Absence of LEDGF/p75 Expression in Astrocytes May Affect HIV-1 Integration Efficiency

Abstract

In spite of effective anti-retroviral therapy, HIV-1 infection may still lead to neurological impairment in patients. The underlying mechanism of neurodegeneration remains mysterious. HIV-1 does not infect neurons, but does infect microglia cells in the brain. It is controversial whether HIV-1 productively infects astrocytes, an abundant glial cell type in the brain. Thirty years of investigation have led to conflicting reports concerning the entry, infection, and production of progeny virions from astrocytes. New models from studies in primary human fetal astrocytes suggest phagocytosis of HIV-1 with little productive infection. The retroviral life cycle requires integration of the viral genome to the host genome. The host protein LEDGF/p75 is required for efficient HIV-1 integration. In the absence of LEDGF/p75, HIV-1 integration and infection efficiency is reduced ten fold. Differentiated astrocytes do not appear to express LEDGF/p75, which suggests these cells are disabled for efficient integration. Phagocytosis of HIV-1 virions and the lack of LEDGF/p75 expression in astrocytes suggest that this cell type is not efficiently infected in vivo.

INTRODUCTION

Human immunodeficiency virus (HIV-1) primarily causes immunodeficiency, but may also lead to neurological disease (reviewed in [1]. HIV-1 associated neurocognitive disorders (HAND) range in severity from mild impairment to more severe HIV-1 associated dementia. In spite of successful anti-retroviral therapy and suppression of viral replication in the periphery, HIV-1 patients may still experience some form of HAND. These neurological disorders are due directly to neurotoxicity rather than opportunistic infections. However, neurons are not infected by HIV-1 and the precise mechanism of neurotoxicity remains elusive.

HIV-1 INFECTION OF ASTROCYTES IS CONTROVERSIAL

HIV-1 tropism is determined by its envelope glycoprotein gp120 [2]. A trimer of viral gp120 binds the host surface protein CD4, which induces a gp120 conformational change and binding of a secondary host receptor CCR5 or CXCR4. These binding events reveal the viral fusion peptide gp41 allowing the viral envelope to fuse with the cellular membrane. The host membrane proteins CD4, CCR5, and CXCR4 are expressed by a subset of T cells and macrophages. In the central nervous system (CNS) microglia are resident phagocytic cells expressing these surface markers and are productively infected by HIV-1 [3]. Neurons do not express these membrane proteins and are not infected. Infection of astrocytes has been more controversial [4, 5]. Astrocytes are abundant cells in the brain that do not express the primary HIV-1 receptor CD4, but do express CCR5 and CXCR4 [6]. Although astrocytes do not express CD4, there have been reports of HIV-1 infection of this cell type through an enigmatic CD4-independent entry mechanism [7, 8].

Following fusion of the viral and cellular membranes, the HIV-1 capsid core enters the host cell cytoplasm [2]. Then the virus reverse transcribes its RNA genome to a linear double stranded complementary DNA (cDNA). The cDNA is part of a poorly defined pre-integration complex that travels to the nucleus. After nuclear entry the viral enzyme integrase covalently joins the viral cDNA to the host genome and is called a provirus. Viral genes are transcribed and translated by the host machinery. Progeny virions assemble at the cellular membrane and bud from the host cell. A productive infection requires the integration of the viral cDNA into the host genome.

There have been multiple conflicting reports concerning the productive infection of astrocytes. Postmortem brain tissues analyzed by in situ hybridization for HIV-1 proteins or in situ PCR for HIV-1 nucleic acids indicated infection of astrocytes [7, 8]. Other reports have suggested that astrocytes are infected by HIV-1 but do not support production of progeny virions [5, 9, 10]. In these reports, it was proposed that HIV-1 integration occurred but the viral life cycle was inhibited before the production of progeny virions.

Using primary human fetal astrocytes or astrocytoma cell lines, more recent models of HIV-1 association with astrocytes have focused on phagocytosis [10–12]. These studies propose that HIV-1 does not enter astrocytes by the classical fusion of the viral envelope with the cell plasma membrane. Instead, astrocytes may phagocytose HIV-1 virions or infected cell debris [10–12]. Virus particles in an endosomal compartment may be destroyed, escape the endosome to infect the cell, or be released from the cell. Viral escape from endosomes to integrate into an astrocyte genome appears to be inefficient [10, 12]. This model supports the notion that previous reports of viral proteins or nucleic acids detected in postmortem brain tissues could be obfuscated by phagocytosed virions or debris from infected cells.

