Identification of the Synthetic Cannabinoid R(+)WIN55,212-2 as a Novel Regulator of IFN Regulatory Factor 3 Activation and IFN-β Expression

β-Interferons (IFN-βs) represent one of the first line treatments for relapsing-remitting multiple sclerosis, slowing disease progression while reducing the frequency of relapses. Despite this, more effective, well tolerated therapeutic strategies are needed. Cannabinoids palliate experimental autoimmune encephalomyelitis (EAE) symptoms and have therapeutic potential in MS patients although the precise molecular mechanism for these effects is not understood. Toll-like receptor (TLR) signaling controls innate immune responses and TLRs are implicated in MS. Here we demonstrate that the synthetic cannabinoid R(+)WIN55,212-2 is a novel regulator of TLR3 and TLR4 signaling by inhibiting the pro-inflammatory signaling axis triggered by TLR3 and TLR4, whereas selectively augmenting TLR3-induced activation of IFN regulatory factor 3 (IRF3) and expression of IFN-β. We present evidence that R(+)WIN55,212-2 strongly promotes the nuclear localization of IRF3. The potentiation of IFN-β expression by R(+)WIN55,212-2 is critical for manifesting its protective effects in the murine MS model EAE as evidenced by its reduced therapeutic efficacy in the presence of an anti-IFN-β antibody. R(+)WIN55,212-2 also induces IFN-β expression in MS patient peripheral blood mononuclear cells, whereas down-regulating inflammatory signaling in these cells. These findings identify R(+)WIN55,212-2 as a novel regulator of TLR3 signaling to IRF3 activation and IFN-β expression and highlights a new mechanism that may be open to exploitation in the development of new therapeutics for the treatment of MS.

␤-Interferons (IFN-␤s) represent one of the first line treatments for relapsing-remitting multiple sclerosis, slowing disease progression while reducing the frequency of relapses. Despite this, more effective, well tolerated therapeutic strategies are needed. Cannabinoids palliate experimental autoimmune encephalomyelitis (EAE) symptoms and have therapeutic potential in MS patients although the precise molecular mechanism for these effects is not understood. Toll-like receptor (TLR) signaling controls innate immune responses and TLRs are implicated in MS. Here we demonstrate that the synthetic cannabinoid R(؉)WIN55,212-2 is a novel regulator of TLR3 and TLR4 signaling by inhibiting the pro-inflammatory signaling axis triggered by TLR3 and TLR4, whereas selectively augmenting TLR3-induced activation of IFN regulatory factor 3 (IRF3) and expression of IFN-␤. We present evidence that R(؉)WIN55,212-2 strongly promotes the nuclear localization of IRF3. The potentiation of IFN-␤ expression by R(؉)WIN55,212-2 is critical for manifesting its protective effects in the murine MS model EAE as evidenced by its reduced therapeutic efficacy in the presence of an anti-IFN-␤ antibody.

R(؉)WIN55,212-2 also induces IFN-␤ expression in MS patient peripheral blood mononuclear cells, whereas down-regulating inflammatory signaling in these cells. These findings identify R(؉)WIN55,212-2 as a novel regulator of TLR3 signaling to IRF3 activation and IFN-␤ expression and highlights a new mechanism that may be open to exploitation in the development of new therapeutics for the treatment of MS.
IFN-␤ is one of several immunomodulatory drugs currently available to treat patients with relapsing-remitting MS 3 (1), dis-playing significant beneficial effects on disability progression (2) and relapse rate (3). The mechanism(s) of action of IFN-␤ is clearly complex with demonstrated effects on antigen presentation, co-stimulatory molecule expression, T-cell proliferation, and leukocyte migration (4). Despite its success in the clinic, IFN-␤ therapy has demonstrated partial efficacy along with various side effects (4), indicating a pressing need for more effective strategies.
Cannabis (Cannabis sativa) has a long history of consumption therapeutically (5). The term "cannabinoid" incorporates the active components of C. sativa, the plant-derived cannabinoids, the endogenous cannabinoids (endocannabinoids), and the synthetic cannabinoid ligands. Cannabinoids are used for the treatment/management of inflammatory conditions including MS (6), arthritis (7), and glaucoma (8). Indeed Sativex (a combination of two plant-derived cannabinoids, tetrahydrocannabinol and cannabidiol) is currently approved for the neuropathic pain and spasticity associated with MS (9). Despite the growing clinical use of cannabinoids their mechanism(s) of therapeutic action are not fully elucidated.
