You are on page 1of 8

Microbes and Infection 4 (2002) 301–308

www.elsevier.com/locate/micinf

Review

HIV-1 and the brain: connections between HIV-1-associated dementia,


neuropathology and neuroimmunology
Diane M. Lawrence *, Eugene O. Major
Laboratory of Molecular Medicine and Neuroscience, National Institute of Neurological Disorders and Stroke, National Institutes of Health,
Building 36, Room 5W21, 36 Convent Drive, MSC 4164, Bethesda, MD 20892-4164, USA

Abstract

AIDS patients frequently exhibit neurological disorders due to the neurotoxic events that result from HIV-1 and/or opportunistic
infections in the brain. This review examines recent clinical findings related to HIV-1-associated dementia, and outlines current areas of
basic research that may clarify how HIV-1-associated encephalopathy produces clinical symptoms of brain dysfunction. © 2002 Éditions
scientifiques et médicales Elsevier SAS. All rights reserved.

Keywords: AIDS dementia complex; HIV-1 encephalopathy; Viral encephalitis

1. Clinical consequences of HIV-1 infection in neuronal atrophy [3–5]. HIV-1-associated dementia is esti-
the brain mated to develop in at least 20% of advanced AIDS cases
[4]. Clinical neurological examination, brain imaging using
Following infection with HIV-1, several years may pass magnetic resonance imaging or X-ray-computed tomogra-
before the virus causes enough damage to produce the phy, and laboratory testing for the presence of HIV-1 and
clinical symptoms of immunosuppression known as AIDS. other infections in the blood and cerebral spinal fluid (CSF)
A wide range of neurological problems may be associated are used in combination for diagnosis [3]. Fig. 1 shows an
with AIDS, including cognitive impairments, motor distur- example of extensive damage to subcortical brain areas in a
bances, behavioral changes, headache, and peripheral neu- case of advanced HIV-1 dementia.
ropathy. These neurological symptoms can be the initial Evidence of neuropathology has been observed in a
manifestation of AIDS in many patients, and current obser- majority of adult AIDS patients, with a higher incidence
vations suggest that more than half of all current AIDS observed in pediatric cases [6]. The hallmarks of HIV-1-
patients will experience some form of neurological abnor- associated neuropathology include: infiltration of macroph-
mality [1]. Many factors may contribute to these symptoms, ages; the formation of microglial nodules and multinucle-
particularly opportunistic brain infections such as crypto- ated giant cells, suggestive of virus-induced fusion of
coccus, Toxoplasma gondii, JC virus (JCV), cytomegalovi- microglia and/or macrophages, in central white matter and
rus, Epstein-Barr virus, varicella zoster virus, and human deep gray matter; widespread reactive astrogliosis, indica-
herpesvirus type 6, which are associated with the immuno- tive of astrocyte activation and damage; the loss of specific
suppression resulting from HIV-1 infection [2]. In the neuron subpopulations, particularly those involved in cog-
absence of opportunistic infections, HIV-1-associated de- nition (hippocampus) and motor function (basal ganglia);
mentia may occur; this disorder is defined by severe the loss of synaptic connections; and myelin pallor, or the
impairment of specific cognitive functions, such as memory loss of myelin surrounding neuronal axons, indicating
recall and speed of information processing, and extensive damage to oligodendrocytes; and the presence of HIV-1
DNA in the CSF [7–9]. Collectively these inflammatory and
degenerative neuropathological events are termed HIV-1-
associated encephalitis, the biological correlate of the clini-
cal symptoms described above. Although severe HIV-1-
* Corresponding author. Tel.: +1-301-594-3225; fax: +1-301-594-5799. associated dementia is typically observed during the late
E-mail address: lawrencd@ninds.nih.gov (D.M. Lawrence). stages of infection, it is becoming increasingly evident that
© 2002 Éditions scientifiques et médicales Elsevier SAS. All rights reserved.
PII: S 1 2 8 6 - 4 5 7 9 ( 0 2 ) 0 1 5 4 2 - 3
302 D.M. Lawrence, E.O. Major / Microbes and Infection 4 (2002) 301–308

