Oncostatin M promotes excitotoxicity by inhibiting glutamate uptake in astrocytes: implications in HIV-associated neurotoxicity

³H-d-aspartate (specific activity = 10–25 Ci/mmol) was obtained from PerkinElmer (MA, USA) and Aquasafe 500 Plus liquid scintillation cocktail from Zinsser Analytic (Frankfurt, Germany). Neurobasal media, Hank’s buffered salt solution (HBSS), phosphate-buffered saline (PBS), sodium pyruvate, l-glutamine, penicillin-streptomycin, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), glutaMAX-1, and B27 supplement were from Gibco (Breda, The Netherlands). Dulbecco’s modified Eagle medium (DMEM) media and fetal calf serum (FCS) were from PAA Laboratories (Cölbe, Germany). Trypsin was obtained from Life Technologies (Breda, The Netherlands). All other cell medium components, recombinant mouse OSM, recombinant mouse IL-6, d-aspartate, l-leucine methyl ester (LME), N-methyl-d-glucamine (NMG), and the dyes used to stain cell nuclei (Hoechst 33342 and propidium iodide) were purchased from Sigma-Aldrich (Zwijndrecht, The Netherlands). DL-threo-β-benzyloxyaspartic acid (TBOA) and MEK1/2 inhibitor U0126 were obtained from Tocris Bioscience (Bristol, UK). Janus kinase (JAK) inhibitor (AG490) and phosphatidylinositol 3-kinase (PI3K) inhibitor (LY294002) were obtained from Calbiochem (CA, USA). Reagents used in immunoblotting experiments were purchased from Bio-Rad Laboratories, except polyvinylidene fluoride (PVDF) membranes that were obtained from Millipore (Amsterdam, The Netherlands). Primary antibodies, mouse monoclonal anti-β-actin, rabbit polyclonal anti-GLT-1/EAAT2, rabbit polyclonal anti-GLAST/EAAT1 (C-terminus), and FITC-conjugated anti-GFP were obtained from Abcam (Cambridge, UK); mouse monoclonal anti-glial fibrillary acidic protein (GFAP) was obtained from Millipore; goat polyclonal anti-OSMR-β (AF662) was from R&D Systems (Minneapolis, USA); mouse monoclonal anti-MAP2 and mouse monoclonal anti-α-tubulin were obtained from Sigma-Aldrich; and primary antibodies against total and phosphorylated signal transducers and activators of transcription (STAT)3, extracellular signal-regulated kinase ½ (ERK1/2), and Akt were obtained from Cell Signaling Technology (Bioke, Leiden, The Netherlands). The fluorescent dye-conjugated secondary antibodies used for Western blot, donkey anti-mouse IR Dye 680, and donkey anti-rabbit IR Dye 800CW were obtained from LI-COR Biosciences (Cambridge, UK). The fluorescent dye-conjugated secondary antibodies used for immunocytochemistry: goat anti-rabbit CY3 was obtained from Jackson ImmunoResearch Laboratories (Uden, The Netherlands); and donkey anti-mouse Alexa Fluor 488 was obtained from Molecular Probes (Breda, The Netherlands).

Primary astrocyte cultures were established from cerebral cortices of postnatal (1–2 days) C57BL/6J mice according to a previously described procedure [41], modified to reduce microglial contamination [42]. Approximately 2 weeks after plating, microglial cells were mechanically separated from the astrocytic monolayer by shake-off at 150 rpm for 1 h. This procedure was repeated twice with an interval of 4 days in vitro between each shake-off, followed by an overnight shake-off at 240 rpm to remove oligodendrocyte precursor cells. Enriched astrocytes were washed with HBSS buffer (in mM: 137 NaCl, 5.3 KCl, 0.3 Na₂HPO₄, 0.4 KH₂PO₄, 4.2 NaHCO₃, 5.6 d-glucose, pH 7.4) containing 1 mM ethylenediaminetetraacetic acid (EDTA) and further detached using 0.1 % trypsin (diluted in HBSS). Cells were reseeded with fresh astrocyte culture medium (DMEM supplemented with 5 % FCS, 2 mM l-glutamine, 1 mM sodium pyruvate, and 50 U/mL penicillin-streptomycin) in multiwell plates (5 × 10⁴ cells/cm²) and maintained in culture to confluence. To further reduce microglial contamination, confluent astrocyte cultures were treated with 5 mM LME, a lysosomotropic agent [43], for 4–5 h. Astrocytes were ready for experiments after 1–2 days. Our cell preparations had a high percentage of astrocytes (≥95 %), which was confirmed by immunostaining against GFAP (astrocyte marker) and CD11b (microglial marker) (Additional file 1: Figure S1) [44]. Although with low abundance, it is possible that the presence of microglial cells (approx. 5 %) in astrocyte cultures may influence some of the observations in this study. Therefore, additional experiments were carried out to verify astrocyte specificity of OSM effects in astrocyte cultures treated with liposomal clodronate (1 mg/mL for 4 h; Encapsula NanoSciences LLC, TN, USA), which resulted in complete depletion of microglial cells (Additional file 2: Figure S2A).

