Deletion Of CB2 Cannabinoid Receptor Induces Schizophrenia-Related Behaviors In Mice

Truth Seeker

New Member
Abstract
The possible role of the CB2 receptor (CB2r) in psychiatric disorders has been considered. Several animal models use knockout (KO) mice that display schizophrenia-like behaviors and this study evaluated the role of CB2r in the regulation of such behaviors. Mice lacking the CB2r (CB2KO) were challenged in open field, light—dark box, elevated plus-maze, tail suspension, step down inhibitory avoidance, and pre-pulse inhibition tests (PPI). Furthermore, the effects of treatment with cocaine and risperidone were evaluated using the OF and the PPI test. Gene expression of dopamine D2 (D2r), adrenergic-α2C (α2Cr), serotonergic 5-HT2A and 5-HT2C receptors (5-HT2Ar and 5-HT2Cr) were studied by RT-PCR in brain regions related to schizophrenia. Deletion of CB2r decreased motor activity in the OF test, but enhanced response to acute cocaine and produced mood-related alterations, PPI deficit, and cognitive impairment. Chronic treatment with risperidone tended to impair PPI in WT mice, whereas it 'normalized' the PPI deficit in CB2KO mice. CB2KO mice presented increased D2r and α2Cr gene expressions in the prefrontal cortex (PFC) and locus coeruleus (LC), decreased 5-HT2Cr gene expression in the dorsal raphe (DR), and 5-HT2Ar gene expression in the PFC. Chronic risperidone treatment in WT mice left α2Cr gene expression unchanged, decreased D2r gene expression (15 μg/kg), and decreased 5-HT2Cr and 5-HT2Ar in PFC and DR. In CB2KO, the gene expression of D2r in the PFC, of α2Cr in the LC, and of 5-HT2Cr and 5-HT2Ar in PFC was reduced; 5-HT2Cr and 5-HT2Ar gene expressions in DR were increased after treatment with risperidone. These results suggest that deletion of CB2r has a relation with schizophrenia-like behaviors. Pharmacological manipulation of CB2r may merit further study as a potential therapeutic target for the treatment of schizophrenia-related disorders.

INTRODUCTION
In recent years, it has been postulated that the endocannabinoid system may be an important element involved in the development of schizophrenia. This assumption is supported by several observations: (1) cannabis use was reported in around 50% of schizophrenic patients (Barnett et al, 2007; Bersani et al, 2002); (2) cannabis use is associated with an increased risk of developing schizophrenia (Andreasson et al, 1987; Zammit et al, 2002); (3) a decrease in the age of onset of schizophrenia has been reported (Sugranyes et al, 2009); and (4) schizophrenia-like symptoms may develop in non-schizophrenic cannabis users (Morgan and Curran, 2008). In addition, schizophrenic patients showed a worsening of both positive and negative symptoms and cognitive deficits after administration of Δ9-THC (D'Souza et al, 2005). Indeed, high levels of anandamide have been found in plasma and cerebrospinal fluid (CSF) in antipsychotic-naive first-episode paranoid schizophrenics (Leweke et al, 1999; Yao et al, 2002), and increased CB1 receptor (CB1r) expression in the dorsolateral prefrontal cortex (PFC) of schizophrenic patients (Dean et al, 2001). It is interesting to note that drugs used to treat schizophrenia modify these abnormal parameters in the endogenous cannabinoid system. In fact, the increased levels of anandamide in the CSF (Giuffrida et al, 2004) and plasma (De Marchi et al, 2003) are normalized by antipsychotic treatment. On the other hand, antipsychotic treatment decrease CB1r immunodensity in the PFC of subjects with schizophrenia as determined in post-mortem studies (Urigüen et al, 2009).

As for the anatomic location of cannabinoid receptors, CB1r is mainly expressed in the CNS, particularly the basal ganglia, hippocampus, cerebellum, and cerebral cortex (Abood et al, 2010a; Herkenham et al, 1990, 1991). In contrast, CB2 receptor (CB2r) has been considered the 'peripheral' cannabinoid receptor owing to its presence in the spleen and lymphocytes (Abood et al, 2010b; Munro et al, 1993). However, considerable functional and anatomic evidence suggests that CB2r are expressed in the nervous system. CB2r were recently found in the brainstem of rat, mouse, and ferret (Van Sickle et al, 2005). Further studies in rats have identified CB2r distributed extensively throughout different brain areas, including the spinal nucleus, hippocampus, olfactory nucleus, cerebral cortex, amygdala, striatum, thalamus, and cerebellum (Atwood and Mackie, 2010; Gong et al, 2006; Onaivi et al, 2006). In addition, CB2r gene expression has been identified in the thalamus, periaqueductal grey matter, cervical and thoracic spinal cord, and different brain nuclei, including the caudate—putamen, nucleus accumbens, cingulate cortex, amygdala, hippocampus, ventromedial hypothalamic nucleus, arcuate nucleus, substantia nigra, and dorsal and medial raphe nuclei (Garcia-Gutierrez et al, 2010; Racz et al, 2008).

