Molecular mimicry between HIV-1 gp41 and an astrocyte isoform of alpha-actinin

A 100-kDa astrocyte antigen previously shown to cross-react with a monoclonal antibody (MAb) generated against amino acids (aa) 598 to 609 of the transmembrane protein gp41 of human immunodeficiency virus type 1 [HIV-1] has now been molecularly characterized and found to be an alpha-actinin (alpha-actinin) related protein. Western blot analyses of human astrocytoma cells fractionated by differential centrifugation and detergent phase separation showed that the antigen was membrane associated. The astrocyte protein was purified to apparent homogeneity by immunoaffinity chromatography. Amino acid analysis of three peptide fragments obtained by cleavage of the purified 100-kDa protein revealed sequence identities of 77, 83 and 100% to a non-muscle isoform of human alpha-actinin. In addition, the aa 598-609 sequence of gp41 recognized by MAb 781.4, and the aa 581-597 sequence recognized by another cross-reactive MAb 781.3, were 73% and 53% similar to regions of alpha-actinin. This molecular mimicry between gp41 and alpha-actinin was supported by antibody cross-reactivity in Western immunoblot and ELISA analyses. Both anti-gp41 and anti-alpha-actinin MAbs bind to the surface of the human astrocytoma cells as detected by a cell surface binding assay and immunofluorescence. Antibodies made against this immunodominant region of gp41 in the serum and CSF of HIV-infected individuals have access to astrocytes within the CNS. The identification of the astrocyte antigen as an alpha-actinin related protein will allow further work to determine how this immunological cross-reactivity could perturb astrocyte function and contribute to HIV neuropathology.

Immunocytochemical quantitation of human immunodeficiency virus in the brain: correlations with dementia

The pathogenesis of human immunodeficiency virus (HIV)-associated dementia is unclear, and the underlying pathological substrate has been a matter of debate. In a prospectively clinically characterized population of acquired immunodeficiency syndrome (AIDS) patients we investigated the relationship between the clinical syndrome of HIV-associated dementia and the presence and relative quantity of immunocytochemical markers for HIV-1 (gp41 antibody), and for macrophages and microglia (HAM-56 antibody). Sections from the basal ganglia and frontal lobes from the brains of 51 patients were studied, and the data were stratified for severity of dementia (16 nondemented, 12 mildly demented, 23 severely demented), rate of dementia progression, duration of AIDS, use of antiretrovirals, and several other demographic features. We found a highly significant correlation between the degree of macrophage staining and the severity of dementia but only a borderline correlation between the presence and amount of gp41-positive cells and dementia. Several nondemented patients showed abundant gp41 immunoreactivity, and some severely demented showed little to no gp41 immunoreactivity. Other correlations with the immunostaining data, including antiretroviral use, were not significant. We conclude that the presence of macrophages and microglia is a better correlate with HIV-associated dementia than is the presence and amount of HIV-infected cells in the brain. These data support the concept that the pathogenesis of HIV-associated dementia is likely due to indirect effects of HIV infection of the brain, possibly through the actions of macrophages and microglia.

Effects of fixation and paraffin embedding on the immunohistological detection of cell-associated HIV-1 by different monoclonal antibodies

This study evaluates a panel of monoclonal antibodies (MAbs) to HIV-1 antigens (DuPont anti-gp120, gp41, p24; Olympus anti-gp120/160, gp41, p24A, p24B, p55, p18A, p18B, reverse transcriptase) for their ability to detect the virus in tissues after exposure to various fixatives (100% acetone, 10% formaldehyde, 2.5% glutaraldehyde, 4% paraformaldehyde/1% glutaraldehyde, Bouin's fluid) and after paraffin embedding. Acetone, 10% formaldehyde, and Bouin's fluid all preserved a wide range of viral epitopes compared with other fixatives. The most robust MAbs were DuPont p24 and Olympus p55, which produced excellent staining regardless of the fixative used. Embedding in paraffin variability influenced the capacity of MAbs to detect HIV-1 epitopes on fixed cells. Certain antibodies (e.g., DuPont gp24, Olympus p24B) produced good staining, whereas other epitopes (e.g., DuPont gp120, formaldehyde) were destroyed. In some cases, paraffin embedding revealed antigenic sites that had been formerly masked (e.g., Olympus gp120 and p24A; formaldehyde and glutaraldehyde fixation). These results indicate that HIV-1 antigens can be detected by immunohistology on cells exposed to most common fixatives. Therefore, retrospective analysis of pathological material is possible, provided that the antibodies are matched to the fixative used to preserve the tissue.

