control in vivo is related to the number but not the fraction of infected cells with viral unspliced RNA.

In the absence of antiretroviral therapy (ART), a subset of individuals, termed HIV controllers, have levels of plasma viremia that are orders of magnitude lower than non-controllers (NC) who are at higher risk for HIV disease progression. In addition to having fewer infected cells resulting in fewer cells with HIV RNA, it is possible that lower levels of plasma viremia in controllers are due to a lower fraction of the infected cells having HIV-1 unspliced RNA (HIV usRNA) compared with NC. To directly test this possibility, we used sensitive and quantitative single-cell sequencing methods to compare the fraction of infected cells that contain one or more copies of HIV usRNA in peripheral blood mononuclear cells (PBMC) obtained from controllers and NC. The fraction of infected cells containing HIV usRNA did not differ between the two groups. Rather, the levels of viremia were strongly associated with the total number of infected cells that had HIV usRNA, as reported by others, with controllers having 34-fold fewer infected cells per million PBMC. These results reveal that viremic control is not associated with a lower fraction of proviruses expressing HIV usRNA, unlike what is reported for elite controllers, but is only related to having fewer infected cells overall, maybe reflecting greater immune clearance of infected cells. Our findings show that proviral silencing is not a key mechanism for viremic control and will help to refine strategies toward achieving HIV remission without ART.

Impact of alemtuzumab-mediated lymphocyte depletion on SIV reservoir establishment and persistence.

Persistence of the rebound-competent viral reservoir (RCVR) within the CD4+ T cell compartment of people living with HIV remains a major barrier to HIV cure. Here, we determined the effects of the pan-lymphocyte-depleting monoclonal antibody (mAb) alemtuzumab on the RCVR in SIVmac239-infected rhesus macaques (RM) receiving antiretroviral therapy (ART). Alemtuzumab administered during chronic ART or at the time of ART initiation induced >95% depletion of circulating CD4+ T cells in peripheral blood and substantial CD4+ T cell depletion in lymph nodes. However, treatment was followed by proliferation and reconstitution of CD4+ T cells in blood, and despite ongoing ART, levels of cell-associated SIV DNA in blood and lymphoid tissues were not substantially different between alemtuzumab-treated and control RM after immune cell reconstitution, irrespective of the time of alemtuzumab treatment. Upon ART cessation, 19 of 22 alemtuzumab-treated RM and 13 of 13 controls rebounded with no difference in the time to rebound between treatment groups. Time to rebound and reactivation rate was associated with plasma viral loads (pVLs) at time of ART initiation, suggesting lymphocyte depletion had no durable impact on the RCVR. However, 3 alemtuzumab-treated RM that had lowest levels of pre-ART viremia, failed to rebound after ART withdrawal, in contrast to controls with similar levels of SIV replication. These observations suggest that alemtuzumab therapy has little to no ability to reduce well-established RCVRs but may facilitate RCVR destabilization when pre-ART virus levels are particularly low.

Engineered deletions of HIV replicate conditionally to reduce disease in nonhuman primates.

Antiviral therapies with reduced frequencies of administration and high barriers to resistance remain a major goal. For HIV, theories have proposed that viral-deletion variants, which conditionally replicate with a basic reproductive ratio [R0] > 1 (termed "therapeutic interfering particles" or "TIPs"), could parasitize wild-type virus to constitute single-administration, escape-resistant antiviral therapies. We report the engineering of a TIP that, in rhesus macaques, reduces viremia of a highly pathogenic model of HIV by >3log10 following a single intravenous injection. Animal lifespan was significantly extended, TIPs conditionally replicated and were continually detected for >6 months, and sequencing data showed no evidence of viral escape. A single TIP injection also suppressed virus replication in humanized mice and cells from persons living with HIV. These data provide proof of concept for a potential new class of single-administration antiviral therapies.

No evidence for ongoing replication on ART in SIV-infected macaques.

The capacity of HIV-1 to replicate during optimal antiretroviral therapy (ART) is challenging to assess directly. To gain greater sensitivity to detect evolution on ART, we used a nonhuman primate (NHP) model providing precise control over the level of pre-ART evolution and more comprehensive analyses than are possible with clinical samples. We infected 21 rhesus macaques (RMs) with the barcoded virus SIVmac239M and initiated ART early to minimize baseline genetic diversity. RMs were treated for 285-1200 days. We used several tests of molecular evolution to compare 1352 near-full-length (nFL) SIV DNA single genome sequences from PBMCs, lymph nodes, and spleen obtained near the time of ART initiation and those present after long-term ART, none of which showed significant changes to the SIV DNA population during ART in any animal. To investigate the possibility of ongoing replication in unsampled putative tissue sanctuaries during ART, we discontinued treatment in four animals and confirmed that none of the 336 nFL SIV RNA sequences obtained from rebound plasma viremia showed evidence of evolution. The rigorous nature of our analyses reinforced the emerging consensus of a lack of appreciable ongoing replication on effective ART and validates the relevance of this NHP model for cure studies.

