A Coat of Many Colors: Neuroimmune Cross Talk in Human Immunodeficiency Virus Infection

Abstract

The use of antiretroviral therapy (ART) has reduced mortality and increased the quality of life of HIV-1 infected people in more developed countries, where access to treatment is more widespread. However, morbidities continue which include HIV-1 associated neurocognitive disorders (HAND). Subtle cognitive abnormalities and low-level viral replication underlie disease. The balance between robust antiviral adaptive immunity, neuronal homeostatic mechanisms, and neuroprotective factors on one hand and toxicities afforded by dysregulated immune activities on the other govern disease. In the ART era, these lead to chronic low-level infection and often subclinical disease. New insights into the pathobiological processes for neuroimmune-linked disease and ways to modulate such activities for therapeutic gain are discussed. Better understanding the complexities of immune regulation during HAND can improve diagnosis and disease outcomes but is also relevant for the pathogenesis of a broad range of neurodegenerative disorders.

Introduction

Human immunodeficiency virus type I (HIV-1) is a lentivirus that infects CD4+ T cells, dendritic cells, monocytes, and macrophages. A progressive immunodeficiency resulting from the destruction of CD4+ T cells and immune dysfunction result in the acquired immunodeficiency syndrome (AIDS), which leads inevitably to broad range of opportunistic infections (OI) and malignancies. It continues to ravage the world; the latest UNAIDS/WHO statistics (July 2008, numbers as of end of 2007) indicate that there are 33 million people living with HIV/AIDS, 2/3rds of whom live in Sub-Saharan Africa, with the majority living in resource-poor regions (http://www.who.int/hiv/pub/epidemiology/pubfacts/en/).

While the immune system disorders are the focus of much study, HIV-1 infection has significant effects on the nervous system, as the virus both infects and affects the brain (Yadav and Collman 2009). Prior to the advent of antiretroviral therapy (ART) OIs were a common complication of advanced HIV-1 infection associated with advanced immunosuppression and reduced CD4+ T cell counts. These included infection of the central nervous system (CNS) with heterologous viruses (cytomegalovirus, herpes and JC viruses), mycobacteria (tuberculosis and atypical mycobacterial infections), fungi (cryptococcus), protozoa (toxoplasmosis), and malignancies including lymphoid cancers linked to Epstein-Barr virus. These still continue with ART albeit reduced (Langford et al. 2003). However functional and pathological abnormalities of the CNS occur in the absence of OIs and have been linked directly or indirectly to HIV-1 itself and the host responses to it (Boisse et al, 2008). This continues as the most common cause of neurologic disease in the setting of progressive HIV infection, although shifts in pathogenesis from florid HIV replication to more diverse mechanisms have recently been noted (Everall et al., 2009). Herein we specifically review recent findings for how HIV can damage the nervous system and how it may be treated with a focus on what was seen on how virus interacts between the nervous and immune systems. As with many issues in biomedical science, the bulk of our knowledge exists from those in regions rich in medical and research infrastructure, and our review reflects these. We note that the problems caused by HIV for the brain, disproportionally affect those in resource-poor areas, and considerable efforts are being mobilized to address this.

The HIV-associated neurocognitive disorder (HAND) comprises a broad range of neurological abnormalities including asymptomatic neurocognitive impairment, HIV-associated mild cognitive motor disorder, and the most severe disease, HIV-1 associated dementia or HAD (Antinori et al. 2007). Severe dementia now affects less than 7% of infected people during the latter stages of disease. Due to the increasing longevity of HIV-1 infected individuals, the incidence of HAD as well as the other cognitive and motor abnormalities associated with HIV-1 infection have declined, although the prevalence of HAND has increased (Antinori et al., 2007; Dore et al., 2003; Sacktor, 2002; Woods et al., 2009). The most severe cognitive, motor and behavioral impairments are now supplanted by milder, less profound cognitive impairment that can have significant negative ramifications on employment, medication adherence, and other activities of daily living (Grant, 2008).

HAD was first described nearly two and a half decades ago and has been associated with an encephalitis characterized by enhanced central nervous system (CNS) viral replication, formation of macrophage-derived multinucleated giant cells, astrogliosis, microgliosis, neuronal loss and myelin pallor. There have been many excellent reviews of neuroAIDS, and readers are referred to these for an in-depth reference (Ellis et al., 2007; Gonzalez-Scarano and Martin-Garcia, 2005; Kaul et al., 2005; McArthur et al., 2005). In the current era of widespread use of ART, HIV-linked CNS disease is associated with more subtle neuropathological changes and appears linked to synaptodendritic damage (Ellis et al., 2007). HIV-1-related brain pathology such as encephalitis is now an infrequent finding (Everall et al., 2009). The neuropathogenesis of HIV-1 infection is most likely a metabolic encephalopathy triggered by viral infection and immune activation of brain mononuclear phagocytes (MP; blood derived perivascular macrophages and microglia) (Zheng and Gendelman, 1997). However, other causes and pathologies are still being sought.

