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 nos. IACUC-19-00049, IACUC-22-00044, and IACUC-24-00004).

Selexipag was purchased from Selleckchem (Cat No. S3726, Houston, TX, USA). Lower doses (0.3 μM and 3.0 μM) of selexipag significantly decrease the expression of the proinflammatory mediator S100A4 in human systemic sclerosis (SSC) skin fibroblasts, whereas a higher dose (30 μM) has no effect on proinflammatory responses (16). Therefore, we chose selexipag doses of 0.5 μM, 1 μM, or 5 μM (in 1% DMSO) to investigate whether selexipag treatment modulates LPS-evoked inflammatory responses in BV2 or primary microglial cells. In addition, selexipag doses of 0.5 mg and 1.0 mg/kg (p.o., twice daily for 3 days) effectively alleviate LPS-mediated lung inflammation in mice, with superior regulatory effects of 1.0 mg/kg compared to 0.5 mg/kg (17). Therefore, we selected a dose of 1.0 mg/kg (dissolved in 1% DMSO in saline) as an optimal concentration for assessing the effects of selexipag treatment on LPS-mediated central neuroinflammation in C57BL/6N mice.

BV2 microglial cells (Elabioscience Biotechnology Inc, Houston, TX, USA) were maintained as previously described (18). To examine the effect of selexipag on LPS-induced neuroinflammatory responses in vitro, BV2 microglial cells (2.0x10⁵/well) were treated with selexipag (0.5, 1 or 5 µM) or vehicle (1% DMSO) for 30 min and then challenged with LPS (200 ng/ml, L2630, Escherichia coli, Sigma-Aldrich, St. Louis, MO, USA) or PBS for 2.5 h, 5.5 h or 23.5 h.

The 3-(4,5-dimethylthiazol-2-yl)-2,5-dipenyltetrazolium bromide (MTT; VWR Chemicals, Solon, OH, USA) assay was performed to evaluate the cytotoxicity of selexipag in vitro. BV2 microglial cells were incubated with vehicle (0.002, 0.01, 0.02, 0.05, or 0.1% DMSO) or selexipag (1, 5, 10, 25, or 50 µM) for 24 h. Next, the treated cells were incubated for 3 h with 0.4 mg/ml MTT solution. DMSO was added to dissolve the formazan crystal product, and 30 min later, the absorbance at 570 nm was measured in a SPECTROstar Nano microplate reader (BMG, Labtech, Ortenberg, Germany).

To examine the effect of selexipag treatment on LPS-induced proinflammatory responses at the protein level, BV2 microglial cells were treated with selexipag (0.5 µM) or vehicle (1% DMSO) for 30 min and then challenged with LPS (200 ng/ml) or PBS for 5.5 h. Next, the cells were lysed on ice for 5 min in lysis buffer (50 mM Tris, pH 7.4, 1% Triton X-100, 2 mM CaCl_2, 2 mM MgCl₂) supplemented with protease inhibitor. The lysate was centrifuged at 12,000 rpm for 15 min, and protein assay reagents (Bio-Rad Laboratories, Hercules, CA, USA) were used to determine the protein concentration in the supernatant. COX-2, IL-1β, IL-6, and TNF-α protein levels were measured using a COX-2 ELISA kit (DYC4198-5; R&D Systems, Minneapolis, MN, USA) and IL-1β, IL-6, and TNF-α ELISA kits (88-7013A-88 for IL-1β, 88-7064–22 for IL-6, 88-7324–22 for TNF-α; Invitrogen, Waltham, Massachusetts, USA) according to the manufacturer’s instructions.

Primary glial cells were isolated from postnatal day 1 or 2 C57BL/6N mice according to an established protocol with minor modifications (19). Briefly, whole mouse brains were collected, mechanically dissociated, and filtered through 70-μm nylon mesh. The resulting mixed glial cultures (MGCs) were incubated in low-glucose DMEM (1000 mg/L glucose) supplemented with 10% FBS, 100 U/mL penicillin, and 100 μg/mL streptomycin. The MGCs were composed mainly of astrocytes and microglia and were maintained in a 5% CO₂ incubator at 37 °C for 3 weeks. On day 21, the MGCs were subjected to mild trypsin digestion (0.5 M EDTA, 1 M CaCl₂, and 0.25% trypsin–EDTA in low-glucose DMEM with 1% penicillin and streptomycin) for 40 min in a 5% CO₂ incubator. After washing to remove the astrocytic layer, primary microglia were detached using trypsin–EDTA (0.25%) and centrifuged twice at 2000 rpm for 10 min before subsequent experiments. To investigate the effect of selexipag on LPS-induced neuroinflammatory responses, primary microglial cells (2.0x10⁵/well) were treated with selexipag (5 µM) or vehicle (1% DMSO) for 30 min and then stimulated with LPS (200 ng/ml) or PBS for 5.5 h.