Immunotoxin-mediated depletion of Gag-specific CD8+ T cells undermines natural control of SIV.

Antibody-mediated depletion studies have demonstrated that CD8+ T cells are required for effective immune control of SIV. However, this approach is potentially confounded by several factors, including reactive CD4+ T cell proliferation, and provides no information on epitope specificity, a likely determinant of CD8+ T cell efficacy. We circumvented these limitations by selectively depleting CD8+ T cells specific for the Gag epitope CTPYDINQM (CM9) via the administration of immunotoxin-conjugated tetrameric complexes of CM9/Mamu-A*01. Immunotoxin administration effectively depleted circulating but not tissue-localized CM9-specific CD8+ T cells, akin to the bulk depletion pattern observed with antibodies directed against CD8. However, we found no evidence to indicate that circulating CM9-specific CD8+ T cells suppressed viral replication in Mamu-A*01+ rhesus macaques during acute or chronic progressive infection with a pathogenic strain of SIV. This observation extended to macaques with established infection during and after continuous antiretroviral therapy. In contrast, natural controller macaques experienced dramatic increases in plasma viremia after immunotoxin administration, highlighting the importance of CD8+ T cell-mediated immunity against CM9. Collectively, these data showed that CM9-specific CD8+ T cells were necessary but not sufficient for robust immune control of SIV in a nonhuman primate model and, more generally, validated an approach that could inform the design of next-generation vaccines against HIV-1.

Innate protection against intrarectal SIV acquisition by a live SHIV vaccine.

Identifying immune correlates of protection is a major challenge in AIDS vaccine development. Anti-Envelope antibodies have been considered critical for protection against SIV/HIV (SHIV) acquisition. Here, we evaluated the efficacy of an SHIV vaccine against SIVmac251 challenge, where the role of antibody was excluded, as there was no cross-reactivity between SIV and SHIV envelope antibodies. After 8 low-dose intrarectal challenges with SIVmac251, 12 SHIV-vaccinated animals demonstrated efficacy, compared with 6 naive controls, suggesting protection was achieved in the absence of anti-envelope antibodies. Interestingly, CD8+ T cells (and some NK cells) were not essential for preventing viral acquisition, as none of the CD8-depleted macaques were infected by SIVmac251 challenges. Initial investigation of protective innate immunity revealed that protected animals had elevated pathways related to platelet aggregation/activation and reduced pathways related to interferon and responses to virus. Moreover, higher expression of platelet factor 4 on circulating platelet-leukocyte aggregates was associated with reduced viral acquisition. Our data highlighted the importance of innate immunity, identified mechanisms, and may provide opportunities for novel HIV vaccines or therapeutic strategy development.

Adeno-associated viral delivery of Env-specific antibodies prevents SIV rebound after discontinuing antiretroviral therapy.

An alternative to lifelong antiretroviral therapy (ART) is needed to achieve durable control of HIV-1. Here we show that adeno-associated virus (AAV)-delivery of two rhesus macaque antibodies to the SIV envelope glycoprotein (Env) with potent neutralization and antibody-dependent cellular cytotoxicity can prevent viral rebound in macaques infected with barcoded SIVmac239M after discontinuing suppressive ART. Following AAV administration, sustained antibody expression with minimal anti-drug antibody responses was achieved in all but one animal. After ART withdrawal, SIV replication rebounded within two weeks in all of the control animals but remained below the threshold of detection in plasma (<15 copies/mL) for more than a year in four of the eight animals that received AAV vectors encoding Env-specific antibodies. Viral sequences from animals with delayed rebound exhibited restricted barcode diversity and antibody escape. Thus, sustained expression of antibodies with potent antiviral activity can afford durable, ART-free containment of pathogenic SIV infection.

Early antiretroviral therapy in SIV-infected rhesus macaques reveals a multiphasic, saturable dynamic accumulation of the rebound competent viral reservoir.