Determinants of human immunodeficiency virus type 1 envelope glycoprotein oligomeric structure

Oligomerization of the human immunodeficiency virus type 1 envelope (env) glycoproteins is mediated by the ectodomain of the transmembrane glycoprotein gp41. We report that deletion of gp41 residues 550 to 561 resulted in gp41 sedimenting as a monomer in sucrose gradients, while the gp160 precursor sedimented as a mixture of monomers and oligomers. Deletion of the nearby residues 571 to 582 did not affect the oligomeric structure of gp41 or gp160, but deletion of both sequences resulted in monomeric gp41 and predominantly monomeric gp160. Deletion of residues 655 to 665, adjacent to the membrane-spanning sequence, partially dissociated the gp41 oligomer while not affecting the gp160 oligomeric structure. In contrast, deletion of residues 510 to 518 from the fusogenic hydrophobic N terminus of gp41 did not affect the env glycoprotein oligomeric structure. Even though the mutant gp160 and gp120 molecules were competent to bind CD4, the mutations impaired fusion function, gp41-gp120 association, and gp160 processing. Furthermore, deletion of residues 550 to 561 or 550 to 561 plus 571 to 582 modified the antigenic properties of the proximal residues 586 to 588 and the distal residues 634 to 664. Our results indicate that residues 550 to 561 are essential for maintaining the gp41 oligomeric structure but that this sequence and additional sequences contribute to the maintenance of gp160 oligomers. Residues 550 to 561 map to the N terminus of a putative amphipathic alpha-helix (residues 550 to 582), whereas residues 571 to 582 map to the C terminus of this sequence.

An approach for an immunoaffinity AIDS sensor using the conservative region of the HIV envelope protein (gp41) and its monoclonal antibody

A monoclonal antibody for the conservative region of gp41, which is one of the HIV envelope proteins, was produced. The antigen determining site of gp41 was examined using the epitope mapping technique, followed by an enzyme linked immunosorbent assay. Some peptides had comparable affinities for the monoclonal antibody, but the peptide EGIEE, having a slightly weaker immunoaffinity than gp41, was the most preferable for the construction of an immunoaffinity AIDS sensor. For the detection of gp41, EGIEE was labelled with catalase and used as a mimic antigen; it was bound to the antibody present on an immuno-membrane and, due to the replacement reaction of the mimic antigen by gp41, indirect quantitative measurement of gp41 was possible using an oxygen electrode. Anti-gp41 antibody was also detected using a mimic antibody, which was chemically modified with polyethylene glycol. An immunoaffinity AIDS sensor was constructed using the mimic molecules which were tailored to have a suitable immunoaffinity for HIV antigen and/or antibody.

Cross-reactivity of anti-human immunodeficiency virus type 1 gp41 antibodies with human astrocytes and astrocytoma cell lines

An antigen expressed by astrocytes in human brain tissue and by various human astrocytoma cell lines was shown to cross-react with a monoclonal antibody generated against amino acids (aa) 584 to 609 of the transmembrane protein gp41 of human immunodeficiency virus type 1 (HIV-1). This region is an immunodominant segment of gp41, and high levels of antibodies against this epitope have been detected in both serum and cerebrospinal fluid of HIV-infected individuals at all stages of HIV infection. Immunohistochemistry with this monoclonal antibody demonstrated the presence of a cross-reactive antigen in human brain tissue, with an increased frequency and intensity of staining in HIV-positive individuals when compared with HIV-negative controls. By using a panel of HIV-positive and -negative sera, we show that antibodies in HIV-positive serum specifically bound to the surfaces of human astrocytoma cells. HIV-positive sera depleted of antibodies recognizing gp41 aa 584 to 609 showed a significant diminution in cell surface binding. Conversely, the serum antibodies that bound to and were eluted from the aa 584 to 609 peptide also bound to the astrocyte cell surface. To identify the target antigen, the immunoreactivity of three astrocytoma cell lines was examined. By immunoprecipitation of metabolically labeled cell lysates and Western blot (immunoblot) analysis, we identified a protein of approximately 100 kDa as the target antigen. Cross-reactive antibodies between HIV proteins and astrocyte epitopes, such as this 100-kDa protein and others previously reported, suggests that an autoimmune response against these target antigens may disrupt the normal functions of astrocytes.

