All experiments were approved by the institutional biosafety committee (IBC) and performed in accordance with approved animal protocols of the Korea Brain Research Institute (KBRI, approval no. IACUC-19-00042).

Gliquidone was purchased from Selleck Chemicals (S3151, Houston, TX, United States) and Tokyo Chemical Industry Co., Ltd. (G0332, Tokyo, Japan). Gliquidone was used at a concentration of 5 μM (in 1% DMSO) in in vitro assays and injected at a dose of 10 or 20 mg/kg (i.p., dissolved in 5% DMSO, 10% PEG, 20% Tween 80) in in vivo experiments.

Male C57BL6/N mice (wild-type; Hana Company, Busan, South Korea) were used in in vivo experiments. A pathogen-free facility with a 12 h light/dark cycle and a temperature of 22°C was used to house all mice. Daily injections of gliquidone (10 or 20 mg/kg) or vehicle (5% DMSO) were administered intraperitoneally (i.p.) on 3 consecutive days. On the third day, 10 mg/kg LPS (i.p.) or PBS was injected. The mice were sacrificed 8 h after the last injection and fixed in 4% paraformaldehyde. Subsequently, mouse brain slices with a thickness of 30 μm were prepared using a Leica CM1850 cryostat (Leica Biosystems, Buffalo Grove, IL, United States).

Mouse brain slices were incubated in 10% normal goat serum (Vector Laboratories) at room temperature for 1 h, followed by immunostaining with rabbit or mouse anti-target antibody at 4°C overnight. The sections were subsequently incubated with an Alexa Fluor 488- or 555-labeled goat IgG secondary antibody for 2 h at room temperature. Sections were mounted on glass slides were covered with a DAPI (Vector Laboratories)-containing mounting solution, and images were obtained by fluorescence microscopy (DMi8, Leica Microsystems, Wetzlar, Germany) for analysis by Image J software (NIH). Detailed information on the primary and secondary antibodies used for IF and immunocytochemistry (ICC) is provided in Table 1.

Wild-type mice were injected daily with gliquidone (10 or 20 mg/kg, i.p.) or vehicle (5% DMSO + 10% PEG + 20% Tween80) for 3 days, followed by injection of LPS (10 mg/kg, i.p.) or PBS 30 min after the last injection on day 3. Eight hours later, blood was sampled and centrifuged at 2,000 rpm for 20 min, and the supernatant (serum) collected and stored at −80°C until analysis. Peripheral IL-1β, IL-6, and TNF-α levels were measured using the appropriate ELISA kit (IL-1β, Cat. no. 88-7013-88; IL-6, Cat. no. 88-7064-88; and TNF-α, Cat no. 88-7324-88; Invitrogen, Waltham, MA, United States) as described by the manufacturer.

BV2 microglial cells (generously provided by Dr. Kyung-Ho Suk) were cultured at 37°C and 5% CO₂ in DMEM high glucose (Invitrogen, Carlsbad, CA, United States) containing 5% fetal bovine serum (FBS, Invitrogen).

Primary astrocytes were prepared from postnatal day 1 (P1) C57BL6/N mice as previously described (Ryu et al., 2019). Briefly, whole brains were dissected and minced through 70 μm mesh, and the mixed cells were grown in low-glucose DMEM (1,000 mg/l glucose) supplemented with 10% FBS, 100 unit/ml penicillin, and 100 μg/ml streptomycin for 2 weeks. On day 14, primary microglial cells were detached by shaking at 250 rpm at room temperature overnight, followed by dissociation of primary astrocytes using trypsin-EDTA. After 3 rounds of washing and centrifugation for 10 min at 2,000 rpm, the pellet containing primary astrocytes was used for experiments. The cells were seeded at 7.0 × 10⁵ cells/well in 12-well plates for q-PCR and 2.5 × 10⁵ cells/well for ICC. After 2 days, the primary astrocytes were fixed in 4% paraformaldehyde at room temperature for 10 min and immunostained with an antibody against GFAP (a marker of astrocytes) at 4°C overnight. Next, the cells were washed thrice with PBS, and secondary antibodies were added for 1 h. Finally, the cells were washed with PBS, incubated with DAPI, and mounted. Images were acquired by fluorescence microscopy (DMi8, Leica Microsystems, Wetzlar, Germany), and primary astrocyte purity was calculated as follows: [Astrocyte purity (%) = (GFAP- and DAPI-positive cells/DAPI-positive cells) × 100].