In addition, recent studies have sketched out the potential role of cannabinoid CB2r in the neurobiology of psychiatric disorders (Ishiguro et al, 2010; Onaivi, 2009). Interestingly, a close relation between diminished CB2r function (polymorphism Q63R) and increased susceptibility to schizophrenia in the presence of other risk factors has been reported (Ishiguro et al, 2010). This observation supports the notion of a relationship between the cannabinoid system and the development of schizophrenia. It is important to point out the role of the CB2r in the control of fundamental neural cell processes, such as proliferation and survival (Fernandez-Ruiz et al, 2007). The proliferation of hippocampal neural progenitors may be modulated through CB2r (Goncalves et al, 2008; Palazuelos et al, 2006) and has resulted defective in CB2KO (knockout) mice (Buckley et al, 2000). Therefore, it can be hypothesized that deletion of the CB2r gene could result in the development of neurochemical abnormalities that may underlie possible behavioral alterations in several experimental paradigms.

On the other hand, several reports have pointed out that the deletion of CB1r produces a behavioral endophenotype consisting of a high degree of anxiety and increased vulnerability to depression-like behaviors (Martin et al, 2002; Urigüen et al, 2004) and decreased memory impairment (Bohme et al, 1999; Reibaud et al, 1999). To date, information on the impact of the elimination of the CB1r on the development of schizophrenia/like behaviors has not been available.

The aim of this study was to determine the response of CB2KO mice in different behavioral and cognitive experimental paradigms and gene expression alterations in brain areas related to schizophrenia. This behavioral and cognitive profile covered motor, anxiety, depression, short- and long-term memory, and attention deficit. Furthermore, the effects of treatment with risperidone were evaluated by using the pre-pulse inhibition test (PPI) in wild-type (WT) and CB2KO mice. The gene expression studies were designed to detect changes in dopamine D2 receptor (D2r) in the PFC, adrenergic-α2C receptor (α2Cr) in the PFC and locus coeruleus (LC), and serotonergic 5-HT2A and 5-HT2C receptors (5-HT2Ar and 5-HT2Cr) gene expression in the PFC and dorsal raphe (DR) by real-time PCR of naive WT and CB2KO mice treated with risperidone.

