The TRAIL: TRAILshort Axis in HIV Immunopathology

Abstract

Accelerated loss of HIV-infected and uninfected CD4 T cells is a hallmark of HIV infection that leads to severe immunodeficiency, rendering the host susceptible to opportunistic infections and malignancies. Obstacles to eradicating HIV involve the virus’s ability to remain in a quiescent state as latent viral reservoirs and manipulate host defenses to benefit viral survival and persistence of the infected reservoir. Several mechanisms cause CD4 T-cell depletion and recent studies demonstrate the role of tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) in this process. Expression of TRAIL and its receptors is upregulated in response to HIV infection. TRAIL interacts with its receptors to activate apoptotic pathways. We recently demonstrated the presence of TRAILshort, a novel splice variant of full-length TRAIL, in the serum of HIV-infected patients. A unique carboxy-terminus allows TRAILshort to bind to death receptors without inducing apoptosis and prevents TRAIL from binding to its receptors, thereby conferring resistance to TRAIL-mediated death. In this review, we describe how the TRAIL: TRAILshort receptor axis modulates apoptosis of different types of immune cells in the context of HIV infection. We also discuss how TRAIL and TRAILshort contribute to the activation of immune cells involved in host defense against HIV and mechanisms that HIV has evolved to manipulate TRAIL for its survival.

I. INTRODUCTION

Finding a cure for HIV infection has been a goal of many infectious disease clinicians and researchers for decades. Obstacles to identifying such a cure, however, are numerous (Fig. 1). First, HIV has developed multiple mechanisms for evading the host immune system in order to benefit the virus and its existence as it causes devastating immunologic effects. Another obstacle to developing a cure is the ability of HIV to establish a latent viral reservoir with no clear distinguishing features between the cells that compose this reservoir and uninfected cells. The viral reservoir contains transcriptionally silent yet replication-competent HIV and these cells can persist for years, reestablishing viremia and the physiological consequences of viremia when anti-retroviral therapy (ART) is discontinued.¹

FIG. 1:

The IFN-α immune response in HIV infection: When a cell is infected with HIV, the pattern recognition receptors (PPRs) such as TLR7 and TLR9 are activated and signal through the myeloid differentiation primary response 88 (MYD88), and in parallel IFI-16 and cyclic GMP-AMP (cGAMP) synthase (cGAS) are activated and signal through stimulator of IFN genes (STING) to phosphorylate and induce homo-oligomerization of IRF-7 and IRF-3, which translocate to the nucleus and bind to the IFN regulatory binding site (IRFBS) and stimulate the production of IFN. IFN is secreted and binds to the IFN receptors IFNR-1 and IFNR-2, activating of the downstream JAK/STAT signaling. The antiviral defense activity of IFN is mediated through the many ISGs, including tetherin, SAMHD-1, APOBEC-3, and TRAIL. The virus has developed many mechanisms to block these protective ISGs. Vpu blocks tetherin, Vpx blocks SAMHD1, and Vif blocks APOBEC-3. The production of TRAILshort blocks the function of TRAIL, thereby preventing the apoptosis of the infected cells.

Members of the tumor necrosis factor (TNF) superfamily of cytokines regulate cell death and play a critical role in maintaining a functional and healthy immune system. In addition to TNF and CD95 (FAS/Apo-1), the TNF-related apoptosis-inducing ligand (TRAIL/APO-2L) is a member of this family that signals through the extrinsic pathway of apoptosis mediated by death receptors.² TRAIL binds to one of the two death-inducing receptors, TRAIL-R1 or TRAIL-R2, initiating the death-signaling cascade that ultimately leads to apoptosis.² TRAIL also can bind to three other receptors: TRAIL-R3, TRAIL-R4, and osteoprotegerin; however, binding to these receptors does not induce cell death.² TRAIL is normally expressed by numerous immune cells, including natural killer cells (NK), NK T cells, CD8 T cells, dendritic cells, and macrophages,³ and thus has been implicated in immune surveillance against malignant transformation and/or infectious diseases. Its role in the immunopathology of HIV infection is important and has been studied extensively.⁴

