| |
Broad CTL response is required to clear latent
HIV-1 due to dominance of escape mutations
|
| |
| |
Download the PDF here
"Our results demonstrate that chronically infected patients retain a broad-spectrum viral-specific CTL response and that appropriate boosting of this response may be required for the elimination of the latent reservoir."
"The seeding of the HIV-1 latent reservoir starts just a few days after infection26, before the development of a robust CTL response14. This is consistent with our finding that patients who initiated treatment early, in the acute infection stage, have few if any CTL escape variants archived in the latent reservoir. However, if treatment was initiated in chronic infection, CTL escape variants became dominant in the latent reservoir, indicating a complete replacement of the initially established 'wild-type' reservoir. The mechanism behind this replacement warrants further investigation, but probably reflects the dynamic nature of the reservoir in untreated infection. In any event, the overwhelming presence of escape variants in the latent reservoir of chronic patients certainly presents an additional barrier to eradication efforts. The striking difference between AP- and CP-treated patients presents another argument for early treatment of HIV-1 infection; early treatment not only reduces the size of the latent reservoir27, but also alters the composition of the reservoir, as shown here, in a way that may enhance the efficacy of potential CTL-based eradication therapies."
----------------------------
Broad CTL response is required to clear latent HIV-1 due to dominance of escape mutations
Nature
(2015)
Nature | Letter
------------------------
NIH Grantees Overcome Hurdle to Kill HIV-Infected Cells Brought out of Hiding
Technique Could Potentially Become Part of HIV Cure Strategy
WHAT:
A major obstacle to curing people of HIV infection is the way the virus hides in a reservoir primarily of dormant immune cells called resting memory CD4+ T cells. One potential approach to curing HIV infection is to awaken these latent CD4+ T cells so they start making HIV proteins. This would alert the immune system that the cells are infected, and, in theory, generate an immune response that kills them. It has been unclear, however, whether typical immune mechanisms for killing virally infected cells would eliminate HIV-infected CD4+ T cells awakened from the HIV reservoir.
To answer this question, NIH grantee Robert F. Siliciano, M.D., Ph.D., of the Howard Hughes Medical Institute and the Johns Hopkins University School of Medicine, and colleagues extracted immune cells and reservoir-based HIV from 25 infected people to study in the laboratory and in mice. Ten of these people had started combination anti-HIV therapy early (within 3 months of infection) and 15 had started late (3 months or more after infection). In the early-treatment group, scientists found that most of the HIV-infected CD4+ T cells in the viral reservoirs were sensitive to detection by killer T cells, the immune cells that seek and destroy infected cells. By contrast, nearly all of the HIV that infected CD4+ T cells in the reservoirs of the late-treatment group had developed mutations that enabled the infected CD4+ T cells to escape detection by the killer T cells that typically dominate the immune response to HIV infection.
Despite this, the scientists discovered that most HIV-infected people in the late-treatment group also had other killer T cells that recognized parts of HIV that had not mutated, but these cells were ineffective at destroying their targets. To boost the killing capacity of these cells, the researchers stimulated them with a mixture of HIV protein fragments before exposing them to unmutated parts of the virus. The boosted cells effectively killed the HIV-infected cells in both the laboratory and mice altered to have human immune systems. This suggests that a therapeutic vaccine that similarly boosts the T-cell response to HIV could be part of a strategy for curing chronic HIV infection.
----------------------------------
Broad CTL response is required to clear latent HIV-1 due to dominance of escape mutations
Nature Jan 7 2015
Kai Deng1, Mihaela Pertea1,2, Anthony Rongvaux3, LeyaoWang4, Christine M. Durand1, Gabriel Ghiaur5, Jun Lai1,
Holly L. McHugh1, Haiping Hao6, Hao Zhang7, Joseph B. Margolick7, Cagan Gurer8, Andrew J. Murphy8, David M. Valenzuela8,
George D. Yancopoulos8, Steven G. Deeks9, Till Strowig3, Priti Kumar10, Janet D. Siliciano1, Steven L. Salzberg2,11,
Richard A. Flavell3,12, Liang Shan3 & Robert F. Siliciano1,13
Despite antiretroviral therapy (ART), human immunodeficiency virus (HIV)-1 persists in a stable latent reservoir1, 2, primarily in resting memory CD4+ T cells3, 4. This reservoir presents a major barrier to the cure of HIV-1 infection. To purge the reservoir, pharmacological reactivation of latent HIV-1 has been proposed5 and tested both in vitro and in vivo6, 7, 8. A key remaining question is whether virus-specific immune mechanisms, including cytotoxic T lymphocytes (CTLs), can clear infected cells in ART-treated patients after latency is reversed. Here we show that there is a striking all or none pattern for CTL escape mutations in HIV-1 Gag epitopes. Unless ART is started early, the vast majority (>98%) of latent viruses carry CTL escape mutations that render infected cells insensitive to CTLs directed at common epitopes. To solve this problem, we identified CTLs that could recognize epitopes from latent HIV-1 that were unmutated in every chronically infected patient tested. Upon stimulation, these CTLs eliminated target cells infected with autologous virus derived from the latent reservoir, both in vitro and in patient-derived humanized mice. The predominance of CTL-resistant viruses in the latent reservoir poses a major challenge to viral eradication. Our results demonstrate that chronically infected patients retain a broad-spectrum viral-specific CTL response and that appropriate boosting of this response may be required for the elimination of the latent reservoir.