Activation of microglia in HIV-1 infected brains is not dependent on the presence of HIV-1 antigens

The activation pattern of microglia in the cerebral cortex of AIDS patients with the neuropathological diagnosis of HIV-1 encephalitis was investigated by immunohistochemistry and morphometry. The number of activated microglial cells in the grey and white matter of five cortical regions was determined. In the grey and white matter of all cortical regions a significant increase in the number of microglial cells was demonstrated in HIV-1 infected brains. Moreover, the activation of microglia was not correlated with the presence of HIV-1 antigen in the brain region. The data show a significantly increased number of microglia in HIV-1 infected brains. These activated microglial cells could, among others, be those cells producing cytotoxic factors which, in turn, cause brain damage.

Brain viral burden in HIV infection

We quantitated the brain viral burden in autopsy material from AIDS patients with and without HIV encephalitis. Central nervous system (CNS) samples from 45 AIDS autopsies with less than 48 hours postmortem autolysis and without significant non-viral opportunistic infections were analyzed using immunocytochemistry (ICC), antigen capture assay (ACA) and polymerase chain reaction (PCR). Approximately three-quarters of the cases contained HIV DNA by PCR. The majority of these had abundant gp41 detected by ICC, but approximately one-third had no HIV p24 detected by ACA. With all assays, HIV proteins and DNA were most abundant in deep gray matter. Approximately one-quarter of the cases contained HIV p24 by ACA in both CNS tissue and cerebrospinal fluid. In more than half of the cases cytomegalovirus was detectable in the brain by PCR, however, only in the basal ganglia of one case was human herpes virus-6 detectable by PCR. In conclusion, HIV infection of the CNS was observed in the majority of AIDS autopsies, however, the quantity of virus was variable between cases and within different neuroanatomical regions. Differences between the quantitation methods could be either technical or biological, however, any of them could be used to compare assessment of HIV burden by different laboratories.

Measurement of CNS HIV burden and its association with neurologic damage

AIDS dementia complex affects between 20 and 30% of terminally ill AIDS patients. The histopathologic substrate of this clinical syndrome is chronic HIV infection of the nervous system termed HIV encephalitis. We hypothesize that the abundance and length of time that HIV is present in the CNS determines the severity of neurologic damage. We compared three different methods of assessing HIV burden in the CNS. While the assays vary in sensitivity, each provides a quantitative estimate of viral burden that can be compared between laboratories. Assays of provirus were readily divided into two groups, while protein assays showed a wider range. HIV-mediated neurotoxicity might be expected to more closely depend upon productive infection.

HIV-1 and HIV-2 isolates differ in their ability to activate the complement system on the surface of infected cells

OBJECTIVE

To analyse the ability of different HIV-1 and HIV-2 isolates to activate the complement system.

DESIGN

H9 cells chronically infected with various HIV isolates and the corresponding purified viruses were tested for complement activation. To identify the molecules responsible for complement activation on the surface of infected cells, the expression of complement inhibitors/regulators and viral proteins on the cell surface was analysed.

METHODS

C3 deposition on the cell surface and the expression of viral and cellular antigens were determined by flow cytometry analysis. Complement activation by purified viruses was measured using a complement consumption assay and a C1 activation assay.

