NeuroAIDS Vol. 2, No. 5, May 1999
Symposium report: HIV and the nervous system
F. Gonzalez-Scarano
University of Pennsylvania, Department of Neurology, Clinical Research Building, 415 Curie Blvd., Philadelphia, Pennsylvania 19104, United States
Address correspondence to:

Goal of the symposium

This symposium, held in Washington, D.C. April 14-16, 1999, was organized by the National Institute of Mental Health and the National Institute of Neurological Disorders and Stroke, in their continuing commitment to the study of neurological disorders in patients infected by HIV.

A major goal of this meeting was to bring together investigators who are working in the problems associated with AIDS in the nervous system and investigators who are familiar with many aspects of the immunological and cellular response in the CNS. In this regard, there were several excellent reviews highlighting the contribution of cellular immune reactions to CNS surveillance, and potentially describing how these interactions could contribute to neuroinvasiveness by HIV.

T-cell and monocyte trafficking into the perivascular space

Reviews by William Hickey (Dartmouth Medical School), Hugh Perry (University of Southampton), and Cedric Raine (Albert Einstein College of Medicine) highlighted the importance of T-cell and monocyte trafficking into the perivascular space, and the distinct role of the perivascular macrophage in initiating autoimmune or infectious processes. HIV appears to be brought into the CNS by circulating lymphocytes, and the first 'endogenous' cell that is infected is the perivascular cell or perivascular macrophage. Traditionally it is has been thought that astrocytes are the exclusive cells lining this space (Virchow-Robin space), but Hickey pointed out that microglial foot processes constitute between 15-20% of the lining, providing a window through which an HIV infection in the perivascular space could be rapidly transmitted to microglial cells. Furthermore, Hickey speculated that perivascular macrophages may re-enter the circulation, providing a mechanism for re-introducing virus from the CNS compartment to the periphery.

Parenchymal microglia

Data from one SIV model (Ken Williams, NERPRC Harvard Medical School) indicates that parenchymal microglia are not infected with HIV - only perivascular cells express viral antigen. This would contradict data that indicates that human microglia can be infected in vitro with many HIV isolates. However, Francoise Gray (Hospital Raimond Poincare, Garches, France) pointed out that - as judged by the expression of MHC and other activation markers -in patients with AIDS microglia are chronically activated, regardless of the presence of HIV encephalitis or clinical dementia. Thus, in the more chronic HIV infection, parenchymal microglia may be primed for infection.

Timing of HIV entry

In this regard, several speakers discussed the timing of HIV entry. Suzanne Gartner (Johns Hopkins) suggested that although HIV can infect the CNS at any point in the course of infection, such infections may be cleared, and establishment of a chronic or permanent infection may only occur later. Howard Gendelman (University of Nebraska) questioned how a chronic infection of microglia could be controlled such that dementia only occurs towards the later part of the infectious process. Howard Fox (Scripps Clinic) indicated that cytolytic T-cells have been identified within the CNS of SIV infected monkeys, and that they, or some other arm of the immune system, could control the microglial production of virions, limiting the spread of infection. Alternatively, some HIV strains may produce a chronic rather than a highly productive infection of microglia, leading to the same type of long-term infection.

Role of chemokines and their receptors

Several investigators discussed the role of chemokines and their receptors in many aspects of HIV infection and in other forms of CNS inflammatory disease. Richard. Ransohoff (Cleveland Clinic) described their expression in inflammatory diseases, particularly Multiple Sclerosis. He showed a surprising and intriguing finding that CCR3 is prominent in the spinal cord lesions in MS, which is somewhat different than lesions in the brain. This appears to be parallel to comments made by other speakers that inflammation in the spinal cord is different than in the brain. It was speculated that the different spinal cord pathology of HIV in the cord could relate to some of these differences in the inflammatory reaction. Joan Berman (Albert Einstein College of Medicine) discussed the role of MCP-1 induction by the viral protein tat in mediating transendothelial migration of monocytes, as well as the potential role of SDF-1a in both monocytic and lymphocytic infiltration via interaction with CXCR4.