MATERIALS AND METHODS
Animals
Male CB2KO mice on a C57BL/6J congenic background (kindly provided by Nancy E Buckley, Cal State Polytechnic University, Pomona, CA) were used. CB2KO founders crossed with outbred CD1 (Charles River, L'Arbresle Cedex, France) background (Buckley et al, 2000). Homozygotes from CB2KO (n=119) and age-matched WT mice (n=108) were used in all experiments. Mice were 2—3 months old and weighed 25—35 g at the beginning of the experiments. All animals were kept at controlled temperature (23±2°C) and light conditions (light—dark cycle switching at 0800 hours and 2000 hours). All studies were conducted in compliance with Spanish Royal Decree 223/1998 of 14 March (BOE. 8 18), the Ministerial Order of 13 October 1989 (BOE 18), and European Council Directive of 24 November 1986 (86/609/EEC) regulating the care of experimental animals. The evaluation of behaviors in the light—dark box, elevated plus-maze, tail suspension, and memory tests were made manually under blind conditions.
Drugs
Cocaine hydrochloride was obtained from the Ministry of Health and Consumer Affairs (Ministerio de Sanidad y Consumo, AGEMED, Madrid, Spain) and was dissolved in sterile 0.9% physiological saline. Mice were administered a single intraperitoneal 15 mg/kg dose 10 min before motor activity assessment. The cocaine dose was selected based on previous experiments from our laboratory not showing stereotyped behavior (data not shown). The atypical antipsychotic risperidone (STADA, Barcelona, Spain) was dissolved in sterile 0.9% physiological saline and administered per os in a volume of 10 ml/kg. Risperidone was administered at doses of 15, 30, and 60 μg/kg, twice a day (0830 and 1800 hours). CB2KO (n=36) and WT (n=36) animals treated with risperidone or saline were used for PPI and gene expression studies.
Motor Activity
Open field test The open field consists of a transparent square cage 25 × 25 × 25 cm3 with a white Plexiglas floor (Urigüen et al, 2004). Mice were individually placed in the center to initiate a 20-min test that was recorded with a video camera and analyzed with the SMART (Spontaneous Motor Activity Recording and Tracking) v.2.5.3 software system (Panlab, Barcelona, Spain). Traveled distance and mean speed were analyzed. At 10 min after cocaine administration, motor activity was measured for 20 min. Results obtained in this test were analyzed in the whole 20-min period and also divided into 5-min periods.
Anxiety-Like Behavior
Light—dark box test This model (Crawley and Goodwin, 1980) consisted of two methacrylate boxes 20 × 20 × 15 cm3, one transparent and one black and opaque, linked by an opaque tunnel (4 cm). Light from a 60 W desk lamp located 25 cm above the light box provided room illumination. Mice were individually placed facing the black box and tested in 5-min sessions. The time spent in the lighted area and the number of transitions was recorded. A mouse that introduced three paws into the opposite side of the box was counted as a transition.
Elevated plus-maze test This paradigm consisted of two open arms and two enclosed horizontal perpendicular arms 50 cm above the floor (Lister, 1987). The junction of four arms formed a central squared platform (5 × 5 cm2). The test began with the animal being placed in the center of the apparatus facing one of the enclosed arms and allowed to explore freely for 5 min. We counted as arm entries the introduction of four paws into the arm. The time spent in the open arms and the number of open-arm entries was recorded.
Depression-Like Behavior
Tail suspension test Mice were individually suspended by the tail at the edge of a lever suspended above the table top (the distance to the table surface was 35 cm), and affixed with adhesive tape placed approximately 1—2 cm from the tip of the tail (Vaugeois et al, 1997). The duration of immobility was measured for 6 min. In this situation, mice develop escape-oriented behaviors interspersed with increasingly longer bouts of immobility.
Evaluation of Short- and Long-Term Memory
Step down inhibitory avoidance The apparatus is a 31 × 19 × 15 cm3 acrylic box with a platform located next to a grid. Mice were placed on the platform and their latency to step down on the grid with all four paws was measured; a modified protocol was followed (Izquierdo et al, 1998). During the training session, immediately after stepping down on the grid, the animals received a 2.0-s, 0.4-mA scrambled foot shock. Retention tests were procedurally identical, except that no foot shock was given and the latency to step down in these conditions was taken as a measure of emotional memory. A ceiling of 180 s was imposed, that is, animals with a test latency of more than 180 s were counted as 180 s. Each animal was tested at 1 and 3 h after training (short-term memory) and at 24 h (long-term memory).
Sensorimotor Gating
Acoustic pre-pulse inhibition Pre-pulse inhibition refers to the reduction in amplitude of the startle reflex that occurs when a brief, sub-threshold stimulus immediately precedes a startle stimulus (Hoffman and Ison, 1980). Startle responses were measured using the CIBERTEC REST 141 system (Madrid, Spain). The testing chamber consisted of a plastic adjustable cover mounted on a platform. Movement of the mice within the cover was detected by a piezoelectric accelerometer attached below the platform. A loudspeaker mounted 15 cm above the cover provided background white noise and both acoustic pulses and pre-pulses. The entire apparatus was housed in a ventilated enclosure. Presentation of acoustic pulse and pre-pulse stimuli was controlled by the MONRS software and interface system, which also digitized, rectified, and recorded the responses from the accelerometer. Mean startle amplitude was determined by averaging ten 100 ms readings taken from the beginning of the pulse stimulus onset. Test sessions consisted of no stimulus, pulse-only, and pre-pulse trials. Each 'pre-pulse' trial consisted of a 20-ms 68, 71, or 77 dB non-startling pre-pulse followed 100 ms later by a 40-ms startling pulse of 120 dB. In contrast, 'pulse-only' trials consisted of the 120-dB stimulus only, and 'no-stimulus' trials contained background noise only. To allow acclimatizing, 3 days before the performance of the test sessions, mice were placed each day in the apparatus for 5 min without background noise. Test sessions began with a 5-min acclimatization period using a background noise of 65 dB. The test sessions consisted of a series of 11 pulse-only trials for habituation purposes, followed by 10 trials of each pre-pulse intensity plus pulse, and 10 no-stimulus trials, all presented in a pseudorandom order with a 7—23 s inter-trial variable interval. The percentage of pre-pulse inhibition was defined as ((startle amplitude on pulse alone trials−startle amplitude on pre-pulse trials) × 100)/startle amplitude on pulse alone trials.
Experimental design Pre-pulse inhibition experiments were carried out in two steps. Firstly, the PPI response was determined under baseline conditions (n=36). After confirming a significant difference in the PPI response, the effect of chronic oral risperidone treatment (twice a day for 12 days) was tested in both CB2KO (n=36) and WT (n=36) mice. Animals were randomly assigned to each treatment group (saline, risperidone 15, 30, and 60 μg/kg) in each genotype. PPI response was evaluated after 4, 8, and 12 days of treatment. Each PPI test session was conducted between 1200 and 1500 hours. At 1 h after the last test session, animals were killed and brains were removed for gene expression studies. The number of animals used for statistical analyses (n=32 per genotype) was slightly lower than the initial number of treated animals owing to software failure (n=4), escape from PPI restraint (n=3), and accidental death after oral administration (n=1).
Gene Expression Analyses
Gene expression studies focused on the main targets of the mechanism of action of risperidone, which is characterized by potent blockade of 5-HT2Ar coupled with the relatively weaker antagonism of the dopamine D2r. In addition, this drug displays high affinity for serotonin 5-HT2Cr and adrenergic α2Cr (Schotte et al, 1995).
Real-time PCR Mice were killed and brains were removed from the skull and frozen over dry ice. Coronal brain sections (500 μm) beginning at plates 19—20 (Paxinos and Franklin, 2001) were obtained in a cryostat (−10°C). The PFC, LC, and DR were microdissected according to a modification of the Palkovits method (Palkovits, 1983) as described previously (Garcia-Gutierrez et al, 2010). Total RNA was isolated from brain tissue micropunches using Trizol reagent (Invitrogen, Madrid, Spain) and subsequently retrotranscribed to cDNA. Quantitative analysis of the relative abundance of 5-HT2Ar, 5-HT2Cr, D2r, and α2Cr gene expressions was performed on the ABI PRISM 7700 Sequence Detector System (Applied Biosystems, Foster City, CA). All reagents were obtained from Applied Biosystems and the manufacturer protocols were followed. The reference gene used was 18S rRNA, detected using Taqman ribosomal RNA control reagents. All primer—probe combinations were optimized and validated for relative quantification of gene expression. Briefly, data for each target gene were normalized to the endogenous reference gene, and the fold change in target gene mRNA abundance was determined using the 2−ΔΔCt method (Schmittgen et al, 2000). This quantification method involves comparing the Ct values of the samples of interest with a control or calibrator, such as a non-treated sample or RNA from normal tissue. The Ct values of both the calibrator and the samples of interest are normalized to an appropriate endogenous housekeeping gene (18S rRNA). CB2KO (n=10) and WT (n=8) intact animals were used to study receptor gene expression under baseline conditions. Not all the samples analyzed resulted in useful data owing to failed reactions during RT-PCR processing. From the initial CB2KO (n=36) and WT (n=36) mice treated with risperidone, eight samples were used per treatment group.
Statistical Analyses
In the open field, light—dark box, elevated plus-maze and tail suspension tests, amplitude of acoustic startle response baseline determination, and gene expression studies under baseline conditions, statistical analysis was performed using the Student's t-test for comparing two groups. One-way analysis of variance (ANOVA) with repeated measures was carried out for both 5-min periods in the open field and step down inhibitory avoidance tests, and pre-pulse inhibition first determination. When appropriate, post hoc individual differences between groups were determined using the Student—Newman—Keuls test. Two-way ANOVA was carried out to evaluate the dose—response effects of risperidone treatment on gene expression. When appropriate, post hoc individual differences between groups were determined using the Student—Newman—Keuls test. Two-way ANOVA with repeated measures was used to analyze the temporal course and dose—response effects of risperidone treatment on the amplitude of the acoustic startle response. Two-way ANOVA with two repeated measures was used to analyze the temporal course and dose—response effects of risperidone treatment on %PPI. Differences were considered significant if the probability of error was less than 5%. SigmaStat v3.11 and SPSS v17 software was used for all statistical analyses.

