Elevating levels of the endocannabinoid 2-arachidonoylglycerol blunts opioid reward but not analgesia

Abstract

Converging findings have established that the endocannabinoid (eCB) system serves as a possible target for the development of new treatments as a complement to opioid-based treatments. Here, we show in male and female mice that enhancing levels of the eCB, 2-arachidonoylglycerol (2-AG), through pharmacological inhibition of its catabolic enzyme, monoacylglycerol lipase (MAGL), either systemically or in the ventral tegmental area (VTA) with JZL184, leads to a substantial attenuation of the rewarding effects of opioids in mice using conditioned place preference and self-administration paradigms, without altering their analgesic properties. These effects are driven by cannabinoid receptor 1 (CB1R) within the VTA, as VTA CB1R conditional knockout counteracts JZL184's effects. Using fiber photometry with fluorescent sensors for calcium and dopamine (DA), we find that enhancing 2-AG levels diminishes opioid reward-related nucleus accumbens (NAc) activity and DA neurotransmission. Together, these findings reveal that 2-AG diminishes the rewarding properties of opioids and provides a potential adjunctive therapeutic strategy for opioid-related analgesic treatments.

Fig. 1.. JZL184 attenuates morphine preference via CB1Rs.

(A) eCB hydrolysis mechanism targeted by JZL184 and PF-385. (B) Timeline of morphine CPP and vehicle or JZL184 pretreatment. PI, phosphatidylinositol; DAG, diacylglycerol; PI, phosphatidylethanolamine; NAPE, N-arachidonoyl phosphatidylethanolamine; AA, arachidonic acid. (C) JZL184 abolished morphine CPP [two-way analysis of variance (ANOVA), significant interaction (treatment × day), F_1,26 = 19.77, P = 0.0001; post hoc: vehicle: test versus pretest, ***P < 0.001; JZL184: test versus pretest, P > 0.9999; test: vehicle versus JZL184, †††P < 0.001; vehicle, N = 7; JZL184, N = 8]. (D) Timeline of morphine CPP and vehicle or PF-3845 pretreatment. (E) PF-3845 had no effect (two-way ANOVA, main effect of day, F_1,32 = 48.49, P < 0.0001; post hoc: vehicle: test versus pretest, ***P < 0.001; PF-3845: test versus pretest, ***P = 0.001; vehicle, N = 9; PF-3845, N = 9). (F) Timeline of morphine CPP and AM251 and/or JZL184 pretreatment. (G) AM251 (3 mg/kg) before JZL184 counteracted JZL184 effects on morphine CPP {three-way ANOVA, significant interaction [AM251 (3 mg/kg): AM251 × days × JZL184, F_1,78 = 5.159, P = 0.0259; post hoc: vehicle-vehicle: test versus pretest, ***P < 0.001; vehicle-JZL184: test versus pretest, P > 0.9999; AM251-vehicle: test versus pretest, **P = 0.0053; AM251-JZL184: test versus pretest, **P = 0.0025; test: vehicle-vehicle versus vehicle-JZL184, †††P < 0.001]; three-way ANOVA, significant interaction [AM251 (1 mg/kg): AM251 × days, F_1,78 = 4.131, P = 0.0456, JZL184 × days, F_1,77 = 16.58, ***P =0.0001; post hoc: vehicle-vehicle: test versus pretest, ***P < 0.001; vehicle-JZL184: test versus pretest, P > 0.9999; AM251-vehicle: test versus pretest, *P = 0.0450; AM251-JZL184: test versus pretest, P > 0.9999; test: vehicle-vehicle versus vehicle-JZL184, †††P < 0.001; test: vehicle-vehicle versus AM251-JZL184, †††P < 0.001]; vehicle-vehicle, N = 24; vehicle-JZL184, N = 26; AM251 [3 mg/kg]–vehicle, N = 15; AM251 [3 mg/kg]–JZL184, N = 17; AM251 [1 mg/kg]–vehicle, N = 15; AM251 [1 mg/kg]–JZL184, N = 16}. Error bars ± SEM. ns, not significant.

