This is the first report to show that treatment with OSM, a member of the IL-6 cytokine family, reduces glutamate uptake in cultured cortical astrocytes and thereby promotes excitotoxic death of cortical neurons in vitro. This effect of OSM is mediated by the down-regulation of the two Na+-dependent glutamate transporters, GLAST, and GLT-1. As shown in Fig. 1b, OSM treatment (10 ng/mL) reduces the expression of GLAST and GLT-1 mRNA in a time-dependent manner. The down-regulation of GLAST by OSM was also confirmed at the protein level; however, the low signals obtained with the GLT-1 antibody in the Western blot analysis precluded any reliable analysis of the impact of OSM on GLT-1 protein levels in cultured astrocytes (Additional file 3: Figure S3C-D). Consistent with the down-regulation of glutamate transporter expression, OSM inhibited 3H-d-aspartate uptake by astrocytes in a concentration-dependent manner (Fig. 2b), which mostly involved the recruitment of the JAK/STAT3 pathway (Fig. 3c), rather than the PI3K or ERK1/2 pathways. This is in agreement with the previously reported ability of OSM to induce STAT3 phosphorylation in astrocytes [65], which we now confirmed (Fig. 3a). We further showed that down-regulation of glutamate transport in astrocytes by OSM decreased survival of cultured cortical neurons. As shown in Fig. 4a, glutamateinc. (see the “Methods” section) from untreated astrocytes did not affect the survival of cortical neurons, suggesting that extracellular glutamate is rapidly taken up by astrocytes. However, glutamateinc. from OSM-treated astrocytes did cause excitotoxicity in cultured cortical neurons (Fig. 4a, b). It has been previously shown that gp130-mediated STAT3 activation precedes reactive gliosis in mouse astrocytes [33], which might lead to nitric oxide-induced inflammatory death of neurons [66]. In our current in vitro study, conditioned media from OSM-treated astrocyte cultures did not affect neuronal survival (Fig. 4a), and OSM treatment did not induce the production of nitric oxide in astrocyte cultures (Additional file 4: Figure S4), thus excluding the possibility of indirect oxidative stress-induced neuronal damage. These observations are supported by a recent study, where the authors showed that nitric oxide synthase is not induced by OSM in primary astrocytes and microglia [63]. Importantly, glutamateinc. from OSM-treated astrocytes did not affect survival of neurons in the absence of neuronal NMDA receptor activity (Fig. 4a), suggesting that the increased neurotoxicity results from the decreased glutamate uptake in OSM-treated astrocytes.
The regulation of extracellular glutamate by astrocytes is determined by the density and activity of both glutamate transporters and glutamine synthetase, the enzyme that converts glutamate to glutamine [49]. Whether or not OSM treatment regulates glutamine synthetase expression or activity has not been addressed in this study. On the other hand, we showed that OSM treatment induces the expression of GFAP, COX2, and OSMR-?, but not gp130, in cultured astrocytes (Additional file 5: Figure S5A-C). In line with our findings, OSM has been previously shown to induce pro-inflammatory factors such as GFAP and COX2, among others in astrocytes [30, 35]. In addition, gp130-mediated STAT3 activation in striatal astrocytes has been reported to be closely associated with neuronal damage in a 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) model of neurodegeneration in vivo [33]. Furthermore, OSM/gp130-mediated STAT3 activation has been shown to mediate methamphetamine-induced astrogliosis [32]. Based on these findings, we provide direct evidence that the activation of OSM receptor, triggering STAT3 signaling in astrocytes, impacts neuronal survival. Thus, blockade of STAT3 signaling in astrocytes might be beneficial to prevent excitotoxic neuronal death in models pertinent to many brain injuries with an inflammatory profile [33].