Cell viability and cytotoxicity were assessed with the CCK-8 assay (Dongin Biotech Co., Ltd., Seoul, South Korea) as recommended by the manufacturer. After seeding in 96-well plates at a density of 2 × 10⁴ cells/well and incubation overnight at 37°C, BV2 cells were treated with gliquidone (0.1, 1, 5, 10, 25, and 50 μM) or vehicle (0.001, 0.01, 0.05, 0.1, 0.25, and 0.5% DMSO) for 6 h at 37°C. Next, CCK solution was added and incubated in the dark for 0.5 h, followed by detection at 450 nm in a SPECTROstar Nano microplate reader (BMG Labtech, Germany).

The cytotoxicity of gliquidone was assessed using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. BV2 cells seeded into 96-well plates without FBS at a density of 4 × 10⁴ cells/well were treated for 24 h with gliquidone (0.1, 1, 5, 10, 25, and 50 μM) or vehicle (0.001, 0.01, 0.05, 0.1, 0.25, and 0.5% DMSO), followed by incubation with 0.5 mg/ml MTT for 3 h in the dark. Finally, DMSO was added to dissolve the formazan crystals with shaking, and the absorbance at 570 nm was measured using a SPECTROstar Nano microplate reader (BMG Labtech, Germany).

Reverse transcription-polymerase chain reaction was performed to assess the effects of gliquidone on LPS-evoked microglial and astroglial inflammatory cytokine levels. First, total RNA was isolated from BV2 microglial cells or primary astrocytes using QIAzol Lysis Reagent (Qiagen, Cat No. 79306) and reverse transcribed into cDNA. The cDNA was subsequently used as the template in RT-PCR using Prime Taq Premix (GeNet Bio) as previously described with the primers shown in Table 2. The amplicons were separated by electrophoresis on a 1.5% agarose gel containing EcoDye (1:5000, Biofact, Daejeon, South Korea), and images were analyzed using the software Fusion Capt Advance (Vilber Lourmat, Eberhardzell, Germany).

The effects of gliquidone on LPS-mediated microglial and astroglial NLRP3 inflammasome activation and subsequent proinflammatory cytokine production were assessed in BV2 microglial cells and mouse primary astrocytes. In the preventive/pretreatment experiments, cells were first treated for 30 min with gliquidone (5 μM) or vehicle (1% DMSO) and then with LPS (200 ng/ml) or PBS for 5.5 h or 23.5 h. In the curative/post-treatment experiments, cells were first treated for 30 min with LPS (200 ng/ml) or PBS and then with gliquidone (5 μM) or vehicle (1% DMSO) for 5.5 h or 23.5 h. Total RNA was then extracted using TRIzol (Invitrogen, Waltham, MA, United States) as recommended by the manufacturer and reverse-transcribed (1 μg) to synthesize cDNA (GeNet Bio, Chungcheongnam-do, South Korea). The cDNA was used with Fast SYBR Green Master Mix (Thermo Fisher Scientific, Waltham, MA, United States) to perform real-time q-PCR in a QuantStudio 5 Real-Time PCR System (Applied Biosystems, Thermo Fisher Scientific, Waltham, MA, United States). The cycle threshold (Ct) values of the mRNA levels of factors related to inflammasome and inflammation were normalized to the Ct value for gapdh, and the fold change relative to the vehicle-treated control was quantified. Used primers are shown in Table 3.