Eight-week-old C57BL/6N male mice (22–24 g; Koatech, Gyeonggi-do, Korea) were maintained in a pathogen-free facility with a 12-hr photoperiod in cages housing 4 mice each and access to food and water ad libitum. All mice were randomly assigned to treatment groups. To determine the effects of selexipag on acutely stimulated LPS-induced neuroinflammatory responses in vivo, we chose a high dose of LPS (10 mg/kg). In addition, since administration of 10 mg/kg LPS reduced the survival rate of mice 20 h after injection, we decided to euthanize the mice 8 h after LPS administration. Our previous studies confirmed that administration of 10 mg/kg LPS induces neuroinflammatory responses in the brains of C57BL/6N mice 8 h after injection (20–23).

We conducted two independent experiments to investigate the effect of selexipag in vivo. In the first experiment, C57BL/6N mice were injected (i.p.) with selexipag (1 mg/kg) or vehicle (1% DMSO) daily for 7 days. On day 7, the mice were treated (i.p.) with LPS (10 mg/kg) or PBS 30 min after the selexipag or vehicle (1% DMSO) injection. Eight hours after the LPS or PBS administration, the mice were sacrificed, and the brain was prepared for immunofluorescence (IF) staining. In the second experiment, C57BL/6N mice were treated as described above, and the bilateral cortical and the hippocampal regions were dissected for real-time PCR, western blotting, or ELISA. The dissected cortical and hippocampal tissues were stored at -80 °C until analysis.

To determine whether selexipag modulates neuroinflammatory responses in vivo, C57BL/6N mice were injected (i.p.) with selexipag (1 mg/kg) or vehicle (1% DMSO) daily for 7 days and sequentially treated with LPS (Sigma Aldrich, St. Louis, MO, USA, Escherichia coli,10 mg/kg, i.p.) or PBS for 8 hr. Subsequently, the mice were anesthetized with 2,2,2-tribromoethanol (2.5% v/v, 150 mg/kg, i.p., Sigma Aldrich, St. Louis, MO, USA) and transcardially perfused with PBS followed by 4% paraformaldehyde (PFA, Chembio, Seoul, Republic of Korea). The brains were removed, post-fixed in 4% PFA at 4 °C overnight and then cryoprotected in 30% sucrose at 4 °C for 2 days. Coronal brain sections (35 μm thick) were prepared using a Leica CM1850 cryostat (Wetzlar, Germany) and permeabilized for 1 h at room temperature in PBS containing 0.2% Triton X-100 (PBST) and 10% normal goat serum. The sections were immunostained for 24–72 h at 4 °C with anti-Iba-1, anti-GFAP, anti-COX-2, anti-IL-1β, or anti-TNF-α antibodies diluted in PBST, washed three times with PBST, and incubated with the secondary antibody at room temperature for 2 h. After a final three washes with PBS, the sections were mounted in DAPI-containing antifade mounting medium (Vector Laboratories, Burlingame, CA, USA). Images were acquired on a DMi8 fluorescence microscope (Leica Microsystems, Wetzlar, Germany) and analyzed by ImageJ (version 1.53e, National Institutes of Health, Bethesda, MD, USA). Supplementary Table 1 provides the details of the primary and secondary antibodies. IF intensity was quantified using the results from 3–4 slices per mouse in each group (n = 5 mice per group x 3–4 slices, n = 19-20).

To measure the intensity of IF staining for Iba-1, GFAP, COX-2, IL-1β, and TNF-α in the cortex and/or hippocampal CA1, CA2, CA3, CA4 and/or DG regions of C57BL/6N mice, regions of interest (ROIs) were first selected in DAPI-stained images. Then, the intensity of IF staining for markers in each ROI was measured semi-automatically using the software ImageJ (version 1.53e, US National Institutes of Health, Bethesda, MD, USA). To quantify Iba-1- and GFAP-labeled regions of the brain, the cortical and hippocampal CA1 and DG regions were selected as the ROIs in DAPI-stained images, and the threshold for immune-reactivity was specified. The percentage immune-labeled area and the number of immune-positive cells in the selected regions were then calculated.