Astrocyte dysfunction resulting in deficient glutamate uptake and metabolism has been reported to be a major contributing factor to the excitotoxic death of neurons in different CNS disease conditions, including HAND [11, 12]. Thus, the identification of essential factors that regulate astrocytic glutamate transporter expression and activity might be beneficial in HAND treatment. It has been reported that HIV-1 could directly inhibit glutamate transporter expression and uptake in human fetal astrocytes, without induction of pro-inflammatory mediators such as TNF-? [64]. The authors of this study showed that the envelope protein, gp120, alone induced effects similar to these of HIV-1 [64]. In our study, we describe a potential gp120-independent mechanism for HIV-induced down-regulation of astrocytic glutamate transport using EcoHIV, a chimeric HIV-1 that can infect mouse cells [39, 40]. EcoHIV itself did not down-regulate GLAST expression in cultured mouse cortical astrocytes (Fig. 7c), whereas it enhanced GLT-1 mRNA expression by 2.5-fold. Interestingly, we show that EcoHIV infection induces a fivefold increase in OSMR-? mRNA and proteins in cultured astrocytes (Fig. 7d, e), whereas OSMR-? proteins in cultured microglia were undetectable by Western blot, as described previously [63]. We also provide evidence that EcoHIV infection induces OSM mRNA expression and protein release in BV2 cells and primary microglia (Figs. 5 and 6), but not in cultured astrocytes (Fig. 7b). In addition, it is noteworthy that the secreted OSM levels in untreated microglial culture supernatants were drastically higher, when compared to that of astrocytic culture supernatants (approximately 2000 and 40 pg/mL, respectively) (Figs. 6b and 7b), suggesting that microglial cells might be a better source for OSM release in the CNS [18, 62]. Further analysis using astrocyte-selective OSMR-? deficient animals or cell cultures would provide better insight into the role of OSM in EcoHIV-mediated neuropathogenesis and/or impaired astrocytic glutamate uptake. Several other pro-inflammatory mediators have been shown to regulate GLAST and GLT-1 expression in astrocytes, including TNF-? and IL-1? [67]. Real-time PCR analysis of TNF-?, IL-1?, cyclooxygenase-2 (COX-2), iNOS, and IL-6 genes in primary microglia showed an induction of these genes following EcoHIV infection (Additional file 6: Figure S6), indicating the complexity of the inflammatory processes that lead to impaired astrocytic glutamate uptake associated with EcoHIV infection.
Comparison of the findings reported in the present study with previous reports suggests that OSM has a complex profile of action in the control of neurodegeneration. In fact, previous studies demonstrated anti-inflammatory as well as neuroprotective properties of OSM, both in vitro and in vivo. For example, OSM inhibits production of pro-inflammatory mediators such as TNF-?, granulocyte macrophage colony-stimulating factor (GM-CSF), and IL-8 [24, 25] and has been shown to suppress inflammatory processes associated with the murine experimental allergic encephalomyelitis model of MS [24]. In addition, we have previously provided evidence for a neuroprotective effect of OSM against glutamate by up-regulating neuromodulatory adenosine A1 receptors [26]. In another study, direct activation of neuronal OSM receptors down-regulated the NR2C subunit of NMDA receptors and thereby prevented NMDA-induced toxicity [27]. More recently, the complex role of OSM signaling was further demonstrated by the reported neuroprotective activity of OSM against ischemic stroke, which is dependent on neuronal OSMR-? expression and activation, with decreased neuronal OSMR-? expression leading to worse stroke outcomes [28]. Taken together these findings and our present study, it may be concluded that the target cell addressed by OSM largely determines the pro- and anti-excitotoxic effects of this cytokine in the CNS.
In the mammalian brain, astrocytes are the predominant players in regulating the glutamate diffusion and spill over from perisynaptic areas, a pre-requisite process to maintain the high signal-to-noise ratio for synaptic communication [6, 49]. Therefore, compromised astrocytic glutamate uptake function caused by the overproduction of cytokines such as OSM, concomitant or resulting from brain injury, might synergistically exacerbate the accumulation of extracellular glutamate at excitotoxic concentrations leading to neuronal damage. In spite of their various suggested roles in astrocytic metabolism, IL-6 family members such as OSM have been scantily explored in their effects on glutamate uptake. There is evidence that CNTF, in contrast to our present findings with OSM, enhances both expression and activity of GLT-1 in astrocytes [68] and thereby promotes survival of neurons against excitotoxicity [69]. Several reports suggest an induction of IL-6 in astrocytes by variety of HIV proteins such as Tat, gp120, Nef, and Vpr [70–73]. Consistently, we show here that infection with EcoHIV virus for 24 h induces several-fold increase in IL-6 secretion in primary astrocytes (Additional file 7: Figure S7). However, IL-6 was shown to have no effect on glutamate uptake on cultured murine astrocytes [74, 75], although it suppressed the increased glutamate uptake induced by prostaglandin E2 (PGE2) [74]. Our preliminary findings show that IL-6 treatment (10 ng/mL for 24 h) did not significantly reduce GLT-1 mRNA expression in cultured astrocytes (p?=?0.08, ?=?5) (Additional file 8: Figure S8). On the other hand, we observed approximately 20 % reduction of GLAST mRNA (p?=?0.04, ?=?5) in IL-6-treated astrocyte culture (Additional file 8: Figure S8). However, this effect of IL-6 is mild, compared to the effect of OSM (10 ng/mL for 24 h) on GLAST gene expression in astrocytes (Fig. 1b and Additional file 2: Figure S2B), and requires further validation at the protein level. Taken together, in this study, we have shown that OSM, through STAT3 activation, impairs the capacity of astrocytes to remove glutamate from extracellular space, which may contribute to excitotoxic neuronal damage. This indicates that a better understanding of OSM signaling mechanisms regulating glutamate transporter level and activity may have important implications for developing novel strategies to limit excitotoxic brain damage in acute and neurodegenerative pathologies.