Cytochrome c dysregulation induced by HIV infection of astrocytes results in bystander apoptosis of uninfected astrocytes by an IP3 and calcium-dependent mechanism

Authors

  • Eliseo A. Eugenin,

    Corresponding author
    1. Public Health Research Institute (PHRI), Newark, New Jersey, USA
    2. Department of Microbiology and Molecular Genetics, Rutgers New Jersey Medical School, Rutgers The State University of New Jersey, Newark, New Jersey, USA
    • Address correspondence and reprint requests to Eliseo Eugenin, Public Health Institute (PHRI) and Department of Microbiology and Molecular Genetics, Rutgers New Jersey Medical School, Rutgers The State University of New Jersey, Newark, NJ 07103, USA. E-mail: eliseo.eugenin@rutgers.edu

    Search for more papers by this author
  • Joan W. Berman

    1. Department of Pathology, Albert Einstein College of Medicine, Bronx, New York, USA
    2. Department of Microbiology/Immunology, Albert Einstein College of Medicine, Bronx, New York, USA
    Search for more papers by this author

Abstract

HIV entry into the CNS is an early event after peripheral infection, resulting in neurologic dysfunction in a significant number of individuals despite successful anti-retroviral therapy. The mechanisms by which HIV mediates CNS dysfunction are not well understood. Our group recently demonstrated that HIV infection of astrocytes results in survival of HIV infected cells and apoptosis of surrounding uninfected astrocytes by the transmission of toxic intracellular signals through gap junctions. In the current report, we characterize the intracellular signaling responsible for this bystander apoptosis. Here, we demonstrate that HIV infection of astrocytes results in release of cytochrome C from the mitochondria into the cytoplasm, and dysregulation of inositol trisphosphate/intracellular calcium that leads to toxicity to neighboring uninfected astrocytes. Blocking these dysregulated pathways results in protection from bystander apoptosis. These secondary messengers that are toxic in uninfected cells are not toxic in HIV infected cells, suggesting that HIV protects these cells from apoptosis. Thus, our data provide novel mechanisms of HIV mediated toxicity and generation of HIV reservoirs. Our findings provide new potential therapeutic targets to reduce the CNS damage resulting from HIV infection and to eradicate the generation of viral reservoirs.

image

We demonstrated that HIV infection of astrocytes protects infected cells from apoptosis but results in cell death of surrounding uninfected astrocytes by a mechanism that is dependent on gap junction channels, dysregulation of mitochondrial cytochrome C (CytC), and cell to cell diffusion of inositol trisphosphate (IP3) and calcium. Our data provide essential information about generation of brain reservoirs and the mechanism of toxicity mediated by the virus.

Abbreviations used
AGA

18-α-glycerritenic acid

CytC

cytochrome c

PBS

phosphate-buffered saline

HIV infection compromises the CNS in a significant number of infected individuals. As individuals infected with HIV live longer because of the success of anti-retroviral therapies, HIV associated neurologic disorders and persistence of viral reservoirs within the CNS have become major public health issues.

