WWL70 attenuates PGE₂ production derived from 2-arachidonoylglycerol in microglia by ABHD6-independent mechanism

BV2 microglial cells [25] were cultured in complete DMEM containing 5% heat-inactivated fetal bovine serum (Life Technologies, Grand Island, NY) under humidified 5% CO₂ environment. The cells were maintained by medium change every other day. One day prior to experiment, cells were plated on 24-well plates to reach 90% confluence at the beginning of experiment.

For primary microglia culture, Sprague–Dawley rats were purchased from Charles River Laboratories (Frederick, MD) and housed in the animal facility of the Uniformed Services University of the Health Sciences (USUHS). All animal protocols were approved by the USUHS Institutional Animal Care and Use Committee. P2 Sprague–Dawley rat pups (male and female) were sacrificed and their brains removed. Mixed glial cultures were then prepared from the cortices after careful removal of the meninges and brain stem. The cortices were then triturated by serial trituration with a 10-ml pipette, 18-G needle (3×), 22-G needle (3×), and finally, a 25-G needle (2×) in culture medium (DMEM containing 10% fetal bovine serum, 1% glutamax, and 1% antibiotic-antimycotic). After trituration, the suspension was filtered using a 70-micron mesh and then pelleted at 168×g for 10 min at room temperature. Cells were resuspended in culture medium, seeded into T75 flasks, and incubated in a CO₂ incubator. The medium was replaced every 2–3 days. After 14 or 15 days in culture, primary microglia were harvested by differential shaking on an orbital shaker for 1 h at 200 rpm in a CO₂ incubator. The medium, containing the detached microglia, was collected and centrifuged at 671×g for 5 min at room temperature. Cells were then resuspended with DMEM containing 10% normal horse serum, 1% glutamax, and 1% streptomycin/penicillin and transferred to uncoated plates at a density of 2.5 × 10⁵ cells/mL.

KT182, an ABHD6 inhibitor, and HT-01, the activity-based protein profiling (ABPP) probe specific for ABHD6, were kindly provided by Drs. Hsu and Cravatt [26]. 2-Arachidonoylglycerol [glycerol-1,2,3-³H] was from American Radiolabeled Chemicals Inc. (Saint Louis, MO). Cyclooxygenase inhibitor assay kits including COX inhibitor and recombinant COX-1 or COX-2 activity assay kit were from Cayman Chemical (Ann Arbor, MI). siRNA (FlexiTube Mm_abhd6_3 and Allstars Negative Control) and HiPerfect transfection reagent were from QIAGEN (Valencia, CA). WWL70, methyl arachidonyl fluorophosphonate (MAFP), SR141716 (SR1), SR144528 (SR2), 2-AG, 2-AG-d₈, and AA were purchased from Cayman Chemical. Other reagents including lipopolysaccharide (LPS) were purchased from Sigma-Aldrich (St. Louis, MO).

A multi-well cell culture plate was prepared 1 or 2 days prior to the test. The cell culture medium was replaced with pre-warmed medium containing the ABHD6 inhibitor WWL70 (10 μM) and incubated for 15 min. The cells were treated with 10 μM of 2-AG for 15 min, followed by addition of 100 ng/ml LPS for BV2, or 2 ng/ml LPS for primary microglia. After incubation for 18 h, the culture medium was collected. Before addition to the enzyme-linked immunoassay (EIA), the medium was centrifuged at 5000 rpm for 2 min with a table top centrifuge to exclude residual cells. To determine the role of WWL70 on PGE₂ production in vivo, EAE was induced by subcutaneous injection of myelin oligodendrocyte glycoprotein peptide 35–55 (MOG_35–55) in 8-week-old female C57BL/6 mice, and the clinical score was assessed as we reported recently [23]. WWL70 (10 mg/kg, i.p.) was given starting at the disease onset and then once a day until the end of the test. The mouse forebrain at 3 weeks post-immunization was dissected and kept frozen at −80 °C until use. The forebrain tissue was homogenized with one-fifth volume of 0.02% trifluoroacetic acid (TFA) and one volume of acetonitrile using a Potter homogenizer at 4 °C. Homogenate was dispersed completely in 2 ml of acetonitrile by vortex and stored at 4 °C overnight. Debris was excluded by centrifugation at ×2000g for 5 min, and then the supernatant was evaporated under the nitrogen gas streaming in a water bath (approximately 35 °C). After reconstitution with the EIA buffer, the levels of PGE₂ were measured following the manufacturer’s protocol (Cayman Chemical, Ann Arbor, MI).

