Δ⁹-Tetrahydrocannabinol Suppresses Cytotoxic T Lymphocyte Function Independent of CB₁ and CB₂, Disrupting Early Activation Events

Abstract

Previously, CD8⁺ T cells were found to be a sensitive target for suppression by Δ⁹-tetrahydrocannabinol (Δ⁹-THC) in a murine model of influenza infection. To study the effect of Δ⁹-THC on CD8⁺ cytotoxic T lymphocytes (CTL), an allogeneic model of MHC I mismatch was used to elicit CTL. In addition, to determine the requirement for the cannabinoid receptors 1 (CB₁) and 2 (CB₂) in Δ⁹-THC-mediated CTL response modulation, mice null for both receptors were used (CB₁ ^−/−CB₂ ^−/−). Δ⁹-THC suppressed CTL function independent of CB₁ and CB₂ as evidenced by reduction of ⁵¹Cr release by CTL generated from CB₁ ^−/−CB₂ ^−/− mice. Furthermore, viability in CD4⁺ and CD8⁺ cells was reduced in a concentration-dependent manner with Δ⁹-THC, independent of CB₁ and CB₂, but no effect of Δ⁹-THC on proliferation was observed, suggesting that Δ⁹-THC decreases the number of T cells initially activated. Δ⁹-THC increased expression of the activation markers, CD69 in CD8⁺ cells and CD25 in CD4⁺ cells in a concentration-dependent manner in cells derived from WT and CB₁ ^−/−CB₂ ^−/− mice. Furthermore, Δ⁹-THC synergized with the calcium ionophore, ionomycin, to increase CD69 expression on both CD4⁺ and CD8⁺ cells. In addition, without stimulation, Δ⁹-THC increased CD69 expression in CD8⁺ cells from CB₁ ^−/−CB₂ ^−/− and WT mice. Overall, these results suggest that CB₁ and CB₂ are dispensable for Δ⁹-THC-mediated suppression and that perturbation of Ca²⁺ signals during Tcell activation plays an important role in the mechanism by which Δ⁹-THC suppresses CTL function.

Introduction

Several cannabinoid compounds including Δ⁹-THC have been implicated in the modulation of immune responses (Howlett et al. 2002; Klein et al. 2003). Cannabinoid compounds exert their activity, in part, by ligation of two identified targets, CB₁ and CB₂ (Maresz et al. 2007). However, studies using antagonists to these receptors, low affinity cannabinoid agonists, and mice lacking CB₁ and/or CB₂, have demonstrated that not all aspects of immune modulation by cannabinoid compounds can be attributed to CB₁ and/or CB₂ (Kaplan et al. 2003; Buchweitz et al. 2008; Springs et al. 2008; De Petrocellis and Di Marzo 2010). Thus, despite the ability to bind CB₁ and CB₂, the mechanism and cellular target(s) for Δ⁹-THC within the immune system remain elusive. In particular, recent studies revealed that in T cells, the effect of Δ⁹-THC on immune competence was found to occur independent of CB₁ and CB₂ (Kaplan et al. 2003; Rao and Kaminski 2006), an observation concordant with CB₁ and CB₂ expression, which is much lower on T cells compared to other leukocyte subpopulations (Galiegue et al. 1995). To further investigate the effects of Δ⁹-THC on Tcell responses, an in vivo influenza challenge model was used and showed that the CD8⁺ T cell population exhibited marked sensitivity to modulation by Δ⁹-THC (Buchweitz et al. 2007). These results suggested that CD8⁺ T cells might be a particularly sensitive T cell subset to cannabinoid-mediated immune modulation and might also represent a useful cellular target to gain novel insights into the molecular mechanisms of cannabinoid action.

The CD8⁺ surface marker distinguishes a lineage of T cells that are capable of interaction with major histocompatibility complex I (MHC I). To become fully functional, naïve CD8⁺ T cells must be activated, proliferate, and differentiate into CTL effectors. CTL are important in antiviral and antitumor immune responses. In addition to releasing cytokines pleiotropically involved in immune defense, CTL release cytotoxic granules or interact by cell contact with target cells to induce apoptosis in order to rid tissues of virus-infected or neoplastic cells (Boissonnas et al. 2007; Stinchcombe and Griffiths 2007). Lysis of target cells and the production of IFNγ are most often used as indicators of CTL function (Slifka et al. 1999; Sun et al. 2003). To experimentally elicit CTL from naïve CD8⁺ T cells, several models have been developed, among them is an allogeneic model using P815 (DBA2-derived) tumor cells to activate lymphocytes from a haplotype mismatched donor, such as C57Bl/6 mice. Using this model, CTL are generated in response to a T cell receptor-MHC mismatch and upon subsequent encounter, their effector function can be assayed (Engers et al. 1975). The objective of this study was to determine the effect of Δ⁹-THC and the involvement of CB₁ and CB₂ on CTL elicitation and function.

Methods

Animals

C57Bl/6 mice were ordered from the National Cancer Institute (NCI) and CB₁ ^−/−CB₂ ^−/− mice on C57Bl/6 background were bred in-house. CB₁ ^−/−CB₂ ^−/− mice were a kind gift of Dr. Andreas Zimmer at the Universität of Bonn, Germany. Mice were housed at Michigan State University in a pathogen free animal research housing facility at 21 to 24°C and 40 to 60% relative humidity with a 12-h light/dark cycle. Water and food (Purina Lab Chow) were available ad libitum. All protocols and procedures were performed in accordance with guidelines set forth by the Institutional Animal Care and Use Committee at Michigan State University. CB₁ ^−/−CB₂ ^−/− mice were routinely genotyped to validate absence of CBs using commercial primers for cnr1 and custom primers for cnr2 from Applied Biosystems (Kaplan et al. 2010).

