Effects of a Novel GPR55 Antagonist on the Arachidonic Acid Cascade in LPS-Activated Primary Microglial Cells

Neuroinflammation is a crucial process to maintain homeostasis in the central nervous system (CNS). However, chronic neuroinflammation is detrimental, and it is described in the pathogenesis of CNS disorders, including Alzheimer’s disease (AD) and depression. This process is characterized by the activation of immune cells, mainly microglia. The role of the orphan G-protein-coupled receptor 55 (GPR55) in inflammation has been reported in different models. However, its role in neuroinflammation in respect to the arachidonic acid (AA) cascade in activated microglia is still lacking of comprehension. Therefore, we synthesized a novel GPR55 antagonist (KIT 10, 0.1–25 µM) and tested its effects on the AA cascade in lipopolysaccharide (LPS, 10 ng / mL)-treated primary rat microglia using Western blot and EIAs. We show here that KIT 10 potently prevented the release of prostaglandin E2 (PGE2), reduced microsomal PGE2 synthase (mPGES-1) and cyclooxygenase-2 (COX-2) synthesis, and inhibited the phosphorylation of Ikappa B-alpha (IκB-α), a crucial upstream step of the inflammation-related nuclear factor-kappaB (NF-κB) signaling pathway. However, no effects were observed on COX-1 and -2 activities and mitogen-activated kinases (MAPK). In summary, the novel GPR55 receptor antagonist KIT 10 reduces neuroinflammatory parameters in microglia by inhibiting the COX-2/PGE2 pathway. Further experiments are necessary to better elucidate its effects and mechanisms. Nevertheless, the modulation of inflammation by GPR55 might be a new therapeutic option to treat CNS disorders with a neuroinflammatory background such as AD or depression.


