We used female Biozzi ABH and ABH mice lacking the CB₁ receptor (Cnr1) gene, susceptible to TMEV-IDD development, gently gifted by Dr. Baker (University College London). Mice were maintained on food and water ad libitum in a 12 hours dark-light cycle. Four-to six week-old mice were inoculated intracerebrally in the right cerebral hemisphere with 10⁶ plaque forming units (PFU) of Daniel's (DA) TMEV strain, in 30 μl of Dulbecco's modified Eagle's medium supplemented with 10% of fetal calf serum (FCS) as previously described 21 25 . Handling of animals was performed in compliance with the guidelines of animal care set by the European Union (86/609/EEC) and the Spanish regulations (BOE67/8509-12; BOE1201/2005) on the use and care of laboratory animals, and approved by the local Animal Care and Ethics Committee of the CSIC.

At the time of TMEV infection, the mice were treated with UCM-707 (3 mg/kg, injected i.p.) twice a day (morning and afternoon) for 3 consecutive days or appropriate vehicle (5% BSA and 0.2% DMSO in phosphate-buffered saline). This dose was chosen on the basis of previous studies in our laboratory 18 .

Animal tissue was processed as previously described 24 . Briefly, mice were perfused transcardially with saline. Brains were fixed in 4% paraformaldehyde in 0.1 M PB, washed in 0.1 M PB, cryoprotected with a 7%, 15% and later 30% solution of sucrose in 0.1 M PB and frozen at -80°C until used. Free-floating coronal brain sections (30 μm thick) were processed as described previously 24 to visualize the adhesion molecule VCAM-1 (anti-VCAM-1 antibody; BD Pharmingen, San Diego, CA) and microglia (Iba-1 antibody; Wako Chemical Pure Industry, GmbH). Immunostaining was visualized with the corresponding secondary antibodies conjugated with avidin-peroxidase (Dako, Barcelona, Spain) and revealed with the choromogen 3.3' diaminobenzidine tetrahydrochloride (DAB; Sigma-Aldrich Inc, St. Louis, MO, USA) followed by counterstaining with toluidine blue. In all cases specificity of staining was confirmed by omitting the primary antibody. To quantify VCAM-1 expression fluorescence secondary antibody was used and six confocal immunofluorescence microphotograps per level were analyzed using the Image J software designed by National Institutes of Health. Results are presented as intensity of staining per vessel in case of VCAM-1 study or percentage of area occupied by CD11b⁺ staining per field in case of microglial analysis.

b.End5: Murine brain endothelial cells (b.End5) which are recognized to present brain endothelium like properties were obtained from European Collection of Cell Cultures (UK). This cell model is an appropriate choice to study blood-brain barrier function 26 27 28 . The cells were grown in Dulbeccos's Modified Eagle's Medium supplemented with 10% heat inactivated fetal bovine serum (FBS); 1% nonessential aminoacid, 1% sodium pyruvate and 1% antibiotic penicillin and streptomycin (all from Gibco, Scotland, UK) and were maintained under standard cell culture conditions at 37°C and 5% CO₂. One hour before experiments, cells were subjected to restricted conditions (1% FBS). In order to assess the possible receptors involved in the effects of AEA, one hour before the treatment with AEA (10 μM) and TMEV (2 × 10⁵ pfu), cells were pre-treated with the cannabinoid receptors antagonists SR141716A (CB₁, 1 μM); AM630 (CB₂, 1 μM); capsazepine (TRPV1, 10 μM) or GW9662 (PPARγ, 100 nM, 1 μM).

Astrocytes: cell cultures were obtained as previously described 29 . Forebrains were dissociated mechanically, filtered through a 150 μm nylon mesh, resuspended in DMEM containing 10% heat-inactivated FCS, 10% heat-inactivated FBS and 1% penicillin/streptomycin and plated on poly-L-lysin-coated (5 μg/ml) 75 cm² flasks (Nunc, Wiesbaden, Germany). After 7 days in culture the flasks were shaken at 260 rpm at 37°C overnight to remove microglia and oligodendrocytes.

Leukocytes: Lymphatic nodes were homogenized in cold PBS with the plunger of a syringe, filtered through a 70 μm cell strainer to obtain a single cell suspension, centrifuged for 5 min at 1200 rpm and resuspended in RPMI supplemented with 10 mM HEPES (pH 7.4), 2 mM glutamine and 10% FCS, β-mercaptoethanol (50 μM).

