A novel allosteric modulator of the cannabinoid CB1 receptor ameliorates hyperdopaminergia endophenotypes in rodent models

The endocannabinoid system (eCBs) encompasses the endocannabinoids, their synthetic and degradative enzymes, and cannabinoid (CB) receptors. The eCBs mediates inhibition of neurotransmitter release and acts as a major homeostatic system. Many aspects of the eCBs are altered in a number of psychiatric disorders including schizophrenia, which is characterized by dysregulation of dopaminergic signaling. The GluN1-Knockdown (GluN1KD) and Dopamine Transporter Knockout (DATKO) mice are models of hyperdopaminergia, which display abnormal psychosis-related behaviors, including hyperlocomotion and changes in pre-pulse inhibition (PPI). Here, we investigate the ability of a novel CB1 receptor (CB1R) allosteric modulator, ABM300, to ameliorate these dysregulated behaviors. ABM300 was characterized in vitro (receptor binding, β-arrestin2 recruitment, ERK1/2 phosphorylation, cAMP inhibition) and in vivo (anxiety-like behaviors, cannabimimetic effects, novel environment exploratory behavior, pre-pulse inhibition, conditioned avoidance response) to assess the effects of the compound in dysregulated behaviors within the transgenic models. In vitro, ABM300 increased CB1R agonist binding but acted as an inhibitor of CB1R agonist induced signaling, including β-arrestin2 translocation, ERK phosphorylation and cAMP inhibition. In vivo, ABM300 did not elicit anxiogenic-like or cannabimimetic effects, but it decreased novelty-induced hyperactivity, exaggerated stereotypy, and vertical exploration in both transgenic models of hyperdopaminergia, as well as normalizing PPI in DATKO mice. The data demonstrate for the first time that a CB1R allosteric modulator ameliorates the behavioral deficits in two models of increased dopamine, warranting further investigation as a potential therapeutic target in psychiatry.


INTRODUCTION
Dysregulation of dopaminergic and glutamatergic signaling are thought to underpin the development of psychosis and schizophrenia [1]. Pharmacological treatment of schizophrenia and psychosis includes the use of antipsychotics, which act as orthosteric receptor antagonists/partial agonists of various GPCR targets, including dopamine receptor D 2 and serotonin receptor 5HT 1A . However, antipsychotics are associated with extrapyramidal side effects, sedation, metabolic syndrome, and weight gain [2,3].
The endocannabinoids, anandamide (AEA) and 2arachidonoylglycerol (2-AG), are orthosteric agonists of the cannabinoid CB 1 receptor (CB 1 R). CB 1 R are expressed presynaptically on various neuronal types, including GABAergic, glutamatergic and serotonergic neurons, where they mediate an inhibition of transmitter release. While not directly expressed on dopaminergic neurons, the endocannabinoid system acts as a crucial filter that integrates both inhibitory and excitatory signaling that modulates dopamine neuron signaling [4]. Furthermore, studies have shown that the endocannabinoid system is a negative modulator of both D 1 and D 2 receptor-mediated behaviors, implicating them in basal ganglia disorders [5]. As such, in combination with the complex dysregulation and circuit-based mechanisms for brain regiondependent alterations in dopaminergic signaling in psychiatry, this suggests that the CB 1 R may be a more attractive, alternative therapeutic target to the classical D 2 receptor antagonism approaches of antipsychotics [4].
Furthermore, there is strong evidence from humans that both endocannabinoid levels and CB 1 R are dysregulated in schizophrenia [6][7][8][9]. Serum and CSF levels of AEA are higher in patients with schizophrenia at all stages of the illness and are normalized after treatment with antipsychotics. CB 1 R expression and binding is higher in post-mortem brain tissues of patients with schizophrenia [6,9]. While the endocannabinoid system seems to be an important potential therapeutic target in psychiatry, targeting CB 1 R at the orthosteric site has not yielded beneficial clinical outcomes. The CB 1 R orthosteric inverse agonist, rimonabant, was effective treating obesity and metabolic syndrome, but caused suicidal ideation and was withdrawn from the market [10][11][12]. Here, we propose to investigate a novel pharmacological approach of targeting the CB 1 R.