The murine microglial BV2 cells were maintained in DMEM media containing 10 % FCS at 37 °C in a 5 % CO₂ incubator. Primary microglia cultures were established following their shake-off (150 rpm for 1 h) from mixed glial culture flasks. The cells were cultured in standard DMEM media (supplemented with 5 % FCS, 2 mM l-glutamine, 1 mM sodium pyruvate and 50 U/mL penicillin-streptomycin), diluted in 1:1 ratio with glial-conditioned media.

Primary cultures of cortical neurons from mouse embryo (~E₁₅) were done as described previously [26]. Briefly, cortices from embryonic brains were dissected in ice-cold HBSS supplemented with 30 % glucose. Meninges were removed, and the tissues were treated with trypsin before they were gently dissociated by trituration in neuronal culture media (neurobasal medium supplemented with 2 % B27, 1 mM sodium pyruvate, 2 mM l-glutamine and 50 U/mL penicillin-streptomycin). Cell suspension was filtered using a cell strainer (70 μm; Falcon, Franklin Lakes, NJ, USA) before centrifugation (800 rpm for 10 min). Cells were then seeded on poly-d-lysine (10 μg/mL)-coated 96-well plates (1 × 10⁵ cells/well) and maintained in neuronal culture media in a humidified atmosphere with 5 % CO₂ at 37 °C. Culture media was refreshed 24 h later to minimize culture debris. The neuronal purity as determined by MAP2 staining was around 98 % [26]. Cultures were used after 5 days in vitro.

In the present work, we used EcoHIV/NL4-3-EGFP (referred to as EcoHIV, for brevity), a chimeric HIV-1 expressing enhanced green fluorescent protein (EGFP) as an indicator, which was constructed on the backbone of HIV-1/NL4-3, as described previously [39]. Infectious EcoHIV stocks were propagated in HEK293TN cells as described [40] and titered for the p24 HIV-1 core antigen by ELISA, following the manufacturer’s instructions (ZeptoMetrix Corporation, NY, USA). Cultured astrocytes, microglia, and BV2 cells were incubated with 35,000 pg of p24 (per 1 × 10⁶ cells) during the incubation time stated in the text. Viral infectivity of cells was assessed by HIV LTR (long-term repeat) gene expression and/or anti-GFP immunocytochemistry.

Cultured cells were lysed in guanidinium isothiocyanate/mono-thioglycerol (GTC⁺) buffer and total RNA was extracted using a phenol-chloroform/iso-amyl alcohol step, followed by DNAse 1 treatment. Purified messenger ribonucleic acid (mRNA) was then transcribed into complementary DNA (cDNA) as described previously [45]. The quality of the cDNA was examined using primers for the housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH; see Table 1). Potential contamination of mRNA samples by genomic deoxyribonucleic acid (DNA) was checked by running reactions without reverse transcriptase, using GAPDH primers for the subsequent PCR amplification.

The expression of OSM, OSMR-β, gp130, GLAST, GLT-1, and HIV LTR genes was analyzed by real-time PCR using an iCycler (Bio-Rad, Veenendaal, The Netherlands) and iQ SYBR Green supermix (Bio-Rad). The housekeeping gene, hypoxanthine phosphoribosyl transferase 1 (HPRT1), was used for normalization, and it showed no variations in response to the experimental treatment (see Table 2 for primer sequences). The comparative C _t method (amount of target amplicon X in sample S, normalized to a reference R and related to a control sample C, calculated by 2 − ((C _t X,S − C _t R,S) − (C _t X,C − C _t R,C)) was used to determine the relative expression levels of all tested genes [46]. Linear regression analysis of the data was performed to understand the effect of OSM treatment over time on the expression of GLAST and GLT-1 mRNA.

