HIV-specific CD8+ T cells and HIV eradication (original) (raw)

There are multiple challenges to harnessing CD8+ T cells to eradicate a reservoir, as outlined above and as evidenced by the inability of these cells to eradicate the reservoir in treated or untreated natural infection. Nevertheless, there are a number of compelling strategies that are worthy of pursuit.

Shock and kill. A promising eradication strategy involves combining LRAs, such as histone deacetylase inhibitors (HDACIs), cytokines, TLR agonists, or others, with CD8+ T cells (or other immune effectors) in order to induce antigen expression from quiescent cells and then eliminate these exposed targets (101). When combined with expanded HIV-specific CD8+ T cell lines, this approach has been shown to drive the elimination of infected cells from a primary cell model of latency and from patient samples in vitro (8, 102). Despite evidence that the administration of certain HDACIs disrupted HIV latency in patients, none of these studies revealed detectable reservoir depletion. One potential explanation for this finding could be that HDACIs impaired CD8+ T cell function in vivo, thereby interfering with the ability of these cells to eliminate exposed target cells. Both panobinostat and romidepsin have been shown to interfere with multiple CD8+ T cell functions, including elimination of HIV-infected cells, when tested in vitro at pharmacologically relevant concentrations (103). Additionally, HDACIs exhibit immunosuppressive activities in animal models of GVHD, experimental autoimmune encephalomyelitis, and other diseases for which they may be of therapeutic benefit (104–106). Encouragingly, ex vivo assessment of CD8+ T cell responses in clinical trials involving HIV-infected participants has thus far shown a lack of detectable impairment following the administration of panobinostat or vorinostat, though increases in CD4+ Treg frequencies were observed (32, 107). The question remains as to whether the degrees of latency reversal observed in these trials were sufficient to expose latently infected cells to immune recognition. If not, then the potential to negatively impact CD8+ T cell function with higher dosing regimens may define an upper limit on the therapeutic windows of these agents. Moreover, as activation through TCR stimulation sensitizes T cells to romidepsin and panobinostat toxicity in vitro (103), it is possible that CD8+ T cells that have been recently boosted by therapeutic vaccination may be preferentially killed by subsequent HDACI treatment. Moving forward, it will be important to continue to assess the potential impact of LRAs on the immune effectors with which they will need to work in concert in order to either mitigate potential interference or capitalize on potential enhancements of immune function. It will also likely be important to combine LRAs with strategies to address the other limitations of CD8+ T cells described above, including epitope escape and the diminished magnitude of responses observed in cART-treated subjects.

Therapeutic immunization. Peripheral blood mononuclear cells (PBMCs) from cART-treated subjects exhibit fairly weak ex vivo CTL-mediated killing of HIV-infected cells. Moreover, cART-associated reductions in viremia skew cells toward a memory phenotype. CD8+ T cell effector activity can be substantially enhanced by short-term expansion with HIV antigens (8, 102, 108). Successful in vitro demonstrations of the shock-and-kill concept have utilized such expanded HIV-specific CD8+ T cell lines (8, 102) and can be replicated in vivo by administering therapeutic vaccines aimed at boosting cellular immunity prior to administering LRAs (reviewed in ref. 109). The issue of immune escape presents an unfortunate complexity, in which — with the exception of subjects whose cART was initiated during their primary infection — immunodominant CD8+ T cell responses that might be preferentially boosted by therapeutic immunization are largely targeted against escaped epitopes and are therefore of no utility to HIV eradication (9). Therapeutic immunization to enhance cure efforts will likely require expanding the breadth of responses to include subdominant epitopes that have not already escaped. A related approach would involve the de novo priming of novel HIV-specific T cell responses that had not been elicited during the untreated infection period. New vaccine technologies, such as peptide-amphiphile vaccines that elicit robust T cell responses to peptides in animal models (110), have the potential to make important contributions to these efforts. Dendritic cells (DCs) are also likely to play critical roles in refocusing the immune response. In vitro studies have demonstrated the ability of DCs to expose and boost CD8+ T cell responses against subdominant autologous variants of HIV epitopes in cART-treated subjects (111). DC vaccines involving the ex vivo manipulation and reinfusion of HIV antigen–loaded DCs have also been shown to boost HIV-specific T cell responses in cART-treated patients, resulting in significantly reduced viral load set points following cART interruption (112). Effective enhancement of CD8+ T cell responses will also almost certainly require augmentation of HIV-specific CD4+ Th cell responses that are both critical for maintaining effective CD8+ T cell function and able to reverse some of the functional defects acquired during prolonged viral exposure (113).