Hydrolysis activity of 2-AG was assessed as previously described [27]. For hydrolysis activity in intact cells, BV2 cells in a 24-well plate were pre-incubated with or without inhibitor for 30 min at 37 °C, followed by washing with pre-warmed DMEM plus 0.15% fatty acid-free BSA. The cells were incubated with 0.2 ml of 10 μM radiolabeled ³[H]-2-AG (33 nCi) in DMEM containing 0.15% fatty acid-free BSA. After 2 min of incubation, the reaction was terminated by mixing with 0.4 ml chilled methanol. The cell extract was transferred to a silanized glass tube, followed by mixing with 0.4 ml of chloroform by brief vortexing. The mixture was centrifuged at ×3000g for 5 min to separate the aqueous and organic phases. Aliquot of the aqueous phase was mixed with scintillation cocktail and measured in a scintillation counter LS6500 (Beckman Coulter, Brea, CA). For hydrolysis activity in membrane fraction, BV2 cell homogenate was centrifuged at ×1400g for 5 min to remove the debris, then ultracentrifuged at ×100,000g at 4 °C for 30 min. The pellet was resuspended with PBS, and protein concentration was determined by DC protein assay kits using BSA as a standard (Bio-Rad). Four hundred microliters of the membrane fraction (20 μg) was pre-incubated in a silanized glass tube with or without 10 μM of WWL70 for 5 min at 37 °C, then mixed with 100 μl of 10 μM ³[H]-2-AG (33 nCi) in PBS containing 0.15% fatty acid-free BSA. After 2 min of incubation, the reaction mixture was mixed with chilled 2 ml methanol/chloroform (1:1). Radioactivity in the aqueous phase was measured as described above.

One milligram per milliliter of the membrane fraction prepared as above was pre-incubated with WWL70 at indicated concentrations for 10 min at room temperature, then mixed with 1 μM HT-01, a serine hydrolase probe that specifically recognizes active ABHD6 enzymes [26], at 37 °C for 40 min. The reaction mixture was added to the SDS-PAGE sample buffer and heated at 95 °C for 5 min. Approximately 10 μg of the protein was loaded on SDS-PAGE. The gel was scanned with a fluorescence imager, Fuji FLA-5100 (Fujifilm, Edison, NJ) with an excitation wavelength at 473 nm using FITC mode. The intensity of the fluorescent band is proportional to the amount of active ABHD6. Subsequently, the gel was transferred onto PVDF membrane for western blotting with anti-calnexin as a loading control for the membrane fraction.

For western blotting, cell lysate or microsome fraction was prepared with RIPA buffer containing 150 mM NaCl, 50 mM Tris-HCl (pH 8.0), 1% Triton X-100, 0.5% Na deoxycholate, 0.1% SDS, 1 mM EDTA, 1 mM EGTA, 1 mM Na₃VO₄, 1 mM β-glycerophosphate, and protease inhibitor cocktail (Roche Applied Sciences) for 5 min on ice, followed by centrifugation at ×12,000g for 5 min at 4 °C to remove the debris. Transferred PVDF membrane was pre-incubated with 5% BSA in PBS + 0.05% Tween-20 (PBST) for 30 min, then incubated with antibodies against β-actin (AC-74, Sigma-Aldrich) at 1:1000, calnexin (T-40, Santa Cruz) at 1:1000, microsomal prostaglandin E2 synthase (mPGES)-1 (#160140, Cayman Chemical) at 1:500, mPGES-2 (#160145, Cayman Chemical) at 1:500, and COX-2 (#160106, Cayman Chemical) at 1:250 in PBST at 4 °C overnight. The transferred membrane was reacted with a secondary antibody conjugated with horseradish peroxidase (Bio-Rad) at 1:2500 for 2 h, followed by visualization with ECL reagent (Thermo Scientific) using the LAS3000 imager (Fujifilm). The relative intensity of COX-2 to β-actin was quantified using ImageJ software (National Institutes of Health), and the fold change compared to the LPS treatment alone was presented.