INTEGRATION OF VIRAL cDNA

HIV-1 integrase binds a cellular cofactor LEDGF/p75, a transcription activator, which appears to tether the integration complex to host chromatin facilitating integration [13]. LEDGF/p75 includes an amino terminal PWWP chromatin binding domain and a carboxyl terminal integrase binding domain. In the absence of LEDGF/p75, integration is reduced ten fold [14, 15]. It has been largely accepted that LEDGF/p75 is ubiquitously expressed in mammalian tissues [16]. Indeed, Western analysis of commercially available multiple tissue blots clearly shows that LEDGF/p75 is expressed in the human brain [17]. However, more careful analysis of LEDGF/p75 expression revealed that this protein is not expressed in differentiated astrocytes [18].

Reports of integration efficiency in cultured astrocytes is variable. HIV-1 virus particles were added to three astrocytoma cell lines [12]. In these studies, integration was not detected by a highly sensitive nested PCR method suggesting that this step is inefficient in astrocytes [12]. To circumvent the entry defect of HIV-1 to astrocytes, HIV-1 particles may be pseudotyped with the vesicular stomatitis virus (VSV) envelope glycoprotein (VSV-G). Addition of VSV-G pseudotyped HIV-1 to human fetal astrocytes produced progeny virions and suggested no intracellular blocks to viral replication, including integration [19]. However, the lack of LEDGF/p75 does not completely prevent HIV-1 integration; it reduces integration ten fold. In contrast VSV-G pseudotyped HIV-1 virions are reportedly 20–130 fold more infectious [20]. These studies employed the same quantity of p24 per cell. The discrepancy in results may be reconciled by the fact that VSV-G pseudotype effectively increased the virions added by possibly two orders of magnitude. The role of differential LEDGF/p75 expression in astrocytes has been overlooked.

TRANSCRIPTION OF UNINTEGRATED VIRAL DNA

In the absence of LEDGF/p75 in fibroblasts or CD4+ T cell lines, most of the HIV-1 viral cDNA remains unintegrated [14, 15]. Inhibition of integration leads to an accumulation of circular forms of the viral cDNA [21]. HIV-1 circular DNA forms are only found in the nucleus and may include one or two viral long terminal repeats (LTRs). During HIV-1 infection, circular viral cDNA is common and long lived [22, 23]. A report from 1990 indicated that unintegrated viral cDNA was readily detected in postmortem brain specimens from patients with HIV-1 associated dementia [23]. This study did not differentiate the cell types with unintegrated viral cDNA. Circular viral cDNA forms could account for the nucleic acids detected by in situ PCR of previous studies suggesting HIV-1 infection of astrocytes.

The HIV-1 promoter is encoded in the LTR and is present in circular viral cDNA [2]. The circles may be transcribed and ultimately produce the viral proteins Tat, Rev, and Nef [21, 24]. The circles are not able to produce progeny virions. It seems unlikely that limited transcription from viral cDNA circles could produce the HIV-1 proteins, such as p24, detected by in situ hybridization in previous reports. Astrocytes in the CNS appear to express both Tat and Nef, which have been reported to be neurotoxic [9, 25]. One model is that HIV-1 enters astrocytes by either fusion or phagocytosis, but is unable to integrate. The resulting accumulation of circular viral cDNA generates the neurotoxic HIV-1 proteins Tat and Nef leading to HAND.

CONCLUSIONS

HIV-1 infection in the CNS can result in neurodegenerative disease [1]. Determination of the underlying mechanisms that lead to neurotoxicity has been hampered by the lack of a tractable animal model of HIV-1 infection. While it seems clear that microglia are infected by HIV-1 and neurons are not infected, the role of astrocytes remains unclear [4]. The phagocytic ability of these cells complicates the interpretation of previous reports detailing the presence of HIV-1 proteins or nucleic acids in astrocytes of postmortem patient brains [10, 11]. However, absence of the host co-factor for integration LEDGF/p75 suggests that integration is inefficient in astrocytes.