HIV-1 establishes latent infection in resting CD4+ T cells3, 4. Recent efforts to eradicate HIV-1 infection have focused on reversing latency without global T-cell activation5. However, inducing HIV-1 gene expression in latently infected cells is not sufficient to cause the death of these cells if they remain in a resting state9. Boosting HIV-1-specific immune responses, including CTL responses, may be required for clearance of the latent reservoir9. CTLs have a significant role in suppressing HIV-1 replication in acute infection10, 11, 12, 13, 14. Because of this strong selective pressure, HIV-1 quickly acquires mutations to evade CTL recognition12, 13, 15, 16, 17, 18. CTL escape has been studied primarily through the analysis of plasma virus12, 13, 16, 18, 19, 20, and CTL-based vaccines have been designed based on conserved epitopes21, 22. A systematic investigation of CTL escape in the latent reservoir will be of great importance to the ongoing CTL-based virus eradication efforts, because latent HIV-1 probably represents the major source of viral rebound after treatment interruption. Earlier studies have suggested the presence of CTL escape mutations in proviral DNA15, 17, but it remains unclear to what extent the latent reservoir in resting CD4+ T cells is affected by CTL escape, whether mutations detected in proviral DNA are representative of the very small fraction of proviruses that are replication competent, and, most importantly, whether the CTL response can recognize and clear infected cells after latency is reversed.
To investigate CTL escape variants in the latent reservoir, we deep sequenced the proviral HIV-1 DNA in resting CD4+ T cells from 25 patients (Extended Data Table 1). Among them, 10 initiated ART during the acute phase (AP; within 3 months of infection) while the other 15 initiated ART during the chronic phase (CP) of infection. The sequencing was focused on Gag because it is an important target of the CTL response23 and is highly conserved, which facilitates the detection of escape variants. Our data show that previously documented CTL escape variants completely dominate the viral reservoirs of nearly all CP-treated patients (Extended Data Fig. 1 and Supplementary Table 1). This trend is especially obvious for several well characterized CTL epitopes: the human leukocyte antigen (HLA)-A2-restricted epitope SLYNTVATL (SL9), the HLA-A3-restricted epitope RLRPGGKKK (RK9) and the HLA-B57/58-restricted epitope TSTLQEQIGW (TW10) (Fig. 1a and Extended Data Fig. 1, highlighted in coloured boxes.). In these epitopes, close to 100% of the sequences harboured escape mutations. Comparison of mutation frequencies between HLA allele-relevant and -irrelevant epitopes in CP-treated patients suggests that the CTL escape mutations identified are specific to each patient's HLA type (Fig. 1b). By contrast, except for SL9 from patient AP01 and RK9 from patient AP08, few if any CTL escape mutations were archived in AP-treated patients (Fig. 1c and Extended Data Fig. 1). The striking difference between AP- and CP-treated patients (Fig. 1c) indicates that, unless treatment is initiated within the first several months of infection, the latent reservoir becomes almost completely dominated by variants resistant to dominant CTL responses.
To confirm that variants detected at high frequency in the latent reservoir represent functional CTL escape mutants, cells from seven CP-treated subjects were tested for reactivity to synthetic peptides representing wild-type and mutant versions of the relevant epitopes. As expected, there were only minimal responses to previously documented CTL escape mutants by patient CD8+ T cells, and no de novo response was detected (Fig. 1d and Extended Data Fig. 2). In contrast, all tested subjects retained a strong response to peptides representing the wild-type epitopes, suggesting that the wild-type virus was initially transmitted, with subsequent evolution of CTL escape variants. Most HIV-1 proviruses detected in patients are defective24. Therefore, to determine whether these CTL escape variants can be reactivated and lead to viral rebound if therapy is stopped, we isolated replication-competent viruses from the latent reservoirs of nine CP-treated patients. We found that all the dominant CTL escape mutations that had been identified in proviruses in resting CD4+ T cells were also present in the replication-competent viruses that grew out after T-cell activation (Fig. 1e and Extended Data Fig. 3), indicating that these CTL escape variants not only dominate the population of proviruses, but can also be released and replicate once latency is reversed.