Cell therapy. Ex vivo expansion and reinfusion of antigen-specific T cells has shown tremendous promise as a safe and effective means of augmenting antiviral immunity to CMV and EBV and as a therapeutic modality for cancer (114–116). A limited number of attempts have been made to translate this approach to HIV (117–121). The sole study that infused oligoclonal-expanded natural T cells into HIV-infected patients was performed in the early days of ARV therapy, when suppression was poor and showed a trend toward increased CD4 counts and decreased viremia in the absence of toxicity (120). The strategy of ex vivo expansion and reinfusion of virus-specific CTLs offers a superior measure of control over epitope specificity and functional characteristics that is particularly well suited to focusing responses against nonescaped epitopes (reviewed in refs. 122, 123). Cell therapy additionally offers the intriguing possibility of addressing issues related to compartmentalization, as particular homing profiles can be imprinted on CD8+ T cells by ex vivo culture conditions. For example, expanding T cells in the presence of retinoic acid results in subsequent homing of these T cells to the gut (124). T cell therapy involving the expansion of natural virus-specific responses has an excellent safety record (125), can be performed for approximately $6,000 per patient (126), and can establish populations of long-lived memory cells.

As an alternative to expanding natural HIV-specific T cell responses, cell therapy products can consist of T cells that have been redirected to recognize HIV-infected cells by genetic modification. This can be achieved by transducing cells with either transgenic HIV-specific TCRs or chimeric antigen receptors (CARs). These approaches offer several potential advantages, including the possibility of engineering high-avidity TCRs that may have enhanced abilities to detect viral reservoirs (127) and freedom from MHC-I restriction in the case of CAR T cells (128). However, unlike the expansion of naturally occurring HIV-specific T cell populations, these approaches must also address safety considerations regarding the possibility of unintentional targeting of self-antigens.

Coinhibitory blockade. Coinhibitory receptors, including PD-1, TIM-3, CD160, 2B4, LAG-3, and CTLA-4, play a critical role in the maintenance of exhaustion (49–55). Blockade of these receptors — either alone or in combination — has enhanced T cell function in vitro and viral control in multiple animal models (49–55, 129–132), providing a rationale for testing coinhibitory pathway blockade as an immunotherapeutic strategy in HIV infection. Additional enthusiasm for this approach can be drawn from advances in cancer immunotherapy, in which Abs that block the PD-1 and CTLA-4 pathways have been highly successful and are considered breakthrough drugs in the treatment of solid tumors (133). While treatment with cART results in some level of downregulation of coinhibitory receptor levels in the majority of HIV-infected subjects, these levels do not fully normalize in peripheral blood T cells, and persistent upregulation may be more pronounced in lymphoid tissues (49–54, 56, 57, 134). Thus, therapeutic blockade of coinhibitory pathways represents a promising approach to enhancing the abilities of CD8+ T cells to clear persistent viral reservoirs.

Additional immunotherapeutics. As an alternative or adjunct to blocking inhibitory pathways, CD8+ T cell function can be enhanced by the provision of cytokines or other immunostimulatory agents. IL-15 agonists are of particular interest in this regard, having been shown to enhance CD8+ T cell and NK cell activity in a number of preclinical models (135–139), and the IL-15 superagonist ALT-803 is moving into a clinical trial in cART-treated HIV-infected subjects (ClinicalTrials.gov identifier: NCT02191098). Other immunostimulatory agents include TLR-2 agonists, which reverse CD8+ T cell exhaustion and enhance both tumor- and pathogen-specific T cell responses in vivo (140–142), and agonistic Abs against 41BB or CD40 (143–145), among others.