NMDA receptor blockade attenuates Japanese encephalitis virus infection-induced microglia activation

Abstract

Neurodegeneration and neuroinflammation are key components in the pathogenesis of Japanese Encephalitis caused by Japanese Encephalitis Virus (JEV) infection. The N-methyl-D-aspartate (NMDA)-type glutamate receptor displays excitatory neurotoxic and pro-inflammatory properties in a cell context-dependent manner. Herein, potential roles of the NMDA receptor in excitatory neurotoxicity and neuroinflammation and effects of NMDA receptor blockade against JEV pathogenesis were investigated in rat microglia, neuron/glia, neuron cultures, and C57BL/6 mice. In microglia, JEV infection induced glutamate release and activated post-receptor NMDA signaling, leading to activation of Ca²⁺ mobilization and Calcium/Calmodulin-dependent Protein Kinase II (CaMKII), accompanied by pro-inflammatory NF-κB and AP-1 activation and cytokine expression. Additionally, increased Dynamin-Related Protein-1 protein phosphorylation, NAPDH Oxidase-2/4 expression, free radical generation, and Endoplasmic Reticulum stress paralleled with the reactive changes of microglia after JEV infection. JEV infection-induced biochemical and molecular changes contributed to microglia reactivity and pro-inflammatory cytokine expression. NMDA receptor antagonists MK801 and memantine alleviated intracellular signaling and pro-inflammatory cytokine expression in JEV-infected microglia. JEV infection induced neuronal cell death in neuron/glia culture associated with the concurrent production of pro-inflammatory cytokines. Conditioned media of JEV-infected microglia compromised neuron viability in neuron culture. JEV infection-associated neuronal cell death was alleviated by MK801 and memantine. Activation of NMDA receptor-related inflammatory changes, microglia activation, and neurodegeneration as well as reversal effects of memantine were revealed in the brains of JEV-infected mice. The current findings highlight a crucial role of the glutamate/NMDA receptor axis in linking excitotoxicity and neuroinflammation during the course of JEV pathogenesis, and proposes the anti-inflammatory and neuroprotective potential of NMDA receptor blockade.

Introduction

Japanese EncephalitisVirus (JEV) is a mosquito-borne neurotropic virus belonging to the Flaviviridae family. Its family members include the Zika Virus, West Nile Virus, Hepatitis C Virus, St. Louis Encephalitis Virus, and Yellow Fever Virus [1]. Its trafficking from the periphery to the Central Nervous System (CNS) and viral amplification in the brain parenchyma both play determinant roles in JEV pathogenesis involving stepwise communication between the host and viral factors. JEV illness is called Japanese Encephalitis (JE), with the associated pathological hallmark characterized by neuronal cell death and inflammation which have close links to JEV pathogenesis. Along with experiencing fatal neuroinfection, survivors of clinical JE cases may suffer from persistent neurological sequelae [2,3,4]. Appropriate immune responses help the host defense viral infection and promote neuronal survival, while aberrant immune responses exacerbate neuronal deterioration. Sustained microglia activation and inflammatory cytokine over-production are deteriorating components of JEV pathogenesis. Interventions aimed at suppressing aberrant immune responses by amelioration of microglia activation and/or inflammatory cytokine over-production are found to have beneficial outcomes in neuronal survival and neurobehavioral functions [5,6,7]. Therefore, inflammation and neuronal cell death are key components to the development of JE and microglia represent targets crucial to therapeutic treatment against JE and its associated complications.

Neurons are the major target of JEV infection, with neuronal cell death being a key component of JEV infection-induced neurodegeneration and neurobehavioral abnormality. Several types of programmed cell death, including apoptosis, autophagy, and necroptosis are found in neurons upon JEV infection [8,9,10]. Along with direct neurotoxicity, bystander damage represents an alternative mechanism attributing to JEV infection-induced neuronal cell death. Severe JE-associated neuronal cell death is accompanied by immune cell accumulation and activation, as well as inflammatory cytokine over-production. Specifically, in vitro cell study demonstrates that JEV-infected microglia release neurotoxic molecules and induce neuronal cell death [7, 9, 11]. These findings indicate that JEV infection causes neuronal cell death, with the compromise of neurons possibly coming from consequences of direct neurotoxicity and/or secondary JEV-infected microglia.

Glutamate is a physiological excitatory neurotransmitter in the nervous system. When seen in excess, glutamate turns into neurotoxic molecules and causes neuronal cell death mainly through the N-methyl-D-aspartate (NMDA) receptor [12]. Glutamate neurotoxicity has been implicated not only in a plethora of neurological disorders but also in virus infection-associated neuronal injury [13, 14]. Beyond being a neurotoxic molecule, glutamate displays additional pharmacological and biochemical functions through its communication with inflammatory responses. Inflammatory stimuli and pro-inflammatory cytokines cause elevation in intracellular glutamate synthesis and extracellular glutamate release [15, 16]. Alternatively, glutamate provides dual effects on pro-inflammatory cytokine expression through its actions on ionotropic and metabotropic glutamate receptors [17,18,19]. These phenomena highlight the role of the interplay between neuronal cell death and inflammation, while underscoring their importance as preventive and therapeutic targets for the control of neurological disorders, including viral encephalitis. Critical to this interplay is that glutamate could act as a key integrator in linking neuronal cell death and inflammation.

Glutamate neurotoxicity has been demonstrated in rodent models of JE. Glutamate levels and NMDA receptor activity were both elevated in the brains of JEV-infected mice and rats. NMDA receptor antagonists, MK801 and memantine, alleviated neuronal cell death and inflammatory responses triggered by JEV infection, while having little effect on viral amplification [4, 20, 21]. Using rat brain primary cultures, our previous study revealed the neuroprotective effects of MK801 against JEV infection-induced neurotoxicity, while also demonstrating the active role microglia play in JEV infection-accompanied glutamate release, pro-inflammatory cytokine expression, and neuronal cell death [15]. Although whether microglia possess NMDA receptor-dependent currents in vivo is of controversy [22], microglia express functional glutamate NMDA receptors and the NMDA receptor blockade is reported to have inhibitory effects on microglia activation and pro-inflammatory cytokine expression [23,24,25]. Along with glutamate, endogenous mimicking ligands and the Toll-Like Receptor 4 (TLR4) have been linked to NMDA receptor signaling leading to microglia activation [26]. JEV infection-induced neuronal cell death and inflammation correlated well with activation of the NMDA receptor axis [4, 20, 21], the specific involvement of the NMDA receptor signaling in JEV infection-induced inflammation has not yet been completely investigated. Microglia are highly dynamic and plastic cells that display multivariate morphological, genetic, metabolic, and functional states in both normal physiology and disease pathophysiology [9, 11, 17]. To extend the scope of JEV pathogenesis study centered on inflammation, this present study investigated the alteration of glutamate/NMDA receptor signaling in infected cells, while also attempting to elucidate change in microglia states reactive to JEV infection by NMDA receptor blockade via microglia culture and infected mice.