Cannabinoids elicit their effects via cannabinoid receptors (CB 1 and CB 2 ) (10,11). However, some cannabinoid-induced effects are mediated independently of these receptors (12). Cannabinoid receptors are localized throughout the central nervous system (CNS) (13) and on immune cells associated with neuroinflammation (14). This is particularly relevant as cannabinoids therapeutically impact diseases associated with a dysregulation of the immune and nervous systems (13). Indeed in experimental autoimmune encephalomyelitis (EAE) cannabinoids attenuate the development of disease (15). The roles of CB 1/2 in mediating these effects varies depending on the pharmacological profile of the cannabinoid (16). Furthermore, whereas CB 1 confers neuroprotection in the CNS (17), the CB 2 receptor plays a pivotal protective role in the periphery by regulating T-cell effector function and myeloid progenitor trafficking into the CNS (16,18).
TLRs are single transmembrane receptors involved in the recognition of bacterial/viral products and induce signaling involving the activation of transcription factors, such as NF-B, and induction of genes encoding IFNs and cytokines (19). To date, 13 mammalian TLRs have been identified, and with the exception of TLR3, all TLRs recruit the adaptor myeloid differentiation factor 88 (MyD88) (20). TLR3 (and TLR4) induces MyD88-independent signaling to regulate NF-B via Toll-interleukin-1 receptor (TIR)-domain-containing adaptor-inducing IFN-␤ (TRIF) protein. Such TRIF-mediated signaling constitutes the MyD88-independent pathway and in addition to stimulating NF-B, this pathway promotes phosphorylation of transcription factors IRF3 and IRF7, via two kinases, TRAF family member-associated NF-B activator (TANK)-binding kinase 1 (TBK1) and inducible IB kinase (21). The phosphorylation of IRF3/7 promotes their nuclear translocation and induction of type I IFNs (19). With respect to MS, specific roles of TLRs have been shown in EAE (22), with changes in TLR expression observed in MS brain lesions (23).
Because IL-1 signaling is sensitive to R(ϩ)WIN55,212-2 (24) and the IL-1R and TLRs contain a homologous Toll/IL-1R (TIR) domain (25), we aimed to evaluate the effects of R(ϩ)WIN55,212-2 on TLR signaling, with particular focus on the molecular mechanism controlling the induction of IFN-␤. Protective roles in EAE have been demonstrated for TLR3 (26) and TLR4 (27) and thus we focused on the effects of R(ϩ)WIN55,212-2 on these pathways. We show that whereas R(ϩ)WIN55,212-2 negatively regulates the activation of NF-B in response to TLR3/4, it enhances TLR3-induced IRF3 activation and IFN-␤ expression. We further show that R(ϩ)WIN55,212-2-induced expression of IFN-␤ mediates its protective effects in EAE. Finally, evidence is presented that the positive effects of R(ϩ)WIN55,212-2 on IFN-␤ is apparent in cells from MS patients. This study thus identifies a novel regulatory pathway that may be open to exploitation in the therapeutic treatment of MS.

EXPERIMENTAL PROCEDURES
Cell Culture-HEK293 cells stably expressing the TLR3 and TLR4 receptors were from InvivoGen (Toulouse, France). Human U373 astrocytoma cells stably transfected with CD14 (U373-CD14) and bone marrow-derived macrophages (BMDMs) from wild type and TRIF-deficient mice were gifts from Dr. Katherine Fitzgerald (University of Massachusetts Medical School, Boston, MA). Cell lines were maintained in DMEM supplemented with 10% FBS, 100 g/ml of penicillin, and 100 g/ml of streptomycin. Cells were maintained in a 37°C humidified atmosphere with 5% CO 2 . The neomycin analog G418 (500 g/ml) was used to select for the stably transfected TLR cell lines and maintenance of CD14 expression. Primary astrocytes were prepared as previously described (28) from the whole brain of 1-day-old C57/BL6 mice in accordance with the guidelines laid down by the local ethical committee (National University of Ireland, Maynooth). Briefly, astrocytes were isolated from mixed glia at days 10 -14 by removing nonadherent cells with mechanical shaking and harvesting by trypsinization (0.25% trypsin, 0.02% EDTA). Cells were centrifuged (2,000 ϫ g for 5 min at 20°C) and the astrocyte-enriched pellet resuspended in DMEM. Astrocytes were plated (2 ϫ 10 5 cells/ml) on 6-or 12-well plates and treated 24 h later. R(ϩ)WIN55,212-2 and S(Ϫ)WIN55,212-2 (Sigma) were initially dissolved in DMSO and stored as 5 mM stock solutions.