HIV-1 infection is now a chronic condition and can exhibit interactions with other neurodegenerative disorders. Older age has indeed been identified as a risk factor for HAD (Valcour et al., 2004), and in HIV-1 infected individuals on ART, extrapyramidal motor signs are linked to age (Valcour et al., 2008). Epidemiological trends now suggest a risk of concomitant neurodegenerative disease such as increased risks for developing Alzheimer's disease (Ikezu, 2009). Notably, increased β-amyloid is found in the brain in HIV-1 infection, and patients with HAND show increased aggregation of β-amyloid within neurons (Achim et al., 2009). The complexities in disease pathobiology for HAND are becoming even broader. Coexisting conditions, such as drug abuse (Bell et al., 1998; Rippeth et al., 2004) and infection with hepatitis C virus (Clifford et al., 2009; Letendre et al., 2007) may play significant roles in HAND. These further serve to complicate disease diagnosis and treatment strategies and warrant a re-evaluation of the importance of disease biology.

MP immunity, viral infection and immune activation

HIV-1 infection of the CNS is known to initiate disease, unlike other neurodegenerative disorders where the cause and effect relationships between innate microglial inflammation and neuronal injury are unclear. HIV-1 enters the brain early after infection. The means to regulate MP inflammatory responses dictate the tempo and progression of cognitive dysfunction. Much is now known in this regard. In neuroAIDS, MP are recruited to the brain. The mechanism for HIV-1 infection of the CNS is commonly described as a “Trojan Horse” in studies of lentiviral pathogenesis (Gendelman, et al., 1985; Haase, 1986; Gendelman et. al., 1986). Herein, virus resides in a restricted state within monocytes or following differentiation into macrophages and buds into endocytic compartments with limited cell surface expression of viral proteins. This allows the cell, in part, to escape immune surveillance (Gendelman et al., 1994), paralleling the ancient mythological Trojan horse that Odysseus devised allowing the Greeks to secretly enter the city of Troy. For lentiviruses, this is the virus's strategy of entering the brain and overcoming the restrictions imposed by a seemingly impermeable blood-brain barrier (BBB). Monocytes attracted into the brain by a chemokine gradient penetrate the BBB and differentiate into macrophages. This infiltration of peripheral cells then allows viral dissemination within the brain.

Monocytes-Macrophages and the BBB

Diapedesis of monocytes across the BBB during HIV-1 infection is a hallmark of neuroAIDS and occurs as a consequence of ongoing brain MP activation. Secretory factors released from immune competent HIV-1 infected macrophages and microglia affect the pathobiologic activities that underlie the cascading immune events leading to the development of HAD/HIV-1 encephalitis (HIVE). MP infection leads to glial activation in the brain. HIV-1 and pro-inflammatory factors first affect the integrity of the BBB. For example, HIV-1 signals through signal transducers and activators of transcription 1 in human brain microvascular endothelial cells in the BBB, inducing interleukin-6 (IL-6) secretion, which speeds monocyte migration (Chaudhuri et al., 2008). This is perpetuated through monocyte chemoattractant protein 1 (MCP-1) secreted by both MP and astrocytes. Indeed, MCP-1 is a potent chemoattractant for macrophages and monocytes, further orchestrating MP brain infiltration (Peng et al., 2008b). Ongoing viral replication as well as the viral proteins themselves induce MCP-1 secretion (Eugenin et al., 2006). MP can also affect the microvascular proteome and vice versa to elicit pro-inflammatory changes required for transendothelial migration (Ricardo-Dukelow et al., 2007). Thus, interplay between viral and cellular factors initiates disease. Once MP gain entry into the brain, the pro-inflammatory environment triggered by ongoing viral infection leads to further MP activation and secretion of pro-inflammatory cytokines, chemokines and arachidonic acid metabolites that serve to exacerbate BBB dysfunction and the neurodegenerative process.