Sample preparation for 2-AG quantification was carried out based on a method published previously [28]. BV2 cells (90% confluence) in 10-cm dishes were treated with WWL70 (10 μM) or MAFP (10 μM) for 1 h. After rinsing with PBS once, the cells were collected by centrifugation at ×5000g for 2 min. The pellet was suspended with 0.1 ml of 0.02% TFA and 1 nmole of 2-AG-d₈ by pipetting and dispersed in 4 ml of acetonitrile in a silanized glass tube to precipitate the debris overnight at −20 °C. The supernatant after centrifuged at ×5000g for 5 min was transferred to a new glass tube and evaporated under a nitrogen gas stream in a mild hot water bath (approximately 35 °C). 2-AG was resuspended with 0.1 ml of acetonitrile and stored at −80 °C until mass analysis.

An HPLC system (1200 Series, Agilent Technologies, Santa Clara, CA) was used with a reverse phase guard column (Wide Pore C18 (ODS), 4 × 2 mm ID; Phenomenex, Torrance, CA), and the column (Sephasil Peptide C18, 5 u, ST, 100 × 4.6 mm ID; Pharmacia Biotech (Amersham), Piscataway, NJ) was maintained at 40 °C. The mobile phase was composed of solvent A: 0.2% formic acid in water, and solvent B: 0.2% formic acid in methanol; the following gradient was used: 62% A/38% B isocratic for 30 s, ramp to 90% B in 60 s, isocratic at 90% B for 18.5 min, ramp back to 62% A/38% B in 60 s, and re-equilibrate at 62% A/38% B for 8 min. The flow rate was 0.4 ml/min. The HPLC output was directed into the TurboV electrospray ionization (ESI) source of a Q-Trap 4000 mass spectrometer (AB Sciex, Framingham, MA). The injection volume was 20 μl. LC-MS/MS analysis was performed in a positive mode with the ion source temperature of 600 °C, a spray voltage of 5.5 kV, and a declustering potential of 45 V. Multiple reaction monitoring (MRM) was performed on the transitions m/z 379.5→287.5 for 2-AG and 386.5→293.5 for 2-AG-d₈. Calibration curves of the ratio of analyte (2-AG) to internal standard (2-AG-d₈) peak areas vs. concentration ratio (at least seven different 2-AG concentrations at a single fixed 2-AG-d₈ concentration) were generated by linear least squares fitting and used to calculate the 2-AG concentrations in the samples (converting the observed y-value (peak area ratio) of the samples into x-values (concentration ratios with a fixed internal standard concentration).

The method for LCMS analysis of PGE₂ and PGE₂-glycerol in positive ion mode was derived from a different previously published procedure [29]. BV2 cells pretreated with or without 0.1 μg/ml of LPS for 8 h were rinsed once with pre-warmed serum-free medium (Opti-MEM, Thermo Fisher Scientific), then incubated with Opti-MEM containing 10 μM 2-AG, 10 μM WWL70, or both for 30 min in CO₂ incubator. The medium was collected and centrifuged at ×1000g for 5 min to remove floating cells and stored at −80 °C until use. One milliliter of medium was mixed with 1.4 ng PGE₂-d4 and 0.5 ng PGE₂-G-d5 (Cayman Chemical) and acidified by glacial acetic acid to 1% at final concentration. It was loaded to a solid-phase extraction column (Oasis HLB 1 cc, Waters Corporation, Milford, MA) equilibrated with 0.5% acetic acid and methanol. After washing with 0.5% acetic acid followed by 0.5% acetic acid with 15% methanol, PGs were eluted with 1.5 ml of methanol. After evaporation in nitrogen streaming under approximately 30 °C, the PGs were reconstituted with acetonitrile/water (1:2).