Microglia/macrophages are the main productively infected cells in the brain (An et al. 1999a,b; Wiley et al. 1999; Cosenza et al. 2002; Wiley 2003). However, in addition, astrocytes are susceptible to low levels of infection, and support minimal to undetectable viral replication (Conant et al. 1994; Tornatore et al. 1994; Trillo-Pazos et al. 2003; Wang et al. 2004; Eugenin and Berman 2007; Churchill et al. 2009; Eugenin et al. 2011, 2012). Despite the fact that these two cell types can be infected in vivo, viral production is generally too low to explain the extensive CNS pathology often observed (Nuovo and Alfieri 1996; Heaton et al. 2011; Jernigan et al. 2011). This observation suggests that mechanisms of amplification of toxicity and inflammation in response to HIV infection, rather than the virus itself, are a major cause of neurological disease. Our previous data demonstrated that gap junction channels contribute to the amplification of apoptosis from HIV infected astrocytes to surrounding uninfected cells. However, these HIV infected astrocytes do not undergo cell death (Eugenin and Berman 2007; Eugenin et al. 2011, 2012). This appears to be because of the generation of toxic signals that lead to bystander apoptosis of the uninfected astrocytes. Here, we demonstrate that HIV infection of astrocytes dysregulates mitochondrial localization, cytochrome c (CytC) mitochondrial localization, intracellular calcium, and IP3 signaling, that together are responsible for the bystander killing of neighboring uninfected cells through gap junction communication. Thus, our data demonstrate that HIV infection of astrocytes generates HIV reservoirs because infected cells do not undergo apoptosis. These infected astrocytes generate intracellular toxic signals, including CytC that alter IP3, and intracellular calcium, resulting in bystander killing of uninfected cells by diffusion through gap junction channels. Our findings identify a new mechanism of toxicity and generation of viral reservoirs in response to HIV infection that are amplified by gap junctions, even in the absence of significant viral replication.

Materials and methods

Materials

Dulbecco's modified Eagle's medium, fetal bovine serum, penicillin/streptomycin (P/S) and trypsin-EDTA were from Invitrogen-GibcoBRL (Grand Island, NY, USA). Monoclonal antibody to Glial fibrillar acid protein (GFAP) and FITC or Cy3-conjugated anti-rabbit IgG, BAPTA-AM, 18-α-glycerritenic acid (AGA), octanol and Cy3 or FITC anti-mouse IgG antibodies were from Sigma (St. Louis, MO, USA). Purified mouse IgG2B and IgG1 myeloma proteins were from Cappel Pharmaceuticals, Inc (Costa Mesa, CA, USA). 4′,6-diamidino-2-phenylindole (DAPI), Fura-2, anti-rabbit, and anti-mouse conjugated to Alexa were from Invitrogen (Eugene, OR, USA). Monoclonal HIV-p24 and CytC antibodies were from Abcam (Cambridge, MA, USA). Antibodies to bcl-2, bcl-x, BAD, BID and caspase-3 were from Santa Cruz (Santa Cruz, CA, USA). The in situ cell death detection kit [Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL)] was from Roche (Mannheim, Germany) and the annexin-5 apoptosis assay was from Clontech (Mountain View, CA, USA). DNKTVTFEEHIKEEHN-BIOPY- 577/618 maleimide, IP3R/CYTC inhibitor, was obtained from Peprotech (Rocky Hill, NJ, USA). Human inositol 1,4,5 trisphosphate (IP3) ELISA kit was obtained from Cusabio (Wuhan, China).

Astrocyte cultures

Cortical human fetal tissue was obtained as part of an ongoing research protocol approved by the Albert Einstein College of Medicine and Rutgers University. The preparation of cultures of primary astrocytes was performed as previously described (Eugenin and Berman 2003, 2007). We also used the astrocytoma cell line, U87 transfected with CD4 and CCR5 (U87CD5CCR5, NIH AIDS repository, Bethesda, MD, USA) to examine the role of HIV infection in astrocytes, because these cells are susceptible to HIV infection (90–95% become infected), instead of 5% of primary human astrocytes.

HIV-infection of astrocyte cultures

Confluent cultures of human astrocytes were infected by incubation with HIVADA (20–50 ng p24/ml/1 × 106 cells), using a previously described protocol (Eugenin and Berman 2003, 2007; Eugenin et al. 2011). Briefly, astrocytes were exposed to the virus for 24 h and washed extensively to eliminate the unbound virus before addition of fresh medium. HIV infection of U87 transfected with CD4 and CCR5 (U87CD4CCR5) was performed for 2 h, to facilitate infection of at least 90–95% of the cells. Immunofluorescence analyses for GFAP and HIV-p24 indicated that the cells infected with HIV were astrocytes and no contamination with microglia, CD68 positive cells, was detected.

