Canine Immune Cells Express High Levels of CB₁ and CB₂ Cannabinoid Receptors and Cannabinoid-Mediated Alteration of Canine Cytokine Production is Vehicle-Dependent

Abstract

With the increased popularity and societal acceptance of marijuana and cannabidiol (CBD) use in humans, there is an interest in using cannabinoids in veterinary medicine. There have been a few placebo-controlled clinical trials in dogs suggesting that cannabis-containing extracts are beneficial for dogs with inflammatory diseases such as osteoarthritis, and there is growing interest in their immunosuppressive potential for the treatment of immune-mediated diseases. Since cannabinoids exhibit anti-inflammatory and immunosuppressive effects in many species, the purpose of these studies was to examine whether the plant-derived cannabinoids, CBD and Δ⁹-tetrahydrocannabinol (THC), would also suppress immune function in canine peripheral blood mononuclear cells (PBMCs). Another goal was to characterize expression of the cannabinoid receptors, CB₁ and CB₂, in canine immune cells. We hypothesized that CBD and THC would suppress stimulated cytokine expression and that both cannabinoid receptors would be expressed in canine immune cells. Surprisingly, cannabinoid suppressive effects in canine PMBCs were quite modest, with the most robust effect occurring at early stimulation times and predominantly by THC. We further showed that cannabinoid-mediated suppression was dog- and vehicle-dependent with CBD and THC delivered in dimethyl sulfoxide (DMSO) producing more immune suppressive effects as compared to ethanol (ETOH). PCR, flow cytometry, and immunohistochemical staining demonstrated that both CB₁ and CB₂ are expressed in canine immune cells. Together these data show that canine immune cells are sensitive to suppression by cannabinoids, but more detailed studies are needed to further understand the mechanisms and broad effects of these compounds in the dog.

Introduction

With the popularity of cannabis use for a variety of ailments in humans, many pet owners have begun exploring use in dogs, especially for its potential to reduce pain or anxiety. Two-thirds of Canadian pet owners believe that marijuana or hemp products are more effective for pain relief than conventional medications [1]. Similarly, nearly half of US veterinarians approve of the medicinal use of hemp/CBD in canine patients [2] and more than half of veterinary students in the US feel that there are potential medicinal/therapeutic benefits to be studied in marijuana [3]. With its increasing popularity and growing number of efficacy studies, clients are expecting their veterinarian to be knowledgeable about medical cannabinoid treatment; however, many veterinarians do not feel sufficiently educated on the topic [2]. A few small studies on the safety and efficacy of cannabinoids in veterinary medicine have been performed [4–7]. Given the limited amount of information despite significant public interest and therapeutic potential, more studies on cannabinoids and on the endocannabinoid system (ECS) in animals are warranted.

The ECS is found in nearly all animal species and is comprised of cannabinoid receptors, cannabinoids, and enzymes involved in endogenous cannabinoid (i.e., endocannabinoid) biosynthesis and degradation [8, 9]. There are two main cannabinoid receptors (CB), CB₁ and CB₂, which are G-protein coupled receptors that play a physiologic role in pain, anxiety, inflammation, immune function, metabolic regulation, neuronal plasticity, and bone growth [8]. CB₁ and CB₂ are encoded by the genes CNR1 and CNR2, respectively. In dogs, CB₁ receptor protein has been identified immunohistochemically in epithelial structures of canine embryos, as well as within the hair follicles, salivary glands, skin, gastrointestinal tract, astrocytes, ependymal cells, neurons, and isolated neuroglial cells [8, 10–16]. Since CB₂ was cloned in dogs [17], it has been identified immunohistochemically in the canine gastrointestinal tract, skin, blood vessels, and astrocytes [13–15, 18].

Cannabinoid compounds derived from Cannabis spp. include THC, which exhibits affinity for both CB₁ and CB₂, and CBD, which does not bind either CB₁ or CB₂ with high affinity [19, 20]. Due to its affinity for CB₁, THC is the psychotropic component of cannabis, whereas CBD is psychoactive but does not produce the euphoric high. Despite the differences in their binding affinities and psychotropic effects, they produce similar effects in the immune system. For instance, it has been demonstrated in mice that CBD and THC reduce levels of many proinflammatory cytokines, including TNF-α, INF-γ, and IL-1β [21, 22]. Both compounds have also been used experimentally in several animal models of autoimmune disease. THC has been shown to be effective in animal models of multiple sclerosis [23], autoimmune diabetes [24], and autoimmune hepatitis [25]. Similarly, CBD has attenuated disease in models of diabetes [26], multiple sclerosis [27], and hepatitis [28]. Furthermore, CBD has been studied in osteoarthritis in dogs and has shown promise in alleviating pain and increasing mobility [4, 29–33].