Fig. 2.. JZL184 attenuates oxycodone preference and self-administration.

(A) Timeline of behavioral protocol for oxycodone CPP and systemic pretreatment of vehicle or JZL184. (B) Systemic JZL184 before each oxycodone conditioning session attenuated oxycodone CPP [two-way ANOVA, significant interaction (treatment × day), F_1,82 = 13.31, P = 0.0005; post hoc: vehicle: test versus pretest ***P < 0.001; JZL184: test versus pretest, no significant; test: vehicle versus JZL184, †††P < 0.001; vehicle, N = 19; JZL184, N = 24]. (C) Timeline of behavioral protocol for oxycodone self-administration and systemic pretreatment of vehicle or JZL184. (D and E) Systemic JZL184 exposure before oxycodone self-administration sessions, attenuated the intake of oxycodone [(D) two-way RM ANOVA, main effect of JZL184 treatment, F_1,21 = 21.43, ***P = 0.0001, main effect of days, F_2.962,62.20 = 8.236, P = 0.0001, interaction F_9,189 = 1.900, P = 0.0541; (E) average of total infusions per animal, t test, t₂₁ = 4.63, ***P < 0.0001; vehicle/saline, N = 12; JZL184, N = 11]. Error bars ± SEM.

Fig. 3.. JZL184 has no effect on morphine analgesia.

(A and C) Experimental timeline of acute systemic vehicle or JZL184 pretreatment before the morphine dose-response in the hot plate test (A) or the tail-flick test (C). (B and D) JZL184 pretreatment has no effect on morphine-induced analgesia during the hot plate test (B; two-way ANOVA, main effect of dose, F_3,68 = 30.65, P < 0.001; vehicle, N = 9; JZL184, N = 10) or the tail-flick test (D; two-way ANOVA, main effect of dose, F_4,60 = 22.17, P < 0.001; vehicle, N = 7; JZL184, N = 7). Error bars ± SEM.

Fig. 4.. Intra-VTA JZL184 attenuates morphine preference, and VTA-CB1R knockout counteracts JZL184.

(A) Experimental timeline for intra-VTA infusion of vehicle or JZL184 and morphine CPP. (B) Exemplar image of guide cannula placement in the VTA. (C) Intra-VTA of JZL184 attenuates morphine preference [two-way ANOVA, significant interaction (treatment × day), F_1,54 = 5.144, P = 0.0274; post hoc: vehicle: test versus pretest, **P = 0.0051, JZL184: test versus pretest, P > 0.999; test: vehicle versus JZL184, †P = 0.018; vehicle, N = 12; JZL184, N = 17]. (D) Experimental timeline for surgery, morphine CPP, and tail-flick test. (E) Representative image of AAV2-Cre-GFP expression in the VTA. (F) Representative image of RNAscope showing Cnr1 mRNA expression (magenta) and GFP-tagged cells (green) in the VTA of Cnr1 ^VTA-WT (left) and Cnr1 ^VTA-KO (right) mice (DAPI = blue). (G) Quantification of Cnr1 puncta in Cnr1^VTA-WT and Cnr1^VTA-KO mice [t test, t₇₀ = 4.508, ***P < 0.001; Cnr1^VTA-WT, N = 3 (42 cells in total); Cnr1^VTA-KO, N = 3 (25 cells in total)]. (H) Focal knockout of Cnr1 in the VTA counteracts JZL184-induced blunting of morphine CPP [three-way ANOVA, significant interaction (days × genotype), F_1,26 = 4.530, P < 0.043; post hoc: Cnr1 ^VTA-WT vehicle: test versus pretest, *P = 0.044; Cnr1 ^VTA-WT JZL184: test versus pretest, P > 0.999; Cnr1 ^VTA-KO vehicle: test versus pretest, *P = 0.020; Cnr1 ^VTA-KO JZL184: test versus pretest, *P = 0.031; test: Cnr1 ^VTA-WT vehicle versus Cnr1 ^VTA-WT JZL184, †P = 0.013; test: Cnr1 ^VTA-WT JZL184 versus Cnr1 ^VTA-KO JZL184, ††P = 0.005; Cnr1 ^VTA-WT vehicle, N = 8; Cnr1 ^VTA-WT JZL184, N = 7; Cnr1 ^VTA-KO vehicle, N = 8; Cnr1 ^VTA-KO JZL184, N = 7].

