Neuronal apoptosis in human immunodeficiency virus infection

Neuronal apoptosis has been shown to occur in HIV infection by a number of in vivo and in vitro studies, however, the cause of neuronal damage in AIDS is still unclear and its relationships with the cognitive disorders characteristic of HIV dementia remain a matter of debate. In this review, based on our experience, we analyse the techniques used to identify neuronal apoptosis on post-mortem AIDS brains and describe the relationships of neuronal apoptosis with the stage of disease, a history of HIV-dementia, the degree of productive HIV infection, microglial activation, blood-brain barrier involvement and axonal damage. We conclude that the severity of neuronal apoptosis in the cerebral cortex correlates with the presence of cerebral atrophy, but not with the cognitive disorders. There is no global quantitative correlation between neuronal apoptosis and HIV encephalitis, microglial activation or axonal damage. However we found some topographical correlation between these changes. We conclude that neuronal apoptosis and consequent neuronal loss, in HIV infected patients, are probably not related to a single cause. It seems likely that microglial activation, directly or indirectly related to HIV infection of the CNS, plays a major role in its causation possibly through the mediation of oxidative stress. Axonal damage, either secondary to microglial activation, or to the intervention of systemic factors may also contribute to neuronal apoptosis.     

Programmed cell death in brains of HIV-1-positive AIDS and pre-AIDS patients

Neuropathological studies have revealed that the brains of HIV-1-infected AIDS patients show the typical encephalitis and, in addition, neuronal loss. More recently, this neuronal cell loss has been thought to take place via programmed cell death (apoptosis) which has been demonstrated by an in situ end labelling (ISEL) technique. In this study 54 brains of HIV-1-positive patients were investigated by the ISEL technique to investigate whether apoptosis is also present in the brains of patients at the asymptomatic stage. Of these, 10 patients suffered from HIV encephalitis (HIVE), 8 had AIDS without neuropathological disorders and 36 were HIV-1-positive pre-AIDS patients. Apoptotic cells were detected in 6 of the 10 HIVE, 1 of the 8 AIDS without central nervous system (CNS) disease and 4 of the 36 asymptomatic individuals. A difference seen between the AIDS and pre-AIDS cases was that, in the latter, apoptotic cells were found in the white matter in all 4 cases, while only 2 of these 4 showed apoptotic neurons. The presence of apoptotic cells in a number, albeit small, of brains of HIV-1-positive pre-AIDS individuals, combined with abnormalities described previously in the same group of patients gives further support to the opinion that brain damage already occurs during the early stages of HIV infection. Received: 19 June 1995 / Revised: 31 July 1995 / Accepted: 6 September 1995     

Neuropathology and neurodegeneration in human immunodeficiency virus infection. Pathogenesis of HIV-induced lesions of the brain, correlations with HIV-associated disorders and modifications according to treatments

A variety of HIV-induced lesions of the central nervous system (CNS) have been described, including HIV encephalitis, HIV leukoencephalopathy, axonal damage, and diffuse poliodystrophy with neuronal loss of variable severity resulting, at least partly, from an apoptotic process. However, no correlation could be established between these changes and HIV dementia (HIVD). From our study of HIV infected patients, it appeared that neuronal apoptosis is probably not related to a single cause. Microglial and glial activation, directly or indirectly related to HIV infection, plays a major role in neuronal apoptosis possibly through the mediation of oxidative stress. In our patients with full-blown AIDS, this mechanism predominated in the basal ganglia and correlated well with HIVD. Axonal damage, either secondary to microglial activation, or to systemic factors also contributes to neuronal apoptosis. Although massive neuronal loss may be responsible for HIVD in occasional cases, we conclude that neuronal apoptosis is a late event and does not represent the main pathological substrate of HIVD. The dementia more likely reflects a specific neuronal dysfunction resulting from the combined effects of several mechanisms, some of which may be reversible. Introduction of highly active antiretroviral therapy dramatically improved patient survival, however, its impact on the incidence and course of HIVD remains debatable. In our series, the incidence of HIVE has dramatically decreased since the introduction of multitherapies, but a number of cases remain whose cognitive disorders persist, despite HAART. The poor CNS penetration of many antiretroviral agents is a possible explanation, but irreversible "burnt out" HIV-induced CNS changes may also be responsible.     