anti-HIV-1 drugs [12,13]. However, despite this improve-


ment, the occurrence of HIV-1-related encephalitis and
neuropathology in postmortem tissue has not changed [14],
suggesting that systemic therapeutic approaches are not
likely to eradicate the neurological consequences of HIV-1
infection. Because some of the antiviral drugs used in
HAART have limited access to the brain, it is likely that
HAART reduces the occurrence of severe neurological
outcomes by indirect means: the peripheral replication of
HIV-1 is blocked, so there is less of the virus available to
enter the brain, and there are fewer opportunistic infections
due to less advanced immunosuppression. Whereas some
believe that this approach to treatment is sufficient to
eliminate HIV-1-associated encephalopathy in the future,
there is evidence that HIV-1 can enter the brain soon after
infection [15,16], perhaps before treatment is begun. Fur-
thermore, up to 85% of the AIDS patients worldwide do not
have access to HAART, and among those patients who
begin and maintain the complex HAART regimen, systemic
HIV-1 infection is resistant to therapy in 20–50% of patients
[16]. Thus, while blocking viral entry into the central
nervous system is essential for the prevention of HIV-1-
associated neurological disorders, the current treatment
approach is not likely to be completely successful.
In order to develop better treatment options for HIV-1-
associated encephalopathy, it is critical to understand that
this disease is not a simple case of virus-induced cell
damage, but rather a chain of events that leads to neuronal
dysfunction. In other words, eliminating virus may not stop
the process of HIV-1-induced cell damage in the brain.
Fig. 1. Magnetic resonance image showing a horizontal section from the
brain of an HIV-1 dementia patient. White areas represent subcortical Understanding the specific mechanisms linking HIV-1 in-
damage (lesions), associated with cognitive deficits. fection and neuronal damage has been far from simple.
Damage to critical populations of neurons in the brain is
moderate neuropsychological, neurophysiological and neu- generally considered the most direct cause of cognitive and
roanatomical abnormalities in the context of HIV-1- asso- motor disorders, yet HIV-1 does not productively infect
ciated encephalitis can be observed well before the end neurons. In addition, the level of viral gene product expres-
stages of AIDS [10]. These observations suggest that HIV-1 sion in the brain does not always correlate with neurological
encephalopathy is a gradual process that, in many patients, symptoms [17], and the distribution of virus in the brain is
may begin early in the course of infection. Recent guide- highly variable [18]. To better understand the basis of
lines for the diagnosis of mild cognitive impairment [11] are HIV-1-associated neurological disorders, it is necessary to
expected to provide a more thorough measure of the clinical examine how HIV-1 enters the brain, what cell types are
consequences of HIV-1 infection in the brain. infected, and what biological events occur in response to
The advent of antiretroviral therapy has brought great brain infection.
hope for prolonging the lives of HIV-1-infected patients,
and for reducing the immunosuppression responsible for the
susceptibility to opportunistic infections that are often the 2. Cell types involved in HIV-1 infection
direct cause of death in AIDS patients. In particular, highly
active antiretroviral therapy (HAART), a combination of at In order for HIV-1 to enter a cell efficiently, it must first
least three anti-HIV-1 drugs including protease inhibitors bind to a CD4 receptor, commonly found on T lymphocytes,
and reverse transcriptase inhibitors, has been most success- blood monocytes, and monocyte derivatives (macrophages).
ful in restoring lymphocyte function and suppressing viral Subsequent binding to one of a family of α- or β-chemokine
replication to nearly undetectable levels within weeks of co-receptors, such as CXCR4 and CCR5, respectively, is
treatment initiation. In two of the most recent studies, the crucial for viral entry. The cell types known to be suscep-
incidence of HIV-1-associated dementia was reduced by tible to HIV-1 infection include CD4+ dendritic cells,
nearly 50% in patients receiving HAART therapy compared lymphocytes and macrophages [19]. Dendritic cells are
to earlier cohorts of patients taking only one or two found in skin and mucosa, but can migrate into other areas;
D.M. Lawrence, E.O. Major / Microbes and Infection 4 (2002) 301–308 303