With the established link between pain and inflammation, the purpose of these studies was to evaluate the effect of CBD and THC on purified canine PBMCs in vitro. First, the degree to which CB₁ and CB₂ protein were expressed was evaluated in canine spleen and brain via immunohistochemistry (IHC) and in PMBCs via flow cytometry. Second, PBMCs were immunologically stimulated with phorbol ester plus calcium ionophore (P/I) and gene expression of 4 pro-inflammatory cytokines (IL2, IFNG, TNFA, and IL17A) were evaluated in response to CBD and THC delivered in ETOH. Next, the effects of CBD and THC delivered in either ETOH or DMSO were assessed on IL-2 and IFN-γ protein using flow cytometry and ELISA. Finally, we present cellular concentrations after in vitro delivery of CBD and THC in canine PBMCs. The mechanism by which cannabinoids alter immune function in dogs is becoming an increasingly important and timely topic as more veterinarians are seeing either toxicity from accidental marijuana ingestion or purposeful administration of CBD or other cannabis preparations for their putative medical benefit. These studies are some of the first to evaluate immune function modulation of canine PBMCs by cannabinoids.

Materials and Methods

PBMC isolation and cell culture

Mississippi State University maintains a research colony of adult Treeing Walker Coonhounds, with normal heath status confirmed via a full physical exam as well as a CBC, chemistry, and urinalysis. Experiment and animal care protocols were approved by Mississippi State University Institutional Animal Care and Use Committee (IACUC # 20–056). Mississippi State University is accredited by the Association for Assessment and Accreditation of Laboratory Animal Care. Each dog was assigned a letter followed by a number to indicate a separate blood draw/experiment with that dog, although not all dogs were used for these data. PBMCs were harvested from buffy coats following differential centrifugation using Histopaque 1077 (Sigma). Cells were then enumerated with a Nexcelom cell counter and suspended in complete media (1X RPMI containing 2% bovine calf serum, 1% penicillin-streptomycin, and 1% Glutamax). 2×10⁶ cells/well were cultured in 1 ml complete medium in triplicate in a 48-well plate. Various concentrations of CBD or THC ( 1–10 μM) were delivered to the cells in either 0.1% ethanol (ETOH) or 0.02% dimethyl sulfoxide (DMSO). Phorbol ester plus calcium ionophore (P/I) at 12.5 ng/mL/0.8 μM was used to stimulate the cells, which was effective in previous studies evaluating pharmacodynamics of immunosuppressant agents in dogs [34].

RNA Isolation and cDNA Synthesis

At harvest, cells were pelleted and washed with 1X PBS. Cell pellets were frozen at −80°C until RNA isolation. RNA was isolated using the RNeasy kit (Qiagen, Germantown, MD). For blood, 250 ml blood was mixed with 500 μl TRI reagent (Sigma) and mixed. For brain tissue, a small amount of tissue was placed in 500 μl TRI reagent and homogenized. Bromochloroproprane (200 μl) was added to the blood or brain homogenate and mixed vigorously. After centrifugation at 10000 × g for 15 min, the aqueous phase was placed in a new tube to which 500 μl 75% isopropanol was added to precipitate the nucleic acids. After centrifugation at 8000 × g for 10 min, the resultant pellet was used in the RNeasy isolation kit (Qiagen). DNase treatment was conducted on samples in which Taqman primer-probes would not span introns (i.e., CNR1 and CNR2). RNA was quantified on a Nanodrop spectrophotometer and adjusted to the same level within each experiment. cDNA was reverse transcribed from RNA using the cDNA High Capacity Kit (Applied Biosystems/Thermo Fisher).

Reverse transcriptase quantitative polymerase chain reaction (RT-qPCR)

cDNA (4 μl) was used in each PCR reaction. Each PCR reaction contained the cDNA, the Taqman primer-probe of interest (all from Thermo Fisher; assay ID numbers were: CNR1, Cf02681726_u1; CNR2, Cf02696139_s1; IL2, 0263317_m1; IFNG, 0263316_m1; TNF, Cf02628237_m1; IL17A, Cf03993800_m1). A housekeeping primer-probe was also used in every Taqman RT-qPCR reaction (either 18S or GAPDH, assay ID Cf0441963_gH). Target gene probes were detected as FAM and housekeeping probes were detected as VIC/JOE. PCR reactions were amplified and detected on a Stratagene MXP3005p or AriaMx. Fold change was calculated on the ΔΔCt method and compared to the unstimulated control.