After treating BV2 microglial cells, primary microglial cells, or C57BL/6N mice with selexipag or vehicle and LPS as described above, TRIzol (Invitrogen, Waltham, MA, USA) was used to extract total RNA from cells or mouse brain tissue (cortex and hippocampus) according to the manufacturer’s recommendations. cDNA was reverse transcribed from the total RNA (1 μg) using the Superscript cDNA Premix Kit II with oligo (dT) primers (GeNetBio, Chungman, Korea). The cDNA was used as the template in real-time qPCR as described previously (20, 22). The sequences of the primers are provided in Supplementary Table 2. The GAPDH value was used to normalize cycle threshold (Ct) values, and fold changes were calculated relative to the control (vehicle-treated group).

To investigate the effect of an IP receptor antagonist itself on LPS-mediated proinflammatory mediator levels, BV2 microglial cells were treated with the IP receptor-selective antagonist BAY 73-1449 (0.01, 0.025 nM; MCE, Monmouth Junction, NJ, USA) or vehicle (1% DMSO) for 30 min followed by LPS (200 ng/ml) or PBS for 5.5 h, and real-time PCR was performed. Next, to determine if selexipag modulates LPS-induced neuroinflammation in an IP receptor (on-target)-dependent manner, BV2 microglial cells were treated first with BAY 73-1449 (1 μM (for ELISA) or 0.01 nM (for real-time PCR) MCE, Monmouth Junction, NJ, USA) or vehicle (1% DMSO) for 30 min, then with selexipag (5 µM) or vehicle (1% DMSO) treatment for 30 min, and finally with LPS (200 ng/ml) or PBS for 5.5 h. After confirming suppression of cAMP levels (IP receptor downstream signaling) by ELISA, proinflammatory mediator mRNA levels were measured by real-time PCR.

To assess whether the anti-inflammatory effects of selexipag treatment against LPS treatment are mediated by suppression of P38 activity, BV2 microglial cells were treated first with the P38 antagonist SB305-580 (10 μM, Calbiochem, Sandiego, CA, USA) or vehicle (1% DMSO) for 30 min, then with selexipag (5 µM) or vehicle (1% DMSO) treatment for 30 min, and finally with LPS (200 ng/ml) or PBS for 5.5 h. Then, proinflammatory mediator mRNA levels were measured by real-time PCR.

To investigate whether selexipag treatment downregulates LPS-stimulated proinflammatory responses via c-Jun, BV2 microglial cells were treated first with the c-Jun-selective antagonist c-Jun peptide (100 μM, MCE, Monmouth Junction, NJ, USA) or vehicle (D.W.) for 30 min, then with selexipag (5 µM) or vehicle (1% DMSO) treatment for 30 min, and finally with LPS (200 ng/ml) or PBS for 5.5 h. Proinflammatory mediator mRNA levels were measured by real-time PCR.

To examine intracellular cAMP levels, BV2 microglial cells treated as described above were lysed in 0.1 M HCl for 10 min at room temperature, and the lysate was centrifuged for 20 min at 4 °C and 12000 rpm. In addition, the cortical and hippocampus regions from C57BL/6N mice treated as described above were dissected and homogenized in RIPA lysis buffer (Merck Millipore, Billerica, MA, USA) containing protease and phosphatase inhibitor cocktail (1%; Thermo Scientific, Waltham, MA, USA). After incubation on ice for 1 h at 4 °C, the lysate was centrifuged for 20 min at 4 °C and 12000 rpm. For both lysates, the supernatant was collected, and cAMP levels were measured by ELISA (cAMP complete ELISA kit, ADI-901-163, Enzo Life Sciences, Farmingdale, NY, U.S.A.) according to the manufacturer’s instructions.

To assess the effect of selexipag treatment on LPS-induced NLRP3, p-P38/P38, p-ERK/ERK, p-c-Jun^Ser73/c-Jun and/or p-STAT3^Y705/STAT3 levels, BV2 microglial cells treated as described above were harvested in lysis buffer (50 mM Tris, pH 7.4, 1% Triton X-100, 2 mM CaCl₂ and 2 mM MgCl₂) with 1% protease inhibitor and phosphatase inhibitors (Thermo Scientific, Waltham, MA, USA), and the lysate was centrifuged for 20 min at 4 °C and 12000 rpm. In addition, the bilateral cortex and hippocampus from C57BL/6N mice treated as described above were dissected and homogenized in RIPA lysis buffer (Merck Millipore, Billerica, MA, USA) supplemented with protease and phosphatase inhibitor cocktail (1%; Thermo Scientific, Waltham, MA, USA). After incubation on ice for 1 h at 4 °C, the lysate was centrifuged for 20 min at 4 °C and 12000 rpm. For both lysates, the supernatant was collected, and the protein concentration was quantified by the BCA method relative to a standard solution of BSA. Next, 10 µg of protein sample was loaded on an 8% SDS-PAGE gel, separated by electrophoresis, and transferred to a polyvinylidene difluoride (PVDF) membrane. After blocking with 5% skim milk at room temperature for 1 h, the membrane was incubated with primary antibody overnight at 4 °C. The next day, the membrane was incubated with the corresponding secondary antibody for 1 h, and then detection was realized by adding ECL Western Blotting Detection Reagent (GE Healthcare, Chicago, IL, USA). Images were acquired and analyzed by Fusion Capt Advance software (Vilber Lourmat). The primary and secondary antibodies used for western blotting are listed in Supplementary Table 3.