To determine the effects of gliquidone on LPS-induced ERK signaling, western blotting was conducted. BV2 cells or primary astrocytes were lysed using lysis buffer (ProPrep, iNtRON Biotechnology, Inc., Seongnam, South Korea), followed by centrifugation at 12,000 rpm for 15 min. After quantifying the protein concentration in the supernatant, 15 μg of protein sample was separated by 8% SDS gel electrophoresis. The proteins in the gel were subsequently transferred to a polyvinylidene difluoride (PVDF) membrane. After blocking with 5% skim milk or 5% BSA at room temperature for 1 h, the membrane was incubated with anti-p-ERK (1:1000, Cell Signaling), anti-ERK (1:1000, Cell Signaling), or β-actin (1:1000, Santa Cruz Biotechnology) at 4°C. Finally, HRP-conjugated goat anti-mouse IgG or HRP-conjugated goat anti-rabbit IgG (both 1:1000, Enzo Life Sciences, Farmingdale, NY, United States) was added to the membrane and incubated for 1 h, and detection was performed using ECL Reagent (GE Healthcare, Chicago, IL, United States). Fusion Capt Advance software was used for image acquisition and analysis (Vilber Lourmat).

BV2 microglial cells and primary astrocytes were fixed in 4% paraformaldehyde for 10 min, washed thrice with PBS, and incubated overnight with anti-target antibody in GDB buffer as described previously (Nam et al., 2018). After washing with PBS, the cells were incubated for 1 h with an Alexa Fluor 488- or 555-conjugated antibody at room temperature. Finally, images of cells mounted in DAPI (Vector Laboratories, CA, United States) were captured using a fluorescence microscope (DMi8, Leica Microsystems, Wetzlar, Germany) and analyzed using ImageJ software.

GraphPad Prism 7 software (GraphPad Software, San Diego, CA, United States) was used for data analyses. Comparisons between two groups were performed using the unpaired two-tailed T-test with Welch’s correction; one-way ANOVA was employed for multiple comparisons (except for hippocampal NLRP3 IF). Tukey’s test or the Newman–Keuls multiple comparisons test was used for multiple comparisons with significance set at p < 0.05. Means ± SD are presented (^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001).

The anti-diabetic drug gliquidone was recently shown to promote a beneficial microglial phenotype via the Kir6-containing ATP-dependent potassium channel in a mouse model of Parkinson’s disease (PD) (Du et al., 2018). However, the effects of gliquidone on LPS-induced proinflammatory responses are unknown. Therefore, we investigated whether gliquidone modulates LPS-mediated microglial activity by injecting wild-type mice daily with gliquidone (10 or 20 mg/kg, i.p.) or vehicle for 3 days; after the final administration of gliquidone, LPS (10 mg/kg, i.p.) or PBS was injected. Eight hours after LPS or PBS injection, IF staining of brain slices was performed with an anti-Iba-1 antibody, and Iba-1 intensity, Iba-1-positive cells, and Iba-1-labeled area were analyzed in the cortex and hippocampus (Figures 1A–E). Treatment with 20 mg/kg gliquidone significantly decreased the LPS-induced increases in Iba-1 immunoreactivity, Iba-1-positive cells, and Iba-1-labeled area in the cortex and hippocampal CA1 and CA3 regions (Figures 1A–E). By contrast, administration of 10 mg/kg gliquidone significantly reduced the LPS-induced increases in Iba-1 intensity, Iba-1-positive cells, and Iba-1-labeled area only in the hippocampal CA3 region (Figures 1C–E). These results indicated that a gliquidone dose of 20 mg/kg more effectively downregulates LPS-mediated microgliosis and changes in microglial kinetics and morphology in the wild-type mouse brain.

To determine if gliquidone alters LPS-evoked astrocyte activation in vivo, we treated wild-type mice according to same paradigm described above and performed IF staining of brain slices with an anti-GFAP antibody. GFAP intensity, GFAP-positive cells, and GFAP-labeled area were measured (Figures 2A–E). At doses of 10 and 20 mg/kg, gliquidone significantly reduced the LPS-induced increases in GFAP immunointensity and GFAP-positive cells in the cortex but not the hippocampus (Figures 2C–E). In addition, administration of 10 or 20 mg/kg gliquidone did not alter GFAP-labeled area in the cortex and hippocampus (Figure 2E). These data indicate that gliquidone modulates LPS-stimulated astrocyte activation in the mouse brain, albeit less effectively than microglial activation.