Neuronal apoptosis in HIV infection in adults

Productive infection of the central nervous system by HIV predominantly involves the white matter and basal ganglia. Involvement of the cerebral cortex with neuronal loss is also described in AIDS patients but not in asymptomatic HIV-positive patients. The mechanism of neuronal damage is unknown. To enquire whether neuronal loss in AIDS may be due to an apoptotic process, we examined the cerebral cortex from 12 patients who died from AIDS using two different methods: in situ end labelling and gel electrophoresis of DNA to demonstrate DNA fragmentation. None of the patients had cerebral opportunistic infection or tumour. Four patients had no significant neuropathological changes, eight patients had variable cerebral atrophy and four of them also had productive HIV infection of the brain. These patients were compared with four HIV-positive asymptomatic patients, five seronegative asymptomatic controls, and two seronegative patients with Alzheimer's disease. We demonstrated neuronal apoptosis in the cortex in all AIDS patients, as well as in the Alzheimer's patients. Apoptosis was not observed in the asymptomatic cases whether seropositive or seronegative. Neuronal apoptosis was more severe in atrophic brains, and did not directly correlate with productive HIV infection, suggesting an indirect mechanism of neuronal damage is most likely.     

Expression of excitatory amino acid transporter-1 in brain macrophages and microglia of HIV-infected patients. A neuroprotective role for activated microglia?

Recent experimental studies showed that activated macrophages/microglia (AMM) express excitatory amino acid transporters (EAATs), suggesting that, in addition to their neurotoxic properties, they also have a neuroprotective role by clearing extracellular glutamate and producing antioxidant glutathione. To test this hypothesis in human, the brain of 12 HIV-positive patients and 3 controls were immunostained for EAAT-1. EAAT-1 was expressed by AMM in all HIV-infected cases but not in HIV-negative controls. Expression varied according to the disease stage. In 5 cases with active HIV-encephalitis (HIVE), AMM strongly expressed EAAT-1 in the white matter and basal ganglia, analogous to HLA-DR and CD68 expression. There was weaker expression in the cortex and perineuronal microglial cells were not involved. In a case with "burnt out" HIVE following highly active antiretroviral therapy (HAART), EAAT-1 expression was mild, identical to that of HLA-DR and CD68 in the white matter and cortex and involved perineuronal microglial cells. In 3 AIDS patients without HIVE and in 3 pre-AIDS cases, EAAT-1 expression in the white matter was weaker than HLA-DR and CD68 expression; there was stronger correlation in the gray matter where perineuronal microglial cells were stained predominantly. Our findings in humans tend to confirm that AMM, particularly perineuronal microglial cells, play a neuroprotective role in the early stages of HIV infection and, possibly, following treatment. This is in keeping with the early microglial activation seen in pre-AIDS cases, and the late occurrence of neuronal loss. It may also explain the reversible cognitive disorders following treatment in some cases.     

Influence of HAART on HIV-related CNS disease and neuroinflammation

Neuroinflammation has an established link with AIDS-related dementia but has not been investigated in the post-highly active anti-retroviral therapy (HAART) era. In this autopsy study we examined post-HAART cases in Edinburgh for the presence of HIV-related pathology and in well-treated cases for evidence of neuroinflammation. We focused on basal ganglia and the hippocampus, 2 key areas of the brain for cognitive functioning and compared pre- and post-HAART cases for neuroinflammatory status. We find evidence, post-HAART, that there is a high level of microglial/macrophage activation that is comparable with the levels seen, pre-HAART, in HIV encephalitis (HIVE) and AIDS cases. This result was maximal in the hippocampus where microglial/macrophage upregulation in the HAART-treated group exceeded that seen in HIVE. In the basal ganglia, HAART-treated cases showed significantly higher levels of CD68-positive microglia/macrophages than in control brains (p = 0.004), and in the hippocampus levels were significantly higher than those seen in control cases, pre-HAART AIDS, and presymptomatic brains (p = 0.01). However, lymphocyte levels in the areas examined were low in HAART-treated cases. We conclude that there is a surprising degree of ongoing neuroinflammation in HAART-treated patients, particularly in the hippocampus. This may pose a threat for the future health of individuals maintained long-term on HAART therapy.     