For culture use, the stock drug was diluted to a final concentration in culture medium and DMSO (Յ0.1%) was used as vehicle control.
Patients and Blood Samples-Healthy donors and MS patients attending outpatient clinics at Queens Medical Centre University Hospital, University of Nottingham, UK, were recruited for this study. Written informed consent was obtained from each patient and the study received ethical approval from the Nottingham Research Ethics Committee. Patients with relapsing-remitting MS were clinically stable with an age ranging between 38 and 56 years (mean 48.4 Ϯ 8.3; n ϭ 3). Patients were naive to any disease modifying therapies including IFN-␤, glatiramer acetate, and natalizumab. Healthy individuals were recruited from the University of Nottingham (mean age 31 Ϯ 2.6; n ϭ 3). Venous blood (30 ml) was obtained from each subject. PBMCs were isolated using the Ficoll-Hypaque isolation technique and plated (1 ϫ 10 6 cells/ml) on 24-well plates.
Histology-Spinal cords were dissected and fixed in 10% formaldehyde saline. Spinal cords were sectioned and stained with hematoxylin and eosin for inflammatory scoring (31). Inflammatory scores were as follows: 0, no inflammatory cells; 1, a few scattered inflammatory cells; 2, perivascular cuffing; 3, perivascular cuffing with extensions into adjacent parenchyma, or parenchymal infiltration without obvious cuffing. Demyelination was assessed on Luxol fast blue-stained spinal cord sections and scored as follows: 0, no evident demyelination; 1, decreased myelination with no foci; 2, obvious demyelination with evident foci; 3: severe demyelination. An investigator blinded to the treatment groups scored all stained sections, with slides labeled by mouse barcode number.
Western Immunoblotting-Astrocytes were seeded in 6-well plates (2 ϫ 10 5 cells/ml). Cells were treated with poly(I⅐C) (25 g/ml) for 5-360 min or pre-treated with R(ϩ)WIN55,212-2 (20 M) for 1 h prior to poly(I⅐C) (25 g/ml) exposure for 1 h. Cells were then washed in ice-cold PBS before being lysed on ice for 10 min in 150 l of lysis buffer (20 mM HEPES, pH 7.4, containing 10 mM KCl, 1.5 mM MgCl 2 , 1 mM EDTA, 1 mM EGTA, 1 mM dithiothreitol, 0.1 mM PMSF, pepstatin A (5 g/ml), leupeptin (2 g/ml), and aprotinin (2 g/ml)). Cell lysates were centrifuged at 13,000 ϫ g for 15 min at 4°C. The supernatant was mixed with SDS-PAGE sample buffer (0.125 Tris-HCl, pH 6.8, 20% (v/v) glycerol, 4% (w/v) SDS, 1.4 M ␤-mercaptoethanol, and 0.0025% (w/v) bromphenol blue). For in vivo experiments samples of spinal cord were homogenized in lysis buffer and the resulting lysate was centrifuged (16,000 ϫ g for 15 min at 4°C). Supernatants were then further centrifuged (100,000 ϫ g for 1 h at 4°C) and the supernatant (cytosolic fraction) added to sample buffer. All samples in sample buffer were boiled for 10 min and separated on 10% SDS-PAGE gels. Proteins were transferred to nitrocellulose membrane (Sigma) and blocked for 1 h in 5% dried milk. Membranes were incubated overnight at 4°C with mouse monoclonal phospho-IB␣ antibody (1:1,000 in 5% dried milk; Cell Signaling Technology Inc., Danvers, MA), rabbit monoclonal phospho-Ser 396 IRF3 antibody (1:750 in 2.5% BSA; Cell Signaling Technology Inc.), rabbit monoclonal total IRF3 antibody (1:1,000 in 2.5% BSA; Cell Signaling Technology Inc.), or mouse monoclonal IB␣ antibody (1:200 in 5% dried milk; Santa Cruz Biotechnology, Santa Cruz, CA). Membranes were washed and incubated with anti-mouse or anti-rabbit IRDye Infrared secondary antibody (1:5,000 in 5% dried milk; Licor Biosciences, Lincoln, NE) for 1 h in the dark at room temperature. The membranes were then washed and immunoreactive bands were detected using the Odyssey Infrared Imaging System (Licor Biosciences). Membranes were stripped and incubated with mouse monoclonal anti-␤-actin antibody (1:10,000; overnight at 4°C, Sigma). Molecular weight markers were used to calculate molecular weights of proteins represented by immunoreactive bands. Densitometry was performed using ImageJ software, and values were normalized for protein loading relative to levels of ␤-actin or total IRF3.