Genetic data implicating a specific chimpanzee species

George Shaw (University of Alabama) presented genetic data implicating a specific chimpanzee species (p. troglodytes troglodytes) in the origin of HIV through cross-species transmission. In this model chimpanzees have adapted to slightly different SIVcpz strains over a long evolutionary period, and a further understanding of this relationship may help define the pathogenesis of HIV in humans. Shaw also presented data suggesting that recombination events in some HIV strains may indicate that cross-species introduction has occurred at several points, even in relatively recent history.

HIV tropism

HIV tropism and its relationship to CNS pathology and neuronal apoptosis was the subject of several talks. Viruses using CXCR4 as a coreceptor have been associated with neuronal apoptosis. This is consistent with data from several groups (discussed at the conference by Dennis Kolson from the University of Pennsylvania) indicating that neurons express CXCR4, and perhaps other chemokine receptors, on their surface.

Viral load and development of cognitive problems

A morning session was devoted to the role of viral load in the development of cognitive problems. Although data regarding viral load in the brain parenchyma is still limited in the number of subjects who have been studied, overall viral load in the CSF appear to relate to CNS disease, but these data are neither sufficiently sensitive nor specific to be useful diagnostically. Nevertheless, persistently elevated HIV RNA levels in the CSF are a clear risk factor for dementia. In children, Pim Browers (Baylor College of Medicine) indicated that CSF viral load appears to correlate with cortical atrophy, a reversible finding in children.

Although the CSF is easier to sample, the parenchymal compartment may be more important in terms of cognitive function. However, investigators from Johns Hopkins (Justin McArthur and Suzanne Gartner) pointed out that HIV encephalitis is a focal disease, and that different regions may have varying levels of HIV replication. This must be taken into account when designing studies to look at viral load in brain.

Neuronal dysfunction

Neuronal dysfunction appears to be the ultimate outcome of HIV infection of the brain, and although it is widely accepted that something produced by macrophages/microglia is responsible for neuronal dysfunction, the exact mechanism is far from understood. Avi Nath (University of Kentucky) discussed the existence of 'virotoxins' - proteins like tat and gp120 that have been implicated in neurotoxiticity. Endogenous macrophage/microglial products like cytokines have also long been suspected of playing a role in neuronal death. Howard Gendelman presented data from investigators from his group indicating that such macrophage supernatants indeed can affect neurophysiological function. In a related portion of his presentation, Hugh Perry suggested that axonal damage, a pathological finding that has been found in association with HIV encephalitis by Francoise Gray and others, could be the result of an active rather than a passive process, akin to apoptosis.


The last session of the meeting was devoted to therapeutics. William Powderly (Washington University School of Medicine) summarized the current status of systemic therapy, and pointed out areas where research is still needed, like the clearance of virus from compartments where macrophages (and microglia) are most abundant. Clearly the CNS is one of those important regions. Thus far the use of protease inhibitors has been clearly associated with increased long-term survival, although some regimens that do not include protease inhibitors may also prove effective. This could be important from the standpoint of the CNS, since in general protease inhibitors penetrate the blood-brain-barrier poorly. During that same session, David Clifford (Washington University School of Medicine) and Stuart Lipton (Harvard Medical School) discussed CNS specific treatments, i.e. those treatments primarily intended to deal with neurological complications in conjunction with anti-retroviral therapy.


There were many other oral and poster contributions that contributed to the success of the meeting. All of the abstracts, several of the presentations and more detailed summaries about the sessions will be available in a forthcoming issue of the Journal of Neurovirology. An informal survey of the participants indicated that the meeting highlighted important issues in HIV, and that it should help focus research into key areas like viral load and the relationship to cognitive deterioration, and the role of chemokines and chemokine receptors in the development of neuronal dysfunction.

Copyright © 2001 by The American Association for the Advancement of Science