Introduction
Neuroinflammation is considered a double-edged sword, in which the acute response is defined as a protective event and, on the other hand, chronic neuroinflammation is detrimental [1]. Several studies demonstrated the contribution of the inflammatory process to the pathophysiology of the different neurological and neuropsychiatric disorders such as Alzheimer's disease (AD), Parkinson's disease (PD), schizophrenia, and depression. Neuroinflammation is well characterized by the activation of glial cells, mainly microglia, which leads to the release of inflammatory mediators [2]. Microglial cells are the first line of defense in the central nervous system (CNS) and exert an anti-inflammatory role being neuroprotective and implicated in the resolution of the inflammatory event by phagocytosis and tissue repair. However, for reasons that are still not clear, microglia have also a detrimental role, where the deregulated microglia response by inducing a pro-inflammatory phenotype releasing pro-inflammatory mediators including cytokines such as tumor necrosis factor (TNF) α, interleukin-6 (IL-6), and IL-1β, prostaglandins (PGs), reactive oxygen species (ROS), and nitric oxide (NO) resulting in progressive neurotoxic consequences and neurodegeneration [3][4][5].
The G-protein-coupled receptor, GPR55, is highly expressed in the CNS [6], but also peripheral tissues [7]. Moreover, our group and others have demonstrated the expression of GPR55 in microglial cells [8,9]. The role of GPR55 in inflammation has been reported in in vitro and in vivo models [10][11][12] and its mediated effects remain controversial and depended on the cell type and stimulation. The activation of GPR55 by its agonist O-1602 increased pro-inflammatory cytokines and cell cytotoxicity in monocytes and natural killer (NK) cells stimulated with lipopolysaccharide (LPS) [11]. Corroborating with this, an antagonist of GPR55, CID16020046, as well as GPR55 −/− knockout mice decreased pro-inflammatory cytokines in mice models of colitis comparable to human inflammatory bowel disease (IBD) [12]. In hyperalgesia associated with inflammatory and neuropathic pain, increased anti-inflammatory cytokines (IL-4 and IL-10) and also pro-inflammatory interferon (IFN) were observed in GPR55 −/− knockout mice [10]. On the other hand, treatment of IL-1β stimulated neuronal cells with O-1602 and chronic systemic inflammation induced by LPS administration in mice demonstrated neuroprotection and an anti-inflammatory profile [13].
GPR55 couples to Gα12/13 proteins which leads to the activation of ras homolog gene family members A (RhoA) and Rho-associated protein kinase (ROCK) triggering the activation of the phospholipase C pathway increasing the intracellular Ca 2+ and phosphorylation of extracellular signal-regulated kinase (ERK) and p38 MAPK [14,15]. The MAPK signaling is involved in the regulation of inflammatory events [16] and we have demonstrated in activated microglia that the activation of JNK, ERK, and p38 MAPK are implicated in the regulation of COX-2/mPGES-1/PGE 2 levels, important proteins involved in neuroinflammation also mediating fever [17][18][19]. We demonstrated that GPR55 may regulate COX-2/mPGES-1/PGE 2 in several steps. GPR55 is endogenously activated by lysophosphatidylinositol (LPI, lysoPI), a bioactive lipid involved in a range of pathologies such as obesity, cancer, and inflammation [20,21], and by different cannabinoids [22]. LPI has been demonstrated to reveal anti-inflammatory effects in microglial cells by decreasing cytokine release and suppressing intracellular generation of reactive oxygen species [23]. Furthermore, a neuroprotective effect in an ex-vivo model of excitotoxicity, mediated by microglial cells, has been shown [24].
These studies demonstrated the involvement of GPR55 in different aspects of inflammation, suggesting this receptor to be an interesting target to modulate neuroinflammation by interfering with the arachidonic acid (AA) cascade. In the present study, a novel GPR55 antagonist based on the coumarin structure (KIT 10) was synthesized, and the effects of KIT 10 on the production of PGE 2 induced by LPS in primary microglial cells was investigated. We have observed the anti-neuroinflammatory effect of KIT 10 by potently inhibiting LPS-induced PGE 2 synthesis in microglia and this effect was associated with the reduction of mPGES-1 and COX-2 protein synthesis, but not through the inhibition of COX-1 and COX-2 enzyme activity. Interestingly, KIT 10 affected the NF-κB pathway, whereas MAPK activation was not affected.

Effects of KIT 10 on Cell Viability
We first evaluated the effects of KIT 10 on cell viability in primary rat microglia using an ATP assay to exclude possible biological inhibitory effects due to a reduction of cell viability. As shown in Figure 1, KIT 10 in all concentrations, as well as LPS, did not change cell viability compared to vehicle only (0.1% DMSO)-treated cells. DMSO (10%) was used as a cell toxic control.

KIT 10 Prevents LPS-Induced PGE 2 Release Primary Microglial Cells
We next studied the effects of KIT 10 on LPS-induced PGE 2 in microglia. We observed a concentration-dependent inhibition of LPS-mediated PGE 2 release starting with 1 µM and maximal effects using 10 and 25 µM with a calculated IC 50 of 2.5 ± 0.93 µM ( Figure 2). Since 10 µM was sufficient to potently prevent LPS-induced PGE 2 synthesis, we used this concentration as a maximal concentration in further experiments. Figure 2. KIT 10 prevents LPS-induced PGE 2 release in primary rat microglial cells. KIT 10 was added 30 min before stimulation with LPS and the amount of PGE 2 in the culture medium was determined after 24 h using an enzyme immunoassay. Each column and error bar represent the mean ± SEM of 3 new cultures/group. ** p < 0.01 and *** p < 0.001 with respect to LPS (One-way ANOVA followed by Newman-Keuls post-test).