To investigate whether the effects of selexipag treatment on LPS-induced neuroinflammatory responses are dependent on in its neuroinflammation-associated molecular target, NLRP3, BV2 microglial cells were transfected with small interfering RNA (siRNA) targeting mouse NLRP3 (Vector Biolabs, Malvern, PA, USA) as previously described with minor modifications (20). Briefly, NLRP3 siRNA or scramble siRNA was diluted in Opti-MEM medium (Thermo Scientific, Waltham, MA, USA) to a concentration of 400 nM, and Lipofectamine^® RNAiMAX reagent (Thermo Scientific, Waltham, MA, USA) was added. Transfection of BV2 microglial cells with the siRNA complex suspension (final siRNA concentration at 40 nM) was performed as previously described (18). Forty-one hour after transfection, the BV2 microglial cells were incubated in serum-free media for 1 h and then treated with selexipag (5 μM) or vehicle (1% DMSO) for 30 min followed by LPS (200 ng/ml) or PBS for 5.5 h. Finally, the BV2 microglial cells were harvested, and NLRP3 knockdown efficiency was analyzed by real-time PCR. After validating NLRP3 siRNA transfection, proinflammatory mediator mRNA levels were measured by real-time PCR.

GraphPad Prism 10 software (GraphPad Software, San Diego, CA, USA) was used for graph generation and statistical analysis. Data are presented as individual data points and means ± standard errors of measurement (SEMs). For comparisons between two groups, Student’s t-test was used. For multiple comparisons, one-way analysis of variance (ANOVA) with Tukey’s post hoc or Newman-Keuls analysis was used. Asterisks indicate significance: *p < 0.05, **p < 0.01, and ***p < 0.001. A detailed statistical analysis is provided in Supplementary Table 4.

Before examining the effects of selexipag treatment on LPS-induced proinflammatory responses, we first evaluated its cytotoxicity in vitro. For this experiments, BV2 microglial cells were treated with vehicle (0.002, 0.01, 0.02, 0.05, or 0.1% DMSO) or selexipag (1, 5, 10, 25, or 50 µM) for 24 h, and cell viability was measured using the MTT assay. We found that selexipag treatment had no cytotoxic effects in BV2 microglial cells, even at the highest concentration (50 µM), compared with vehicle treatment (Figure 1A).

To investigate whether selexipag treatment affects LPS-evoked proinflammatory mediator levels in a dose-dependent manner in vitro, BV2 microglial cells were treated first with 0.5 µM, 1 µM or 5 µM selexipag or vehicle (1% DMSO) for 30 min and then with 200 ng/ml LPS or PBS for 5.5 h. Next, proinflammatory mediator mRNA levels were quantified by real-time PCR. All three doses of selexipag significantly reduced the LPS-induced proinflammatory mediators IL-1β, IL-6, COX-2, and TNF-α mRNA levels (Figures 1B–D).

Next, we performed ELISA to examine whether selexipag treatment alters the protein levels of LPS-induced proinflammatory mediators. To test this, BV2 microglial cells were treated with 0.5 µM selexipag or vehicle (1% DMSO) for 30 min, followed by 200 ng/ml LPS or PBS for 5.5 h, and the protein levels of IL-1β, IL-6, COX-2, and TNF-α were quantified by ELISA. Treatment with 0.5 µM selexipag significantly reduced the LPS-induced protein levels of the proinflammatory mediators IL-1β, IL-6, COX-2, and TNF-α in BV2 microglial cells (Figure 1E).

We then investigated the effects of selexipag on LPS-mediated proinflammatory responses in a time dependent manner. For these experiments, BV2 microglial cells were treated with 0.5 µM selexipag or vehicle (1% DMSO) for 30 min, followed by 200 ng/ml LPS or PBS for 2.5 h or 23.5 h. Quantification of IL-1β, IL-6, COX-2, and TNF-α mRNA levels by real-time PCR showed that treatment with We found that selexipag (0.5 µM) did not affect LPS-induced proinflammatory mediators mRNA expression when the total treatment time was for 3 h (Figure 1F), whereas the LPS-induced mRNA levels of the proinflammatory mediators IL-1β, IL-6, COX-2, and TNF-α were significantly reduced when the total exposure time to selexipag was 24 h (Figure 1G). These results suggest that selexipag attenuates LPS-induced proinflammatory mediator expression in BV2 microglial cells in a time-dependent manner.