Western blotting of primary astrocyte and microglia cell lysates was performed as previously described [26]. Protein calibration controls were performed to ensure that the chosen working quantities were at non-saturating conditions (Additional file 3: Figure S3B-C). Briefly, equal amounts of protein (30 μg) were loaded onto 12.5 % sodium dodecyl sulfate-polyacrylamide gels and subsequently transferred to PVDF membranes. The membranes were blocked using Odyssey blocking buffer (OBB; diluted 1:1 in PBS) for 1 h and incubated overnight at 4 °C with different combinations of primary antibodies (diluted in 1:1 OBB and PBS-T (PBS + 0.1 % Tween 20)): goat anti-OSMR-β, rabbit anti-GLAST, rabbit anti-GLT-1, mouse anti-β-actin, mouse anti-α-tubulin, mouse anti-STAT3, mouse anti-p44/42 mitogen-activated protein kinase (MAPK), mouse anti-Akt (Pan), rabbit anti-phospho STAT3 (Tyr705), rabbit anti-phospho-p44/42 MAPK (Thr202/Tyr204), and rabbit anti-phospho-Akt (Ser473). The next day, membranes were washed in PBS-T (4 × 5 min) and incubated for 1 h at room temperature (with gentle shaking in the dark) with appropriate fluorescent dye-conjugated secondary antibodies (diluted in PBS-T): donkey anti-goat IR Dye 800CW, donkey anti-mouse IR Dye 680, and donkey anti-rabbit IR Dye 800CW. Membranes were washed again in PBS-T (4 × 5 min), and fluorescent bands were detected using LI-COR’s Odyssey infrared imaging system. The densitometry analysis of protein bands was performed using ImageJ software (NIH) [47].

Supernatants from cultured BV2, primary shake-off microglia, and primary astrocytes were collected following EcoHIV infections, to measure secreted OSM levels using a mouse OSM ELISA-kit (USCN Life Science Inc., TX, USA) following the manufacturer’s instructions. Secreted IL-6 levels in astrocyte culture supernatants were measured using a mouse IL-6 ELISA Ready-SET-Go kit (Affymetrix, eBioscience, CA, USA), following the manufacturer’s instructions.

d-Aspartate was used as index of glutamate uptake in this study as it is a substrate for the high-affinity glutamate transporters with slow intracellular metabolization, thus allowing a more precise measure of glutamate transporter activity [48‐50]. The analysis of ³H-d-aspartate uptake was evaluated as previously described [41]. Briefly, cultured astrocytes were incubated with Krebs buffer (in mM: 132 NaCl, 4 KCl, 1.2 Na₂HPO₄, 1.4 MgCl₂, 6 glucose, 10 HEPES, 1 CaCl₂, pH 7.4) containing ³H-d-aspartate (0.1 μCi/mL) and 50 μM d-aspartate for 10 min at 37 °C. Subsequently, the medium was removed and the cultured cells were placed on ice and washed twice with cold NMG buffer (where NaCl is replaced by N-methylglucamine, NMG) to terminate the uptake process. The cells were lysed with 0.5 M NaOH and transferred to a scintillation vial to be mixed with liquid scintillation cocktail. The radioactivity content (disintegrations per minute) was determined using liquid scintillation counting on a TRICARB® 2900TR analyzer. The remaining cell suspension was used to determine the protein content with the bicinchoninic acid (BCA) method (Pierce Technology) [51]. The uptake rate was expressed as the uptake per minute per milligram of protein. For saturation kinetics assays, the total d-aspartate concentrations ranged from 5 to 200 μM. The kinetic constants (i.e., maximum velocity, V _max, and Michaelis-Menten constant, K _M ) were determined by using nonlinear regression fit of the data with a rectangular hyperbola, using the GraphPad Prism software (version 5.02, GraphPad Software Inc, La Jolla, CA, USA).

Confluent astrocyte cultures were treated without or with OSM (10 ng/mL for 24 h) in the presence or absence of STAT3 activation inhibitor AG490 (25 μM, added 2 h before OSM). The cultures were then washed with warm HBSS and incubated with glutamate (100 μM, diluted in neuronal culture media) at 37 °C for 30 min. The supernatants from astrocytes incubated with glutamate, hereinafter referred to as glutamate^inc., were collected and applied (diluted 1:1 with neuronal media) onto 5 days old embryonic cortical neuron cultures and incubated for 1 h at 37 °C. Some neuron cultures were also treated with 50 μM glutamate for 1 h (positive control for excitotoxicity). Where indicated, neurons were pre-incubated with the NMDA receptor antagonist, dizocilpine (MK 801; 30 μM) for 30 min before the glutamate or the glutamate^inc. treatment. Following glutamate treatment, the neuronal medium was refreshed and cultures were incubated for the indicated period of time before they were assessed for cellular dysfunction and cytotoxicity using MTT assay and propidium iodide labeling, respectively.

