DMF, adenosine 5’-triphosphate disodium hydrate, LPS (LPS 055: B5), ML385, Fetal bovine serum (FBS), RPMI 1640 cell culture media, L-Glutamine, penicillin/streptomycin, phosphate-buffered saline (PBS), and trypsin/EDTA were purchased from Sigma-Aldrich (St. Louis, USA). Ultra-Pure LPS (LPS 0111: B4) were purchased from In vivoGen (San Diego, USA). Supplementary Table 1 contains all antibodies used in this research.

Murine N9 microglia were provided by Dr. Paola Ricciardi-Castagnoli (Cellular Pharmacology Center, Milan, Italy) (29). Cells were sustained in RPMI 1640 containing 10% FBS and 2 mM L-Glutamine, 100 U/ml penicillin, and 100 µg/ml streptomycin at 37°C with 5% CO₂. Treatments were introduced as follows unless a different experimental setup is implemented; DMF (10 µM) for 1 hour, LPS (1 μg/ml) for 4 hours, and ATP (5 mM) for 1 hour.

The cytotoxicity of N9 microglia was determined by checking LDH release with Cytotoxicity Detection Kit^PLUS (Roche, Germany) by following the manufacturer’s protocol. Microplate reader Varioskan (Thermo, USA) was used to obtain the colorimetric change of cells at 492 nm with a reference wavelength of 630 nm. Cytotoxicity percentages were calculated as follows: Cytotoxicity = (OD Sample – OD Low Control) (OD Maximal Release –OD Low Control) x 100

According to the manufacturer’s protocol, N9 microglia viability was investigated with Cell Counting Kit-8 (Sigma Aldrich, St. Louis, USA). Microplate reader Varioskan (Thermo, USA) was used to determine the colorimetric change of cells at 450 nm with a reference wavelength of 630 nm. Cell viability percentages was calculated as the percentage of untreated cells.

Secreted IL-1β and IL-18 levels were determined by IL-1β, and IL-18 sandwich enzyme-linked immunosorbent assay (ELISA) kits (Invitrogen, USA) according to the manufacturer’s instructions.

Total RNA was extracted by using the Nucleospin RNA II Kit (Macherey-Nagel, Germany) according to the manufacturer’s protocol. Total RNA concentration was measured by the Nanodrop spectrophotometer (Thermo Fischer, USA), and 1 µg of RNA was used for reverse transcription with cDNA Revert-Aid First Strand cDNA Synthesis Kit (Thermo Scientific, USA) by following the manufacturer’s protocol. GoTaq^® qPCR Master Mix (Promega, USA) was used for performing quantitative real-time PCR with LightCycler ^® 480 Instrument II (Roche Life Science, USA) according to the manufacturer’s protocol.

For miRNA studies, RNA was isolated by the miRNeasy mini kit (Qiagen, Germany). miScript II RT Kit and QuantiTect SYBR Green PCR Kits (Qiagen, Germany) were used for cDNA synthesis and qPCR, respectively. All the primers utilized in experiments are listed in Supplementary Table 2 . Relative expression levels were calculated with the 2^−ΔΔCt method.

Protein extraction and western blot analysis were carried out as described before (30). Briefly, RIPA lysis buffer (50mM Tris-HCL, pH 7.4, 150mM NaCl, 0.25% deoxycholic acid, 1% Nonidet P-40, 1mM EDTA) supplemented with protease and phosphatase inhibitor (Thermo Scientific, Massachusetts, USA) was used for total protein isolation from both cell lysate and supernatant. NE-PER, Nuclear and Cytoplasmic Extraction Reagents (Thermo Scientific, USA) were used to isolate nuclear and cytosolic fractions. For Western Blot, equal amounts of protein samples were run on a 8-15% SDS–PAGE gel depending on the protein of interest and then was transferred to the PVDF membrane. Depending on the antibodies used, the PVDF membrane was blocked with 5% BSA or 5% dry milk in TBS-T. The membrane was incubated with a primary antibody overnight at 4°C and a secondary antibody for 1 hour at room temperature. At the end of incubations, antigen-antibody complexes were screened with Luminata Forte Western HRP substrate (Merck Millipore, USA) using a densitometer (Vilber Lourmat Gel Imager System, CA). ImageJ software (National Institutes of Health, USA) was used for band density analysis (27). Results were normalized to β-actin and Lamin A/C for quantification or given as arbitrary unit (a.u.) for secreted proteins present in the supernatant.