Immunofluorescence and apoptosis assay

Primary astrocytes or the astrocytoma cell line, U87 transfected with CD4 and CCR5, were grown on coverslips, fixed and permeabilized in 70% ethanol for 20 min at −20°C. Cells were incubated in TUNEL reaction mixture at 37°C for 1 h or mito tracker, washed three times in phosphate-buffered saline (PBS), and incubated in blocking solution for 30 min at 25°C. Cells were incubated in diluted primary antibody (anti-HIV-p24, anti-CytC and anti-GFAP: 1 : 50, 1 : 200 and 1 : 800, respectively) overnight at 4°C. Cells were washed several times with PBS at and incubated with the appropriate secondary antibodies conjugated to Alexa-350, FITC, Cy3, or Cy5 for 1 h at 25°C, followed by another wash in PBS for 1 h. Coverslips were then mounted using anti-fade reagent with DAPI and cells were examined by confocal microscopy (Leica Microsystems, Wetzlar, Germany). Antibody specificity was confirmed by replacing the primary antibody with a non-specific myeloma protein of the same isotype or non-immune serum. The percentage of apoptotic astrocytes was determined by double immunofluorescence staining for TUNEL and GFAP. Quantification of staining and colocalization were determined using imaging software (NIS Elements Advance Research, Nikon, Kanagawa, Japan).

Intracellular quantification of IP3 and calcium quantification

Intracellular IP3 was determined using a human IP3 ELISA kit according to the protocols provided by the manufacture. Intracellular calcium was determined by quantifying the ratio of fluorescence as previously described (Saez et al. 1989). Because of the small numbers of infected cells in the cultures of astrocytes, to demonstrate that CytC results in release of IP3 and calcium despite the presence or absence of blockers, we electroporated astrocytes with CytC using a Bio-Rad Laboratories, Hercules, CA, USA system for transfection (150 V, 960 μF for 30 s).

Microinjection

CytC was injected into the cytoplasm of the astrocytes by single cell electroporation (Molecular Devices, Sunnyvale, CA, USA, see details in http://www.moleculardevices.com/Products/Instruments/Conventional-Patch-Clamp/Axon-Axoporator.html) to maintain the survival of the microinjected cell, since this method minimally compromises the plasma membrane. In addition, to assure proper microinjection, a fluorescent dye (Lucifer yellow or Dextran-FITC) was included inside of the pipette. After 24 h post-microinjection, apoptosis in the microinjected and neighboring cells was quantified using HIV-p24 staining to identify HIV infected and uninfected astrocytes.

Statistical analysis

Student′s two-tail, paired t-test was used. A value of p < 0.05 was considered significant.

Results

HIV infection of primary astrocytes results in bystander apoptosis of uninfected cells and survival of HIV infected astrocytes. Our previous (Eugenin and Berman 2007; Eugenin et al. 2011, 2012) and current data indicate that HIV infection of human primary astrocytes with HIVADA (20–50 ng/mL), results in infection of few cells in the cultures (4.7 ± 2.8%) and low to undetectable level of viral production, 0–100 pg/mL (Eugenin and Berman 2007; Eugenin et al. 2011, 2012). Minimal apoptosis was detected in uninfected cultures of primary astrocytes after 120 days in culture (Fig. 1, lines with circles). HIV infection of primary astrocyte cultures evidenced increased apoptosis up to 40 days post-infection, with ~ 20% apoptosis or 80% survival (Fig. 1, lines with squares). After 40 days post-infection a significant decrease in survival was detected (Fig. 1, lines with squares) as compared to uninfected cultures (Fig. 1, lines with circles). Staining for HIV-p24 protein to detect only infected cells in these cultures as well as TUNEL to quantify apoptosis indicated that HIV infected astrocytes (HIV-p24 positive cells) did not undergo apoptosis, suggesting that HIV infection protects these cells from apoptosis (Fig. 1 lines with triangles). Thus, apoptotic cells were mostly uninfected astrocytes. Our data indicate that HIV infection maintains the survival of infected astrocytes, which may result in the establishment of astrocyte viral reservoirs. To address the mechanism by which infected cells generate toxic signals, we examined mitochondrial integrity.