Chemicals and reagents

Δ⁹-THC was obtained from the National Institute on Drug Abuse (Bethesda, MD). Ethanol was purchased from Decon Labs (King of Prussia, PA). Ionomycin (Io) and dimethylsulfoxide (DMSO) were purchased from Sigma Chemical Co. (St. Louis, MO). RPMI media was obtained from Gibco Invitrogen (Carlsbad, CA), and ⁵¹Cr as sodium chromate was obtained from Perkin Elmer (Waltham, MA).

T cell elicitation for generation of functional CTL

C57Bl/6 (WT) and CB₁ ^−/−CB₂ ^−/− mice were euthanized, the spleens isolated in a sterile environment, and splenocytes enumerated using a Coulter Counter (Beckman Coulter, Brea, CA). P815 cells were irradiated with 3000 rads to prevent proliferation, washed 3 times with RPMI and counted using a hemacytometer. Splenocytes and irradiated P815 cells were combined at 1×10⁶ and 1×10⁵ cells, respectively, in RPMI 1640 supplemented with 5% bovine calf serum (BCS) in a total volume of 200 μL in a round bottom 96 well plate. The plates were incubated in a humidified incubator with 5% CO₂ at 37°C for the indicated amounts of time.

Drug treatment

At the time of co-culture of splenocytes and irradiated P815, Δ⁹-THC (1, 5, 10 μM), vehicle (VH, 0.1% ethanol) or RPMI (NA) was added. All Δ⁹-THC treatments had the same ethanol content (0.1%) as vehicle control.

⁵¹Cr release assay

After elicitation, cells were harvested and washed twice with RPMI 1640 media without serum. P815 cells were washed once and 1×10⁶ cells were incubated in the presence of Na₂ ⁵¹CrO₄ for 1 h in 10% fetal bovine serum (FBS) supplemented RPMI 1640 media in a volume of less than 50 μL. After incubation P815 cells were washed 3 times using RPMI 1640 media without serum. ⁵¹Cr-labeled P815 cells were adjusted to 1×10⁵ cells/mL in 2% FBS RPMI media. Elicited CTL were adjusted in 2% FCS complete RPMI media to ratios ranging from 50 (5×10⁵) to 1 (1×10⁴ cells) : 1 (1×10⁴) P815 cells, depending on the experimental design, in a volume of 200 μL. After co-culture, elicited CTL and P815 were added to a 96 well round bottom plate and centrifuged at 200 g for 1 min to force cellular interactions. Control wells for spontaneous release (200 μL of P815 only) and total release (1% Triton-X 100 in 200 μL of P815 cells in RPMI) were used to determine the range of experimental release. After 5 h of co-culture in a humidified incubator with 5% CO₂ at 37°C, cell lysis was assessed by aliquoting 100 μL of supernatant from each well, which represents the experimental release. The cytolytic activity was calculated as follows: % Release = (experimental release − spontaneous release)/(experimental release − total release)× 100.

IFNγ T cell functional analysis

CTL were elicited as described above for generation of CTL. After 5 days, cells were harvested and co-cultured with P815 at a ratio of 10:1 (see above) for 12 h in the presence of brefeldin A to prevent IFNγ release and allow for detection by fluorescently labeled antibody. After co-culture, cells were prepared for fluorescent antibody staining (described below).

Proliferation assay

Prior to elicitation, splenocytes were incubated with Cell Trace carboxyfluorescein succinimidyl ester (CFSE) dye (Invitrogen, Carlsbad, CA) according to manufacturer’s instructions. CTL were elicited as described above for generation of CTL. Dilution of dye staining is indicative of proliferation and staining profiles fromelicited samples were compared to a 24 h SPLC only control, incubated without P815.

T cell activation assays

CTL were elicited as described for generation of CTL; however, RPMI supplemented with 2% FCS was used instead of 5%. For direct stimulation, 0.5 μM Io dissolved in 0.01% DMSO in RPMI was used as the Ca²⁺ ionophore to induce intracellular Ca²⁺ levels, thereby increasing CD69 expression. After 6 h, CD69 surface staining was performed using antibodies specific for CD69 and after 24 h for CD25 (staining described in detail below).

Procedure for immunofluorescent staining for flow cytometry

CTL were elicited as described above. Cells were washed in 96 well round bottom plates with HBSS once and incubated with Near IR LIVE/DEAD (Invitrogen) dye to assess viability according to manufacturer’s instructions. This dye stains cells that have lost membrane integrity. Cells were washed twice with FACS buffer (1X HBSS, 1% bovine serum albumin (BSA), 0.1% sodium azide). Subsequently, Fc receptors were blocked using Purified Rat Anti-Mouse CD16/CD32 (BD Pharmingen, San Diego, CA). To stain surface proteins, cells were incubated for 20 min with 0.25–0.5 μg of CD4 (clone RM4-5, Biolegend, San Diego, CA) and CD8 (53–6.7, Biolegend). For the T cell activation experiments, cells were additionally stained with CD69 (H1.2 F3, Biolegend) and CD25 (PC61, Biolegend). Cells were then washed twice and fixed with Cytofix (BD, Franklin Lakes, NJ). After re-stimulation (described above), on the day of flow cytometric analysis, cells were washed once, then permeabilized with Perm/Wash Buffer (BD, Franklin Lakes, NJ) by incubation for 20 min and stained with IFNγ (XMG1.2, Biolegend) for 30 min. All experiments included single stain controls to compensate for fluorescence overlap between detectors. Samples were individually analyzed for percent positive staining as defined by a gate to perform statistical analysis and concatenated by treatment group for graphing using FlowJo v.8.8.6. for Macintosh.