KIT 10 Didn't Affect the COX Enzyme Activities
To verify that KIT 10 mediated inhibition of PGE 2 release is additionally due to direct suppression of COX enzymatic activities, we performed a COX enzyme activity assay. As shown in Figure 4, COX-1 and -2 activities were not affected by all concentrations of KIT 10. In contrast, COX enzyme activities were strongly inhibited by classical inhibitors of COX-1 (SC560, Figure 4A) and COX-2 (diclofenac sodium, Figure 4B). To assess COX-2 activity, cells were stimulated for 24 h with LPS (10 ng/mL) and then treated with different concentrations of KIT 10 for 15 min. After this incubation time, 15 µM of AA were added and PGE 2 in the supernatants was measured by enzyme immune assay. Data are expressed as mean ± SEM of at least three new cultures/group. * p < 0.05 and *** p < 0.001 with respect to control (One-way ANOVA followed by the Newman-Keuls post-test).

Effects of KIT 10 on IκB-α Phosphorylation in LPS-Stimulated Primary Microglial Cells
Multiple evidence indicates that the phosphorylation and degradation of Ikappa B-alpha (IκB-α), as well as the phosphorylation of IKK-alpha/beta, are primarily responsible for the activation of the NF-kappaB (NF-κB) signaling pathway [7]. Therefore, we studied the effects of KIT 10 on IκB-α phosphorylation, a crucial upstream step of the inflammation-related NF-κB signaling pathway. As shown in Figure 5, LPS-induces prominent phosphorylation of IκB-α. Treatment of microglial cells with increasing concentrations of KIT 10 reversed LPS-induced IκB-α phosphorylation in the concentrations of 5 and 10 µM. Microglia were pre-treated with KIT 10 for 30 min, afterwards, cells were incubated with or without LPS (10 ng/mL) for the next 15 min, and SDS gel electrophoresis followed by western blot performed. Representative original blots are included in Appendix A ( Figure A1). Data are expressed as mean ± S.E.M. of three independent experiments. * p < 0.05 with respect to the LPS group (One-way ANOVA followed by the Newman-Keuls post-test).

Effects of KIT 10 on Mitogen-Activated Kinases (MAP Kinases): JNK, ERK 1/2 and p38 MAPK in LPS-Stimulated Primary Microglial Cells
MAPK signaling is involved in the regulation of inflammatory events [30] and we have demonstrated that the activation of JNK, ERK, and p38 MAPK are implicated in the regulation of COX-2/PGE 2 levels in activated microglia [17][18][19]. We, therefore, explored the effects of KIT 10 on MAPK signaling pathways in LPS-stimulated primary microglial cells. However, none of the LPS-phosphorylated and thus activated MAPKs [JNK ( Figure 6A), ERK 1/2 ( Figure 6B), p38 MAPK ( Figure 6C)] were significantly affected by KIT 10 although a tendency of weak inhibition was overserved for p38 MAPK and JNK