Figure 1.

HIV infection of astrocytes results in apoptosis of uninfected astrocytes, but not in infected cells. Using immunofluorescence staining of uninfected and HIVADA infected cultures of astrocytes for HIV-p24 (to demonstrate HIV infection), TUNEL (to determine apoptosis) and DAPI (to quantify total numbers of cells); cell death was quantified for up to 120 days post-infection. HIV infection of primary cultures of astrocytes resulted in apoptosis, mostly of uninfected cells (lines with squares). Apoptosis was minimally detected in uninfected cultures of primary astrocytes (control cells, lines with circles). Quantification of apoptosis in just HIV-p24 positive astrocytes indicated that most of infected cells survive infection (lines with triangles). n = 5, *p < 0.005, represents significance as compared to control uninfected cultures (control cells).

HIV infection of astrocytes dysregulates mitochondrial cellular localization and induces the release of CytC from the mitochondria into the cytoplasm

Our published data indicated that HIV or SIV infection of cultured primary astrocytes results in infection of ~ 5–8% of the astrocytes with undetectable to low levels of viral replication (Eugenin and Berman 2007; Eugenin et al. 2011). We also demonstrated that HIV infected astrocytes, despite this low to undetectable replication, triggered apoptosis in surrounding uninfected cells by a gap junction dependent mechanism (Eugenin and Berman 2007; Eugenin et al. 2011). However, HIV protects infected astrocytes from cell death (Eugenin and Berman 2003, 2007; Eugenin et al. 2011). However, our previous data did not identify the toxic mediators that result in apoptosis in uninfected cells nor the mechanism by which these toxic factors are not apoptotic in HIV infected astrocytes.

Mitochondria are key organelles in the regulation of energy production as well as in apoptosis (Boehning et al. 2003, 2004). Thus, to examine the contribution of mitochondria to HIV induced apoptosis we determined their localization and function in primary human astrocytes and the U87CD4CCR5 cell line. We used the U87CD4CCR5 cell line to achieve 90–95% of HIV infection instead of the 5% obtained in the primary astrocytes. Staining of both cell types for CytC and mitochondria (mito-tracker dye) shows perfect colocalization between the two markers in uninfected conditions (Fig. 2, control). Quantification of the colocalization indicated 100% of CytC localized inside of mitochondria (Fig. 2, black bars). HIV infection of astrocytes, primary and cell line, resulted in movement of mitochondria into HIV positive areas (HIV-p24), and loss of colocalization with CytC (Fig. 2, HIV). In HIV infected cells CytC was mostly distributed in the cytoplasm with minimal localization within mitochondria. Quantification of colocalization between CytC and mito-tracker indicated that 50% of CytC was released into the cytoplasm (Fig. 2, white bars). However, despite mitochondrial dysregulation and release of CytC into the cytoplasm in response to HIV infection, no apoptosis was detected in HIV infected primary astrocytes (98 ± 1.7% survival, n = 12), suggesting that the virus protected HIV infected astrocytes from cell death. In addition, to determine whether mitochondria integrity was fully compromised, we examined the mRNA and protein levels of bcl-2, bcl-x, BAD, BID and Caspase-3, and no changes were detected (data not shown), suggesting that mitochondrial integrity was not fully compromised. Thus, HIV infected cells survived for extended periods of time with no clear signs of stress, supporting the idea that mitochondria are still functional, despite the loss of CytC into the cytoplasm.

Figure 2.