Existing literature demonstrates that, in mice, OSM acts on target cells through a receptor complex consisting of a ligand recognition subunit (OSMR-β) and a signal transducing subunit (gp130) [53, 54]. Furthermore, all receptor components required for OSM signaling are expressed in human astrocytes [36]. In the present work, we first confirmed expression of gp130 and OSMR-β mRNA in cultured cortical astrocytes established from neonatal (P2) mouse brains (Fig. 1a). We next investigated whether activation of OSM receptors in cortical astrocytes regulates expression of GLT-1 and GLAST genes in vitro. Cultured astrocytes treated with recombinant mouse OSM (10 ng/mL) for different time periods (2, 4, 8, 12, and 24 h) were analyzed for GLT-1 and GLAST mRNA expression by real-time PCR (Fig. 1b). Linear regression analysis showed a time-dependent reduction of the expression of GLT-1 (R ² = 0.415, p = 0.001, n = 3) and GLAST (R ² = 0.525, p < 0.001, n = 3) mRNA in OSM-treated astrocytes compared to the control; this effect was apparent after 4 and 8 h of OSM incubation, for GLAST and GLT-1, respectively (Fig. 1b). In order to understand whether the 5 % microglial cells present in astrocyte cultures influence this effect of OSM, we depleted microglial cells with liposomal clodronate (1 mg/mL for 4 h) before addition of OSM (Additional file 2: Figure S2A-B). Treatment with OSM (10 ng/mL for 24 h) of microglia-depleted astrocyte cultures reduced both GLT-1 and GLAST mRNA (p < 0.001, n = 3) (Additional file 2: Figure S2B), and this effect of OSM was comparable to the observations made from astrocyte cultures containing 5 % microglia (Fig. 1b) further establishing that OSM receptors are primarily expressed in astrocytes.

It is well known that astrocyte cultures, due to the absence of soluble neuronal factors, express GLT-1 protein at very low levels [55, 56]. Consistently, in our study, GLT-1 protein levels in cultured cortical astrocytes were minimally detected by Western blot (Additional file 3: Figure S3C-D) and thus making them difficult to quantify. In contrast, GLAST proteins were robustly expressed in astrocyte cultures (Additional file 3: Figure S3B). Treatment with OSM (10 ng/mL) induced a time-dependent reduction of GLAST protein density (Fig. 1c), consistent with the reduction of GLAST mRNA levels (Fig. 1b). When compared to the untreated control, a 5-h OSM treatment significantly decreased of GLAST protein density (30.0 ± 7.0 %, n = 3, p < 0.01), while a 24-h OSM treatment further decreased GLAST protein density (60.1 ± 2.4 %, n = 3, p < 0.001) (Fig. 1c).

We next investigated whether the down-regulation of GLT-1 (mRNA) and GLAST (mRNA and protein) by OSM had consequences on the ability of astrocytes to clear extracellular glutamate. The rate of ³H-d-aspartate uptake, as determined by the saturation isotherm, was maximal when astrocytes were incubated with 50 μM of ³H-d-aspartate for 10 min at room temperature (Fig. 2a). As shown in Fig. 2b, cultured astrocytes treated with OSM (1 or 10 ng/mL, for 24 h) displayed a concentration-dependent decrease of ³H-d-aspartate uptake (OSM 1 ng/mL: 12.2 ± 2.2 % decrease, n = 3, p < 0.05; OSM 10 ng/mL: 24 ± 4 % decrease, n = 3, p < 0.01).