lymphocytes and macrophages are present in various tis- of increased virus levels in the CNS, which would result in
sues, including the brain [19]. Over the long course of both direct and indirect effects on brain function.
infection, HIV-1 undergoes dramatic changes, particularly Other resident brain cells may be infected by HIV-1, but
in the nucleotide sequence encoding envelope glycoproteins at an extremely low level. While most studies have found a
(gp120 and gp41, cleavage products of gp160). The differ- lack of infection of neurons, there are some reports of HIV-1
ences in the resulting envelope proteins determine which DNA and protein expression in neurons [25,26]. The pos-
co-receptors the virus uses to enter cells, and thus, which sibility of neuronal infection certainly would be an impor-
cells become infected. Initially, during a systemic infection, tant factor in neuropathology, and infected neurons, like
there is a predominance of the less virulent ‘R5’ virus, astrocytes, may provide a reservoir of virus with the
which enters cells using the CCR5 co-receptor (usually capacity for reactivation. Oligodendrocytes, cells that create
expressed by macrophages, monocytes and dendritic cells). the insulating myelin sheath critical for neuronal function,
During the later stages of disease, the more virulent ‘X4’ are believed to be minimally involved in HIV-1 dementia.
strains appear, using the CXCR4 co-receptor (expressed by However, the loss of these cells may be a significant factor
T lymphocytes, for example) [20]. in AIDS-related neurological disorders, particularly because
In the brain, the vast majority of infected cells that they are targets of HIV-1-induced cellular toxins and oppor-
actively produce HIV-1 particles are monocyte-derived tunistic infections, including JCV. Analysis of brain tissue
macrophages and brain-resident microglia [7,19]. HIV-1 is from an AIDS patient with progressive multifocal leukoen-
thought to enter the brain by the infiltration of infected cephalopathy showed the presence of proteins from both
monocytes, which then differentiate into macrophages. Viral HIV-1 and JCV [27]. Although the oligodendrocytes in
replication within these cells is the source of viral particles these brains did not appear to be infected with HIV, the
that can infect other cells, such as microglia, the brain- HIV-1 Tat protein and JCV genome were detected in these
resident macrophages. Several studies have shown substan- cells. Tat can upregulate the transcription of JCV genes,
tial differences in nucleotide sequence and biological char- which speeds the lytic infection of the target oligodendro-
acteristics of HIV-1 derived from brain and blood of the cytes, leading to demyelination and dysfunction of neurons.
same patients [21], indicating that the brain may harbor and Oligodendrocytes are also susceptible to damage by viral
allow virus modification for several years. The numbers of envelope proteins and cellular factors produced by infected
infected macrophages in the brains of patients with HIV-1- macrophages in the brain [28]. Thus, as cells critical for
associated dementia, as well as the total amounts of virus neuronal function, oligodendrocytes must be considered as
production within the brain tissues, are highly variable. factors in HIV-1-related neurological disease, particularly in
However, these numbers are generally too low to explain the context of opportunistic pathogens, such as infection
the extent of pathology observed, suggesting that the with JCV.
cellular responses to HIV-1 infection of brain macrophages
and microglia are the primary factors contributing to the
process of HIV-1 encephalitis, rather than viral replication 3. Viral and cellular sources of neuronal damage
and virus-mediated cell lysis [7–9].
In addition to macrophages and microglia, astrocytes and Because the low number of infected cells in the brain
capillary endothelial cells have been found to contain HIV-1 cannot explain the extent of damage observed in HIV-1
protein and/or DNA in the brains of HIV-1 patients [20,22]. encephalopathy, it is widely accepted that viral proteins
Studies of cultured human astrocytes have shown that shed by infected cells, as well as a variety of toxic products
HIV-1 infection is initially productive yet noncytopathic, secreted by activated cells (infected or uninfected), are the
and ultimately is converted to a latent stage with little major factors involved in the underlying neuropathology.
expression of viral protein [22]. Although the proportion of One well-studied viral protein with putative neurotoxic
HIV-1-infected astrocytes is very low, approximately 70% effects is the HIV-1 envelope glycoprotein, gp160. This
of the cells in the brain are astrocytes, so this cell population protein can be cleaved into two products that remain
may be a significant factor in HIV-1-mediated neuropatho- noncovalently associated: gp120 and gp41. The cleaved
genesis. Interestingly, astrocytes can be infected in the gp120 is soluble and can be shed from infected cells, and at
absence of the classical CD4 receptor [23], and it is unclear low concentrations, gp120 can damage neurons cultured
whether the CCR5, CXCR4, or other chemokine receptors from various regions of the brain [29,30]. For several years
are involved in HIV-1 entry in these cells. Astrocyte it was believed that this damage was a direct effect of gp120
function, critical for the survival of neurons, may be on the neurons, but recent evidence indicates that gp120-
impaired in the context of HIV-1 infection; in addition, mediated neurotoxicity depends on the presence of other
infected astrocytes can produce cellular factors that may cells, either macrophages, microglia, or astrocytes, suggest-
adversely affect neuronal survival [22]. Under certain con- ing an indirect mechanism involving toxic intermediates
ditions, latent HIV-1 infection of astrocytes can be reacti- [30]. When macrophages or microglia are stimulated with
vated, e.g. by an inflammatory response or contact with gp120, these cells produce inflammatory cytokines and
susceptible cells [24]. This activation could produce a wave arachidonic acid metabolites; the neurotoxic effects of these
304 D.M. Lawrence, E.O. Major / Microbes and Infection 4 (2002) 301–308