Cytochrome c (CytC), normally concentrated inside of the mitochondria, is released into the cytoplasm of HIV infected astrocytes. We stained CytC (FITC staining, green staining), mitochondria with Mito-tracker (Mtrack, to observe mitochondria, red staining), HIV-p24 (Cy5 staining, to identify HIV infected astrocytes, white), and DAPI (to identify nucleus, blue staining) in uninfected and HIV infected primary cultures of astrocytes. Colocalization of mitochondria and CytC was quantified by using an imaging software, NIS Elements. Mitochondria in uninfected astrocytes (control) were localized throughout the cytoplasm and 100% of CytC was localized inside these structures (black bars and pictures labeled uninfected cells, perfect colocalization in the merge picture). Staining of HIV infected primary cultures (HIV) indicated that cells are positive for HIV-p24, and CytC did not totally colocalized with mito-tracker (see graph, white bars). The loss of colocalization was because of the release of CytC into the cytoplasm of infected astrocytes. Pictures in the right side correspond to a high magnification to observe the colocalization between the different markers. Despite the dysregulation of CytC, no apoptosis was detected in HIV infected cells as described in Fig. 1. n = 6, *p < 0.0042, represents significance as compared to control uninfected cultures. Bar: 50 μm to uninfected cells and 15 to HIV-infected cells.

Bystander apoptosis of uninfected cells induced by a few HIV infected astrocytes is mediated by intracellular calcium signaling, and by a mechanism involving IP3 and CytC

To identify the signaling involved in bystander apoptosis from the few infected astrocytes into the uninfected surrounding astrocytes, we focused our research on intracellular molecules that can diffuse through gap junction channels, specifically calcium, IP3 and cyclic nucleotides.

Blocking gap junction channels using 18-α-glycerritenic acid (AGA, 35 μM, Fig. 3) or octanol (500 μM, data not shown) blocked bystander apoptosis. Previous published data from other groups indicated that dysregulation of CytC has been associated with alterations in IP3, IP3 receptors and calcium dysregulation (Gafni et al. 1997; Boehning et al. 2004, 2005). Thus, we hypothesized that a similar mechanism occurs during HIV infection of astrocytes.

Figure 3.

Blocking cytochrome c (CytC), IP3 and the increase in intracellular calcium induced by the virus reduces bystander toxicity from HIV infected to uninfected astrocytes. Control cultures have minimal apoptosis (Control). HIV infection of astrocytes cultures with HIVADA, resulted as we described previously, in bystader apoptosis of uninfected astrocytes after 7–14 days post-infection (Eugenin and Berman 2007; Eugenin et al. 2011) (HIVADA). Blocking gap junction channels using 18-α-glycerritenic acid (AGA, 35 μM) to reduce the spread of toxic signals generated from HIV infected to uninfected astrocytes completely abrogates bystander apoptosis (HIVADA+AGA). Blocking increases in intracellular calcium with BAPTA-AM (HIVADA+BAPTA-AM) or CytC induced activation of IP3 and calcium release with a specific peptide reduced bystander apoptosis induced by the virus to control levels (HIVADA+IP3/CytC pep). Scramble peptides did not alter bystander apoptosis (data not shown). Calcium ionophore (A23187, 5 μM), cell permeable cAMP (8Br-cAMP, 1 mM) and cGMP (8Br-cGMP, 1 mM) did not alter bystander apoptosis (HIVADA+ cAMP or cGMP). All of these factors can cross gap junction channels. (n = 5, *p < 0.007, represent significance from control conditions, and #p < 0.001, represent significance from HIVADA infected cultures).

The use of BAPTA-AM (5 μM) to abolish increase in intracellular calcium by HIV infection reduced bystander apoptosis to control conditions (Fig. 3). To block CytC induced activation of IP3 and intracellular Ca+2 release blocking peptide, DNKTVTFEEHIKEEHN-BIOPY- 577/618 maleimide, IP3/CytC pep, was used to inhibit the apoptosis cascade wherein CytC binds to IP3 receptors early in apoptosis as described previously (Boehning et al. 2005). Treatment of HIV infected cultures with this peptide resulted in significant reduction of bystander apoptosis of uninfected astrocytes to control, uninfected cell levels (Fig. 3, IP3/CytC pep). In addition to calcium and IP3, cyclic nucleotides, such as cAMP and cGMP, freely cross gap junction channels (Harris 2008). However, increasing the intracellular concentration of cAMP by using 8Br-cAMP (1 mM) or cGMP by using 8Br-cGMP (1 mM) also did not alter bystander apoptosis (Fig. 3). Surprisingly, a non-specific increase in intracellular calcium by using the calcium ionophore, A23187 (1 μg/mL), did not alter bystander apoptosis or survival of HIV infected astrocytes.