RESULTS

H9 cells infected with different HIV-1 and HIV-2 isolates showed varying degrees of complement activation on the cell surface, ranging from strong activation and deposition of large amounts of C3 to no increased C3 deposition compared to uninfected cells. The C3 deposition was eliminated by EDTA and reduced in the presence of EGTA. In contrast, all purified viral isolates tested activated the complement system in a comparable manner. While the expression of MCP, DAF and CD59 was not modified after infection with different viral isolates, the reaction of the infected cells with a monoclonal antibody (3D6) directed against a gp41 epitope (amino acids 601-620) was found to correlate with the complement activation on the cell surface.

CONCLUSIONS

Some HIV-1 as well as HIV-2 isolates activate the complement system on the surface of infected cells independent of anti-HIV antibodies, while other isolates fail to do so. Complement activation on the cell surface is mediated by the alternative and, to a lesser extent, the classical pathway. The differences in complement activation on the cell surface are not caused by a modified expression of membrane-bound complement inhibitors/regulators. C3 deposition on the cell surface correlates with the expression of an epitope lying within the major complement activating domain of gp41 (amino acids 591-620). These results suggest a role of gp41 for complement activation on HIV-infected cells as has been described previously for purified HIV.

Quantitation of human immunodeficiency virus, immune activation factors, and quinolinic acid in AIDS brains

HIV encephalitis is unusual in that neurologic damage occurs in the absence of significant infection of neuronal or glial cells. Because the predominant infected cell in the brain is the macrophage, it has been proposed that release of viral or immune activation factors from macrophages may mediate neurologic damage. Numerous studies have examined the concentration of immune activation factors in the cerebrospinal fluid (CSF), however, there has been no correlation between these CSF measurements and severity of HIV encephalitis (Wiley, C.A., C.L. Achim, R.D. Schrier, M.P. Heyes, J.A. McCutchen, and I. Grant. 1992. AIDS (Phila.). 6:1299-1307. Because CSF measurements may not represent tissue concentrations of these factors, we examined the concentrations of HIV p24, quinolinic acid (QUIN), IL-1, IL-3, IL-6, TNF-alpha, and GMCSF within the brains of 10 AIDS autopsies. Homogenization and extraction of cortical gray, cortical white and deep gray matter showed a good correlation between the amount of HIV gp41 immunostaining and extracted HIV gag protein p24. The concentrations of cytokines were low in the tissue extracts and showed no correlation with severity of HIV encephalitis. Brain extracts from mild cases of HIV encephalitis showed elevated levels of TNF-alpha in deep gray matter, while in more severe cases, elevated TNF-alpha levels were also found within cortical white and cortical gray matter. Brain tissue and CSF QUIN concentrations were substantially increased compared to control values. QUIN concentrations were not correlated with the severity of HIV encephalitis. We conclude that increased tissue levels of TNF-alpha and QUIN may have a role in the etiology of HIV-related neurologic dysfunction.

Conversion of an immunogenic human immunodeficiency virus (HIV) envelope synthetic peptide to a tolerogen in chimpanzees by the fusogenic domain of HIV gp41 envelope protein

The fusogenic (F) domain of human immunodeficiency virus (HIV) gp41 envelope (env) protein has sequence similarities to many virus and mediates the fusion of HIV-infected cells. During a survey of the immunogenicity of HIV env peptides in chimpanzees, we have observed that HIV peptide immunogenicity was dramatically altered by the NH2-terminal synthesis of the gp41 F domain to an otherwise immunogenic peptide. We compared two hybrid peptide types comprised of T helper (Th) and B cell epitopes of HIV gp120 env protein for their immunogenicity in chimpanzees. The Th-B epitope hybrid peptides contained the HIV gp120 Th cell determinant, T1 (amino acids [aa] 428-440)-synthesized NH2 terminal to gp120 V3 loop peptides, which contain B cell epitopes that induce anti-HIV-neutralizing antibodies (SP10IIIB [aa 303-321] and SP10IIIB [A] [aa 303-327]). The F-Th-B peptide contained the HIV gp41 F domain of HIVIIIB gp41 (aa 519-530)-synthesized NH2 terminal to the Th-B peptide. Whereas Th-B peptides were potent immunogens for chimpanzee antibody and T cell-proliferative responses, the F-Th-B peptide induced lower anti-HIV gp120 T and B cell responses. Moreover, immunization of chimpanzees with F-Th-B peptide but not Th-B peptides induced a significant decrease in peripheral blood T lymphocytes (mean decrease during immunization, 52%; p < 0.02). Chimpanzees previously immunized with F-Th-B peptide did not respond well to immunization with Th-B peptide with T or B cell responses to HIV peptides, demonstrating that the F-Th-B peptide induced immune hyporesponsiveness to Th and B HIV gp120 env determinants. These observations raise the hypothesis that the HIV gp41 env F domain may be a biologically active immunoregulatory peptide in vivo, and by an as yet uncharacterized mechanism, promotes primate immune system hyporesponsiveness to otherwise immunogenic peptides.