RESULTS
Assessment of Motor Activity: Open Field Test
Analyses of 20-min periods: CB2KO mice (n=14) presented significantly shorter traveled distances compared with their respective controls (n=10) (Student's t-test, t=2.482, p=0.021, 22 d.f.) (Figure 1a). Interestingly, the administration of cocaine (15 mg/kg, intraperitoneal) significantly increased the traveled distance in CB2KO mice (n=14) compared with WT mice (n=10) (Student's t-test, t=−5.410, p<0.001, 22 d.f.) (Figure 1c).

Analyses of 5-min periods: Under baseline conditions, CB2KO mice (n=14) presented significantly shorter traveled distances compared with their corresponding control (n=10) in the open field test. One-way ANOVA with repeated measures showed the significant main effects of genotype (F(1,22)=6.150, p=0.021) and time course (F(1,22)=124.922, p<0.001), but not significant effect of the genotype × time-course interaction (F(1,22)=0.097, p=0.759) (Figure 1b). The administration of cocaine (15 mg/kg, intraperitoneal) significantly increased the traveled distance in CB2KO (n=14) compared with WT mice (n=10). One-way ANOVA with repeated measures showed significant main effects of genotype (F(1,22)=29.271, p<0.001) and time course (F(1,22)=36.851, p<0.001), but no significant effect of the genotype × time-course interaction (F(1,22)=0.773, p=0.389) (Figure 1d).