In 2005, we discovered the CB 1 R allosteric site and the original, prototype allosteric modulator, Org27569. This compound has served as a tool compound to characterize the allosteric site but is not a drug candidate. Org275, and related compounds, display an atypical, complex allosteric profile at CB 1 R [13,14]. Org275 increases the B max of [ 3 H] CB 1 R agonist binding but functionally acts as an inhibitor of CB 1 R agonist-mediated signaling [13,15]. Importantly, in October 2019, Shao et al. [16] elucidated the ternary crystal structure of CB 1 R in complex with agonist and Org275. The structure shows that Org275 binds to a cholesterol-binding site on the CB 1 R, suggesting that the compound works by partitioning into the bilayer and competing with endogenous cholesterol for this surface. Previous studies have demonstrated that cholesterol may act as an endogenous modulator of CB 1 R [17]. There is growing evidence that, instead of targeting the orthosteric site of CB 1 R, the allosteric site may have key advantages [15,18,19]. By modulating the effects of the endogenous ligand, normal physiological tone (spatial and temporal effects of ligand binding to the receptor) are maintained, as opposed to the non-physiological binding and distribution seen with exogenous direct ligands such as orthosteric agonists or antagonists.
Since discovering the CB 1 R allosteric site in 2005, and identification of Org275 as the first CB 1 R-negative allosteric modulator [13], we, and others, have worked to develop both CB 1 R-negative and -positive allosteric modulators. The positive allosteric modulators developed by us, and others, have shown efficacy in the treatment of neuropathic pain [20] and other therapeutic indications [21]. As Org275, and related compounds, have insufficient metabolic stability, in order to further investigate the potential of this unique class of CB 1 R allosteric modulator, we embarked on a chemistry campaign, with the goal of generating new molecules with improved drug-like characteristics that are more suitable for in vivo testing and clinical development. Our working hypothesis is that the unique pharmacological profile of CB 1 R allosterics provides a distinctive pharmacological approach for modulation of the endocannabinoid system in complex disorders, and offers an alternative to CB 1 R orthosteric antagonists [22]. The in vivo outcomes of this complex mechanism are yet to be elucidated, particularly in models in which the endocannabinoid system is dysregulated.

METHODS AND MATERIALS
Animal ethics Animal housing and experimentation were carried out in accordance with the Canadian Council in Animal Care (CCAC) guidelines for the care and use of animals and following protocols approved by the Faculty of Medicine and Pharmacy Animal Care Committee at the University of Toronto and the University of Guelph Animal Care Committee, respectively.
Compound synthesis See Supplementary Information for details.

Pharmacokinetic analyses
Microsomal stability assays were conducted by Cyprotex Ltd (Macclesfield, UK). The in vitro metabolic stability of ABM300 was measured in the presence of human or rat liver microsomes by determination of the rate of compound disappearance. Single dose in vivo PK studies were conducted by Sai Life Ltd (Pune, India) to investigate the plasma pharmacokinetics and brain distribution of ABM300 in male C57Bl/6 mice following a single intraperitoneal (i.p.) administration of a 10 mg/kg dose.
Predictive model of ABM300 bound to CB 1 R The crystal structure of the CB 1 R-CP55940-Org275 complex (PDB code 6kqi) [16] was loaded into ICM (Molsoft, San Diego, CA), hydrogens were added, and rotameric states of hydroxy groups, histidine, asparagine, and glutamine side-chains optimized. ICM's ligand editor was used to strip Org275 to the indole core scaffold shared with ABM300, and to incrementally grow the scaffold into ABM300, with a Monte Carlo-based energy minimization in the internal coordinate space at each step [32].
Cyclic AMP (cAMP) assays A DiscoverX HitHunter® cAMP Assay for Small Molecules (DiscoverX, Fremont, CA) was used to quantify cAMP levels as per the manufacturer's instructions. hCB 1 R CHO cells were seeded at a density of 40,000 cells/well (white 96-well), and after 24 h were serum-starved in the presence of 0.1% Bovine Serum Albumin for 60 min prior to pretreatment for 30 min with increasing concentrations of ABM300 (10 −10 -10 −5 M) and a final 30 min incubation with 40 nM CP55,940 and 5 μM forskolin. The cAMP BRET experiments were performed as previously described [33,34]. Briefly, hCB 1 R HEK293 cells were transfected with 5 µg/10 cm dish of the CAMYEL biosensor using polyethylenimine. After 24 h, the cells were plated into white 96-well plates (PerkinElmer) at a density of 60,000 cells/well.