To examine whether selexipag treatment modulates LPS-induced neuroinflammatory responses in primary microglial cells, the cells were treated with 5 µM selexipag or vehicle (1% DMSO) for 30 min, followed by treatment with 200 ng/ml LPS or PBS for 5.5 h. Subsequent real-time PCR analysis showed that selexipag treatment significantly reduced LPS-induced IL-1β, IL-6, COX-2, and TNF-α mRNA levels in primary microglial cells (Figures 1H–K), indicating that selexipag treatment diminishes LPS-mediated proinflammatory responses in primary microglial cells.

Since selexipag treatment downregulated LPS-induced proinflammatory mediator levels in BV2 microglial cells and primary microglial cells, we assessed the effects of selexipag administration on LPS-induced microglial activation in vivo. C57BL/6N mice were injected with 1 mg/kg selexipag (i.p.) or vehicle (1% DMSO) daily for 7 days; on day 7, the mice were injected with 10 mg/kg LPS (i.p.) or PBS after selexipag or vehicle injection. Subsequent IF staining of brain sections with an anti-Iba-1 antibody demonstrated that selexipag treatment reduced the LPS-induced increases in Iba-1 fluorescence intensity in the cortex and hippocampal CA1 and DG regions but not in the hippocampal CA2, CA3 and CA4 regions (Figures 2A–C). In addition, selexipag administration significantly reduced the LPS-induced Iba-1-labeled area in the cortex and hippocampal CA1 region, whereas no significant changes were observed in the hippocampal CA2, CA3, CA4 and DG regions (Figures 2A, B, D). However, selexipag treatment did not alter the LPS-induced increase in Iba-1-positive cell numbers in the brain (Figures 2A, B, E). These results demonstrate that selexipag selectively attenuates LPS-induced microglial activation in brain subregions rather than exerting a uniform effect throughout the hippocampus.

We then examined the effects of selexipag treatment on astrogliosis in LPS-treated C57BL/6N mice. IF staining with an anti-GFAP antibody showed that selexipag treatment significantly suppressed LPS-induced GFAP fluorescence intensity in cortical layers III–V and the hippocampal CA1, CA2, and DG regions but not in the hippocampal CA3 and CA4 regions (Figures 3A–C). In addition, selexipag treatment significantly downregulated the LPS-induced increase in GFAP-labeled area in cortical layers III–V and the hippocampal CA1 and CA2 regions but not in the hippocampal CA3, CA4, and DG regions (Figures 3A, B, D). Moreover, selexipag treatment significantly decreased the number of GFAP-positive cells in the hippocampal DG region but not in cortical layers III–V and the hippocampal CA1–CA4 regions compared with mice treated with vehicle (Figures 3A, B, E). Collectively, these results suggest that selexipag treatment diminishes LPS-induced astrocyte activation, hypertrophy, and migration in the brain in C57BL/6N mice.

Since selexipag treatment abolished LPS-induced microglial and astrocyte activation in C57BL/6N mice, we next investigated the effects of selexipag treatment on LPS-induced proinflammatory mediator levels in vivo. C57BL/6N mice were treated as described above, and the cortex and hippocampus were dissected for real-time PCR analysis. The results revealed that selexipag treatment significantly reduced LPS-induced COX-2 and IL-1β mRNA levels in the hippocampus but not in the cortex (Figures 4A–D).

To validate these findings, IF staining of brain sections with an anti-COX-2 antibody was performed. The results demonstrated that selexipag treatment significantly attenuated LPS-induced COX-2 fluorescence intensity in the hippocampal CA1, CA3, CA4 and DG regions but not in the hippocampal CA2 region (Figures 4E, F). Additionally, selexipag treatment significantly decreased the LPS-induced increase in IL-1β fluorescence intensity in the hippocampus (Figures 4G, H).

