Harris A Gelbard

Human immunodeficiency virus type 1 (HIV-1) is a member of the lentiviral family that includes simian immunodeficiency virus, visna virus, and feline immunodeficiency virus. All lentivirus infections are characterized by multiorgan disease, including infection and inflammation of brain tissue to varying degrees. Current estimates suggest there are approximately 21.8 million individuals worldwide infected with HIV-1, and of this group, at least one million are children. Furthermore, HIV-1-associated dementia (HIV-D) is the most frequent cause of neurologic disease in young adults in the U.S.1 HIV-D is estimated to occur at an annual rate of 7% in people with AIDS.2 HIV-1 replicates continuously through the asymptomatic and symptomatic phases of infection. This poses a special problem for the central nervous system (CNS), because new antiretroviral therapies and treatment for opportunistic infections have lengthened the survival time, but have not eradicated the virus from the CNS. This is largely due to impaired immune surveillance and poor penetration of antiretroviral agents through the blood-brain barrier (BBB) and the blood-cerebrospinal fluid (CSF) barrier.

HIV-1 infection of the CNS occurs early in the course of disease. Neurologic disease, including dementia, occurs relatively late in the course of infection and is usually associated with immunosuppression. HIV-D may be associated with the neuropathologic correlates of HIV-1 encephalitis (HIVE).3 HIV-1 productively infects brain-resident macrophages and microglia.4,5 In the absence of opportunistic infections of brain, features of HIVE include microglial activation, multinucleated giant cells, astrocytosis, and myelin pallor (decreased staining for myelin). HIV-1-infected children (usually through vertical transmission) have the most striking evidence of this disease complex.5 Using polymerase chain reaction/in situ hydridization techniques, no productive or latent HIV-1 infection of endothelial cells or oligodendrocytes has been demonstrated in the CNS.6

However, HIV-1 can "restrictively" infect astrocytes in children and adults; that is, only regulatory gene products such as Tat and nef are made without production of progeny virus.7,8,9 The actual percentage of restrictively infected astrocytes in brains of patients with HIVE is unknown, but thought to be relatively small. The pathophysiologic significance of restricted infection in astrocytes remains unknown.

Productive infection of neurons with HIV-1 has never been demonstrated. Nuovo et al. demonstrated latent proviral HIV-1 infection of neurons in brain tissue from a single patient with severe HIV-D of 5 years' duration using in situ polymerase chain reaction (ISPCR).10 However, using similar techniques, Takahashi et al. failed to demonstrate HIV-1 infection in neurons from patients with HIV-1-associated dementia.6 Nevertheless, it is accepted that there is focal neuronal loss in retina, neocortex, and in subcortical brain regions, including putamen, substantia nigra, and cerebellum. There is also a decrease in synaptic density and vacuolation of dendritic spines in affected brain regions.11,12,13,14,15 However, quantitative analysis of neuronal loss from postmortem studies of patients with HIV-1 and neurologic dysfunction has not demonstrated a clear correlation between the magnitude of neuronal loss and neurologic disease.16

