HIV-associated dementia eventually develops in approximately half of children and a quarter of adults infected with HIV-1 (7). Neuropathological features that may accompany this cognitive-motor complex include dendritic and synaptic damage, apoptosis and frank loss of neurons, myelin pallor, astrocytosis, and infiltration of macrophages, microglia, and multinucleated giant cells (7,20,21).
Addiction to drugs, such as heroin, cocaine, and methamphetamine, is considered to be a major risk factor for HIV infection (22). Furthermore, it has been proposed that abuse of substances that cause by themselves profound alterations in CNS function could also affect the development of HIV-associated dementia (23). An autopsy study detected evidence of HIV encephalitis, including multinucleated giant cells and HIV p24 antigen, more frequently in HIV-positive drug users than HIV-infected homosexual man (24), and cocaine has been implicated in facilitation of HIV entry into the brain (25,26). However, although an effect of substance abuse on HIV-associated dementia appears likely, direct links between drug addiction and HIV-induced brain injury remain to be elucidated (23).
Macrophages and microglia play a crucial role in HIV-associated dementia because they are the predominant cells productively infected with HIV-1 in AIDS brains (7). Infection of astrocytes has also been rarely observed in pediatric cases (reviewed in ref. 27). It is currently held that HIV-1 infected macrophages migrate into the brain (28), thus allowing viral entry into the CNS, and the presence of macrophages/microglia has been reported to correlate with the severity of HIV-associated dementia (29). Furthermore, we and our colleagues have shown that HIV-1-infected or immune stimulated macrophages/microglia produce neurotoxins (7,30).
The mechanisms that initiate and mediate HIV-associated dementia are not completely understood. HIV-1 apparently enters the CNS soon after peripheral infection, and the virus primarily resides in microglia and macrophages, especially in those located in the perivascular space (28). It is not clear if the migration of infected monocytes and macrophages represents the only pathway for viral entry into the brain. Additionally, infection of monocytoid cells per se may not be sufficient to initiate the dementing process (28). In the CNS, HIV-1 is thought to cause immune activation of macrophages/microglia, changes in expression of cytokines, chemokines, and their receptors, and upregulation of endothelial adhesion molecules (reviewed in ref. 28). However, these observations may be the result of the process rather than the inciting event for HIV-1-associated brain pathology. Therefore, it has been proposed that peripheral (non-HIV) infection or other factors may trigger events leading to dementia after HIV-1 infection in the CNS has been established. One such factor could be the increased number of activated monocytes in the circulation that express CD16 and CD69. These activated cells could possibly adhere to the normal endothelium of the brain microvas-culature, transmigrate, and then trigger a number of deleterious processes (28).
Infection of cells by HIV-1 can occur after binding of the viral envelope protein gp120 to one of several possible chemokine receptors in conjunction with CD4. Depending on the exact type of gp120, different HIV-1 strains may use CXCR4, CCR3, CCR2, CCR5, or a combination of these chemokine receptors to enter target cells (31,32). Microglia are infected by HIV-1 primarily via CCR5 and CCR3, and possibly via CXCR4 (33). CCR5 and CXCR4, among other chemokine receptors, are also present on neurons and astrocytes, and, in particular, CXCR4 and CCR5 are highly expressed on neurons of macaques and humans (34). In vitro studies strongly suggest that chemokine receptors are directly involved in HIV-associated neuronal damage (17,35,36).
Even in the absence of intact virus, the HIV proteins gp120, gp41, gp160, Tat, Nef, Rev, and Vpr have been reported to initiate neuronal damage both in vitro and in vivo (37-41). In this regard, the viral envelope protein gp120 has been of particular interest, as it is essential for selective binding and signaling of HIV-1 to its target cell and for viral infection (17,33,35,36,39). Additionally, evidence has been provided that gp41, the membrane-spanning region of the viral envelope protein, correlates with the expression of immunologic/type II NOS (iNOS) as well as the degree of HIV-associated dementia (40).
A recurring question has been whether HIV-1 or its component proteins induce neuronal damage predominantly by an indirect route (e.g., via toxins produced by infected or immune-stimulated macrophages and/or astrocytes) or by a direct route (e.g., via binding to neuronal receptors) (7,17,42). Several lines of evidence suggest that HIV-associated neuronal injury involves predominantly an indirect route and resulting excessive activation of NMDARs with consequent excitotoxicity (30,43,44). Analysis of specimens from AIDS patients (44) as well as in vivo and in vitro experiments indicate that HIV-1 infection creates excitotoxic conditions, most probably indirectly via induction of soluble factors in macrophage/microglia and/or astrocytes, such as glutamatelike molecules, viral proteins, cytokines, chemokines, and arachidonic acid metabolites (7,42,45).
However, it has also been suggested that HIV-1 or its protein components can directly interact with neurons and also modulate NMDAR function, at least under some conditions (35,46). Picomolar concentrations of soluble HIV/gp 120 induce injury, both in vitro and in vivo, and this can lead to apoptosis in primary rodent and human neurons (37,39). Additionally, our group and subsequently several others have shown that gp120 contributes to NMDAR-mediated neurotoxicity (43). Both voltage-gated Ca2+ channel blockers and NMDAR antagonists can ameliorate gp120-induced neuronal cell death in vitro, although NMDAR antagonists are more effective in their protection (43,47). Transgenic mice expressing gp120 manifest neuropathological features that are similar in many ways to the findings in brains of AIDS patients, and in these mice, neuronal damage is ameliorated by the NMDAR antagonist memantine (38,48). Memantine acts as an open-channel blocker of the NMDAR-coupled ion channel, but the drug only manifests significant action when the NMDAR is excessively activated, leaving relatively spared normal synaptic transmission. Hence, memantine has proven to be clinically tolerated in a number of human trials and is already avialable in Europe for other clinical indications. It is also conceivable that other glutamate receptors in addition to NMDARs influence HIV-associated neuronal damage. For example, disparate mGluRs have been found to up- or down-modulate excitotoxic signals triggered by NMDARs (16).