Discussion
Neuroinflammation is an important process to maintain homeostasis in the CNS. It is involved in the response against different invaders and deregulators such as pathogens, toxins, infections, and the accumulation of aggregated or modified proteins [31].
The neuroinflammatory response is mediated by glial cells, especially microglia. Microglial cells are the resident macrophages of the CNS, responsible for the initiation, amplification, and/or equalization of the inflammatory response by the synthesis of inflammatory mediators, like cytokines, prostaglandins (PGs), and free radicals [3][4][5].
In the present study, we demonstrated the anti-neuroinflammatory effects of a novel GPR55 antagonist, KIT 10, in LPS-activated primary microglial cells by potently inhibiting LPS-induced PGE 2 synthesis. This effect was associated with the reduction of protein synthesis of mPGES-1 and COX-2, and not through the inhibition of COX-1/2 enzyme activity. Interestingly, KIT 10 affected the NF-κB pathway, but not MAPKs.
With the intention to understand the effects of the novel GPR55 antagonist KIT 10 in neuroinflammation, especially in microglia, we first eliminated the possibility that any effects observed may be caused by toxic effects of the compound. Therefore, we evaluated its effects on cell viability and observed that the compound is not toxic at all concentrations tested.
We next investigated the effects of KIT 10 on LPS-induced PGE 2 release. Microglial cells are the most important source of PGE 2 in neuroinflammation. The stimulation of primary microglia with LPS strongly increases PGE 2 and its synthesizing enzymes COX-2 and mPGES-1 [32][33][34]. We have previously shown that GPR antagonists are inhibiting LPSinduced COX-2 and mPGES-1 protein levels and PGE 2 synthesis in primary microglia [8]. Corroborating with our findings, the GPR55 antagonist CID16020046 was demonstrated to reduce COX-2 mRNA and protein levels and decreased the release of PGE 2 of advanced glycation end products (AGEs) induced chondrocytes activation in chondrogenic cell line ATDC5 [35]. In mice models of intestinal inflammation (chronic drinking of 2.5% dextran sulfate sodium or by a single intrarectal application of trinitrobenzene sulfonic acid in C57BL/6 mice), chronic treatment with CID16020046 (20 mg/kg) significantly decreased COX-2 levels [12]. Furthermore, GPR55 knockout mice (GPR55 -/-) revealed a decrease in COX-2 levels in a mouse model of colorectal cancer, using azoxymethane and dextran sulfate sodium [36].
We further investigated if the reduction of PGE 2 levels may be mediated by a decrease of the enzymatic activity of COX, the mechanism of action of most nonsteroidal anti-inflammatory drugs (NSAIDs). Two COX isoforms exist, COX-1 and COX-2. COX-1 is constitutively expressed in almost every cell and under unstimulated conditions; primary microglial cells only express the COX-1 isoform [37]. The other isoform COX-2 is induced by cytokines and endotoxins but is also constitutive in some other cells such as neurons, endothelial, or smooth muscle cells [38,39]. In our previous study [8], we have demonstrated that GPR55 antagonists did not affect COX-2 activity, but increased COX-1 activity, being a beneficial property as COX-1 is responsible for gastric mucosa protection [40,41]. Here, we showed that KIT 10 did not affect both COX activities, demonstrating that the reduction on the PGE 2 levels by KIT 10 is not due to direct inhibition of COX enzyme activity.
The activity of the transcription factor Nuclear factor-kappa B (NF-κB) is increased in acute CNS diseases such as stroke and severe epileptic seizures, as well as in chronic neurodegenerative conditions, including AD, PD, and amyotrophic lateral sclerosis [42]. Moreover, several studies indicate that degradation of IκBα, as well as the phosphorylation of IKKα/β, are primarily responsible for the activation of the NF-κB signaling pathway [43][44][45]. The endogenous ligand of GPR55, LPI, has been shown to activate NF-κB and nuclear factors of activated T-cells (NFAT). On the other hand, the GPR55 antagonist CID16020046 concentration-dependently inhibited the LPI effects on these transcription factors [46]. In GPR55 knockout mice (GPR55 -/-), NF-κB levels are decreased in a mouse model of colorectal cancer [36]. In addition, the GPR55 antagonist CID16020046 showed inhibition on IκBα phosphorylation on AGEs induced activation of the chondrogenic cell line ATDC5 [35]. Further studies have to be performed to elucidate the upstream step on which KIT10 exactly prevents the phosphorylation of IκBα. Besides interfering with the activation of IL-1R-associated kinase (IRAK) 1/2 and IKK, GPR55 antagonists might directly act at recruited adapters of TLR4 such as myeloid differentiation factor (MyD) 88 and TNF-receptor-associated factor (TRAF6).
MAPK signaling is involved in the regulation of inflammatory events [16] and we have demonstrated that the activation of JNK, ERK, and p38 MAPK are implicated in the regulation of COX-2/mPGES-1/PGE2 levels in activated microglia [17][18][19]. Several studies have demonstrated that the activation of GPR55 induced the phosphorylation of ERK 1/2 [9,20,47,48] and p38 MAPK [30]. However, LPS-induced phosphorylation of JNK, p38 MAPK, and ERK was not affected by KIT 10.
In summary, we provided evidence that KIT 10 interferes in several steps of the synthesis of PGE 2 in LPS-activated microglia. This study provides significant insights on the potential anti-inflammatory activity of this GPR55 antagonist and described a novel compound that modulates prostaglandin production by microglia.

