Effects of pro-inflammatory cytokines on cannabinoid CB1 and CB2 receptors in immune cells
Abstract
Aims
To investigate the regulation of cannabinoid receptors CB1 and CB2 on immune cells by pro-inflammatory cytokines and its potential relevance to the inflammatory neurological disease, multiple sclerosis (MS).
CB1 and CB2 signalling may be anti-inflammatory and neuroprotective in neuroinflammatory diseases. Cannabinoids can suppress inflammatory cytokines but the effects of these cytokines on CB1 and CB2 expression and function are unknown.
Methods
Immune cells from peripheral blood were obtained from healthy volunteers and patients with MS. Expression of CB1 and CB2 mRNA in whole blood cells, peripheral blood mononuclear cells (PBMC) and T cells was determined by quantitative real-time polymerase chain reaction (qRT-PCR). Expression of CB1 and CB2 protein was determined by flow cytometry. CB1 and CB2 signalling in PBMC was determined by Western blotting for Erk1/2.
Results
Pro-inflammatory cytokines IL-1β, IL-6 and TNF-α (the latter likely NF-κB dependently) can upregulate CB1 and CB2 on human whole blood and peripheral blood mononuclear cells (PBMC). We also demonstrate upregulation of CB1 and CB2 and increased IL-1β, IL-6 and TNF-α mRNA in blood of patients with MS compared with controls.
Conclusion
The levels of CB1 and CB2 can be upregulated by inflammatory cytokines, which can explain their increase in inflammatory conditions including MS.
Activation of cannabinoid receptors, CB1 and CB2 by endocannabinoids or exogenous ligands can suppress inflammation (Tanasescu & Constantinescu 2010). The endocannabinoid (EC) system consists of the G-protein-coupled receptors CB1 and CB2, the endogenous ligands such as anandamide and 2-arachidonylglycerol and the synthesizing and metabolizing enzymes (Alexander & Kendall 2007). Exogenous cannabinoid compounds such as phytocannabinoids Δ9 tetrahydrocannabinol and cannabidiol have immunomodulatory properties (Tanasescu & Constantinescu 2010). They inhibit pro-inflammatory cytokines (Klein et al. 2000); (Jean-Gilles et al. 2010), suppress T-cell proliferation (Katona et al. 2005), induce T-cell apoptosis (Rieder et al. 2010) and reduce migration and adhesion of immune cells. CB receptors are expressed on immune cells including B cells, T cells, neutrophils and NK cells, with CB2 levels known to be higher than those of CB1 (Galiegue et al. 1995). Their widespread location in the immune system allows CB receptor activation to regulate cytokine levels.
Both CB receptors are also expressed in CNS cells such as neurones, oligodendrocytes, astrocytes and microglia (Molina-Holgado et al. 2002, Sheng et al. 2005, Stella 2009, Pertwee 2010). All these cells can secrete inflammatory mediators that are implicated in the pathogenesis of neuro-inflammatory and neurodegenerative diseases such as multiple sclerosis (MS) (Hemmer et al. 2006). MS is a chronic inflammatory disease of the CNS white and grey matter which is characterized by demyelination, axonal injury and gliosis (Compston & Coles 2008). Inflammatory cytokines such as tumour necrosis factor (TNF)-α, interleukin (IL)-1β and IL-6 are elevated in lesions, cerebrospinal fluid and blood of patients with MS (Kahl et al. 2002, Filion et al. 2003, Kleine et al. 2003, Malamud et al. 2003). These cytokines are thought to play a role both in the CNS and in the periphery in neuro-inflammatory disease including MS (Compston & Coles 2008).
Neuroprotective and immunomodulatory effects of cannabinoids have been demonstrated in the animal model of MS, experimental autoimmune encephalomyelitis (EAE) (Pryce et al. 2003) (Arevalo-Martin et al. 2003). Beneficial effects were associated with diminished CNS mRNA expression of pro-inflammatory cytokines (Croxford & Miller 2003). Cannabinoids have also shown symptomatic benefit in human MS clinical trials (Pertwee 2002, Zajicek et al. 2003, Kavia et al. 2010, Rog 2010) where reduced pain, sleep disturbance, bladder overactivity, tremor and spasticity have been observed.