Statistical analysis

Samples were analyzed using ANOVA or non-parametric equivalent Kruskal-Wallis. A p-value of 0.05 or less was deemed statistically significant between experimental and relative control. Regression analysis was performed treating values of each independent variable as a single datum point. All Δ⁹-THC treated samples were compared to vehicle (VH, 0.1% ethanol) control and comparisons between WT and CB₁ ^−/−CB₂ ^−/− were performed in naïve (NA) groups. Stimulation effect was determined between resting (NA–0 h) and NA at the indicated time point of sample collection. Statistical analyses were performed using Graph Pad Prism v4.03.

Results

Kinetics of CTL response

In order to determine the kinetics of the P815-induced CTL response, CTL were elicited for 3, 5, 7, and 9 days after co-culture with irradiated P815 cells. Splenocytes from WTand CB₁ ^−/−CB₂ ^−/− mice were used as the source of naïve CD8⁺ T cells. Elicited CTL were assayed for CTL activity by ⁵¹Cr release assay using ⁵¹Cr-labelled P815 target cells. A ratio of 50:1 CTL:P815 cells was used to maximize release, thereby increasing sensitivity on non-peak days. ⁵¹Cr release is a result of CTL-dependent lysis of P815 cells and the peak day of CTL activity occurred at day 5 post elicitation (Fig. 1). At the peak day of CTL elicitation, CB₁ ^−/−CB₂ ^−/− splenocytes had a statistically significant reduction in CTL activity, when compared to WT (p≤0.01). The same trend was observed for non-peak days, although it was not deemed statistically significant. Subsequent experiments determining CTL effector activity under various conditions were performed at day 5 post elicitation.

Fig. 1.

CTL activity peaks at day 5 after elicitation. To determine the peak day of CTL activity, splenocytes from WT and CB₁ ^−/−CB₂ ^−/− mice were co-cultured with irradiated P815 cells to elicit an allogeneic CTL response. On various days after elicitation, CTL were harvested and assayed for target (P815) lysing activity by reincubation at a ratio of 50:1 (Effector CTL : P815 Targets) with ⁵¹Cr labeled P815 cells. Percent release was calculated as described in Methods (n=4). The experiment was performed once, while more replicate experiments were performed for WTonly. ## p≤0.01 indicates difference between WT and CB₁ ^−/−CB₂ ^−/−

Δ⁹-THC suppressed CTL responses during elicitation but not during the effector response

To determine the sensitivity of CTL function to Δ⁹-THC treatment in the P815 allogeneic model, Δ⁹-THC and/or vehicle (0.1% ethanol) were added directly to culture during either the elicitation or effector phase of the CTL response. A concentration of 10 μM Δ⁹-THC was used to determine sensitivity of CTL to immune modulation, a concentration that historically impaired T cell immune responses in the absence of direct cytotoxicity (Springs et al. 2008). Δ⁹-THC suppressed CTL activity only when added prior to, but not after, the elicitation phase (Fig. 2). Likewise, repeated addition of Δ⁹-THC during elicitation and effector phases did not further reduce CTL activity below single addition of the drug during the elicitation phase (Fig. 2).

Fig. 2.

Δ⁹-THC suppresses CTL activity during elicitation but not effector phase. Splenocytes were treated with Δ⁹-THC and/or VH (0.1% ethanol) and stimulated with irradiated P815 cells for 5 days. After harvest, CTL were reincubated at a 10:1 ratio with ⁵¹Cr labeled P815 cells in the presence of Δ⁹-THC and appropriate controls (NA–naïve: no treatment; VH–0.1% ethanol) (n=4). Data shown are representative of two repeat experiments. * p≤0.05 indicates difference as compared to VH

Concentration-dependent suppression of CTL effector function

Next, we assessed whether CTL responses were suppressed by Δ⁹-THC in a concentration-dependent manner. At E:T ratios ranging from 10:1 to 1:1, Δ⁹-THC suppressed CTL responses in a concentration-dependent manner with statistically significant suppression at concentrations as low as 1 μM (Fig. 3a). Also, lower E:T ratios resulted in lower CTL activity, demonstrating that the elicited cell populations are responsible for the cytolytic activity and that Δ⁹-THC suppressed CTL activity independent of the E:T ratio (Fig. 3a). Regression analysis was performed on CTL activity results and a statistically significant difference was observed from zero regression values for E:T ratios 10:1 (r²= 0.8650, p<0.0001), 5:1 (r²=0.7246, p=0.0004), 2.5:1 (r²= 0.7461, p=0.0003), and 1:1 (r²=0.7692, p=0.0002). Next, we assessed whether the effect by Δ⁹-THC was dependent on the presence of CB₁ and/or CB₂. Δ⁹-THC suppressed CTL activity in a concentration-dependent manner in CTL elicited from CB₁ ^−/−CB₂ ^−/− splenocytes with significantly reduced CTL activity at concentrations of 5 μM or more as in WT (Fig. 3b). Regression analysis revealed a statistically significant difference from zero regression of E:T ratio of 10:1 in WT (r²=0.6893, p=0.0008) and showed the same trendinCB₁ ^−/−CB₂ ^−/− albeit not significant (r²=0.2905, p=0.0706). Consistent with previous observations, CTL activity of CB₁ ^−/−CB₂ ^−/− splenocytes was lower in magnitude compared to WT, although differences in genotypes were not deemed statistically significant.