products are discussed below. Astrocytes exposed to gp120 [33]. While glutamate stimulation is necessary for neuronal
exhibit both cytoskeletal and functional changes, including function and development, excessive levels can affect the
induction of nitric oxide synthase [9]; these changes most homeostasis of the neuronal environment, particularly by
likely impair the ability of these cells to protect neurons altering the intracellular calcium concentration and thus
from damage. reducing the neuron’s capacity to produce action potentials
The gp41 molecule remains inserted in the membrane of and release neurotransmitters [22]. Normally astrocytes
infected cells, yet it has also been implicated in HIV-1 have ‘buffering’ functions to help maintain optimal extra-
dementia [31]. A recent report found that gp41 is directly cellular levels of calcium and glutamate. These functions
toxic to neurons, and that this toxicity was prevented when include calcium uptake, glutamate uptake and catabolism,
inducible nitric oxide synthase was not functional, suggest- and production of kynurienic acid, which blocks glutamate
ing that gp41-mediated induction of nitric oxide may be receptors. However, astrocytes are also susceptible to the
responsible for neuropathology in the HIV-1-infected brain toxic products of activated and/or HIV-1-infected macroph-
[31]. The HIV-1 transactivator protein, Tat, which is ac- ages, and their function may be altered by latent HIV-1
tively secreted by infected cells, has the capacity to induce infection and/or reactivation. Thus, astrocytic dysfunction
neuronal death both directly and indirectly [9,32]. The may play a major role in HIV-1 dementia by the loss of
mechanism of Tat-induced neurotoxicity involves prolonged neuroprotective functions [34,35].
increases in intracellular calcium, which stimulates the
production of toxic reactive oxygen intermediates and
caspase activation of the apoptosis pathway [9]. Tat may 4. The brain immune response in HIV-1 encephalitis
also induce neuronal death indirectly, by stimulating mac-
rophages to secrete matrix metalloproteinases, which induce One of the areas most intensely studied to elucidate the
neuronal apoptosis and are found in the brain of HIV-1 mechanisms of HIV-1 encephalitis is that of inflammatory
dementia patients [32]. cytokine production in the CNS. Whereas profound defi-
In addition to viral proteins, other sources of neuronal ciency of cellular immunity is the hallmark of AIDS, it is
damage in the infected brain include the cellular products of clear that inflammatory responses from activated monocytes
HIV-1-infected cells. Macrophages and microglia that are and lymphocytes occur during HIV-1 infection. Antibody
activated by infection secrete several products that can synthesis occurs systemically and in the brain early after
cause neuronal damage directly (typically by altering the HIV-1 infection, during the asymptomatic phase. Increased
intracellular calcium homeostasis), indirectly by stimulating levels of pro-inflammatory cytokines such as IL-1, IFN-γ,
uninfected macrophages and microglia (amplifying the and especially TNF-α, anti-inflammatory cytokines includ-
response), and/or indirectly by damaging or overstimulating ing TGF-β and IL-6, and soluble cytokine receptors are
other cell types in the CNS, such as astrocytes and oligo- found in the CSF of AIDS patients [36,37]. Importantly, the
dendrocytes. Activated macrophages and microglia, levels of cytokine production in the CNS correlate with the
whether infected or not, can produce nitric oxide, superox- severity of neurologic symptoms [36,37]. It is suspected that
ide anions, platelet activating factor, arachidonic acid me- activated macrophages are the source of these pro-
tabolites, matrix metalloproteinases, and glutamate receptor inflammatory cytokines in the CNS, but astrocytes, neurons,
agonists [7,9]. These factors are implicated not only in and endothelial cells, particularly when stimulated by acti-
HIV-1-associated encephalopathy, but also in other neuro- vated macrophages or HIV-1-Tat protein, have the capacity
degenerative diseases, such as Alzheimer’s disease and for cytokine production as well [9,36,37].
Parkinson’s disease. Nitric oxide has potent antiviral effects It is unclear exactly how these cytokines contribute to
and acts as a neurotransmitter at moderate levels, but at high brain pathology, but it has been shown that TNF-α and
concentrations it can stimulate a biochemical cascade that is IL-1β enhance HIV-1 replication through increased produc-
toxic to neurons [9]. Nitric oxide combined with superoxide tion and viral DNA binding of the NFjΒ transcription factor
anions can produce direct neurotoxicity, possibly due to in macrophages [37]. These same cytokines can activate
excess calcium influx into the neuron. Arachidonic acid, a HIV-1 replication in latently infected astrocytes [38]. Cy-
precursor of prostaglandins, leukotrienes, and thrombox- tokines are also known to stimulate macrophages and
anes, is involved in many biological functions including microglia to produce the cellular toxins described above
inflammation; in the brain, these products may also cause [9,36]. TNF-α can affect astrocytes by stimulating the
neuronal damage by NMDA receptor overstimulation. production of other cytokines and chemokines [36], and can
Glutamate-mediated neurotoxicity, or overstimulation of inhibit the ability of astrocytes to buffer extracellular
glutamate receptors including the NMDA subtype, is a glutamate [39]. Cytokines such as TNF-α and IL-1 in the
common basis for toxicity in many neurodegenerative brain can adversely affect oligodendrocyte function and
diseases. In the case of HIV-1 dementia, high levels of pituitary hormone secretion, as well as induce fever, nausea
quinolinic acid, a glutamate receptor agonist produced by and loss of appetite, explaining some of the more general
activated and/or HIV-1-infected macrophages, are found in symptoms associated with HIV-1-associated dementia [36].
the CSF and are associated with increased brain atrophy IFN-γ can stimulate macrophage production of quinolinic
D.M. Lawrence, E.O. Major / Microbes and Infection 4 (2002) 301–308 305