We then examined whether selexipag treatment affects the levels of other proinflammatory cytokine in LPS-treated C57BL/6N mice by performing IF staining and real-time PCR. IF staining with an anti-TNF-α antibody showed that selexipag treatment significantly attenuated LPS-induced TNF-α levels only in the hippocampal CA1 region and not in the hippocampal CA2, CA3, CA4, and DG regions or cortex (Figures 5A–C). Furthermore, selexipag administration significantly suppressed LPS-induced TNF-α mRNA levels in the hippocampus but not the cortex (Figure 5D). These data indicate that selexipag treatment attenuates proinflammatory mediator COX-2, IL-1β and TNF-α mRNA and protein expression in LPS-treated C57BL/6N mice.

Since selexipag treatment significantly reduced LPS-induced microglial/astroglial activation and proinflammatory mediator levels in C57BL/6N mice, we examined its effects on the expression of markers of microglia- and astroglia-associated neuroinflammatory dynamics in vivo. To address this, C57BL/6N mice were treated as described above, and the cortex and hippocampus were dissected for real-time PCR analysis. We found that selexipag treatment significantly decreased LPS-induced CXCL10 [a marker of reactive astrocytes (RAs)] mRNA levels in the cortex and hippocampus (Figure 6A). In addition, selexipag treatment significantly downregulated LPS-mediated SERPINA3N [a marker of reactive astrocytes (RAs)] mRNA levels in the cortex but not the hippocampus (Figure 6B). Moreover, selexipag treatment downregulated LPS-induced GBP2 and CHI3L1 (markers of RAs) mRNA levels in the hippocampus but not the cortex (Figures 6C, D).

Next, we assessed the effects of selexipag treatment on microglia-associated neuroinflammatory dynamics in LPS-treated C57BL/6N mice. The results of real-time PCR showed that selexipag treatment significantly reduced LPS-induced CD44 [a marker of disease-associated microglia (DAMs)] mRNA levels in the cortex and hippocampus (Figure 6E). However, selexipag treatment did not alter P2RY12 (a marker of homeostatic microglia) mRNA levels in the cortex and hippocampus (Figure 6F). These data suggest that selexipag treatment selectively reduces LPS-induced microglia- and astrocyte-associated neuroinflammatory dynamics-related mRNA levels in C57BL/6N mice.

To determine the mechanisms by which selexipag alters neuroinflammatory responses in vivo, C57BL/6N mice were treated as described above, and the cortex and hippocampus were dissected for real-time PCR analysis. We found that selexipag treatment significantly downregulated LPS-induced NLRP3 mRNA levels in the cortex and hippocampus (Figure 7A) and reduced LPS-induced pro-IL-1β mRNA levels only in the hippocampus (Figure 7B).

Next, we assessed whether selexipag affects LPS-mediated NLRP3 protein expression and NLRP3 inflammasome activation in vitro. BV2 microglial cells were treated with vehicle (1% DMSO) or selexipag (5 μM) for 30 min, followed by treatment with PBS or LPS (200 ng/ml) for 5.5 or 23.5 h. Analysis by western blotting and real-time PCR showed that selexipag treatment significantly downregulated LPS-induced NLRP3 mRNA and protein levels in BV2 microglial cells (Figures 7C–E). In addition, we found that selexipag-treated BV2 microglial cells significantly decreased LPS-mediated caspase-1 and pro-IL-1β mRNA levels (Figures 7F, G).

We then investigated the effects of selexipag treatment on NLRP3 inflammasome activation in primary microglial cells treated with vehicle (1% DMSO) or selexipag (5 μM) for 30 min followed by LPS (200 ng/ml) or PBS for 5.5 h. We found that selexipag treatment significantly reduced LPS-induced NLRP3, CASPASE-1 and pro-IL-1β mRNA levels in primary microglial cells (Figures 7H–J).

In C. albicans-infected macrophages, the IP receptor antagonist CAY10441 decreases COX-2 protein levels, consistent with COX-2’s role as a principal enzyme in prostacyclin (PGI₂) biosynthesis (24, 25). Since the IP receptor agonist selexipag significantly suppressed LPS-induced proinflammatory mediator levels in vitro and in vivo, we wanted to determine if these effects are modulated via the IP receptor (on-target of selexipag). To address this, we first investigated whether IP receptor antagonist treatment alters LPS-induced proinflammatory mediator levels in BV2 microglial cells. BV2 microglial cells were treated with vehicle (1% DMSO) or the selective IP receptor antagonist BAY 73-1449 (0.01 or 0.025 nM) for 30 min followed by PBS or LPS (200 ng/mL) for 5.5 h, and real-time PCR was conducted. We found that BAY 73–1449 and LPS (BAY 73-1449 + LPS) treatment significantly decreased LPS-induced COX-2 mRNA levels in BV2 microglial cells but not IL-1β mRNA levels (Supplementary Figures 2A, B).