Although CB signalling and the regulation of cytokines by cannabinoids are well documented (De Jesus et al. 2006, Rubio-Araiz et al. 2008), only a few recent studies, in rodents mostly, have looked at the effects of inflammatory cytokines on the cannabinoid system. An increase in interferon (IFN)-γ modulated microglial CB2 receptors in neuropathic pain and MS mouse models (Maresz et al. 2005, Racz et al. 2008). Also, CB2 upregulation in chronic EAE correlated highly with the production of pro-inflammatory cytokines (Loría et al. 2008). CB1 receptors were also upregulated by the Th2 cytokine IL-4 (which downregulates both Th1 and Th17 cytokines involved in the pathogenesis of MS and EAE) and by cannabinoids themselves in human T lymphocytes (Börner et al. 2007, Borner et al. 2008) while IFN-γ was not shown to have any effect on CB1 expression by human T cells (Börner et al. 2007). In other studies in EAE, CB2 was shown to be upregulated in a progressive manner on specific post-immunization days in the presence of cytokines IFN-γ, IL-17, IL-4, IL-10, IL-1β, IL-6 and TNF-α (Lou et al. 2011), while CB1 receptor expression was downregulated on same days (Lou et al. 2012). Those studies suggest that manipulation of CB and CB receptors may have therapeutic value in MS.
In this study, we investigate the effects of the above pro-inflammatory cytokines on the expression of CB1 and CB2 on immune cells from both healthy subjects and patients with MS. We also show an upregulation of mRNA for CB receptors and for TNF-α, IL-1 and IL-6 in peripheral blood of people with MS compared to healthy controls.
Materials and methods
Donors and RNA extraction from blood
The study conforms with guidelines for Acta Physiologica (Persson 2013). Whole blood was obtained from 20 patients with MS (relapsing remitting, n = 12; secondary progressive, n = 6; and primary progressive, n = 2). These included 13 females and 7 males aged 28–64 years; mean Expanded Disability Status Scale (EDSS) score (Kurtzke 1983) = 4.6; range, 2.0–7.5. The patients were free of MS-specific immunomodulatory, immunosuppressive or monoclonal antibody therapies for at least 2 months.
Nineteen age- and gender-matched healthy volunteer subjects (14 females and 5 males, aged 27–62 years), not taking medicines, were used as controls. Peripheral blood mononuclear cells (PBMC) were isolated from heparinized whole blood obtained from healthy adult donors (8 females, 8 males; age 21–56 years) by density gradient centrifugation. Cells were resuspended in 10% FCS/RPMI, transferred to 24-well plates at 1 × 106/well and unstimulated or stimulated with IL-6 (100 ng mL−1), IL-1β (100 ng mL−1) or TNF-α (25 ng mL−1) (Peprotech, London, UK) in the absence or the presence of the NF-κB inhibitor SN50 (18 μm) (Calbiochem, Nottingham, UK) and incubated at 37 °C (5% CO2) for 18 h.
In a separate experiment comparing MS and control PBMC responses to pro-inflammatory cytokines, PBMC were isolated as above from 12 patients with MS different from the above (10 females, 2 males; mean ± SD age 48.3 ± 13.8 years) and 12 healthy controls (7 females, 5 males, males; mean ± SD age 42.8 ± 10.2 years). Five were not receiving and 7 were receiving disease-modifying treatment (5 interferon, 2 glatiramer acetate). Frozen PBMC were thawed and unstimulated or stimulated with IL-6, IL-1β or TNF-α (100 pg mL−1). Expression of CB1 and CB2 mRNA and protein was analysed by qRT-PCR and flow cytometry respectively.
Generation of polyclonally activated T cells (T-cell blasts)
Phytohemagglutinin (PHA)/IL-2-stimulated T cells were obtained from PBMC as shown before (Fahey et al. 2007) and after 24-h serum starvation stimulated with IL-6, IL-1β (100 ng mL−1) or TNF-α (25 ng mL−1) for a further 18 h.
RNA extraction and quantification
This was performed as previously described (Fahey et al. 2007, Jean-Gilles et al. 2009).