The rebound competent viral reservoir (RCVR)-virus that persists during antiretroviral treatment (ART) and can reignite systemic infection when treatment is stopped-is the primary barrier to eradicating HIV. We used time to initiation of ART during primary infection of rhesus macaques (RMs) after intravenous challenge with barcoded SIVmac239 as a means to elucidate the dynamics of RCVR establishment in groups of RMs by creating a multi-log range of pre-ART viral loads and then assessed viral time-to-rebound and reactivation rates resulting from the discontinuation of ART after one year. RMs started on ART on days 3, 4, 5, 6, 7, 9 or 12 post-infection showed a nearly 10-fold difference in pre-ART viral measurements for successive ART-initiation timepoints. Only 1 of 8 RMs initiating ART on days 3 and 4 rebounded after ART interruption despite measurable pre-ART plasma viremia. Rebounding plasma from the 1 rebounding RM contained only a single barcode lineage detected at day 50 post-ART. All RMs starting ART on days 5 and 6 rebounded between 14- and 50-days post-ART with 1-2 rebounding variants each. RMs starting ART on days 7, 9, and 12 had similar time-to-measurable plasma rebound kinetics despite multiple log differences in pre-ART plasma viral load (pVL), with all RMs rebounding between 7- and 16-days post-ART with 3-28 rebounding lineages. Calculated reactivation rates per pre-ART pVL were highest for RMs starting ART on days 5, 6, and 7 after which the rate of accumulation of the RCVR markedly decreased for RMs treated on days 9 and 12, consistent with multiphasic establishment and near saturation of the RCVR within 2 weeks post infection. Taken together, these data highlight the heterogeneity of the RCVR between RMs, the stochastic establishment of the very early RCVR, and the saturability of the RCVR prior to peak viral infection.

TGF-ß blockade drives a transitional effector phenotype in T cells reversing SIV latency and decreasing SIV reservoirs in vivo.

HIV-1 persistence during ART is due to the establishment of long-lived viral reservoirs in resting immune cells. Using an NHP model of barcoded SIVmac239 intravenous infection and therapeutic dosing of anti-TGFBR1 inhibitor galunisertib (LY2157299), we confirm the latency reversal properties of in vivo TGF-ß blockade, decrease viral reservoirs and stimulate immune responses. Treatment of eight female, SIV-infected macaques on ART with four 2-weeks cycles of galunisertib leads to viral reactivation as indicated by plasma viral load and immunoPET/CT with a 64Cu-DOTA-F(ab')2-p7D3-probe. Post-galunisertib, lymph nodes, gut and PBMC exhibit lower cell-associated (CA-)SIV DNA and lower intact pro-virus (PBMC). Galunisertib does not lead to systemic increase in inflammatory cytokines. High-dimensional cytometry, bulk, and single-cell (sc)RNAseq reveal a galunisertib-driven shift toward an effector phenotype in T and NK cells characterized by a progressive downregulation in TCF1. In summary, we demonstrate that galunisertib, a clinical stage TGF-ß inhibitor, reverses SIV latency and decreases SIV reservoirs by driving T cells toward an effector phenotype, enhancing immune responses in vivo in absence of toxicity.

Potent antibody-dependent cellular cytotoxicity of a V2-specific antibody is not sufficient for protection of macaques against SIV challenge.

Fc-mediated antibody effector functions, such as antibody-dependent cellular cytotoxicity (ADCC), can contribute to the containment HIV-1 replication but whether such activities are sufficient for protection is unclear. We previously identified an antibody to the variable 2 (V2) apex of the HIV-1 Env trimer (PGT145) that potently directs the lysis of SIV-infected cells by NK cells but poorly neutralizes SIV infectivity. To determine if ADCC is sufficient for protection, separate groups of six rhesus macaques were treated with PGT145 or a control antibody (DEN3) by intravenous infusion followed five days later by intrarectal challenge with SIVmac239. Despite high concentrations of PGT145 and potent ADCC activity in plasma on the day of challenge, all animals became infected and viral loads did not differ between the PGT145- and DEN3-treated animals. To determine if PGT145 can protect against a neutralization-sensitive virus, two additional groups of six macaques were treated with PGT145 and DEN3 and challenged with an SIVmac239 variant with a single amino acid change in Env (K180S) that increases PGT145 binding and renders the virus susceptible to neutralization by this antibody. Although there was no difference in virus acquisition, peak and chronic phase viral loads were significantly lower and time to peak viremia was significantly delayed in the PGT145-treated animals compared to the DEN3-treated control animals. Env changes were also selected in the PGT145-treated animals that confer resistance to both neutralization and ADCC. These results show that ADCC is not sufficient for protection by this V2-specific antibody. However, protection may be achieved by increasing the affinity of antibody binding to Env above the threshold required for neutralization.