Assessment of Anxiety-Like Behaviors
CB2KO mice (n=12) spent significantly less time in the light box compared with WT mice (n=14) (Student's t-test, t=4.367, p<0.001, 24 d.f.) (Figure 2a). No differences were observed between the two genotypes in the number of transitions (Student's t-test, t=0.341, p=0.736, 24 d.f.) (Figure 2b).

In the elevated plus-maze test, the percentage of time spent in the open arms significantly decreased in CB2KO (n=12) compared with WT mice (n=14) (Student's t-test, t=5.440, p<0.001, 24 d.f.) (Figure 2c). No differences were observed in the number of transitions between compartments (Student's t-test, t=−1.024, p=0.316, 24 d.f.) (Figure 2d).

Assessment of Depressive-Like Behaviors: Tail Suspension Test
Exposure to the tail suspension induced a higher immobility time in CB2KO mice (n=8) than in WT littermates (n=10) (Student's t-test, t=−10.047, p<0.001, 16 d.f.), revealing increased despair behavior in the mice that lacked CB2r (Figure 2e).
Assessment of Memory Impairment: Step Down Inhibitory Avoidance
No difference was observed in the pre-training session between CB2KO (n=11) and WT (n=8) mice. Interestingly, the evaluation of representative parameters of short-term memory revealed that CB2KO presented a shorter latency time at 1 and 3 h after scrambled foot shock. Moreover, CB2KO mice had a lower latency time at 24 h (long-term memory) (one-way ANOVA with repeated measures followed by Student—Newman—Keuls test, genotype F(1,17)=7.446, p=0.014; time F(1,17)=6.297, p=0.023; genotype × time F(1,17)=8.069, p=0.011) (Figure 3).

Assessment of Sensorimotor Gating: Pre-pulse Inhibition
CB2KO mice (n=36) did not show differences in the acoustic startle response amplitude compared with WT mice (n=36) (Student's t-test, t=0.265, p=0.792, 70 d.f.) (Figure 4a). On the other hand, PPI was significantly decreased in CB2KO compared with WT mice. One-way ANOVA with repeated measures revealed the significant main effects of genotype (F(1,70)=16.762, p<0.001) and pre-pulse intensity (F(1,70)=230.285, p<0.001), whereas the genotype × pre-pulse intensity interaction was not significant (F(1,70)=0.155, p=0.695) (Figure 4b). Therefore, the reduction of PPI observed in CB2KO mice was independent of the pre-pulse intensity tested.

Time Course and Dose Response of the Effects of Treatment with Risperidone on Acoustic Startle Response Amplitude
Risperidone treatment had a statistically significant effect, whereas genotype factor (n=32) did not alter the acoustic startle response amplitude (Figure 5, Table 1). The time of treatment induced significant changes in startle amplitude.

Time Course and Dose Response of the Effects of Treatment with Risperidone on %PPI
The complexity of this experimental design did not allow the main factors of variation, genotype, and risperidone treatment to reach statistical significance (Figure 6, Table 2). However, the two factors of repetition, pre-pulse intensity and time of treatment, have emerged as relevant factors of variation. In the case of pre-pulse intensity, more intense pre-pulse induced a higher %PPI value. In addition, the significant interaction between these two factors and genotype indicates that the evolution of %PPI differed in WT and CB2KO mice. Thus, WT mice treated with saline showed a tendency toward lower %PPI (10—15%) over the course of treatment. In contrast, CB2KO mice treated with saline showed a tendency toward higher %PPI (5—15%) in the same period. Although CB2KO mice had a lower %PPI in the baseline determination, this slight increase could indicate that these mice maintain capacity to improve pre-attention. The differences between genotypes in the temporal course of %PPI were also observed in risperidone-treated animals. Thus, risperidone treatment enhanced the reduction of %PPI observed in saline-treated WT mice. The lowest and medium doses resulted in a 12—17%, decrease, and the highest dose in a 16—21% decrease. Consequently, risperidone treatment tended to reduce pre-pulse inhibition. On the other hand, risperidone treatment increased %PPI in CB2KO mice. The percentage PPI increased by 25—36% (15 μg/kg), 25—30% (30 μg/kg), and 32—34% (60 μg/kg). In CB2KO mice, risperidone treatment tended to improve PPI response (Figure 6, Table 2).