Apoptotic neurons in brains from paediatric patients with HIV-I encephalitis and progressive encephalopathy

The pathogenesis of human immunodeficiency virus type 1 (HIV-1) associated dementia in adults involves neuronal loss from discrete areas of the neocortex and subcortical regions, but the mechanism for neuronal death is poorly understood. Gene-directed cell death resulting in apoptosis is thought to be a normal feature of neuronal development, but little is known about neuronal apoptosis in disease states. We investigated whether HIV-1 infection of the central nervous system is spatially associated with apoptosis of neurons. Using an in situ technique to identify newly cleaved 3'-OH ends of DNA as a marker for apoptosis, we demonstrate the presence of apoptotic neurons in cerebral cortex and basal ganglia of children that had HIV-1 encephalitis with progressive encephalopathy. Furthermore, an association was observed between the localization of apoptotic neurons and perivascular inflammatory cell infiltrates containing HIV-1 infected macrophages and multinucleated giant cells. Apoptotic neurons and p24–positive macrophages were observed infrequently in cerebral cortex and basal ganglia in children with HIV-1 infection without encephalitis or clinical encephalopathy. In nine control (HIV-1 negative) brains, ranging from the first post-natal month of life to 16.5 years of age, infrequent neuronal apoptosis was observed in three cases. These findings suggest that neuronal apoptosis is unlikely to be associated with post-natal development except in early post-natal germinal matrix, and that it may instead represent the end result of specific pathological processes, such as HIV-1 encephalitis.     

Damage and repair of DNA in HIV encephalitis

Neuronal damage and dementia are common sequelae of HIV encephalitis. The mechanism by which HIV infection of CNS macrophages results in neuronal damage is not known. We examined the brains from 15 AIDS autopsies (8 with HIV encephalitis and 7 without) and 4 non-infected control autopsies for the presence of DNA strand breaks, for associated changes in the expression of the DNA repair enzymes KU80 and Poly (ADP-ribose) polymerase (PARP), and for accumulation of amyloid precursor protein (APP). Abundant DNA damage was observed with terminal transferase-mediated dUTP nick end-labeling (TUNEL), however, there was no morphologic evidence of significant neuroglial apoptosis. The DNA repair enzyme KU80 was immunocytochemically detectable in neuronal and glial cells in autopsy brains from patients with and without HIV encephalitis; however, in cases with HIV encephalitis the staining was more prominent than in the infected or non-infected controls without encephalitis. There was no difference in KU80 immunostaining in oligodendroglia from autopsies with and without encephalitis. Immunostaining for PARP was more intense in gray and white matter of cases with HIV encephalitis. No clear spatial relationship existed between expression of DNA repair enzymes and the spatial proximity of microglial nodules or HIV-infected macrophages. The cytoplasm of cortical and subcortical neurons immunostained for APP Stronger cortical neuronal APP staining was observed in cases without HIV encephalitis. Staining of deep gray matter neurons was similar, irrespective of the presence or absence of encephalitis. While foci of intense APP staining were noted in white matter not related to HIV infection, they were associated with foci of opportunistic infections (e.g. due to CMV or PML). We conclude that damaged DNA and altered patterns of expression of DNA repair proteins and APP are common findings in the brains of AIDS patients at autopsy, but do not have a spatial relationship to HIV-infected macrophages.     