Our studies have added a new dimension to understanding the relationship between TRAIL and HIV infection. The discovery of TRAILshort, a TRAIL antagonist, in the serum of HIV-infected patients confounds understanding of the role of TRAIL in HIV pathogenesis.⁵ TRAILshort is a splice variant of full-length TRAIL that lacks exons 3 and 4. The splicing event introduces a frame shift in exon 5, resulting in a novel 11-amino acid carboxy-terminus. TRAILshort preferentially binds to TRAIL-R1 and TRAIL-R2 and in the process prevents full-length TRAIL from binding to TRAIL receptors to activate the apoptotic cascade.⁵ Importantly, TRAILshort lacks the zinc-binding cysteine at position 230.⁶ Although we first discovered this isoform only in cells from HIV-infected patients, we now know that TRAILshort is produced by both HIV-infected and uninfected cells, raising the possibility that TRAILshort might also play a role in diseases other than HIV.⁷ HIV-infected cells were found to produce TRAILshort as a mechanism to establish the virus’s own persistence.⁵ In this review, we discuss the role of the TRAIL: TRAILshort axis in the immunological dysfunction associated with HIV infection.

A. Role of Apoptosis in Human Biology and Disease

Apoptosis is an essential physiological process that involves the predetermined removal of cells during development or disease. Apoptosis occurs during embryogenesis to shape the organs and continues throughout growth and adult development to maintain homeostasis by balancing cell turnover against the regeneration of billions of cells daily as needed to maintain healthy cell populations in tissues.⁸ Apoptosis also eliminates dysfunctional cells such as defective synaptic connections during neuronal development or autoreactive cells during immune maturation.⁸ Apoptosis is an immune effector response against infection, disease, or noxious agents. Additionally, apoptosis kills compromised cells without creating debris, thereby preventing undesirable autoimmune responses.⁹

Various mechanisms have evolved to ensure tight control of apoptosis at different levels. Cellular damage triggers death receptors such as Fas and TNF receptors to oligomerize and recruit a multi-protein complex called the death-inducing signaling complex to regulate the expression of specific genes, further activating the apoptotic effector molecules such as caspases, which execute cell death via extrinsic apoptotic pathways. Inhibitor of apoptosis proteins (IAPs) suppress the apoptotic pathway; the B-cell lymphoma-2 (Bcl-2) family, encompassing over 20 pro-apoptotic and anti-apoptotic molecules, alters the intrinsic mitochondrial pathway to modulate apoptosis by stimulating or inhibiting the release of mitochondrial content.¹⁰

In order for a cell to survive in an environment optimal for its growth and functionality, it must constantly survey its surroundings to detect signals of danger and initiate programmed cell death when necessary.¹¹ At the same time, production of anti-apoptotic signals to prevent unwanted cell death is another means of maintaining a healthy population of cells.¹² Several groups have shown that the outcome of a cell’s exposure to varying degrees of sublethal doses of stress differs depending on the expression of anti-apoptotic proteins, including the Bcl-2 family, the FLICE-inhibitory proteins, and the IAPs.¹²

The pathophysiology of many diseases is explained by an imbalance in apoptosis (Table 1).¹⁰ Defective or insufficient apoptosis is one of the hallmarks of carcinogenesis.¹³ Many anti-apoptotic/pro-apoptotic proteins are key to the transformation of a normal cell into a cancerous cell. Overexpression of anti-apoptotic proteins such as Bcl-2 and Survivin is implicated in the development and progression of many tumors.¹⁴ Conversely, reduced levels of B-cell lymphoma-associated X protein (BAX), a pro-apoptotic protein, cause loss of apoptosis and accelerate tumorigenesis in B-cell lymphomas.¹⁵ Excessive Bcl-2 expression, however, is associated with a less malignant phenotype and a more favorable outcome in breast cancer and gastric carcinoma.^{16, 17}

TABLE 1:

Diseases involving impaired apoptosis

Disease	Type of Impaired Apoptosis
Cancer	Reduced apoptosis leading to accumulation of aberrant cells
Autoimmune disorders	Impaired apoptosis of autoreactive immune cells
Viral and bacterial infections	Decreased apoptosis of some virally infected cells leads to persistence, as in hepatitis B virus or HIV infection Excessive apoptosis causing viral clearance and inflammation, as in West Nile virus, Helicobacter pylori infection, or bacterial meningitis
Atherosclerosis	Reduced apoptosis leading to accumulation of vascular smooth muscle and endothelial cells inside the blood vessels
Degenerative disorders	Excessive apoptosis of functional neurons, as in Alzheimer’s disease, Parkinson’s disease, or amyotrophic lateral sclerosis
Cardiovascular diseases	Increased apoptosis during stroke, myocardial infarction, or coronary artery disease