MP and Neurodegenerative Responses

How MPs can affect neurodegeneration is certainly complex and can occur through a number of mechanisms each of which is not mutually exclusive. For example, HIV-1gp120 is shed from the surface of progeny virus and can elicit neurotoxicity (Cornblath and Hoke, 2006) through the production of cytokines from astrocytes and microglia. HIV-1gp120 over-expression in transgenic mice elicits neural damage, which correlates with gp120 protein concentrations (Okamoto et al., 2007) and is dependent on activation of c-Jun N-terminal kinase (JNK) and p42 extracellular-regulated kinase. HIV-1gp120 is also responsible for signaling the release of a broad range of pro-inflammatory factors from macrophages, in part, through phosphatidylinositol-3 kinase and mitogen-activated protein kinase signaling (Cheung et al., 2008; Lee et al., 2005).

While HIV-1gp120 induces cytokine and chemokine secretion and facilitates disease progression, this is just a portion of the many mechanisms of neuroinflammatory activities seen in disease. Indeed, viral infection itself can lead to MP activation and the secretion of a broad range of neurotoxic factors. Cytokines and chemokines released as a consequence further enhance MP activation during HIV-1 infection resulting in local neuronal dysfunction/death as well as additional monocyte-macrophage recruitment across the BBB (Brabers and Nottet, 2006; Persidsky et al., 2006). For example, TNF-α increases the permeability of the BBB and increases adhesion molecules on the epithelial cells required for cell transmigration. TNF-α induces a inflammatory cascade resulting in macrophage cytoskeleton remodeling and further increasing cytokine secretion (Kadiu et al., 2007). IL-1β shares neurotoxic properties with TNF-α, including NMDA receptor-mediated neurotoxicity (Brabers and Nottet, 2006; Strijbos and Rothwell, 1995). In the context of HIV-1 infection in the CNS, TNF-α and IL-1β are primarily considered toxic (Brabers and Nottet, 2006), and interestingly affect IL-8 secretion by astrocytes within the CNS (Saas et al., 2002). Although IL-8 is a chemokine involved in neutrophil chemoattraction, it is increased in the CSF in HAD (Zheng et al., 2008). These are amongst many conjoint innate immune responses that are operative in disease.

M1 and M2 MP phenotypes and disease regulation

One hypothesis commonly espoused for HAND is the control of the MP phenotype. Viral infection and immune activation drive M1 responses, while antiviral responses (drugs and immune surveillance mechanisms) drive M2. Certainly, cytokine and chemokine secretion by MP are in large part controlled by the MP polarization state, either as an M1 or M2 phenotype (Benoit et al., 2008). M1 polarized macrophages are activated through classical pathways and secrete pro-inflammatory cytokines such as TNF-α, IL-1β and interferon-gamma (IFN-γ). M2 polarized macrophages are activated through an alternative pathway by IL-4 and IL-13 resulting in the secretion of anti-inflammatory cytokines such as IL-10 and IFN-β. M1 polarization, which is relatively short-lived, affects the levels of HIV-1 replication and neurotoxicity, while M2 polarization results in long-term decreased viral production and neural homeostasis despite no change in proviral DNA and viral protein levels (Cassol et al., 2009) (Figure 1). During disease, MP are recruited into the brain and depending on the neural environment, can affect the activation state of perivascular cells. These blood borne macrophages are a major source of virus spread as well as neurotoxic mediators. The ongoing inflammatory environment in affected brain tissue dictates the tempo of disease and affects both macrophage and microglial innate immune responses that regulate neuronal function (Figure 1). Successful control of the macrophage and microglial phenotype remains a target in development of new therapies that affect neuroinflammation-induced neurodegenerative activities.

Figure 1. Neuropathogenesis of HIV-1 Infection.

A schematic of the cascade of neuropathologic events during progressive HIV-1 infection of the CNS. Activated monocytes are drawn to the brain by chemokines. CNS infection serves to activate astrocytes and resident microglia leading resulting in spread of infection to neighboring cells with concomitant neuroinflammation. Upon activation microglia polarize to either an M1 or M2 phenotype, leading to a neurotoxic or neurotrophic microglial phenotype during HIV-1 infection of the CNS.

Glial Cross Talk

While abundant studies support the role of macrophages and microglia in HIV-1-induced CNS damage, other glia such as astrocytes likely play a crucial role. HIV-1-exposed astrocytes up-regulate complement C3 in an effort to clear the infectious agent (Speth et al., 2002). This upregulation of complement C3 was demonstrated in proteomic studies of cerebrospinal fluid (CSF) in HIV-1 and SIV (Pendyala et al., 2009; Rozek et al., 2007). Intriguingly, the classic complement cascade (including C3) mediates neuronal synaptic destruction during development, and when activated later, may contribute to neurodegeneration (Stevens et al., 2007). HIV-1 binding to astrocytes and/or the local production of MP-derived proinflammatory factors can induce robust pro-inflammatory responses. (Li et al., 2007). Astrocytes secrete chemokine ligand-10 (CXCL10), a potent chemoattractant, which acts synergistically in the presence of pro-inflammatory cytokines and HIV-1 recruiting immune cells (Williams et al., 2009).