Neuronal apoptosis in Creutzfeldt-Jakob disease

Neuronal loss is a salient feature of prion diseases; however, its causes and mechanisms are unclear The possibility that it could occur through an apoptotic process has been postulated and is consistent with the lack of inflammation in prion disorders as supported by experimental studies. In order to test this hypothesis in humans, we examined samples of frontal and temporal cerebral cortex, striatum, thalamus, and cerebellum from 16 patients who died from Creutzfeldt-Jakob disease. They included 5 sporadic cases, 5 familial, 3 iatrogenic, and 3 cases with the new variant. These were compared with age and sex matched controls. Using in situ end labelling, we identified apoptotic neurons in all the cases of Creutzfeldt-Jakob disease. A single labelled neuron was found in the eldest control. Apoptotic neurons were mostly found in damaged regions and their presence and abundance seemed to correlate closely with neuronal loss. This supports the view that apoptosis of neurons is a feature of prion diseases and may contribute to the neuronal loss which is one of the main characteristics of these conditions. Neuronal apoptosis also correlated well with microglial activation, as demonstrated by the expression of major histocompatibility complex class II antigens, and axonal damage, as identified by beta-amyloid protein precursor immunostaining. In contrast, we found no obvious relationship between the topography and severity of neuronal apoptosis and the type, topography, and abundance of prion protein deposits as demonstrated by immunocytochemistry.     

Apoptosis in cerebral autosomal-dominant arteriopathy with subcortical infarcts and leukoencephalopathy

To test the hypothesis that an apoptotic process plays a role in the pathogenesis of cerebral lesions in cerebral autosomal-dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL), we examined samples from frontal, temporal, insular, and occipital regions, basal ganglia, and cerebellum from 4 patients with CADASIL, 2 with Binswanger disease, and 3 controls. Apoptotic cells were identified using in situ end labeling and activated caspase 3 immunostaining. Immunolabeling for Notch3, the beta-amyloid protein precursor, and phosphorylated neurofilament protein was performed on successive sections. Apoptosis of vascular cells was markedly increased in status cribrosus in CADASIL, both in basal ganglia and subcortical white matter, suggesting that concomitantly with Notch3 deposition it may play a causative role in the dilatation of Virchow-Robin spaces. Neuronal apoptosis was found in CADASIL, mostly in cortical layers 3 and 5. Its severity correlated semiquantitatively with the extent of ischemic lesions and axonal damage in the underlying white matter. It was more severe in demented patients. Only occasional apoptotic neurons were found in the Binswanger cases and none in the controls. This supports the view that neuronal apoptosis may contribute to cortical atrophy and cognitive impairment in patients with CADASIL and that it may, at least partly, result from axonal damage in the underlying white matter.     

Central nervous system correlates of behavioral deficits following simian immunodeficiency virus infection

Despite the high incidence of cognitive and motor impairment in acquired immunodeficiency syndrome (AIDS) patients, the mechanisms of AIDS-related central nervous system (CNS) pathology are not completely understood. Infection with simian immunodeficiency virus (SIV) in macaques provides an excellent model of AIDS, including human immunodeficiency virus (HIV)-induced CNS pathology and cognitive/behavioral impairment. Co-inoculation with two SIV strains, SIV/17E-Fr and SIV/DeltaB670, accelerates SIV CNS disease, producing SIV encephalitis in over 90% of pig-tailed macaques within 3 months. In the present study, this SIV model was employed to identify cellular and viral correlates of behavioral impairment following SIV infection. Measures of psychomotor speed (simple reaction time), fine motor control (bimanual motor task), and general motor activity (home cage movement) were all adversely affected by SIV disease. Prior to euthanasia, performance was significantly impaired in both a simple reaction time task in 6 of 12 monkeys and a bimanual motor task in 5 of 6 monkeys. All monkeys evaluated (11 of 11) showed significant reductions in spontaneous motor activity. Significant correlations were found between impaired performance on the bimanual motor test and axonal damage (accumulation of beta-amyloid precursor protein in the corpus callosum) as well as increased microglial activation and macrophage infiltration (levels of CD68 and Ham56 immunostaining). These results suggest that axonal damage is related to the behavioral impairment induced by infection with SIV. The axonal damage may result from neuroimmune responses, including microglial and macrophage activation. Therefore, axonal damage may be a morphologic manifestation of neuronal dysfunction that underlies development of behavioral impairment in HIV/SIV CNS infection.     