Evaluation of D2r, α2Cr, 5-HT2Cr, and 5-HT2Ar Gene Expressions
Dopamine D2r gene expression was studied in the PFC under baseline conditions and after 12 days of treatment with risperidone (15, 30, and 60 μg/kg, per os) or saline in WT and CB2KO mice. D2r gene expression was significantly increased in the PFC of CB2KO mice compared with WT mice (Student's t-test, t=−2.933, p=0.013, 12 d.f.) (n=6—8) (Figure 7a). After 12 days, risperidone treatment modified D2r gene expression in the PFC (two-way ANOVA, genotype: F(1,63)=5.397, p=0.024; risperidone treatment: F(3,63)=6.805, p<0.001; genotype × treatment interaction: F(3,63)=6.856, p<0.001). In WT mice (n=8 per group), only the 15 μg/kg risperidone dose significantly reduced D2r gene expression in the PFC (Student—Newman—Keuls test, p<0.05) (Figure 7b). In CB2KO mice (n=8 per group), treatment with all risperidone doses (15, 30, and 60 μg/kg, per os) significantly reduced D2r gene expression in the PFC (Student—Newman—Keuls test, p<0.05) (Figure 7b). Saline treatment did not modify the observed differences (under baseline conditions) in D2r gene expression between CB2KO and WT mice (Student—Newman—Keuls test, p<0.05) (Figure 7b). D2r gene expression at doses of 30 and 60 μg/kg of risperidone was significantly reduced in CB2KO compared with risperidone-treated WT mice (Student—Newman—Keuls test, p<0.05) (Figure 7b).

Secondly, α2Cr gene expression was analyzed in the PFC and the LC under baseline conditions and after 12 days of treatment with risperidone (15, 30, and 60 μg/kg, per os) or saline in WT and CB2KO mice. The α2Cr gene expression was significantly increased in the PFC (Student's t-test, t=−4.262, p=0.003, 9 d.f.) (n=5—6) (Figure 8a) and LC (Student's t-test, t=−2.420, p=0.039, 10 d.f.) (n=6) (Figure 8c) of CB2KO compared with WT mice under baseline conditions. In the PFC, risperidone or saline treatment did not modify the α2Cr gene expression found under baseline conditions (two-way ANOVA, genotype: F(1,63)=960.581, p<0.001; risperidone treatment: F(3,63)=0.426, p=0.735; genotype × treatment interaction: F(3,63)=1.447, p=0.239) (Figure 8b). In the LC, risperidone treatment reduced α2Cr gene expression only in CB2KO mice (n=8 per group), whereas it failed to alter α2Cr gene expression in WT mice (n=8 per group) (two-way ANOVA, genotype: F(1,63)=114.593, p<0.001; risperidone treatment: F(3,63)=5.204, p=0.003; genotype × treatment interaction: F(3,63)=5.077, p=0.004) (Figure 8d). Treatment with all the doses studied of risperidone in CB2KO significantly reduced α2Cr gene expression in the LC (Student—Newman—Keuls test, p<0.05) (Figure 8d). Although treatment with all the doses of risperidone studied reduced gene expression level in CB2KO, these levels were significantly higher than those observed in WT mice (Student—Newman—Keuls, p<0.05) (Figure 8d).

Thirdly, 5-HT2Cr gene expression was examined in the PFC and DR under baseline conditions and after 12 days of risperidone treatment (15, 30, and 60 μg/kg, per os) or saline in WT and CB2KO mice. Under baseline conditions, no difference was observed between the two genotypes in the PFC (Student's t-test, t=−0.025, p=0.981, 10 d.f.) (n=6) (Figure 9a). However, risperidone treatment reduced 5-HT2Cr gene expression in the PFC in both genotypes (two-way ANOVA, genotype: F(1,63)=237.027, p<0.001; risperidone treatment: F(3,63)=36.346, p<0.001; genotype × treatment interaction: F(3,63)=37.532, p<0.001). In WT mice (n=8 per group), oral risperidone treatment (15, 30, and 60 μg/kg, per os) significantly reduced 5-HT2Cr gene expression in the PFC (Student—Newman—Keuls test, p<0.05) (Figure 9b). In CB2KO mice (n=8 per group), only risperidone at the dose of 60 μg/kg (per os) significantly reduced 5-HT2Cr gene expression in the PFC compared with the saline group (Student—Newman—Keuls test, p<0.05) (Figure 9b). The reduction observed in the risperidone-treated WT mice was significant compared with risperidone-treated CB2KO mice at all doses (Student—Newman—Keuls test, p<0.05) (Figure 9b). On the other hand, under baseline conditions, CB2KO presented significantly reduced 5-HT2Cr gene expression in the DR compared with WT mice (Student's t-test, t=2.919, p=0.011, 14 d.f.) (n=7—9) (Figure 9c). The effect of risperidone treatment on 5-HT2Cr gene expression in the DR was different in WT and CB2KO mice (two-way ANOVA, genotype: F(1,63)=4.901, p=0.031; risperidone treatment: F(3,63)=4.164, p=0.010; genotype × treatment interaction: F(3,63)=14.026, p<0.001). In WT mice (n=8 per group), 30 and 60 μg/kg risperidone doses significantly reduced 5-HT2Cr gene expression in the DR (Student—Newman—Keuls test, p<0.05) (Figure 9d). In contrast, in CB2KO mice (n=8 per group) the 60 μg/kg risperidone dose significantly increased 5-HT2Cr gene expression in the DR (Student—Newman—Keuls test, p<0.05) (Figure 9d). Significant differences between CB2KO and WT mice were observed in saline-treated and 30 and 60 μg/kg risperidone-treated mice (Student—Newman—Keuls test, p<0.05) (Figure 9d).