Astrocytes are also known to contribute to glutamate excitotoxicity during HIV-1 infection [comprehensively reviewed by (Erdmann et al., 2006)]. Glutamate causes oxidative stress in glial cells, which can negatively affect neurons. HIV-1 also has been found to induce astrocyte expression of matrix metallopeptidase 9, which is known to contribute to degradation of the BBB (Ju et al., 2009).

Astrocytes also produce high levels of stromal cell-derived factor-1α (SDF-1α, CXCL12), induced by IL-1β. Evidence has been provided for the effects of SDF-1 during HAND to be both beneficial and detrimental; however, the true role is still unclear (Peng et al., 2006). Finally, astrocyte secretion of low levels of TNF-α and IL-1β by HIVE patients provide further evidence for the contribution of proinflammatory astrocytes to the toxic inflammatory environment (Xing et al., 2009). Further exploration of the role of astrocytes during HIV-1 infection is vital to understanding the mechanism(s) that leads to the development of HAND in HIV-1 infected individuals.

Considerable evidence now exists for cooperation amongst glial cells in affecting neurotoxic activities. Co-cultivation of astrocytes and microglia has yielded great insight into glial cell communication as it occurs during HIV-1 infection. Astrocytes can affect the microglial phenotype during HIV-1 infection by altering expression of microglia structural proteins, chemoattractants, levels of viral replication and neurotoxins (Wang et al., 2008a). HIV-1-infected astrocytes also accelerate MP cell death and induce oxidative stress (Wang et al., 2008b). The ability of astrocytes to regulate HIV-1 infected microglial neurotoxicity serves to define the tempo of HAND (Figure 2).

Figure 2. Innate and Adaptive Immunity in HIV-1 Neuropathogenesis.

Within the infected brain, both proinflammatory (Teff, CTL, M1 microglia) and protective (M2 microglia, Treg) aspects of immunity are evident [adapted from (Huang et al., 2009)]. When the inflammatory environment becomes dominant disease is manifest.

Adaptive Immunity, Viral Infection, and Disease Control

Cytotoxic T Cells (CTL)

Along with the innate response, there is good evidence of an adaptive response to HIV-1 (and in the monkey SIV) in the brain. Within weeks following SIV inoculation, a significant increase in the number of CD8+ T cells is present within the brains of infected animals (Marcondes et al., 2001). These cells have distinct patterns of T cell receptor clonal rearrangements relative to those found in the lymphoid tissues, indicating regional specificity in the adaptive cellular response (Marcondes et al., 2007). Such brain infiltrating cells have been shown to be SIV-specific CTL functionally, and analyses with SIV specific tetramer reagents confirm the specific CTL presence and enrichment in the CNS (Marcondes et al., 2007). In SIV encephalitis, these CD8+ T cells are often found associated with infected macrophages (Kim et al., 2004), further indicating their specificity and activity. Interestingly, combined antiretroviral treatment, at least in the short-term, does not lower the number of these cells in the brain (Marcondes et al., 2009).

CD8+ T cells are found in the brains of patients with HIVE (Petito et al., 2003; Petito et al., 2006). Furthermore, similar to the studies in monkeys, these cells can be found in the brains before the development of disease (McCrossan et al., 2006). During the course of infection, increased CD8+ T cell activation is found in the CSF (Neuenburg et al., 2005) including HIV-1-specific CTL (Jassoy et al., 1993; Sadagopal et al., 2008). CTL response to HIV-1 in the CNS and elsewhere is critical in the body's protection from the virus. In a distinctly specialized organ such as the brain, CTL activity may also have harmful effects, such as bystander damage to neurons or support cells. CTL accumulate and persist in the brain, likely throughout the now lengthened period of infection. As with the innate response, these adaptive CTL may be double-edged swords – needed for viral control but capable of leading to brain injury.

Regulatory T Cells and Neuroprotection

Natural regulatory T cells (Treg) have been identified as key contributors to the control of self-tolerance [reviewed by (Sakaguchi, 2000)]. More recently their role in immune suppression has been investigated in many neurodegenerative disorders, such as Parkinson's disease (PD) (Reynolds et al., 2007), amyotrophic lateral sclerosis (ALS) (Banerjee et al., 2008) and stroke (Liesz et al., 2009). HIV-1 infected individuals have increased Treg populations (Cao et al., 2009; Terzieva et al., 2009), and increases in Treg are seen even as other T cell populations are decreasing during later stages of infection (Rallon et al., 2009). In contrast, effector T cells (Teff) can exacerbate immune activation and perpetuate the progression of HIV-1 infections, while Treg have a profound anti-inflammatory response decreasing astrogliosis and microglial inflammation leading to neuroprotection (Liu et al., 2009) (Figure 2). The balance between Treg and Teff responses can affect the tempo and progression of HIV-1 neuropathogenesis. This, however, is not limited to neuroAIDS.