Given that BAY 73–1449 treatment suppresses cAMP signaling, downstream of the IP receptor (26), we tested whether BAY 73–1449 administration modulates intracellular cAMP levels in LPS-treated BV2 microglial cells. To test this, BV2 microglial cells were first treated with vehicle (1% DMSO) or BAY 73-1449 (1 μM) for 30 min, then with vehicle (1% DMSO) or selexipag (5 μM) for 30 min, and finally with PBS or LPS (200 ng/mL) for 5.5 h. We found that IP receptor agonist selexipag and LPS (selexipag + LPS) treatment significantly increased cAMP levels compared to LPS-treated BV2 microglial cells. By contrast, BAY 73–1449 and LPS (BAY 73-1449 + LPS) treatment significantly reduced cAMP levels compared to LPS treatment in BV2 microglial cells, confirming successful blockade of IP receptor signaling by BAY 73–1449 treatment (Supplementary Figure 2C).

Next, to investigate whether selexipag modulates LPS-mediated IL-1β mRNA levels through the IP receptor, BV2 microglial cells were treated first with vehicle (1% DMSO) or BAY 73-1449 (0.01 nM) for 30 min, then with vehicle (1% DMSO) or selexipag (5 μM) for 30 min, and finally with PBS or LPS (200 ng/mL) for 5.5 h. We found that treatment with selexipag and LPS (Selexipag + LPS) significantly reduced LPS-evoked IL-1β mRNA levels in BV2 microglial cells compared with treatment with LPS alone (Supplementary Figure 2D). Importantly, BAY 73-1449, selexipag and LPS (BAY 73-1449 + selexipag + LPS) treatment did not further decrease IL-1β mRNA levels compared to treatment with BAY 73–1449 and LPS (BAY 73-1449 + LPS or treatment with Selexipag and LPS (Selexipag + LPS) in BV2 microglial cells (Supplementary Figure 2D). These findings suggest that selexipag downregulates LPS-induced IL-1β mRNA levels in an IP receptor (on target)-dependent manner.

To determine whether the neuroinflammatory effects of selexipag are NLRP3 dependent, BV2 microglial cells were transfected with either NLRP3 siRNA or scramble siRNA complex for 41 h and then incubated in serum-free media for 1 h. Next, the cells were treated with 5 μM selexipag or vehicle (1% DMSO) for 30 min followed by LPS (200 ng/ml) or PBS for 5.5 h. Real-time PCR analysis showed that NLRP3 siRNA transfection reduced NLRP3 mRNA levels in BV2 microglial cells by 71.69% compared with scramble siRNA transfection (Figure 8A). We then examined whether NLRP3 silencing influences the effect of selexipag on LPS-induced proinflammatory responses in BV2 microglial cells. We found that selexipag treatment significantly reduced LPS-induced COX-2, IL-1β, IL-6 and TNF-α mRNA levels in scramble (control) siRNA-transfected BV2 microglial cells compared to LPS treatment (Figures 8B–E). By contrast, in NLRP3 siRNA-transfected BV2 microglial cells, selexipag treatment did not alter LPS-mediated COX-2, IL-1β, IL-6 and TNF-α mRNA levels compared to LPS treatment (Figures 8B–E). These data suggest that selexipag treatment downregulates LPS-evoked proinflammatory responses in BV2 microglial cells via an NLRP3-dependent mechanism.

To examine the underlying signaling pathway by which selexipag modulates LPS-induced proinflammatory responses, C57BL/6N mice were intraperitoneally (i.p.) administered selexipag (1 mg/kg) or vehicle (1% DMSO) and thereafter injected with PBS or LPS (10 mg/kg, i.p.) as described above. The hippocampus and cortex were subsequently dissected, and cAMP levels were measured by ELISA, which showed that selexipag treatment significantly increased cAMP levels in the hippocampus but not the cortex in LPS-treated C57B/L6 mice (Figures 9A, B).

We then investigated the effect of selexipag treatment on LPS-mediated downstream signaling. BV2 microglial cells were treated with vehicle (1% DMSO) or selexipag (5 μM) for 45 min followed by PBS or LPS (200 ng/ml) for 45 min. Western blotting with anti-p-P38 or anti-p-ERK antibodies revealed that selexipag treatment significantly reduced LPS-mediated P38 phosphorylation in BV2 microglial cells (Figures 9C–E) but not p-ERK levels (Supplementary Figure 1).