Autoimmune lymphoproliferative syndrome is another example of a disease characterized by defective apoptosis of T cells and proliferation of B cells, producing self-antibodies.¹⁸ Likewise, extensive apoptosis is the underlying mechanism of several diseases. Neurodegenerative disorders such as Alzheimer’s disease, Parkinsonism, and Huntington’s disease are associated with aberrant and excessive apoptosis.¹⁹ Cell death also plays a role in the pathogenesis of ischemic injury of the heart and the brain.²⁰

Viruses, obligate intracellular parasites, have developed a variety of mechanisms to prevent cell death and to ensure their own persistence. However, many viruses are cytotoxic and death of the virally infected cell releases the virions to infect new cells. The death of large number of CD4 cells during HIV infection and the resulting immunodeficiency explains the underlying pathology of infection.²¹ Although both HIV-infected and uninfected cells are targeted for HIV-induced apoptosis, death of uninfected cells is more extensive.²¹ Upon infection, HIV expresses proteins such as Gp120, Tat, Nef, viral protein U (Vpu), and Vpr, which enhance apoptosis through increased expression of FAS, FASL, TNF-alpha (TNF-α), and the TRAIL receptors death receptor 4 (DR-4) and DR-5 and reduced expression of anti-apoptotic Bcl-2 and Bcl-XL.⁴ We have shown that HIV protease causes apoptosis in virally infected cells through the generation of Casp8p41, which binds and activates BAK, leading to cell death.²² Other viruses such as herpesviruses, cytomegalovirus, adenoviruses, and pox viruses also produce Bcl-2-homologous proteins that have an anti-apoptotic effect.²³ Hepatitis B virus can induce apoptosis through its large and middle surface proteins, hepatitis B X protein and hepatitis B spliced protein,²⁴ ultimately leading to fulminant hepatitis and liver failure. Conversely, hepatitis C virus has been shown to induce the surface expression of both TRAIL and FASL.²⁵

B. Biology of Apoptosis and Therapeutic Opportunities

Altering apoptotic pathways has been proposed as therapy for diseases including cancer and viral infection. Studies showing that tumor cells expressing high levels of Bcl-2 are chemoresistant led to the hypothesis that targeting prosurvival molecules might increase sensitivity to other chemotherapeutic agents or that they may lead to cell death on their own.²⁶ BH-3-mimetic compounds were developed to overcome Bcl-2’s anti-apoptotic effects.²⁶ ABT-737 was the first drug in this class, which binds and inhibits Bcl-2, Bcl-XL, and Bcl-W in a manner similar to BH3. This compound was shown to be effective in killing tumor cells both in vivo and in vitro as a stand-alone agent and also in combination with other chemotherapeutics.²⁷ ABT-263 (navitoclax), which has a similar mechanism of action, was used successfully in a clinical trial to reduce the tumor burden in patients with chronic lymphocytic leukemia.²⁸ However, the severe thrombocytopenia that developed from inhibition of Bcl-XL led to the use of more specific Bcl-2 inhibitors.²⁹ ABT-199 (venetoclax) is currently used to treat refractory chronic lymphocytic leukemia; this compound also is in clinical trials alone and in combination with other agents to treat acute myeloid leukemia, acute lymphocytic leukemia, and many B-cell lymphomas.²⁹ Congruent with this, we postulated that Bcl2 antagonism might favor HIV-infected cell death induced by Casp8p41 following HIV reactivation from latency.

Indeed, we have shown that central memory T cells express high levels of Bcl-2³⁰ and that, when HIV reactivates from latency, Casp8p41 is produced. Therefore, when CD4 cells were isolated from ART-suppressed HIV patients; priming these cells using venetoclax followed by HIV reactivation ex vivo significantly reduced cell-associated HIV DNA compared with HIV reactivation in the absence of venetoclax.³⁰ In another study using a model of homeostatic proliferation of HIV and acute HIV infection, venetoclax caused preferential killing of HIV-expressing cells; furthermore, during new rounds of infection, it selectively induced apoptosis in the HIV-infected cells.³¹ This observation raises the possibility of using a selective Bcl-2 antagonist either alone or in combination with other drugs as an approach to curing HIV.