Fig. 3.

Δ⁹-THC suppresses CTL activity in a concentration-dependent manner. Splenocytes from WT mice were treated with 1, 5 and 10 μM of Δ⁹-THC and/or VH (control) and co-cultured for 5 days with irradiated P815 cells. After harvest, CTL were reincubated with ⁵¹Cr labeled P815 cells at indicated ratios of 10:1 to 1:1 (a) (n=4). In a second experiment splenocytes from WTand CB₁ ^−/−CB₂ ^−/− mice were elicited and restimulated with ⁵¹Cr labeled P815 cells at a ratio of 10:1 (b) (n=4). Data shown are representative of two repeat experiments. * p≤0.05, ** p≤0.01 indicate differences as compared to VH

Δ⁹-THC-mediated suppression of CTL generation

Although MHC I mismatch is a strong stimulus for driving the expansion of differentiated CD8⁺ T cell effectors, the overall number of viable Tcells within control cultures (NA or VH treatment groups) on day 5 of co-culture with P815 allogenic target cells was modest. This demonstrated that T cell proliferation and clonal expansion in response to MHC I-restricted alloantigens in this assay system induced a relatively small number of Tcells from the spleen. To assess the effects of Δ⁹-THC on its ability to impair alloantigen-induced expansion of T cell effectors, the number of viable CD4⁺ and CD8⁺ T cells was enumerated on Day 5 of the elicitation phase of the CTL response. Viability was determined using LIVE/DEAD dye (Invitrogen), which permeates dead cells and thus low staining is observed in viable cell populations. The percentage of viable CD4⁺ (Fig. 4a and b) and CD8⁺ Tcells (Fig. 4c and d) was markedly reduced by Δ⁹-THC treatment 5 days after elicitation with concentrations as low as 1 μM in both WTand CB₁ ^−/−CB₂ ^−/− samples. No significant differences were found in the percent of viable CD8⁺ cells between CB₁ ^−/−CB₂ ^−/− and WT splenocytes (Fig. 4c and d); however, a greater number of viable CD4⁺ T cells was observed in CB₁ ^−/−CB₂ ^−/− compared to WT splenocytes (Fig. 4a and b).

Fig. 4.

Lower viability of CD4⁺ and CD8⁺ cells following Δ⁹-THC treatment. CTL were elicited from splenocytes of WTand CB₁ ^−/−CB₂ ^−/− mice in the presence or absence of VH (0.1% ethanol) or Δ⁹-THC (1, 5, and 10 μM). After 5 days in culture, CD4 and CD8 surface staining was performed and viability was assessed using LIVE/DEAD staining. Viability was determined within CD4⁺ (a, b) and CD8⁺ (c, d) cells after singlet and lymphocyte size gating (n=4). Data shown are representative of four repeat experiments. * p≤0.05, ** p≤0.01 indicate differences as compared to VH, ## p≤0.01 as compared to WT

IFNγ production by CD8⁺ T cells is often used as a surrogate for CTL activity. After gating for single live cells (Fig. 5a), CD4⁺ cells did not produce significant amounts of IFNγ as P815 cells are MHC I-mismatched to C57Bl/6 cells (Fig. 5b). However, significant IFNγ production was detected by CD8⁺ cells, which was enhanced in CTL elicited from WTcompared to CB₁ ^−/−CB₂ ^−/− (Fig. 5c, d and e). In the presence of Δ⁹-THC, enhanced IFNγ production was observed in CD8⁺ T cells derived from WT and the same trend was observed at higher Δ⁹-THC concentrations in CB₁ ^−/−CB₂ ^−/− mice, although not deemed statistically significant (Fig. 5c, d and e). Also, IFNγ production was reduced in samples from CB₁ ^−/−CB₂ ^−/− compared to WT mice as previously observed in ⁵¹Cr release assays (Fig. 5c, d and e). Analysis of CD8⁺ IFNγ-secreting cells, as a percentage of viable cells, takes into account the reduced viability as a result of Δ⁹-THC addition (see Fig. 4). However, while there was an enhancement of IFNγ-secreting cells with Δ⁹-THC treatment, there were also fewer viable cells after Δ⁹-THC treatment, thus overall the percentage of CD8⁺ cells secreting IFNγ within the total population did not change (Fig. 5f).

Fig. 5.