Flow cytometry

PBMCs were washed with 1X PBS and incubated with Near-IR fixable viability dye (APC-Cy7) for 20 min at RT in the dark. Following a wash with 1X PBS, samples were incubated with Fc block in FCM buffer (IX Hank’s Buffered Saline Solution, HBSS, containing 1% BSA). We initially conducted a comparison between mouse Fc block (BD Biosciences) and human FcX (Biolegend) and determined that similar staining with antibodies was achieved, so human FcX was used for subsequent analyses. After incubation with Fc block, primary polyclonal anti-CB₁ or anti-CB₂ (Novus Biologicals) or PE-Cy7-CD4 or APC-CD8 (eBioscience) was added for 30 min at RT in the dark. After a wash in FCM buffer, the secondary antibody conjugated to FITC for CB₁ and CB₂ (Biolegend) was added for 30 min at RT in the dark. Since a single secondary antibody was used to detect CB₁ and CB₂, they were analyzed in separate samples (although this prevented evaluation of the percentage of cells that expressed both receptors). Cells were then fixed with BD Cytofix (BD Biosciences) for 15 min at RT in the dark. Cells were washed and suspended in FCM buffer prior to analysis on an ACEA Novocyte (ACEA/Agilent). For intracellular cytokine staining, fixed cells previously stained for CD4 or CD8 were permeabilized with eBioscience permeabilization buffer, then incubated with RPE-IFN (BioRad) and biotinylated IL-2 (R and D Systems) for 30 min. IL-2 required a streptavidin secondary conjugated to Brilliant Violet 650 (Biolegend) that was also delivered in permeabilization buffer for another 30 min. As with extracellular stains, cells were washed and suspended in FCM buffer prior to analysis on an ACEA Novocyte. Cells were gated on live single leukocytes. Staining index for the CB₁ and CB₂ antibody titration is presented as the ratio of the positive to negative signal using the secondary antibody alone to discriminate the populations. Intracellular cytokines were gated on live single leukocytes expressing CD4 or CD8. Detailed antibody information can be found in Supplemental Table 1.

Immunohistochemistry (IHC)

Samples were acquired from the MSU-CVM Pathology Department and included archived tissues from necropsy cases of 5 clinically healthy dogs of varied signalments (Chihuahua, Dachshund, Golden Retriever, Beagle, and mixed breed, ranging from 4–16 years of age). 5 μm sections of formalin-fixed, paraffin-embedded (FFPE) cerebrum and spleen were examined. All samples were grossly and histologically normal. Slides were deparaffinized and rehydrated by placing them consecutively in xylene, 100% ethanol, 95% ethanol, and de-ionized water for 2 min each. Antigen retrieval was achieved via treatment with target retrieval solution (TRS; Dako) in a steamer for 20 min. Slides were then air dried for approximately 10 min and rinsed with PBS. Endogenous peroxidases were blocked by treating slides with 3% H₂O₂ for 30 min, followed by three PBS washes. Nonspecific binding of primary antibody was blocked by incubating the slides with a solution of 1% BSA and 10% goat serum (Vector Labs) for 60 minutes. Slides were incubated overnight (~18 hours) at 4°C with polyclonal rabbit anti-CB₁ or anti-CB₂ (Novus Biologicals) in PBS/1%BSA/0.1% Triton X-100 as the primary antibody. Numerous primary antibody dilutions were attempted to achieve signal optimization in the brain and spleen (1:50, 1:100, 1:250, 1:400, 1:500, 1:600 or 1:800). Final dilutions were 1:50 for CB₁ in the brain and 1:800 and 1:250 for CB₁ and CB₂ in the spleen, respectively. Slides were rinsed 3 times with PBS, then treated at room temperature for 2 hr with biotinylated goat anti-rabbit (Vector Labs) in PBS/1% BSA/0.1% Tx (1:100). Slides were rinsed with distilled water three times. The avidin-biotin complex method was used for detection per kit instructions (Vectastain ABC kit, Vector Labs). Incubation with diaminobenzidine (DAB; Dako) was applied for 5 min and halted by dipping slides in distilled water. Slides were counterstained with hematoxylin for 4 min, rinsed in distilled water, then dipped in 0.3% ammonia water. Dehydration was achieved by sequentially placing the slides in distilled water, 100% ethanol, and xylene for 1 min each, and slides were cover-slipped. Detailed antibody information can be found in Table 1.

ELISA:

Plates were coated with anti-canine IL-2, IFN-γ (both R and D Systems), or IL-17A (Bio-Rad) overnight. Plates were washed 3 times with 0.05% tween-20 in PBS followed by 3 times with deionized water (water washes were omitted for IL-2 and IFN-γ). Plates were blocked for at least 1 hr with 3% bovine serum albumin (BSA) in 1X PBS. A standard curve generated with recombinant canine IL-2, IFN-γ, or IL-17A and samples were incubated for 1 hr at RT. After plate washing, biotinylated matched antibodies were added to the respective plates for 1 hr at RT. Color-based detection was achieved by subsequent incubation steps with HRP Avidin (Biolegend) and TMB substrate, with washes in between incubations. Reactions were terminated with 2N H₂SO₄, and absorbance was detected at 450 nm. Detailed antibody information can be found in Table 1.