As previously noted, cytolytic infection of neurons by HIV-1 is highly unlikely. Thus, changes in neuronal architecture, signal transduction, and number that underlie neurologic disease are likely due to indirect mechanisms; that is, soluble, diffusible factors, including HIV-1 gene products and cellular metabolites released from productively infected macrophages or microglia, present in focal inflammatory infiltrates, mediate dysfunction and death in vulnerable neurons. Numerous reports have demonstrated that HIV-1-infected macrophages and microglia produce soluble neurotoxic factors. Levels of some of these neurotoxins can markedly increase after antigenic stimulation.17,18 Neurotoxic factors include the HIV-1 coat protein gp120, gp41, and Tat,19,20,21 as well as cellular metabolites, including eicosanoids (i.e., arachidonic acid and its metabolites), the phospholipid mediator platelet activating factor (PAF), the proinflammatory cytokine tumor necrosis factor alpha (TNF-a), and an as yet unidentified NMDA receptor agonist, NTox, that has been tentatively identified as a phenolic amine with lipophilic properties that lacks the carboxyl groups of quinolinic acid.22,23,17,24 With the exception of NTox, most of these neurotoxins act by indirect mechanisms that involve dysregulation of a normal cellular process, such as arachidonic acid- and TNF-a-mediated decreases in glutamate uptake in affected astrocytes or gp120-mediated increases in glutamate efflux from affected astrocytes. All of these effects may then result in excess levels of glutamate and overstimulation of NMDA and non-NMDA receptors on vulnerable neurons.25,26 Dysregulation of high-affinity glutamate uptake in astrocytes could be further amplified by PAF because microglial PAF receptor activation can result in arachidonic acid release.27 Other indirect mechanisms of neurotoxicity include: activation of autocrine loops in adjacent uninfected macrophages or microglia, such as gp41- and gp120-mediated release of arachidonic acid, cytokines, and nitric oxide;28,29 and immune activation of uninfected monocytes by proinflammatory cytokines or antigenic stimulation to produce the quinolinic acid, which in turn can directly activate NMDA receptors.18 Furthermore, glucocorticoids, adrenal hormones released during stress, can potentiate gp120-induced neurotoxicity.30

Tat, TNF-a, and PAF may also have direct effects on vulnerable neurons. Tat has been shown to induce hippocampal CA1 neuron depolarization, probably via a non-NMDA receptor mechanism,31 and induce neuronal apo-ptosis.32 TNF-a can induce neuronal cell death, in part by activation of non-NMDA (i.e., AMPA) receptors.33 This TNF-a-induced neuronal death has the biochemical and morphologic features of apoptosis, involves oxidative stress, and is independent of nuclear factor kappa B (NFkB) activation.34 PAF can also induce neuronal cell death via a mechanism that involves activation of NMDA receptor channels,17 and results in neuronal apoptosis.35 Because these three neurotoxins appear to work in part through activation of non-NMDA and NMDA receptors, it is important to note that NMDA receptor activation and production of nitric oxide (potentially mediated via gp120) can also induce neuronal apoptosis.36,37

The hypothesis that a lentiviral infection such as HIV-1 causes cell loss and subsequent tissue atrophy in the brain and the immune system by inappropriate expression of genes involved in programmed cell death (PCD) was first promulgated in 1990.38 In particular, the authors speculated that HIV-1 is able to modify normal inter- and intracellular signaling to induce pathologic programmed cell death of CD4+ lymphocytes. This hypothesis was confirmed in two subsequent studies.49,40 A later study demonstrated that CD4+ and CD8+ lymphocytes from patients with both asymptomatic and symptomatic HIV-1 infection have increased levels of reactive oxygen species and a decreased mitochondrial transmembrane potential, which appears to be an early, irreversible step in activation of programmed cell death.41

As new techniques became available to identify free 3'-OH ends of newly cleaved DNA in situ,42 coupled with the light microscopic identification of some of the morphologic hallmarks of apoptosis (i.e., chromatin condensation), determination of whether apoptosis of neural cells occurred in postmortem archival formalin-fixed brain tissue from patients with HIV-1 infection became feasible. Using the TUNEL (Terminal deoxynucleotidyl dUTP Nick End Labeling) technique, we demonstrated that apoptotic neurons were present in the cerebral cortex and basal ganglia of children that had HIVE and progressive encephalopathy (HIVE/PE).43 Double-labeling immunocytochemistry for the TUNEL reagent and the HIV-1 P24 antigen in brain tissue from children with HIVE/PE revealed a spatial association between apoptotic neurons and perivascular inflammatory cell infiltrates containing HIV-1-infected macrophages and multinucleated giant cells (Figure 5.1A-C).43 Quantitative morphometric analysis of apoptotic neurons


Montage of cerebral cortex from a pediatric patient with HIVE/PE stained with antisera to p24 (new fuchsin, red immunostain) and ApopTag (i.e., TUNEL) reagent (DAB, brown immun-ostain). Panel A shows migration of P24-positive perivascular, TUNEL-negative macrophages into brain tissue, Panel B shows an adjacent field of TUNEL-positive neurons (large arrows) and TUNEL-positive microglia (small arrows). Panel C is a higher power magnification of Panel B.