Whole blood preparation for flow cytometry
0.5 mL of heparinized venous blood was unstimulated or stimulated with IL-1β, IL-6 (100 ng mL−1) or TNF-α (25 ng mL−1) and incubated for 18 h (37 °C; 5%CO2). 2 mmol L−1 EDTA (pH 8) was added to tubes and incubated for 15 min (RT). Aliquotes were exposed to FACS lysing and FACS permeabilizing solution (BD Biosciences, Oxford, UK) and stained with antibodies CB1 [1 : 50] (PA1-745), CB2 [1 : 50] (PA1-744) (Affinity Bioreagents, Golden, CO, USA) or control normal rabbit IgG [1 : 40] (ProSci, Poway, CA, USA), then incubated with PE-conjugated goat anti-rabbit IgG [1 : 10] (Invitrogen, Paisley, UK) and were fixed in 1% formaldehyde/PBS for flow cytometry (Beckman-Coulter Altra Sorter, High Wycombe, UK).
Reverse-transcriptase reaction and quantitative real-time polymerase chain reaction
These studies were performed as previously described (Fahey et al. 2007) on a Mx4000 or Mx3005 platform (Stratagene, La Jolla, CA, USA). Reaction conditions are available on request. Primer sequences are shown in Table 1. Beta-2-microglobulin (B2MG) (PBMC, T cells) and large human ribosomal protein-PO (RPLPO) (whole blood) were used as housekeeping genes for normalization of data.
| B2MG for | 5′-CTC CGT GGC CTT AGC TGT G-3′ |
| B2MG rev | 5′-TTT GGA GTA CGC TGG ATA GCC T-3′ |
| RPLPO for | 5′-CCA CGC TGC TGA ACA TGC T-3′ |
| RPLPO rev | 5′-TCG AAC ACC TGC TGG ATG AC-3′ |
| CB1 for | 5′-CTG GAA CTG CGA GAA ACT G-3′ |
| CB1 rev | 5′-CGC ATA CAC GAT GAA CAG AAG-3′ |
| CB2 for | 5′-GCC TCT TCC CAA TTT AAA CAA C-3′ |
| CB2 rev | 5′-AGT CAG TCC CAA CAC TCA TC-3′ |
| IL-6 for | 5′-TCA ATG AGG AGA CTT GCC TG-3′ |
| IL-6 rev | 5′-GAT GAG TTG TCA TGT CCT GC-3′ |
| IL-1β for | 5′-GGA TAT GGA GCA ACA AGT GG-3′ |
| IL-1β rev | 5′-ATG TAC CAG TTG GGG AAC TG-3′ |
| TNF-α for | 5′-ACA AGC CTG TAG CCC ATG TT-3′ |
| TNF-α rev | 5′-AAA GTA GAC CTG CCC AGA CT-3′ |
- for, forward; rev, reverse.
mRNA results were based on normalized ratio to the internal standard expression (RPLPO for blood or B2MG for PBMC and T cells), and significance (P ≤ 0.05) in differences between the groups was determined by t-test and one-way repeated measures anova.
Intracellular staining and flow cytometry
Peripheral blood mononuclear cells thawed from 12 patients with MS and 12 healthy controls were unstimulated or stimulated with cytokines then incubated for 24 h. For some experiments (not shown), conjugated anti-CD4, CD8 (T cells), CD3-, -CD14, -CD16 and -CD19 (PBMC) mouse monoclonal antibodies were added. After fixation and permeabilization, they were stained with CB1-APC (and isotype control) and CB2 primary antibody (R&D), incubated for 30 min and then stained with FITC secondary antibody and isotype control (R&D). Data were analysed using Weasel v3.1.
For T-cell experiments, the same above reagents as for whole blood were used. CB1 and CB2 protein abundance was using winmdi software. Changes in mean fluorescence intensity [geometric (g) mean] and/or % of positive cells were calculated by comparing negative control (no anti-CB1 or CB2 antibodies used), unstimulated and stimulated (with cytokines) cells. Positive cells were gated in two-parameter histograms and/or shifts (increase in mean or median fluorescence intensity (MFI) were gated within marker M1 for single-parameter histograms and averaged across whole population.