acid, which as described above can overstimulate and receptors on glial cells [20]. Fractalkine, a chemokine that is
destroy neurons via glutamate receptor activation [7]. Thus, highly expressed in the central nervous system, is upregu-
in addition to promoting the survival of the virus, cytokines lated in the context of HIV-1 encephalitis, and protects
in the HIV-1-infected brain serve as messengers between neurons from the toxic effects of HIV-1-Tat protein and
various cell types to amplify the production of molecules platelet activating factor [44]. Thus, some cytokines and
that are toxic to neurons. chemokines may provide neuroprotection by competing
Another aspect of the immune response currently con- with viral particles or viral envelope proteins for binding to
sidered to be very important in the process of HIV-1 target cells, as well as by preventing cytokine-mediated or
encephalitis is chemokine production in the brain. Chemok- cellular toxin-mediated effects in the HIV-1-infected brain.
ines, a family of small cytokines typically produced by
damaged or infected cells, work in a gradient fashion to
activate and attract particular subsets of cells from the 5. Conclusions
immune system to the site of damage. Cells expressing
receptors that recognize a specific chemokine respond to Subsequent to systemic viral infection in humans, HIV-1
low concentrations, and move in the direction of increasing clearly can invade the brain, where it has the capacity to
concentrations of the chemokine until the source is reached. induce inflammatory responses, neuropathology, and pro-
Chemokines can also increase adhesion molecules on vari- gressive symptoms of neurological disorders. There is
ous cell types, including the responding cells and the substantial evidence that neurons susceptible to damage in
endothelial cells of the blood–brain barrier, which helps the HIV-1 associated encephalopathy are rarely infected, and
responding cells enter the damaged tissue more easily from the numbers of infected macrophages and microglia as well
the blood. In some cases, an infected cell upregulates the as the amount of virus they produce are too low to explain
production of a particular chemokine in order to attract cells the widespread areas of brain pathology. It is thus apparent
that can bring a protective immune response into the tissue. that the cellular responses to infection, and the effects of
However, if the magnitude of this response is too great, these responses on different cell populations in the brain, are
damage to resident cells, such as neurons, can occur. the links between HIV-1 infection, HIV-1 encephalopathy,
Several studies have shown a specific upregulation of and the HIV-1-associated dementia complex. Fig. 2 depicts
monocyte chemoattractant protein-1 (MCP-1) in the CSF of what is currently understood about the complex and indirect
HIV-1 dementia patients [40,41] and in cultured astrocytes mechanisms of HIV-1-induced neuronal damage. Virus
and microglia exposed to the HIV-1 regulatory protein Tat enters the brain via infected monocyte infiltration; the
[41,42]. TNF-α is a known inducer of MCP-1, and may play monocytes differentiate into macrophages, and virus can
a role in the upregulation observed in HIV-1 dementia [40]. replicate and spread to other macrophages, microglia, and
Production of other chemokines, such as macrophage in- astrocytes. Viral proteins shed from infected cells may
hibitory protein (MIP)-1α, MIP–1β, or RANTES, has also directly affect neuronal function, or they can stimulate other
been detected in the CSF of HIV-1 dementia patients [40]. cells in the CNS to produce cytokines or toxic products.
In HIV-1 encephalopathy, it is likely that these chemokines These cellular products can cause damage to specific
contribute to pathogenesis by bringing into the brain in- neuronal populations, or can further stimulate different cell
creased numbers of infected and/or activated monocytes, types in the CNS, providing amplification of this toxic
another source of both viral and cellular toxins that affect inflammatory response. Some cells may respond by produc-
neuronal function and survival [41]. In addition, resident ing defensive molecules, providing a possibility of antiviral
brain cells, including neurons, express chemokine receptors or anti-inflammatory effects. Normally functioning astro-
and it is possible that chemokines act directly on neurons to cytes and oligodendrocytes act to promote neuronal func-
induce death [19]. tion and survival, but if these cells are damaged by the
Not all immune responses in the brain have damaging inflammatory response, their neuroprotective functions are
effects, however. In addition to the toxic effects of TNF-α lost.
described earlier, the same molecule may have neuroprotec- Taken together, it is clear that a critical balance exists in
tive effects under certain conditions [9]. TGF-β, produced in the HIV-1-infected brain. Virus replication and viral prod-
response to IL-1, is a classically ‘anti-inflammatory’ cytok- ucts, antiviral and neuroprotective responses, and damaging
ine that can stabilize calcium homeostasis and stimulate inflammatory responses all compete over the course of
anti-apoptotic pathways in neurons. Both TGF-β and IL-1, many years, and whichever of these three dominates deter-
with seemingly opposing actions, are upregulated in HIV- mines the outcome of infection. As discussed, actual viral
1-infected brains [36]. TGF-β-mediated prevention of replication seems relatively minor in the brain environment,
HIV-1 gp120-induced neuronal death was demonstrated in yet it is indeed this replication and the production of
mixed neuron–glia cultures [43]. The chemokines MIP-1α extracellular viral products that stimulates the cascade of
and RANTES, which can recruit inflammatory cells into the events that leads to both neuroprotection and neurotoxicity.
brain, also prevent gp120-induced neuronal death, most In addition, as the rapidly changing virus becomes more
likely due to competition with HIV-1 binding to chemokine virulent over time, HIV-1 replication in the brain may
306 D.M. Lawrence, E.O. Major / Microbes and Infection 4 (2002) 301–308