We next asked whether the host CTL response could recognize and eliminate the cells infected with these escape variants. We infected activated CD4+ T cells from these patients with autologous, replication-competent virus derived from the latent reservoir (Extended Data Fig. 4a). The infected cells were then co-cultured with autologous CD8+ T cells, either unstimulated or pre-stimulated, to assess HIV-1-specific cytolytic activity. Non-specific activation of CD8+ T cells was not observed after co-culture with phytohaemagglutinin (PHA)-activated CD4+ T cells (Extended Data Fig. 4b). From all 13 CP-treated subjects tested, CD8+ T cells pre-stimulated by a Gag peptide mixture efficiently killed autologous infected CD4+ T cells (median 61% elimination), while unstimulated CD8+ T cells from most subjects had significantly less effect (median 23% elimination) (Fig. 2a and Extended Data Fig. 4c, d). CD8+ T cells from 7/7 healthy donors completely failed to eliminate autologous infected cells (Fig. 2a), confirming that the observed killing was HIV-1 specific. The killing effect was enhanced by increasing the effector to target ratio (Extended Data Fig. 5a), and was cell-cell contact dependent (Extended Data Fig. 5b). When the co-culture was maintained over time in the absence of ART, viral replication was significantly reduced, but not completely inhibited by pre-stimulated CD8+ T cells (Fig. 2b). We found that peptide mixtures from other HIV-1 proteins (Nef, Tat, Rev and Env) could also boost CTL responses and facilitate the elimination of infected cells (Fig. 2c), and that CTLs pre-stimulated by Gag peptides generally had the highest activity. Together, these results demonstrate that chronically infected patients retain CTL clones that can recognize and eliminate autologous infected CD4+ T cells, despite the presence of CTL escape mutations in dominant epitopes. However, these clones require stimulation with antigen for optimal activity.
To characterize further which CTL population contributed to the elimination of cells infected by CTL escape variants, we compared the killing activity of two specific CTL populations: the population that targets epitopes in which escape has been identified and the one that targets unmutated epitopes (Fig. 3a). CD8+ T cells from patients CP36 and CP39 were pre-stimulated with interleukin (IL)-2 and different synthetic peptides representing the wild-type forms of the relevant epitopes. After incubation for 6 days, each CTL population exhibited significant proliferation compared with no treatment or IL-2 alone (Fig. 3b, c). Pentamer staining for three available epitopes revealed that the number of epitope-specific CD8+ T cells increased dramatically after stimulation with wild-type peptides (Fig. 3b). After co-culture with autologous target cells infected with latent reservoir-derived viruses, CTLs targeting unmutated epitopes clearly showed stronger cytolytic activity than the IL-2 only controls, while CTLs targeting epitopes with identified escaped mutations showed no significant killing (Fig. 3d). CTLs pre-stimulated by the Gag peptide mixture exhibited stronger killing than all single-peptide-stimulated populations (Fig. 3d).