Increased IFNγ production in live but not total cells after Δ⁹-THC treatment. CTL of WTand CB₁ ^−/−CB₂ ^−/− were elicited as before with and without VH (0.1% ethanol) or Δ⁹-THC (1, 5, 10 μM) treatment for 5 days. Cells were restimulated for 12 h in the presence of Brefeldin A and stained for CD4⁺, CD8⁺, LIVE/DEAD and IFNγ. Gating scheme is shown (a) for populations of CD4⁺ (b) and CD8⁺ (c) cells and dot plots indicate concatenated samples (n=4). Bar graph for CD8⁺ cells are shown within live populations % (d), MFI (e), and in % of total populations (f) (n=4). Data shown are representative of three repeat experiments. ** p≤0.01 indicate differences as compared to VH, ## p≤0.01 as compared to WT

Δ⁹-THC does not affect CD8⁺ T cell proliferation induced by P815 co-culture

Prior to becoming functional CTL, proliferation occurs in T cells to expand the effector pool. Thus, we focused on proliferation as a potential endpoint leading to decreased viability and decreased CTL response. Splenocytes were labeled with CFSE dye to determine proliferation by dye dilution. No statistically significant differences in proliferation were observed between VH and Δ⁹-THC-treated groups neither in CD8⁺ nor in CD4⁺ cells from WTor CB₁ ^−/−CB₂ ^−/− mice (Fig. 6a, b, c and d). In comparison to WT, CD4⁺ cells obtained from CB₁ ^−/−CB₂ ^−/− spleens underwent much greater proliferation. CD4⁺ T cells were not directly stimulated in this model, suggesting that greater proliferation in CB₁ ^−/−CB₂ ^−/− CD4⁺ cells might either be due to higher basal proliferation in CB₁ ^−/−CB₂ ^−/− cells or a result of the cytokines produced by CD8⁺ cells. In contrast, notably more CD8⁺ cells were elicited in samples from WT compared to CB₁ ^−/−CB₂ ^−/− spleens (Fig. 6c and d).

Fig. 6.

No effect of Δ⁹-THC on proliferation. Splenocytes were isolated from WT and CB₁ ^−/−CB₂ ^−/− mice and labeled with Cell Trace CFSE according to manufacturer’s instructions. Four days after co-culture with irradiated P815 in the presence or absence of VH (0.1% ethanol) and Δ⁹-THC (1, 5, 10 μM), surface staining with CD4 and CD8 were performed and CFSE fluorescence was assessed by FACS. Shown are concatenated samples (n=4) of CFSE staining as a result of proliferation in CD4⁺ (a, b) and CD8⁺ cells (c, d). Data shown are representative of two repeat experiments. * p≤0.05 indicate differences as compared to VH, ## p≤0.01 as compared to WT

Increase of CD69 expression on CD8⁺ cells with Δ⁹-THC treatment

In an effort to determine events leading to the Δ⁹-THC-mediated reduction in cell viability and CTL activity seen at day 5, without any apparent effects on proliferation, we focused on T cell activation. Two markers of cellular activation found on both CD4⁺ and CD8⁺ T cells are the expression of CD69 and CD25. Whereas CD69 is thought to be one of the earliest markers of activation and is rapidly upregulated, CD25 expression increases gradually on activated T cells and is observed later than CD69 on activated T cells (Lopez-Cabrera et al. 1995; Brenchley et al. 2002). We focused on CD69 expression at 6 h and CD25 expression at 12 h after co-culture with irradiated P815 cells. Cells were gated on singlets, by size, and then on CD4 or CD8 (Fig. 7a). In CD4⁺ cells, CD69 expression was unaffected by co-culture with P815 cells; however, cells from CB₁ ^−/−CB₂ ^−/− mice showed higher basal expression compared to WT mice (Fig. 7b and c). In addition, no effect with Δ⁹-THC treatment was observed on CD69 expression on CD4⁺ cells (Fig. 7b and c). In contrast, the expression of CD69 on CD8⁺ was increased by co-culture with P815 cells and CD8⁺ T cells from CB₁ ^−/−CB₂ ^−/− expressed higher levels of CD69 than WT CD8⁺ T cells (Fig. 7d and e). CD69 expression was increased in a concentration-dependent manner by Δ⁹-THC in CD8⁺ T cells from both WT and CB₁ ^−/−CB₂ ^−/− spleens, which was significant at concentrations as low as 5 μM in CB₁ ^−/−CB₂ ^−/−-derived cells (Fig. 7d and e).

Fig. 7.

Δ⁹-THC increases CD69 expression on CD8⁺ cells in a concentration-dependent manner. Splenocytes from of WT and CB₁ ^−/− CB₂^−/− mice were incubated with irradiated P815 cells for 6 h to induce CD69 surface expression, in the presence or absence of VH (0.1% ethanol) or Δ⁹-THC (1, 5, and 10 μM). Cells were gated on singlets and lymphocyte populations by size and then on CD4 and CD8 (a). CD4⁺ (b, c) and CD8⁺ (d, e) cells positive as defined by the box gate for CD69 in % are shown (n=4). Data shown are representative of three repeat experiments. Difference due to stimulation by P815 is indicated by ++ p≤0.01, due to genotypes by ## p≤0.01, due to Δ⁹-THC by * p≤0.05,** p≤0.01

CD25 expression on CD4⁺ T cells was increased as a result of stimulation with P815 cells and increased levels of CD25 expression were observed in NA unstimulated samples from CB₁ ^−/−CB₂ ^−/− compared to WT mice (Fig. 8a and b). There was a trend towards a Δ⁹-THC-dependent increase in CD25 expression in both WT and CB₁ ^−/−CB₂ ^−/− CD4⁺ cell populations, but this trend was not statistically significant (Fig. 8b). In CD8⁺ cells, the expression of CD25 was much lower than in CD4⁺ cells (Fig. 8a, b, c and d). Expression of CD25 in CD8⁺ cells of both genotypes was increased by co-culture with P815 cells (Fig. 8c and d). In CD8⁺ cells from CB₁ ^−/−CB₂ ^−/− mice, a concentration-dependent decrease with Δ⁹-THC treatment was observed, while in WT Δ⁹-THC had no effect (Fig. 8c and d).