Next, we examined whether selexipag modulates LPS-induced proinflammatory responses via P38 signaling. BV2 microglial cells were treated with vehicle (1% DMSO) or the P38 inhibitor SB203-580 (10 μM) for 30 min, followed by vehicle (1% DMSO) or selexipag (5 μM) for 30 min. Finally, the cells were stimulated with PBS or LPS (200 ng/mL) for 5.5 h, and real-time PCR was performed. We found that selexipag or SB203–580 treatment significantly reduced LPS-induced IL-1β, IL-6, COX-2, and TNF-α mRNA levels in BV2 microglial cells, indicating effective inhibition of P38 activity (Figures 9F–I). Importantly, combined SB203-580, selexipag, and LPS (SB203-580 + selexipag + LPS) treatment did not further alter proinflammatory mediator mRNA levels in BV2 microglial cells compared to treatment with SB203–580 and LPS (SB203-580 + LPS) (Figures 9F–I). These findings indicate that the anti-inflammatory effects of selexipag against LPS stimulation are P38 dependent.

To further explore the downstream transcriptional signaling by which selexipag downregulates LPS-induced proinflammatory responses, C57BL/6N mice were administered selexipag (1 mg/kg, i.p.) or vehicle (1% DMSO) and thereafter injected with PBS or LPS (10 mg/kg, i.p.). Eight hours after the LPS/PBS injection, the hippocampus and cortex were subsequently dissected, and western blotting was performed with anti-p-c-Jun^s73 and anti-c-Jun antibodies. We found that selexipag treatment significantly decreased LPS-induced p-c-Jun^S73 expression in the hippocampus but not the cortex in C57BL/6N mice (Supplementary Figure 3). However, selexipag treatment did not affect LPS-induced p-STAT3^Y705 and total STAT3 expression in the hippocampus and cortex (Supplementary Figure 4). These findings indicate that selexipag treatment downregulates c-Jun phosphorylation in LPS-treated C57BL/6N mice.

We further investigated whether selexipag downregulates LPS-induced proinflammatory responses via c-Jun by using a c-Jun antagonist in vitro. To test this, BV2 microglial cells were treated with vehicle (D.W.) or c-Jun peptide (a selective c-Jun antagonist, 100 μM) for 30 min followed by vehicle (1% DMSO) or selexipag (5 μM) for 30 min. Finally, the cells were stimulated with PBS or LPS (200 ng/mL) for 5.5 h, and Western blotting or real-time PCR was conducted. Because c-Jun peptide interferes with the JNK-c-Jun interaction, thereby reducing c-Jun phosphorylation (27), we first investigated the effect of c-Jun peptide treatment on inhibition of c-Jun phosphorylation by performing Western blotting with an anti-p-c-Jun^S73 and anti-PCNA antibodies. We found that treatment with c-Jun peptide and LPS (c-Jun peptide + LPS) significantly reduced c-Jun phosphorylation compared with LPS treatment in BV2 microglial cells, indicating effective inhibition of c-Jun phosphorylation (Supplementary Figures 5A, B).

Next, we examined whether selexipag treatment downregulates LPS-mediated proinflammatory mediator levels via c-Jun signaling. For this experiment, BV2 microglial cells were treated as described above, and real-time PCR was conducted. We found that treatment with selexipag and LPS (selexipag + LPS) significantly reduced IL-1β, IL-6, COX-2, and TNF-α mRNA levels compared with LPS treatment (Supplementary Figures 5C–F). Surprisingly, c-Jun peptide + LPS significantly increased IL-1β and COX-2 mRNA levels but not IL-6 and TNF-α mRNA levels compared to LPS treatment in BV2 microglial cells (Supplementary Figures 5C, E). However, c-Jun peptide, selexipag and LPS (c-Jun peptide + selexipag + LPS) treatment further decreased IL-1β, IL-6, COX-2, and TNF-α mRNA levels compared to c-Jun peptide + LPS in BV2 microglial cells (Supplementary Figures 5C–F). These data indicate that selexipag treatment reduces LPS-induced proinflammatory mediator levels in a c-Jun-independent manner in BV2 microglial cells.

To determine whether selexipag treatment modulates pyroptosis in vivo, C57BL/6N mice were treated as described above, and the cortex and hippocampus were dissected for real-time PCR analysis. Selexipag treatment significantly suppressed LPS-induced GSDMD mRNA levels in the cortex and hippocampus in C57BL/6N mice (Supplementary Figure 6A). However, selexipag administration did not alter the mRNA levels of other pyroptosis-related genes, including NLRP6, CASPASE-1, ASC, IL-18, and HMGB1, in the cortex and hippocampus of LPS-treated C57BL/6N mice (Supplementary Figures 6B–F). These data suggest that selexipag selectively downregulates pyroptosis-associated GSDMD mRNA levels in LPS-treated C57BL/6N mice.