Western blotting
Peripheral blood mononuclear cells were in 2%FCS/RPMI with or without cytokines, CB1/CB2 agonist HU-210 (Tocris, Bristol, UK) [100 nm] in the presence or absence of CB1 antagonist SR141716A (SR1, rimonabant) and CB2 antagonist SR144528 (SR2) [300 nm] (gifts from Sanofi, Guildford, UK) and were centrifuged for 10 min (300 g, 4 °C). The supernatant was discarded and lysis buffer [with 20 mmol L−1 Tris, 1 mmol L−1 EGTA and protease inhibitor (P8340; Sigma, Gillingham, UK)] was added. Samples were homogenized for 30 min (4 °C) and supernatant extracted. To quantify protein, Lowry test was conducted for homogenates and human recombinant CB1 and CB2 proteins (Axxora, UK) used as positive controls, and protein was equally loaded onto gels. 2× solubilization buffer [containing Tris (20 mmol L−1) (Invitrogen), glycerol, sodium dodecyl sulphate (10%) (SDS), β-mercaptoethanol and bromophenol blue (Sigma)] was added [1 : 2], and samples were heated to 100 °C for 5 min. Proteins (20 μg) were loaded onto 10% SDS gels and resolved by SDS-polyacrylamide gel electrophoresis with a standard marker ladder of known molecular weights run alongside (45 min, 200 V). Following transfer onto nitrocellulose membranes (Amersham Biosciences, Little Chalfont, UK) (60 min, 100 V, 4 °C) and staining with Ponceau S (Sigma), blots were blocked in 1 g/20 mL non-fat dry milk in 0.1% Tris-buffered saline–Tween 20 (pH 7.6) for 60 min. Separated membranes were incubated overnight (4 °C) with anti-CB1 (1 : 500), anti-CB2 (1 : 1000) (Affinity Bioreagents), total or phosphorylated (p)-mitogen-activated protein kinase (MAPK) 42/44 (Erk1/2) (1 : 1000) antisera (Cell Signalling Biotechnology, Hitchin, UK). Blots were then washed and incubated with secondary antibody horseradish peroxidase-conjugated goat anti-rabbit IgG (Amersham Biosciences) in blocking buffer (1 : 2000) for 60 min (RT) and reprobed with β-actin (1 : 4000) antibody (Abcam, Cambridge, UK) as a loading control using enhanced chemoluminescent reagents, hyperfilm (Amersham) and Kodak reagents for development. Specific immunoreactive bands were analysed using a Bio-Rad GS-710 imaging scanning densitometer. Protein differences between samples were analysed based on optical density (OD) of each band and by two-tailed t-test with P ≤ 0.05. Differences were explored using one-way repeated measurements anova.
Results
CB1 and CB2 expression in cytokine-stimulated whole blood, PBMC and PHA/IL-2 induced T cells of normal subjects
The effects of IL-6, IL-1β and TNF-α on CB1 and CB2 receptor gene expression in whole blood, PBMC and T cells are shown in Figure 1. TNF-α significantly induced CB1 and CB2 mRNA levels in whole blood (P = 0.02, P = 0.01), T cells (P = 0.04, P = 0.04) and PBMC (P = 0.01, P = 0.04) compared to unstimulated cells. In each cell population, the expression of CB receptor mRNA levels induced by TNF-α was 2–3 times higher than levels induced by IL-6 and IL-1β. Although IL-1β and IL-6 significantly increased CB1 (P = 0.04, P = 0.04) and CB2 (P = 0.03, P = 0.02) mRNA levels in whole blood, induction by these cytokines in T cells did not reach significance. In PBMC, a non-significant increase was noted for CB2 (P = 0.06 for IL-6; P = 0.07 for IL-1β) and for CB1 (P = 0.12 for IL-6; P = 0.19 for IL-1β). These differences may reflect a direct or indirect contribution of CB mRNA induction in polymorphonuclear leucocytes.