Since gliquidone downregulated LPS-induced micro- and astrogliosis in wild-type mice, we next investigated whether gliquidone modulates LPS-stimulated proinflammatory cytokine levels in vivo. Wild-type mice were treated as described above, and brain slices were subjected to IF staining with anti-COX-2 or anti-IL-6 antibodies. Treatment with 20 mg/kg gliquidone significantly diminished LPS-stimulated levels of COX-2 in the cortex and hippocampal CA3 region but not the hippocampal CA1 and DG regions (Figures 3A–C). At a dose of 10 mg/kg, gliquidone significantly decreased the LPS-evoked increase in COX-2 levels in the cortex and hippocampal CA3 region (Figures 3A–C). Moreover, injection of 10 or 20 mg/kg gliquidone following LPS administration markedly suppressed IL-6 levels in the cortex and hippocampus in wild-type mice, and the effects of 10 mg/kg gliquidone were significantly greater than those of 20 mg/kg gliquidone (Figures 4A–C). These findings suggest that gliquidone suppresses LPS-stimulated COX-2 and IL-6 levels in a model of neuroinflammatory disease.

The induction of proinflammatory cytokines by LPS is closely linked to microglia- and/or astrocyte-associated inflammation as well as inflammasome formation (Su et al., 2021; Zhao et al., 2021). Thus, we examined whether gliquidone affects LPS-induced NLRP3 inflammasome activation in vivo. Wild-type mice were treated as described above, and IF staining of brain slices was conducted with an anti-NLRP3 antibody. Injection of 10 mg/kg gliquidone significantly decreased LPS-induced NLRP3 activation in the cortex and hippocampus (Figures 5A–C), whereas 20 mg/kg gliquidone effectively diminished LPS-induced NLRP3 levels in the cortex but not the hippocampus (Figures 5A–C). These data indicate that gliquidone decreases LPS-stimulated NLRP3 inflammasome formation in the mouse brain.

To investigate whether gliquidone alters peripheral proinflammatory cytokine levels in vivo, wild-type mice were treated as described above, followed by measurement of proinflammatory cytokine levels in peripheral blood by ELISA. Treatment of wild-type mice with 10 or 20 mg/kg gliquidone significantly abolished the LPS-induced increases in serum IL-1β and IL-6 levels but not TNF-α (Supplementary Figures 1A–C). These data indicate that gliquidone selectively alters peripheral proinflammatory cytokine levels in LPS-treated wild-type mice.

Since gliquidone regulates micro/astroglial activation, proinflammatory cytokine levels, and NLRP3 inflammasome activation in LPS-treated wild-type mice, we next verified the effects of gliquidone on LPS-mediated proinflammatory cytokine levels in vitro. We first assessed gliquidone cytotoxicity and induction of mitochondrial arrest using CCK or MTT assays in BV2 microglial cells. No cytotoxicity was observed in BV2 microglial cells after treatment with up to 50 μM gliquidone for 6 h (Figures 6A,B). In BV2 microglial cells treated for 24 h, gliquidone had no cytotoxicity up to 25 μM, but 50 μM gliquidone significantly diminished cell viability (Figure 6C). Based on these data, a gliquidone concentration of 5 μM, 10-fold lower than the concentration producing mitochondrial dysfunction, was used in subsequent experiments.

Gliquidone is a ligand of SUR1, which is part of the Kir6 channel (Wu et al., 2018), and thus we assessed the effects of gliquidone on SUR1 mRNA levels as well as proinflammatory cytokine levels by q-PCR. Treatment of BV2 microglial cells with gliquidone alone for 6 or 24 h did not alter SUR1 mRNA levels (Figure 6D). In addition, gliquidone alone reduced COX-2 mRNA levels but did not affect IL-1β, iNOS, and IL-6 mRNA levels in BV2 microglial cells (Figure 6E). These results indicate that 5 μM gliquidone itself selectively affects mRNA levels of the proinflammatory cytokine COX-2.