To test whether CTLs that recognize unmutated viral epitopes can inhibit HIV-1 replication and clear infected cells in vivo, we generated patient-derived humanized mice using an improved version of a recently reported mouse system named MISTRG25. Whereas the previously reported MISTRG mice bear a bacterial artificial chromosome (BAC) transgene encoding human SIRP-α, the newly generated MIS(KI)TRG mice harbour a knock-in replacement of the endogenous mouse Sirpa gene with a humanized version. With humanization by knock-in replacement of the Csf1, Csf2, Il3, Tpo and Sirpa genes in the Rag2-/- Il2rg-/- genetic background, MIS(KI)TRG mice are highly permissive for human haematopoiesis and support the reconstitution of robust human lymphoid and myelomonocytic systems. With the demonstrated development of functional T lymphocytes and monocytes/macrophages, MIS(KI)TRG mice provide a useful humanized mouse host for HIV-1 infection studies. Bone marrow biopsies were obtained from study participants and purified CD34+ cells were used to reconstitute the MIS(KI)TRG mice. We infected these patient-derived humanized mice with primary HIV-1 isolates grown from resting CD4+ T cells from the same patient and then evaluated the antiviral effect of autologous CD8+ T cells (Fig. 4a). MIS(KI)TRG mice engrafted with bone marrow CD34+ cells from patient CP18 successfully developed human T-lymphocyte and monocyte/macrophage subsets (Fig. 4b, c), which were sufficient to support HIV-1 infection (Fig. 4d). Plasma HIV-1 RNA levels peaked 20-30 days after infection (Extended Data Fig. 6a). Depletion of CD4+ T cells was clearly evident 12 days after infection in peripheral blood and spleen (Fig. 4e and Extended Data Fig. 6b, c). Cell-associated HIV-1 RNA was detected in both T cells and macrophages/monocytes (Extended Data Fig. 6d). Viral infection was also observed in various tissues in which a large number of memory CD4+ T cells were detected (Extended Data Fig. 7). In control mice or mice that received autologous patient CD8+ T cells pre-stimulated with a peptide representing the unmutated dominant SL9 epitope, levels of plasma HIV-1 RNA and proviral DNA in peripheral blood continued to increase from day 14 to day 29 after infection (Fig. 4f). In sharp contrast, mice that received CD8+ T cells pre-stimulated with unmutated epitopes (Gag mix or WF9) had a significantly lower level of viral replication (Fig. 4g, h). Dramatic decreases in plasma HIV-1 RNA of 100- to 1,000-fold were observed in all three mice that received CD8+ T cells pre-stimulated with the mixture of Gag peptides including dominant and subdominant epitopes. Two of three mice had undetectable levels of plasma HIV-1 RNA and proviral DNA in peripheral blood measured at three time points (Fig. 4g). We performed the same experiments using patient CP36-derived humanized mice and a reduction of peripheral HIV-1 RNA and DNA levels was also observed in mice that received CP36 CD8+ T cells pre-stimulated with the mixture of Gag peptides (Extended Data Fig. 8). Since the post-engraftment lifespan of MIS(KI)TRG mice is only 10-12 weeks25, we were only able to investigate the acute phase of HIV-1 infection and demonstrate the in vivo functionality of patient CD8+ T cells. Future developments of the MIS(KI)TRG model will prolong the post-engraftment lifespan of these mice and allow studies of the establishment and clearance of the HIV-1 latent reservoir in vivo. Together, our in vitro and in vivo experiments demonstrate that only CTL clones targeting unmutated epitopes are effective against cells infected with the viral variants that are likely to represent the major source of rebound HIV-1 after reversal of latency.
The seeding of the HIV-1 latent reservoir starts just a few days after infection26, before the development of a robust CTL response14. This is consistent with our finding that patients who initiated treatment early, in the acute infection stage, have few if any CTL escape variants archived in the latent reservoir. However, if treatment was initiated in chronic infection, CTL escape variants became dominant in the latent reservoir, indicating a complete replacement of the initially established 'wild-type' reservoir. The mechanism behind this replacement warrants further investigation, but probably reflects the dynamic nature of the reservoir in untreated infection. In any event, the overwhelming presence of escape variants in the latent reservoir of chronic patients certainly presents an additional barrier to eradication efforts. The striking difference between AP- and CP-treated patients presents another argument for early treatment of HIV-1 infection; early treatment not only reduces the size of the latent reservoir27, but also alters the composition of the reservoir, as shown here, in a way that may enhance the efficacy of potential CTL-based eradication therapies.
The hierarchy of HIV-1-specific CTL response in acute infection appears to have an important role in initial viral suppression, as demonstrated by the fact that certain immunodominant CTL populations are frequently linked to lower set point viraemia later in infection17, 28. These immunodominant responses in acute infection have been identified as the major selection force driving the development of CTL escape mutations13, 20. Here we show that these immunodominant response-driven mutations are not only archived in the latent reservoir, but also in fact dominate the latent provirus population in CP-treated patients. Therefore, directing CTL responses to unmutated viral epitopes is essential to clear latent HIV-1. Owing to bias in antigen presentation or recognition29, common vaccination strategies will probably re-stimulate immunodominant CTL clones that do not kill infected cells after the reversal of latency. Stimulation of CTL responses with viral peptides circumvents antigen processing and is able to elicit broad-spectrum CTL responses against unmutated regions of viral proteins. Our study suggests that latent HIV-1 can be eliminated in chronically infected patients despite the overwhelming presence of CTL escape variants. Future directions in therapeutic vaccine design need to focus on boosting broad CTL responses, as also reported elsewhere30, and/or manipulating immunodominance.
|
|
| |
| |
|
|
|