Expression and antigenecity of human immunodeficiency virus type-1 transmembrane protein gp41 in insect cells

The HIV-1 transmembrane protein, gp41, is processed together with the envelope glycoprotein, gp120, from the same precursor, gp160, during the virus maturation. We used a baculovirus expression system to demonstrate that gp41 could be properly expressed without the preceding gp120 sequence. Two constructs with slight differences in the N-terminal region of gp41 were generated: one with a deletion of the first 7 hydrophobic residues of gp41, which have been suggested to be in a region important for membrane fusion and penetration, whereas the second with a complete sequence of gp41 except that a nonconserved leucine was substituted with a glutamine during DNA manipulation. Results from Western blotting with specific antisera confirm the gp41 identity. The sizes of gp41 were sensitive to tunicamycin treatment, indicating that N-linked glycosylation did occur. Further immunoblotting analyses with 90 different serum samples from HIV-1-infected individuals gave similar reaction patterns, suggesting that gp120 as well as the N-terminal region of gp41 are not critical for the expression and antigenecity of gp41. These eucaryotic constructs should provide valuable gp41 sources for detailed characterization of gp41 functions.

HIV envelope protein-induced neuronal damage and retardation of behavioral development in rat neonates

Cognitive and motor impairment are common symptoms among patients infected with the human immunodeficiency virus (HIV), including children who suffer neurological deficits and are frequently developmentally impaired. The HIV envelope protein, gp120, which has been shown to be toxic to neurons in culture, is shed in abundance by infected cells, and thus may play a significant role in the neuropathology of AIDS. To test this possible mechanism, neonatal rats were injected systemically with purified gp120 and the following consequences were observed: (1) radiolabeled gp120 and toxic fragments thereof were recovered in brain homogenates; (2) dystrophic changes were produced in pyramidal neurons of cerebral cortex; (3) retardation was evident in developmental milestones associated with complex motor behaviors. In parallel studies, co-treatment with peptide T, a gp120-derived peptide having a pentapeptide sequence homologous with vasoactive intestinal peptide, prevented or attenuated the morphological damage and behavioral delays associated with gp120 treatment. These studies suggest that gp120 and gp120-derived toxic fragments may contribute to the neurological and neuropsychiatric impairment related to HIV infection, and that peptide T appears to be effective in preventing gp120-associated neurotoxicity in developing rodents.

A human lymphoid recombinant cell line with functional human immunodeficiency virus type 1 envelope

Our goal has been to develop a safe and effective system that would allow us to explore the functions of the human immunodeficiency virus (HIV) envelope. We have generated a human lymphoid cell line (TF228.1.16) that stably expresses functional HIV envelope proteins on its cell surface, and therefore closely mimics the viral envelope and virus-infected cells. The TF228.1.16 line forms syncytia with human cells of the CD4+ phenotype and provides a facile virus-free cell-based assay for examining the mechanism of syncytia formation and for evaluating novel agents that may disrupt this process. The TF228.1.16 cells also provide an opportunity to present the HIV envelope proteins to the immune system in cellular form. In vitro immunization of human peripheral blood mononuclear cells (PBMC) and in vivo immunization of rhesus monkeys with this reagent results in the production of antibodies with neutralizing (anti-syncytia) activities. When the HIV envelope is expressed against the background of human lymphoid cells, it may exhibit immune protection with unique properties that have not yet been explored. Our results indicate that a virus-free cell system can play an important role in exploring the biology and function of HIV-envelope proteins without the interference of other viral components present in infected cells. This paper discusses these results, and examines the potential use of TF228.1.16 as a vaccine.