In PD, Treg that have been exposed to nitrated alpha-synuclein are capable of slowing the progression of PD through modulation of the microglial phenotype (Reynolds et al., 2009a, b). In ALS, Treg populations are decreased (Mantovani et al., 2009). Also, in animal models of ALS, transferring activated Treg to ALS animals delays neurological symptom onset and loss of motor function with extended survival (Banerjee et al., 2008). The benefits of Treg in CNS HIV-1 infection, whereby a tolerance response is induced instead of activating the immune system in response to infection and neurodegeneration, has recently been reviewed (Huang et al., 2009). Prophylactic induction of Treg is thus becoming recognized as an important potential therapeutic approach for neurodegenerative diseases. Recent studies support the notion that controlling the immune system to regulate the inflammatory response is essential to decreasing neural damage caused by secretion of pro-inflammatory cytokines and that are exacerbated to Teff responses. The balance is not easily seen as Teff responses serve to control viral infection. Thus, the balance between control of inflammation and viral replication is critical in control of HIV-1 neuroimmunity and neurotoxicity (Figure 2).

Immune reconstitution inflammatory syndrome (IRIS)

IRIS is a condition observed in some immunosuppressed advanced HIV infected people during the time the affected patient's immune system begins to recover after ART and responds to prior OIs. This is manifest by a robust inflammatory response that paradoxically speeds disease. Most patients will recover without treatment. However, when manifest within the CNS, a common CD8+ T cell lymphocyte infiltration can cause significant morbidity and sometimes mortality (Gray et al., 2005; Venkataramana et al., 2006). CNS manifestations highlight the role of prior infections and how a newly competent immune response can affect disease in the brain.

Mechanisms of Neuronal Demise

Apoptosis has long been thought to represent the end-result of neuronal damage in neuroAIDS. Recently, it has been found that there is a strong interdependent relationship between apoptosis and autophagy mechanism (Maiuri et al., 2007). Autophagy and the ubiquitin-proteasome system (UPS) are responsible for the clearance of intracellular proteins. While apoptosis is a synchronized process leading to cell death, autophagy is now largely considered a mechanism to protect from toxicity in stress conditions. Strikingly, inhibition of autophagy can also induce apoptosis (Boya et al., 2005). In cells such as neurons removal of abnormal cellular components, such as aggregated proteins or damage organelles, is necessary to prevent cellular damage that can trigger apoptosis. In vivo, inhibition of autophagy in neurons leads to neurodegeneration (Hara et al., 2006; Komatsu et al., 2006).

Studies using the SIV-infected monkey model have revealed inhibition of neuronal autophagy in the presence of encephalitis (Alirezaei et al., 2008a; Alirezaei et al., 2008b). Supernatants from SIV-infected microglia inhibit autophagy in primary neuronal cultures and induce their death; furthermore accumulation of p62/sequestosome 1, a protein cleared by autophagy, is present in neurons in the brains of infected monkeys as well as HIV-infected people. Recent studies have confirmed findings of decreased autophagy in the brains of HIV-1-infected people and feline immunodeficiency virus infected cats, as well as in vitro studies with human neurons (Zhu et al., 2009).

Other reports indicate that alterations in the UPS also occur in HIVE (Gelman and Nguyen, 2009; Gelman and Schuenke, 2004) In cells such as neurons, autophagy is especially important where the UPS apparatus is insufficient; thus, the processes related to intracellular clearance in neurons may be an Achilles' heel in the untoward effects of HIV-1 on neurons. It is also interesting that lithium, which increases autophagy by inhibiting inositol monophosphatase (Sarkar et al., 2005), has been shown to have efficacy in preclinical models of neuroAIDS (Dou et al., 2005). Lithium is now being clinically tested in those with neuroAIDS (Letendre et al., 2006; Schifitto et al., 2009), potentially linking some of the newer adjunctive therapies being tested in neuroAIDS (see below) with autophagy.

