Astrocyte-derived MCP-1 mediates neuroprotective effects of noradrenaline

Abstract

The neurotransmitter noradrenaline (NA) can provide neuroprotection against insults including inflammatory stimuli and excitotoxicity, which may involve paracrine effects of neighboring glial cells. Astrocytes express and secrete a variety of inflammatory and anti-inflammatory molecules; however, the effects of NA on astrocyte chemokine expression have not been well characterized. In primary astrocytes, NA increased expression of chemokine CCL2 (MCP-1) at the mRNA and protein levels. NA increased activation of an MCP-1 promoter driving luciferase expression, which was replicated by beta-adrenergic receptor agonists and a cAMP analog, and blocked by a specific beta2-adrenergic receptor antagonist. In primary neurons, addition of MCP-1 reduced NMDA-dependent glutamate release as well as glutamate-dependent Ca(2+) entry. Similarly, conditioned media from NA-treated astrocytes reduced glutamate release, an effect that was blocked by neutralizing antibody to MCP-1, whereas MCP-1 dose-dependently reduced neuronal damage attributable to NMDA or to glutamate. MCP-1 significantly reduced lactate dehydrogenase release from neurons after oxygen-glucose deprivation (OGD) and prevented the loss of ATP levels that occurred after OGD or treatment with glutamate. Incubation of neurons with astrocytes separated by a membrane to prevent physical contact showed that NA induced astrocyte release of sufficient MCP-1 to reduce neuronal damage attributable to OGD. These findings indicate that the neuroprotective effects of NA are mediated, at least in part, by induction and release of astrocyte MCP-1.

Figure 1.

Noradrenaline increases MCP-1 in astrocytes. A, Enriched cultures of astrocytes (Ast) or microglia (μGlia) were incubated with 0 or 10 μm NA for 24 h, then MCP-1 mRNA levels measured by QPCR. Data are mean ± SE of MCP-1 mRNA levels normalized to values for β-actin and are expressed relative to control (100%) values. **p < 0.001, ***p < 0.0001 versus control; n = 8 replicates per group. B, Astrocytes were incubated with 0–25 μm NA, and MCP-1 levels in the media were assessed after 24 h by ELISA. Data are means ± SE of n = 8 replicates per group. *p < 0.05, **p < 0.001, ***p < 0.0005 versus control. C, Astrocytes were transfected with an pMCP-1 luc reporter (0.2 μg) and Renilla control phRG-TK vector. After 24 h, cells were treated with the indicated concentrations of NA, dbcAMP, or the β-AR agonist Isoproterenol, or (D) 10 μm NA together with varying concentration of the β-AR agonist propranolol or the β2-AR agonist ICI118551. MCP-1 promoter activity was measured after 4 h, and differences in transfection efficiencies were corrected for by normalization to renilla luciferase activity. The data are mean ± SE of n = 3–4 replicates and are relative to the MCP-1 activity measured in nontreated cells. *p < 0.05, **p < 0.005 versus control values.

Figure 2.

MCP-1 decreases extracellular glutamate. A, Primary neurons were incubated with NMDA (20 μm) in the presence or absence of MCP-1 (10 ng/ml), and after 48 h, the extracellular concentration of glutamate was determined. The data are mean ± SE of n = 10 replicates and is shown relative to values measured in control (Ctrl) cells (100% = 40.5 ± 0.7 μm) ^§p < 0.05 versus control, *p < 0.05 versus NMDA. B, Primary neurons were incubated with NMDA (20 μm) in the presence of conditioned media from astrocytes treated without (CM-C) or with (CM-NA) NA (75 μm) for 24 h and in the absence or presence of a neutralizing antibody against MCP-1 (10 μg/ml). After 48 h, the extracellular concentration of glutamate was determined. The data are mean ± SE of n = 10 replicates and is shown relative to values measured in CM-C cells (100% = 40.5 ± 0.7 μm) ^§§§p < 0.0005 versus CM-C, *p < 0.05 versus CM-NA.

Figure 3.

MCP-1 reduces excitotoxicity. A, Primary neurons were treated for 24 h with the indicated concentrations of MCP-1, then preloaded with Fluo-4 A.M. and incubated with glutamate (200 μm). As a control neurons were incubated with glutamate together with the NMDA-R antagonist MK801 (“MK,” 10 μm). The fluorescence signal was measured after 3 min. The data are mean ± SE of n = 12 replicates and is Ca²⁺ entry relative to that measured in the presence of glutamate alone. *p < 0.05, **p < 0.005, ***p < 0.0001 versus glutamate alone. B, Primary neurons were incubated with 0 (Ctrl) or 200 μm glutamate and the indicated concentrations of MCP-1, and the cellular ATP levels were measured 24 h later. Data are mean ± SE of n = 12 replicates and is the ATP level relative to control cells. ^§§§p < 0.0001 versus control; *p < 0.05, **p < 0.005, ***p < 0.0001 versus no MCP-1. C, Primary neurons were incubated with NMDA (20 μm) or glutamate (100 μm) in the presence of the indicated concentrations of MCP-1. After 48 h, neuronal viability was assessed by measurement of LDH in the media. Data are mean ± SE of n = 8 replicates and is the LDH released compared with control (CTL; nontreated cells). ^§p < 0.05 versus control; ^§§§p < 0.0005 vs control; **p < 0.005 versus NMDA alone or versus glutamate alone. In control cells, 100% LDH reflected 20 ± 1.4% of total LDH released after 48 h (mean ± SE of n = 15 individual measurements).

Figure 4.

MCP-1 reduces ischemic damage. A, Primary neurons were preincubated with the indicated concentrations of MCP-1 for 24 h, then exposed to OGD or kept under normoxic conditions [control (Ctrl); open bar], after which the media was replaced with fresh media (containing the same amounts of MCP-1). Neuronal viability was assessed 48 h later by measurement of LDH release. Data are mean ± SE of n = 12 replicates and shown relative to LDH release attributable to OGD alone. ^§§§p < 0.0005 versus control cells; **p < 0.005, ***p < 0.0001 versus OGD. B, Primary astrocytes growing on transwell membranes were transferred to wells containing primary neurons. After 24 h, the cocultures were treated for a further 24 h with fresh media (none), 10 μm NA (NA), or 10 μm NA and 10 μg/ml of MCP-1 neutralizing antibody (NA+αMCP1). After treatment, the astrocytes were removed, the media replaced, the neurons were exposed OGD, and neuronal viability assessed after 24 h. Control neurons (Ctrl; open bar) were kept normoxic. The data are the mean ± SE on n = 8 replicates and is LDH release relative to OGD treated only cells. ***p < 0.0005 versus control cells; *p < 0.05 versus none; ^§p < 0.05 versus NA. In control cells, 100% LDH reflected 20 ± 1.4% of total LDH released after 48 h (mean ± SE of n = 15 individual measurements).