Some HIV-infected CD4 T cells harbor replication-competent integrated provirus and convert into resting memory T cells. These resting memory cells resist apoptosis, have long half-lives, and provide a means for HIV to persist. Importantly, when ART is discontinued, HIV rebounds from the central memory reservoir, so the presence of these cells presents a challenge to HIV eradication.^32,33 One approach that has shown promise in preclinical studies relates to TRAIL, which enhances the apoptosis of T cells, including those that harbor latent viral DNA in vitro, resulting in significant reduction in the amount of replication-competent HIV and integrated provirus in resting memory cells.³²

More than a decade ago, TRAIL agonists were evaluated as a therapy to selectively kill cancer cells without much success.³⁴ Several studies tried to explain why many tumors are either TRAIL resistant or develop resistance shortly after treatment.³⁴ Some have suggested evaluating the level of TRAIL receptors in tumor biopsies prior to starting treatment.³⁵ Although this might seem simple and logical, it is complicated by multiple factors, including the following.

First, there is a lack of knowledge regarding the contributions of the individual TRAIL receptors DR4 and DR5 to apoptosis in a given cancer.³⁵ Some studies have suggested that DR4 is more predominant in acute myeloid leukemia, chronic lymphocytic leukemia, and pancreatic cancers, whereas DR5 is more effective in epithelial cancers.³⁶ These studies were performed in vitro and many cancers known to be TRAIL sensitive have not been tested. Also, these studies were performed more than a decade ago using the TRAIL agonists or agonistic antibodies available at the time, which may have had poor affinity.³⁵

Second, others have noted that the level of expression of the death-inducing TRAIL receptors is of limited use but emphasize the relative expression of the apoptotic TRAIL-R1 and TRAIL-R2 versus the non-apoptotic TRAIL-R3 and TRAIL-R4 in different tumor types.³⁴ The hypothesis supposes that tumors with higher expression of the decoy receptors are more TRAIL resistant.

Finally, TRAIL signaling is extremely complicated and TRAIL receptors represent only one component of this signaling cascade.

The recent description of TRAILshort complicates the understanding of the TRAIL: TRAIL receptor axis; we believe that the expression of TRAILshort may be an important factor in determining resistance to TRAIL treatment (Fig. 2). This novel isoform was not considered in previous clinical trials. By blocking TRAILshort, we can overcome TRAIL resistance.⁷ Because TRAILshort is a secreted protein that can be detected in a patient’s serum, measuring the levels of TRAILshort in the plasma would be a simple and noninvasive technique to evaluate patients’ response to TRAIL treatment. Furthermore, the beneficial effects of antagonizing TRAILshort might enhance the innate immune response and increase the cytotoxic capacity of the patient’s NK cells. TRAILshort antagonism could be used in combination with other immunotherapeutic agents such as programmed cell death protein-1 (PD-1) antibodies to produce a synergistic effect.

FIG. 2:

TRAILshort negatively affects the immune response to HIV infection: TRAILshort is a splice variant of full-length TRAIL that lacks exons 3 and 4. The splicing event introduces a frame shift in exon 5, resulting in a novel 11-amino acid carboxy-terminus. The production of TRAIL short will reduce the apoptosis in the infected CD4 cells and decrease the cytotoxic capacity of NK cells and CD8 T cells, creating a favorable environment for HIV replication.

C. HIV and TRAIL: It’s Complicated

1. Involvement of TRAIL in T- and B-Cell Apoptosis

Shortly after the discovery and cloning of TRAIL, it was established that it plays a role in the activation-induced cell death of CD4 cells from HIV-infected patients.³⁷ Activation-induced cell death is defined as increased susceptibility of activated lymphocytes to death stimuli under the conditions of inappropriate immune activation.³⁷ CD4 T cells isolated from HIV-infected patients and activated through their T-cell receptor proliferate and are more sensitive to cell death by both TRAIL and Fas ligation.³⁷ Following this discovery, TRAIL expression was shown to be increased on the surface of CD4 T cells containing HIV compared with uninfected cells from HIV-infected patients.³⁸ One study showed that plasmacytoid dendritic cells (pDCs) collected from HIV-infected subjects after interruption of ART showed elevated levels of TRAIL surface expression. These cells were able to induce cell death in the HIV-infected cell line Sup-T1, yet they were unable to kill autologous infected CD4 cells.³⁹ The investigators explained this observation by the lack of sustained high levels of interferon-alpha (IFN-α) in chronic HIV-1 infection.³⁹ Another group suggested that TRAIL expression is upregulated on the surface of pDCs from viremic patients and correlates with viral load.³⁸ They also suggested that TRAIL expression on the surface of these cells contributes to CD4 T-cell depletion.³⁸ These pDCs killed only activated T cells and lacked any significant effect on naive T cells.³⁸ Loss of memory B cells in chronic HIV infection was shown to be mediated by TRAIL-induced apoptosis and the expression of FOXO3.⁴⁰ Studies have shown that loss of memory B cells during HIV infection causes impairment of long-term serological memory to non-HIV antigens such as measles and Streptococcus pneumoniae, thereby contributing to the diminished response to vaccines observed in these individuals (Fig. 2).⁴¹