We then examined the effects of gliquidone on LPS-stimulated microglial SUR1 and proinflammatory cytokine mRNA levels in BV2 microglial cells. Cells were pretreated for 30 min with 200 ng/ml LPS or PBS and treated for 5.5 or 23.5 h with 5 μM gliquidone or vehicle (1% DMSO). Treatment with 5 μM gliquidone for 23.5 h but not 5.5 h significantly abolished the LPS-induced increase in SUR1 mRNA levels in BV2 microglial cells (Figure 6F). In addition, RT-PCR analysis showed that gliquidone post-treatment significantly reduced the LPS-induced increases in IL-1β, IL-6, iNOS, and COX-2 mRNA levels (Figures 6G,H). Confirming these results, q-PCR showed significant reductions in LPS-mediated IL-1β, IL-6, and COX-2 mRNA levels, with a trend toward decreased iNOS mRNA levels (Figure 6I). Moreover, we found that pretreatment with 5 μM gliquidone for 30 min significantly reduced the increases in IL-1β, IL-6, iNOS, and COX-2 mRNA levels in BV2 microglial cells induced by treatment with 200 ng/ml LPS for 5.5 h (Supplementary Figures 2A–C). Taken together, these results show that post- and pretreatment with gliquidone modulates LPS-induced microglial proinflammatory responses in vitro.

In microglial cells, LPS stimulates TLR4 and ATP-sensitive K⁺ channels via ERK/STAT3 or NF-κB signaling (Rodriguez-Gomez et al., 2020). To examine whether gliquidone alone regulates ERK signaling mediated by ATP-sensitive K⁺ channels, BV2 microglial cells were treated sequentially with PBS for 45 min and 5 μM gliquidone or vehicle (1% DMSO) for 45 min. Subsequently, Western blotting was conducted with anti-p-ERK, anti-ERK, and anti-β-actin antibodies. We found that gliquidone alone did not alter p-ERK and total ERK levels in BV2 microglial cells (Figures 7A,B).

To examine whether gliquidone affects LPS-stimulated microglial ERK signaling, BV2 microglial cells treated with 200 ng/ml LPS or PBS for 45 min were post-treated with 5 μM gliquidone or vehicle (1% DMSO) for 45 min. Subsequently, Western blotting was conducted with anti-p-ERK, anti-ERK, and anti-β-actin antibodies. Gliquidone significantly decreased the LPS-induced increase in p-ERK levels but did not alter total ERK or β-actin levels (Figures 7C,D).

To assess the effects of gliquidone itself on nuclear STAT3 or NF-κB phosphorylation, BV2 microglial cells were sequentially treated with PBS for 30 min and 5 μM gliquidone or vehicle (1% DMSO) for 5.5 h, and ICC was performed with anti- p-STAT3^s727 or anti-p-NF-κB antibodies. We found that gliquidone alone did not alter nuclear p-STAT3 and p-NF-κB levels in BV2 microglial cells (Figures 7E–H).

We then tested whether gliquidone alters LPS-mediated nuclear STAT3/NF-κB phosphorylation, BV2 microglial cells were sequentially treated with 200 ng/ml LPS or PBS for 30 min and 5 μM gliquidone or vehicle (1% DMSO) for 5.5 h. ICC assay demonstrated that gliquidone significantly reduced the LPS-mediated increases in nuclear STAT3 and NF-κB phosphorylation in BV2 microglial cells (Figures 7I–L). These results indicate that gliquidone modulates downstream ERK/STAT3 or NF-κB signaling regulated by TLR4 and ATP-sensitive K⁺ channels in microglial cells in response to neuroinflammation.

Since gliquidone downregulated LPS-evoked NLRP3 inflammasome formation in wild-type mice, we investigated whether gliquidone itself affects NLRP3 inflammasome activation. Treatment of BV2 microglial cells with PBS for 30 min followed by 5 μM gliquidone for 5.5 or 23.5 h did not alter NLRP3 and pro-IL-1β mRNA levels (Figures 8A–C).