In our hands, the predominant mode of HIV-1- or gp120-induced neurotoxicity of cerebrocortical neurons requires the presence of macrophages/microglia; HIV-1-infected or gp 120-stimulated mononuclear phagocytes have been shown to release neurotoxins that lead to excessive stimulation of NMDARs (17,30). These macrophage toxic factors include molecules that directly or indirectly act as NMDAR agonists, such as quinolinic acid, cysteine, and arachidonic acid and its metabolites, such as platelet-activating factor (PAF), a low-molecular-weight amine designated NTox, and perhaps glutamate itself (7,45,49).
Additionally, activated macrophages/microglia and possibly astrocytes produce inflammatory cytokines, including tumor necrosis factor (TNF)-a and interleukin (IL)-1p, arachidonic acid metabolites, and free radicals (ROS and NO) that may contribute to excitotoxic neuronal damage (7,45). TNF-a and IL-1P may amplify neurotoxin production by stimulating adjacent glial cells and by increasing iNOS activity [(45); Fig. 2].
In contrast to these indirect neurotoxic pathways, it has been reported that gp120 can directly interact with neurons when the neurons are exposed to gp120 in the absence of glial cells. Recently,
Fig. 2. Current model of HIV-associated neuronal injury. Immune activated and HIV-infected brain macrophages/microglia release potentially neurotoxic substances. These substances, emanating from macrophages and also possibly from reactive astrocytes, contribute to neuronal injury and apoptosis as well as to proliferation and activation of astrocytes (astrocytosis). A major mode of entry of HIV-1 into monocytoid cells occurs via the binding of gp120 and, therefore, it is not surprising that gp120 (or a fragment thereof) is capable of activating uninfected macrophages to release similar factors to those secreted in response to frank HIV infection. Macrophages bear CCR5 and possibly CXCR4 chemokine receptors on their surface in addition to CD4, and gp120 binds via these receptors. Some populations of neurons and astrocytes have been reported to also bear CXCR4 and CCR5 receptors on their surface, raising the possibility of direct interaction with gp120. Macrophages and astrocytes have mutual feedback loops (signified by the reciprocal arrows). Cytokines participate in this cellular network in several ways. For example, HIV infection or gp 120 stimulation of macrophages enhances their production of TNF-a and IL-1p (solid arrow). The TNF-a and IL-1p produced by macrophages stimulate astrocytosis. Neuronal injury is primarily mediated by overactivation of NMDARs with resultant excessive intracellular Ca2+ levels. This, in turn, leads to overactivation of a variety of potentially harmful enzyme systems, the formation of free radicals, and release of the neurotransmitter—glutamate. Glutamate subsequently overstimulates additional NMDARs on neighboring neurons, resulting in further injury. This final common pathway of neurotoxic action can be blocked in large measure by NMDAR antagonists. For certain neurons, depending on their exact repertoire of ionic channels, this form of damage can also be ameliorated to some degree by calcium channel antagonists or non-NMDAR antagonists. Additionally, agonists of p-chemokine receptors, which are present in the CNS on neurons, astrocytes, and microglia, can confer partial protection against neuronal apoptosis induced by HIV/gp120 or NMDA. IFN-interferon; IL-interleukin; NO^-nitric oxide; O2^-superoxide anion; TNF-tumor necrosis factor. (Color illustration in insert following p. 142.)
picomolar levels of gp120 was found to act at chemokine receptors on neurons to induce their death (35). Additionally, higher (nanomolar) concentrations of gp120 have been reported to interact with the glycine-binding site of the NMDAR (50), although it is not clear if gp120 exists at this high a level in the HIV-infected brain. Furthermore, gp120 may produce a direct excitotoxic influence via NMDAR-mediated Ca2+ oscillations in rat hippocampal neurons (51) and may bind to noradrenergic axon terminals in neocortex, where it possibly potentiates NMDA-evoked noradrenaline release (52). Nonetheless, it must be emphasized that many, if not all, of these direct effects on neurons were observed in vitro in the absence of glial cells. However, glial cells are known to modify these death pathways. Thus, we feel that based on work in mixed neuronal/glial systems that simulate the conditions that exist in vivo, the indirect route to neuronal injury is the predominant one.
Along these lines, gp120 has been found to aggravate excitotoxic conditions by impairing astrocyte uptake of glutamate via arachidonic acid that is released from activated macrophages/microglia (42). Metabolites of arachidonic acid, such as prostaglandins, also stimulate a Ca2+-dependent release of Glu by astrocytes (53). Moreover, HIV-1 can induce astrocytic expression of the P-chemokine known as macrophage chemotactic protein-1 (MCP-1). This P-chemokine, in turn, attracts additional mononuclear phagocytes and microglia to further enhance the potential for indirect neuronal injury via release of macrophage toxins (54).
Existing evidence suggests that HIV-1 infection and its associated neurological dysfunction involve both chemokine receptors and NMDAR-mediated excitotoxicity. This dual-receptor involvement raises the question of whether G-protein-coupled chemokine receptors and ionotropic glutamate receptors might influence each other's activity. Indeed, the P-chemokine known as "regulated upon activation T-cell expressed and secreted" (RANTES), which binds to the chemokine receptors CCR1, CCR3, and CCR5, can abrogate neurotoxicity induced by gp120 (17) or by excessive NMDAR stimulation (55). In turn, excitotoxic stimulation can enhance expression of CCR5 (56). Whether or not these findings reflect a mechanism of feedback or crosstalk of these receptors remains to be elucidated.
Was this article helpful?