Materials and methods

Cell cultures

For preparation of the primary cultures, cerebral cortices were collected from postnatal Day 1 Sprague-Dawley rats, which then proceeded to tissue trituration and cell isolation according to our previously reported methods [11]. The experimental protocols surrounding animal studies were reviewed and approved by The Animal Experimental Committee of Taichung Veterans General Hospital (IACUC approval code: La-1061509). The dissociated cells were then transferred to poly-D-lysine-coated (30 µg/mL) dishes/plates for cultivation. Neuron/glia cultures consisting of 30–40% neurons, 40–50% astrocytes, and 10–15% microglia were prepared by maintaining the cells in a Minimum Essential Medium, supplemented with 10% Fetal Bovine Serum (FBS) and 10% Horse Serum for 10–12 days. Neurobasal Medium supplemented with B27 and cytosine arabinoside (10 µM) was used to culture the neurons. Neuron cultures (more than 95% purity) were ready for use in the experiments after 10–12 days in vitro. Cultivation of the cells for 2 weeks was performed in Dulbecco’s Modified Eagle Medium (DMEM)/F12 containing 10% FBS enriched astrocytes and microglia. Microglia (more than 90% purity) were separated from astrocytes by shaking and centrifugation. The detached microglia were seeded onto poly-D-lysine-coated (30 µg/mL) dishes/plates and then maintained in DMEM/F12 containing 10% FBS until analyses could be performed. One batch of cell preparation was referred to one biological replicate. Usually, three to four biological replicates were included for each experiment.

Virus

The propagation of JEV strain NT113 was performed in C6/36 cells (BCRC-60114, Bioresource Collection and Research Center, Hsinchu, Taiwan), with viral titers counted using Baby Hamster Kidney cells (BHK21, BCRC-60041, Bioresource Collection and Research Center, Hsinchu, Taiwan) [11]. A mock virus was prepared from C6/36 cells without JEV infection, in parallel with propagation of JEV virus stock. For virus inactivation, virus stocks were manipulated with UV exposure (254 nm exposure for 30 min, JEV/UV-inactivated) or boiling (94 °C incubation for 15 min, JEV/Heat-inactivated). To infect the primary cultures, cells were inoculated with JEV stock or manipulated JEV at a Multiplicity of Infection (MOI) 5 for 1 h at 37 °C. After gentle washing with Phosphate-Buffered Saline (PBS) to remove the unbound viruses, a fresh DMEM/F12 containing 2% FBS was added for further incubation. For treatment, pharmacological agents were added with the refreshing medium.

Cytotoxicity assessment

Cytotoxicity was assessed using the Pierce™ Lactate Dehydrogenase (LDH) Cytotoxicity Assay Kit (Thermo Fisher Scientific, Waltham, MA, USA), according to the manufacturer’s instructions.

Immunocytochemical staining

Cells were sequentially treated with 4% paraformaldehyde, 0.1% Triton X-100, and 5% nonfat milk for fixation, permeabilization, and blocking, respectively. To visualize the neurons, astrocytes and microglia, cells were incubated with each corresponding recognizing antibody, Microtubule-Associated Protein 2 (MAP-2, 1:500, sc-390543), Glial Fibrillary Acidic Protein (GFAP, 1:500, sc-33673) (Santa Cruz Biotechnology, Santa Cruz, CA, USA), and Cluster of Differentiation 68 (CD68, 1:500, ab125212) (Abcam, Cambridge, UK), followed by diaminobenzidine color development. Under a light microscope, viable and MAP-2 immunoreactive neurons were defined by a visible cell body having a length of neurite process greater than two cell bodies. For quantitation, the numbers of positive cells were counted in three randomly selected optical fields (100X magnification) in the well of a 24-well plate, with two wells being included. Data from three different batches of cell preparation were combined for quantitative analyses.

Enzyme-linked immunosorbent assay (ELISA)

The levels of Nitric Oxide (NO, nitrite/nitrate), Tumor Necrosis Factor-α (TNF-α), and Interleukin-1β (IL-1β) in the cell supernatants were measured using a Griess Reagent Kit (Thermo Fisher Scientific, Waltham, MA, USA) and the corresponding ELISA kits (R&D Systems, Minneapolis, MN, USA).

Measurement of reactive oxygen species (ROS)

To measure intracellular ROS, cells were incubated with the cell permeable fluorogenic dye 2’,7’-Dichlorofluorescein Diacetate (DCFDA) (Thermo Fisher Scientific, Waltham, MA, USA). The fluorescence signals of reacted DCFDA were measured using a fluorometer with excitation/emission at 485 nm/535 nm.

RNA isolation and quantitative real-time reverse transcriptase polymerase chain reaction (RT-PCR)

Cellular RNA extraction andcomplementary DNA synthesis were completed using conventional procedures. Quantitative real-time PCR was performed on ABI StepOne™ (Applied Biosystems, Foster City, CA, USA), withthe data calculated based on the ΔΔCT method according to our previously reported methods [11]. The oligonucleotides used for amplification were: iNOS, 5’-ACAACGTGGAGAAAACCCCAGGTG and 5’-ACAGCTCCGGGCATCGAAGACC; TNF-α, 5’-CCCTCACACTCAGATCATCTTCTCAA and 5’-TCTAAGTACTTGGGCAGGTTGACCTC; IL-1β, 5’-CACCTCTCAAGCAGAGCACAG and 5’-GGGTTCCATGGTGAAGTCAAC; JEV, 5’-AGAGCACCAAGGGAATGAAATAGTand 5’-AATAGGTTGTAGTTGGGCACTCTG; and β-actin, 5’-AAGTCCCTCACCCTCCCAAAAG and 5’-AAGCAATGCTGTCACCTTCCC.

Measurement of glutamate

The levels of glutamate in the cell supernatants were measured using the High Performance Liquid Chromatography (HPLC) technique, with chromatographic data processed using HP Chem Station (Hewlett Packard, Palo Alto, CA, USA), according to our previously reported methods [15].

Measurement of cytosolic Ca2+ levels

The levels of cytosolic Ca²⁺ were measured using Fluo-4 AM according to our previously reported methods with modifications [15]. The fluorescence signals were measured in a fluorometer with excitation wavelength at 485 nm and the emission wavelength at 510 nm.

Western blot analysis

Conventional protein extraction, separation, and SDS-PAGE were performed according to our previously reported methods [11]. The targeted proteins in the membranes were recognized by sequential incubation with primary antibodies, horseradish peroxidase-labeled IgG and enhanced chemiluminescence Western blotting reagents. The chemiluminescence on the membranes were visualized using the G: BOX mini Multi fluorescence and chemiluminescence imaging system (Syngene, Frederick, MD, USA), and then quantified by Image J software (National Institute of Health, Bethesda, MD, USA). Targets of the used antibodies included JEV nonstructural protein 1 (NS1, 1:2000, ab41651), CD68 (1:1000, ab125212), phospho-IRE1 (Serine-724, 1:500, ab48187) (Abcam, Cambridge, UK), c-Jun (1:1000, sc-74543), MAP-2 (1:1000, sc-390543), GFAP (1:1500, sc-33673), Calcium/Calmodulin-dependent Protein Kinase II (CaMKII, 1:1000, sc-5306), NF-κB p65 (1:1000, sc-514451), phospho-NF-κB p65 (Serine-536, 1:500, sc-136548), c-Fos (1:1000, sc-8047), Extracellular Signal-Regulated Kinase (ERK, 1:1000, sc514302), c-Jun N-terminal Kinase (JNK, 1:1000, sc-7345), phospho-p38 (Tyrosine-182, 1:500, sc-166182), Interferon Regulatory Factor 5 (IRF5, 1:1000, sc-390364), P2X purinoceptor 7 (P2 × 7R, 1:1000, sc-514962), Dynamin-Related Protein 1 (Drp1, 1:1000, sc-271583), NADPH Oxidase 2 (NOX2, 1:1000, sc-130543), Protein Kinase RNA-like Endoplasmic Reticulum Kinase (PERK, 1:1000, sc-377400), eIF2α (1:1000, sc-133132), IRE1 (1:1000, sc-390960), Glyceraldehyde 3-Phosphate Dehydrogenase (GAPDH, 1:3000, sc-32233) (Santa Cruz Biotechnology, Santa Cruz, CA, USA), Transforming Growth Factor β-Activated Kinase-1 (TAK1, 1:1000, 12330-2-AP), NOX4 (1:1000, 14347-1-AP) (Proteintech, Rosemont, IL, USA), phospho-c-Jun (Serine-63, 1:500, PA5-17890), phospho-TAK1 (Serine-192, 1:500, BS-5435R), phospho-Drp1 (Serine-616, 1:500, PA5-64821), phospho-PERK (Threonine-982, 1:500, PA5-40294) (ThermoFisher Scientific, Waltham, MA, USA), NMDA GluN1 receptor (1:1000, #5704), phospho-GluN1 (Serine-896, 1:500, #3384), NMDA GluN2A receptor (1:1000, #4205), phospho-GluN2A (Tyrosine-1246, 1:500, #4206), NMDA GluN2B receptor (1:1000, #4207), phospho-GluN2B (Serine-1303, 1:500, #71335), phospho-eIF2α (Serine-51, 1:500, #3597), phospho-CaMKII (Threonine-286, 1:500, #12716), phospho-ERK (Threonine-202/Tyrosine-204, 1:500, #9101), phospho-JNK (Threonine-183/Tyrosine-185, 1:500, #9255), p38(1:1000, #9212) (Cell Signaling Technology, Danvers, MA, USA).