Fig. 5.. JZL184 attenuates NAc neural activity time-locked to the morphine-paired chamber.

(A) Brain schematic with an optic fiber in the NAc expressing GCaMP6s and recording apparatus. (B) Representative image of GCaMP6s expression in the NAc. aca, anterior commissure; AcbC, NAc core; AcbSh, NAc shell. (C) Experimental timeline for surgery, morphine CPP, and FP recording of GCaMP6s. (D) Schematic of the CPP box depicting a mouse approaching the saline (left) or the morphine chamber (right). Dashed line represents the position in the CPP box where the FP GCaMP6s or dlight1.2 signal was time-locked to the approach behavior. (E) Schematic of the CPP box depicting a mouse entering the saline (left) or the morphine chamber (right). Dashed line represents the entry point in the saline or morphine chamber to which the FP signal was time-locked. (F and G) Average trace of NAc activity in (F) vehicle-pretreated animals time-locked to the approach of the morphine-paired (black) or the saline-paired chamber (blue) or average trace of NAc activity in (G) JZL184-pretreated animals time-locked to the approach of the morphine-paired (red) or the saline-paired chamber (gray). Gray shading indicates window used for quantification in (H). (H) Mean NAc signal during approach (0- to 5-s window) of saline- or morphine-paired chambers in vehicle or JZL184-pretreated animals. In vehicle animals, there was a significant difference in signal during approach of morphine- compared to saline-paired chambers (N = 6 mice, 96 morphine chamber approaches, 95 saline chamber approaches; linear mixed effect model ANOVA marginal test, significant effect of group F_1,189 = 12.1; ***P = 0.0006). There was no difference in signal in the JZL184-pretreated animals (N = 9 mice, 166 morphine chamber approaches, 163 saline chamber approaches; linear mixed effect model—ANOVA marginal test, no effect of group F _1,327 = 0.37; P = 0.54). Error bars ± SEM (number of trials). LED, light-emitting diode.

Fig. 6.. JZL184 pretreatment attenuates NAc DA dynamics time-locked to the morphine-paired chamber.

(A) Brain schematic with an optic fiber implanted into the NAc expressing dLight1.2 and recording apparatus. (B) Representative image of dLight1.2 expression in the NAc. aca, anterior commissure; AcbC, NAc core; AcbSh, NAc shell. (C) Experimental timeline for surgery, morphine CPP, and FP recording of DA (dLight1.2) on CPP test day 6. (D) Average trace of DA signal in vehicle-pretreated animals time-locked to entry into either the morphine-paired chamber (black) or the saline-paired chamber (blue). Gray shading indicates window used for quantification in (F). N = 16 mice, 286 morphine chamber entries, 300 saline chamber entries. (E) Average trace of DA signal in JZL184-pretreated animals time-locked to entry into either the morphine-paired chamber (red) or the saline-paired chamber (gray). Gray shading indicates window used for quantification in (F). N = 12 mice, 217 morphine chamber entries, 241 saline chamber entries. (F) Mean DA signal during entry (0- to 5-s window) of either saline- or morphine-paired chambers in either vehicle or JZL184-pretreated animals. In vehicle animals, there was a significant difference in signal during entry into morphine- compared to saline-paired chambers (N = 16 mice, 286 morphine chamber entries, 300 saline chamber entries; linear mixed effect model—ANOVA marginal test, significant effect of group F_1,584 = 8.67; **P = 0.0034. There was no significant difference in signal in the JZL184-pretreated animals (N = 12 mice, 217 morphine chamber entries, 241 saline chamber entries; linear mixed effect model—ANOVA marginal test, no effect of group F_1,456 = 2.8; P = 0.094). Error bars ± SEM (number of trials).