Fig. 8.

Δ⁹-THC decreases CD25 expression on CD8⁺ cells from CB₁^−/− CB₂^−/− mice in a concentration-dependent manner. Splenocytes from WT and CB₁^−/− CB₂^−/− mice were co-cultured with irradiated P815 cells for 12 h to induce CD25 levels, in the presence or absence of VH (0.1% ethanol) or Δ⁹-THC (1,5, and 10 μM). Cells were gated on singlets and lymphocyte populations by size and cells positive for CD25 in % are shown (n=4). Data shown are representative of two repeat experiments. Difference due to stimulation by P815 is indicated by ++ p≤0.01, due to genotypes by ## p≤0.01, due to Δ⁹-THC by * p≤0.05

Δ⁹-THC synergizes with Io to induce CD69 expression

In light of the rapid Δ⁹-THC-mediated increase in CD69 expression of CD8⁺ T cells and its inverse correlation with Δ⁹-THC-mediated suppression of CTL activity, we focused on intracellular signals leading to CD69 protein expression. Previously, it was determined that independent of CB₁ and CB₂, Δ⁹-THC partially mediated its effects through the TRPC1 channel, which causes an increase in intracellular Ca²⁺ levels (Rao and Kaminski 2006). Using the calcium ionophore, Io, we tested whether a rise in intracellular Ca²⁺ might contribute to the observed increase in CD69 expression. In addition, the effect of Δ⁹-THC on Io-induced CD69 expression was evaluated. Again, cells were gated as described above (Fig. 7a). In CD4⁺ cells, Δ⁹-THC increased expression of CD69 moderately in the absence of an activating stimulus when isolated from CB₁ ^−/−CB₂ ^−/−, but not WT mice (Fig. 9a and c). Δ⁹-THC increased CD69 expression in CD8⁺ splenocytes from both WT and CB₁ ^−/−CB₂ ^−/− mice in the absence of an activating stimulus (Fig. 9a and e). After stimulation with Io, CD69 expression was dramatically increased in CD4⁺ and CD8⁺ cells from both genotypes (Fig. 9b, d and f). While CD69 expression was similar between WTand CB₁ ^−/−CB₂ ^−/− CD8⁺ cells (Fig. 9b and f), a marked reduction in CD69 expression was observed in CB₁ ^−/−CB₂ ^−/− CD4⁺ cells, when compared to WT CD4⁺ cells (Fig. 9b and d). Most importantly, Δ⁹-THC synergized with Io to further increase CD69 surface expression in CD4⁺ and CD8⁺ cells derived from both genotypes (Fig. 9b, d and f).

Fig. 9.

Δ⁹-THC synergizes with Io to upregulate CD69 expression. Splenocytes were incubated in the presence of Io (0.5 μM) or VH (0.01% DMSO) in the presence or absence of VH (0.1% ethanol) or Δ⁹-THC (1, 5, and 10 μM). For NA-0 h, CD69 staining was performed after isolation of a single cell suspension from the spleen otherwise 6 h after co-culture. Cells were gated on singlet, lymphocytes and within CD4 (a–d) or CD8 (a, b, e, f) positive populations. Data shown in histograms are concatenated (n=4) and are representative of two repeat experiments. Difference due to stimulation by Io is indicated by ++ p≤0.01, due to genotypes by ## p≤0.01, due to Δ⁹-THC by ** p≤0.01

Discussion

Collectively, these studies provide evidence that Δ⁹-THC-mediated suppression of CD8⁺ T cell function occurs independent of CB₁ and CB₂. Moreover, Δ⁹-THC alters early signaling events that involve putative changes in intracellular Ca²⁺ levels, eventually resulting in lower numbers of CTL effectors. Finally, this study introduces an inherent role for CB₁ and/or CB₂ in the elicitation of CTL, because elicitation and function of CTL was reduced in CB₁ ^−/−CB₂ ^−/− mice in vitro.