Luminometric Caspase-Glo^® 1 Inflammasome Assay (Promega, USA) was used for measuring caspase-1 activity by following the manufacturer’s protocol. Centro XS3 lb 960 microplate luminometer (Berthold Technologies, Germany) was used to measure the luminescence of cells. Caspase-1 percentages were calculated as the percentage of untreated cells.

Cells were treated with 50 μg/ml propidium iodide (PI) stain (Thermo Scientific, USA) in the last 15 minutes of ATP treatment. Fluorescent images were captured using the fluorescent microscopy system (Olympus IX-71, Japan). PI-positive and negative cells were counted by ImageJ software (National Institutes of Health, USA) (31). PI-positive cell percentages were calculated by comparing them to untreated cells.

Mitochondrial ROS (mtROS) was determined with the MitoSOX reagent. After treatment, microglia were treated with 5 μM MitoSOX (Invitrogen, USA) for 15 minutes at 37°C, and fluorescence absorbance values were measured at 530 nm with a reference wavelength of 590 nm with the microplate reader Varioskan (Thermo scientific, Massachusetts, USA). For immunofluorescence assay, cells were treated with 5 μM MitoSOX for 15 minutes at 37°C and 0.1 μM DAPI for 2 minutes at 37°C. Images were obtained under an inverted fluorescent microscope, Olympus IX-71 (Olympus, Japan).

Total intracellular ROS was investigated with CM-H2DCFDA (Invitrogen, USA) according to the manufacturer’s instructions. The fluorometric measurements were done at 495 nm with a reference wavelength of 527 nm with a Varioskan Flash (Thermo Scientific, USA) microplate reader.

JC-1 (Thermo, USA) stain was used for investigating mitochondria membrane potential measurement of microglia according to the manufacturer’s instructions. FACS Canto II analyzer (Becton Dickinson, USA) using a 488 nm laser (Becton Dickinson, USA) was used for assessing the mitochondria membrane potential.

All the animal experiments and animal care were strictly performed according to the Izmir International Biomedicine and Genome Institute Local Ethic Committee for Animal Experiments (IBG-AELEC), protocol number: 03/2017. For in vivo study, male BALB/c, aged 12–14 weeks, were used. All the animals were housed and maintained in IBG-Vivarium under controlled conditions (22 ± 2°C; 12 hours light/dark periods) with access to food and water ad libitum. In the experimental setup, mice were grouped randomly into four different groups: Control, LPS, DMF + LPS, and ML385 + DMF + LPS (n= 7 mice/group). Before the experimental procedures, all mice were weighed and controlled. Mice were intraperitoneally injected with ML385 (30 mg/kg) 24 hours before LPS injection, DMF (30 mg/kg) 1 hour before LPS injection, and LPS (5 mg/kg). After 24 hours from LPS injection, all animals proceeded with the clinical scoring, weighing, and behavioral tests. Animals were sacrificed by decapitation, and their brains were collected.

To assess DMF’s effects on sickness behaviors, the following evaluations and behavioral tests were carried out.

Microglia isolation was performed as previously described (33). Briefly, animals were anesthetized and transcardially perfused with ice-cold PBS using a peristaltic pump. After removing meninges, brains were dissected and collected in cold buffer HBSS (Hanks’ balanced salt solution, Sigma, USA) containing 1% BSA and 1mM EDTA. Brains were chopped in a petri dish containing HBSS with the help of a razor blade and then filtered through a 70 µm cell strainer into a 50 ml falcon tube. Collected samples were centrifuged for 12 minutes at 300 x g and 10°C. The pellets were resuspended in 5 ml cold 37% Percoll in PBS, underlaid with 4 ml 70% Percoll and overlayed with 4 ml 30% Percoll in a 15 ml Falcon Tube. These falcons were centrifuged for 40 minutes at 600 x g and 10°C with 0 acceleration and 0 brakes. Finally, the cell layer was collected from the 70% and 37% Percoll interface, transferred into an equal volume of PBS containing 1% FBS, and centrifuged for 5 minutes, 400 x g, 4°C. The supernatant was discarded carefully, and 400 µl PBS was added to the tube, and cells were used in further experiments.

For immunostaining (ASC speck, NF-κB, and Nrf2 staining), cells were fixed with 4% paraformaldehyde (PFA) at 37°C for 15 minutes, then washed with PBS. Permeabilization and blocking were done using PBS containing 10% Goat Serum and 0.5% Triton-X-100 at 37°C for 30 minutes. Cells were incubated with primary antibody overnight, and the following day, the secondary antibody was incubated for 1 hour. Images were acquired with LSM 880 Confocal microscopy (Zeiss, Germany) or fluorescent microscope Olympus BX-61 (Olympus, Japan).