FIGURE 5.1 (continued)

present in basal ganglia from children with HIVE/PE revealed a 12-fold increase in apoptotic neurons relative to children that were seronegative for HIV-1 infection, and a 3-fold increase in apoptotic neurons relative to children that had HIV-1 infection without the neuropathologic features of HIVE, or a pre-mortem diagnosis of PE.44

However, apoptotic neurons were infrequently identified in a third of the cases from pediatric HIV-1 seronegative controls.43 HIV-1 negative control tissue ranged in age between 3 weeks and 16.5 years. Taken together, these studies suggested that neuronal apoptosis was unlikely to be associated with postnatal development, but instead may be the end result of a disease process such as HIVE. Importantly, neuronal apoptosis in the CNS of adult patients with HIV-1 infection was confirmed in four separate reports.45,46,47,48 Two reports demonstrate that neuronal apoptosis occurs in a small number of cases prior to the onset of clinical AIDS or dementia.5,38

Four studies noted the presence of apoptotic macrophages and microglia45,18,38,45 in the brains of patients with HIVE, but only one report45

(Petito et al., 1995) noted the presence of apoptotic astrocytes in 2/7 brains of patients with HIVE. This finding was only observed with a DNA poly-merase technique, not the TUNEL technique. Interestingly, only one in vitro study has demonstrated apoptosis of astrocytes after HIV-1 infection of primary human brain cultures.25

In an attempt to reconcile these discrepancies and to further investigate the fate of glial cells in the brains of pediatric patients with HIVE, we have recently analyzed the in situ expression of pro- (Bax) and anti-apoptosis (Bcl-


Human Bax expression in white matter containing inflammatory infiltrate from a pediatric patient with HIVE/PE. The chromagen is VIP (purple immunostain). Note that virtually all perivascular macrophages (also identified by CD68 immunostaining, data not shown) are strongly immunoreactive for cytoplasmic Bax.


Human Bax expression in white matter containing inflammatory infiltrate from a pediatric patient with HIVE/PE. The chromagen is VIP (purple immunostain). Note that virtually all perivascular macrophages (also identified by CD68 immunostaining, data not shown) are strongly immunoreactive for cytoplasmic Bax.

2) gene products in cerebral cortex and basal ganglia.29 Markedly elevated numbers of microglia, and macrophages immunoreactive for bax were present in the pons, basal ganglia and cerebral cortex of children with HIVE/PE (Figure 5.2), in comparison to HIV-1 infected children and HIV-1 seronegative children. Marked increased numbers of Bax-positive microglia were also observed in an adult brain with HIVE.29 In contrast, astrocytes in brain tissue from patients with HIVE or HIV-1 had little or no Bax immun-ostaining.

Additional findings from Krajewski et al.29 demonstrated that patients with HIVE/PE, but not HIV-1 or seronegative controls, had increased expression of Bcl-2 in reactive astrocytes in cortex and basal ganglia. Thus, the lack of TUNEL staining and the presence of increased expression of Bcl-2 suggest that astrocytes may be resistant to apoptosis in brain tissue that has the neuropathologic hallmarks of HIVE. In contrast, the findings of TUNEL staining, increased Bax expression, and decreased or absent Bcl-2 expression in brain-resident macrophages and microglia suggest that these cells are more prone to undergo apoptosis in patients with HIVE. These findings are consonant with a cellular defense mechanism to limit microglial activation, thereby decreasing the spread of productive HIV-1 infection in brain-resident macrophages and microglia in the CNS of children with HIVE. In support of this hypothesis, a recent in vitro study demonstrated that caspase inhibitors


Human Bcl-2 expression in white matter containing inflammatory infiltrate from the same pediatric patient with HIVE/PE (see Figure 5.2). The chromagen is VIP (purple immunostain). Note astrocytes (arrows; also identified by glial fibrillary acidic protein immunostaining, data not shown) are strongly immunoreactive for cytoplasmic Bcl-2.