In contrast to other reports describing enhanced immune responses in CB₁ ^−/−CB₂ ^−/− mice (Karsak et al. 2007; Buchweitz et al. 2008), in this in vitro model of alloantigen-induced CTL elicitation and effector function, CTL activity in cells from CB₁ ^−/−CB₂ ^−/− mice was reduced when compared to WT mice. Although this was consistent with our previous study in which we observed a reduction in the T cell-dependent humoral immune response to sheep erythrocytes in vitro in CB₁ ^−/−/CB₂ ^−/− as compared to WT mice (Springs et al. 2008), there are several studies demonstrating enhanced immune function in CB₁ ^−/−CB₂ ^−/− mice (Karsak et al. 2007; Buchweitz et al. 2008). In a previous study from our laboratory (Buchweitz et al. 2008), it was demonstrated that CB₁ ^−/−CB₂ ^−/− mice, in addition to responding with a greater magnitude to immune stimuli, have altered immune homeostasis compared to WT mice as evidenced by greater numbers of CD4⁺ T cells and higher levels of TNF-α in the bronchoalveolar lavage fluid in the absence of influenza challenge. In addition, there appeared to be mechanisms in CB₁ ^−/−CB₂ ^−/− mice that might compensate for the loss of CB₁ and/or CB₂, as evidenced by greater levels of the endocannabinoids, 2-arachidonoyl glycerol and anandamide (Karsak et al. 2007), and increased transcripts of cnr2 using primers specific for the undeleted portion of the gene (Liu et al. 2009). Although CB₁ ^−/−CB₂ ^−/− mice seem to be particularly vulnerable to proinflammatory insult, there were no drastic differences observed in an adjuvant-free airway hypersensitivity model between CB₁ ^−/−CB₂ ^−/− and WT mice (Kaplan et al. 2010). Reasons for reduced function of CTL from CB₁^−/− CB₂^−/− mice compared to WT mice, although not altogether clear, might be due to the present studies being conducted in vitro, and/or the nature in which the immune response was initiated, in this case by direct stimulation of CD8⁺ T cells with an alloantigen.

Greater proliferation was detected in CD4⁺, but lower proliferation in CD8⁺ cells, from CB₁^−/− CB₂^−/− compared to WT mice. This lower proliferation in CD8⁺ cells might contribute to the reduced CTL activity in CB₁^−/− CB₂^−/− compared to WT mice. Furthermore, higher percentages of CD4⁺CD25⁺ were observed in unstimulated splenocytes of CB₁^−/− CB₂^−/− compared to WT mice. These CD4⁺CD25⁺ cells are thought to be regulatory T cells (Treg) (Suri-Payer et al. 1998), which are involved in the suppression of the immune responses. It has been demonstrated that an increase in CD4⁺CD25⁺ cells leads to reduced CTL responses; thereby Tregs could potentially contribute to the reduced activity in vitro in CTL generated from CB₁^−/− CB₂^−/− mice (Piccirillo and Shevach 2001). This in vitro model, due to direct stimulation of CTL, lacks a proinflammatory stimulus and does not produce IL-6, which has been implicated in overriding Treg responses, thus Treg responses might be maintained during the in vitro culture (Bettelli et al. 2006). As our focus was on the Δ⁹-THC-mediated suppression of CTL responses, we did not further characterize these CD4⁺CD25⁺ cells for expression of Foxp3 or other cellular markers of the Treg lineage (Fontenot and Rudensky 2005). Additional studies are required to determine the role of Tregs in CB₁^−/−/CB₂^−/− mice.

Addition of Δ⁹-THC during the elicitation and effector response showed that only NA CD8⁺ cells, but not differentiated CTL, were sensitive to Δ⁹-THC-mediated suppression of CTL function as assessed by ⁵¹Cr release assay, suggesting Δ⁹-THC-mediated immune modulation during elicitation. Arguably, the concentrations of Δ⁹-THC at which altered CTL activity was observed in this investigation are higher than detected in the serum of recreational marijuana smokers. However, it is important to emphasize two points. The first is that in addition to Δ⁹-THC, marijuana smoke contains over 60 other structurally-related cannabinoids, a number of which are also well established immune modulators and the majority of which remain to be evaluated for immunomodulatory activity (Ashton 2001; Friedman et al. 2003). Second, and equally important, the goal of this study was to relate the decrease in viral clearance observed after Δ⁹-THC treatment of animal with influenza-infected airways (Buchweitz et al. 2007) and the possibility that this may be due to cannabinoid-mediated suppression in CTL activity. In the case of inhaled marijuana smoke, the cannabinoid concentrations attained in the lung are significantly higher than systemic serum Δ⁹-THC concentrations and may be more closely related to those used in the current study (Azorlosa et al. 1992; Huestis and Cone 2004). The timeframe of sensitivity to Δ⁹-THC might reflect a high level of T cell plasticity prior to differentiation and reduced plasticity after differentiation (Sundrud et al. 2003). Other immunomodulatory compounds such as 2,3,7,8-tetrachlorodibenzo- p-dioxin and glucocorticoids exert their greatest effect on early stages of immune cell differentiation and tend to have a lower efficacy if added after the onset of the differentiation program (Piccolella et al. 1985; Tucker et al. 1986). Specifically, cyclosporin A is very efficacious in suppressing in vitro culture of allogeneic cells only when added prior to elicitation (Hess and Tutschka 1980). Overall, this suggests a critical window of sensitivity for immunomodulatory compounds, including Δ⁹-THC, for immune suppression, specifically during the transition of NA CD8⁺ T cells to CTL during elicitation.

Using IFNγ as a surrogate for CTL activity demonstrated that CD8⁺ cells are the sole producers of IFNγ after allogeneic stimulation, suggesting that only CTL directly contribute to the ⁵¹Cr release from labeled target cells. Although it appeared that Δ⁹-THC increased IFNγ expression in CTL, the cells were gated on the live cell populations. Upon examination of the results from the total population, there was no effect of Δ⁹-THC on IFNγ, demonstrating alternative interpretations of the data. Taken together, our results suggest that while there seemed to be a population of CTL whose activity cannot be suppressed by Δ⁹-THC, which become the IFNγ-secreting cells, the remaining cells are sensitive to Δ⁹-THC and were not capable of surviving the five-day culture period. Thus, the suppression by Δ⁹-THC does not appear to reduce the amount of IFNγ each cell produces, but rather impaired the differentiation of a select pool of naïve cells from becoming effector CTL.