For primary microglia staining, isolated cells were centrifuged for 4 minutes, 1600 RPM at RT, fixed with 4% PFA for 15 minutes at 37°C, then washed 1-2 times with PBS, and permeabilized with PBS containing %0,1 Triton-x100 for 10 minutes at RT. After washing twice, cells were blocked with PBS containing 5% donkey serum for 30 minutes at RT and incubated for an overnight hour at 4°C with primary antibody IBA1 and NLRP3/Caspase-1 diluted in PBS. On the next day, the cells were rewashed with PBS twice and incubated with secondary antibody diluted in PBS for 1 hour at RT. In the last 20 minutes of incubation, Hoechst was added for counterstaining. In the end, the cells were washed PBS twice, and slides were mounted. Visualization of the slides was done using a fluorescent microscope Olympus BX-61 (Olympus, Japan).

All images were analyzed with ImageJ software (National Institutes of Health, USA). In the analysis of ASC speck images, ASC speck positive cells were counted, and data were given as the percentage of total cells. Image analysis was conducted bling to avoid experimental bias.

Graphpad Prism 8.0 (Graphpad Software Inc., CA, USA) was used for all Statistical analyses. Shapiro–Wilk test was followed for the normality test. Accordingly, Mann–Whitney U-test and Student’s t-test were utilized for non-parametric and parametric data, respectively. For comparison of multiple groups with normal distribution, one-way ANOVA with Bonferroni multiple comparison corrections was used. On the other hand, for the comparisons with three or more group without normal distribution, Kruskal-Wallis test with Dunn’s multiple comparison was utilized. All data were provided as mean ± standard error of the mean. A p-value less than 0.05 (p<0.05) was considered as statistically significant in all experiments. All in vitro experiments were repeated three times, unless otherwise indicated.

Ameliorative effects of DMF on secreted NLRP3 inflammasome-related cytokines IL-1β and IL-18 were tested with ELISA. We found that DMF significantly decreased the amounts of IL-1β ( Figure 1A ) and IL-18 ( Figure 1B ), which were induced by LPS and ATP treatment. Next, expression levels of IL-1β and IL-18 were evaluated with the RT-qPCR method. More than a 3-fold significant decline in IL-1β mRNA level ( Figure 1C ) and approximately 2-fold significant decline in IL-18 mRNA level ( Figure 1D ) were observed with DMF pretreatment. Furthermore, we observed that DMF significantly reduced both pro-IL-1β ( Figures 1E, F ) and secreted mature IL-1β protein levels ( Figures 1E, G ) against NLRP3 inflammasome activation. Additionally, we found that DMF significantly decreased High Mobility Group Box 1 (HMGB1) expression in LPS and ATP-induced microglia ( Figures 1E, H ).

Next, we investigated DMF’s effect on the activated NLRP3 protein complex consisting of NLRP3, ASC adaptor protein, and Caspase-1. The intracellular level of p45 caspase-1 protein showed no significant difference in all three groups ( Figures 2A, B ). Active caspase-1 (p20) protein level significantly decreased in DMF pretreated cells compared to LPS and ATP-induced cells ( Figures 2A, C ). Additionally, Caspase-1 activity was evaluated and, there was a 2-fold increase with LPS and ATP treatment as it decreased 25% with DMF pretreatment ( Figure 2D ). Next, our investigations demonstrated that the DMF pretreated group showed an approximately 2-fold significant reduction in protein ( Figures 2E, F ) level and more than 2.5-fold significant reduction of the NLRP3 mRNA level ( Figure 2G ) as compared to the LPS and ATP treated group. Adaptor protein ASC is critically essential for the assembly of the NLRP3 inflammasome complex. Immunofluorescence staining results indicated that DMF diminished ASC speck formation induced by LPS and ATP administration ( Figures 2H, I ).

We further checked the effects of DMF on pyroptotic cell death stimulated by LPS and ATP administration. In the LDH assay ( Figure 3A ), we observed a significant decrease in cell death with DMF pretreatment compared to microglia with LPS and ATP treatment. Additionally, we revealed significant augmentation in microglia viability with DMF pretreatment ( Figure 3B ) We further carried out PI staining, a DNA intercalating dye that stains pyroptotic cells. We found a significant decrease in the number of pyroptotic cells with DMF pretreatment in PI staining ( Figures 3C, D ). Lastly, we evaluated the effect of DMF on GSDMD cleavage with western blotting and found a significant decline in cleaved GSDMD levels in DMF pretreated group against only the LPS and ATP treated group ( Figures 3E, F ).