Therapeutic Milestones

Antiretroviral Drugs and NeuroAIDS

Multi-modality antiretroviral treatment has emerged as the mainstay for effective treatment of HIV-1 infected patients reducing both morbidities and disease mortality. For CNS HIV-1 infections, ART-naïve, HAD patients with significant cognitive dysfunction show dramatic improvement in neurologic function following ART initiation (Gendelman et al., 1998), which is coincident with and thought to be a function of lowered viral loads. Following therapy, levels of neurotoxins in plasma and CSF elevated at presentation are decreased, which also parallel reductions in plasma and CSF neurotoxicity. These data serve to support the notion that HAND is a reversible metabolic encephalopathy (Gendelman et al., 1998). Overall, ART has decreased the incidence of HAD along with reductions in more common opportunistic infection and other AIDS-related illnesses.

Despite the increased improvement in HAD, thought due to lower systemic viral loads, the lack of ART penetration into the CNS still allows continued viral replication within the CNS (Letendre et al., 2008). This may be corrected through the administration of CNS penetrating drugs directed to control CNS viral replication (Letendre et al., 2004). While this most likely will not prevent development of HAND, it would slow disease progression by decreasing levels of viral proteins and secretory products from HIV-1 infected activated macrophages. Most antiretroviral drugs cannot traverse the BBB or are quickly effluxed via P-glycoproteins (P-gp) or other efflux transporters. P-gp inhibitors have shown some success in animal models of HIVE; blockade of the P-gp efflux transporter allows retention of antiretroviral drugs in the CNS and permits increased bioavailability and activity within the CNS (Spitzenberger et al., 2007). Strategies like these allow therapeutic modalities, which would otherwise never pass the BBB, to act at sites of uncontrolled viral growth within the CNS. There is also the added benefit of preventing further complications that are associated with direct administration into the CNS.

However, it should be noted that a recent study (ACTG 736) on HAND found that while ART regimens with good penetration led to lower CSF viral loads, they were also associated with poorer neurocognitive performance (Marra et al., 2009). While larger controlled clinical trials need to be performed, this highlights the current status of HAND; it is not simply a low-level version of HAD/HIVE; and in the now chronic condition of HIV-1 infection, our knowledge of the direct and indirect effects on the brain are under evolution. In addition, the effects of chronic treatment with ART on CNS function is unknown

Adjunctive therapies for neuroAIDS

Because HAND shares many symptoms with other neurodegenerative disorders, drugs that were previously used in other diseases have more recently been utilized. For many years memantine, currently approved for use in the treatment of AD, has been suggested for treatment of HAND due to its observed ability to block gp120 and Tat induced neurotoxicity (Gendelman, 2007; Holden et al., 1999; Nath et al., 2000). It acts as a noncompetitive antagonist of NMDA receptors and also has been found to increase levels of brain-derived neurotrophic factor and to protect dopamine function (Meisner et al., 2008). While memantine administration has proved to be safe, no differences in cognitive performance have been observed in HIV-1 infected patients, despite positive changes seen by magnetic resonance spectroscopy (Schifitto et al., 2007).

Copolymer-1 (Cop-1), a known immune modulator responsible for decreasing TNF-α and IL-12 levels, has shown neuroprotective capabilities in a murine model of HIVE, resulting in decreased neuronal damage, astrogliosis and microgliosis (Gorantla et al., 2007a; Gorantla et al., 2008). Cop-1 acts directly on neurons providing protection through protein kinase Cα anti-apoptotic signaling (Liu et al., 2007). Sodium valproate, has been administered in clinical trials during early stage infection and appears to be mildly beneficial in recovery from HAND (Schifitto et al., 2006). While not affecting viral replication, it supports modest improvement in cognitive function (Schifitto et al., 2006); however, future studies are necessary to determine the extent of improvement possible with long-term administration.

Glycogen synthase kinase-3beta (GSK-3β) is activated by Tat and platelet activating factors and controls apoptosis through inhibition of NF-κB (Schifitto et al., 2009). GSK-3β is inhibited by lithium and provides an impetus to study GSK-3β inhibitors for HAND (Dou et al., 2004; Gendelman, 2007; Schifitto et al., 2009). A pilot study to test the effects of lithium on HAND showed improved neuropsychological performance in ART-treated, neurologically impaired individuals (Letendre et al., 2006). Further clinical trials for this and other GSK-3β inhibitors need to be pursued to establish the nature of lithium effects on HAND (Crews et al., 2009).

Growth factors affect neuronal survival. Platelet-derived growth factor (PDGF) is one such factor known to protect neurons from HIV-1-induced cytotoxicity (Peng et al., 2008a). PDGF down-regulates Bcl-2-associated Bax protein, an apoptosis inducing protein, and inhibits HIV-1gp120 induced cytochrome c release (Peng et al., 2008b). These and other trials of adjunctive therapies are necessary as antiretroviral drugs alone have failed to eliminate HAND in HIV-1 patients (Cysique and Brew, 2009).