Confluent brain endothelial cell monolayer infected with TMEV (2 × 10⁵ pfu) and treated with AEA (10 μM) was subjected or not to the cannabinoid receptors antagonist by pre-treatment for 1 hour with the CB₁ or CB₂ selective receptor antagonist, SR141716A (SR1, 1 μM) or AM630 (1 μM), respectively. After 6 hours, 2.5 × 10⁵ leukocytes stained with calcein acetoxymethyl ester (AM) (5 μM) (Sigma-Aldrich Inc, St. Louis, MO, USA) were allowed to adhere to endothelial monolayer for 20 hours. Lapsed this time non bound leukocytes were removed, five microphotographs/field, fluorescence and phase contrast, were used for counting adhered leukocytes by Metamorph software. The assay was performed in triplicate for each value and was repeated 3 times. Calcein acetoxymethyl ester is a vital dye what is membrane permeable but becomes membrane impermeable and fluorescent when cleaved by intracellular sterases.

Blood brain barrier model was performed as described previously 30 with modifications. Briefly, transwell filters (surface area 6.4 mm; pore size, 8 μm; BD Falcon™ Cell Culture Inserts) were coated with colagen type I (50 μg/ml; BD Falcon) and fibronectine (50 μg/ml; Invitrogen, Barcelona, Spain). Astrocytes (5 × 10⁴ cell/well) were allowed to adhere to the bottom of the filter for 10 minutes in DMEM with 10% FBS, 10% FCS and 1% penicillin/streptomycin. Contamination of adherent astrocytes on the bottom of the well was avoided. After 24 hours, brain endothelial cells (b.End5) were seeded on the top of the filter at a density of 5 × 10⁴ cell/well in DMEM with 10% FBS, 1% non-essential aminoacid, 1% sodium pyruvate and 1% penicillin/streptomycin. We consider that BBB was established when transendothelial electrical resistance was close to 200Ω/cm² 30 . Once confluent, endothelial cells were infected with TMEV (2 × 10⁵ pfu) and treated with AEA (10 μM). To study the involvement of cannabinoid receptor, cells were pre-treated with the CB₁ or CB₂ receptors antagonist, SR1 (1 μM) or AM630 (1 μM) respectively. Following stimulation for 6 hours, 2.5 × 10⁵ leukocytes were added on the top of the insert for 20 hours. The entire transmigrating cell populations present in the bottom chamber were collected and counted by using a hemocytometer. For schematic illustration of the BBB model see additional file 1A and additional file 2.

Permeability assay was performed as described 31 . Briefly, after rinsed in phenol-red-free DMEM the top and the bottom of the filter, 400 μl of 10% FBS/phenol-red-free DMEM and 200 μl of 0.45% albumin conjugated to Evan's blue dye were added to the bottom and the top of the well, respectively and incubated at 37°C for 30 min. Absorbance of the bottom medium was read at 620 nm [see Additional file 1B].

To visualize the tight junction zonula occludens-1 (ZO-1) in the blood brain barrier model, cells were fixed with 4% paraformaldehyde, washed with PBS and incubated overnight at 4°C with the primary antibody (ZO-1, Zymed Laboratorioes, Carlsbad, CA) in PBS containing 5% NGS and 0,1 Triton X-100. After washing with PBS, cells were incubated for 1 h at RT with secondary anti-rabbit antibody IgGs, conjugated with Alexa 488 (Molecular Probes, Eugene, OR, USA) washed with PBS and mounted on glass slides with fluorescent mounting medium. In all cases, specificity of staining was confirmed by omitting the primary antibody [see Additional file 1C].

Soluble fraction VCAM-1 (sVCAM-1) content in endothelial cells supernatants was measured by solid phase sandwich ELISA, using a monoclonal antibody specific for mouse sVCAM-1 (R & D Systems Inc., MN, USA), according to the manufacturer's instructions. The assay sensitivity was 30 pg/ml.

All results are presented as mean ± SEM. For in vitro experiments the n value corresponds at least to three independent experiments; with triplicate determinations in each experiment. One-way ANOVA, followed by a post hoc Tukey's multiple comparison tests was used to examine the statistical significance of in vitro assays. Repeated measure test and post hoc Duncan test was used to analyze the statistical significance of VCAM-1 and CD11b studies. p values < 0.05 were considered significant.

The endothelial blood brain barrier protects the CNS from the changing environment in both physiologic and pathologic conditions. Previous work in our lab has demonstrated that sVCAM-1 is constitutively expressed on b.End5 cells and increased by TMEV infection 24 . We first analyzed the effect of AEA on the production of VCAM-1 by TMEV-infected brain endothelial cells. Dose response studies of AEA on sVCAM-1 production showed that 10 μM was the most effective dose to prevent the expression of VCAM-1 induced by TMEV at 20 hours postinfection (Figure 1A). AEA also inhibited VCAM-1 expression in resting cells (data not shown). Down-regulation of VCAM-1 induced by AEA (10 μM) was partially reversed by the addition of the CB₁ receptor antagonist, SR141716A (SR1) but not by the CB₂ receptor antagonist AM630 (Figure 1B). The doses used for CB antagonists were 1 μM on the basis of their capability for antagonizing CB effects in our previous work. To examine if vanilloid receptors expressed in brain endothelial cells 32 were involved in AEA inhibition of VCAM-1 we pretreated the cells with capsazepine (10 μM). As shown in Figure 1C, the blockade of vanilloid receptors did not modify the inhibitory effect of AEA on VCAM-1 expression. In addition we explored the role of PPARγ receptors as it has been described to mediate some of the actions of AEA [reviewed by 33 ]. In our study the treatment with the inhibitor of PPARγ, GW9662 (at nanomolar and micromolar doses) did not prevent AEA-induced VCAM-1 inhibition (Figure 1D). In conclusion, AEA-induced inhibition of VCAM-1 in brain endothelial cells implies the activation of CB₁ receptors.

