Here we found that, overall, CBD did not reverse schizophrenia-relevant behaviors in Nrg1 mutant mice. CBD increased locomotion at baseline and increased social behaviors under THC in all mice without affecting SI at baseline. Acute THC increased anxiety-like behavior and decreased exploration and fine motor movements in the OF, and increased social behaviors. CBD and THC combined impaired startle habituation specifically in Nrg1 mutants. CBD affected some glutamatergic and GABAergic markers in Nrg1 mice, where CBD increased GluA1 and decreased GAD67 levels in Nrg1 mutant mice but not in WT controls.

Cannabidiol did not reduce the expression of schizophrenia-relevant behaviors in Nrg1 TM HET mice. Nrg1 mice showed schizophrenia-relevant behaviors including hyperlocomotion, some reduced social behaviors, and impaired fear-associated memory recall (reported previously in Karl et al., 2007; Duffy et al., 2010), but this was unaffected by chronic adolescent CBD treatment. Contrary to our findings, previous work shows chronic adolescent CBD prevented poly I:C-induced cognitive deficits and social withdrawal in rats and mice (Osborne et al., 2017, 2019a; da Silva et al., 2020). Differences between these studies and ours include the model system used, i.e., gene mutation versus neonatal poly I:C administration, which may impact the effectiveness of CBD as a treatment candidate. For example, CBD is anti-inflammatory (Burstein, 2015) and in poly I:C rats, there can be significant and sustained neuroinflammation in adult offspring (Ding et al., 2019), whereas neuroinflammation is less pronounced in Nrg1 TM HET mice, with only small changes to serum cytokine levels are evident in adult Nrg1 male mutants (Desbonnet et al., 2012, 2017). Thus, CBD may be effective in model systems and individuals with schizophrenia with high levels of neuroinflammation (Fillman et al., 2013, 2016; North et al., 2021), but may be less effective when schizophrenia-induced behaviors are less dependent on neuroinflammation.

The current study did not find social deficits present in adolescence in male Nrg1 mutant animals in baseline testing. While this is the case, social deficits were present some weeks later under acute THC treatment, suggesting deficits of social behavior may develop later in this model, after the adolescent period. This is supported by a previous treatment study that did not find social deficits in male adolescent Nrg1 mutants (Long et al., 2013).

Cannabinoids such as THC and CBD have differential effects on Nrg1 mutants during adolescence and adulthood (Arnold et al., 2007; Boucher et al., 2007; Long et al., 2013), and while chronic CBD has been assessed previously in adult male mutants (Long et al., 2012), it is not known whether chronic CBD in adolescence could be different. Due to its key role in neurodevelopment, adolescence may provide a window for utilizing CBD’s protective effects as a preventative. Adolescent CBD induced a mild hyperlocomotive phenotype in the first OF test in all animals, regardless of genotype. This is the first investigation of chronic adolescent CBD on locomotor activity. Interestingly, our findings contrast with data from chronic CBD in adult animals, where CBD did not induce hyperlocomotion itself (Moreira and Guimarães, 2005; Long et al., 2010b,2012; Gururajan et al., 2012) or even reduced locomotion in adult C57BL/6 mice (Schleicher et al., 2019). It is possible that this effect of CBD is only relevant to adolescent CBD exposure as we did not detect increased locomotor activity in CBD-treated mice under acute THC which was conducted in later adolescence, suggesting pro-locomotor effects of CBD may be highly age-dependent.

CBD had different effects on social behavior depending on treatment duration: CBD increased social behaviors in both genotypes but only when given longer-term (i.e., 6 weeks versus 3 weeks). This is the first time CBD has been found to increase social behaviors in control animals. Previous work found similar effects only in rodent models of schizophrenia and Dravet syndrome (Long et al., 2012; Osborne et al., 2017; Patra et al., 2020), and one study actually reported adolescent CBD decreasing social behaviors in control female rats (Osborne et al., 2019a). It is possible that CBD treatment takes longer than three weeks to improve social behaviors in control animals, and this may be due to sensitization of the serotonergic effects of CBD. There is evidence for sensitization of the serotonergic system with chronic CBD, as chronic, but not acute CBD increases serotonin release in the ventromedial PFC (Linge et al., 2016). Importantly, prosocial behavior is associated with higher serotonin levels (Dölen et al., 2013), and brain regions including the PFC mediate social behaviors (Ko, 2017). CBD is a 5-HT1A agonist and prosocial effects of CBD can be mediated by 5-HT1A receptors (Hartmann et al., 2019). Thus, it is possible that prosocial effects of chronic CBD may be due to a sensitized serotonergic response (e.g., increased serotonin release, potential upregulation of 5-HT1A receptor expression) which only occurs after extended CBD administration. As this was the first study to investigate how > 3 weeks of CBD treatment affects social behavior, future studies should consider how extended CBD treatment durations can affect prosocial behaviors.

Our study is the first to determine that adolescent CBD does not affect sensorimotor gating in Nrg1 TM HET mice. Effects of CBD in adult animals appear dependent on treatment duration, as acute CBD robustly reverses PPI deficits of mice after acute MK-801 or amphetamine challenge (Long et al., 2006; Pedrazzi et al., 2015, 2021), but chronic CBD is less effective and only reverses PPI deficits at one prepulse intensity in a chronic MK-801 mouse model (Gomes et al., 2015). Interestingly, PPI was elevated in Nrg1 mutant mice in the first PPI test (i.e., baseline), but not the second PPI test (i.e., during the THC battery). This is opposite to previous findings of reduced PPI in Nrg1 TM HET male mice (Stefansson et al., 2002; Boucher et al., 2007, 2011; Desbonnet et al., 2012; Long et al., 2013) but some studies have reported elevated PPI in adult Nrg1 TM HET male (Karl et al., 2011) and female mice (Long et al., 2010a) in the past. This phenotype variability is in line with earlier (van den Buuse et al., 2009; Karl et al., 2011) and late adolescence (Long et al., 2013), and PPI deficits can disappear with repeated testing (Boucher et al., 2011).