No effects of Δ⁹-THC treatment on the proliferation of CD4⁺ or CD8⁺ cells in WT or CB₁^−/− CB₂^−/− mice were observed. To link the functional outcome of Δ⁹-THC treatment to earlier T cell signaling, the expression of activation markers, CD25, forming the high affinity IL-2 receptor and CD69, the earliest known lymphocyte activation marker (Lopez-Cabrera et al. 1995) were assessed. While CD25 expression was increased as a result of P815 co-culture, only CD8⁺ cells from CB₁^−/− CB₂^−/− mice had reduced CD25 expression as a result of Δ⁹-THC treatment. CD69 expression was also increased as a result of co-culture with P815 cells in both WT and CB₁^−/− CB₂^−/− CD8⁺ cells and Δ⁹-THC treatment further enhanced the P815-induced rise in CD69 expression in both genotypes. The ability of Δ⁹-THC to increase CD69 expression on the surface of CD8⁺ T cells from WT and CB₁^−/− CB₂^−/− after P815 cell stimulation provides further evidence for CB₁ and/or CB₂-independent mechanism(s) of Δ⁹-THC immune modulation. Notably, expression of CD69 was inversely correlated with CTL activity. For example, the highest CD69 expression and lowest CTL activity were observed in cells treated with the highest concentration of Δ⁹-THC (10 μM) from CB₁^−/− CB₂^−/− mice.

CD69 expression is controlled by several transcription factors, including AP-1 in the proximal promoter (Castellanos et al. 1997) as well as EGR, ATF/CREB (Castellanos Mdel et al. 2002), and NFκB (Lopez-Cabrera et al. 1995). However, the pathways regulating the expression of CD69 are different depending on the cell type (Vazquez et al. 2009). It is not entirely clear what cellular signals contribute to CD69 induction in CD8⁺ T cells, but it is known that Ca²⁺ influx into T cells increases CD69 expression (Testi et al. 1994) and, at least in thymocytes, extracellular Ca²⁺ was found to be critical for CD69 induction (Rodrigues Mascarenhas et al. 2003). In these studies Δ⁹-THC synergized with Io to increase CD69 expression suggesting that Ca²⁺ plays an important role in the Δ⁹-THC-mediated increase of CD69 expression. Previously, our laboratory reported that Δ⁹-THC strongly induced intracellular Ca²⁺ in T cells, which was mediated by TRPC1 (Rao et al. 2004; Rao and Kaminski 2006) independent of the requirement for T cell activation. These studies, taken together with previous results, suggest that modulation of an activation stimulus by concurrent induction of Ca²⁺ through Δ⁹-THC-mediated opening of TRPC1 channels contributes directly to the suppression of CTL activity. In addition, there are reports indicating that Δ⁹-THC has a high binding affinity for GPR55 (Ryberg et al. 2007) and causes increases in intracellular Ca²⁺ levels via GPR55 (Lauckner et al. 2008), but there are also conflicting reports demonstrating that Δ⁹-THC possesses a low affinity for GPR55 (Oka et al. 2007). The role of GPR55 in the present study is unknown; however, due to the expression of GPR55 in the spleen (Ryberg et al. 2007) and its ability to modulate Ca²⁺ currents it is tempting to speculate that GPR55 contributes to alterations in Ca²⁺ signaling induced by Δ⁹-THC. Ca²⁺ is a ubiquitous intracellular signaling molecule and in addition to activation, Ca²⁺ is central to the induction of anergy and cell death, which are seemingly separate processes, yet involve shared mechanisms and endpoints (Macian et al. 2002; Parish et al. 2009). Thus, due to the role of Ca²⁺ in the generation of CTL, it seems plausible that Δ⁹-THC-mediated alterations in Ca²⁺-signaling are involved in the suppression of CTL responses.

Collectively, the reduced cytolytic activity, decreased number of CTL effectors, and the increase in CD69 expression, although temporally distinct, might be interrelated, and a consequence of early Δ⁹-THC-induced changes in Ca²⁺ signaling. Concordantly, immune cells appear to initially undergo greater activation (CD69), but become unresponsive to secondary stimulation; in essence, anergic. The critical role of calcium in leukocyte activation and subsequent differentiation and effector function is well established. Δ⁹-THC treatment prior to cellular activation significantly elevates intracellular Ca²⁺ levels, which in turn interferes with either entry or execution of the T cell differentiation program. The specific downstream signaling pathways affected through the Δ⁹-THC-mediated elevation of intracellular Ca²⁺ levels affecting T cell function remain to be elucidated but likely involve p38 MAPK, JNK, and ERK. Indeed, prior studies by our laboratory demonstrated decreased ERK phosphorylation by cannabinol, another cannabinoid acting independently of CB₁ and CB₂ (Faubert Kaplan and Kaminski 2003). Thus, the immediate perturbation of intracellular Ca²⁺ levels by cannabinoids, including Δ⁹-THC, seems to be the initial event triggering a cascade of events, eventually resulting in immune suppression. Signaling by cannabinoids seems to alter the regulation of Ca²⁺ that is central to the initiation of a T cell response to impair immune function.