Macrophages treated with particulate matter PM2.5 induce selective neurotoxicity through glutaminase-mediated glutamate generation

Abstract

Exposure to atmospheric particulate matter PM2.5 (aerodynamic diameter ≤ 2.5 μm) has been epidemiologically associated with respiratory illnesses. However, recent data have suggested that PM2.5 is able to infiltrate into circulation and elicit a systemic inflammatory response. Potential adverse effects of air pollutants to the central nervous system (CNS) have raised concerns, but whether PM2.5 causes neurotoxicity remains unclear. In this study, we have demonstrated that PM2.5 impairs the tight junction of endothelial cells and increases permeability and monocyte transmigration across endothelial monolayer in vitro, indicating that PM2.5 is able to disrupt blood-brain barrier integrity and gain access to the CNS. Exposure of primary neuronal cultures to PM2.5 resulted in decrease in cell viability and loss of neuronal antigens. Furthermore, supernatants collected from PM2.5 -treated macrophages and microglia were also neurotoxic. These macrophages and microglia significantly increased extracellular levels of glutamate following PM2.5 exposure, which were negatively correlated with neuronal viability. Pre-treatment with NMDA receptor antagonist MK801 alleviated neuron loss, suggesting that PM2.5 neurotoxicity is mediated by glutamate. To determine the potential source of excess glutamate production, we investigated glutaminase, the main enzyme for glutamate generation. Glutaminase was reduced in PM2.5 -treated macrophages and increased in extracellular vesicles, suggesting that PM2.5 induces glutaminase release through extracellular vesicles. In conclusion, these findings indicate PM2.5 as a potential neurotoxic factor, crucial to understanding the effects of air pollution on the CNS.

Keywords: PM 2.5; blood-brain barrier; glutamate; glutaminase; macrophage; neurotoxicity.

Fig. 1. PM_2.5 reduces the levels of TJ proteins in endothelial monolayer

A–B: The levels of TJ proteins ZO-1 and ZO-2 in HUVEC (A) and HBMEC (B) in both PM_2.5-treated groups and mock-treated control group were determined by western blot. β-actin was used as the loading control. C–F: Protein levels of ZO-1 and ZO-2 were normalized as a ratio to β-actin after densitometric quantification and presented as fold change relative to control. Results are representative of three independent experiments. *p < 0.05, **p < 0.01 compared with the untreated control group.

Fig. 2. PM_2.5 increases permeability and monocytes transmigration cross endothelial monolayer

A: Scheme of BBB permeability. B: HUVEC monolayer permeability after treatment with PM_2.5 (25 μg/ml) for 24 h. The monolayer-free upper chamber group was set as the positive control (PC). C: Scheme of monocyte migration. Cell types and their placement in the transwell system were indicated. D: Monocytes migrated through PM_2.5-treated HUVEC monolayers in the transwells were determined by immunostaining with DAPI. E: Quantification of migrated monocytes was performed by counting the number of DAPI-positive nuclei in each microscope field (n = 10). Scale bar, 50 μm. *p < 0.05, ***p < 0.001 compared with the untreated control group.

Fig. 3. PM_2.5 induces direct neurotoxicity

A–B: Cell viability of human neurons (A) and RCN (B) after treatment with PM_2.5 for 24 h. C–D: Caspase3 cleavage in human neurons (C) and RCN (D) after treatment with PM_2.5 for 24 h were determined by western blot. β-actin was used as loading control. E: Representative MAP2 staining of primary RCN after treatment with PM_2.5 or staurosporine for 24 h. Staurosporine (STS, 1 μM) was used as a positive control. F: Quantification of MAP2 fluorescence in panel E. Scale bar, 50 μm. Results are representative of three independent experiments. *p < 0.05, **p < 0.01 compared with the untreated control group.

Fig. 4. PM_2.5 increases macrophage-mediated neurotoxicity

A: Enrichment of macrophage cultures was evaluated by CD68 immunostaining. B: Cell viability of macrophages after PM_2.5 exposure for 24 h. C: Morphology of macrophages (CD68-staining pictures) and microglia (bright-field pictures) after PM_2.5 exposure. Typical rounded macrophages and microglia after PM_2.5 exposure were marked with red arrows. D: Representative MAP2 staining of RCN after treatment with MCM. E: Quantification of MAP2 fluorescence in panel D. Scale bar, 50 μm. F–G: Caspase 3 cleavage in RCN after treatment with MCM (F) or MGCM (G) was determined by western blot. β-actin was used as loading control. H–I: Cell viability of RCN after treatment with MCM (H) or MGCM (I). Results are representative of three independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001 compared with the control MCM-treated group.

Fig. 5. Neurotoxicity of PM_2.5-treated macrophages is dependent on glutamate production

A–B: HPLC determination of glutamate production in macrophages (A) and microglia (B) supernatants after PM_2.5 treatment. C–D: Correlation of neuronal viability with extracellular glutamate levels. E: Representative MAP2 staining of RCN after treatment with MCM. RCN were pre-treated with 3 μM MK801 (NMDA receptor antagonist) for 2 h before treatment with MCM. Scale bar, 50 μm. F: Quantification of MAP2 fluorescence in panel E. Results are representative of three independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001 compared with the control MCM-treated group.

Fig. 6. PM_2.5 induces glutaminase release in the extracellular vesicles

A: HPLC determination of glutamate production in macrophage supernatants. Bis-2-(5-phenylacetamido-1,2,4-thiadiazol-2-yl)ethyl sulfide (BPTES, 1 μM) was added to the culture along with PM_2.5. B: Protein levels of GAC and KGA in macrophages after exposure to PM_2.5 were determined through western blot. C-D: Quantification of GAC and KGA in panel B. E: Extracelluler glutaminase activity. Glutamate levels in supernatants were determined through HPLC after incubation with (black) or without (white) glutamine for 48 h at 37°C. F: Extracelluler glutaminase level. KGA and GAC protein levels in both whole cell lysates and extracellular vesicles were determined by western blot. Results are representative of three independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001 compared with the untreated control group.