Current strategies to induce selective killing of HIV‐1‐infected cells

Abstract Although combination antiretroviral therapy (ART) has led to significant HIV‐1 suppression and improvement in immune function, persistent viral reservoirs remain that are refractory to intensified ART. ART poses many challenges such as adherence to drug regimens, the emergence of resistant virus, and cumulative toxicity resulting from long‐term therapy. Moreover, latent HIV‐1 reservoir cells can be stochastically activated to produce viral particles despite effective ART and contribute to the rapid viral rebound that typically occurs within 2 weeks of ART interruption; thus, lifelong ART is required for continued viral suppression. Several strategies have been proposed to address the HIV‐1 reservoir such as reactivation of HIV‐1 transcription using latency reactivating agents with a combination of ART, host immune clearance and HIV‐1‐cytotoxicity to purge the infected cells—a “shock and kill” strategy. However, these approaches do not take into account the multiple transcriptional and translational blocks that contribute to HIV‐1 latency or the complex heterogeneity of the HIV‐1 reservoir, and clinical trials have thus far failed to produce the desired results. Here, we describe alternative strategies being pursued that are designed to kill selectively HIV‐1‐infected cells while sparing uninfected cells in the absence of enhanced humoral or adaptive immune responses.


INTRODUCTION
HIV-1 infects and establishes a productive infection in CD4+ T cells, macrophages, microglia, and hematopoietic stem cells during the acute phase of infection through the integration of replication-competent proviruses into, for the most part, actively expressed genes in the host genome. [1][2][3][4] Despite much research, the precise molecular mechanisms that govern whether an infected cell is destined to become latently (harbors replication-competent virus but does not produce infectious viruses unless activated) or productively infected are unknown. Moreover, the mechanisms of latency may vary both between persons living with HIV-1 (PLWH) and between cells within a single patient. A number of epigenetic modulators have been proposed to be responsible for latency, most of which work to maintain a transcriptionally silent chromatin architecture at the HIV-1 promoter, 5,6 but other studies do not support these findings. 7 These studies suggest that viral transcription still occurs during latency, and that the principal blocks to HIV-1 transcription in vivo are downstream and inhibit proximal elongation, distal transcription, polyadenylation, splicing, and the nuclear export of viral RNAs, which require host cellular machinery working in concert with HIV-1 proteins to overcome. 5,7,8 As an example, circulating resting CD4+ T cells, the largest and best characterized HIV-1 reservoir, have been found to express low levels of matrin 3 (MATR3), an essential cofactor for HIV-1 Rev-mediated RNA export, 9 and express high levels of cellular microRNAs (miR-28, miR-125b, miR-150, miR-223, and miR-382) that bind to the 3′ untranslated region of viral mRNAs, inhibiting their translation. 10 It is generally accepted that during untreated progressive infection, HIV-1 continually replicates in secondary lymphoid tissues. 11 In contrast, there is much debate regarding ongoing viral replication under suppressive antiretroviral therapy (ART). [11][12][13][14][15][16][17][18][19] However, cell-associated RNA transcripts are detectable in PLWH receiving fully suppressive ART that have undetectable replication competent viremia, 7,16,19 and the virus remains abundant in lymphatic tissues where ART concentrations may not reach therapeutic levels, and where HIV-1 appears to infect new cells. 18 Although HIV-1 establishes a productive infection in CD4+ T cells, macrophages, microglia, and hematopoietic stem cells, there is also much debate as to which cells constitute the HIV-1 reservoir-the cell population that allows replication-competent HIV-1 to persist for years despite suppressive ART. Although latent HIV-1-infected resting CD4+ T cells were thought to be the only population to fulfill this role, 20 accumulating evidence suggests that hematopoietic stem and progenitor cells, 21 macrophages, and microglia 22,23 are all also distinct HIV-1 reservoirs. A number of factors are thought to account for the stability and persistence of the HIV-1 reservoir despite effective antiviral immunity and/or ART. 24 These include the viral cytopathic effect, immune clearance, and clonal expansion of latent HIV-1-infected CD4+ T cells. 25,26 The precise mechanism(s) by which latent HIV-1infected cell clonal expansion occurs are not completely understood, but are thought to include general immune activation, antigen driven expansion and contraction of infected CD4+ T cell clones, homeostatic proliferation, and HIV-1 integration site-dependent provirus-driven clonal expansion. 4,11,19,[27][28][29][30][31][32][33] Although ART suppresses HIV-1 replication, improves immune   function, reduces comorbidities, and has led to significant improvements in both longevity and quality of life in PLWH, it does not   completely reverse the significant loss in virus-specific immune functions observed in PLWH, 34 nor does it eliminate the preexisting HIV-1 reservoir. 24 Additionally, ART presents challenges such as adherence to drug regimens, the emergence of resistant viruses, and cumulative ART toxicity. Moreover, the latent cells that remain are a source of viral resurgence upon ART interruption or failure. 17,35,36 Thus, a strategy to cure HIV-1 is an urgent and unmet need. Although cure has been achieved in a few persons, in each of these cases patients underwent allogeneic hematopoietic stem-cell transplantation with HIV-1 resistant CCR5Δ32/Δ32 CD34+ peripheral blood stem cells. 37,38 Although this procedure is logistically not unfeasible for deployment across the almost 38 million PLWH who do not have other conditions that make them candidates for such a procedure, it demonstrates that a cure is possible.