Quantification and statistical analysis For in vitro assays, results were analyzed by non-linear regression analysis of sigmoidal dose-response curves. For in vivo assays, statistical parameters, the definition of measures and statistical significance are reported in the figures and the figure legends. Data are represented as mean ± SEM. Studies and data analysis were not blinded. Differences in means were considered statistically significant at p < 0.05. All data analyses were performed using GraphPad Prism 6.0 or 8.0 software (San Diego, CA) and/or IBM SPSS 23.0 Software (Armonk, NY).
Docking model of ABM300 bound to CB 1 R A model of ABM300 bound to CB 1 R was derived from the CB 1 -CP55940-Org275 ternary complex [16] (Fig. 1). ABM300 recapitulates hydrophobic interactions observed with Org275 in the crystal structure. Owing to its increased rigidity, ABM300 needs to shift by 2 Å towards the wall formed by I141 to accommodate the phenyl ring that abuts I245. The docked conformation optimally occupies the pocket, the oxadiazole ring is stacked against the indole of W241, and the secondary amine bridging the phenyl and oxadiazole rings is engaged in a hydrogen-bond with C238.
In vitro pharmacology: ABM300 increases agonist binding but inhibits CB 1 R orthosteric agonist signaling In line with our previous studies using a related compound (Org275) [13], we find that ABM300 causes a significant and concentration-dependent increase in the specific binding of [ 3 H] CP55,940 to hCB 1 R CHO (Fig. 2a) with an E max value of 328 ± 47% and a EC 50 value of 132 nM (pEC 50 6.90 ± 0.09) and an α value of 4.33 ± 0.79 (logα 0.622 ± 0.08) (Supplementary Table S2).
We further characterized the real-time kinetic effect of ABM300 using a cAMP BRET sensor assay in HEK293 cells. Similar to Fig. 1 Molecular structure and docking model of ABM300 bound to CB 1 R. a Overall structure of CB 1 R (PDB code: 6kqi) with bound Org275 (orange). The pregnenolone allosteric-binding site is highlighted in cyan [60], b Molecular structures of ABM300 and Org275. c Overall model showing ABM300 and surrounding side-chains. d Space filling representation showing that ABM300 occupies optimally the negative allosteric modulator-binding pocket. e ABM300 superimposed with the crystal structure of Org275 (orange) bound to CB 1 R (PDB code: 6kqi). The molecular surface of CB 1 R is color-coded based on binding properties. Green: hydrophobic. Red: hydrogen-bond acceptor. Blue: hydrogenbond donor.
previous observations with Org275 [38], ABM300 produced a complex, concentration and time-dependent modulation of agonist-mediated regulation of cAMP levels (Fig. 2g, h). Levels of cAMP were measured over time with a high concentration of CP55,940 (1 μM) in the presence of varying concentrations of ABM300. Consistent with Org275 observations, ABM300 did not affect the initial inhibition of cAMP by CP55,940 but, following a concentration-dependent "lag" in drug onset, inhibition of the agonist effect became apparent. At high concentrations, the ABM300 inhibitory effect overcame the CP55,940 effect and further enhanced cAMP levels above those produced by forskolin alone-reflecting inverse agonism.  (Supplementary Fig. S2).