We then investigated the effects of gliquidone on LPS-mediated microglial NLRP3 inflammasome activation. After treatment for 30 min with 200 ng/ml LPS or PBS, BV2 microglial cells were treated with 5 μM gliquidone or vehicle (1% DMSO) for 5.5 h or 23.5 h, and NLRP3/pro-IL-1β mRNA levels were detected by q-PCR. Post-treatment with gliquidone for 5.5 h significantly reduced LPS-enhanced pro-IL-1β mRNA levels, and a trend toward decreased NLRP3 mRNA levels was observed (Figures 8D,E). In addition, LPS-induced NLRP3 mRNA levels were markedly decreased by post-treatment with gliquidone for 23.5 h (Figure 8F). ICC analysis showed that gliquidone significantly downregulated LPS-induced microglial NLRP3 protein levels (Figures 8G,H). These data imply that gliquidone regulates the LPS-induced neuroinflammatory responses by decreasing NLRP3 inflammasome formation in BV2 microglial cells.

Since gliquidone downregulates LPS-mediated microglial neuroinflammatory responses in BV2 microglial cells, we next examined the effects of gliquidone on LPS-induced astrocytic neuroinflammatory responses. Mouse primary astrocytes were treated with 200 ng/ml LPS or PBS for 0.5 h and post-treated with 5 μM gliquidone or vehicle (1% DMSO) for 5.5 or 23.5 h. q-PCR analysis showed that gliquidone post-treatment for 5.5 or 23.5 h did not alter SUR1 mRNA levels (Figures 9A,B). In addition, gliquidone post-treatment significantly decreased LPS-stimulated IL-1β and IL-6 mRNA levels but did not alter iNOS and COX-2 mRNA levels (Figures 9C–F). These results suggest that gliquidone modulates proinflammatory cytokines less effectively in astrocytes than in microglia.

Given the significant reduction of downstream ERK signaling in BV2 microglial cells treated with gliquidone (Figure 7), we then investigated the effects of gliquidone on LPS-mediated astrocytic ERK signaling. Initially, we evaluated the purity of the primary astrocytes culture by measuring the GFAP/DAPI ratio, indicating an astrocyte purity of greater than 80% (Supplementary Figures 3A,B). Subsequently, mouse primary astrocytes were treated with 200 ng/ml LPS or PBS for 0.5 h followed by 5 μM gliquidone or vehicle (1% DMSO) for 5.5 h, and ICC was performed with an anti-p-ERK antibody. Gliquidone treatment did not affect LPS-evoked ERK phosphorylation in primary astrocytes (Supplementary Figures 3C,D).

Next, we examined the effects of gliquidone on LPS-induced astrocytic STAT3 or NF-κB phosphorylation in the nucleus. Again, we first confirmed that our primary astrocytes had a purity of greater than 80% (Figures 10A,B). Subsequent ICC analysis of cells treated as described above with anti-CD11b and anti-p-STAT3^S727 or anti-CD11b and p-NF-κB^S536 antibodies indicated that gliquidone significantly reduced the LPS-stimulated increases in nuclear STAT3 and NF-κB phosphorylation in primary astrocytes (Figures 10C–F). These results indicate that gliquidone modulates LPS-mediated astrocytic STAT3 or NF-κB signaling in a p-ERK-independent manner in response to neuroinflammation.

Since gliquidone downregulated LPS-evoked proinflammatory cytokine levels and nuclear STAT3 and NF-κB phosphorylation in primary astrocytes, we further examined the effects of gliquidone on LPS-induced astrocytic NLRP3 inflammasome activation. Mouse primary astrocytes were treated with 200 ng/ml LPS or PBS for 0.5 h and post-treated with 5 μM gliquidone or vehicle (1% DMSO) for 5.5 h or 23.5 h, and NLRP3 and pro-IL-1β were detected by q-PCR or ICC. Gliquidone post-treatment for 5.5 h significantly diminished the LPS-induced increase in pro-IL-1β mRNA levels, whereas no changes in NLRP3 mRNA and protein levels were observed after gliquidone post-treatment for 5.5 or 23.5 h (Supplementary Figures 4A–D). These results indicate that gliquidone regulates LPS-mediated astrocytic proinflammatory cytokine levels by inhibiting STAT3/NF-κB signaling but does not modulate ERK signaling and NLRP3 inflammasome activation in primary astrocytes in vitro.