Finally, 5-HT2Ar gene expression was examined in the PFC and DR under baseline conditions and after 12 days of risperidone (15, 30, and 60 μg/kg, per os) or saline treatment in WT and CB2KO mice. Under baseline conditions, gene expression of 5-HT2Ar was significantly reduced in the PFC of CB2KO compared with WT mice (Student's t-test, t=2.502, p=0.024, 15 d.f.) (n=7—10) (Figure 10a). Risperidone treatment had not significant effect of on 5-HT2Ar gene expression in the PFC (two-way ANOVA, genotype: F(1,63)=13.838, p<0.001; risperidone treatment: F(3,63)=2.756, p=0.051; genotype × treatment interaction: F(3,63)=0.274, p=0.844). The differences observed in CB2KO and WT mice treated with risperidone (n=8 per group) were due to the genotype factor (Figure 10b). In the DR under baseline conditions, no significant difference was observed in 5-HT2Ar gene expression between CB2KO and WT mice (Student's t-test, t=1.533, p=0.164, 8 d.f.) (n=5) (Figure 10c). However, risperidone treatment induced different effects on 5-HT2Ar gene expression in the DR of WT and CB2KO mice (two-way ANOVA, genotype: F(1,63)=36.216, p<0.001; risperidone treatment: F(3,63)=4.012, p=0.012; genotype × treatment interaction: F(3,63)=23.454, p<0.001). In WT mice (n=8 per group), treatment with risperidone doses of 30 and 60 μg/kg significantly reduced 5-HT2Ar gene expression in the DR (Student—Newman—Keuls test, p<0.05) (Figure 10d). In CB2KO mice (n=8 per group), the 60 μg/kg risperidone dose significantly increased 5-HT2Ar gene expression in the DR (Student—Newman—Keuls test, p<0.05) (Figure 10d). Significant differences between CB2KO and WT were observed in mice treated with 30 and 60 μg/kg risperidone (Student—Newman—Keuls test, p<0.05) (Figure 10d).

DISCUSSION
The results of this study provide new information about a possible role of the CB2r in the regulation of schizophrenia-like behaviors. This claim is supported by several observations: (1) CB2KO mice exhibited decreased spontaneous motor activity and increased sensitivity to the motor stimulant effects of acute cocaine administration in the open field test; (2) deletion of CB2r gene produced an anxiogenic-like response in the light—dark box and elevated plus-maze tests, and a depressogenic-like response in the tail suspension test; (3) CB2KO mice showed disrupted short- and long-term memory consolidation in the step down inhibitory avoidance paradigm; (4) the PPI of the acoustic startle response was significantly lower in CB2KO mice compared with WT mice; (5) PPI was markedly enhanced after chronic oral treatment with the antipsychotic drug risperidone in CB2KO mice, but was not affected in WT mice; (6) deletion of CB2r's increased D2r and α2Cr gene expression in the PFC and LC and decreased 5-HT2Cr gene expression in the DR and 5-HT2Ar gene expression in the PFC of CB2KO compared with WT mice; (7) oral risperidone treatment of WT mice did not affect α2Cr gene expression, decreased D2r (15 μg/kg) in the PFC, and decreased 5-HT2Cr and 5-HT2Ar in the PFC and DR; and (8) treatment with risperidone in CB2KO mice reduced gene expressions of D2r in the PFC, α2Cr in the LC, and 5-HT2Cr and 5-HT2Ar in the PFC; risperidone treatment increased 5-HT2Cr and 5-HT2Ar gene expression in the DR.