Host cell membrane proteins on human immunodeficiency virus type 1 after in vitro infection of H9 cells and blood mononuclear cells. An immuno-electron microscopic study

Human immunodeficiency virus type 1 (HIV-1)-infected H9 and blood mononuclear cells (MNCs) were studied by immunogold electron microscopy for the presence of HIV-1 gag p24 protein, env gp41 and gp120 proteins, and host cell molecules CD4, CD11a, CD25, CD54, CD63, HLA class I and HLA-DR. Uninfected H9 cells and MNC membranes labelled for CD4, HLA class I and class II, and, at low density, CD11a and CD54; lysosomal structures in the cytoplasm labelled for CD63. The infected cell surface showed immunolabelling for HIV-1 proteins, as did budding particle-like structures. Immunogold labelling of the cell membrane for CD4 was almost non-existent. The level of immunolabelling for CD11a and CD54 on infected cells was greater than that on uninfected cells; this is presumably related to a state of activation during virus synthesis. Budding particle-like structures and free virions in the intercellular space were immunogold-labelled for all host cell markers investigated. This was confirmed by double immunogold labelling using combinations of HIV-1 gag p24 labelling and labelling for the respective host cell molecule. We conclude that virions generated in HIV-1-infected cells concentrate host-derived molecules on their envelope. Also molecules with a prime function in cellular adhesion concentrate on the virion.

Selective neuronal vulnerability in HIV encephalitis

Recent studies of human immunodeficiency virus type 1 (HIV-1) encephalitis have shown that in addition to well established white matter damage, the neocortex shows thinning, loss of large neurons and dendritic damage. In order to identify neuronal populations affected in HIV encephalitis and to determine how neuronal damage relates to the severity of HIV infection within the nervous system, we quantified parvalbumin (PV+) and neurofilament (NF+) immunoreactive neurons in the frontal cortex and hippocampus. We found that in the neocortex, the density of NF+ and PV+ neurons was independent of severity of HIV encephalitis, and therefore changes in these neuronal subsets did not account for previously reported neuronal loss. However, neuritic processes of PV+ neurons were fragmented, atrophic and in some cases distended. In contrast to the frontal cortex, there was a trend toward decreased density of PV+ neurons in the hippocampus which only reached significance in the CA3 layer where there was a 50-90% decrease in PV+ neurons. This decrease was closely correlated with the severity of HIV encephalitis. Double-label immunocytochemical analysis confirmed neuritic damage to interneurons. These results suggest that HIV encephalitis differentially involves specific subpopulations of neurons. Since direct HIV infection of neuronal cells was not detected, damage to PV+ cells and fibers may be indirectly mediated by cytokines released by HIV-infected microglia.

Detection of HIV-1 protein and nucleic acid in enterochromaffin cells of HIV-1-seropositive patients

Diarrhea contributes significantly to the morbidity and mortality of patients with the acquired immunodeficiency syndrome (AIDS). Up to 50% of AIDS patients have diarrhea, and an etiologic agent for this cannot be identified in all of them. Recent evidence suggests that enterochromaffin cells may be infected by the human immunodeficiency virus type 1 (HIV-1) and may contribute to the unexplained diarrhea. To test this hypothesis further, endoscopic biopsies of duodena from 22 HIV-1 seropositive patients [17 with diarrhea (> 500 g/day and > 3 bowel movements/day), five without diarrhea] and from 15 normal controls (no HIV risk factors) without diarrhea were studied. Formalin-fixed and paraffin-embedded 5-microns sections were examined by immunocytochemistry, using a monoclonal antibody to the HIV-1 gp41 protein, and by in situ hybridization with a full-length biotinylated HIV-1 DNA probe. Positive staining for gp41 was detected in crypt cells, consistent with the location, size, and morphology of enterochromaffin cells, in 11 of 17 HIV-1-seropositive patients with diarrhea, and in none of five without diarrhea. Nucleic acid hybridization staining was performed in five of the 11 patients who had positive gp41 staining; all showed HIV nucleic acid sequences in similar cells. All three of the five patients with positive staining for HIV nucleic acid sequences had diarrhea for which no etiologic agent for diarrhea could be found, and one each had cryptosporidia or microsporidia. No staining was observed in any of the samples from normal control tissues. These results suggest that HIV-1 may infect enterochromaffin cells and possibly alter their function. This, in turn, may contribute to the diarrhea associated with AIDS.