ABM300 has no effect in the cannabinoid-induced tetrad alone and does not display anxiogenic-like effects We confirmed that ABM300 (10 mg/kg) did not have agonist activity via the cannabinoid-induced tetrad, compared to THC (10 mg/kg). In all four tetrad measures (Fig. 3), THC, but not ABM300, produced effects (p < 0.0001 for all outputs). Previous reports demonstrated adverse effects observed with CB 1 R orthosteric agonists/inverse agonists [39,40]. Possible anxiogenic effects of ABM300 (10 mg/kg) were investigated using the EPM and compared to rimonabant (10 mg/kg) (Supplementary Fig. S3). ABM300 did not affect time spent in the open arms (p = 0.3335), Fig. 2 ABM300 increases agonist binding but inhibits CB 1 R orthosteric agonist signaling through arrestin recruitment, ERK phosphorylation and cAMP signaling. a ABM300 increases [ 3 H]CP55,940 binding to hCB 1 R CHO cell membranes. b ABM300 concentration-dependently decreases CP55,940 (10 nM)-mediated arrestin recruitment, with the PathHunter® β-arrestin assay. c ABM300 concentration-dependently decreases ERK phosphorylation at the EC 80 concentration of CP55,940 (40 nM), using the AlphaScreen® SureFire® ERK1/2 phosphorylation kit in hCB 1 R CHO cells. d ABM300 has no effect alone, but decreases the E max for CP55,940-stimulated ERK phosphorylation in a concentration-dependent manner in hCB 1 R CHO cells. e ABM300 decreases the E max of AEA-stimulated ERK1/2 phosphorylation in a concentration-dependent manner in hCB 1 R CHO cells. f ABM300 concentration-dependently inhibits CP55,940-(EC 80 of 40 nM) mediated inhibition of forskolin-stimulated cAMP signaling in hCB 1 R CHO cells. Data shown as mean ± SEM from 3-5 independent experiments conducted in triplicate. g BRET CAMYEL real-time cAMP signaling data in hCB 1 R HEK cells, showing ABM300 concentrationdependently inhibiting the reduction in cAMP level induced by 5 µM forskolin and 1 µM CP55,940; the time-dependent activity of ABM300 is particularly apparent at moderate concentrations (1 µM and 100 nM), in which the onset of the ABM300 effect is delayed (representative experiment). h Area-under-the-curve analysis of g, showing that ABM300 concentration-dependently inhibits CP55,940-mediated cAMP reductions in HEK cells. At high concentrations of ABM300, cAMP levels are increased above forskolin alone (100%).  ABM300 decreases novelty-induced hyperactivity, exaggerated stereotypy, and vertical exploration in GluN1KD mice GluN1KD mice display hyperactivity, increased stereotypy and vertical exploration patterns, along with impairment in sensorimotor gating [26][27][28][29]. GluN1KD mice do not display a difference in Cnr1 mRNA expression in key brain regions mediating these behaviors ( Supplementary Fig. S4). ABM300 decreased the number of dysregulated behaviors in the GluN1KD model of hyperdopaminergia (Fig. 4). Hyperactivity (Fig. 4a, b) was affected by genotype (F [1,94] = 100.7, p < 0.0001), and GluN1KD mice responded to ABM300 (p < 0.0001) when compared to vehicle. Significant effects of genotype were observed for stereotypy and vertical exploration (Fig. 4c, d) (Fig. 4c), as well as phenotypic increased rearing behavior (Fig. 4d) Fig. S5).
ABM300 decreases novelty-induced hyperactivity, exaggerated stereotypy, vertical exploration, and normalizes pre-pulse inhibition (PPI) in DATKO mice The DATKO mouse, another model of hyperdopaminergia, displays hyperactivity, increased stereotypy and vertical exploration, with an impairment in sensorimotor gating [24,31]. Similar to the GluN1KD model, DATKO mice showed no change in Cnr1 mRNA expression ( Supplementary Fig. S6). ABM300 normalized dysregulated hyperactivity, stereotypic movements, vertical exploration and sensorimotor gating in the DATKO mice (Fig. 5), a pattern of findings similar to what was seen in the GluN1KD model (Fig. 4). For the hyperactivity measure (Fig. 5a) (Fig. 5b, c). Furthermore, the actions of ABM300 extended to sensorimotor gating deficits present in DATKO (Fig. 5d), with a rescue of the PPI deficit at the 16 dB pre-pulse interval (p = 0.018). Neither genotype, nor treatment with ABM300,  ABM300 has no effect on the conditioned avoidance response (CAR) The CAR task has been used as a test to infer antipsychotic efficacy via the selective suppression of the avoidance response [41,42]. Administration of olanzapine (1 mg/kg), but not ABM300 (10 mg/kg), attenuated avoidance behavior in CAR testing ( Supplementary  Fig. S7). A main effect of drug was observed (p = 0.007). Furthermore, olanzapine, but not ABM300, enhanced escape responding during CAR ( Supplementary Fig. S7B), with a main effect of drug treatment (p = 0.011) and an interaction between drug treatment and testing day (p = 0.034). Lastly, neither ABM300 nor olanzapine affected escape failures (Supplementary Fig. S7C). ABM300 did not induce catalepsy, or changes in body temperature ( Supplementary Fig. S7D,E).