SHOCK AND KILL
The most prominent approach to achieve a HIV-1 cure is the "shock and kill" strategy. This strategy attempts to induce the reactivation  and an autophagy-dependent decrease in both HIV-1 release from infected macrophages and cell-associated HIV-1 capsid abundance in the absence of cell death. 58,59 Current research is now examining the "shock and kill" approach in the context of supporting strategies that include broadly neutralizing or engineered bispecific antibodies, therapeutic vaccines, chimeric antigen receptors, or checkpoint inhibitors.
However, the results from recent trials using this strategy have not been promising. 53,55,60 Fundamentally, these supporting strategies still do not address each of the multiple blocks in viral transcription and translation such that they do not induce outgrowth of latent HIV-1 to elicit immune mediated clearance. Moreover, they do not address the heterogeneous HIV-1 reservoir in various tissue compartments and sanctuary sites, the homeostatic proliferation of latently infected cells, or the presence of noninducible HIV-1. 61 Thus, eradication of the HIV-1 reservoir using a "shock and kill" strategy might be very difficult to achieve. 55

Apoptosis
There are 2 general pathways towards apoptosis, the extrinsic pathway and the intrinsic pathway ( Figure 2 The initiation of apoptosis is a multistep process that can be either extrinsic or intrinsic. In the extrinsic pathway, cognate ligand binding to TNFRSF members can either promote the direct recruitment of TRADD (as is the case for TNFRSF25 or TNFR1) or FADD (as is the case for FAS and TNFRSF10A [TRAILR1]). In the case of TRADD recruitment, this either promotes the recruitment of FADD leading to a caspase signaling cascade resulting in the activation of caspase-3 and apoptosis, or TRADD recruits RIPK1 and TRAF2, triggering NF-κB activation, cell survival and a proinflammatory response. In both cases, the extrinsic pathway can be either mitochondria independent (type I death receptor pathway) or converge with the intrinsic pathway and be mitochondria dependent (type II death receptor pathway) when BID is activated by caspase-8, which then oligomerizes BAK1 leading to mitochondrial outer membrane permeabilization (MOMP) and the release of cytochrome c from the mitochondrial intermembrane space. The key event in the intrinsic pathway is the formation of the apoptosome and subsequent activation of caspase-9 after MOMP. Also released during MOMP are DIABLO/SMAC, which promotes apoptosis indirectly by inhibiting IAPs, and AIF, which promotes parthanatos, a caspase-independent form of cell death (not shown) autoubiquitination and proteasomal degradation. 84,85 The latter prevents the BIRC2-and BIRC3-mediated ubiquitination of RIPK1, which results in switching the TNFR superfamily signaling from prosurvival to proapoptotic through the formation of a FADD-caspase-8-containing complex that leads to caspase-8 activation. 86 Moreover, IAPs inhibit FAS ligand-induced cell death by limiting RIPK1 recruitment to FAS. 80 Apoptosis-inducing factor (AIF) is also released from the mitochondria, which can stimulate parthanatos, a caspase independent form of regulated cell death.

Macroautophagy
Macroautophagy is 1 of 3 principal types of autophagy (the other 2 being microautophagy and chaperone-mediated autophagy), an evolutionarily conserved major catabolic degradative process that occurs in all eukaryotic cells and is constitutively functional at low levels.
Autophagy is up-regulated under conditions of cellular stress including nutrient deprivation and infection to recycle cytoplasmic components to generate biologic macromolecule monomers and energy,