Preparation of nuclear extracts and Electrophoretic mobility Shift Assay (EMSA)

An NE-PER Nuclear and Cytoplasmic Extraction Kit (ThermoFisher Scientific, Waltham, MA, USA) and an EMSA kit (LightShift™ Chemiluminescent EMSA Kit, ThermoFisher Scientific, Waltham, MA, USA) were used to extract nuclear proteins and evaluate transcription factor DNA binding activity, respectively, according to the manufacturer’s instructions. As with our previous studies [11], nuclear extracts (5 µg) were reacted with either AP-1 oligonucleotide (5’-CGCTTGATGAGTCAGCCGGAA) or NF-κB oligonucleotide (5’-AGTTGAGGGGACTTTCCCAGGC), with the complexes visualized using enhanced chemiluminescence Western blotting reagents. The chemiluminescence on the membranes were visualized using a G: BOX mini Multi fluorescence and chemiluminescence imaging system (Syngene, Frederick, MD, USA), and quantified by Image J software (National Institute of Health, Bethesda, MD, USA).

Caspase 3 activity assay

Proteins were extracted from neuron/glia cultures and caspase 3 activity was measured using a commercially available kit (Fluorometric Assay Kit, BioVision, Mountain View, CA, USA), according to the manufacturer’s instructions.

Animals

Adult male C57BL/6 mice (8 weeks old) were purchased from BioLASCO (Taipei, Taiwan). The experimental protocols surrounding animal studies were reviewed and approved by The Animal Experimental Committee of Taichung Veterans General Hospital (IACUC approval code: La-1061509). Eighteen mice were divided into three groups receiving either PBS (n = 6) or JEV (2 × 10³ PFU in 200 µL PBS, n = 12) through intraperitoneal injection according to our previously reported methods [27]. Memantine (20 mg/kg) was delivered into JEV group (n = 6) through drinking water starting from 1 day before JEV infection and lasting for 8 days according to related study with slight modifications [4]. On day 7 post-viral infection, mice were euthanized and the brain tissueswere collected for further studies.

Immunofluorescence staining

Mice (n = 3) were anesthetized and transcardially perfused with ice-cold saline followed by 4% paraformaldehyde. The brains were removed and post-fixed with 4% paraformaldehyde, cryoprotected in sucrose, and frozen. Immunoflurescence staining was performed according to previouslyreported methods with modifications [27]. Frozen sections were incubated with MAP-2 (1:200, sc-390543), GFAP (1:200, sc-33673) (Santa Cruz Biotechnology, Santa Cruz, CA, USA), Iba1 (1:200, #17198), and phospho-GluN2A (Tyrosine-1246) (1:200, #4206) (Cell Signaling Technology, Danvers, MA, USA) antibody andsequentially with rhodamine- or fluorescein isothiocyanate(FITC)-conjugated secondary antibody. The cell nuclei were counterstained with 4’,6-diamidino-2-phenylindole (DAPI). The immunoreactive signals were observed under an epifluorescence microscope.

Statistical analysis

Data are presented as mean value ± standard deviationfrom the three to four independent experiments.Variances between more than two experimental groups and control group were analyzed by one-way Analysis of Variance (ANOVA) and assessed using either the Bonferroni or Tukey post hoc test. Student’s t-test was used if two groups were compared. A level of p < 0.05 was considered statistically significant.

Results

JEV infection activated NMDA receptor signaling in microglia

To demonstrate the effects of JEV infection on glutamate biological activity, alterations in NMDA receptor and post-receptor signaling molecules were investigated in cultured microglia. When cultured in poly-D-lysine-coated plates, microglia exhibited a process-bearing morphology. JEV infection transformed the microglia from process-bearing to cytoplasm extension with a rounded, darkly stained morphology (Fig. 1A). Upon wild type JEV infection, microglia continuously synthesized and released glutamate to the extracellular space (Fig. 1B). At the same time, elevation of intracellular free labile Ca²⁺ was found (Fig. 1C). Our previous study showed that wild type JEV is critical to infection-induced biological responses when compared to the corresponding mock infection, UV-inactivated JEV, and Heat-inactivated JEV [15]. In consisting with previous findings, we further found that wild type, not Heat-inactivated and UV-inactivated JEV infection, caused glutamate efflux (Fig. 1D) and Ca²⁺ elevation (Fig. 1E) in microglia. Elevation of intracellular free labile Ca²⁺ caused by JEV infection was alleviated by non-competitive NMDA receptor antagonists, MK801 and memantine (Fig. 1F). Changes in extracellular glutamate and intracellular free labile Ca²⁺ paralleled with increased protein phosphorylation in NMDA receptor GluN1, GluN2A, and GluN2B subunit components, as well as Ca²⁺-activated CaMKII (Fig. 1G). MK801 and memantine decreased JEV infection-increased protein phosphorylation (Fig. 1G). These findings indicate that functional microglial glutamate/NMDA receptor signaling could be activated by JEV infection. For further investigation, MK801 and memantine were applied at concentrations of 10 µM and 5 µM, respectively.