2. TRAIL and T-Cell Activation and Priming

In addition to inducing apoptosis, TRAIL has been also shown to block antigen-specific T-cell activation.⁴² This non-apoptotic role for TRAIL was found to be independent from that of antigen-presenting cells and was attributed to the suppression of calcium-dependent lymphocyte activation.⁴² Furthermore, another study showed that exposure of pDCs to HIV-1 in vitro led activated T cells to inhibit the priming of naive T cells.⁴³ The investigators attributed this inhibition to the upregulation of TRAIL, PD-1, and cytotoxic T-lymphocyte associated protein 4 (CTLA-4) on the surface of the T cells, creating a “suppressor” T-cell phenotype.⁴³

3. HIV Resistance to TRAIL Treatment

Elevated surface expression of TRAIL receptors on HIV-infected macrophages encouraged researchers to use recombinant TRAIL to treat HIV-infected cells in vitro. The experiments demonstrated the ability of TRAIL to induce apoptosis in HIV-infected peripheral blood lymphocytes and monocyte-derived macrophages.⁴⁴ However, the concentrations of the leucine zipper TRAIL used in these in vitro studies would be impossible to replicate in vivo without producing significant cytotoxicity. HIV also has developed mechanisms to subvert the TRAIL-mediated apoptosis of the infected cells. Swingler et al. showed that HIV-infected macrophages acquire resistance to TRAIL-induced apoptosis by reducing levels of TRAIL receptors and increasing levels of anti-apoptotic proteins such as BFl1 and Mcl1 under the influence of HIV glycoprotein.⁴⁵ Results from us and others also demonstrated a role for TRAIL-mediated killing in the apoptosis of hepatocytes during HIV/hepatitis C virus coinfection.^{25, 46} Although increased serum levels of TRAIL could be detected in the serum of HIV-infected patients, initiation of ART reduced but did not normalize them.⁴⁶ Particularly, DR5/TRAIL-R2 expression was increased on the surface of T cells in the lymphoid tissue of HIV-infected persons. Successful ART treatment reduced the expression of TRAIL-R2.^46,47 The increase in the level of DR5 expression can be caused by Gp120 alone, although other in vivo signals are likely too.⁴⁸

D. Effects of the TRAIL: TRAILshort Axis on HIV Immune Response

1. TRAIL: TRAILshort and IFN-α

When IFN was first discovered in 1957, it was thought to be the cure for all viral infections and cancer because it could result in death of virally infected or cancerous cells. However, with time and extensive research, the controversy around this vital cytokine has increased.

Type 1 IFNs have 12 different subtypes that produce variable biological effects in the cell.⁴⁹ During acute HIV infection, IFNs are produced in response to pathogen-associated molecular patterns. Viral pathogen-associated molecular patterns are sensed by numerous pattern recognition receptors such as Toll-like receptors (TLRs) and retinoic acid inducible gene 1 (RIG-1) or by intracellular pattern recognition receptors such as IFN-gamma (IFN-γ)-inducible protein 16 (IFI-16) and cyclic GMP–AMP synthase. Pathogen recognition receptors play an essential role in developing an innate immune response to the virus, leading to IFN production.^50,51 In the acute phase of the infection, type 1 IFN is produced mainly by pDCs in the mucosa.⁵² Type 1 IFNs then bind to its surface receptors, IFNAR-1 and IFNAR-2.