CBD analysis:

Cells were collected, washed with 1X PBS, then deproteinized by adding 500 μl of cold 1:1 acetonitrile/methanol (ACN/MeOH) containing 20 nM CBD-d9 deuterated internal standard (exact amount added was 10 pmol) to each sample. Samples were mixed by inverting the tubes then vortexing for ~30 sec. Samples were then incubated at −80°C for 15 min to ensure maximal protein precipitation. Samples were then centrifuged at 16,100 × g for 10 min at 4°C. The supernatant was carefully transferred to a sample analysis vial. 10 μl of each sample was injected onto a Waters Acquity UPLC system that was interfaced with a Thermo Quantum Access Max triple-quadrupole mass spectrometer. Chromatographic separation was carried out using an Acquity UPLC BEH C18 column (2.1 mm × 100 mm, 1.7 μm) equipped with a precolumn (2.1 mm × 5 mm, 1.7 μm) at 40°C. The mobile phases were 95/5 (v/v) H₂O/ACN + 0.1% acetic acid (mobile phase A) and ACN + 0.1% acetic acid (mobile phase B). Mobile phase gradient conditions were as follows: hold at 95% A and 5% B for 0.5 min, linear increase of B to 99% over 6 min, hold at 99% B for 1.5 min, then decrease B to 5% in 0.5 min and re-equilibrate column at the initial conditions for 4 min. The overall run time was 12 min, and the mobile phase flow rate was set at 0.2 mL/min between 0–0.5 min, stepped up to 0.25 mL/min at 0.51 min and maintained at that rate for 7 min, followed by a step down to 0.2 mL/min between 8–12 min. CBD was analyzed in the positive ion mode by electrospray ionization and selected reaction monitoring (SRM). The respective SRMs for CBD and CBD-d9 were 315.1>193.1 and 324.2>202.2 (collision energy, 19; tube lens voltage, 89; arbitrary units). A nine-point CBD calibration curve (ranging from 0.3–500 nM) was prepared in MeOH in the exact same manner as the unknowns.

Statistical analysis

Data are presented as mean ± SD. Fold change values were transformed using natural log (fold change + 1) prior to ANOVA analyses. For vehicle comparisons, the drug-treated samples from triplicate cultures were normalized to an average of the respective vehicle to obtain % vehicle. The drug-treated % vehicle values were then averaged for each dog (i.e., separate culture set up) then plotted as N=1. Percent data were log transformed prior to ANOVA. One-way ANOVA was performed followed by a post hoc test to determine statistical differences from control (stimulation plus VH) at p < 0.05.

Results

RT-PCR and IHC demonstrate expression of CB₁ and CB₂ in canine immune cells.

A detailed analysis of cannabinoid receptor expression was performed in various canine tissues and PBMCs. First, RT-PCR was used to detect gene expression. In the brain, there were similar levels of CNR1 and CNR2 gene expression. Higher CNR1 expression was expected in the brain as compared to CNR2; however, the brain samples were from frozen, healthy canine brains obtained in routine necropsies, so the brains were not perfused. It is likely then that the relatively high CNR2 gene expression in the brain came from blood. Indeed, an examination of whole blood demonstrated high expression of CNR2 but almost no detectable CNR1 (Fig. 1). Next, IHC was used to examine cannabinoid receptor protein expression in brain and spleen. In the brain, CB₁ was expressed, whereas CB₂ was not expressed at detectable limits (Fig. 2). There was strong CB₁ immunostaining of the cytoplasm of neurons, as well as scattered immunopositive astrocytes. The cerebral cortex, at the level of the hippocampus, was stained in all animals, and staining was scattered and not localized to a specific location. In the spleen, CB₁ and CB₂ were both expressed (Fig. 3). Both CB₁ and CB₂ immunostaining were present in the cytoplasm of lymphocytes, as well as some scattered macrophages. For both receptors, staining was scattered and variably distributed throughout the sections, and in both the spleen and brain, staining abundance varied between dogs.

Figure 1. RT-PCR analysis of CNR1 and CNR2.

RNA was isolated from whole blood from healthy untreated dogs or healthy brain tissue obtained in routine necropsies. RT-qPCR was performed, and fold change was calculated as compared to CNR1 expression in blood.

Figure 2: Immunohistochemical staining for CB₁ receptor in the brain.

A, B: CB₁ receptor was multifocally expressed in the cytoplasm of neurons within the cortex (black arrows). C: Scattered astrocytes were immunopositive for CB₁ (yellow arrows). D: Negative control. Magnification: 400X.

Figure 3. Immunohistochemical staining for CB₁ and CB₂ receptors in the spleen.

A, B: CB₁ receptor was multifocally expressed in the cytoplasm of lymphocytes and scattered macrophages. D, E: CB₂ receptor was multifocally expressed in the cytoplasm of lymphocytes and scattered macrophages. C, F: Negative control. Magnification: 400x.

PBMC yields.

Using the colony of healthy Treeing Walker Coonhounds, approximately 50 ml of whole blood was obtained from one dog at one time. At the time of blood collection, the dogs were not receiving any other treatments. After differential centrifugation with Histopaque 1077, total viable PBMCs were enumerated and showed that between 10–200 × 10⁶ cells were obtained, with an average of approximately 25 × 10⁶ PBMCs (Fig. 4).

Figure 4. PBMC counts.