Fig. 1

JEV infection activated NMDA receptor signaling in microglia. (A) Cultured microglia were infected with mock or JEV for 24 h. Microglia were examined by immunocytochemical staining with antibody recognizing CD68. Representative photomicrograph is shown. Bar, 50 μm. Microglia cultures were infected with mock or JEV over time. The supernatants were collected and subjected to HPLC for the measurement of glutamate concentration at the indicated times (B). The changes in cytosolic Ca²⁺ concentration were determined by Fluo-4 AM fluorescence (arbitrary unit) measurements at the indicated times (C). Microglia cultures were infected with mock, JEV, JEV/UV-inactivated, or JEV/Heat-inactivatedfor 12 h. The supernatants were collected and subjected to HPLC for the measurement of glutamate concentration (D). The changes in cytosolic Ca²⁺ concentration were determined by Fluo-4 AM fluorescence (arbitrary unit) measurements (E).(F) Microglia cultures were infected with mock or JEV in the absence or presence of various concentrations of MK801 (0–40 µM) or memantine (0–10 µM) for 12 h.The changes in cytosolic Ca²⁺ concentration were determined by Fluo-4 AM fluorescence (arbitrary unit) measurements. (G) Microglia cultures were infected with mock or JEV in the absence or presence of MK801 (10 µM) or memantine (5 µM) for 8 h. Total cellular proteins were extracted and subjected to Western blot analysis with the indicated antibodies. One representative blot of three independent culture batches and the quantitative results of relative protein contents (arbitrary unit) are shown. *p < 0.05 vs. Mock control and #p < 0.05 vs. JEV control, n = 4

NMDA receptor antagonists alleviated JEV infection-induced cytokine expression and glutamate release in microglia

It has been reported that NMDA receptors transduce Lipopolysaccharide (LPS) signals to trigger cytokine expression in microglia [25, 26, 28]. Whether the NMDA receptors play the same role to deliver pro-inflammatory signaling in JEV-infected microglia, was determined by the expression levels of pro-inflammatory cytokines. There were elevated levels of NO, TNF-α, IL-1β, and glutamate in the supernatants of wild type JEV-infected cells (Fig. 2A and B), as well as their corresponding mRNAs (Fig. 2C). UV-inactivated and Heat-inactivated JEV lost ability to stimulate microglia releasing of NO, TNF-α, and IL-1β (Fig. 2A). Reduction of those JEV infection-induced cytokine and glutamate changes was found in microglia treated with MK801 and memantine (Fig. 2B and C). Although glutamate itself had a marginal effect on cytokine production during the current study conditions, its presence augmented the JEV infection-induced production of NO, TNF-α, and IL-1β in microglia (Fig. 2D). All the findings reveal the substantial role NMDA receptors signaling play in JEV infection-induced pro-inflammatory cytokine expression through its interaction with concurrent intracellular changes upon JEV infection.

Fig. 2

MK801 and memantine alleviated cytokine expression in microglia. (A) Microglia cultures were infected with mock, JEV, JEV/UV-inactivated, or JEV/Heat-inactivated for 24 h. The supernatants were collected and subjected to ELISA for the measurement of NO, TNF-α, and IL-1β. Microglia cultures were infected with mock or JEV in the absence or presence of MK801 (10 µM) or memantine (5 µM) over time. The supernatants were collected and subjected to the measurement of NO, TNF-α, IL-1β (24 h), and glutamate (12 h) (B). Total cellular RNAs were extracted (8 h) and subjected to quantitative RT-PCR for the measurement of iNOS, TNF-α, and IL-1β mRNA. Quantitative results of the relative mRNA levels (arbitrary unit) are shown (C). (D) Microglia cultures were infected with mock or JEV in the absence or presence of glutamate (500 µM) for 24 h. The supernatants were collected and subjected to ELISA for the measurement of NO, TNF-α, and IL-1β. *p < 0.05 vs. Mock control and #p < 0.05 vs. JEV control, n = 4

NMDA receptor antagonists alleviated JEV infection-activated NF-κB and AP-1 signaling in microglia

NF-κB and AP-1govern pro-inflammatory axes lie downstream from the activation of the NMDA receptors and play key roles in JEV infection-induced microglia activation and cytokine expression [11, 15, 25]. Corresponding DNA binding activity and relevant upstream regulators of NF-κB and AP-1 for activation in microglia were assessed using EMSA and Western blotting, respectively. JEV infection caused an increase in the DNA binding activity of NF-κB and AP-1 (Fig. 3A) and upstream regulatory molecules of NF-κB and AP-1, including the protein phosphorylation of TAK1, ERK, JNK, p38, p65 and c-Jun, and the protein expression of c-Fos (Fig. 3B). JEV infection-increased DNA binding activity (Fig. 3A), protein phosphorylation, and protein expression (Fig. 3B) were all decreased by MK801 and memantine. The purinergic receptor P2 × 7 (P2 × 7R) is an ATP-gated non-selective cation channel that is highly expressed in microglia and can mediate oxidative stress and inflammatory responses. IRF5, a member of the intracellular mediators for type I interferon, is a transcription factor. Its expression in microglia participates in inflammatory responses [11]. In JEV-infected microglia, there were elevated expression of IRF5 and P2 × 7R. Protein levels of IRF5 and P2 × 7R in JEV-infected microglia were reduced by MK801 and memantine (Fig. 3B). Current findings imply that the anti-inflammatory effects of NMDA receptor blockade against JEV infection may come from the inhibition of the axis of NF-κB and AP-1 signaling as well as suppression of microglia reactive to JEV infection.

Fig. 3

MK801 and memantine alleviated NF-κB and AP-1 activation in microglia. Microglia cultures were infected with mock or JEV in the absence or presence of MK801 (10 µM) or memantine (5 µM) for 8 h. Nuclear proteins were extracted and subjected to EMSA for the measurement of NF-κB and AP-1 DNA binding activity. One representative blot of three independent culture batches and the quantitative results (arbitrary unit) are shown (A). Total cellular proteins were extracted and subjected to Western blot analysis with indicated antibodies. One representative blot of three independent culture batches and the quantitative results of relative protein contents (arbitrary unit) are shown (B). *p < 0.05 vs. Mock control and #p < 0.05 vs. JEV control, n = 3

NMDA receptor antagonists alleviated JEV infection-activated free radical generation in microglia

The NOX-mediated free radical generation can come from the signals of the NMDA receptors and contribute to pro-inflammatory activation involving mitochondrial dysfunction. Dynamin-associated GTPase Drp1 mediates mitochondrial fragmentation and involves in inflammatory responses and microglia activation [29,30,31]. Using the redox-sensitive fluorogenic probe, the levels of DCFDA fluorescence (Fig. 4A) increased in microglia after JEV infection. At the protein levels, JEV infection increased NOX2 and NOX4 protein expression as well as mitochondrial fission-related Drp1 protein phosphorylation in microglia (Fig. 4B). MK801 and memantine alleviated JEV infection-induced changes in fluorescence (Fig. 4A) and protein levels (Fig. 4B). To further examine the biological implications of free radical generation and Drp1 activation, effects of the corresponding antioxidant N-Acetyl Cysteine (NAC) and Drp1 inhibitor Mdivi-1 were evaluated. NAC and Mdivi-1 alleviated JEV infection-induced increases of DCFDA fluorescence (Fig. 4C) and expressions of NO, TNF-α, and IL-1β (Fig. 4D). Thus, JEV infection increases NOX expression, induces intracellular free radical generation, and activates mitochondrial Drp1, contributing to pro-inflammatory cytokine expression. The aforementioned changes are alleviated by MK801 and memantine.