The same LCMS system described above was used with a Higgins Analytical TARGA C18 3 μm 2.1 × 100 mm column (Nest Group Inc., Southborough, MA). The HPLC gradient was as follows: the column was equilibrated at 80% 5 mM ammonium formate in water (buffer A)/20% 4:1 acetonitrile/water with 5 mM ammonium formate (buffer B); after sample injection, the gradient for elution of analytes went from 20 to 80% buffer B in 15 min. The column was re-equilibrated by ramping to 100% buffer B in 1 min, holding at 100% buffer B for 2 min to flush, ramping back down to the initial HPLC buffer composition (80%/20% A:B) in 2 min, and holding at the initial buffer composition for 8 min before the next injection. The flow rate was 0.2 mL/min, the column temperature was 40 °C, and the injection volume was 25 μL. The ion source temperature was 550 °C, and the declustering potential was 40 V. Since the HPLC buffer contained ammonium ions, the precursor ions for PGE₂ and PGE₂-glycerol were ammonium adducts (M + 18). The following MRM transitions were used: for PGE₂, precursor ion mass 370.3 Da to fragment ion mass 335.2 Da (collision energy (CE) 12.5 V) and 317.1 Da (CE 15 V); for d4-PGE₂, 374.3 to 339.2 Da (CE 12.5 V) and 321.3 Da (CE 15 V); for PGE₂-glycerol, 444.4 to 409.2 Da (CE 12.5 V) and 299.2 Da (CE 26 V); and for d5-PGE₂-glycerol, 449.4 to 414.1 Da (CE 12.5 V) and 299.2 Da (CE 26 V).

Total RNA from BV2 cells or primary microglia treated with reagents for 8 h was isolated using TRIzol (Life Technologies) according to the manufacturer’s protocol. Five hundred nanograms of RNA were used for cDNA synthesis using MAXIMA First Strand cDNA synthesis kit with dsDNAse (Thermo Fisher Scientific) following the manufacturer’s protocol. The Cq value of negative control without reverse transcription was >40 due to the treatment with DNase. Complementary DNA from 10 ng RNA was employed for qPCR in the presence of 250 nM of gene-specific primers (mPGES-1 forward, 5′-TGTCCAAATCCTGTCTTCCA-3′, reverse, 5′-GGTTCTGGAGCACACCCTAT-3′; mPGES-2 forward, 5′-GAAATGGCTGCAGAATTGAA-3′, reverse, 5′-AAGGAGAATGGTGCTCCAAG-3′; COX-2 forward, 5′-GAAATGGCTGCAGAATTGAA-3′, reverse, 5′-AAGGAGAATGGTGCTCCAAG-3′; ABHD6 forward, 5′-CATTCCAATCCTGGCATTTGTTG-3′, reverse, 5′-ATGGTGTGCGTAGCGAACTT-3′; and GAPDH forward, 5′-AGGTCGGTGTGAACGGATTTG-3′, reverse, 5′-TGTAGACCATGTAGTTGAGGTCA-3′) using Power SYBR Green PCR master mix (Life Technologies) in 12 μl reaction mixture. Thermal cycling condition was 95 °C for 10 min, followed by 40 cycles of 95 °C × 15 s and 60 °C × 60 s performed by LightCycler 480 II (Roche Life Science, Indianapolis, IN). Gene-specific PCR amplification was confirmed by the melting curve profile using LightCycler 480 system program showing 80 to 85 °C of single melting temperature for all the genes tested. GAPDH was used as a reference gene because its Cq value was constant among the cells tested (Cq value is 15.7 ± 0.1, mean ± SD). The PCR amplification linearity was confirmed using serial diluted cDNA showing the consistent amplification in the concentration range tested (r ² > 0.998).