Fig. 2. Schematic of the processes and cell types that interact during HIV infection of the brain. Leukocytes in the bloodstream include monocytes (MO),
T lymphocytes (T), and B lymphocytes (B). HIV-1 infection is represented by a dotted pattern. Virus is transported into the brain via infected monocytes,
which differentiate into macrophages (MAC). HIV-1 can replicate and spread to other macrophages, microglia (MG), and astrocytes (A). 1. Viral products
shed from infected cells can be toxic to neurons (N), either directly or indirectly through stimulation of uninfected macrophages, microglia, or astrocytes.
2. Cellular products from either infected or uninfected but activated macrophages and microglia can stimulate and/or damage neurons, astrocytes,
oligodendrocytes (O), or other macrophages and microglia. Chemokine production, particularly from stimulated and/or infected astrocytes, may help recruit
more infected and/or activated monocytes into the brain tissue. 3. Astrocytes and oligodendrocytes provide neuroprotective functions, such as homeostasis
of calcium and glutamate levels, myelin production, and possible stimulation of anti-inflammatory molecules.

become more of a factor in the course of disease. Research perhaps regulate the immune response, such that the balance
in many areas of neurodegeneration has shown that some is swayed toward neuroprotection.
immune stimulation in the brain can be beneficial, both for
their antiviral effects and direct neuroprotective effects.
However, if the immune response is prolonged, or if Acknowledgements
infection induces a chemokine that can specifically attract
more damaging cells, the outcome is neuropathology that is The authors wish to thank Peter N. Jensen for helpful
potentially fatal. In HIV-1 encephalitis, the damage done by discussions during the preparation of this manuscript.
an excessive immune response in the brain contrasts with
the immunosuppression defining AIDS, and this contrast
highlights the complexities that must be understood in order References
to develop effective treatments for both. While the current
therapeutic approach to HIV-1 infection does seem to have [1] D.J. Lanska, Epidemiology of human immunodeficiency virus infec-
a beneficial effect for HIV-1 dementia, it is clearly not a tion and associated neurologic illness, Semin. Neurol. 19 (1999)
cure. A better understanding of the mechanisms in HIV-1- 105–111.
induced neuropathogenesis is critical for designing treat- [2] D.B. Clifford, Opportunistic viral infections in the setting of human
ment and prevention options that access the brain and immunodeficiency virus, Semin. Neurol. 19 (1999) 185–192.
D.M. Lawrence, E.O. Major / Microbes and Infection 4 (2002) 301–308 307