Approximately 50 ml of blood was obtained from healthy, untreated dogs. PBMCs were isolated from buffy coats following differential centrifugation with Histopaque 1077 and enumerated. Each dog was assigned a different letter and dots represent separate draws from 1–5.

Flow cytometric analysis of CB₁ and CB₂ in canine PBMCs.

First, antibody titrations were performed for both antibodies for flow cytometry and the highest staining was obtained using 0.5 μl of antibody, providing a distinct separation from a negative population (Figs. 5A and 6). After gating on live single lymphocytes (Fig. 5B), it was shown that canine PBMCs expressed both CB₁ and CB₂ on their cell surface, which was expected (Fig. 6). Although it was difficult to directly compare relative levels of expression since staining had to be performed separately (both CB₁ and CB₂ are rabbit antibodies which prevents distinguishing CB₁ from CB₂ expression if done simultaneously), the PBMC staining using flow cytometry was consistent with the staining patterns observed in the spleen with IHC (Fig. 3).

Figure 5. CB₁ and CB₂ receptor flow cytometry staining index and gating strategy.

Freshly-isolated PBMCs were washed with 1X PBS and stained with NIR fixable viability dye (FVD). PBMCs were then treated with Fc block and incubated for 30 min with either CB₁ or CB₂ rabbit polyclonal antibodies. After a wash step, PBMCs were incubated for 30 min with 0.5 μl donkey anti-rabbit secondary antibody conjugated to AlexaFluor 488 (detected in the FITC channel). A, Staining index. Cells were incubated with varying concentrations of either the CB₁ or CB₂ primary antibodies, and staining index was calculated. B, Cells are gated on live, lymphocytes, and single cells before setting up the FITC gate to detect either CB₁ or CB₂ receptors (done separately).

Figure 6. CB₁ and CB₂ receptor expression in dog PBMCs.

Freshly-isolated PBMCs were washed with 1X PBS and stained with NIR fixable viability dye (FVD). PBMCs were then treated with Fc block, primary antibodies for either CB₁ or CB₂ receptors, and the secondary FITC antibody. Expression histograms for 3 separate dogs show that PBMCs express both CB₁ and CB₂ receptors.

Effect of CBD and THC on cytokine gene expression in canine PBMCs.

With the demonstration that canine PBMCs expressed both CB₁ and CB₂, the effects of CBD and THC were examined. Cells were treated with cannabinoids for 15 min and stimulated with P/I for 5 hr. There was robust stimulation of IL2, IFNG, IL17A and TNFA with P/I, and there was significant suppression of all genes by THC but not CBD in at least one dog (Fig. 7). Overall, the results were a bit underwhelming as both THC and CBD have been reported to significantly affect cytokine gene expression in a wide variety of human and animal cells [35] so longer times of stimulation with P/I or Concanavalin A (ConA) were also examined (Supplemental figures 1 and 2). Again, only modest effects with either CBD or THC were observed at these longer times. These results also suggested that even these modest cannabinoid effects were dog-dependent.

Figure 7. RT-PCR analysis of cytokines after 5 hr stimulation.

PBMCs (~2×10⁶ cells/well) were treated with vehicle (VH, 0.1% ethanol) or CBD or THC for 30 min prior to cellular activation with P/I (12.5 ng/ml/0.8 μM) for 5 hr. Total RNA was isolated, and RT-PCR was performed. Each letter is a different dog, and the number refers to a different blood draw for that particular dog. Fold change was calculated as compared to untreated (untx) samples. Lower case letters indicate a statistically significant difference at p < 0.05 from the VH-treated stimulated sample. Three dog cultures were used for all analyses.

Effect of CBD and THC on cytokine protein expression in canine PBMCs.

Since the effect of cannabinoids on cytokine gene expression was modest, the effects of cannabinoids on cytokine protein were examined using intracellular staining and detection by flow cytometry and cytokine secretion by ELISA. With these experiments, the hypothesis that delivery vehicles would influence cannabinoid effects was also tested. All assessments to this point were conducted using 0.1% ethanol as the vehicle. Using DMSO for the vehicle allowed for a lower percent vehicle (i.e., 0.02% DMSO) since cannabinoids have higher solubility in DMSO as compared to ethanol. First, intracellular IL-2 and IFN-γ were examined in CD4⁺ and CD8⁺ T cells. As seen in Figure 8, CBD and THC produced significant suppression of percent IFN-γ-expressing cells only when delivered in DMSO as compared to ETOH. There were little to no effects of CBD or THC on percent of IL-2-expressing T cells, although there were a few data points in which CBD or THC delivered in DMSO trended toward suppression (Figure 9). Similar effects were seen on the mean fluorescence intensity values for IFN-γ and IL-2 (Supplemental Figures 3 and 4), although effects were modest. It was also interesting to again note the variability of the data points since each point represents data from 3–5 dogs, indicating that cannabinoid effects are dog-dependent. Second, cytokine secretion was examined by ELISA. IL-17A secretion was not detected at 5 hr post P/I (data not shown) but there was not much of an effect of CBD or THC on IL-2 or IFN-γ secretion at this stimulation time (Figure 10).