Subsequent phosphorylation of Janus kinase 1 (JAK-1) and tyrosine kinase 2 (TYK-2)⁵² leads to the recruitment, phosphorylation, dimerization, and nuclear translocation of multiple signal transducer and activator transcription (STAT) proteins.⁵¹ Phosphorylated STAT-1 and STAT-2 and IFN-regulatory factor 9 (IRF-9) form the IFN-stimulated gene factor 3 (ISGF-3) complex, which binds to the IFN-stimulated response element (ISRE) sequence in the nucleus to induce hundreds of IFN-stimulated genes (ISGs) that have multiple roles in viral restriction and antiviral defense (Fig. 1).⁴⁹

Apolipoprotein B mRNA editing enzyme catalytic-like 3 (APOBEC-3) is one of the important ISGs induced by acute HIV infection. APOBEC-3 induces mutations in the viral genome, allowing insertion of an missense or stop mutations that impair or block viral replication.⁵¹ Sterile alpha motif (SAM) domain- and HD domain-containing protein 1 (SAMHD-1) limits HIV reverse transcription through converting intracellular deoxyribose containing nucleoside triphosphates (dNTPs) to deoxyribonucleosides (DNs) and inorganic triphosphates.⁵¹ Tetherin (CD137), another IFN-inducible ISG, is inserted in the cell membrane of virally infected cells to prevent budding of new virions.⁵¹

The list of ISGs induced by HIV infection and its antiviral effects is long and beyond the scope of this review; however, it is significant that HIV has evolved different counter-mechanisms against many of these restricting factors. The HIV-1 accessory protein viral infectivity factor (Vif) degrades APOBEC-3 and the viral protein X (Vpx) degrades SAMHD1 by directing them to the proteasome.⁵³ Vpu downregulates tetherin expression on the surface of virus-producing cells.⁵³ It seems that the virus has evolved to block the action of many of these ISGs to facilitate its own replication, so the IFN-α response, while beneficial, is inefficient in preventing the development of chronic infection (Fig. 1).

During chronic infection, the source of IFN is less clear, although all nucleated cells can produce IFN.⁴⁹ The induction of high levels of IFN during persistent HIV infection may function for the benefit of the virus itself and become a pathogenic mechanism of disease progression. Several studies have shown the detrimental effects of chronic immune activation on viral control and inhibition of replication.⁴⁹ High levels of type 1 IFNs enhance the activation-induced proliferation (via T-cell receptor stimulation) but inhibit the homoeostatic proliferation (interleukin-7 [IL-7]-induced) of T-cells, which in turn causes increased T-cell death.⁵⁴ Thymic dysfunction, another consequence of high IFN, limits CD4 T-cell recovery.⁵⁴ A state of persistent IFN production also was shown to stimulate programmed death ligand 1 (PDL-1), a marker of immune exhaustion on the surface of HIV-infected pDCs.⁵⁵

TRAIL is another of the well-known ISGs; we and others have shown that IFN stimulation induces TRAIL production at both the RNA and protein level.⁵⁶ With this in mind, we wondered what drives TRAILshort production. After treatment of uninfected peripheral blood mononuclear cells with a variety of cytokines known to induce TRAI, as early as 24 hours after treatment, we observed a significant increase in the mRNA of TRAILshort in response to type 1 IFNs, in particular IFN-α14 (Fig. 1).⁷

Despite the controversy associated with the effects of IFN on HIV infection, many have attempted to use type 1 IFNs as a treatment for HIV.⁴⁹ Monotherapy using PEGylated IFN-α2a reduces viral replication and cell-associated integrated HIV DNA.⁵⁷ IFN-α treatment also delays viral rebound after cessation of therapy in simian immunodeficiency virus (SIV)-infected macaques.⁵⁸ Another study in macaques reported that IFN-α treatment suppresses viral replication during the acute phase of the infection. Cumulatively, these findings suggest that IFN-α2a displays significant anti-HIV effects. As with other biological treatments, however, continuous treatment with IFN caused desensitization and subsequently increased the size of the viral reservoir.⁵⁹ In more recent studies, certain subtypes of IFN, such as IFN-α14 and IFN-α11, showed superior antiviral effects with lesser side effects compared with IFN-α2a.^50,60 Interestingly, recent studies using humanized mouse models have shown that, during the chronic phase of HIV infection, blockade of IFN receptor reduces the level of T-cell activation, reduces expression of receptors PD-1 and TIM-3, and improves IFN-γ production by CD8⁺ T cells.^61,62 This blockade also reduced levels of HIV replication and remarkably reduced the size of the viral reservoir.^61,62 Although the investigators did not suggest a mechanism for these observations, we speculate, based on our research, that blocking IFN signaling may lead to a reduction in the levels of TRAILshort, which may affect the results.