The effects of IL-6, IL-1β and TNF-α on the CB1 and CB2 receptor at the protein level assessed using flow cytometry are shown in Figure 2. Although all three pro-inflammatory cytokines increased CB1 and CB2 protein expression levels, stimulation with TNF-α induced higher levels than IL-6 or IL-1β in whole blood and PBMC (Fig. 2a,b). CB1 and CB2 immunopositivity in whole blood gated for monocytes/lymphocytes was confirmed to be increased by TNF-α (CB1 MFI = 973 ± 243, P = 0.001 and CB2 MFI = 206 ± 56, P = 0.02) when compared to unstimulated samples (MFI = 42.1 ± 14.1). CB1 protein MFI was also increased in whole blood stimulated with IL-1β (317.4 ± 64.5, P = 0.009) and IL-6 (594.6 ± 130.8, P = 0.005). CB2 protein MFI induced by IL-1β showed a non-significant increase (81.5 ± 20.4, P = 0.06), and with IL-6, it was significantly increased (106.3 ± 28.7, P = 0.05) (Fig. 2a). PBMC stimulated with cytokines showed increases in CB1 (MFI: TNF-α = 84.9 ± 17.8; P = 0.02); IL-6 (49.1 ± 1.1; P = 0.04); IL-1β (32.1 ± 7.0; P = 0.04). CB2 protein levels (MFI for TNF-α = 13.5 ± 4.0; P = 0.08; IL-6: 11.9 ± 3.0; P = 0.09; IL-1β: 10.5 ± 2.9; P = 0.12) only showed an insignificant increase compared to unstimulated (MFI = 9.0 ± 2.7) samples (Fig. 2b). CB receptor protein expression on T cells showed similar regulation by cytokines to that of PBMC (data not shown).
MAPK (Erk1/2) activation in HU-210-stimulated PBMC
To confirm that CB receptors were functional in PBMC, HU-210 (100 nm), a non-selective CB1 and CB2 receptor agonist, was used to activate them. HU-210 was chosen due to its potency in preference to weaker selective agonists (Alexander & Kendall 2007). HU-210-stimulated PBMC showed a transient enhancement of pMAPK42/44 levels (23% increase; P = 0.03) compared to control (FCS 2%; ODavg = 0.22 ± 0.04) to peak at 15 min (Fig. 3) although no significant increase in total MAPK was observed. Immunoreactivity of p42 was more prominent than p44. Beta-actin immunoreactivity was unchanged.
MAPK (Erk1/2) activation in cytokine-stimulated PBMC
Total and phosphorylated (p)MAPK42/44 immunoreactivity were also assessed in cytokine-stimulated PBMC, with or without exposure to the cannabinoid receptor antagonists SR 141716 and SR 144528 (SR1 and SR2), by reprobing of blots (Fig. 4). Our preliminary work showed that the combination of SR1 and SR2 rather than individual antagonist use is typically needed to antagonize HU-210 effectively in immune cells. No significant change was observed in total MAPK42/44 immunoreactivity (ODavg = 0.20 ± 0.03) whereas pMAPK42/44 was increased in the presence of cytokines, potentially reflecting CB receptor response to cytokine stimulation in PBMC. These levels were decreased close to basal levels (ODavg = 0.21 ± 0.05) in the presence of SR1 and SR2 (Fig. 4). TNF-α induced the highest level of pMAPK42/44 immunoreactivity (ODavg = 0.72 ± 0.14; P = 0.03); SR1/SR2 reduced it (ODavg = 0.53 ± 0.08; P = 0.04) while remaining above basal levels. Beta-actin immunoreactivity remained unchanged.
CB1 and CB2 immunoreactivity and functional response in cytokine-stimulated PBMC
CB1 and CB2 immunoreactivity was assessed by Western blot in cytokine-stimulated PBMC with or without the cannabinoid receptor antagonists SR141716A (SR1) and SR144528 (SR2) (300 nm each). Two bands were detected at around 60 and 37 kDa possibly corresponding to CB1 dimer and monomer, respectively, and a band migrating at approx. 60 kDa representing CB2 monomer (Fig. 5). These band sizes were similar to those obtained in the previous literature (Nunez et al. 2004, De Jesus et al. 2006) and to those of reference human CB1 recombinant protein (Fig. 5, left lane). CB1 and CB2 protein expression was upregulated by exposure to cytokines and downregulated in the presence of SR1 and SR2. The maximum OD increase was 55% (P = 0.02) for CB1 and 17% for CB2 (P = 0.07), possibly because of a higher baseline expression of CB2. No significant changes were observed in beta-actin immunoreactivity.
Effect of NF-kB inhibitor treatment on cannabinoid receptor and TNF-α gene expression in PBMC
NF-κB is a key transcription factor for several inflammatory mediators including TNF-α. Given that this cytokine induced the highest levels of CB1 and CB2 expression (see above), we investigated its influence on cannabinoid receptor and TNF-α gene expression in PBMC unstimulated or stimulated with TNF-α, in the presence or absence of the NF-κB inhibitor SN50 (18 μm). SN50 significantly decreased the levels of CB1, CB2 and TNF-α mRNA (P = 0.004, P = 0.002 and P = 0.004, respectively) compared to those of unstimulated cells and also decreased their expression in TNF-α-stimulated PBMC (all P < 0.05) (Fig. 6). These results indicate that TNF-α induction of CB1, CB2 and TNF-α itself involves the NF-κB signalling pathways.