It has been previously shown that OSM binding to the OSMR-β/gp130 receptor complex results in phosphorylation of tyrosine residues of the gp130 receptor; this triggers the recruitment, phosphorylation, and activation of STAT proteins, mainly STAT1, STAT3, and STAT5 [57‐59]. In addition, OSM activates other signaling pathways such as phosphatidylinositol 3-kinase (PI3K/Akt) [60] and the MAPK, namely, ERK1/2, p38, and c-jun N-terminal kinases/stress-activated protein kinases (JNK/SAPK) [61]. We next investigated whether treatment with OSM induced phosphorylation of the three principal downstream pathways: JAK/STAT3, PI3K/Akt, and ERK1/2-MAPK [35] in cultured astrocytes. As shown in Fig. 3a, basal activation of ERK1/2 and Akt proteins, but not STAT3, was observed in untreated (control) primary astrocytes. OSM treatment (10 ng/mL for 1 h) induced the phosphorylation at Tyr705, Ser473, and Thr202/Tyr204 residues of STAT3, Akt, and ERK1/2 proteins, respectively (Fig. 3a). Treatment of cultured astrocytes (2 h prior to OSM) with selective inhibitors of PI3K (LY294002, 25 μM), MEK1/2 (U0126, 5 μM), and JAK (AG490, 25 μM) inhibited OSM-induced phosphorylation of Akt, ERK1/2, and STAT3 proteins, respectively (Fig. 3a). In the subsequent experiments, we investigated how selective inhibition of PI3K/Akt, ERK1/2, and JAK/STAT3 signaling pathways regulated OSM-induced reduction of glutamate transporter activity in cultured astrocytes. Treatment with OSM or AG490 did not affect the survival of cultured astrocytes (Fig. 3b), whereas both LY294002 and U0126 induced about 20 % reduction (p < 0.01) in astrocyte survival (Fig. 3b). The blockade of Akt, ERK1/2, and STAT3 signaling pathways occluded the OSM-induced inhibition of ³H-d-aspartate uptake (Fig. 3c). However, whereas AG490 did not modify astrocytic ³H-d-aspartate uptake, both LY294002 and U0126 decreased per se the uptake of ³H-d-aspartate (Fig. 3c). Taken together, these results indicate that STAT3 activation is sufficient to mediate the ability of OSM to impair glutamate uptake in mouse cortical astrocytes.

We next tested if this ability of OSM to inhibit astrocytic glutamate uptake translated into a detrimental effect of OSM on neurons. In order to investigate this, cultured astrocytes with or without OSM pre-treatment (10 ng/mL for 24 h) were incubated with glutamate (100 μM for 30 min). The glutamate^inc. (see Methods) was then applied to neuronal cultures. After 1 h of incubation at 37 °C, the medium was refreshed and neuronal viability was assessed after 4 and 24 h by immunocytochemistry and MTT assay, respectively (Fig. 4a, b). As a positive control, we used a direct addition of glutamate (50 μM) to neuronal cultures, which decreased neuronal viability by approximately 45 % (Fig. 4a) [26]. This glutamate-induced neuronal excitotoxicity was NMDA receptor-dependent since MK 801 (NMDA receptor antagonist; 30 μM) abolished the toxic effect of glutamate (F = 13.321; p < 0.001) (Fig. 4a). Finally, we observed that glutamate^inc. from OSM-treated astrocytes, but not untreated astrocytes, also reduced neuronal viability (OSM treated 67.1 ± 14.24 % survival, n = 4, p < 0.001 and untreated 98.4 ± 5.4 % survival, n = 4), when compared to the untreated control (Fig. 4a). These results were further confirmed by direct measure of neurotoxicity using PI staining (Fig. 4b). By contrast, neuronal viability was preserved when the cultures were co-treated with MK-801 (Fig. 4a), confirming that the neurotoxicity induced by glutamate^inc. from OSM-treated astrocytes was dependent on neuronal NMDA receptor activation. Since OSM-induced inhibition of glutamate uptake in cultured astrocytes was reversed by the JAK/STAT3 signaling inhibitor, AG490 (Fig. 3c), glutamate^inc. from cultured astrocytes that were pre-treated simultaneously with OSM and AG490 should not cause neuronal cell death. Indeed, neuronal viability was unchanged by glutamate^inc. from OSM-treated astrocytes where JAK/STAT3 activation was blocked (F = 13.321; p < 0.001) (Fig. 4a), reinforcing that OSM-induced inhibition of astrocytic glutamate transporter activity is regulated by the JAK/STAT3 signaling pathway.