Figure 8. Intracellular staining of IFN-γ after 5 hr stimulation.

PBMCs (~2×10⁶ cells/well) were treated with vehicle (VH; either ETOH, 0.1% or DMSO, 0.02%) or CBD or THC (1, 5 or 10 μM) for 30 min prior to cellular activation with P/I (12.5 ng/ml/0.8 μM) for 5 hr. Cells were washed and stained with viability dye and antibodies directed against extracellular CD4 or CD8. After fixing, cells were permeabilized and stained with antibodies directed against intracellular IFN-γ. Symbols are mean percent IFN-γ ± SD using separate cultures that were normalized to the respective VH within each culture. * p < 0.05 as compared to the respective VH. Three to five dog cultures were used for analyses.

Figure 9. Intracellular staining of IL-2 after 5 hr stimulation.

PBMCs (~2×10⁶ cells/well) were treated with vehicle (VH; either ETOH, 0.1% or DMSO, 0.02%) or CBD or THC (1, 5 or 10 μM) for 30 min prior to cellular activation with P/I (12.5 ng/ml/0.8 μM) for 5 hr. Cells were washed and stained with viability dye and antibodies directed against extracellular CD4 or CD8. After fixing, cells were permeabilized and stained with antibodies directed against intracellular IL-2. Symbols are mean percent IL-2 ± SD using separate cultures that were normalized to the respective VH within each culture. * p < 0.05 as compared to the respective VH. Three to five dog cultures were used for analyses.

Figure 10. ELISA of secreted IL-2 or IFN-γ after 5 hr stimulation.

PBMCs (~2×10⁶ cells/well) were treated with vehicle (VH; either ETOH, 0.1% or DMSO, 0.02%) or CBD or THC (1, 5 or 10 μM) for 30 min prior to cellular activation with P/I (12.5 ng/ml/0.8 μM) for 5 hr. Supernatants were collected and used without further dilution to quantify the concentration of IL-2 or IFN-γ using a standard curve in ELISA. Symbols are mean pg/ml IL-2 or IFN-γ using separate cultures that were normalized to the respective VH within each culture. None were significant. Three to five dog cultures were used for analyses.

Influence of vehicle on cannabinoid-mediated alteration of cytokine gene expression.

With the suggestion that CBD or THC delivered in DMSO produced more robust suppression of percent of T cells expressing IFN-γ intracellularly, gene expression was re-evaluated. Using cultures from two different dogs, both CBD and THC modestly suppressed IFNG gene expression when delivered in DMSO as compared to ETOH (Figure 11).

Figure 11. RT-PCR analysis of cytokines after 5 hr stimulation.

PBMCs (~2×10⁶ cells/well) were treated with vehicle (VH; either ETOH, 0.1% or DMSO, 0.02%) or CBD or THC (10 μM) for 30 min prior to cellular activation with P/I (12.5 ng/ml/0.8 μM) for 5 hr. Total RNA was isolated, and RT-PCR was performed. Fold change was calculated as compared to untreated (untx) samples then samples were normalized to the respective VH within each culture. Three dog cultures were used for all analyses.

Influence of vehicle on cannabinoid delivery to canine PBMCs.

To determine if ETOH or DMSO affected the concentration of cannabinoids that was in the cells, LC-MS/MS analysis of cell pellets was conducted. Cells were treated with CBD or THC for 30 min to mimic the pre-treatment time before cells were stimulated with P/I. Figure 12 reveals that delivery vehicle did not alter cellular concentrations of cannabinoids, but there was more THC in/associated with the cells as compared to CBD at the same concentration, regardless of vehicle.

Figure 12. Cellular cannabinoid concentrations after 30 min treatment.

PBMCs (~2×10⁶ cells/well) were treated with vehicle (VH; either ETOH, 0.1% or DMSO, 0.02%) or CBD or THC (1, 5 or 10 μM) for 30 min. Cell pellets were collected, washed with PBS, and extracted using 100% methanol. Extracts were analyzed using LC-MS/MS against cannabinoid standard curves. Bars are mean ± SD from triplicate wells. One dog culture was used for analysis.

Discussion

These data are some of the first data to characterize the expression of CB₁ and CB₂ in canine PBMCs. We also examined expression of both CB₁ and CB₂ in canine spleen and brain by flow cytometry and IHC. These data showed that CB₁ and CB₂ are both expressed in spleen and PBMCs, and CB₁ is exclusively expressed in brain. These results are consistent with other species, and these results confirm what others have shown, especially for CB₁ in dogs [12, 15].

These data further demonstrated that CBD and THC could suppress pro-inflammatory gene expression but that the response was dog- and vehicle-dependent. Of all the conditions initially tested with cannabinoids delivered in ETOH, THC was able to suppress all genes following P/I stimulation in at least one dog at the 5-hr timepoint. This manner of stimulation has been used routinely in whole blood to evaluate time course effects of various immunosuppressants in dogs [34]. However, it was a bit surprising that the cannabinoid effects on cytokine gene expression were quite modest. One explanation could be that only a single breed of healthy dogs was used, so the possibility exists that other breeds or the presence of various disease states might influence how immune responses are affected by cannabinoids.