Screening of Cannabinoid Receptor Expression-Total cellular RNA was prepared from HEK293 cells, cDNA was generated as above and PCR amplification was performed to selectively amplify regions of CB 1 , CB 2 , and GAPDH cDNA.
cAMP Assay-HEK293 cells were pre-treated with or without PTX (100 ng/ml; 24 h), SR141716 (SR1; 1 M for 1 h), and SR144528 (SR2; 1 M for 1 h) prior to treatment with the selective CB 1 agonist ACEA (100 nM for 1 h; Tocris Bioscience, Bristol, UK) or the selective CB 2 agonist JWH133 (100 nM for 1 h; Sigma). Cells were then incubated with the potent cAMP phosphodiesterase inhibitor, 3-isobutyl-1-methylxanthine (500 M for 15 min; Sigma) and stimulated with forskolin (30 M for 30 min; Sigma) to induce cAMP. Lysates were harvested and assessed for levels of intracellular cAMP using a cAMP parameter kit as per the manufacturer's instructions (R&D Systems).
Confocal Microscopic Analysis of IRF3-For characterizaton of endogenous IRF3, primary astrocytes were seeded (1 ϫ 10 5 cells/ml) in 4-well chamber slides (Lab-Tek, Roskilde, Denmark) and grown for 24 h. Cells were pre-treated with R(ϩ)WIN55,212-2 (20 M) or S(Ϫ)WIN55,212-2 (20 M) for 1 h prior to poly(I⅐C) (25 g/ml) exposure for 1 h. Cells were fixed in 4% paraformaldehyde, permeabilized with 0.2% Triton X-100 in PBS for 10 min at room temperature, and blocked with 10% goat serum (Vector Laboratories, Peterborough, UK) for 2 h. Cells were treated overnight at 4°C with rabbit polyclonal IRF3 antibody (1:200 in 5% goat serum; Santa Cruz Biotechnology). Cells were washed and incubated with goat anti-rabbit Alexa 488 secondary antibody (1:500 in 5% goat serum; Invitrogen) and DAPI (1.5 g/ml) in PBS, washed, and mounted (Vectashield, Vector Laboratories). All samples were viewed using an Olympus FluoView FV1000 confocal laser scanning microscope equipped with the appropriate filter sets. Acquired images were analyzed using the Olympus FV-10 ASW imaging software. Negative control experiments were performed by replacing the primary antibody with isotype controls (Millipore) and using equal gain settings during acquisition and analysis.
Statistical Analysis-Data are expressed as mean Ϯ S.E., and the results represent two or three independent experiments. Statistical comparisons of different treatments were done by a one-way analysis of variance using a post hoc Student's Newman-Keuls test. Differences with a p value less than 0.05 were considered statistically significant.