To discount any non-specific effects of the blockers used, we quantified the intracellular amount of IP3 and calcium produced in response to electroporation of astrocytes with CytC. Fura-2 fluorescence and ELISA for IP3 [Figure S1, (a) calcium, (b) IP3 generation] were determined at 5, 10, 15, 30, and 60 min post-electroporation of CytC. Electroporation of PBS did not alter intracellular release of IP3 in control conditions, but increased minimally intracellular calcium release (Figure S1a). However, electroporation of CytC resulted in a sustained increase of IP3 and intracellular calcium that was not altered by the blockers used to reduce bystander apoptosis, AGA and IP3/CytC pep (Fig. 1a and b). BAPTA-AM reduced in 38 ± 12.1% the increase in calcium induced by CytC electroporation by 5–15 min, but did not alter production of IP3 (data not shown).

Thus, CytC, IP3 and the localized increase in intracellular calcium are important mediators in the bystander killing triggered by HIV infected astrocytes.

Direct intracellular microinjection of CytC into the cytoplasm of astrocytes results in apoptosis of uninfected, but not of HIV infected astrocytes

To determine whether CytC is toxic inside of uninfected and HIV infected astrocytes, we microinjected CytC by single cell electroporation into the cytoplasm of uninfected and HIV infected astrocytes (U87CD4CCR5 and primary human cultures of astrocytes). After 24 h, apoptosis in the injected and surrounding cells was determined by TUNEL staining (Fig. 4). In control conditions, PBS was microinjected by single cell electroporation and minimal apoptosis was detected (Fig. 4, control conditions). In contrast, microinjection of CytC into the cytoplasm of uninfected cells resulted in almost 100% apoptosis in the microinjected cells, indicating that soluble CytC is highly toxic (Fig. 4a and b, U87CD4CCR5 and primary cultures, respectively). Quantification of apoptosis after microinjection of PBS into the cytoplasm of HIV infected cultures did not result in significant apoptosis in U87CD4CCR5 cells (90–95% of the cells are HIV infected) or primary cultures of astrocytes (5.9 ± 3.2% of the cells are HIV infected).

Figure 4.

HIV infection protects infected astrocytes from apoptosis induced by cytoplasmic cytochrome c (CytC). (a) Microinjection by single cell electroporation of CytC into uninfected U87CD4CCR5 astrocytes resulted in apoptosis as determined by TUNEL staining (control+CytC) as compared to cells microinjected with phosphate-buffered saline (PBS) only (control). Microinjection of CytC into the cytoplasm of HIV-infected U87CD4CCR5 astrocytes (95% of the cells are infected) did not result in apoptosis, suggesting that HIVADA infection protects these cells from the toxicity of cytoplasmic CytC (HIV+CytC). Microinjection of PBS into cells did not trigger apoptosis (HIV). (b) Primary cultures of human astrocytes were microinjected with PBS and this did not result in apoptosis (control). However, microinjection of CytC into uninfected astrocytes resulted in 100% apoptosis, supporting the high toxicity of free cytoplasmic CytC. Microinjection of HIV infected primary cultures, where 5.9 ± 3.2% of the cells are infected with HIV, with PBS resulted in minimal apoptosis (HIV). However, microinjection of CytC into the cytoplasm of HIV infected cultures resulted in a decrease in apoptosis of 6.2 ± 2.9% that corresponds to the fraction of cells infected with HIV (HIV+CytC). n = 6, *p < 0.0005 as compared to control conditions and #p < 0.05 as compared to control+CytC.