Chemistry
We have synthesized a novel GPR55 antagonist inspired by tetrahydrocannabinol, which we have called KIT 10. It differs from the compound KIT 17 (3-benzyl-6-hydroxy-5,7,8-trimethyl-2H-chromen-2-one) used in our previous study [8], which is a small generic coumarin and structurally different from KIT 10 (7-(1-butylcyclohexyl)-3-(4-fluorobenzyl)-5-hydroxy-2H-chromen-2-one). The position of the hydroxy groups (meta/para) causes a different acidity and thus different binding to proteins. Also, the bulky cyclohexyl ring in KIT 10 increases the hydrophilic character of the western portion of this compound. Thus, the compounds are very different from each other.
Commercially available reagents and solvents for the synthesis were obtained from commercial suppliers and used without further purification. The solvents used (dichloromethane and toluene) were dried before use (dichloromethane by distillation over calcium hydride; toluene by distillation over sodium metal with benzophenone as an indicator). The purity of the final compound was determined to be >98%. Nuclear magnetic resonance (NMR) spectra were recorded on an Avance 400 ( 1 H, 400 MHz; 13 C, 100 MHz; 19 F, 376 MHz) instrument (Bruker, Rheinstetten, Germany) in deuterated chloroform. The chemical shift of the residual protons was used as an internal standard: 1 H, 7.26 ppm; 13 C, 77.0 ppm. The chemical shifts of 19 F peaks were calculated by the instrument without an additional standard. Chemical shifts (δ) are expressed in parts per million (ppm); Coupling constants (J) are given in Hertz (Hz). The reactions were monitored by thin-layer chromatography (TLC) on aluminum sheets with silica gel 60 F254 (Merck, Darmstadt, Germany). Electron ionization mass spectra (EI-MS) and high-resolution mass spectra (HRMS) were recorded on a MAT 95 system (Finnigan, now Thermo Fisher, Waltham, MA, USA) and infrared spectra (IR) on a Bruker Alpha P instrument.  1, Compound 4).

Ethics Statement
Sprague-Dawley rats were obtained from the Center for Experimental Models and Transgenic Services-Freiburg (CEMT-FR). All the experiments were approved and conducted according to the guidelines of the ethics committee of the University of Freiburg under protocol Nr. X-13/06A (approved on 15 July 2013) and the study were carefully planned to minimize the number of animals used and their suffering.

Chemicals
LPS from Salmonella typhimurium (Sigma Aldrich, Deissenhofen, Germany) was resuspended in sterile phosphate-buffered saline (PBS, 5 mg/mL) as stock. Based on our previous studies [8,49,50], the final concentration of LPS used in the cultures was 10 ng/mL. KIT 10 was dissolved in DMSO.

Primary Rat Microglia Cultures
As described in our previous studies [49,50], primary mixed glial cell cultures were prepared from cerebral cortices of 1-day neonatal Sprague-Dawley rats. Under sterile conditions, brains were carefully taken, and the cerebral cortices were isolated and the meninges removed. Then, the cortices were gently dissociated and filtered through a 70 µm nylon cell strainer (BD Biosciences, Heidelberg, Germany). After centrifugation at 1000× g for 10 min, cells were collected and resuspended in Dulbecco's modified Eagle's medium (DMEM from Gibco Life Technologies, Darmstadt, Germany) containing 10% fetal calf serum (Biochrom AG, Berlin, Germany) and antibiotics (40 U/mL penicillin and 40 µg/mL streptomycin, both from Sigma-Aldrich GmbH, Taufkirchen, Germany). Cells were cultured on 10 cm cell culture dishes (Falcon, Heidelberg, Germany) with a density of 5 × 10 5 cells/mL in 10% CO 2 at 37 • C (Heracell 240i, Thermo Scientific, distributed by Omnilab, Munich, Germany). With this protocol, it is possible to isolate glial cells (microglia and astrocytes), in which microglia grow on top of the astrocyte layer. These mixed cell cultures reach confluence approximately 2 weeks post-seeding. After 12 days in vitro, microglial cells, which can be seen floating in the medium, were harvested by gentle physical shaking and re-seeded into cell culture plates according to the experimental setup. On the next day, the medium was changed to remove non-adherent cells and after 1 h, the cells were stimulated for respective experiments.