Long-term oral CBD treatment reduced freezing in the cue test of all females regardless of genotype. While it is well-established that acute systemic CBD can impair fear memory consolidation (Stern et al., 2018; Shallcross et al., 2019; Han et al., 2022), including in female mice (Montoya et al., 2020), effects of chronic CBD on fear memory have had limited investigation and chronic CBD does not appear to affect fear memory acquisition (Cheng et al., 2014, 2014c). Considering CBD-induced differences in freezing were very limited in this study, future research should consider evaluating the effects of long-term CBD on fear learning in more detail.

We replicated sedative and anxiogenic-like effects of THC in the OF, and increased sensitivity to acute THC in Nrg1 TM HET mice. In the OF, acute THC increased anxiety-like behavior and decreased exploration across genotypes and treatment groups, similar to previous work (Schramm-Sapyta et al., 2007; Long et al., 2010b; Schreiber et al., 2019). Importantly, we extend prior research to demonstrate that anxiogenic and exploration-inhibiting effects of THC can be found at lower doses than previously reported i.e., 3 mg/kg (Long et al., 2010b). Interestingly, THC increased social behavior overall, where higher doses, e.g., 5 mg/kg have been shown to decrease social behavior (Arnold et al., 2007), and this effect was more pronounced in Nrg1 mutants, similar to our previous findings of increased THC susceptibility of Nrg1 mutant mice (Boucher et al., 2007). While acute THC increasing social behaviors does not reflect clinical research (Haney et al., 1999), adolescent Nrg1 mutants have been shown to be protected against a reduction in social behaviors caused by chronic 10 mg/kg THC where WT mice were not (Long et al., 2013). It is possible that a lower acute dose of THC may increase social behaviors selectively in Nrg1 mutants. Acute THC increasing social behaviors is however a novel finding, and requires replication and further investigation to understand the mechanisms driving this finding (e.g., if it is related to anxiety-like behavior), as one recent study has found a low dose (3.2 and 6.4 mg/kg) of THC can decrease anxiety measures in the elevated plus maze (Liu et al., 2022). Certainly, an increase in SI has been linked to decreased anxiety in previous work (File and Hyde, 1978), and this could be the mechanism by which this is occurring in the current study.

Few changes in THC sensitivity by chronic CBD were found in this study. Indeed, the only change detected was impaired startle habituation in Nrg1 mutants treated with both CBD and THC. We found increased hippocampal GluA1 levels in Nrg1 mutants following adolescent CBD, suggesting increased hippocampal excitability. Intra-hippocampal NMDA receptor antagonist infusions disrupt PPI in rats (Shoemaker et al., 2005), and GluA1 receptors are necessary for short-term habituation to recently experienced stimuli (Sanderson et al., 2010; Sanderson and Bannerman, 2012), such as a startle pulse, suggesting a role for hippocampal glutamatergic receptors, and possibly GluA1 receptors in regulating startle habituation. As we did not see CBD-induced changes to startle or PPI in Nrg1 TM HET mice, it is possible that CBD had subthreshold effects on sensorimotor gating, and a combination of CBD and THC was needed to exacerbate glutamatergic receptor imbalance in Nrg1 TM HET mice thereby impairing startle habituation. Indeed, in cannabis users, the combination of acute THC and CBD reduces mismatch negativity (an endophenotype for schizophrenia) more than acute CBD or THC alone (Greenwood et al., 2022). Considering THC can increase hippocampal GluA1 receptor levels in rats (Rubino et al., 2015; Zamberletti et al., 2016), it is possible THC could have exacerbated effects of CBD on startle habituation in Nrg1 TM HET mice, via a GluA1-based mechanism.

We found increased hippocampal GAD67 levels in VEH-treated Nrg1 mutant mice compared to VEH WTs, which were decreased by chronic CBD treatment. In other studies, decreased PFC and hippocampal GAD67 protein and mRNA has been reported in individuals with schizophrenia (Akbarian and Huang, 2006). However, the Nrg1 TM mouse model has been proposed as a gain of function model (Long et al., 2015), thus elevated Nrg1 protein levels may increase GAD67 protein levels. Indeed, acute Nrg1 treatment has been found to increase GAD67 protein levels in a ketamine rat model of schizophrenia (Wang et al., 2014). Furthermore, recent work has shown that reduced hippocampal GAD67 binding of male poly I:C offspring can be normalized to control levels by chronic adolescent CBD (Osborne et al., 2019b) and CBD increased hippocampal GAD67 in both control and poly I:C female offspring (Osborne et al., 2019a).

In summary, this study suggests chronic adolescent CBD does not limit the development of schizophrenia-relevant behaviors in a Nrg1 mouse model in young adulthood, and may therefore not be a potent preventative therapeutic candidate for patients harboring this mutation. Nonetheless, an increase in social behaviors after chronic adolescent CBD does suggest some therapeutic potential, perhaps for the treatment of social withdrawal. Future research should expand on our findings and consider also testing female mice and other CBD dosing regimes. Other genetic schizophrenia model systems should be considered also as CBD continues to show promise as an intervention for schizophrenia-relevant behaviors in other preclinical model systems (Osborne et al., 2017, 2019a; da Silva et al., 2020).