F I G U R E 3
Regulation of autophagy. The initiation of autophagy is a multistep process, the main regulators of which are MTOR, an inhibitor, and AMPK an activator. MTORC1 inhibition drives the formation of the phagophore through the formation of the ULK1 complex that directly activates the PIK3C3 complex. This complex translocates to endoplasmic reticulum sites and produces PI3P, which recruits WIPI2 and ZFYVE1. In the ATG12 conjugation system, ATG12 is covalently attached to ATG5, which is then attached to ATG16L1. WIPI2b acts immediately upstream of ATG16L1 and recruits ATG12-ATG5-ATG16L1 to PI3P-tagged phagophores. The ATG12-ATG5-ATG16L1 complex then promotes conjugation of ATG8 proteins with phosphatidylethanolamine leading to their incorporation into the phagophore membranes where they interact with cargo receptors harboring LC3-interacting motifs that recruit and incorporate ubiquitin-decorated cargoes into the nascent autophagosome. After detachment of ATG factors, the phagophore closes through scission forming the autophagosome, which then fuse with lysosomes resulting in the degradation of the engulfed components and to remove unneeded and/or damaged organelles and protein complexes to maintain cellular homeostasis and survival, as well as the degradation and elimination of toxic proteins and invasive microorganisms to control infection and inflammation. In addition, autophagy plays a critical role in thymic selection, antigen presentation, cytokine production, lymphocyte homeostasis and survival, and immunometabolism. 87 In macroautophagy, a dynamic complex signaling pathway initiates the deployment of proteins, lipids, and membranes to form a double-membrane sequestering phagophore on membrane sites predominately located on ER (other membrane sites have also been implicated). 88 Unlike the formation of secretory vesicles, phagophore formation, and by extension autophagosome biogenesis, is thought to be a de novo process, although there is still some debate on this point. 89  In mammalian cells, selectivity is often conferred by cargo receptor proteins such as calcium binding and coiled-coil domain 2 (NDP52), NBR1 autophagy cargo receptor, optineurin, or SQSTM1 (p62), which act as a link to tether specific cargo to a nascent autophagosome, 90 or a resident protein expressed on the cargo surface such as BCL2L13 (MIL1) on the surface of mitochondria. 91 The sequestering phagophore then continues to expand and sequester cargo before closing and forming an autophagosome through membrane scission. 92 Once formed, the mature autophagosome then fuses with the membrane of lysosomes, forming autolysosomes wherein the autophagic cargo are degraded and/or processed by lysosomal hydrolases (Figure 3).

Autosis
Autosis is a form of autophagy-dependent cell death that is dependent upon Na+,K+-ATPase. This is a nonapoptotic, nonnecrotic form of cell death characterized by an increase in autophagosome, autolysosome, and empty vacuole numbers that coincides with the dilation and fragmentation of the ER. This is quickly followed by nuclear membrane convolution, focal ballooning of the perinuclear space, the depletion of ER, autophagosome and autolysosome numbers, swelling of mitochondria, and the focal rupture of the plasma membrane. 93 The molecular processes that lead to autosis are currently outwith our ken.

Prime, shock, and kill
A number of compounds that target cell death pathways including BCL2 antagonists, PI3K inhibitors, DDX3 inhibitors, polo like kinase 1 inhibitors, and genetic manipulation of long noncoding RNAs with and without an additional LRA have been assessed for their potential to specifically kill reactivated HIV-1-infected cells (Figures 1(B) and 1(C)). 56,[94][95][96][97][98] Notably, these approaches all induce selective cell microglia cell line while also restricting viral production. 101 However, since sodium nitroprusside is a known stimulator of TNF and NO production and has excitotoxicity within the brain parenchyma, this approach could do more harm than good within the central nervous system.

Targeting TREM1
TREM1 is a 30 kDa IgV glycoprotein expressed on the surface of macrophages, microglia, and neutrophils and is up-regulated during HIV-1 infection. 66

IAP inhibition
Although TREM1 is highly expressed in HIV-1-infected macrophages, it is poorly expressed in latent HIV-1-infected CD4+ T cells. Conversely, both in vitro latent HIV-1-infected CD4+ T cells and macrophages have increased expression of IAPs, which inhibit HIV-1 transcription, 105,106 reduce autophagy through the ubiquitination of BECN1 78 and MDM2 proto-oncogene (MDM2), 107 inhibit apoptosis through the antagonization, ubiquitination, and neddylation of apoptosis caspases, 108,109 and increase the ubiquitination of RIPK1. 78

CONCLUDING REMARKS
The development of a strategy that selectively kills latent HIV-1-infected cells while also reducing replication, transmission, and immune hyperactivation is the goal of HIV-1 cure research. The "shock and kill" approaches have received the most attention, but to date their results are disappointing. However, the "prime, shock, and kill" strategy and the selective killing strategies described above still suffer from the same drawbacks as the "shock and kill" strategies. Namely, they do not address the heterogeneousness of the latent reservoir in various tissue compartments and sanctuary sites that have suboptimal drug penetration, 130,131 nor do they address the homeostatic proliferation of these reservoir cells, or the presence of noninducible HIV-1 reservoirs. 61 As HIV-1 infection is highly complex and involves multiple cellular and tissue compartments, it is unlikely that a single agent approach will be successful. Despite this, the targeted killing of HIV-1 reservoir cells is attractive, as it does not depend on the elicitation of a secondary immune response.
A number of important issues still need to be addressed through well-designed preclinical and clinical trials. These include demonstrating efficacy and safety in vivo, defining the relationship between these multidrug regimens and residual viremia in tissues, including the CNS, assessing the ability of these drugs to penetrate the blood-brain barrier, assessing their effect on immune activation and/or chronic inflammation as well as off-target effects, and whether escape mutations will evolve to avoid immunologic control. Despite these limitations, it is likely that targeted killing of HIV-1-infected cells will be involved in an HIV-1 cure.

DISCLOSURE
The authors declare that they have no commercial or financial relationships that could be construed as a potential conflict of interest.