Cytokine, CB1 and CB2 mRNA expression in blood from MS patients
mRNA levels for IL-6, IL-1β and TNF-α were significantly increased in patient with MS blood compared to that of controls. IL-6 showed the largest increase (17-fold; P = 0.01) in the MS group while IL-1β showed fivefold (P = 0.02) and TNF-α showed sixfold (P = 0.01) increase (Fig. 7a). CB1 and CB2 mRNA were also significantly elevated in MS blood compared to that of controls. CB1 and CB2 gene expression was increased approximately four- (P = 0.02) and fivefold (P = 0.01), respectively, in patients with MS (Fig. 7b).
CB1 and CB2 mRNA and protein expression in MS vs. healthy control cytokine-stimulated PBMC
To compare inducibility of CB receptors by cytokines in normal human immune cells and those of patients with MS, we investigated the effects of IL-1β, IL-6 and TNF-α on CB mRNA of PBMC from 12 normal control subjects and 12 age- and sex-matched patients with MS. We used low concentration ranges (100 pg mL−1) that are close to those found in body fluids in inflammatory circumstances. In this sample, the same trends as with the higher concentrations used above in normal controls were noted (data not shown). Statistical significance and a trend is seen for CB1 and CB2 increase at baseline in patients with MS (P = 0.04 and 0.07 respectively). IL-6 increased both CB1 and CB2 mRNA in healthy controls (P = 0.018 and 0.013 respectively). The results partially confirm those using higher cytokine concentrations, but indicate that the lower concentrations are less potent. The data suggest that MS patients’ PBMC tend to respond similarly to those of healthy subjects in terms of CB induction, however, starting from higher baseline expression levels. No significant differences were noted between untreated patients and those on immunomodulatory treatment.
At the protein levels, similar trends were observed. Significance was noted in CB2 levels being higher in all cytokine-stimulated PBMC of MS than in healthy controls (Fig. 8).
Discussion
Our study contributes to the evidence that the interaction between the cannabinoid system and the cytokine network is bidirectional (Roche et al. 2008, Rubio-Araiz et al. 2008) and extends this evidence to human biology. CB1 and CB2 expression levels were most significantly induced by TNF-α. This occurs, at least in part, through NF-κB activation as its inhibition suppressed TNF-α-enhanced CB1, CB2 as well as TNF-α gene expression itself. Activation of NF-κB could have been induced by stimulation of the TNF receptor. Upon activation, NF-κB translocates to the nucleus where it binds DNA and triggers the transcription of target genes (Ledeboer et al. 2005), some of which encode inflammatory proteins and may also include the cannabinoid receptor genes. It has been shown that the CB receptors are involved in TNF-α-mediated NF-κB activation (Zheng et al. 2008). NF-κB mediates the TNF-α-induced transcription of CB1 gene in neurones (Borner et al. 2012). Thus, cytokine-responsive cells such as T cells, macrophages, neurones, astrocytes and microglia may enhance CB1 and CB2 expression through the NF-κB pathway. A study by Börner et al. (2007) provides support for this, showing that the activation of human T cells upregulates CB1 but not CB2, transcription, and that CD3/CD28-mediated upregulation is in part NF-κB dependent through activation of CB1 promoter.
On the other hand, caution is invited in interpretation of our results, as SN50 may not be highly specific for NF-κB, being also a proteasome and AP-1 inhibitor (Boothby 2001). Of note, in a separate but similar experimental set-up, we were unable to block the induction of CB1 and CB2 in the presence of AP-1 inhibitor NGDA, suggesting at least that the effect of SN50 was not due to AP-1 blockade (Jean-Gilles L, Constantinescu CS, unpublished observations).