Fig. 4

MK801 and memantine alleviated free radical generation in microglia. Microglia cultures were infected with mock or JEV in the absence or presence of MK801 (10 µM) or memantine (5 µM) for 8 h. The levels of intracellular free radicals were measured using the fluorogenic probe DCFDA (arbitrary unit) (A). Total cellular proteins were extracted and subjected to Western blot analysis with the indicated antibodies. One representative blot of three independent culture batches and the quantitative results of relative protein contents (arbitrary unit) are shown (B). Microglia cultures were infected with mock or JEV in the absence or presence of NAC (500 µM) or Mdivi-1(50µM) over time.The levels of intracellular free radicals (8 h) were measured using the fluorogenic probe DCFDA (arbitrary unit) (C). The supernatants were collected (24 h) and subjected to ELISA for the measurement of NO, TNF-α, and IL-1β (D). *p < 0.05 vs. Mock control and #p < 0.05 vs. JEV control, n = 4

NMDA receptor antagonists alleviated JEV infection-activated endoplasmic reticulum (ER) stress in microglia

NMDA receptor signaling possesses the ability to induce Unfolded Protein Response (UPS) and ER stress, resulting in pro-inflammatory activation [30, 32]. Parameters of ER stress and potential contribution to JEV infection-induced changes in microglia were then assessed. Upon JEV infection, microglia increased protein phosphorylation in PERK, eIF2α, and IRE1. In the presence of MK801 and memantine, there was a decline in protein phosphorylation (Fig. 5A). ER stress inhibitor, Salubrinal, had an alleviative effect on JEV infection-induced increases of DCFDA fluorescence (Fig. 5C) and expressions of NO, TNF-α, and IL-1β (Fig. 5D). These findings reveal the role NMDA receptor signaling plays in JEV infection-induced ER stress and its accompanied pro-inflammatory effects.

Fig. 5

MK801 and memantine alleviated ER stress in microglia. (A) Microglia cultures were infected with mock or JEV in the absence or presence of MK801 (10 µM) or memantine (5 µM) for 8 h. Total cellular proteins were extracted and subjected to Western blot analysis with indicated antibodies. One representative blot of three independent culture batches and the quantitative results of relative protein contents (arbitrary unit) are shown. Microglia cultures were infected with mock or JEV in the absence or presence of Salubrinal (10 µM) over time.The levels of intracellular free radicals (8 h) were measured using the fluorogenic probe DCFDA (arbitrary unit) (B). The supernatants were collected (24 h) and subjected to ELISA for the measurement of NO, TNF-α, and IL-1β (C).*p < 0.05 vs. Mock control and #p < 0.05 vs. JEV control, n = 4

Crosstalk among intracellular signaling in microglia

The aforementioned findings indicate the active role which NMDA receptor signaling plays in linking JEV infection and cytokine expression; probably involving CaMKII, oxidative stress, and ER stress. The central role and importance of CaMKII were next assessed using its pharmacological inhibitor KN93 in microglia. CaMKII inhibitor KN93 decreased NO, TNF-α, and IL-1β production in JEV-infected microglia (Fig. 6A). Their alleviation in cytokine expression paralleled with changes in DCFDA fluorescence (Fig. 6B), the protein expression of c-Fos, IRF5, P2 × 7R, NOX2, and NOX4, the protein phosphorylation of TAK1, ERK, JNK, p38, p65, c-Jun, Drp1, PERK, eIF2α, and IRE1 (Fig. 6C), as well as the DNA binding activity of NF-κB and AP-1 (Fig. 6D). These findings suggest that the Ca²⁺-activated CaMKII axis is complicated and plays a central role in linking the glutamate/NMDA receptor and NF-κB/AP-1 pro-inflammatory program in JEV-infected microglia.

Fig. 6

Effects on intracellular signaling cascades in microglia. Microglia cultures were infected with mock or JEV in the absence or presence of KN93 (5µM) over time. The supernatants (24 h) were collected and subjected to ELISA for the measurement of NO, TNF-α, and IL-1β (A). The levels of intracellular free radicals (8 h) were measured using the fluorogenic probe DCFDA (arbitrary unit) (B). Total cellular proteins (8 h) were extracted and subjected to Western blot analysis with indicated antibodies. One representative blot of three independent culture batches and the quantitative results of relative protein contents (arbitrary unit) are shown (C). Nuclear proteins (8 h) were extracted and subjected to EMSA for the measurement of NF-κB and AP-1 DNA binding activity. One representative blot of three independent culture batches and the quantitative results (arbitrary unit) are shown (D). *p < 0.05 vs. Mock control and #p < 0.05 vs. JEV control, n = 4

NMDA receptor antagonists protected against JEV infection-induced neuronal cell death and cytokine expression in neuron/glia and neuron

To implicate the consequences of MK801- and memantine-inhibited microglia activation in neuronal cell death, neuron/glia cultures consisting of neurons, astrocytes, and microglia were prepared for investigation. JEV infection increased CD68 immunoreactivity (Fig. 7A and B), CD68 protein content (Fig. 8A), and NO, TNF-α, and IL-1β production in neuron/glia cultures (Fig. 8B). As with microglia, MK801 and memantine alleviated expression in all pro-inflammatory parameters as well (Figs. 7A and B and 8A, and 8B). The degeneration of MAP-2 immunoreactive neurons (Fig. 7A and B) and reduction of MAP-2 protein content (Fig. 8A) in JEV-infected neuron/glia cultures were improved by MK801 and memantine. In contrast, the immunoreactivity of astrocytic GFAP (Fig. 7A and B) and GFAP protein content (Fig. 8A) were not markedly altered by either JEV infection, MK801, or memantine. JEV infection caused an elevation of apoptosis-related caspase 3 activity in neuron/glia cultures. MK801 and memantine alleviated the elevation of caspase 3 activity (Fig. 7C). Wild type JEV infection caused cell damage as evidenced by LDH efflux and the cytotoxicity was not seen in neuron/glia cultures infected with UV-inactivated and Heat-inactivated JEV (Fig. 8C). The LDH efflux in JEV-infected neuron/glia cultures was improved by MK801 and memantine (Fig. 8D). Cultured neurons were modeled for further neurotoxic study (Fig. 9A). The conditioned media of JEV-infected microglia caused elevated LDH efflux in neuron cultures, while the conditioned media of MK801- and memantine-treated microglia upon JEV infection decreased ability in causing LDH efflux in neuron cultures (Fig. 9B). Additionally, the presence of MK801 and memantine in neuron cultures alleviated JEV-infected microglia conditioned media-induced LDH efflux (Fig. 9C). These findings show that JEV infection causes neuronal cell death and microglial activation, and that MK801 and memantine offer anti-inflammatory and neuroprotective effects.

Fig. 7

MK801 and memantine alleviated neuronal cell death in Neuron/glia. Neuron/glia cultures were infected with mock or JEV in the absence or presence of MK801 (10 µM) or memantine (5 µM) for 48 h. Cells were subjected to immunocytochemical staining with antibodies recognizing MAP-2, GFAP, and CD68. Representative photomicrographs are shown. Bar, 60 μm (A). The numbers of viable MAP-2 immunoreactive neurons, GFAP immunoreactive astrocytes, and CD68 immunoreactive microgliaare depicted (B). Total cellular proteins were isolated and subjected to the measurement of caspase 3 activity (arbitrary unit) (C). *p < 0.05 vs. mock control and #p < 0.05 vs. JEV control, n = 4

Fig. 8

MK801 and memantine alleviated cytokine expression in Neuron/glia. Neuron/glia cultures were infected with mock or JEV in the absence or presence of MK801 (10 µM) or memantine (5 µM) for 48 h.Total cellular proteins were extracted and subjected to Western blot analysis with indicated antibodies. One representative blot of three independent culture batches and the relative protein contents (arbitrary unit) are shown (A).The supernatants were collected and subjected to ELISA for the measurement of NO, TNF-α, and IL-1β (B). (C) Neuron/glia cultures were infected with mock, JEV, JEV/UV-inactivated, or JEV/Heat-inactivatedfor 48 h. Cell damage was measured by LDH efflux assay. (D) Neuron/glia cultures were infected with mock or JEV in the absence or presence of MK801 (10 µM) or memantine (5 µM) for 48 h. Cell damage was measured by LDH efflux assay. *p < 0.05 vs. mock control and #p < 0.05 vs. JEV control, n = 4