[3] R.S. Janssen, D.R. Cornblath, L.G. Epstein, R.P. Foa, J.C. McArthur, [25] G.J. Nuovo, F. Gallery, P. MacConnell, A. Braun, In situ detection of
R.W. Price, Nomenclature and research case definitions for neuro- polymerase chain reaction-amplified HIV-1 nucleic acids and tumor
logic manifestations of human immunodeficiency virus-type 1 necrosis factor-alpha RNA in the central nervous system, Am.
(HIV-1) infection, Neurology 41 (1991) 778–785. J. Pathol. 144 (1994) 659–666.
[4] J.C. McArthur, N. Sacktor, O. Selnes, Human immunodeficiency [26] O. Bagasra, E. Lavi, L. Bobroski, K. Khalili, J.P. Pestaner,
virus-associated dementia, Semin. Neurol. 19 (1999) 129–150. R. Tawadros, R.J. Pomerantz, Cellular reservoirs of HIV-1 in the
[5] B.J. Brew, AIDS dementia complex, Neurol. Clin. 17 (1999) central nervous system of infected individuals: identification by the
861–881. combination of in situ polymerase chain reaction and immunohis-
[6] D.M. Simpson, Human immunodeficiency virus-associated dementia: tochemistry, AIDS 10 (1996) 573–585.
review of pathogenesis, prophylaxis, and treatment studies of zidovu- [27] L.D. Valle, S. Croul, S. Morgello, S. Amini, J. Rappaport, K. Khalili,
dine therapy, Clin. Infect. Dis. 29 (1999) 19–34. Detection of HIV-1 Tat and JCV capsid protein, VP1, in AIDS brain
[7] H.E. Gendelman, S.A. Lipton, M. Tardieu, M.I. Bukrinsky, with progressive multifocal leukoencephalopathy, J. Neurovirol. 6
H.S.L.M. Nottet, The neuropathogenesis of HIV-1 infection, J. Leu- (2000) 221–228.
koc. Biol. 56 (1994) 389–398.
[28] A. Bernardo, C. Agresti, G. Levi, HIV-gp120 affects the functional
[8] J.E. Bell, The neuropathology of adult HIV infection, Rev. Neurol.
activity of oligodendrocytes and their susceptibility to complement,
154 (1998) 816–829.
J. Neurosci. Res. 50 (1997) 946–957.
[9] A. Nath, Pathobiology of human immunodeficiency virus dementia,
Semin. Neurol. 19 (1999) 113–127. [29] E.B. Dreyer, P.K. Kaiser, J.T. Offermann, S.A. Lipton, HIV-1 coat
[10] R.K. Heaton, I. Grant, N. Butters, D.A. White, D. Kirson, J.H. Atkin- protein neurotoxicity prevented by calcium channel antagonists,
son, J.A. McCutchan, M.J. Taylor, M.D. Kelly, R.J. Ellis, et al., The Science 248 (1990) 364–367.
HNRC 500—neuropsychology of HIV infection at different disease [30] M. Kaul, S.A. Lipton, Chemokines and activated macrophages in HIV
stages, J. Int. Neuropsychol. Soc. 1 (1995) 231–251. gp120-induced neuronal apoptosis, Proc. Natl. Acad. Sci. USA 96
[11] R.C. Petersen, J.C. Stevens, M. Ganguli, E.G. Tangalos, J.L. Cum- (1999) 8212–8216.
mings, S.T. DeKosky, Practice parameter: Early detection of demen- [31] D.C. Adamson, K.L. Kopnisky, T.M. Dawson, V.L. Dawson, Mecha-
tia: Mild cognitive impairment (an evidence based review), Neurol- nisms and structural determinants of HIV-1 coat protein, gp41-
ogy 56 (2001) 1133–1142. induced neurotoxicity, J. Neurosci. 19 (1999) 64–71.
[12] M. Maschke, O. Kastrup, S. Esser, B. Ross, U. Hengge, A. Hufnagel, [32] J.B. Johnston, K. Zhang, C. Silva, D.R. Shalinsky, K. Conant, W. Ni,
Incidence and prevalence of neurological disorders associated with D. Corbett, V.W. Yong, C. Power, HIV-1 Tat neurotoxicity is
HIV since the introduction of highly activated antiretroviral therapy prevented by matrix metalloproteinase inhibitors, Ann. Neurol. 49
(HAART), J. Neurol. Neurosurg. Psych. 69 (2000) 376–380. (2001) 230–241.
[13] N. Sacktor, R.H. Lyles, R. Skolasky, C. Kleeberger, O.A. Selnes,
[33] M.P. Heyes, R.J. Ellis, L. Ryan, M.E. Childers, I. Grant, T. Wolfson,
E.N. Miller, J.T. Becker, B. Cohen, J.C. McArthur, HIV-associated
S. Archibald, T.L. Jernigan, Elevated cerebrospinal fluid quinolinic
neurologic disease incidence changes: Multicenter AIDS Cohort
acid levels are associated with region-specific cerebral volume loss in
Study, 1990–1998, Neurology 56 (2001) (1990) 257–260.
HIV infection, Brain 124 (2001) 1033–1042.
[14] E. Masliah, R.M. DeTeresa, M.E. Mallory, L.A. Hansen, Changes in
pathological findings at autopsy in AIDS cases for the last 15 years, [34] L. Vitkovic, A. da Cunha, Role for astrocytosis in HIV-1-associated
AIDS 14 (2000) 69–74. dementia, Curr. Top. Microbiol. Immunol. 202 (1995) 105–116.
[15] F. Gray, F. Scaravilli, I. Everall, F. Chretien, S. An, D. Boche, [35] H.S. Nottet, M. Jett, C.R. Flanagan, Q.H. Zhai, Y. Persidsky,