Activation of MAPK reflects CB receptor activity. Recently, CB2 was shown to be upregulated by stimulation with IL-1β, TNF-α or lipopolysaccharide in synovial fibroblasts of patients with rheumatoid arthritis, and the selective CB2 agonist HU-308 inhibited IL-1β-induced production of IL-6 and IL-1β-induced activation of Erk 1/2 and p38 MAPK (Gui et al. 2014). In our study, the upregulation of CB1 and CB2 mRNA and protein was paralleled by activation of MAPK upon exposure to pro-inflammatory cytokines. This confirms that the upregulation of CB receptors in immune cells has functional consequences. However, the similarity of MAPK activation by cannabinoid receptor activation and by inflammatory cytokines in PBMC does not necessarily imply that these phenomena are causally linked.
Our finding of upregulation of CB by inflammatory cytokines and in MS, together with the dysregulation of the endocannabinoid system in MS (Jean-Gilles et al. 2009), raises the possibility of a therapeutic effect of cannabinoids in inflammatory diseases.
We show that cannabinoid receptors are functional in PBMC, as stimulation with HU-210-enhanced phosphorylation of MAPK 42/44 (also known as ERK1/2). Cannabinoids have previously been shown to activate these kinases through the activation of cannabinoid receptors (Upham et al. 2003). The fact that exposure of cytokine-stimulated cells to cannabinoid receptor antagonists, SR1 and SR2 synergistically suppressed the functional response of the cannabinoid receptors to cytokines suggests that these responses were direct and specific.
The present study also demonstrates that CB1 and CB2 upregulation in the blood of patients with MS. This is in line with very recent data by Sanchez Lopez et al. (2015) which shows an increased expression of cannabinoid receptors in peripheral blood lymphocytes from patients with MS. Of note, in the second set of experiments comparing PBMC from healthy controls and patients with MS, the upregulation in MS was less prominent. The difference from baseline levels in Figure 7b may represent difference between whole blood and PBMC expression. This may also reflect the use of frozen cells (3–5 months) or the blunting of the overall expression due to the 58% of patients being on disease-modifying treatments, as interferon treatment reduces CB levels (Sanchez Lopez et al. 2015). In our study, there is baseline variation in the levels of expression of these receptors, and some discrepancies between results at the mRNA and protein level; between results of whole blood testing vs. PBMC; and between results using higher vs. lower cytokine concentrations. Regulation of expression at the protein and mRNA level can differ, in particular for cytokines, through post-translational or mRNA stability changes (e.g. Kazanecki et al. 2007). Nevertheless, the tendency towards an increase in CB expression by cytokines and an increase in MS compared to controls was reasonably consistent at the mRNA, protein and signalling level. The responses of immune cells of patients with MS to inflammatory cytokines, while, in some instances, more vigorous than those of controls, tend to be similar. This was also observed in immune cell subsets of a separate set of patients with MS vs. controls (n = 5 and 3, respectively) (data not shown); however, the small sample size precludes definitive conclusions.
Upregulation of CB in MS is associated with a simultaneous upregulation of pro-inflammatory cytokine expression in immune cells as previously demonstrated in MS mouse models (Maresz et al. 2005, Loría et al. 2008). Our current findings are consistent with previous reports demonstrating that pro-inflammatory cytokines are upregulated in patients with MS (Kahl et al. 2002, Filion et al. 2003).
We recognize that the association of the upregulation of CB1 and CB2 to that of cytokines in MS does not imply causation, in particular as concentration of cytokines likely to be found in body fluids in MS was less potent than the doses used in vitro on control cells in the first part of the study. The increased expression of CB1 and CB2 in MS may be in part mediated by individual pro-inflammatory cytokines or their combined effect. In addition, other factors in the MS blood or inflammatory/neurodegenerative micro-environment may also contribute to CB upregulation, as a more general damage signal. Endocannabinoids themselves can upregulate their own receptors. We have previously demonstrated upregulation of endocannabinoid levels in blood (Jean-Gilles et al. 2009), which can contribute to CB1 and CB2 upregulation.
This study is among the first thorough investigations of cannabinoid receptor regulation by prototypical, antigen-presenting cell-derived pro-inflammatory cytokines in humans and of their expression in MS blood. Further investigations are required to better understand the mechanisms entailed in CB1 and CB2 receptor regulation by cytokines, including studies to determine whether there is a direct transcriptional effect of these cytokines on the CB receptor genes as shown for transcriptional control of the CB1 gene by IL-4 (Borner et al. 2008). Further investigations in MS patients, in different phases of MS with different levels of disease activity and under different treatments, will also be important. Very recently, it was shown that untreated patients with MS have increased gene expression levels of cannabinoid receptors in different types of immune cells, and those levels decrease significantly over the course of 1 year of interferon-beta (IFN-β) treatment (Sanchez Lopez et al. 2015). This observation is consistent with our findings, given the suppressive effect of IFN-β on proinflammatory cytokines.