Microinjection of CytC into HIV infected cultures of U87CD4CCR5 cells (90–95% become infected with HIV) did not result in apoptosis (Fig. 4, HIV+CytC). In primary cultures in which only a small percentage of the cells were HIV infected, apoptosis was reduced in similar numbers as the percentage of HIV infected cells (6.2 ± 2.9, n = 6, # as compared to CytC injection into uninfected cells). Thus, despite the high toxicity of cytoplasmic CytC in uninfected astrocytes, no toxicity was detected in HIV infected astrocytes, suggesting that HIV has a mechanism of cellular protection against apoptosis to maintain its survival and to establish viral reservoirs.

Discussion

In our previous reports, we demonstrated that HIV infection of astrocytes, despite the low numbers of infected cells (~ 5%) and low to undetectable level of viral replication, results in the spread of toxic signals from the few infected astrocytes into the neighboring uninfected cells through gap junctions (Eugenin and Berman 2007; Eugenin et al. 2011, 2012). Here, we characterized the toxic intracellular signals generated by the few infected cells as subproducts of mitochondrial dysregulation including release of CytC from the mitochondria into the cytoplasm of the infected cells. Free cytoplasmic CytC is highly toxic in uninfected cells, but HIV infection protects infected astrocytes from apoptosis induced by this intracellular mediator. However, because the molecular size of CytC is too large to pass through gap junction channels, it cannot be the signal that triggers bystander apoptosis in neighboring uninfected astrocytes. We demonstrated that the intracellular signals mediating these toxic effects in uninfected astrocytes are most likely intracellular calcium, and IP3.

We demonstrated that HIV infection of astrocytes has profound effects on dysregulation of CytC and protection against apoptosis. Despite mitochondrial dysregulation, and loss of CytC into the cytoplasm, HIV infected astrocytes survive for extended periods of time, underscoring that these cells are important viral reservoirs by maintain their survival. Experiments in the retina demonstrated that cytoplasmatic CytC induces bystander killing between retina cells through a gap junction dependent mechanism (Cusato et al. 2003). However, the size of CytC is approximately 30 kDa; thus, it cannot cross gap junction channels (the maximal size for these channels is 1.2 kDa). Therefore, other second messengers must mediate the bystander killing of surrounding uninfected cells. Our data indicated that HIV infection of astrocytes resulted in dysregulation of CytC and subsequent IP3 and calcium signaling in correlation with apoptosis of uninfected astrocytes. In agreement, studies in muscle cells indicated a close correlation between mitochondrial dysregulation, loss of CytC into the cytoplasm and alterations in IP3 signaling resulting in toxicity and apoptosis (Li et al. 1997; Szalai et al. 1999; Pacher and Hajnoczky 2001; Pacher et al. 2001; Boehning et al. 2003, 2005; Mattson and Chan 2003).

In general, mechanisms that mediate apoptosis can be divided into two major pathways: extrinsic, that is activated by ligands such as member of the tumor necrosis factor (TNF)-α family, and intrinsic, mediated by dysregulation of mitochrondria/CytC. The intrinsic pathway results in the activation of caspases and the cleavage of several intracellular substrates. Cytotoxic agents induce CytC release from mitochondria, which binds to cytoplasmic apoptotic protease activating factor-1 (Li et al. 1997). The complex CytC/apoptotic protease activating factor-1 recruits and activates caspase-9, caspase-3 and raises intracellular levels of calcium resulting in formation of the apoptosome (Boehning et al. 2004). In physiological conditions and in non-excitable cells, including astrocytes, IP3 interacts with IP3R, resulting in the release of calcium from intracellular stores that control physiological signaling without toxic effects. However, in response to apoptotic events, the amount of intracellular calcium released by IP3R into the cytosol is higher than in normal conditions, resulting in calcium accumulation inside of mitochondria and subsequent CytC release (Szalai et al. 1999; Pacher and Hajnoczky 2001; Pacher et al. 2001). The interaction between these two components, IP3R and CytC and their role in apoptosis has been demonstrated by several studies (Boehning et al. 2003, 2005; Mattson and Chan 2003). This binding of CytC to IP3R blocks the calcium dependent inhibition of this ion channel, resulting in extended calcium oscillations. This interaction can be controlled by specific bcl-2 family members. In our system, we did not observe any changes in protein or mRNA levels for these proteins, suggesting a different mechanism of control of IP3R activation. Using this model of interaction, several groups proposed the concept of mitochondrial waves that involves the release of toxic factors that affect other mitochondria in neighboring cells, but these toxic signals are unknown. In our study we identified that calcium and IP3 amplify toxicity through gap junction channels from few HIV infected cells into surrounding uninfected cells resulting in bystander apoptosis. Blocking either signaling component completely reduced bystander apoptosis. Thus, we propose that a similar system of apoptosis described in muscle cells (Boehning et al. 2003, 2005; Mattson and Chan 2003) is also present in HIV infected astrocytes and is amplified by gap junction channels to neighboring uninfected cells.