H. Adle-Biassette, L. Wingertsmann, M. Durigon, B. Hurtrel, A. Rizzino, E.W. Bernton, P. Genis, T. Baldwin, J. Schwartz, A
F. Chiodi, J. Bell, P. Lantos, Neuropathology of early HIV-1 infection, regulatory role for astrocytes in HIV-1 encephalitis. An overexpres-
Brain Pathol. 6 (1996) 1–15. sion of eicosanoids, platelet activating factor, and tumor necrosis
[16] W.G. Powderly, Current approaches to treatment for HIV-1 infection, factor-alpha by activated HIV-1-infected monocytes is attenuated by
J. Neurovirol. 6 (suppl. 1) (2000) S8–S13. primary human astrocytes, J. Immunol. 154 (1995) 3567–3581.
[17] J.D. Glass, H. Fedor, S.L. Wesselingh, J.C. McArthur, Immunocy- [36] M. Yoshioka, W.G. Bradley, P. Shapshak, I. Nagano, R.V. Stewart,
tochemical quantitation of human immunodeficiency virus in the K.Q. Xin, A.K. Srivastava, S. Nakamura, Role of immune activation
brain: correlations with dementia, Ann. Neurol. 38 (1995) 755–762. and cytokine expression in HIV-1-associated neurologic diseases,
[18] C.A. Wiley, V. Soontornniyomkij, L. Radhakrishnan, E. Masliah, Adv. Neuroimmunol. 5 (1995) 335–358.
J. Mellors, S.A. Hermann, P. Dailey, C.L. Achim, Distribution of [37] D.E. Griffin, Cytokines in the brain during viral infection: clues to
brain HIV load in AIDS, Brain Pathol. 8 (1998) 277–284. HIV-associated dementia, J. Clin. Invest. 100 (1997) 2948–2951.
[19] R.J. Miller, O. Meucci, AIDS and the brain: is there a chemokine
[38] C. Tornatore, A. Nath, K. Amemiya, E.O. Major, Persistent human
connection? Trends Neurosci. 22 (1999) 471–479.
immunodeficiency virus type 1 infection in human fetal glial cells
[20] D. Gabuzda, J. Wang, Chemokine receptors and mechanisms of cell
reactivated by T-cell factor(s) or by the cytokines tumor necrosis
death in HIV neuropathogenesis, J. Neurovirol. 6 (2000) S24–S32.
factor alpha and interleukin-1 beta, J. Virol. 65 (1991) 6094–6100.
[21] J.K. Wong, C.C. Ignacio, F. Torriani, D. Havlir, N.J. Fitch, D.D. Rich-
man, In vivo compartmentalization of human immunodeficiency [39] S.M. Fine, R.A. Angel, S.W. Perry, L.G. Epstein, J.D. Rothstein,
virus: evidence from the examination of pol sequences from autopsy S. Dewhurst, H.A. Gelbard, Tumor necrosis factor alpha inhibits
tissues, J. Virol. 71 (1997) 2059–2071. glutamate uptake by primary human astrocytes: implications for
[22] R. Brack-Werner, Astrocytes: HIV cellular reservoirs and important pathogenesis of HIV-1 dementia, J. Biol. Chem. 271 (1996)
participants in neuropathogenesis, AIDS 13 (1999) 1–22. 15303–15306.
[23] F. Sabri, E. Tresoldi, M. DiStefano, S. Polo, M.C. Monaco, A. Verani, [40] W. Kelder, J.C. McArthur, T. Nance-Sproson, D. McClernon,
J.R. Fiore, P. Lusso, E. Major, F. Chiodi, G. Scarlatti, Nonproductive D.E. Griffin, β-Chemokines MCP-1 and RANTES are selectively
human immunodeficiency virus type 1 infection of human fetal increased in cerebrospinal fluid of patients with human immunodefi-
astrocytes: independence from CD4 and major chemokine receptors, ciency virus-associated dementia, Ann. Neurol. 44 (1998) 831–835.
Virology 264 (1999) 370–384. [41] K. Conant, A. Garzino-Demo, A. Nath, J.C. McArthur, W. Halliday,
[24] C. Tornatore, K. Meyers, W. Atwood, K. Conant, E. Major, Temporal C. Power, R.C. Gallo, E.O. Major, Induction of monocyte chemoat-
patterns of human immunodeficiency type 1 transcripts in human fetal tractant protein-1 in HIV-1 Tat-stimulated astrocytes and elevation in
astrocytes, J. Virol. 68 (1994) 93–102. AIDS dementia, Proc. Natl. Acad. Sci. USA 95 (1998) 3117–3121.
308 D.M. Lawrence, E.O. Major / Microbes and Infection 4 (2002) 301–308

[42] C.M. McManus, K. Weidenheim, S.E. Woodman, J. Nunez, J. Hes- [44] N. Tong, S.W. Perry, Q. Zhang, H.J. James, H. Guo, A. Brooks,
selgesser, A. Nath, J.W. Berman, Chemokine and chemokine-receptor H. Bal, S.A. Kinnear, S. Fine, L.G. Epstein, D. Dairaghi, T.J. Schall,
expression in human glial elements: induction by the HIV protein, H.E. Gendelman, S. Dewhurst, L.R. Sharer, H.A. Gelbard, Neuronal
Tat, and chemokine autoregulation, Am. J. Pathol. 156 (2000) fractalkine expression in HIV-1 encephalitis: roles for macrophage
1441–1453. recruitment and neuroprotection in the central nervous system,
[43] O. Meucci, R.J. Miller, gp120-Induced neurotoxicity in hippocampal J. Immunol. 164 (2000) 1333–1339.
pyramidal neuron cultures: protective action of TGF-β1, J. Neurosci.
16 (1996) 4080–4088.

You might also like