There may be potential limitations of our study. The specificity of CB1 and CB2 antibodies used in the first part of the study has been subject to some debate, in part explaining some of the discrepancies found in the literature (Grimsey et al. 2008). However, well-conducted human studies have detected bands of the same size as in our Western blots, notably De Jesus et al. (2006), who detected pronounced bands of 60 and 37 kDa ascribed to CB1 (the latter potentially representing a result of deglycosylation or proteolysis of the latter); these bands were confirmed in other studies, for example Xu et al. (2005). For CB2, a 60 kDa band has also been confirmed, for example by Nunez et al. (2004). In addition, in preliminary studies, we showed significant suppression of both CB1 and CB2 positivity by an average of 74% after blocking with recombinant CB1 and CB2 (data not shown). This indicates that the antibodies used here were specific. It must be noted that densitometry on Western blots performed with HRP reaction, while not fully quantitative, allows for comparisons within experiments.
The lower cytokine concentrations were less potent than the high concentrations used initially. We need to take into account the fact that the in vivo inflammatory environment is much more complex and combination of lower concentrations may have similar effects to individual higher concentrations in vitro. For example, cytokines are known to influence the expression of one another and a multitude of other immune factors could further modulate the expression levels of cytokines or cannabinoid receptors. This is in part the reason we included experiments on whole blood in this study, to reproduce better the in vivo conditions and to minimize alterations that may result from cell separation. Indeed, there were differences between whole blood and PBMC results, which can be explained by the different micro-environment. Thus, caveats of translating in vitro results to the complex in vivo context must be taken into consideration.
There may be multiple functional implications of the CB regulation by inflammatory cytokines in particular TNF-α, but also IL-6 and IL-1. As well as cannabinoids themselves, these cytokines have both pro- and anti-inflammatory properties, depending on a variety of factors including concentration, micro-environment, timing of appearance and clearance. TNF-α has both neuroprotective and neurodegenerative effects. Such dual effects are also thought to exist in CB signalling. Interesting data show a high level of coregulation between TNF-α and cannabinoid receptors in the CNS, in the process of neural stem cell migration. TNF-α provides a crucial signal for stem cell migration, and this signal appears to be dependent on CB1/CB2 signalling (Rubio-Araiz et al. 2008). Our data provide a plausible mechanistic explanation for this important biological phenomenon, whereby TNF-α induces the required CB1 and CB2 expression.
Previous studies have shown analgesic effects of CB agonists in neuropathic, but not normal, rats (Sagar et al. 2005). Upregulated CB1 expression in these animals contributes to increased analgesic efficacy of cannabinoids (Siegling et al. 2001). Moreover, CB2-mediated control of neuropathic pain in mice is IFN-γ dependent (Racz et al. 2008). Also, it has been shown that expression of endocannabinoid ligands is increased in a viral model of MS and inflammatory lesions of patients with MS (Eljaschewitsch et al. 2006, Loría et al. 2008) and in plasma of patients with different types of MS (Jean-Gilles et al. 2009). Such studies may help to elucidate cannabinoid mechanisms involved in controlling MS symptoms such as spasticity, overactive bladder and pain. In addition, polymorphisms in the CB genes have been associated both with MS subtypes and with surface CB expression (Woolmore et al. 2008, Ramil et al. 2010, Rossi et al. 2013). This may explain the heterogeneity at baseline and upon stimulation in our study. Future studies may determine which patients with MS are most likely to show benefit from modulation of the cannabinoid system. If the regulation of cannabinoid receptors in CNS cells and immune cells is similar, our results could contribute to the future studies of cannabinoid neuroprotective mechanisms and possibly future treatment of inflammatory and neurodegenerative diseases.
Conflict of interest
The authors report no conflict of interests.
This study was supported in part by the MS Society of Great Britain and Northern Ireland and by the University of Nottingham Neuroscience at Nottingham, N@N). LJE was supported by a Patrick Berthoud Fellowship. RT was supported by a Visiting Fellowship to University of Nottingham from the European Neurological Society.