Early Neurodegeneration after Hypoxia-Ischemia in Neonatal Rat Is Necrosis while Delayed Neuronal Death Is Apoptosis

We used silver staining to demonstrate neuronal cell body, axonal, and terminal degeneration in brains from p7 rat pups recovered for 0, 1.5, 3, 6, 24, 48, 72 h, and 6 days following hypoxia-ischemia. We found that initial injury is evident in ipsilateral forebrain by 3 h following hypoxia-ischemia, while injury in ventral basal thalamus develops at 24 h. A secondary phase of injury occurs at 48 h in ipsilateral cortex, but not until 6 days in basal ganglia. Initial injury in striatum and cortex is necrosis, but in thalamus the neurodegeneration is primarily apoptosis. Degeneration also occurs in bilateral white matter tracts, and in synaptic terminal fields associated with apoptosis in regions remote from the primary injury. These results show that hypoxia-ischemia in the developing brain causes both early and delayed neurodegeneration in specific systems in which the morphology of neuronal death is determined by time, region, and potentially by patterns of neuronal connectivity.     

Altered P-glycoprotein expression in AIDS patients with HIV encephalitis

Penetrance of anti-retroviral drugs into the CNS depends partly on the activity of P-glycoprotein (P-gp), an ATP-dependent efflux pump involved in restricting entry of lipophilic drugs into the brain. The present study characterizes the patterns of P-gp expression in the brains of AIDS patients and examines its relationship with clinical and neuropathological indicators of HIV encephalitis (HIVE). For this purpose, brain tissue collected at autopsy from 26 subjects with a history of HIV (9 without HIVE; 17 with HIVE) was analyzed. Immunocytochemical staining and Western blot analyses for regional P-gp expression were performed and levels were correlated with neuropathological indicators and with HIV RNA. Double labeling experiments were performed with antibodies against astroglial (GFAP), endothelial (CD31), microglial (CD45) and neuronal (MAP2) cell markers. In the HIVE-negative cases, P-gp immunoreactivity was associated primarily with endothelial cells. HIVE-positive cases showed extensive immunolabeling of astroglial and microglial cells, but relatively less endothelial cell immunolabeling. No neuronal P-gp immunostaining was detected in brain tissue from any cases in the study. In the HIVE-positive cases with extensive astroglial labeling, the most intense immunoreactivity was detected in white matter. A subset of HIVE-positive cases displayed intense P-gp immunostaining of astrocytes closely associated with blood vessels in the cortex. Both the immunocytochemical and Western blot analyses showed a significant correlation between P-gp expression and HIV RNA levels. In conclusion, P-gp immunoreactivity was detected largely in glial cells in tissue from HIVE-positive patients. Furthermore, in HIVE-positive patients, brain viral burden and P-gp levels were significantly higher than those in HIVE-negative patients. Taken together, our data suggest that P-gp may be part of a central pathway mediating viral compartmentalization in the brains of HIV-infected individuals and may play a significant part in HIV disease progression in the brain.     