We studied mitochondrial ROS production with MitoSOX. Significantly elevated mtROS levels induced by LPS and ATP treatment were significantly reduced with DMF pretreatment ( Figures 4A, B ). After that, we investigated intracellular total ROS production of microglia with DCFDA. DMF pretreatment significantly alleviated, nearly 2.5-fold, total ROS production compared to the LPS, and ATP treated group ( Figure 4C ). Lastly, we checked the mitochondrial membrane potential with JC-1 staining. There was a significant decrease in the green ratio in DMF pretreated group compared only to the LPS, and ATP treated group via flow cytometry ( Figures 4D, F ); green signal indicates deteriorated mitochondrial membrane potential, whereas the red signal denotes healthy mitochondrial membrane potential. Furthermore, we confirmed the improvement of mitochondrial membrane potential with DMF pretreatment under a fluorescent microscope with JC-1 staining. We observed an increased green signal with LPS and ATP introduction and, conversely, an increased red signal and reduced green signal with DMF pretreatment ( Figure 4E ).

As the NF-κB pathway is the upstream pathway of the NLRP3 inflammasome pathway, we evaluated DMF’s effect on the NF-κB pathway with western blotting and immunostaining. Firstly, we found that DMF significantly increased the IκB-α protein level, the expression of which decreased by LPS treatment ( Figures 5A, B ). Next, we checked DMF’s effects on NF-κB subunits p65 and p50. We demonstrated that DMF significantly reduced nuclear p-p65 compared to the LPS treated group ( Figures 5A, C ). Furthermore, DMF pretreatment significantly reduced increased protein levels of nuclear p65 and p50 reduced with LPS treatment ( Figures 5A, D, E ). We also confirmed our result with immunostaining against the p65 and the p50 subunit of NF-κB as DMF decreased p50 and p65 translocation into the nucleus ( Figures 5F, G ). Lastly, we further checked the expression levels of miR-155 and miR-146a with the RT-qPCR method, as these miRNAs are known to act upstream of NF-κB and be related to inflammasome activation. We showed that DMF significantly decreases the expression levels of miR-155 ( Figure 5H ) and miR-146a ( Figure 5I ) in both in vitro and in vivo models of NLRP3 inflammasome activation.

Given that DMF is a well-known activator of the Nrf2 signaling pathway, we further checked DMF’s effect on Nrf2 on microglia. First, we showed that DMF induced Nrf2 activation and led to the translocation of Nrf2 to the nucleus ( Figures 6A–C ). Next, we determined the expression levels of Nrf2 target genes, including Ho-1, Nqo1, Gclc, Gclm, Srxn1, and its inactivator Keap-1 by RT-qPCR. The results demonstrated that treatment with DMF increased the expression levels of the target genes and decreased Keap-1 expression ( Figure 6D ). We further confirmed our results with an in vivo study. In particular, microglia isolated from mice were stained against NLRP3 and Caspase-1. The results demonstrated that DMF decreased NLRP3 and Caspase-1 protein expression in those mice injected with DMF. However, ML385, a Nrf2 inhibitor, reversed this effect and increased the expression levels showing that Nrf2 mediates the protective effects of DMF ( Figures 6E – H ).

In order to assess the effects of DMF on LPS-induced sickness behaviors, behavioral tests and clinical evaluation on mice were carried out ( Figure 7A ). In this regard, we conducted clinical scoring, weighed the mice, and performed OFT, FST, and TST. Results demonstrated that injection of LPS resulted in sickness behaviors in BALB/c mice while treatment with DMF reversed this situation ( Figure 7B ). On the other hand, pretreatment with Nrf2 inhibitor, ML385, reversed DMF’s protective effects on the LPS-induced sickness phenotype. We also weighed the animals before and after treatment. Results indicated that treatment with DMF decreased weight loss and inhibition of Nrf2 activation worsened this situation ( Figure 7C ). Next, we performed behavioral tests. In OFT, the animals treated with DMF crossed a significantly higher number of squares compared to LPS. However, the ones injected with ML385 traveled less in the open field ( Figure 7D ). Again, in TST and FST, DMF increased the latency to immobility. These effects also seem dependent on Nrf2 activation as the use of ML385 decreased the latency to immobility ( Figures 7E, F ). As a result, DMF significantly restored sickness behaviors, reversed clinical course, and prevented weight loss induced by LPS.