Previous Article | Next Article ![]()
Journal of Virology, January 2006, p. 541-544, Vol. 80, No. 1
0022-538X/06/$08.00+0 doi:10.1128/JVI.80.1.541-544.2006
Copyright © 2006, American Society for Microbiology. All Rights Reserved.
Department of Immunology and Microbiology, Rush University Medical Center, Chicago, Illinois
Received 6 July 2005/ Accepted 8 October 2005
|
|
|---|
), granulocyte-macrophage colony-stimulating factor, and tumor necrosis factor alpha, on priming astrocytes for HIV infection. We demonstrate that IFN-
was the most potent in its ability to facilitate substantial productive HIV infection of an astroglioma cell line (U87MG) and human fetal astrocytes (HFA). The mechanism of IFN-
-mediated priming of HIV in HFA is unlikely to be at the level of up-regulation of receptors and coreceptors relevant to HIV entry. These data demonstrate that cytokine priming can alter HIV replication in astrocytes. |
|
|---|
We previously demonstrated that cytokine pretreatment of otherwise non-HIV-permissive lymphocytes renders them permissive to HIV productive replication (1, 26). Therefore, we evaluated whether key cytokines can prime astrocytes to support productive HIV infection. The candidate cytokines, chosen on the basis of their relevance to HIV neuropathogenesis, are tumor necrosis factor alpha (TNF-
), granulocyte-macrophage colony-stimulating factor (GM-CSF), and gamma interferon (IFN-
). TNF-
is secreted by glial cells (3) and linked to demyelination (16), impaired glutamate scavenging by astrocytes (8), and limited induction of HIV from astrocytes in culture (28). GM-CSF is also secreted by glial cells (32) and correlates with elevated viral load in the cerebral spinal fluid (CSF) of HIV-infected individuals and in mixed brain cell culture aggregates (14). GM-CSF induces the expression of the mannose receptor on astrocytes (5), which was identified as an alternate receptor for HIV on astrocytes (17). IFN-
is secreted by lymphocytes which have infiltrated the CNS. IFN-
induces the expression of chemokine coreceptors on simian astrocytes (6, 25) and HIV infection in microglia (13).
To evaluate the impact of TNF-
, GM-CSF, and IFN-
on modulating HIV replication in astrocytes, an astroglioma cell line (U87MG) and primary human fetal astrocytes (HFA) were used. U87MG cells were obtained from the NIH AIDS Research and Reference Reagents Program (Germantown, MD). HFA were purified from second-trimester-aborted fetuses, as described previously (23). A typical level of HFA purity is shown in Fig. 1, where the cultures were >95% positive for the astrocyte-specific marker glial fibrillary acidic protein (Fig. 1A); less than 1% positive for the microglial marker CD68 (Fig. 1A); 11% positive for nestin, a marker for precursor neural cells (Fig. 1B); and 4% positive for neurons as indicated by MAP2 immunostaining (Fig. 1C). Neurons do not support productive HIV replication (22) and eventually die in these cultures. Microglia are extremely adherent and are depleted with the continuous passage of HFA, which was critical given that microglia replicate HIV efficiently (10).
![]() View larger version (29K): [in a new window] |
FIG. 1. Purity of HFA cultures. Flow cytometry was performed to evaluate the purity of the HFA cultures. Cells were stained for an astrocyte marker (glial fibrillary acidic protein [GFAP]) (A, B), a microglia marker (CD68) (A, C), an astrocyte precursor marker (nestin) (B), or a neuron marker (MAP2) (C). Dot plots shown are representative of the typical purity of HFA cultures used throughout this study. APC, allophycocyanin; PE, phycoerythrin; FITC, fluorescein isothiocyanate.
|
, and TNF-
can augment HIV replication of astrocytes, we determined whether U87MG cells and HFA express the respective cytokine receptors. U87MG cells and HFA were treated with or without the respective cytokine for 24 h, and cytokine receptor expression was evaluated by flow cytometry. Approximately 35% of U87MG cells expressed the IFN-
receptor and low levels of the GM-CSF and TNF-
receptors (<8%) (Fig. 2A). After a 24-hour treatment of U87MG cells with IFN-
, IFN-
receptor expression was down-regulated by twofold to approximately 15% (Fig. 2A), while neither GM-CSF nor TNF-
treatment had an effect on their respective receptor expression (Fig. 2A). Untreated HFA expressed only low levels of IFN-
receptor (6%), and this expression was down-regulated with IFN-
treatment (3%). GM-CSF and TNF-
receptors were expressed at background levels on HFA (Fig. 2B).