Another possibility for the modest effect of cannabinoids on cytokines might be that efficacy is dependent on delivery vehicle. For the initial gene expression studies, we used 0.1% ETOH, but we then showed that equimolar concentrations of THC or CBD produced more pronounced immune suppression on percentage of cells expressing intracellular cytokines or IFNG gene expression when delivered in DMSO. The mechanism for the more pronounced immune suppressive effects by which cannabinoids when delivered in DMSO as compared to ETOH is not clear. We did investigate if cannabinoid delivery in DMSO as opposed to ETOH would in fact deliver more cannabinoid/cell, but this was not the case. Interestingly these studies revealed that more THC gets to or into the cells as compared to equimolar concentrations of CBD. These results might suggest that there is more THC bound to extracellular or intracellular receptors in canine PBMCs. One possibility is that THC is bound to peroxisome proliferator-activated receptor (PPAR)-γ [36, 37]. Even though THC has low affinity for PPAR-γ, it is higher than CBD [36], suggesting that the relatively higher cellular concentration of THC could be due to higher binding of THC to intracellular PPAR-γ.

The cannabinoid-mediated suppression of cytokine gene expression might account, at least in part, for the reported efficacy of cannabinoids in canine immune-mediated diseases. For instance, there have been several placebo-controlled cannabinoid studies conducted in dogs with osteoarthritis. In the first, a hemp product containing CBD, cannabigerol, and cannabichromene was administered orally to osteoarthritic dogs for 10 weeks (4-week administrations separated by a 2-week washout in between). After treatment, dogs demonstrated a decrease in pain and increase in activity as assessed by the canine brief pain inventory and Hudson scale; however, there were no changes in weight bearing load on the affected limb(s) [29]. In the second study, naked CBD and liposomal CBD were compared at doses of 20–50 mg/kg/day. Each dog was evaluated by a veterinarian team on day 0 and day 30 and evaluated by the owner on days 0, 30, and 45 using the Helsinki chronic pain index, which is an 11-point evaluation of chronic pain. Based on owner and veterinarian reports, dogs receiving the 50 mg/day dose of naked CBD or the 20 mg/day liposomal CBD showed significant reductions in pain up to 15 days after the last dose was given [30]. In the third study, a CBD oil product containing highly purified CBD in addition to hemp seed oil was given to dogs with naturally occurring osteoarthritis of the carpus, elbow, shoulder, tarsus, stifle, or hip joint at 2.5 mg/kg CBD orally twice daily. While there were no statistically significant changes by CBD noted on gait analysis, this was a small study with various dog breeds, with some dogs also taking other anti-inflammatory medications concurrently [31]. In a fourth study, a cannabis extract containing CBD in medium chain triglyceride was administered to dogs with OA at 2 mg/kg every 12 hr for 12 weeks [33]. In this randomized placebo-controlled study, there was decreased pain perception and increased quality of life [33]. Many of these studies were included as part of a systematic review of the safety and efficacy of cannabinoids for various diseases in dogs [4]. The major finding from the systematic review was that cannabis appeared to reduce pain and increase activity in dogs with osteoarthritis with minimal adverse effects [4] but that more research is necessary in more dogs and more breeds. The need for additional studies examining the potential anti-inflammatory efficacy of cannabinoids in dogs is further supported by the findings of dog-dependent responsiveness to cannabinoids in vitro in this study.

Additional evidence that modulation of the ECS is effective in dogs comes from a placebo-controlled trial in dogs with atopic dermatitis. In this study, a topical endocannabinoid reuptake inhibitor was applied for 21 days during which a dust mite allergen challenge was introduced every 3 days to elicit pruritis. Using the Canine Atopic Dermatitis and Extent Severity Index (CADESI-03) and blinded assessment of pruritic acts, the topical agent reduced severity of flares after dermatitis challenge [38]. CBD also reduced scratching and allowed reduction of other therapeutics in a small study in dogs with atopic dermatitis [39].

A few studies have demonstrated modulation of cannabinoid receptors in inflamed tissues; however, the results are not consistent. In dogs with atopic dermatitis, it appeared that cannabinoid receptor protein expression was increased [13]. CB₂ protein expression was also increased in spinal cord lesions of dogs with intraspinal spirocercosis, an inflammatory parasitic disease in the central nervous system [40]. On the other hand, in dogs with chronic colonic dysmotility (which is often associated with inflammatory bowel-like disease in dogs), cannabinoid receptor gene expression was downregulated as compared to control dogs [41]. Together these data show that the ECS is altered in dogs with inflammatory diseases, but it will be important to clarify this in future studies by comparing cannabinoid receptor expression in activated versus non-activated PBMCs and/or in cells derived from healthy dogs versus those with immune-mediated diseases. Especially if we observe consistent downregulation of cannabinoid receptors in an activated (or inflamed) state, these data might suggest that the ECS plays a role in immune homeostasis in the dog, as has been suggested in other animals and humans [35]. It will also be important to further characterize cannabinoid effects on additional immune function endpoints using broad concentration responses of cannabinoids, extensive time course studies, and alternate modes of cellular activation. Furthermore, it will be important to determine whether the THC effects are mediated by CB₁, CB₂, both, or neither receptor since we have established that canine PBMCs do express both cannabinoid receptors.