R(ϩ)WIN55,212-2 Augments TLR3-induced IRF3 Activation and IFN-␤ Induction in a Cannabinoid Receptor-independent
Manner-We next characterized the cannabinoid pharmacology underlying the above effects. Receptor expression was first confirmed on HEK293 cells (Fig. 3A) and receptor involvement was addressed using the CB 1 and CB 2 antagonists, SR141716 and SR144528, respectively. Pre-exposure to SR141716 (Fig. 3, B and C) or SR144528 (Fig. 3, D and E), failed to attenuate the ability of R(ϩ)WIN55,212-2 to potentiate poly(I⅐C)-induced activation of IRF3 (Fig. 3, B and D) and expression of IFN-␤ mRNA (Fig. 3, C and E). This indicates that R(ϩ)WIN55,212-2 impacts the TLR3-IRF3-IFN-␤ axis independently of CB 1/2 . Because CB 1/2 receptors signal via G i proteins, we employed the G i inhibitor PTX to validate this finding. PTX had no effect on the stimulatory effect of R(ϩ)WIN55,212-2 on poly(I⅐C)-induced activation of IRF3 (Fig. 3F) and expression of IFN-␤ (Fig.  3G), confirming that R(ϩ)WIN55,212-2 is acting in a cannabinoid receptor-independent manner. Both CB 1 and CB 2 antagonists and PTX were active in our system as they prevented the inhibitory effects of specific CB 1 and CB 2 agonists on forskolininduced cAMP production (Fig. 3H).

DISCUSSION
Here we aimed to understand the molecular mechanisms of the immunomodulatory effects of the cannabinoid R(ϩ)WIN55,212-2 and in so doing we have identified an important regulatory pathway that may be able to control pathogenesis in MS. We propose that R(ϩ)WIN55,212-2 controls the expression of IFN-␤. In addition to ameliorating proinflammatory signaling induced by TLR3/4, R(ϩ)WIN55,212-2 augments TLR3 signaling, enhancing IFN-␤ expression that ameliorates the pathology associated with EAE. We also demonstrate that cells from MS patients are especially sensitive to R(ϩ)WIN55,212-2 in terms of increased expression of endog- Transfected cells were left overnight and treated in the absence or presence of R(ϩ)WIN55,212-2 (5-50 M) for 6 h. Cell lysates were assayed for firefly luciferase activity and normalized for transfection efficiency using Renilla luciferase activity. Data are presented relative to vehicle-treated cells and represent the mean Ϯ S.E. of triplicate determinations from three independent experiments. B, TRIF-deficient BMDMs were pre-treated (1 h) with R(ϩ)WIN55,212-2 (20 M) and then stimulated with poly(I⅐C) (25 g/ml) for 18 h. cDNA was generated and assayed by quantitative real time PCR for levels of Ifn-␤ mRNA. The expression level of Ifn-␤ was normalized relative to expression of the housekeeping gene Gapdh and represents the mean Ϯ S.E. of triplicate determinations from three independent experiments. C and D, primary mouse astrocytes were seeded into 6-well plates and treated with poly(I⅐C) (25 g/ml) (C) for various time points (5-360 min) or pre-treated with R(ϩ)WIN55,212-2 (20 M; 1 h) (D) prior to stimulation with poly(I⅐C) (25 g/ml) for 1 h. Cell lysates were subsequently subjected to Western immunoblotting using anti-phospho-Ser 396 IRF3, anti-total IRF3, and anti-␤-actin antibodies (lower panels). All immunoblots were subjected to densitometric analysis with levels of phospho-IRF3 normalized to total levels of IRF3 (upper panels). Densitometic data are representative of 8 (C) and 6 (D) independent experiments. *, p Ͻ 0.05; **, p Ͻ 0.01; and ***, p Ͻ 0.001 versus non-transfected (A) and vehicle-treated cells (C and D). ϩ, p Ͻ 0.05; and ϩϩϩ, p Ͻ 0.001 compared with vehicle treated TRIF-transfected cells. E, primary mouse astrocytes were grown in chamber slides and pre-treated (1 h) with R(ϩ)WIN55,212-2 (20 M) or S(Ϫ)WIN55,212-2 (20 M) for 1 h prior to poly(I⅐C) (25 g/ml) exposure for 1 h. Cells were fixed, mounted in anti-fade medium with DAPI, and visualized using confocal microscopy. Confocal images were captured using a UV Zeiss 510 Meta System laser scanning microscope equipped with the appropriate filter sets. Data analysis was performed using the LSM 5 browser imaging software. Images are