SDF-1alpha is expressed in astrocytes and neurons in the AIDS dementia complex: an in vivo and in vitro study

Recent in vitro studies suggest that the alpha chemokine stromal-derived factor-1alpha (SDF-1alpha) and its receptor CXCR-4 may contribute to neuronal apoptosis in HIV infection of the brain. The cellular and regional expression of this chemokine and its relationship to the AIDS dementia complex (ADC), however, have remained undetermined. Using immunohistochemistry and semiquantitative RT-PCR, we examined the expression of SDF-1alpha in the frontal cortex (FC), the adjacent deep white matter (DWM). and the basal ganglia (BG) of 17 patients with ADC and 5 normal controls, and the FC and temporal cortex of 6 patients with Alzheimer disease (AD). Additionally, SDF-1alpha expression was studied in 3 different neuronal cultures: differentiated SK-N-MC cells, primary human fetal neuronal, and mouse hippocampal cultures. SDF-1alpha staining was predominantly localized to astrocytes in all 3 groups in the gray matter of the FC and the BG, often in the vicinity of cortical and basal ganglia neurons, but was generally absent in the DWM. Further, the number of positive neurons was significantly greater in the BG of AIDS subjects with advanced brain disease compared to subjects with lesser disease (p = 0.029). All cultures showed prominent SDF-1alpha staining of neurons within the cytoplasm and in neurites, whereas preferential expression in GABA-ergic neurons was found in hippocampal cultures. This is the first study to show that SDF-1alpha is constitutively expressed in astrocytes of the deep and cortical gray matter as well as in neurons of the human brain. Its increased expression in basal ganglia neurons of patients with advanced HIV CNS disease suggests it may also contribute to pathogenesis.     

Early entry and widespread cellular involvement of HIV-1 DNA in brains of HIV-1 positive asymptomatic individuals

There is overwhelming evidence that invasion of the central nervous system (CNS) by HIV-1 takes place at an early stage of the infection. It has been demonstrated that HIV-1 DNA is present in brains of asymptomatic individuals. Evidence of immune activation and increased expression of cytokines suggested that neuropathological changes and neuronal and axonal damage could be the effect of the presence of the virus. The purpose of the study is to ascertain whether target cells for HIV-1 in brain of patients at early stage of the infection are the same as those found in AIDS sufferers or if the distribution seen in AIDS patients results from the late spreading of the infection from cells considered traditionally the reservoir of the virus, i.e. microglial cells. Eighteen brains, all HIV-1 DNA positive, as shown by nested polymerase chain reaction (PCR), were selected among the group of HIV-1 positive asymptomatic cases. In 6 of them, HIV-1 DNA was detected by PCR in situ. Positive cells included astrocytes and endothelial cells, in addition to microglial cells. We conclude that astrocytes and endothelial cells are already infected at an early (asymptomatic) stage of the infection and suggest that they might contribute to the damage of the CNS.     

Perinatal human hypoxia-ischemia vulnerability correlates with brain calcification

Deregulation of intracellular calcium homeostasis is widely considered as one of the underlying pathophysiological mechanisms of hypoxic-ischemic brain injury. Whether this alteration can result in cerebral calcification was investigated in basal ganglia, cerebral cortex, and hippocampus of human premature and term neonates together with glial reaction. In all samples nonarteriosclerotic calcifications were observed, their number and size were area-specific and increased in term neonates. Basal ganglia always presented the highest degree of calcification and hippocampus the lowest, located mainly in the CA1 subfield. In all cases, neuronal damage was associated with astroglial reaction and calcium precipitates, with microglial reaction only in basal ganglia and cerebral cortex, and argues for the participation of excitatory amino acid receptors in hypoxia-ischemia damage. These data correlate with hypoxia-ischemia vulnerability in the perinatal period. The clinical relevance of these precipitates and the neuroprotective interest of non-NMDA receptor manipulation are discussed in the light of our results.     