VCAM-1 is critically involved in leukocyte transmigration into the CNS. Therefore, our next step was to assess the functional relevance of AEA-induced VCAM-1 inhibition in leukocyte transmigration. First, we showed that leukocyte adhesion to TMEV-infected endothelium was significantly increased (p < 0.01) in comparison to cell adhering to resting cell monolayer. Importantly, the treatment with AEA (10 μM), at the time of virus infection diminished leukocyte adhesion (p < 0.01) (Figure 2A). In agreement with the involvement of CB₁ receptors in AEA-induced VCAM-1 inhibition, the pretreatment with the CB₁ receptor antagonist (SR1), but not with the CB₂ antagonist (AM630), reversed the inhibitory effect of AEA on leukocyte adhesion (Figure 2A). Quantification is presented as a ratio of number of leukocyte adhered to the endothelial cell monolayer in each group normalized to control group (Figure 2B).

Next, we analyzed whether the effect of AEA on the adhesion of leukocytes interferes on leukocyte transmigration through the BBB model. As expected TMEV-infection of brain endothelial cells increased the number of leukocytes crossed the BBB model referred to control (Figure 2C). Accordingly to our results in the experiments of leukocytes adhesion the treatment of endothelial cells with AEA (10 μM) diminished leukocyte crossing by a mechanism that involves CB₁ receptors. Figure 2D shows the quantification data on the number of leukocytes that cross the BBB model.

On the basis of our in vitro results, we next analyze the effect of the pharmacological modulation of the AEA tone on VCAM-1 response against TMEV infection in vivo, using wild type and CB₁ knockout mice (Cnr1^-/-). Accordingly to other studies 24 , VCAM-1 expression was not detected, by immunohistochemistry, in the brains of sham animals in both type of mice, Cnr1^+/+ or Cnr1^-/-. The intracranial injection of TMEV induced the expression of VCAM-1 in the ipsilateral cerebral cortex surrounding blood vessels close to the site of injection in both type of mice, Cnr1^+/+ as well as Cnr1^-/- mice (Figure 3A). Corroborating our in vitro findings, the treatment with the inhibitor of AEA uptake UCM-707 induced a significant reduction of VCAM-1 expression in TMEV-infected mice (Figure 3A). Although, UCM-707 decreased VCAM-1 expression in Cnr1^+/+ and Cnr1^-/- mice, quantification analysis (Figure 3B) revealed that the degree of VCAM-1 reduction in the ipsilateral cerebral cortex of Cnr1^+/+ mice was significantly higher than that observed in Cnr1^-/- (p < 0.05). This observation suggests the participation of CB₁ receptors in the effects of UCM-707 treatment. Additionally, when we analysed the contralateral hemisphere (Figure 3C) we found that only mice lacking CB₁ receptors showed increased VCAM-1 expression in the vasculature in response to TMEV that was significantly inhibited by the treatment with UCM-707 as revealed the quantification of staining intensity (Figure 3D).

The intracranial injection of Theiler's virus induced an increase of microglia with reactive morphology in the cerebral cortex at the level of infection (medium level) but only in the ipsilateral infected hemisphere (Figure 4A). Interestingly, the microglial response was exacerbated in Cnr1^-/- mice (Figure 4B), now extending from prefrontal cortex (rostral level) to hippocampal level (caudal level). When we analyzed the contralateral hemisphere we found that microglial cells did not show reactive morphology at the three brain levels examined in the wild type mice as well as in Cnr1^-/- mice. Therefore, in response to TMEV infection, activation of microglial cells only occurred in the ipsilateral hemisphere. Quantification analysis of percentage of area occupied by microglia per field was summarized in Figure 4C.

The treatment with UCM-707 significantly (p < 0.01) reduced the presence of microglia with reactive morphology in Cnr1^+/+ mice (Figure 5B) at the medium level close to the site of injection (Figure 5A). Cerebral cortex sections from Cnr1^-/- mice showed a tendency toward diminishing microglia reactivity but without reaching statistical significance (p = 0,07; Figure 5C). The analysis of the contralateral hemispheres didn't reveal the presence of microglia with reactive morphology (data not shown).