DISCUSSION
There is growing interest in the possible therapeutic potential of CB 1 R allosteric molecules [21]. Here, we show for the first time that a novel CB 1 R allosteric modulator ameliorates select disrupted behaviors in two distinct models of hyperdopaminergia. The rescue of these phenotypes by ABM300 occurred without adverse anxiogenic-like or cannabimimetic effects traditionally observed with orthosteric CB 1 R inverse agonists. Thus, these findings represent the first study demonstrating the potential for the use of a CB 1 R allosteric modulator as a therapeutic strategy for the treatment of hyperdopaminergic states, such as psychosis and mania. Our molecular docking data indicate that ABM300 binds to the recently elucidated binding site for the original allosteric modulator Org275 [16]. The Org275-binding site on CB 1 R overlaps with a cholesterol-binding site, which is an extrahelical site within the inner leaflet of the membrane. The model is broadly in agreement with literature demonstrating that Org275 apparently stabilizes a high affinity, agonist bound CB 1 R [43], which may impede activation of selected downstream signaling pathways. The structure proposes a mechanism by which Org275, binding to the cholesterol site, captures an intermediate conformation that binds the orthosteric agonist and also inhibits G protein coupling. The model accommodates the complex effects of Org275 on CB 1 orthosteric ligand binding, including the increase in B max of agonists. ABM300 displays a similar in vitro pharmacological profile to Org275 [13,15], whereby it increases CB 1 R agonist binding with an α value > 1 (4.33 ± 0.79) [44], but acts as a Fig. 5 The effectiveness of ABM300 is recapitulated in a second, distinct, mouse model of hyperdopaminergia, the DATKO model, with additional restoration of sensorimotor deficits. ABM300 (10 mg/kg) decreases novelty-induced hyperactivity (total distance traveled; cm) (a), aberrant stereotypic movements (b), and mania-like vertical exploration (c) in the open-field test. ABM300 ameliorates sensorimotor gating deficits (d), rescuing the PPI deficit at 16 dB pre-pulse. All tests balanced for sex, drugs administered 30 min before test via i.p. Data shown as mean ± SEM, *p ≤ 0.05 compared to vehicle (within genotype), *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, two-way ANOVA, multiple comparisons, post-hoc Sidak's test. functional inhibitor of CB 1 R agonist-mediated β-arrestin recruitment (IC 50 , 50 nM), ERK phosphorylation (IC 50 , 47 nM), and cAMP inhibition (IC 50 , 380 nM).
As previously described [38], the unique time-dependent mechanism of action of ABM300 (Fig. 2g) posits that at moderate concentrations (high nM/low µM), early agonist signaling may remain unaffected; with inhibition initiating after a lag. These observations indicate that the molecular mechanism of action of ABM300 is related to previously characterized CB 1 R allosteric modulators such as Org275 and PSNCBAM-1 [38]. This may introduce potential for a unique kinetic profile of modulation of endocannabinoid signaling. The consequences of such effects at a network level and in a disease state are highly complex but may underlie the beneficial effects observed with ABM300 in the genetic models of hyperdopaminergia. Furthermore, expression levels, affinity, and pre-coupling of CB 1 Rs can significantly differ in various neuronal cell types; CB 1 R on GABAergic interneurons have significantly higher agonist affinity than those found on glutamatergic terminals, but the coupling efficacy of glutamatergic CB 1 R is significantly higher [45,46]. This expression and affinity profile leads to the complex nature in the function of the endocannabinoid system, which explains the diverse effects of certain cannabinoid drugs, and the opposing effects in different illnesses [46]. We hypothesize that, in the genetic murine models of hyperdopaminergia, ABM300 acts as a modulator of endogenous CB 1 R signaling in vivo and, potentially, selectively modulates the endocannabinoid system in specific neurotransmitter system pathways to ameliorate the dysregulated behaviors observed in these mice.
Studies have demonstrated that administration of compounds that increase endocannabinoid levels modulate a number of schizophrenia-like responses in mice [47]. Studies in cultured cells have shown that the related compound, Org275, causes migration of CB 1 R to the soma, while cholesterol, which binds to the same site as ABM300 and acts as a positive modulator, allows for the enrichment of CB 1 R at the axon [48]. Thus, in addition to complex effects on endocannabinoid affinity and signaling, there is the potential for the modulation of topological CB 1 R membrane localization by CB 1 R allosterics. The consequences of this in a complex neuronal network, and pathological states, are yet to be fully elucidated.