Tumor necrosis factor-α, microglia and astrocytes in AIDS dementia complex

The pathogenesis of HIV-associated cognitive changes is poorly understood. Cytokines such as tumor necrosis factor-α (TNF-α) have been postulated to contribute to the mechanism of the neurological complications of HIV infection. One of the effects of TNF-α is to induce astrocyte proliferation in vitro. The purpose of this study was to look for a correlation between the expression of TNF-α, astrogliosis and the degree of cognitive impairment in 12 prospectively assessed AIDS cases without focal brain lesion, 8 of whom were demented. They were compared with 6 control patients without neurological disease. Neuropathological examination showed myelin pallor in 5 of the 8 demented patients. TNF-α expression was detected by immunohistochemistry in the midfrontal cortex, subcortical and deep white matter, and basal ganglia. Not only perivascular macrophages but also some microglial and endothelial cells were labeled. Most TNF-α-positive cells were in close contact with glial fibrillary acidic protein-positive astrocytes. They were more numerous than gp41-positive cells. Their density increased with increasing cognitive impairment and in parallel to the astrogliosis in the frontal cortex, basal ganglia and deep white matter. These findings further support the hypotheses that lesions of the deep white matter, driven by TNF-α, are associated with cognitive alteration, and that indirect effects of HIV infection in the brain participate in the development of HIV-associated dementia through a diffuse immune activation, mediated by cytokines. Received: 22 May 1996 / Revised: 13 September 1996 / Accepted: 11 November 1996     

Accumulation of β-amyloid precursor protein in HIV encephalitis: Relationship with neuropsychological abnormalities

The pathogenesis of neuropsychological abnormalities in patients with human immunodeficiency virus type 1 (HIV-1) encephalitis is obscure because neurons are not the target of infection and severe neuronal loss occurs only late during the disease. Moreover, there is evidence indicating that HIV dementia is not a homogeneous entity and could partially reverse after treatment with zidovudine. The finding that impaired axonal flow, evidenced by β-amyloid precursor protein immunoreactivity, could contribute to the neuropsychological deficits prompted the present study. Brains of patients with full-blown acquired immunodeficiency syndrome (AIDS) were studied and findings compared with those of normal and abnormal control subjects. The presence of HIV-1 DNA was investigated by nested polymerase chain reaction; axonal abnormalities were detected by β-amyloid precursor protein, ubiquitin immunohistochemistry, and silver staining. Accumulation of β-amyloid precursor protein was observed in all the HIV encephalitis brains studied; the appearance of the immunostaining varied from globular structures to bundles of parallel formations. In 2 AIDS brains without pathological abnormalities, only the latter pattern was detected. The brains with trauma were strongly reactive with β-amyloid precursor protein antibody and the different reactivity within them correlated with posttrauma survival, only globular structures being detected in the older cases. No correlation was found between the different pattern of β-amyloid precursor protein reactivity and dementia in AIDS patients. These results show that widespread axonal injury is a constant feature in AIDS brains and suggest that it could play a role in the pathogenesis of the neuropsychological abnormalities of these patients.     

DNA breaks detected by in situ end-labelling in dorsal root ganglia of patients with AIDS

Distal sensory axonal polyneuropathy (DSP) is the most frequent HIV-associated peripheral neuropathy. DSPs tend to occur in full-blown AIDS and worsen as CD4 cell counts decrease in blood. To assess a possible role for apoptosis in the pathogenesis of the neuropathy, we used in situ end-labelling (ISEL) detecting DNA strand breaks in DRG neurons of 19 HIV-infected patients, of whom nine had axonal polyneuropathy, and 11 controls. Sensory neurons with ISEL-assessed DNA breaks were observed in 9/19 patients with AIDS, 0/3 patients with pre-AIDS, and 1/11 controls. The prevalence of DNA breaks in neurons was higher in AIDS patients than in controls ( P <0.05). Among AIDS patients, DNA breaks in neurons were more abundant in patients with peripheral neuropathy ( P <0.04). It is possible that DNA breaks of DRG neurons induce the axonopathy and consequently play a role in the pathogenesis of DSP. It cannot be excluded, however, that DNA breaks could represent the result rather than the cause of axonopathy. We suggest that ISEL may detect neurons that were primed to apoptosis before death in relation with the HIV infection, and undergo DNA fragmentation at time of death, rather than neurons that underwent pre-mortem both priming and triggering steps of the apoptotic process. This hypothesis could explain why most ISEL-positive neurons lack typical apoptotic morphology and why normal controls do not show ISEL positive cells.