Fig. 9

MK801 and memantine alleviated neuronal cell death in neurons. (A) Cultured neurons were examined by immunocytochemical staining with antibody recognizing MAP-2. Representative photomicrograph is shown. Bar, 60 μm. (B) Microglia cultures were infected with mock or JEV in the absence or presence of MK801 (10 µM) or memantine (5 µM) for 48 h. The supernatants were collected and subjected to UV-inactivation. The resultant media (Mock CM and JEV CM) were mixed with an equal volume of fresh medium and then added to the cultured neurons for 24 h. Cell damage was measured by LDH efflux assay. (C) Additionally, the manipulated media form mock- (Mock CM) and JEV-infected microglia (JEV CM) were added to cultured neurons in the presence of MK801 (10 µM) or memantine (5 µM) for 24 h. Cell damage was measured by LDH efflux assay. *p < 0.05 vs. mock control and #p < 0.05 vs. JEV control, n = 4

Memantine protected against JEV infection-induced changes in the brains of mice

To further verify the in vitro findings, a C57BL/6 mouse model of JEV infection was produced for in vivo investigation [27]. It is reported that severe adverse effects of MK801 preclude its clinical application, while memantine is well tolerated in clinical use [33]. Therefore, memantine represents a preferable candidate of NMDA receptor antagonists for in vivo study. At day 7 post-infection, JEV-infected mice showed abnormal gait, stiffness, and paralysis without death. We found that JEV infection caused neuronal cell degeneration and disrupted morphological integrity in the brains of mice, as evidenced by MAP-2 immunofluorescence in the hippocampal regions. Memantine alleviated JEV infection-induced neuronal cell degeneration (Fig. 10A). Upon JEV infection, Iba1-labelled microglia became reactive and the alteration was alleviated by memantine (Fig. 10A). However, GFAP immunofluorescence astrocytes were not remarkably altered by JEV infection or memantine (Fig. 10A). To substantiate histological findings, quantitative measurement of corresponding MAP-2, CD68, and GFAP in protein content revealed the same changes as those in immunofluorescence study (Fig. 10B). In JEV-infected brains, contents of JEV genomic RNA (Fig. 10C) and JEV NS1 protein (Fig. 10D) in memantine group slightly decreased. However, the difference failed to reach statistical significance. The findings indicate that JEV infection induces neuronal cell degeneration and microglia activation in brains of mice and that memantine offers protective effects.

Fig. 10

Memantine alleviated cell alterations in the brains of JEV-infected mice. Eight-week-old male C57BL/6 mice were infected with mock or JEV through intraperitoneal injection. Mice were access to drinking water containing vehicle or memantine (20 mg/kg) starting from 1 day before infection and lasting for 8 days. On post-viral infection day 7, mice were euthanized and the brains were isolated. (A) The frozen sections of brain tissues (n = 3) were subjected to immunohistochemistry with fluorescent labeling for the detection of MAP-2-immunoreactive (FITC), Iba1-immunoreactive (FITC), and GFAP-immunoreactive (FITC) cells. The cell nuclei were counterstained with DAPI. Representative photomicrographs of hippocampal regions are shown. Scale bar, 50 μm. (B) Total proteins were extracted from prefrontal cortices and hippocampus (n = 3) and subjected to Western blot analysis with indicated antibodies. One representative blot and the relative protein contents (arbitrary unit) are shown. (C) Total RNAs (n = 3) were isolated from prefrontalcortices and hippocampus and subjected to quantitative RT-PCR for the measurement of JEV RNA (arbitrary unit). (D) Total proteins were extracted from prefrontal cortices and hippocampus (n = 3) and subjected to Western blot analysis with indicated antibodies. One representative blot and the relative protein contents (arbitrary unit) are shown. *p < 0.05 vs. mockvehicle and #p < 0.05 vs. JEVvehicle, n = 3

Memantine alleviated JEV infection-induced inflammatory changes in the brains of mice

Data of in vitro microglia study revealed an active role of the NMDA receptor signaling in inflammatory responses caused by JEV infection. Thus, parameters of the identified inflammatory axis were determined in the brains of JEV-infected mice. Upon JEV infection, protein phosphorylation in GluN1, GluN2A, GluN2B, and CaMKII increased and the increments were alleviated by memantine (Fig. 11A). Double immunofluorescence assay in the brains of hippocampal regions highlighted that some signals of P-GluN2A fluorescence were colocalized with signals of Iba1 fluorescence (Fig. 11B), indicating an activation of the NMDA receptor signaling in microglia after JEV infection. The results were consistent with the in vitro microglia findings, with brains from JEV-infected mice showing increased mRNA contents of iNOS, TNF-α, and IL-1β (Fig. 12A), DNA binding activity of NF-κB and AP-1 (Fig. 12B), protein contents of NOX2, NOX4, c-Fos, IRF5, and P2 × 7R, and protein phosphorylation of ERK, JNK, p38, PERK, eIF2α, IRE1, TAK1, p65, c-Jun, and Drp1 (Fig. 12C). Memantine alleviated those changes (Fig. 12C). The findings further confirm the activation of a NMDA receptor-related inflammatory response and an alleviative effect of memantine in the brains of JEV-infected mice.

Fig. 11

Memantine alleviated NMDA receptor signaling in the brains of JEV-infected mice. Eight-week-old male C57BL/6 mice were infected with mock or JEV through intraperitoneal injection. Mice were access to drinking water containing vehicle or memantine (20 mg/kg) starting from 1 day before infection and lasting for 8 days. On post-viral infection day 7, mice were euthanized and the brains were isolated. (A) Total proteins were extracted from prefrontal cortices and hippocampus (n = 3) and subjected to Western blot analysis with indicated antibodies. One representative blot and the relative protein contents (arbitrary unit) are shown. (B) The frozen sections of brain tissues (n = 3) were subjected to immunohistochemistry with fluorescent labeling for the detection of Iba1-immunoreactive (FITC) and P-GluN2A-immunoreactive (Rhodamine) cells. The cell nuclei were counterstained with DAPI. Representative photomicrographs of hippocampal regions and merged images are shown. Scale bar, 50 μm. *p < 0.05 vs. mockvehicle and #p < 0.05 vs. JEVvehicle, n = 3

Fig. 12

Memantine alleviated signaling alterations in the brains of JEV-infected mice. Eight-week-old male C57BL/6 mice were infected with mock or JEV through intraperitoneal injection. Mice were access to drinking water containing vehicle or memantine (20 mg/kg) starting from 1 day before infection and lasting for 8 days. On post-viral infection day 7, mice were euthanized and the brains were isolated. (A) Total RNAs (n = 3) were isolated from prefrontalcortices and hippocampus and subjectedto quantitative RT-PCR for the measurement of iNOS, TNF-α, and IL-1β mRNA (arbitrary unit). (B) Nuclear proteins (n = 3) were extracted from prefrontalcortices and hippocampus and subjected to EMSA for the measurement of NF-κB and AP-1 DNA binding activity. One representative blot and the quantitative results (arbitrary unit) are shown. (C) Total proteins were extracted from prefrontal cortices and hippocampus (n = 3) and subjected to Western blot analysis with indicated antibodies. One representative blot and the relative protein contents (arbitrary unit) are shown. *p < 0.05 vs. mockvehicle and #p < 0.05 vs. JEVvehicle, n = 3