Lymphoid tissues contribute to plasma viral clonotypes early after antiretroviral therapy interruption in SIV-infected rhesus macaques.

The rebound-competent viral reservoir, composed of a virus that is able to persist during antiretroviral therapy (ART) and mediate reactivation of systemic viral replication and rebound viremia after ART interruption (ATI), remains the biggest obstacle to treating HIV infection. A better understanding of the cellular and tissue origins and the dynamics of viral populations that initiate rebound upon ATI could help develop therapeutic strategies for reducing the rebound-competent viral reservoir. In this study, barcoded simian immunodeficiency virus (SIV), SIVmac239M, was used to infect rhesus macaques to enable monitoring of viral barcode clonotypes contributing to virus detectable in plasma after ATI. Blood and tissues from secondary lymphoid organs (spleen, mesenteric lymph nodes, and inguinal lymph nodes) and from the colon, ileum, lung, liver, and brain were analyzed using viral barcode sequencing, intact proviral DNA assay, single-cell RNA sequencing, and combined CODEX and RNAscope in situ hybridization. Four of seven animals had viral barcodes detectable by deep sequencing of plasma at necropsy, although plasma viral RNA remained below 22 copies per milliliter. Among the tissues studied, mesenteric lymph nodes, inguinal lymph nodes, and spleen contained viral barcodes detected in plasma. CD4+ T cells were the main cell type harboring viral RNA after ATI. Furthermore, T cell zones in lymphoid tissues showed higher viral RNA abundance than B cell zones for most animals. These findings are consistent with lymphoid tissues contributing to the virus present in plasma early after ATI.

TGF-ß blockade drives a transitional effector phenotype in T cells reversing SIV latency and decreasing SIV reservoirs in vivo.

HIV-1 persistence during ART is due to the establishment of long-lived viral reservoirs in resting immune cells. Using an NHP model of barcoded SIVmac239 intravenous infection and therapeutic dosing of the anti-TGFBR1 inhibitor galunisertib (LY2157299), we confirmed the latency reversal properties of in vivo TGF-ß blockade, decreased viral reservoirs and stimulated immune responses. Eight SIV-infected macaques on suppressive ART were treated with 4 2-week cycles of galunisertib. ART was discontinued 3 weeks after the last dose, and macaques euthanized 6 weeks after ART-interruption(ATI). One macaque did not rebound, while the remaining rebounded between week 2 and 6 post-ATI. Galunisertib led to viral reactivation as indicated by plasma viral load and immunoPET/CT with the 64Cu-DOTA-F(ab')2-p7D3-probe. Half to 1 Log decrease in cell-associated (CA-)SIV DNA was detected in lymph nodes, gut and PBMC, while intact pro-virus in PBMC decreased by 3-fold. No systemic increase in inflammatory cytokines was observed. High-dimensions cytometry, bulk and single-cell RNAseq revealed a shift toward an effector phenotype in T and NK cells. In summary, we demonstrated that galunisertib, a clinical stage TGFß inhibitor, reverses SIV latency and decreases SIV reservoirs by driving T cells toward an effector phenotype, enhancing immune responses in vivo in absence of toxicity.

Timing of initiation of anti-retroviral therapy predicts post-treatment control of SIV replication.

One approach to 'functional cure' of HIV infection is to induce durable control of HIV replication after the interruption of antiretroviral therapy (ART). However, the major factors that determine the viral 'setpoint' level after treatment interruption are not well understood. Here we combine data on ART interruption following SIV infection for 124 total animals from 10 independent studies across 3 institutional cohorts to understand the dynamics and predictors of post-treatment viral control. We find that the timing of treatment initiation is an important determinant of both the peak and early setpoint viral levels after treatment interruption. During the first 3 weeks of infection, every day of delay in treatment initiation is associated with a 0.22 log10 copies/ml decrease in post-rebound peak and setpoint viral levels. However, delay in initiation of ART beyond 3 weeks of infection is associated with higher post-rebound setpoint viral levels. For animals treated beyond 3 weeks post-infection, viral load at ART initiation was the primary predictor of post-rebound setpoint viral levels. Potential alternative predictors of post-rebound setpoint viral loads including cell-associated DNA or RNA, time from treatment interruption to rebound, and pre-interruption CD8+ T cell responses were also examined in the studies where these data were available. This analysis suggests that optimal timing of treatment initiation may be an important determinant of post-treatment control of HIV.