![]() View larger version (14K): [in a new window] |
FIG. 2. Cytokine receptor expression on astrocytes. U87MG cells (A) or HFA (B) were left untreated or treated with GM-CSF (50 ng/ml), IFN- (100 ng/ml), or TNF- (100 ng/ml). Cytokine concentrations were based on the concentrations determined to induce maximal HIV expression, except for TNF- , which did not induce HIV expression (Fig. 3). Percentages of the expression of the GM-CSF receptor (R), IFN- receptor, and TNF- receptor was measured 24 h poststimulation by flow cytometry. Data are representative of three experiments. Error bars represent standard deviations of the means.
|
, GM-CSF, or IFN-
at various concentrations (0 to 1,000 ng/ml) and then infected with HIV BAL at 10 ng of HIV p24/1 x 106 cells for 24 h. The cells were then either washed extensively or trypsinized in some experiments to remove bound virus and propagated in the presence of the respective cytokine concentration. HIV infection was monitored by p24 enzyme-linked immunosorbent assay 7 days postinfection. The choice of the duration of cytokine pretreatment was based on time kinetics indicating that 24-hour prestimulation yielded the highest level of HIV infection (data not shown). For U87MG cells, the effect of IFN-
appears to be dose dependent, with maximal induction of HIV replication being 10-fold higher than that in infected but untreated cultures observed at 250 ng/ml (Fig. 3A). GM-CSF prestimulation induced a fivefold induction of HIV replication at 10 ng/ml, but higher doses were not as effective (Fig. 3A). TNF-
prestimulation did not induce HIV replication (data not shown). In HFA, IFN-
induced a twofold induction of HIV replication, while GM-CSF (Fig. 3B) and TNF-
(data not shown) had no effect. Combining IFN-
, GM-CSF, and TNF-
to prestimulate U87MG cells and HFA prior to HIV infection had no additive or synergistic effects (data not shown).
![]() View larger version (12K): [in a new window] |
FIG. 3. Effects of GM-CSF and IFN- prestimulation on HIV replication in astrocytes. U87MG cells (A) and HFA (B) were prestimulated with increasing doses of either GM-CSF or IFN- (IFN-g) for 24 h followed by HIV BAL infection for 24 h. HIV replication was determined by measuring HIV p24 levels by enzyme-linked immunosorbent assay 7 days postinfection. The dotted line represents the basal level of HIV infection of astrocytes without cytokine pretreatment. Data are representative of three experiments performed in quadruplicate. Error bar represent standard deviations of the means. Asterisks denote significant values (P < 0.05) in comparison to values for untreated cultures as evaluated by analysis of variance.
|
was the most potent in inducing the HIV infection of astrocytes, we determined whether IFN-
treatment modulates key receptors (CD4) and mannose receptor (17) and coreceptors (CCR1, CCR2, CCR3, and CCR5) for HIV infection. In both U87MG cells and HFA, CD4 and the mannose receptor were not expressed and their expression was not augmented by IFN-
or GM-CSF (Fig. 4). CCR1 expression on U87MG cells was up-regulated by fourfold after treatment with IFN-
, while GM-CSF had no effect on its expression. HFA were positive for CXCR4 (12%), which is consistent with previous reports (15, 21), but its expression was not augmented by GM-CSF or IFN-
treatment (Fig. 4B). IFN-
was reported to enhance CXCR4 and CCR5 expression on simian adult astrocytes, which is in contrast to our report on human fetal astrocytes highlighting a discrepancy between simian adult and human fetal astrocytes (6). Although IFN-
alone does not modulate CCR5 and CXCR4 expression, it can synergize with TNF-
to up-regulate these chemokine coreceptors in human fetal astrocytes (6). Infection of U87MG cells and HFA with a T-tropic isolate (IIIB), while it was productive, was not enhanced by cytokine pretreatment (data not shown). This may be a consequence of the laboratory-adapted HIV IIIB strain; therefore, primary T-tropic or dualtropic strains may still be responsive to IFN-
pretreatment of astrocytes. Recently, the CC chemokine coreceptor D6 was identified as a functional coreceptor for dualtropic HIV infection of astrocytes (19).
![]() View larger version (13K): [in a new window] |
FIG. 4. Effect of GM-CSF and IFN- on the expression of receptors and chemokine coreceptors relevant to HIV replication. U87MG cells (A) and HFA (B) were left untreated or treated with GM-CSF (50 ng/ml) or IFN- (100 ng/ml) for 24 h. Levels of expression of receptors on the x axis were measured by flow cytometry. Isotype refers to mouse immunoglobulin G. Data are representative of three experiments. Error bars represent standard deviations of the means. hMR, human mannose receptor; IFN-g, gamma IFN.
|
pretreatment in particular was an effective inducer of HIV infection in U87MG cells and the only cytokine tested to induce HIV infection in primary fetal astrocytes. Down-regulation of IFN-
receptor after IFN-
stimulation confirms the responsiveness of the cells to this stimulus, which is consistent with the documented feedback mechanism utilized by IFN receptors after treatment with IFNs (9). Given that, in primary fetal astrocytes, IFN-
did not induce the expression of any of the HIV receptors that we evaluated, the mechanism by which IFN-
can induce HIV replication in primary astrocytes may be at postentry events, perhaps by inducing key transcriptional factors in astrocytes. Taken together, our data show that the presence of IFN-
, and possibly other soluble factors found in the CNS milieu in vivo, may account for the discrepancy between postmortem data identifying viral transcripts within astrocytes (2) and in vitro studies demonstrating astrocytes to be resistant to HIV infection (4, 12). The finding that astrocytes are productively infected with HIV is significant because astrocytes make up approximately 50%, while microglia make up 10 to 20%, of the CNS cell population (18, 31). Therefore, even if the efficiency of infection in astrocytes is less than that in microglia, the total viral output from astrocytes may contribute significantly to the CNS viral load.
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»