Discussion

Activation of the neuronal NMDA receptor and the neuroprotective effects of NMDA receptor blockade have been demonstrated in rodent models of JE [4, 20, 21]. Using microglia cultures, this study demonstrated that JEV infection caused activation of the NMDA receptor, together with elevated signals of glutamate release, Ca²⁺ mobilization, CaMKII, MAPKs, NOXs, ER stress, Drp1, and NF-κB/AP-1. Data of molecular, biochemical, and pharmacological studies highlighted a crucial role of CaMKII in transducing the NMDA receptor signaling in JEV-infected microglia. CaMKII can communicate with regulators of NF-κB/AP-1 to induce microglia M1 polarization and pro-inflammatory cytokine expression involving phosphorylatory modification, ER stress, and mitochondrial dysfunction. Pro-inflammatory cytokines released from JEV-infected microglia and NMDA receptor activation provoke neurotoxic potential in cultured neurons and neuron/glia. NMDA receptor antagonists, MK801 and memantine, have anti-inflammatory and neuroprotective effects in JEV-infected cells (Fig. 13). Activation of NMDA receptor-related inflammatory changes, microglia activation, and neurodegeneration as well as reversal effects of memantine are revealed in the brains of JEV-infected mice, as well. Although activation of the NMDA receptor, CaMKII, and NF-κB/AP-1 has been implicated in the degeneration of neurons [34], currently, their changes in JEV-infected neurons were not investigated. The cell and rodent studies presented here further extend earlier findings that showed JEV infection induces NMDA receptor activation in microglia as well, while providing new insight into the anti-inflammatory and neuroprotective potential of NMDA receptor blockade in combating viral pathogenesis, such as that seen in JE.

Fig. 13

A possible schema of actions of NMDA receptor elicited by JEV infection is proposed. This schematic diagram indicates the cellular molecules, signaling molecules, and cascades used in mediating microglia M1 polarization and pro-inflammatory cytokine expression as well as neuronal cell death after JEV infection. JEV infection-activated microglia have ability to cause neuron toxicity. NMDA receptor coordinates cellular changes in neuron and microglia after JEV infection and NMDA receptor antagonists, MK801 and memantine, have alleviative effects. Some additional signaling molecules and cascades have been omitted for the sake of clarity

Most cells express glutamate ionotropic receptors, metabotropic receptors, and transporters, including neurons, astrocytes, and microglia [18, 35, 36]. Microglia represent an emerging source of extracellular glutamate release through the transporter efflux system upon inflammatory stimulation [15, 35, 36]. Both glutamate release and neuronal NMDA receptor activation have been implicated in JE-associated neuronal cell death and neuroinflammation [4, 20, 21]. Previously, we demonstrated that JEV infection caused neuronal cell death and a protective effect of MK801, implying a neurotoxic involvement of glutamate [15]. There are both multiple and interrelated mechanisms which cause neuronal cell death and contribute to JEV pathogenesis. Besides well-known neurotoxicity of abnormal neuronal NMDA receptor activation, in the present study, we further demonstrated the activation of the NMDA receptor in microglia upon JEV infection, and provided additional evidence implicating the role the dysregulated glutamate/NMDA receptor axis plays in JE-associated neuronal cell death and neuroinflammation.

The interplay between excitotoxicity and neuroinflammation is a key component to the development of neurodegeneration. Injection of NMDA in rat brains reveals an induction of excitotoxic neuronal damage and neuroinflammation [37]. Evidence indicates an emerging role where pro-inflammatory cytokines and TNF-α and IL-1β represent dominant candidates as linking molecules. TNF-α and IL-1β activate neuronal glutaminase expression and glutamate extracellular release, leading to glutamate production and glutamate-mediated neurotoxicity [16]. TNF-α potentiates glutamate-mediated neurotoxicity through the inhibition of glutamate uptake on astrocytes and the promotion of surface expression in the neuronal NMDA receptor [38]. A vicious crosstalk between glutamate and pro-inflammatory cytokines in JEV-infected microglia has been demonstrated in both the current study and our previously reported works. Upon JEV infection, microglia-released TNF-α increased both microglia glutaminase expression and glutamate production, resulting in glutamate extracellular release [15]. We found that activation of the NMDA receptor and accompanied Ca²⁺ mobilization led to both microglia activation and pro-inflammatory cytokine production by cultured microglia infected with JEV. The exogenous addition of glutamate augmented both JEV infection-induced microglia activation and pro-inflammatory cytokine production. Additionally, we further provided evidence to demonstrate the substantial roles excitotoxicity and neuroinflammation play in JEV pathogenesis. Therefore, our findings revealed the crucial role microglia plays in the amplification of driving molecules critical to excitotoxicity and neuroinflammation.

Unlike neurons, microglia are tolerated by glutamate excitotoxicity. On the contrary, microglia switch the glutamate/NMDA receptor signals to promote cell proliferation and initiate an inflammatory program. In addition to endotoxins and neurotoxic compounds, viral infections and viral proteins also orchestrate a pro-inflammatory state of microglia througha NMDA receptor mechanism involving the Ca²⁺ current, CaMKII, Mitogen-Activated Protein Kinases (MAPKs), TAK1 and NF-κB/AP-1 [24, 25, 39,40,41]. Upon JEV infection, elevation in the NMDA receptor component protein phosphorylation, labile free Ca²⁺, CaMKII phosphorylation, TAK1 phosphorylation, ERK phosphorylation, JNK phosphorylation, p38 phosphorylation, NF-κB regulators and AP-1 regulators paralleled with microglia reactivity and pro-inflammatory cytokine expression, with those accompanied changes ameliorated by MK801 and memantine. Therefore, JEV infection shifts microglia to a pro-inflammatory state, at least in part, through a typical NMDA receptor/CaMKII/TAK1/MAPKs/NF-κB, AP-1 axis.

Transcription factors, such as NF-κB and AP-1, determine transcriptional programs for orchestrating the pro-inflammatory state of microglia and governing pro-inflammatory cytokine expression [11, 15, 25]. There are additional intracellular signaling cascades lying downstream from the axis of NMDA receptor/Ca²⁺/CaMKII, where ER stress, mitochondrial dysfunction, and NOXs are commonly investigated [29,30,31,32]. Microglial NOXs, particularly NOX2 and NOX4, have been implicated in the generation of free radicals, closely linking to ER stress, mitochondrial Drp1 activation, and redox-sensitive MAPKs, NF-κB, and AP-1 [31, 42,43,44]. An adequate ER stress will alert the cells to take on a defensive response for adaptation, while sustained or uncontrolled ER stress delivers signals to effectors, which has an impact on diverse cell functions. Regarding microglial activation, ER stress-accompanied JNK and p38 have substantial roles in pro-inflammatory transcriptional programs and inflammasome assembly/activation [32, 45, 46]. NOX inhibitors have inhibitory effects on Matrix Metalloprotease-9 (MMP-9) expression through JEV-infected astrocytes [47]. The induction and role of ER stress are characterized and demonstrated in JEV infection-induced neuronal cell apoptosis and autophagy [10, 48]. Using cultured rat microglia, increases in intracellular free radical generation, NOX2/NOX4 protein expression, and ER stress marker expression were revealed after JEV infection. Parallel elevations in JEV-infected microglia occurred in redox-sensitive MAPKs, NF-κB, and AP-1 signaling, along with pro-inflammatory cytokine expression. Their crosstalk and specific contribution to JEV infection-acquired pro-inflammatory state of microglia and pro-inflammatory cytokine expression were further demonstrated using corresponding pharmacological inhibitors. Although a complicated and mutually interactive intracellular signaling cascade has been drawn in this study, the exact step-by-step crosstalk was not fully investigated in detail due to a lack of temporal study.