Lymphoid tissues contribute to viral clonotypes present in plasma at early post-ATI in SIV-infected rhesus macaques.

The rebound-competent viral reservoir (RCVR), comprised of virus that is able to persist during antiretroviral therapy (ART) and mediate reactivation of systemic viral replication and rebound viremia after antiretroviral therapy interruption (ATI), remains the biggest obstacle to the eradication of HIV infection. A better understanding of the cellular and tissue origins and the dynamics of viral populations that initiate rebound upon ATI could help develop targeted therapeutic strategies for reducing the RCVR. In this study, barcoded SIVmac239M was used to infect rhesus macaques to enable monitoring of viral barcode clonotypes contributing to virus detectable in plasma after ATI. Blood, lymphoid tissues (LTs, spleen, mesenteric and inguinal lymph nodes), and non-lymphoid tissues (NLTs, colon, ileum, lung, liver, and brain) were analyzed using viral barcode sequencing, intact proviral DNA assay, single-cell RNA sequencing, and combined CODEX/RNAscope/ in situ hybridization. Four of seven animals had viral barcodes detectable by deep sequencing of plasma at necropsy although plasma viral RNA remained < 22 copies/mL. Among the tissues studied, mesenteric and inguinal lymph nodes, and spleen contained viral barcodes detected in plasma, and trended to have higher cell-associated viral loads, higher intact provirus levels, and greater diversity of viral barcodes. CD4+ T cells were the main cell type harboring viral RNA (vRNA) after ATI. Further, T cell zones in LTs showed higher vRNA levels than B cell zones for most animals. These findings are consistent with LTs contributing to virus present in plasma early after ATI. One Sentence Summary: The reemerging of SIV clonotypes at early post-ATI are likely from the secondary lymphoid tissues.

Immune reconstitution inflammatory syndrome drives emergence of HIV drug resistance from multiple anatomic compartments in a person living with HIV.

Reservoirs of HIV maintained in anatomic compartments during antiretroviral therapy prevent HIV eradication. However, mechanisms driving their persistence and interventions to control them remain elusive. Here we report the presence of an inducible HIV reservoir within antigen-specific CD4+T cells in the central nervous system of a 59-year-old male with progressive multifocal leukoencephalopathy immune reconstitution inflammatory syndrome (PML-IRIS). HIV production during PML-IRIS was suppressed by modulating inflammation with corticosteroids; selection of HIV drug resistance caused subsequent breakthrough viremia. Therefore, inflammation can influence the composition, distribution and induction of HIV reservoirs, warranting it as a key consideration for developing effective HIV remission strategies.

Correction: Mitigation of endemic GI-tract pathogen-mediated inflammation through development of multimodal treatment regimen and its impact on SIV acquisition in rhesus macaques.

[This corrects the article DOI: 10.1371/journal.ppat.1009565.].

Experimental bacterial dysbiosis with consequent immune alterations increase intrarectal SIV acquisition susceptibility.

Variations in the composition of the intestinal bacterial microbiome correlate with acquisition of some sexually transmitted pathogens. To experimentally assess the contribution of intestinal dysbiosis to rectal lentiviral acquisition, we induce dysbiosis in rhesus macaques (RMs) with the antibiotic vancomycin prior to repeated low-dose intrarectal challenge with simian immunodeficiency virus (SIV) SIVmac239X. Vancomycin administration reduces T helper 17 (TH17) and TH22 frequencies, increases expression of host bacterial sensors and antibacterial peptides, and increases numbers of transmitted-founder (T/F) variants detected upon SIV acquisition. We observe that SIV acquisition does not correlate with measures of dysbiosis but rather associates with perturbations in the host antimicrobial program. These findings establish a functional association between the intestinal microbiome and susceptibility to lentiviral acquisition across the rectal epithelial barrier.

CD8+ lymphocytes do not impact SIV reservoir establishment under ART.