A recent study showed that PBMCs isolated from neurologically compromised HIV-1-infected patients spontaneously secreted high levels of OSM, whereas OSM levels secreted from PBMCs of neurologically asymptomatic HIV-1 patients were comparable to that of age-matched healthy subjects [20]. Furthermore, this OSM produced by PBMCs isolated from HIV-1-infected individuals induces toxicity in cultured primary human fetal neurons [29]. These findings strongly suggest an important role for OSM in neuropathogenesis and/or neurocognitive impairments observed in HIV-1-infected patients. However, it is not known if HIV-1 infection induces OSM production in microglial cells, a potential source for OSM in the CNS [18, 62]. In order to address this question, we next investigated the expression and release of OSM in cultured microglial cells that were infected with EcoHIV. The viral infection of BV2 cells for 4 h increased GFP-immunoreactivity (Fig. 5a), indicating successful viral infection. The viral infectivity of BV2 cells increased with time, as HIV LTR gene expression following 24 h of EcoHIV incubation was significantly higher when compared to that at 4 h of viral incubation (approx. 10³-fold increase at 4 h, p < 0.001; and 10⁶-fold increase at 24 h, p < 0.0001, n = 6) (Fig. 5b). Next, real-time PCR analysis of OSM gene in EcoHIV-infected BV2 cells showed a fivefold increase in expression of OSM mRNA both at 4 h (p < 0.05) and 24 h (p < 0.01) of viral incubation, when compared to uninfected controls (Fig. 5c). In addition, EcoHIV induced OSM release from BV2 cells, since the supernatants from infected cells (24 h following viral incubation) contained higher levels of secreted OSM (n = 2 in triplicates, p < 0.01), compared to the uninfected controls (Fig. 5d). We further confirmed the above findings in primary mouse microglia, as the culture supernatants contained elevated OSM levels (n = 6, p < 0.05) following 24 h of EcoHIV infection (Fig. 6a, b).

We next analyzed whether EcoHIV infection regulates expression of OSM receptor subunits (gp130 and OSMR-β) in cultured primary microglia. As shown in Fig. 6c, EcoHIV-infected microglia (following 24 h of treatment) showed a 2.4-fold increase in gp130 mRNA (n = 3, p < 0.05), when compared to the uninfected control. On the other hand, OSMR-β mRNA expression in primary microglia was not affected by EcoHIV infection (Fig. 6c). We further analyzed OSMR-β protein expression in primary microglia by Western blot. We observed that neither control nor EcoHIV-infected microglia express detectable levels of OSMR-β proteins (Fig. 6d). These observations are in line with a recent study that showed OSMR-β proteins are absent in microglial cells [63].

Like microglia, astrocytes may also serve as a source for OSM in the CNS [18], since OSM immunoreactivity was detected in brains from MS patients in microglia, reactive astrocytes, and infiltrating leukocytes [18]. We next investigated whether EcoHIV induces OSM release from astrocytes. Cultured cortical astrocytes were infected for 24 h with EcoHIV, as confirmed by the HIV LTR gene expression (Fig. 7a), and culture supernatants were analyzed for secreted OSM levels. Control astrocytes secreted low levels of OSM (approximately 40 pg/mL) (Fig. 7b). However, in contrast to microglia, EcoHIV infection did not increase OSM secretion from cultured astrocytes (Fig. 7b).

It has been shown that HIV-1 directly inhibits glutamate transporter expression and glutamate uptake in human fetal astrocytes, without induction of pro-inflammatory mediators [64], and this effect was largely reproduced by treatment with the HIV-1 envelope protein gp120 alone [64]. The chimeric EcoHIV that we used in the present study does not contain gp120, as it is replaced with gp80 from murine leukemia virus [39, 40]. Thus, EcoHIV serve as a good model to investigate gp120-independent effects of HIV-1 in murine cells. We next investigated how EcoHIV infection regulated GLAST and GLT-1 genes in cultured cortical astrocytes. Interestingly, cultured astrocytes infected for 24 h with EcoHIV showed a 2.5-fold increase in GLT-1 mRNA (n = 3, p < 0.05) (Fig. 7c). By contrast, GLAST gene expression was unchanged in EcoHIV-infected astrocytes, when compared to the untreated control (Fig. 7c). We next investigated whether EcoHIV-infected astrocytes displayed an enhanced gene expression of OSM receptor subunits, OSMR-β, and gp130. As shown in Fig. 7d, EcoHIV infection did not affect the expression of gp130 mRNA in cultured astrocytes. On the other hand, EcoHIV-infected astrocytes showed a fivefold increase in the expression of OSMR-β mRNA (n = 3, p < 0.01), compared to the uninfected control (Fig. 7d). Enhanced astrocytic expression of OSMR-β following EcoHIV infection was also observed at the protein level (n = 3, p < 0.05) (Fig. 7e). Taken together, our findings suggest that HIV-1 may additionally inhibit astrocytic glutamate uptake by a gp120-independent mechanism that involves a combined alteration of GLT-1 expression and the induction of OSMR-β expression and signaling in astrocytes.