There have been several pharmacokinetic studies using various cannabinoid preparations in dogs [5–7, 42–44] so it is well established that CBD and THC are bioavailable in dogs. For instance, in one study in which CBD was delivered by various routes, peak plasma CBD was highest following oral administration as compared to intranasal or intrarectal (oral > nasal > rectal) [5]. Given the likely immunosuppressive effects of cannabinoids, cannabinoids might have a potential role in the treatment of many immune-mediated diseases, such as hemolytic anemia, immune-mediated thrombocytopenia, and myasthenia gravis. Consideration should be given to using cannabinoid-based medicines as adjuvant therapies or in cases for which all other therapeutic options have been exhausted. This study was limited to evaluation of a few pro-inflammatory cytokines at an early time point following stimulation, so future work will include extensive assessment of immune function in response to various stimuli over various times. The studies also revealed a difference in how much THC gets in/into the cells, which suggests a possible intracellular receptor. Regardless, these studies represent the first steps in understanding the endocannabinoid system in canine immune responses and might also provide pre-clinical data to support the development of cannabinoid-based drugs in veterinary medicine.

Supplementary Material

1

Supplemental Figure 1. RT-PCR analysis of IL2 and IFNG after overnight stimulation. PBMCs (~2×10⁶ cells/well) were treated with vehicle (VH, 0.1% ethanol) or CBD or THC for 30 min prior to cellular activation with P/I (40 nM/0.5 μM) or ConA (10 μg/ml) overnight (18–24 hr). Total RNA was isolated, and RT-PCR was performed. Each letter is a different dog, and the number refers to a different blood draw for that particular dog. Fold change was calculated as compared to untreated (untx) samples. Lower case letters indicate a statistically significant difference at p < 0.05 from the VH-treated stimulated sample.

2

Supplemental Figure 2. RT-PCR analysis of IL17A and TNF after overnight stimulation. PBMCs (~2×10⁶ cells/well) were treated with vehicle (VH, 0.1% ethanol) or CBD or THC for 30 min prior to cellular activation with P/I (40 nM/0.5 μM) or ConA (10 μg/ml) overnight (18–24 hr). Total RNA was isolated, and RT-PCR was performed. Each letter is a different dog, and the number refers to a different blood draw for that particular dog. Fold change was calculated as compared to untreated (untx) samples. Lower case letters indicate a statistically significant difference at p < 0.05 from the VH-treated stimulated sample.

3

Supplemental Figure 3. Intracellular staining of IFN-γ after 5 hr stimulation. PBMCs (~2×10⁶ cells/well) were treated with vehicle (VH; either ETOH, 0.1% or DMSO, 0.02%) or CBD or THC (1, 5 or 10 μM) for 30 min prior to cellular activation with P/I (12.5 ng/ml/0.8 μM) for 5 hr. Cells were washed and stained with viability dye and antibodies directed against extracellular CD4 or CD8. After fixing, cells were permeabilized and stained with antibodies directed against intracellular IFN-γ. Symbols are mean MFI (mean fluorescence intensity) IFN-γ ± SD using separate cultures that were normalized to the respective VH within each culture. * p < 0.05 as compared to the respective VH. Three to five dog cultures were used for analyses.

4

Supplemental Figure 4. Intracellular staining of IL-2 after 5 hr stimulation. PBMCs (~2×10⁶ cells/well) were treated with vehicle (VH; either ETOH, 0.1% or DMSO, 0.02%) or CBD or THC (1, 5 or 10 μM) for 30 min prior to cellular activation with P/I (12.5 ng/ml/0.8 μM) for 5 hr. Cells were washed and stained with viability dye and antibodies directed against extracellular CD4 or CD8. After fixing, cells were permeabilized and stained with antibodies directed against intracellular IL-2. Symbols are mean MFI (mean fluorescence intensity) IL-2 ± SD using separate cultures that were normalized to the respective VH within each culture. * p < 0.05 as compared to the respective VH. Three to five dog cultures were used for analyses.

Highlights.

CB₁, but not CB₂, protein expression was detected in canine brain

CB₁ and CB₂ are expressed in canine spleen as assessed by IHC

CB₁ and CB₂ were also detected in canine PBMCs as assessed by flow cytometry

Cannabinoid-mediated suppression of cytokine production was dog and vehicle-dependent

Abbreviations:

CB

cannabinoid receptor

IFN

interferon

IL

interleukin

TNF

tumor necrosis factor