Persistence of the human immunodeficiency virus type-1 (HIV-1) latent reservoir in infected individuals remains a problem despite fully suppressive antiretroviral therapy (ART). While reservoir formation begins during acute infection, the mechanisms responsible for its establishment remain unclear. CD8+ T cells are important during the initial control of viral replication. Here we examined the effect of CD8+ T cells on formation of the latent reservoir in simian immunodeficiency virus (SIV)-infected macaques by performing experimental CD8+ depletion either before infection or before early (that is, day 14 post-infection) ART initiation. We found that CD8+ depletion resulted in slower decline of viremia, indicating that CD8+ lymphocytes reduce the average lifespan of productively infected cells during acute infection and early ART, presumably through SIV-specific cytotoxic T lymphocyte (CTL) activity. However, CD8+ depletion did not change the frequency of infected CD4+ T cells in the blood or lymph node as measured by the total cell-associated viral DNA or intact provirus DNA assay. In addition, the size of the persistent reservoir remained the same when measuring the kinetics of virus rebound after ART interruption. These data indicate that during early SIV infection, the viral reservoir that persists under ART is established largely independent of CTL control.

Natural Killer Cells Regulate Acute SIV Replication, Dissemination, and Inflammation, but Do Not Impact Independent Transmission Events.

Natural killer (NK) cells are potent effector cells of the innate immune system possessing both cytotoxic and immunoregulatory capabilities, which contribute to their crucial role in controlling human immunodeficiency virus (HIV) and simian immunodeficiency virus (SIV) infections. However, despite significant evidence for NK cell modulation of HIV disease, their specific contribution to transmission and control of acute infection remains less clear. To elucidate the contribution of NK cells during acute SIV infection, we performed an acute necropsy study, where rhesus macaques (RM) were subjected to preinfection depletion of systemic NK cells using established methods of IL-15 neutralization, followed by subsequent challenge with barcoded SIVmac239X. Our study showed that depletion was highly effective, resulting in near total ablation of all NK cell subsets in blood, liver, oral, and rectal mucosae, and lymph nodes (LN) that persisted through the duration of the study. Meanwhile, frequencies and phenotypes of T cells remained virtually unchanged, indicating that our method of NK cell depletion had minimal off-target effects. Importantly, NK cell-depleted RM demonstrated an early and sustained 1 to 2 log increase in viremia over controls, but sequence analysis suggested no difference in the number of independent transmission events. Acute bulk, central memory (CM), and CCR5+ CD4+ T cell depletion was similar between experimental and control groups, while CD8+ T cell activation was higher in NK cell-depleted RM as measured by Ki67 and PD-1 expression. Using 27-plex Luminex analyses, we also found modestly increased inflammatory cytokines in NK cell-depleted RM compared to control animals. In the effort to determine the impact of NK cells on HIV/SIV transmission and acute viremia, future studies will be necessary to better harness these cells for future viral therapies. Collectively, these data suggest NK cells are important modulators of lentivirus dissemination and disease but may not have the capacity to independently eliminate individual transmission events. IMPORTANCE Natural killer (NK) cells as major effector cells of the innate immune system can contribute significantly to human immunodeficiency virus (HIV) and simian immunodeficiency virus (SIV) control. However, a specific role for NK cells in blocking lentivirus transmission remains incompletely clear. In this study, we depleted NK cells prior to challenge with a barcoded SIV. Importantly, our studied showed systemic NK cell depletion was associated with a significant increase in acute viremia, but did not impact the number of independent transmission events. Collectively, these data suggest NK cells are critical modulators of early lentivirus replication but may not regulate individual transmission events at mucosal portals of entry.

A cellular trafficking signal in the SIV envelope protein cytoplasmic domain is strongly selected for in pathogenic infection.

The HIV/SIV envelope glycoprotein (Env) cytoplasmic domain contains a highly conserved Tyr-based trafficking signal that mediates both clathrin-dependent endocytosis and polarized sorting. Despite extensive analysis, the role of these functions in viral infection and pathogenesis is unclear. An SIV molecular clone (SIVmac239) in which this signal is inactivated by deletion of Gly-720 and Tyr-721 (SIVmac239ΔGY), replicates acutely to high levels in pigtail macaques (PTM) but is rapidly controlled. However, we previously reported that rhesus macaques and PTM can progress to AIDS following SIVmac239ΔGY infection in association with novel amino acid changes in the Env cytoplasmic domain. These included an R722G flanking the ΔGY deletion and a nine nucleotide deletion encoding amino acids 734-736 (ΔQTH) that overlaps the rev and tat open reading frames. We show that molecular clones containing these mutations reconstitute signals for both endocytosis and polarized sorting. In one PTM, a novel genotype was selected that generated a new signal for polarized sorting but not endocytosis. This genotype, together with the ΔGY mutation, was conserved in association with high viral loads for several months when introduced into naïve PTMs. For the first time, our findings reveal strong selection pressure for Env endocytosis and particularly for polarized sorting during pathogenic SIV infection in vivo.