JE-associated neuronal cell death is a complicated mechanism involving diverse alterations, including apoptosis, autophagy, necroptosis, ER stress, oxidative stress, Receptor Interacting Serine/Threonine-Protein Kinase 3 (RIPK3), Mixed-Lineage Kinase Domain-Like Protein (MLKL), NMDA and Caveolin-1 [5, 8,9,10, 21, 47, 48]. The protective potential of NMDA receptor antagonists has been demonstrated in rodent models of JE [4, 20, 21]. Tethering with our previous report [15], the neuroprotective effects of NMDA receptor blockade were further demonstrated in cultured neurons and neuron/glia by challenging with either JEV infection or conditioned media of JEV-infected microglia. There are studies showing microglial activation upon JEV infection through numerous mechanisms, including TLRs, MLKL, impaired autophagy flux, CLEC5A and Heat Shock Protein (HSP) [8, 9, 11, 50, 51]. Herein, we have added the NMDA receptor axis to the list of mechanisms contributing to JEV infection-induced microglial activation. Like other stressed conditions [24, 25, 39,40,41], the activation of NMDA receptor signaling through component protein phosphorylatory mechanisms and their associated Ca²⁺ current and CaMKII was demonstrated in JEV-infected microglia with concurrent glutamate extracellular release. Along with glutamate and endogenous mimicking ligands, Src, Protein Kinase A (PKA), PKC, ERK and TLR4 protein interaction all play positive roles in the NMDA receptor subunit components phosphorylation and activation [26, 52]. The activation of TLR4, Src, ERK, PKA and PKC has been reported in JEV-infected cells, particularly in microglia [2, 11, 15, 27, 53]. Therefore, JEV infection-activated microglial NMDA receptor signaling could be mediated through multifactorial mechanisms via ligand engagement and/or ligand-independent subunit component phosphorylation. Although phosphorylation and activation of the NMDA receptor had been revealed in JEV-infected microglia, neuron/glia, and mice brains, the detailed mechanisms underlying the NMDA receptor activation in JEV-infected cells are yet to be identified.

The NMDA receptor-controlled inflammatory responses, neuronal cell death, and lethality had been demonstrated in mouse model of JEV infection. NMDA receptor blockade abrogated JEV infection-induced neuronal cell death and glial activation in mice brain tissues and delayed the death of mice, while having little effect on viral amplification [4, 21]. Similar post-infection findings were seen in our JEV-infected mice treated with memantine. The results of molecular and biochemical studies in brain tissues were consistent with the in vitro findings, confirming the activation of a NMDA receptor-related inflammatory response and an alleviative effect of memantine. Importantly, data of double immunofluorescence staining centered on Iba1 and P-GluN2A suggested an activation of the NMDA receptor component in microglia in responding to JEV infection. The expression of NMDA receptor components in lectin positive amoeboid microglia was also revealed in rat brain tissues following hypoxic exposure [18]. Direct inhibition of viral amplification or targeting virus-induced signaling changes represents druggable options for the treatment of viral diseases. Current in vitro and in vivo findings highlighted an active role of the NMDA receptor signaling in JEV infection-induced microglia activation and inflammation and revealed that NMDA receptor blockade provided protection against neuroinflammation and neurodegeneration during the JEV infection. Since the pathogenic role of the neuronal NMDA receptor in JE was reported [20] and the colocalization of P-GluN2A immunoreactivity with additional cells were not examined in the current study, the detailed action mechanisms and cell types involved in the protective effects of the NMDA receptor blockade required further investigation.

Although we provided interesting findings, current study was still being suffered from some limitations. First, activation and functional execution of the NMDA receptor have been demonstrated in cultured microglia and microglia cell lines [25, 54, 55], however, there is no detectable NMDA receptor-dependent current in microglia using model of brain slices [22]. Despite the demonstration that Iba1-positive microglia expressed hyperphosphorylated GluN2A in the brain tissues of JEV-infected mice, in vivo expression of functional NMDA receptor in microglia should be evaluated with additional approaches. Second, Ca²⁺ channels are governed by several receptor families, including the ionic glutamate receptors, metabotropic glutamate receptors, and purinergic ionotropic receptors [17, 56]. Constitutive changes of Ca²⁺ elevation in JEV-infected microglia implied a complicated regulatory mechanism. As with NMDA receptor antagonists MK801 and memantine, group I metabotropic glutamate receptor antagonist 1-Aminoindan-1,5-Dicarboxylic Acid (AIDA), metabotropic glutamate receptor subtype 5 antagonist 2-Methyl-6-(phenylethynyl)pyridine (MPEP), and purinergic ionotropic P2 × 7 receptor antagonist A438079 decreased JEV infection-induced Ca²⁺ elevation and cytokine production in microglia (unpublished data). Currently, the roles of metabotropic glutamate receptors and purinergic ionotropic receptors in JEV infection-induced microglia activation and neuroinflammation are actively investigated. Due to complicated actions, use of single agent such as MK801 or memantine is unlikely to offer sufficient anti-inflammation and neuroprotection against JEV infection. Third, current in vitro and in vivo findings only represented a simple and one-sided description of the changes of signaling molecules and pharmacological effects of the NMDA receptor antagonists in JEV infection-induced microglia activation and inflammation. Genetic manipulation of the NMDA receptor components is an alternative approach to avoid off-target effects caused by pharmacological agents. Besides, a signaling network investigation in both neurons and microglia is necessary to get a better action illustration. Therefore, these limitations should be carefully considered in the interpretation of current results and in future research on the mechanisms of JEV infection-associated microglia activation and neuroinflammation.

Glutamate neurotoxicity is commonly seen in neurological disorders, with no exception for viral encephalitis [13, 14, 57]. Through this study, we have provided experimental evidence demonstrating the pro-inflammatory potential of the NMDA receptor axis in JEV-infected microglia, as well as the anti-inflammatory effects of NMDA receptor blockade by MK801 and memantine. Upon JEV infection, extracellular released glutamate and its microglia NMDA receptor counterpart transduce diverse signals to turn on Ca²⁺ current, CaMKII, MAPKs, NOXs, ER stress, and Drp1 finally converging to NF-κB/AP-1 transcriptional program. Both glutamate and pro-inflammatory cytokines released from JEV-infected microglia provoke neurotoxic programs and neuronal cell death. Therefore, the current findings highlight the crucial role the glutamate/NMDA receptor axis plays in linking excitotoxicity and neuroinflammation during the course of JEV pathogenesis. The NMDA receptor blockade offers benefits in rodent models of JE [4, 20, 21]. Our data further elicit the anti-inflammatory mechanisms and suggest that NMDA receptor blockade is a proposed anti-inflammatory and neuroprotective strategy for the treatment of neuroinflammation- and neuronal cell death-accompanied brain pathogenesis, such as JE. Since the inflammatory potential of glutamate receptors and transporters varies and only the NMDA-type of receptor has been investigated, a deeper investigative insight into glutamate’s actions is still required.