Recent studies have examined the ability of minocycline, an antibiotic with potent anti-inflammatory and neuroprotective properties, to protect against SIV encephalitis and neurodegeneration. Minocycline-treated macaques had less severe encephalitis with reduced brain expression of neuroinflammatory markers of p38 mitogen-activated protein kinase activation and MCP-1, as well as reduced axonal degeneration and CNS virus replication (Zink et al., 2005). The findings have led to ongoing clinical studies now operative in resource-limited settings.

All together, adjunctive therapies shown successful in management of other neurodegenerative disorders need also be assessed in HAND for their efficacy beyond the mere control of viral load. The therapeutic solution relies on the means to harness the immune system for therapeutic benefit rather than simple control of viral replication. It is for this reason that working to understand intercellular communication among different encephalic cells during infection is a vital part in prospective HAND therapies.

Nanomedicine and CNS Drug Delivery

A broad range of nanomedicines are being developed to improve drug delivery for CNS disorders. These nanoformulations are designed to circumvent the BBB (Gendelman et al., 2008). BBB penetrance is dependent on nanoparticle size, shape, and protein and lipid coatings. These all affect drug uptake, release and ingress across the barrier. The BBB poses limitations on antiretroviral therapies and their pharmacokinetic and biodistribution properties. Recent reports support the notion that nanotechnology can serve to improve the delivery of antiretroviral medicines, called nanoART, across the BBB and affect biodistribution and clinical benefit for HIV-1 disease (Nowacek and Gendelman, 2009). To test CNS-penetrating ART, cell-based nanoparticle delivery systems were developed in an effort to transport ART across the BBB. Taking advantage of the Trojan horse principle, these nanoparticles are taken up by monocytes and sheltered within the monocytes as they are carried across the BBB. The drug is then released to act within the CNS (Dou et al., 2006; Dou et al., 2009). Other laboratories have conjugated nanoparticles with Tat, which has an affinity for nuclear transport mechanisms (Berry et al., 2007). This results in a nanoparticle that has high CNS penetrability while still bypassing efflux transporters to prolong exposure within the CNS.

Animal models for HIV-1 Neuropathogenesis

A specific challenge in finding ways to test new therapeutics for HAND is the inability of HIV-1 to infect other species. The most studied means to address this is through the SIV model for neuroAIDS. SIV has many parallels to HIV-1; and therefore, infection of monkeys provides a non-human primate model for researching HIV-1 infection and HAND/HAD/HIVE. To examine the affects of HIV-1 genes in vivo, portions of HIV-1 can be inserted into SIV creating hybrid simian-human immunodeficiency virus (SHIV). There are many recent reviews of SIV and SHIV models for HIV-1 neuropathogenesis (Clements et al., 2008; Crews et al., 2008; Fox, 2008; Williams et al., 2008). Most studies focus on the development of terminal disease with SIV encephalitis and much has been learned about its development and pathology, including its in-depth molecular characterization (Roberts et al., 2003)

Several studies have recently examined the brains of SIV-infected monkeys during chronic infection and during the administration of effective ART, in the absence of the confounding conditions and experimental limitations that exist in humans. In chronic SIV infection (animals infected for approximately 2 years but with no obvious disease) evidence of an active viral-host interaction with both virus and CD8+ T cells present in the brain, and increased expression of a number of immune modulatory molecules (Roberts et al., 2006). Two studies have examined the effects of combination antiretroviral treatment. In one, markers of CNS inflammation were reduced but viral DNA persisted in the brain, indicating the possibility of the brain remaining a reservoir for the virus (Clements et al., 2005). In the other, treatment reduced viral RNA in the brain and prevented the development of SIV-induced neurophysiological abnormalities, but did not affect the number of CD8+ T cell in the brain; and varied effects were seen on production of immune modulatory molecules (Marcondes et al., 2009). Clearly, given the current clinical state, further study of the chronically infected brain in the presence of treatment is needed to understand the substrates of HAND.

However, the expense of this system and limitations in use has led to the development of model systems that use HIV-1 in “humanized” animal models. For example such “humanized” mice have been developed through reconstituting irradiated BALB/c-Rag2^−/−γ_c^−/− mice with umbilical cord blood-derived CD34+ hematopoietic stem cells (Gorantla et al., 2007b). Direct injection of HIV-1-infected monocyte-derived macrophages (MDM) intracerebrally induces HIVE, providing a good model to study therapeutics that target HIVE. Other humanized rodent models, such as those with transplanted human bone marrow, liver and thymus have been used for studies on HIV-1 prevention and pathogenesis (Wege et al., 2008) but not yet for HIV neuropathogenesis. The application of these and other rodent models to current issues in HAND, including its chronicity, will be a useful addition to our research resources.