representative of three independent experiments. Scale bars are 20 m. F, primary astrocytes were pre-treated with or without R(ϩ)WIN55,212-2 (20 M) for 1 h prior to stimulation in the absence or presence of poly(I⅐C) (25 g/ml; 1 h). Cytosolic and nuclear fractions were prepared and subsequently subjected to Western immunoblotting using anti-total IRF3 and anti-␤-actin antibodies. Blots are representative of data obtained from 6 animals. post-immunization, with disease severity peaking on day 16 followed by a relapse on day 26. Mice treated with R(ϩ)WIN55,212-2 (administered (20 mg/kg) intraperitoneally on days 0, 1, 2, 3, 4, and 5 after immunization) showed delayed development of EAE and attenuated disease severity. PLP-immunized mice treated with R(ϩ)WIN55,212-2 and an anti-IFN-␤ antibody (administered intraperitoneally (2 ϫ 10 3 neutralizing units) on days 3 and 5 after PLP immunization) were not protected. B, representative images of Luxol fast blue-stained spinal cord sections from untreated mice, PLP-treated, PLP ϩ WIN-treated, and PLP ϩ WIN ϩ ␣IFN␤-treated mice illustrating the extent of demyelination and lymphocytic inflammation. The posterior funiculi of the spinal cord were observed under high power (right panels). Images are representative of data from 4 to 8 animals per treatment group. Scale bars are 200 and 50 m. Spinal cords were sectioned and stained with hematoxylin and eosin and quantified for spinal cord inflammation (C) and extent of demyelination (D) using Luxol fast blue-stained spinal cord sections in treated groups. cDNA was generated from spinal cords and assayed by quantitative real time PCR for relative levels of Gfap mRNA (E) and Cd11b mRNA (F) from vehicle-treated, PLP-treated, PLP ϩ WIN-treated, and PLP ϩ WIN ϩ ␣IFN␤-treated mice. The expression level of Gfap and Cd11b was normalized relative to expression of the housekeeping gene Gapdh and represent the mean Ϯ S.E. of triplicate determinations from 4 to 8 animals per treatment group. Cytosolic fractions were prepared from the spinal cord of vehicle-treated, PLP-treated, PLP ϩ WIN-treated, and PLP ϩ WIN ϩ ␣IFN␤-treated mice. Cell lysates were subsequently subjected to Western immunoblotting using anti-phospho IB␣, anti-total IB␣, and anti-␤-actin antibodies. Blots are representative of data from 4 to 8 animals per treatment group. *, p Ͻ 0.05; **, p Ͻ 0.01; and ***, p Ͻ 0.001 for differences between WIN-treated mice and other groups.
enous IFN-␤ and this strongly indicates the mechanism described has relevance to treatment of MS.
The study highlights the anti-inflammatory potential of R(ϩ)WIN55,212-2 by virtue of its inhibitory effects on the NF-B pathway. We have previously shown that R(ϩ)WIN55,212-2 blocks the IL-1 pathway leading to NF-B (24) and here we demonstrate for the first time that it can inhibit TLR3/4-induced activation of NF-B. This likely makes a major contribution to the inhibitory effects of R(ϩ)WIN55,212-2 on pro-inflammatory gene expression. Indeed, we demonstrate that R(ϩ)WIN55,212-2 blunts TLR3/4 induction of TNF-␣. Such effects translate into strong antiinflammatory activity in vivo. Thus R(ϩ)WIN55,212-2 blunts neutrophil migration in a mouse peritonitis model (42), whereas R(ϩ)WIN55,212-2 abrogates the clinical development of EAE (30). The inhibitory effects of R(ϩ)WIN55,212-2 on leukocyte adhesion to endothelia is likely to contribute to its therapeutic properties in EAE (43). However, whereas these direct anti-inflammatory effects of R(ϩ)WIN55,212-2 are pivotal, the present study highlights a novel mechanistic basis to its protective effects by virtue of its ability to induce endogenous IFN-␤.
We provide evidence for the first time that IRF3 is a target for synthetic cannabinoids. We propose that R(ϩ)WIN55,212-2 can enhance IRF3 nuclear localization and positively impact on IFN-␤ expression in response to TLR3 signaling. Intriguingly, R(ϩ)WIN55,212-2 exerts differential effects on LPS-and poly(I⅐C)-induced activation of IRF3 and expression of IFN-␤.