Microglial regulation of immunological and neuroprotective functions of astroglia

Abstract

Microglia and astroglia play critical roles in the development, function, and survival of neurons in the CNS. However, under inflammatory conditions the role of astrogliosis in the inflammatory process and its effects on neurons remains unclear. Here, we used several types of cell cultures treated with the bacterial inflammogen LPS to address these questions. We found that the presence of astroglia reduced inflammation-driven neurotoxicity, suggesting that astrogliosis is principally neuroprotective. Neutralization of supernatant glial cell line-derived neurotrophic factor (GDNF) released from astroglia significantly reduced this neuroprotective effect during inflammation. To determine the immunological role of astroglia, we optimized a highly-enriched astroglial culture protocol and demonstrated that LPS failed to induce the synthesis and release of TNF-α and iNOS/NO. Instead we found significant enhancement of TNF-α and iNOS expression in highly-enriched astroglial cultures required the presence of 0.5-1% microglia, respectively. Thus suggesting that microglial-astroglial interactions are required for LPS to induce the expression of pro-inflammatory factors and GDNF from astroglia. Specifically, we found that microglia-derived TNF-α plays a pivotal role as a paracrine signal to regulate the neuroprotective functions of astrogliosis. Taken together, these findings suggest that astroglia may not possess the ability to directly recognize the innate immune stimuli LPS, but rather depend on crosstalk with microglia to elicit release of neurotrophic factors as a counterbalance to support neuronal survival from the collateral damage generated by activated microglia during neuroinflammation.

Keywords: astroglia; glial cell line-derived neurotrophic factor; glial interaction; microglia; neuroinflammation; neuroprotection.

Figure 1

Astroglia improve the outcome of LPS-mediated DAnergic neurodegeneration in neuron-glia cultures. (A) Neuron-glia and neuron-microglia cultures were treated with vehicle or of LPS (15 ng/ml). After 7 days, [³H] DA uptake assay was performed to assess DA neuron function. DA uptake function of DAnergic neurons had significantly less reduction in DA uptake in LPS-treated (15 ng/ml for seven days) neuron-glia (containing neuron, microglia and astroglia) cultures compared to neuron-microglia (absence of astroglia) cultures (F_(1,12)= 188.4, P< 0.0001; t=19.41, *P < 0.05, post hoc analysis by Bonferroni t-test). (B) After 7 days, cultures were also subjected to immunocytochemical staining with anti-tyrosine hydroxylase (TH) antibody. The number of TH immunoreactive cells were significantly less depleted in LPS-treated neuron-glia cultures than neuron-microglia (F_(1,12)= 60.71, P< 0.0001; t=11.02, *P < 0.05 post hoc analysis by Bonferroni t-test). Data show mean ± SEM from four experiments, each done in triplicate.

Figure 2

Neuron-glia cultures incubated an hour prior to LPS stimulation (15 ng/ml for seven days) with 20 μg/ml of anti-GDNF antibody displayed a significant reduction in DA uptake function (A) (LPS vs. GDNF Ab, t=6.334, *P < 0.05; GDNF Ab vs. isotype control IgG, t=10.40, *P < 0.05, post hoc analysis by Bonferroni t-test), but not altered the culture supernatant TNF-α level (B) compared to stimulated cultures with and without isotype control antibodies. Data show mean ± SEM from four independent experiments done in duplicate.

Figure 3

Assessment of microglia presence in highly-enriched astroglial cultures. (A) Three days after seeding, primary rat mixed glia cultures were treated with various concentrations of LME (1 mM) for 4 days. Cultures were immunostained using antibodies against GFAP (green, for astroglia) and Iba-1 (red, for microglia). Scale bar: 50 μm. (B) LME was removed from highly-enriched astroglial cultures and Iba1-positive cells were counted at different times thereafter to assess the repopulation of contaminating microglia. t=17.53, *** P < 0.0001, post hoc analysis by Bonferroni t-test compared to Day 0.

Figure 4

Astroglial function in the presence of LME was assessed by their ability to stimulate greater GDNF mRNA levels by the HDAC inhibitor suberoylanilide hydroxamic acid (SAHA) (Chen et al. 2012). Mixed glial cultures (containing both microglia and astroglia) and highly-enriched astroglial cultures were treated with 1.25 μM of SAHA for 12 hours and changes in GDNF (A) and BDNF (B) mRNA expression levels were assessed. The basal level of GDNF or BDNF mRNA in mixed glia and highly-enriched astroglial cultures is comparable. To emphasize the effect of SAHA treatment, we set vehicle-treated group as 100%. Values were present as mean ± SEM from three independent experiments, with duplicates. ** P <0.001, *** P < 0.0001, Bonferroni t-test compared to basal expression levels of vehicle-treated mixed glia culture.