Human Bcl-2 expression in white matter containing inflammatory infiltrate from the same pediatric patient with HIVE/PE (see Figure 5.2). The chromagen is VIP (purple immunostain). Note astrocytes (arrows; also identified by glial fibrillary acidic protein immunostaining, data not shown) are strongly immunoreactive for cytoplasmic Bcl-2.

actually stimulated HIV-1 production in activated peripheral blood mononuclear cells obtained from HIV-1-infected asymptomatic individuals.10

Curiously, TUNEL-positive neurons in basal ganglia and cerebral cortex of children with HIVE did not have cytoplasmic expression of Bax with double label immunocytochemistry. Neurons in tissue sections from basal ganglia and cerebral cortex of children with HIVE were not immunoreactive for Bax alone. Lack of neuronal Bax immunostaining in brain tissue from patients with HIVE may be due to fixation artifacts and suboptimal processing conditions secondary to delays in postmortem fixation. Because micro-glia and brain-resident macrophages were intensely immunoreactive for Bax in brains of patients with HIVE, a more likely explanation would be that neurons in the brains of patients with HIVE may undergo apoptosis secondary to dysregulation of other proapoptotic genes including Bak, Bcl-xs and Bad.43

Several studies have focused on neuronal changes that occur during SIV infection. One report demonstrated that hippocampal neuronal atrophy occurs in rhesus macaques as early as 3 months following SIV inoculation.31 In younger monkeys there was a positive association between a reduction in neuronal density and duration of infection. A more recent study by Adam-son et al.,1 using a neurovirulent strain of SIV, demonstrated a spatial association between apoptotic neurons and perivascular inflammatory cell infiltrates containing SIV-infected macrophages and multinucleated giant cells. These findings are consonant with previous reports in patients with HIVE.3,18 However, this report also noted glial cell apoptosis, in agreement with Petito et al.45

Neuronal apoptosis has also been demonstrated in a severe combined immunodeficiency (SCID) mouse model of HIV-1 encephalitis. Here HIV-1-infected monocytes were exogenously infected with a neurovirulent macrophage-tropic strain of HIV-1 (ADA) and stereotactically injected into brain parenchyma, resulting in microglial activation, astrocyte proliferation, and TUNEL-stained neurons.44 These findings were specific for HIV-1-infected monocytes, since stereotactically injected uninfected monocytes did not elicit the same inflammatory response or induce neuronal apoptosis.

In summary, the available data from in vitro, in vivo, and postmortem studies suggest that neurons in CNS with a lentiviral infection such as HIV-1 and SIV-1 undergo apoptosis as opposed to necrosis. Loss of vulnerable neurons by apoptosis is likely to be a gradual process, mediated by a complex, multifactorial network of proinflammatory mediators and excitotoxins that result in impaired glial-neuronal signaling, and ultimately neuronal death. With the development of in vivo models of HIVE such as the SCID mouse model,44 the relative pathophysiologic significance of the many HIV-1-induced neurotoxins in mediating neuronal dysfunction and death can be ascertained. This may in turn help us to design more rational therapeutic strategies to ameliorate the neurologic disease associated with HIV-1 infection of the CNS.

All About Alzheimers

All About Alzheimers

The comprehensive new ebook All About Alzheimers puts everything into perspective. Youll gain insight and awareness into the disease. Learn how to maintain the patients emotional health. Discover tactics you can use to deal with constant life changes. Find out how counselors can help, and when they should intervene. Learn safety precautions that can protect you, your family and your loved one. All About Alzheimers will truly empower you.

Get My Free Ebook

Post a comment