Cell Viability Assay
Viability of primary rat microglia after treatment with KIT 10 was measured by the CellTiter-Glo ® Luminescent Cell Viability Assay (Promega, Mannheim, Germany), which is used to determine the number of metabolically active and viable cells in cell culture based on quantitation of the ATP present in the cells. Briefly, cells were cultured in 96 well plates at the density of 25 × 10 3 cells/well for 24 h. Then, the medium was changed and after at least 1 h, the cells were incubated with KIT 10 for 24 h. Since this compound was dissolved in DMSO, the solvent was used in the negative control wells at a final concentration of 0.10% and as a positive control in a higher concentration (10%). The concentration of ATP was measured after 24 h of incubation by adding 100 µL of a reconstituted substrate and incubating for 10 min. Luminescence was measured using a Modulus TM II Microplate Multimode Reader (Turner BioSystems/Promega, Mannheim, Germany).

Determination of PGE 2 Release from LPS-Activated Microglia
Cultured primary rat microglia were incubated with KIT 10 (0.1-25 µM) for 30 min. Afterward, the cells were treated with or without LPS (10 ng/mL) for the next 24 h. Supernatants were harvested and levels of PGE 2 were measured using a commercially available enzyme immunoassay (EIA) kit (Cayman Chemical, Ann Arbor, MI, USA). Standard concentrations of 39-2500 pg/mL were used and the sensitivity of the assay was 36 pg/mL. The results were normalized to LPS and presented as a percentage of change in PGE 2 levels of at least three independent experiments.

Cyclooxygenases Activities Assay in Primary Microglia Culture
The COX enzymes are responsible for the formation of prostanoids through arachidonic acid (AA). Under an inflammatory stimulus, the enzyme COX-2 is one of the enzymes responsible to synthesize prostaglandins [29]. Under unstimulated conditions, primary microglial cells only express the COX-1 isoform [37].
To measure COX-1 activity, primary rat microglial cells were plated in 24-well cell culture plates. After 24 h, the medium was removed and replaced with a serum-free medium. KIT 10 (0.1-10 µM) or the selective COX-1 inhibitor SC560 (0.1-10 µM, reversible inhibitor) were added and left for 15 min. Then, 15 µM of AA (substrate) was supplemented for another 15 min. Supernatants were then collected and used for the determination of PGE 2 .
To measure COX-2 activity, primary rat microglial cells were plated in 24-well cell culture plates, and pre-incubated with LPS (10 ng/mL) for 24 h. Then, the medium was removed, and replaced with a serum-free medium. KIT 10 (0.1-10 µM) or diclofenac sodium (non-selective COX inhibitor, 10 µM) were added and left for 15 min. Then, 15 µM of AA (substrate) was added for another 15 min. Supernatants were then collected and used for the determination of PGE 2 .

Statistical Analysis
Statistical analyses were performed using Prism 5 software (GraphPad Software Inc., San Diego, CA, USA). Values of all experiments were represented as mean ± SEM of at least three independent experiments. Raw values were converted to a percentage and positive control was considered as 100%. Values were compared using one-way ANOVA with post hoc Student-Newman-Keuls test (multiple comparisons). The level of significance was set at * p < 0.05, ** p < 0.01, and *** p < 0.001. IC 50 was calculated using Prism 5 software following the nonlinear regression, "Dose-Response-Inhibition".