To assess psychosis-like behaviors, and therapeutic efficacy of ABM300, we focused on two main behavioral categories: exploratory behavior (hyperactivity) and sensorimotor gating (disruption in PPI). The literature related to psychosis-like behaviors in genetically modified mouse models has classically focused on the same two categories [49]. Although these behaviors do not directly translate to human symptoms present in the disease, they do mimic the neurotransmitter changes that are involved in psychotic symptoms. It has been accepted that subcortical hyperdopaminergia is implicated in psychosis, confirmed further by the neuropharmacological action of antipsychotic drugs currently available; all licensed pharmacological treatments of psychosis (antipsychotics) require interactions with the dopamine D 2 receptor [50]. Therefore, by assessing locomotor behavior within this study, we can infer that, driven by subcortical dopamine levels, an increase in dopamine leads to enhanced motor activity (either horizonal, rearing, and/or stereotypy) [49]. Meanwhile, disruption in PPI allows for a more straightforward phenotypic analysis of psychosis, as it has been reported in a number of psychiatric diseases, particularly schizophrenia and psychosis [51].
Here, we show that ABM300 restores dysregulated dopaminemediated exploratory activity in both genetic models: decreasing exaggerated hyperactivity, stereotypy and rearing. Furthermore, ABM300 rescues PPI deficits in the DATKO model. Since psychosis symptomology is never present alone in a disease state such as schizophrenia, it would be intriguing to investigate the effects of ABM300 in other symptomatic domains, such as cognition. CB 1 R antagonism and loss of function may enhance some forms of learning and memory, and it is possible that CB 1 R-negative allosteric modulators may be pro-cognitive in preclinical schizophrenic models [52,53].
We saw no effect of the compound in CAR in rat even though olanzapine produced significant suppression of CAR as has been shown in previous studies [54]. CAR is an extensively validated preclinical test used to predict therapeutic efficacy of antipsychotics that directly target the dopaminergic and serotoninergic receptor systems [42,55]. This experiment further supports our assertion that ABM300's antipsychotic effects are not mediated via dopamine D2 receptors, as considerable occupancy of striatal D2 receptors is required to see suppression of CAR (65-80% for typical antipsychotics,~50% for atypical antipsychotics like clozapine) [42].
It is important to note that we do not have direct evidence for the involvement of the CB 1 R in the in vivo effect of ABM300 in genetic models. Directly implicating CB 1 R is challenging. One approach would be to create a double knockout of CB 1 R-/-and DAT-/-or GluN1-/-. However, given the crucial role of the CB 1 R in dopaminergic and glutamatergic signaling, there is a strong likelihood that the double knockouts could have a novel phenotype (such as seizures) or that previous phenotypes would be exacerbated [56,57]. Furthermore, genetic manipulation of the CB 1 R has been shown to have effects on dopamine signaling [58,59]. Another approach would be to investigate whether a CB 1 R orthosteric antagonist would block the effects of ABM300, thereby implicating CB 1 R. However, there is also doubt as to whether an orthosteric CB 1 R antagonist would block the effects of a negative allosteric inhibitor (antagonist blocking inhibitor); it is conceivable that a synergistic effect might be observed. Taken together, these limitations highlight the complexity of directly implicating CB 1 R in the mechanism of action of ABM300 in vivo in these genetic models. We plan to make this the subject of future investigations involving a variety of in vivo and ex vivo approaches. The pharmacological profiling and assessment of the potential for off-target interactions of ABM300 in binding screens suggest that the compound is CB 1 Rselective. Taken together, the potent inhibitory effect of the compound on CB 1 R signaling at nM concentrations in vitro and the favorable PK and brain penetration data, there is substantial support for the hypothesis that the CB 1 R mediates the effects of ABM300 in vivo in the genetic mouse models of hyperdopaminergia. It is also notable that the present study only included acute administration of one dose of ABM300. Future experiments will focus on various dose regimens, chronic dosing, development of tolerance and the effects of ABM300 in combination with an antipsychotic.
The data presented here offer the first evidence that acute administration of a CB 1 R allosteric modulator, with a unique pharmacological profile, effectively ameliorates certain behavioral deficits in two distinct models of increased dopamine. These data highlight that the allosteric-binding pocket on the CB 1 R warrants further investigation as a potentially important therapeutic target in psychiatry. would like to gratefully acknowledge Wendy Horsfall for mouse colony maintenance. The work was funded by grants to RAR from CIHR (PPP-125784, PP2-139101), CIHR funding to AJR (MOP119298) and CIHR funding to AS (PJT-15619).