Figure 5

Microglia-derived soluble factors from LPS stimulated cultures are required in the upregulation of GDNF mRNA during astrogliosis. (A) GDNF mRNA was significantly increased 3 and 6 hours after LPS stimulation (15 ng/ml) in mixed glial cultures (containing both microglia and astroglia) [3 hours, t=13.46, ***P<0.0001; 6 hours, t=6.402, **P<0.01, post hoc analysis by Bonferroni t-test compared to 0 hour value] but not in enriched microglia or highly-enriched astroglial cultures. Due to the great difference of GDNF mRNA level among these cultures, we set maximum value as 100%. Data were presented as percent of GDNF mRNA expression with respect to the expression of LPS-treated mixed glia cultures at 3 hours. (B) Highly-enriched astroglial cultures treated for 3 hours with microglia conditioned medium (MCM) from LPS-stimulated enriched microglia collected after 24 hours showed a significant increase in GDNF mRNA expression (t=28.47, ***P<0.0001, post hoc analysis by Bonferroni t-test compared to vehicle treated cultures). No changes were observed when treated with boiled MCM, unstimulated microglial conditioned medium or unstimulated microglial conditioned media plus LPS (15 ng/ml) compared to the vehicle control. To emphasize the effect of MCM treatment, we set vehicle-treated group as 100%. Data were presented as percent of GDNF mRNA expression with respect to the expression of the vehicle control. Data show mean ± SEM from three independent experiments, each done in duplicate.

Figure 6

TNF-α is pivotal in LPS-induced increase in astroglial GDNF expression. Mixed glia cultures prepared from (A) TNF-α KO or (B) TNF R1/R2 KO mice showed a significant decrease in GDNF mRNA expression 3 hours after LPS (15 ng/ml) stimulation compared to cultures from their respective wild type controls [TNF-α KO, t=5.031, **P<0.001; TNF R1/R2 KO, t=31.96, ***P<0.0001, post hoc analysis by Bonferroni t-test compared to wild type treated with LPS for 3 hour cultures]. Due to the great difference of GDNF mRNA level between 0 and 3 hours after LPS stimulation, we set maximum value as 100%. Data were presented as percent of GDNF mRNA expression with respect to the expression of the wild type cultures treated with LPS for 3 hours. Data shows mean ± SEM from three independent experiments, each done in duplicate.

Figure 7

LPS was unable to elicit the release of TNF-α or NO from either highly-enriched astroglial or astrocytoma cultures. Highly-enriched astroglia (LME-treated method) and C6 astrocytoma cultures treated with 100 ng/ml of LPS secreted significantly negligible levels of NO (A) and TNF-α (B) compared to mixed glia or enriched astroglia (shake-off method) cultures (** P < 0.001, *** P < 0.0001, post hoc analysis by Bonferroni t-test). A similar trend for TNF-α (C) and iNOS (D) mRNA expression was observed in enriched microglia, mixed glia and highly-enriched astroglial cultures (* P < 0.05, *** P < 0.0001, post hoc analysis by Bonferroni t-test compared to highly-enriched astroglial cultures). Due to the great difference of TNF and iNOS mRNA levels among these cultures, we set maximum value as 100%. Data were normalized with respect to the maximum value of microglia enriched culture. Data show mean ± SEM from three independent experiments, each with triplicates.

Figure 8

Microglia are the major source of LPS-elicited TNF-α and NO in enriched ‘astroglial’ cultures. Reconstituted highly-enriched astroglial cultures with 0.5 to 20% microglia were stimulated with LPS (100 ng/ml) and secreted TNF-α (A) and NO (B) levels were determined 3 and 24 hours after LPS treatment, respectively. Data were presented as mean ± SEM from three independent experiments, each done in triplicates. *** P < 0.0001, one-way ANOVA followed by Bonferroni post hoc multiple comparison test.

Figure 9

Activated microglia are required for the expression of proinflammatory factors during LPS-induced astrogliosis. Highly-enriched astroglial (A) and mixed glial (B) cultures were treated with 100 ng/ml of LPS for 72 hours. Cultures were immunostained with antibodies against Iba-1 (green, for microglia), iNOS (red) and DAPI (blue) in upper panel; or GFAP (green, for astroglia), iNOS (red) and DAPI (blue) in middle panel; or S100b (green, for astroglia), iNOS (red) and DAPI (blue) in lower panel. Scale bar: 50 μm. Mixed glia and highly-enriched astroglial cultures were treated with vehicle, LPS, microglia conditioned medium which was collected from enriched microglia cultures at 24 hours after LPS treatment, and control conditioned medium (ctrl). After 3 and 6 hours, cultures were subjected to real-time PCR assay for (C) TNF-α and (D) iNOS mRNA expressions. Value of mixed glia cultures-treated with LPS was set as 100%. Data show mean ± SEM from three independent experiments, each done in duplicate. ***P<0.001, **P<0.01, post hoc analysis by Bonferroni t-test compared to vehicle treated highly-enriched astroglial cultures.

Figure 9

Activated microglia are required for the expression of proinflammatory factors during LPS-induced astrogliosis. Highly-enriched astroglial (A) and mixed glial (B) cultures were treated with 100 ng/ml of LPS for 72 hours. Cultures were immunostained with antibodies against Iba-1 (green, for microglia), iNOS (red) and DAPI (blue) in upper panel; or GFAP (green, for astroglia), iNOS (red) and DAPI (blue) in middle panel; or S100b (green, for astroglia), iNOS (red) and DAPI (blue) in lower panel. Scale bar: 50 μm. Mixed glia and highly-enriched astroglial cultures were treated with vehicle, LPS, microglia conditioned medium which was collected from enriched microglia cultures at 24 hours after LPS treatment, and control conditioned medium (ctrl). After 3 and 6 hours, cultures were subjected to real-time PCR assay for (C) TNF-α and (D) iNOS mRNA expressions. Value of mixed glia cultures-treated with LPS was set as 100%. Data show mean ± SEM from three independent experiments, each done in duplicate. ***P<0.001, **P<0.01, post hoc analysis by Bonferroni t-test compared to vehicle treated highly-enriched astroglial cultures.