Maternal antibodies to gp120 V3 sequence do not correlate with protection against vertical transmission of human immunodeficiency virus

A retrospective study of sera from mothers infected with human immunodeficiency virus (HIV-1) was undertaken to investigate whether the titers or affinities of antibodies against the third hypervariable region (V3 loop) of gp120 correlated with transmission of the virus from mother to child. The cohort comprised 7 mothers who transmitted HIV-1 to their children and 20 who did not. Sera were screened for reactivity against two synthetic peptides, one encompassing the entire V3 loop of gp120 (amino acids 297-330) and the other containing an immunodominant epitope from gp41 (amino acids 596-614). Doubling dilutions of sera were tested to obtain antibody titers against both peptides: Anti-gp41 titers were used to normalize the anti-V3 titers. Maternal sera were also screened for the presence of high-affinity antibodies against the V3 peptide. No differences were observed in either titers or affinities of maternal antibodies to the V3 sequence from transmitters and nontransmitters.

Spectrum of human immunodeficiency virus-associated neocortical damage

A spectrum of neurocognitive defects, termed human immunodeficiency virus type 1 (HIV-1)-associated cognitive/motor complex, has been described in patients with acquired immunodeficiency syndrome (AIDS). AIDS dementia complex (ADC) is a severe form of this disease seen in 20 to 30% of terminally ill patients. The etiology of this complex is distinct from commonly observed opportunistic infections seen in brains of patients with AIDS and has been attributed to HIV infection within the brain. At autopsy, the brains of patients with ADC contain numerous HIV-infected macrophages/microglia with prominent subcortical damage, together termed HIV encephalitis. We retrospectively analyzed all 107 brains from a three-year period (1988-1990) of AIDS autopsies using immunocytochemistry to detect HIV. Rather than breaking into distinct groups of HIV encephalitis versus non-HIV encephalitis, the specimens revealed a spectrum of severity of HIV infection. Although only 16% of the brains showed the histological hallmarks of HIV encephalitis, more than 50% of the autopsies showed moderate to severe HIV infection. In a subset of 23 AIDS autopsies during which short postmortem times and absence of significant opportunistic infection permitted quantitative analysis of dendritic and synaptic complexities, we identified a strong correlation between neocortical dendritic and presynaptic damage and abundance of HIV envelope protein in the neocortical gray and deep white matter. This correlation suggests that the presence of HIV-1 in the neocortex may be responsible by direct or indirect mechanisms for dendritic and synaptic damage.

HIV-1 in postmortem brain tissue from patients with AIDS: a comparison of different detection techniques

OBJECTIVE

The presence of HIV-1 in postmortem brain tissue from 31 patients with AIDS and 12 HIV-1-negative controls was investigated.

DESIGN

Most laboratories have access to the methods used. We readily applied in situ hybridization and immunohistochemistry to archival formalin-fixed paraffin-embedded (FFPE) brain specimens.

METHODS

The techniques used to detect HIV-1 were explant culture, in situ hybridization with 35S-labeled polymerase (pol) gene riboprobes and immunohistochemistry with monoclonal antibody to gp41.

RESULTS

HIV-1 was isolated from explant cultures in 13 out of 20 (65%) patients, whereas HIV-1-infected cells were detected in FFPE brain tissue from nine out of 26 (35%) patients examined by in situ hybridization and in seven out of 26 (27%) patients examined by immunohistochemistry.

CONCLUSIONS

Although the isolation technique was the most sensitive of the three techniques tested, infected cells may be identified with in situ hybridization in conjunction with immunohistochemistry.