We propose that survival of HIV infected astrocytes generates CNS viral reservoirs in astrocytes. Similar mechanisms have been described in peripheral cells. CD4+ T lymphocytes are major reservoirs of HIV in the periphery (Chomont et al. 2009; Richman et al. 2009). Long lived infected CD4+ T cells are generated by increased proliferation as well as increased survival (Chomont et al. 2009; Richman et al. 2009). However, the mechanisms that support the formation of these reservoirs are correlated with IL-7 mediated homeostatic proliferation. The molecular mechanisms by which these cells have longer survival and generate viral reservoirs are unknown and requires further investigation.

Thus, we demonstrated that few HIV infected astrocytes trigger bystander apoptosis by a mechanism depend upon gap junctions and mitochondrial dysregulation. We have shown that HIV infection of astrocytes protects these cells from apoptosis and dysregulates mitochondrial integrity, resulting in release of CytC into the cytoplasm, and altering intracellular calcium signaling, and IP3/IP3R response. We propose that the sequence of events in bystander apoptosis involves HIV infection, mitochondrial dysfunction, secretion of CytC into the cytoplasm, dysregulation of IP3/IP3R signaling, impaired intracellular calcium metabolism, and diffusion of second messenger (Ca+2 and IP3) into neighboring uninfected cells, resulting in apoptosis of uninfected astrocytes as shown in Fig. 5. These toxic factors trigger apoptosis in uninfected astrocytes. However, we propose that these toxic signals also require specific protective mechanisms provided by HIV infection to maintain the survival of infected cells, such as loss of IP3R sensitivity to IP3, desensitization to high intracellular Ca+2 or alterations in downstream apoptosis pathways. Our data represent a novel mechanism of toxicity in NeuroAIDS and in the generation of viral CNS reservoirs. Our studies underscore the importance of HIV infection of astrocytes in the development and persistence of the virus within the CNS, and indicate new avenues for therapeutic interventions to reduce and eradicate these cells from the CNS.

Figure 5.

Schematic representation of the mechanism of generation of HIV reservoirs and amplification of toxicity to uninfected astrocytes by a bystander mechanism. Our proposed model is that, cytochrome c (CytC), inositol trisphosphate (IP3) and calcium mediate apoptosis of uninfected astrocytes. HIV infection of astrocytes dysregulates mitochondrial function and CytC metabolism resulting in release of CytC into the cytoplasm that acts as a toxic factor in neighboring uninfected cells. IP3 and calcium can diffuse through gap junctions resulting in bystander apoptosis of uninfected astrocytes that do not have the protection provided by HIV. These mechanisms assure the perpetuation of HIV CNS reservoirs and result in amplification of toxicity in uninfected cells. The crosses in the cartoon denote the potential target pathways for the virus to maintain the survival of the infected astrocytes.

Acknowledgments

We thank the Fetal Tissue Repository at the Albert Einstein College of Medicine. This work was supported by the National Institutes of Mental Health and drug abuse grants, MH096625 (to E.A.E), MH090958, MH075679 and DA025567 (J.W.B). The authors had no financial interest.

Ancillary