Drug Abuse and NeuroAIDS

The influence of concurrent drug abuse in AIDS is significant. Neuropathologically drug abusers show greater levels of neuroinflammation as evidenced by microglial activation, astrocytosis and CD8 lymphocytic brain infiltration (Anthony et al., 2005; Tomlinson et al., 1999). Alcohol, marijuana, nitrite inhalants, amphetamines, cocaine and hallucinogens are the drugs most commonly associated with HIV-1. These have been the subjects of many reviews in recent years (Anthony and Bell, 2008; Nath et al., 2008).

Morphine is a well-known opioid analgesic used extensively in clinical practice with significant dampening effects on the immune system (Roy et al., 2001; Roy et al., 2006). Studies in rhesus macaques that were chronically administered opioids have yielded conflicting results on the effects of morphine on SIV pathogenesis (Donahoe et al., 2009; Kumar et al., 2006).

Cocaine has been one of the most widely abused drugs worldwide. Epidemiological studies on abused drug users and AIDS link abuse of cocaine (by different routes) to increased incidence of HIV-1 seroprevalence and progression to AIDS (Baldwin et al., 1997). In laboratory studies, cocaine exposure has been shown to aggravate neurotoxic effects of HIV-1 proteins Tat and gp120, thus mediating accelerated neuronal apoptosis (Aksenov et al., 2006; Turchan-Cholewo et al., 2006).

Cocaine has multiple immunomodulatory effects, including the ability to influence cytokine release in immunoeffector cells. Cocaine was reported to increase the production of p24 in HIV-1-infected leukocytes by a mechanism involving TGF-β (Peterson et al., 1990; Pettoello-Mantovani et al., 1998). The interaction between cocaine and HIV-1 has also been evaluated in vivo using a hybrid-mouse model (huPBL SCID mouse) infected with HIV-1, in which cocaine administration decreases CD4+ cells and increases circulating virus loads (Roth et al., 2002). In addition, cocaine is known to enhance the secretion of MCP-1 in brain endothelial cells (Zhang et al., 1998), thereby facilitating increased influx of infected cells across the BBB (Dhillon et al., 2008). These findings underscore the role of cocaine as a cofactor for neuroAIDS.

HIV-infected individuals who are dependent on methamphetamine (METH) have a higher rate of neuropsychological impairment than those who do not use the drug (Rippeth et al., 2004). HIV-1-infected METH users on ART also manifest a higher plasma viral load than infected individuals on ART who are not using METH (Cadet et al., 2003; Langford et al., 2003). METH administration is known to result in long lasting dopamine depletion in humans and animals (Davidson et al., 2001). Intriguingly there is evidence of dopaminergic dysfunction in neuroAIDS (Kumar et al., 2009), leading to the possibility of interaction with the dopamine-affecting drugs cocaine and METH. Neuropathological examinations of brains of METH users with HIVE revealed a higher degree of loss of calbindin interneurons in the neocortex and increased microgliosis (Chana et al., 2006). Loss of such interneurons, including parvalbumin interneurons, correlated well with memory abnormalities in METH users with HIV-1 (Langford et al., 2003). Interestingly, transcriptional profiling studies revealed that in the brains of those with HIVE, METH users had significantly increased expression of a group of IFN-inducible genes (Everall et al., 2005).

Conclusions

HIV-1 commonly induces a spectrum of neurologic disorders that include opportunistic infections caused as a consequence of immune compromise and those resulting from direct virus-associated damage to the nervous system. These are common and debilitating and continue to occur despite usage of ART. Interestingly, neurologic disease parallels disordered regulation of both the innate and adaptive immunity and is driven through cell cross talk that includes glialneuronal interactions. Viral and cellular mediators linked to MP activation, growth factors, and adaptive immune surveillance that includes cytotoxic and regulatory T cells underlie disease pathogenesis. The control of antiretroviral and neuroprotective T cell functions and macrophage polarization are essential in combating the clinical consequences of HAND. Understanding the causes of neurodegeneration during HIV-1 infection and the parameters whereby certain individuals develop disease can provide researches with new therapeutic targets to positively affect disease outcomes. New insights into immunoregulation, drug carriage, neuroprotective strategies and intercellular communications will certainly lead to improved mental health outcomes by facilitating drug access to the CNS and control of pathologic events. Moreover, studies of neural immune processes in the setting of neuroAIDS will also provide novel ideas of broad implication to a range of neurodegenerative diseases.