NEURONAL CELL DEATH IN HIV-1 INFECTION

Despite the absence of productive infection of neurons, HIV-1 infection has been associated with neuronal loss in distinct regions of the brain, including the frontal cortex,^65-67 the substantia nigra,⁶⁸ the cerebellum,⁶⁹ and the putamen.⁷⁰ Neuronal loss in the frontal cortex and in other regions of the neocortex has been estimated between 18 and 38%.^65-67 Select neuronal subpopulations, including larger pyramidal cells within the cortex, are at greater risk of cell death and, similarly, neuronal populations that express certain neurotransmitters (e.g., GABA) and proteins (e.g., parvalbumin and calbindin) are more likely to be diminished in the HIV-1-infected brain.⁷¹

Cellular injury and loss in the brain during HIV-1 infection are also evident from magnetic resonance imaging and spectroscopy studies that revealed brain atrophy in HIV-1-infected patients occurring within cortical, central, brain stem, and cerebellar regions.^72-76 In some studies, the severity of atrophy was correlated with advanced stages of HIV-1 infection^72,73 or with clinical signs of cognitive impairment.^72,73,75 Although necrosis of neurons was demonstrated in HAD patients,⁷⁷ it was hypothesized as early as 1990 that the neuronal loss observed in HIV-1-infected patients may also be due to apoptosis.^65,66 Apoptosis is an active process of cell death, characterized by cell shrinkage, chromatin aggregation with genomic fragmentation, and nuclear pyknosis.^78,79 In vivo, phagocytic cells normally sequester antigenically modified apoptotic cells, preventing inflammation and damage to the surrounding tissue.⁸⁰ In contrast, necrosis is characterized by passive cell swelling and disruption of internal homeostasis, resulting in membrane lysis and release of intracellular constituents that induces a local inflammatory reaction with subsequent injury to the surrounding tissue.⁸¹ If apoptotic neurons are not phagocytosed in a timely fashion, however, secondary necrosis may occur, producing ultrastructural phenotypes characteristic of both apoptosis and necrosis.^82-84 It has been suggested that the intensity of the initial insult dictates whether neurons undergo apoptosis or necrosis; a severe initial challenge will result in necrosis, whereas a milder challenge triggers apoptotic signals via p38 MAPK.⁸²

With the development of TUNEL (terminal deoxynucleotidyl dUTP nick end labeling), a technique that identifies the free 3-OH ends of newly cleaved DNA in situ,⁸⁵ several groups in the mid-1990s demonstrated that neuronal damage in both adult and pediatric AIDS was, in part, due to apoptosis.^86-89 However, neuronal apoptosis seems to be a feature of AIDS, as only rare apoptotic neurons were demonstrated in a few pre-AIDS cases.^87,90 Whether a direct correlation exists between the degree of neuronal apoptosis and the severity of neurological cognitive disorders observed in HIV-1 infection remains unclear.

TUNEL analysis also identified DNA strand breaks in the DRG of HIV-1-infected individuals.⁹² TUNEL-positive neurons were observed only in the DRG of patients with AIDS and were more abundant in patients with peripheral neuropathy. However, TUNEL-positive neurons did not cor­relate with neuronal loss or demonstrate morphological features of apoptosis, suggesting that these neurons might be “primed” to apoptosis rather than fully committed to apoptosis.

In addition to neuronal cell death, apoptosis of other CNS cell types in the brains of HIV/AIDS patients was reported, including astrocytes^86,89,93 and, more rarely, multinucleated giant cells.⁸⁶ In addition, astrocyte apoptosis was demonstrated in primary human brain cultures following exposure to HIV-1.⁸⁹ Significantly greater numbers of apoptotic astrocytes were detected in the brains of HIV/AIDS patients with rapidly progressing dementia compared with slow progressors,⁹³ and detection of apoptotic astrocytes seemed to be more common in HAD patients compared with nondemented HIV/AIDS patients,⁸⁹ suggesting a role for astrocyte cell loss in the neuropathogenesis of HAD.

Although there is no clear consensus as to the underlying mechanism of HIV-induced neuro­pathogenesis, two main theories dominate (Figure 22.1). First, although HIV-1 does not infect neurons productively, if at all, HIV-1-encoded proteins may injure neurons directly without requiring the intermediary functions of nonneuronal cells. Alternatively, neuronal apoptosis may result indi­rectly from the secretion of neurotoxic factors by resident brain macrophages or microglia in response to HIV-1 infection, stimulation by viral proteins, or immune activation. These two theories are in no way mutually exclusive, and current data exist to support each hypothesis.