BV2 cells in a 24-well plate (50 to 70% confluent) were transfected with siRNA for the mouse ABHD6 gene (Mm_abhd6_3, QIAGEN) or the negative control using HiPerfect transfection reagent according to the manufacturer’s protocol. Briefly, 37.5 ng siRNA was mixed with 100 μl DMEM without FBS, then added 6 μl of the transfection reagent and incubated for 10 min at room temperature. The siRNA complex suspension was added to the cells drop-wise. Total RNA was isolated from the cells after 1 day of incubation and qRT-PCR was performed to examine the expression of ABHD6 and GAPDH. Otherwise, the transfected cells after 2 days of incubation were used to prepare membrane fraction for ABPP to assess ABHD6 hydrolytic activity, followed by western blot with the calnexin antibody as a loading control.

The COX assay was carried out using the COX inhibitor screening assay kit from Cayman Chemical. Briefly, ovine recombinant COX-1 or human recombinant COX-2 was pre-incubated in 0.1 M Tris-HCl (pH 8.0) and EDTA (1 mM) buffer with WWL70 (10 μM), SC-560 (0.33 μM), or Dup-697 (0.3 μM) for 5 min at 27 °C. Then, 10 μM of AA was added and incubated for 1 min at 27 °C. The reaction was stopped by adding 1 N HCl. Stannous acid was then added to convert PGH₂ to the stable product PGF_2α. The resultant mixture was applied to a prostaglandin EIA to measure the total amount of prostaglandins including PGE₂ and PGF_2α. The PGE₂ biosynthesis assay was performed using microsomes derived from BV2 cells treated with LPS for 8 h or brain tissue from EAE mouse at day 21 post-immunization [23]. BV2 cells were suspended with 0.1 M potassium phosphate (pH 7.4), 0.25 M sucrose, 1 mM β-glycerophosphate, 1 mM sodium vanadate, 1 μM MAFP, and protease inhibitor cocktail (Roche) and disrupted by sonication for 15 s, while the brain tissue was homogenized with a Potter homogenizer with the potassium phosphate buffer at 4 °C. These homogenates were centrifuged at ×1400g for 10 min at 4 °C to remove the debris, followed by further centrifugation at ×10,000g for 10 min at 4 °C. The supernatant was ultracentrifuged at ×170,000g for 1 h at 4 °C to precipitate the microsome fraction. The pellet was suspended with potassium phosphate buffer containing glutathione (2.5 mM) by brief sonication. One hundred micrograms per milliliter of BV2 microsomes was pre-incubated with WWL70, MK886 (3 μM), SC-560 (1 μM), or NS398 (10 μM) for 5 min at 23 °C, then mixed with 10 μM of AA for 1 min at 23 °C. 500μg/ml brain microsomes were incubated with 10 μM of AA for 2 min at 23 °C. The reaction was stopped by mixing with stannous acid (5 mg/ml in 0.1 N HCl) to deactivate the enzyme and convert intermediate PGH₂ to PGF_2α, followed by the measurement of PGE₂ concentration by EIA as described above. The activity was determined after subtraction with the amount of PGE₂ in the microsome fraction incubated without substrate.

All data are expressed as mean ± SD. The GraphPad Prism 7 (GraphPad Software Inc., San Diego, CA) was used for statistical analysis. The statistical comparison among the drug-treated groups (different drugs or the same drug at various concentrations) was performed using one-way ANOVA followed by Turkey’s test; while in experiments with both LPS, LPS + 2-AG, or LPS + AA groups, two-way ANOVA was performed. An unpaired t test was used for data comparison between two groups. Statistical significance was set at P < 0.05, with single, double, and triple asterisks to denote p < 0.05, 0.01, and 0.001, respectively.