2. TRAIL: TRAILshort and NK Cells

High levels of IFN-α produced during acute HIV infection leads to a significant increase in the number of activated NK cells. These cytotoxic cells produce their effects through the delivery of granzyme B after the formation of the immunological synapse and also through the expression of death ligands such as FASL and TRAIL on the surface of the NK cells.⁶³ IFN-α has been shown to increase the cytotoxicity of NK cells and to increase their surface expression of TRAIL.⁶³ The initial expansion of activated NK cells is followed by the accumulation of NK cells with markedly impaired function. NK function is very delicately organized and the ability of these cells to produce their cytotoxic effects depends on a balance between numerous activating and inhibitory receptors.⁶⁴ Killer immunoglobulin receptors are inhibitory receptors expressed on the surface of NK cells; the interaction between these receptors and the corresponding human leukocyte antigen (HLA) ligands from the host are extremely important in determining the activation status of the NK cells.⁶⁵ NKp46 is an activating natural cytotoxicity receptor that plays a vital role in controlling HIV. Another activating receptor, NKG2D, is found in the early stages of infection; HIV-infected cells express the NKG2D ligand, which targets them for cytotoxic lysis by activated NK cells.⁶⁵ Interestingly, it has been shown that IFN-activated NK cells can cause lysis of autologous HIV-infected CD4 cells,^63,64 which is mediated through NKp46 and NKG2D. In fact, high expression of NKG2D is associated with superior control of HIV.⁶⁴

TRAIL expression and associated cytotoxic function was first thought to be limited to the early stages of development of NK cells.⁶⁶ Other groups then described the relevance and significance of TRAIL-expressing NK cells in the liver in destroying self-hepatocytes.⁶⁶ Recently, two studies demonstrated that, in mice, surface expression of TRAIL on NK cells is dependent on mNKp46.^66,67 In the context of viral infection, one group has shown that, during chronic cytomegalovirus infection, TRAIL-expressing NK cells accumulate in the salivary gland.⁶⁸ These cells caused the depletion of activated CD4 T cells, which promotes viral persistence but also prevents the development of autoimmune disease.⁶⁸ Another study has shown that apoptosis of primary hepatocytes from the livers of hepatitis B virus-infected patients, but not healthy controls, was partially TRAIL dependent.⁶⁹ More recently, two distinct subtypes, tissue-resident and tumor-associated NK cells, have been shown to express high levels of TRAIL on their surface with low levels of perforin and granzyme.⁶⁷ Unlike circulating NK cells, which are defined by the presence of CD56, identification of these NK cells is dependent on the expression of a number of surface receptors, namely NKp46, Tbet, Eomes, CD49a, CD49b, CD103, and CD69.⁶⁷

In terms of the cytokines necessary for the control of NK development, function, and homeostasis, IL-15 is considered one of the most important. Interest in the role of IL-15 in promoting the anti-tumor and anti-viral role of NK cells has grown significantly in past years.⁷⁰ Many clinical trials already are testing the ability of IL-15 and its agonists to promote NK proliferation and cytotoxicity.^70,71 Interestingly, studies from us and others have shown the critical role of IL-15 in upregulating TRAIL expression on the surface of NK cells.^71,72 Binding of IL-15 to its receptor promotes NK toxicity in a TRAIL-dependent and independent manner via the activation of several downstream pathways, including the JAK/STAT pathway.⁷²

Despite growing evidence demonstrating the expression of TRAIL by many different subsets of NK cells, the extent of its effect on NK-mediated killing remains largely unknown. We have shown that incubation of primary NK cells with plasma from HIV-infected patients significantly increases the expression of TRAILshort on the surface of NK cells.⁷³ This also led to a decrease in cytotoxic function of NK cells (unpublished data).

3. TRAIL: TRAILshort and CTLs

CD8 T cells play an important role in control of viral infection. The initial decline in viremia occurs only after the appearance of the viral-specific CD8 cells and depletion of CD8 cells in SIV-infected macaques caused accelerated disease progression.^74,75 Once the naive CD8 cells are activated, clonal expansion is required to produce the cells that recognize specific HIV epitopes.⁷⁴ The early CD8 response is usually directed towards Nef and Env, which are variable epitopes.⁷⁵ Unfortunately, this initial differentiation of CD8 cells and the development of an effector population do not result in viral control in many cases.⁷⁵ To further understand the factors that are relevant in appropriate control of HIV by these CTLs, several studies looked into the CTL response in HIV elite controllers, a subset of HIV-infected patients characterized by their ability to naturally control the virus.⁷⁴ The CD8 cells of HIV elite controllers have several characteristics that are thought to contribute to their superior ability to control the virus: (1) polyfunctionality, meaning that they express and produce CD107a, IFN-γ, MIP-1β, IL-2, and TNF-α simultaneously, and the ability to sustain this polyfunctional phenotype⁷⁶; (2) higher cytotoxic capacity, explained by the level of Tbet expression and the ability of these cells to efficiently load lytic granules and deliver granzyme B⁷⁷; (3) increased proliferation⁷⁶; and (4) predominance of an activation phenotype associated with viral control, characterized by the expression of HLA-DR and the absence of CD38.⁷⁸ Host factors such as HLA-I alleles including B*27 and B*57 also affect CTL function and have been found to be extremely significant for viral control by CTLs.⁷⁹ These HLA alleles determine the HIV epitopes that are presented to the immune system; it is noteworthy that these protective HLA molecules target highly conserved epitopes in Gag.⁷⁹

Through the course of the infection and with persistent exposure to the HIV antigen, the T cells become “exhausted.”⁸⁰ During the early phase of this T-cell exhaustion, viral-specific CD8 cells show decreased proliferation and cytokine production and the development of this exhausted phenotype associated with the expression of markers such as PD-1, CTLA-4, and TIM3 significantly impairs the antiviral capacity of these cells.⁸⁰ Although the role of TRAIL produced by CD8 in control of several viruses such as influenza and West Nile virus has been studied,⁸⁰ its role in HIV infection is less well defined. A study by Kuerten et al. shows that, despite the high numbers of TRAIL-expressing CD8⁺ T cells in advanced HIV infection (i.e., in HIV⁺ CD4^low subjects), these CD8⁺ T cells do not appear to be able to control the infection.⁸¹ Although the investigators hypothesize that it is the absence of HIV-antigen-specific CD4⁺ T cells that renders HIV-specific CD8⁺ T cells helpless, the presence of a TRAIL antagonist preventing these CD8 from properly performing their function cannot be ruled out.

Despite the complex nature of the problem and the tiny number (n = 1) of patients in whom HIV has been cured,⁸² hope for a cure still is strong in the minds of many researchers and in the hearts of many patients living with the virus.

II. FUTURE DIRECTIONS AND UNANSWERED QUESTIONS

It is clear that TRAIL is an important player in HIV pathogenesis and the presence of TRAILshort to antagonize the function of TRAIL poses many questions. Can the expression of TRAILshort be used as a marker to predict response to treatment? Given its role in HIV immunopathogenesis, could TRAILshort expression correlate with the strength of a patients’ immune response to HIV? One would expect patients with a stronger immune response to show less TRAILshort. With the previously disappointing results of TRAIL agonists as a treatment option for cancer and chronic viral infections, could blocking TRAILshort help to renew hope for therapeutic benefits of this molecule? If used, would the inhibition of TRAILshort cause any significant effects on healthy cells expressing this protein? These and many other unanswered questions point to the ongoing need for continued research into various potential roles of TRAILshort in the immunopathology and treatment of HIV infection.

ABBREVIATIONS:

APOBE-3

Apolipoprotein B mRNA editing enzyme catalytic-like 3

ART

antiretroviral therapy

BCL-2

B-cell lymphoma-2

CTL

cytotoxic T lymphocytes

DR5

death receptor 5

HIV

human immunodeficiency virus

HLA

human leukocyte antigen

IFN

interferon

ISG

interferon stimulated gene

NK

natural killer

PD-1

programmed cell death protein-1

pDC

plasmacytoid dendritic cell

SAMHD1

sterile alpha motif (SAM) domain- and HD domain-containing protein 1

STAT

signal transducer and activator transcription

TNF

tumor necrosis factor

TRAIL

tumor necrosis factor-related apoptosis inducing ligand