It is known that schizophrenia is associated with brain abnormalities induced during CNS development (Rapoport et al, 2005; Ross et al, 2006). A number of findings suggest a pro-neurogenic role of CB2r in the control of fundamental neural cell processes (Galve-Roperh et al, 2008; Harkany et al, 2007; Katona and Freund, 2008). Therefore, it can be hypothesized that the lack of CB2r might impair neural development, thus inducing relevant alterations in several brain areas. In this context, it can be postulated that these alterations could be the substrate underlying the behavioral modifications observed in CB2KO mice. However, the existence of developmental compensatory mechanisms related to the lack of this receptor cannot be excluded. In addition, it is well known that the systemic administration of receptor antagonists does not mimic the situation of a KO mouse for the same receptor. The blockade of the CB2r has been involved in the prevention of alcohol preference development (Ishiguro et al, 2006), the inhibition of food consumption, and the enhancement of spontaneous activity and stereotyped behavior in C57BL/6 and DBA/2 mice (Onaivi et al, 2008). In contrast, pre-pulse inhibition or locomotor activity was not affected in mice (Ishiguro et al, 2010). Indeed, transient inactivation of the receptor by CB2r antagonist use in WT mice does not induce the behavioral abnormalities observed in CB2KO mice.

In recent years, a considerable number of susceptibility genes for schizophrenia have been described (Desbonnet et al, 2009). The generation of mice with targeted mutation of these genes was focused on the development of genetic models of the putative regulation of the pathophysiological mechanisms (Gray et al, 2009; Halene et al, 2009; Han et al, 2009; Perry et al, 2009; Powell et al, 2008; Rojas et al, 2007; Sakae et al, 2008; Tanda et al, 2009; Wiedholz et al, 2008) and genetic models of risk for schizophrenia (Bégou et al, 2008; Dyck et al, 2009; Karl et al, 2007; Kvajo et al, 2008; Willi et al, 2010). Schizophrenia includes abnormalities in coordination and movement including extrapyramidal paradigms (tending toward higher motor activity) and catatonia (usually lower motor activity) (Peralta et al, 2010). Schizophrenia manifests a spectrum of different levels of motor activity observable as endophenotypes of this disease. Mice with mutations in the genes of susceptibility to schizophrenia exhibited increased (Gray et al, 2009; Powell et al, 2008; Sakae et al, 2008; Wiedholz et al, 2008), decreased (Hattori et al, 2008), or failure to alter (Wolinsky et al, 2007) motor activity. The results found in this study revealed that deletion of the CB2r gene significantly reduced spontaneous motor activity in the open field and increased sensitivity to the motor stimulant effects of acute cocaine administration. Similarly, PKCI/HINT1 KO mice presented diminished spontaneous motor activity with increased sensitivity to the motor action of amphetamine (Barbier et al, 2007). In contrast, mice that were genetically modified at the NMDA or dopamine receptors presented increased spontaneous motor activity and reduced reactivity to dizocilpine, phencyclidine, amphetamine, or cocaine (Gainetdinov et al, 1999; Mohn et al, 1999). The limitations of this paradigm seem to be related to the lack of specificity and predictive value of the effects on negative or cognitive symptoms. In addition, it has been reported that individuals with bipolar disorder and schizophrenia have distinctive profiles of exploratory behavior (Perry et al, 2009). Patients with bipolar mania present high motor activity and increased object interaction, whereas patients with schizophrenia exhibit normal object interaction (Perry et al, 2009). As patients with bipolar disorder present reduced PPI (Giakoumaki et al, 2007) and increased motor activity (Perry et al, 2009), it is possible that mutant mice with motor hyperactivity and reduced PPI are more closely related to animal models of bipolar disorder than to schizophrenia.

The response of CB2KO mice has been studied in animal models of anxiety and depression. It is important to note that schizophrenic patients report depression and anxiety as the most frequent early signs and symptoms occurring before the first psychotic episode (Iyer et al, 2008). Thus, the characterization of clinically important phenomena observed before the onset of psychosis could be important in the diagnostic process. CB2KO mice display increased anxiogenic-like response in the light—dark box and elevated plus-maze tests. Similarly, mutant mice like the heterozygous YWHAE and homozygous sandy show moderately enhanced anxiety-like behavior in the elevated plus-maze test (Hattori et al, 2008; Ikeda et al, 2008), whereas other mice with deletion of schizophrenia-related genes did not show any variation in anxiety-like behaviors in the light—dark box and elevated plus-maze tests compared with WT mice (Barbier and Wang, 2009; Hsu et al, 2007; Wolinsky et al, 2007). On the other hand, mice with overexpression of CB2r presented reduced anxiety-like behaviors using the same experimental paradigm (García-Gutiérrez and Manzanares, 2011). Deletion of the CB2r gene induced depressive-like responses in the tail suspension test. Similarly, Homer1-KO mice and STOP-null mice show increased depressive-like behavior in the Porsolt test (Delotterie et al, 2010; Szumlinski et al, 2005). However, other animal models based on schizophrenia susceptibility gene mutations showed an antidepressant-like response (Barbier and Wang, 2009; Perona et al, 2008; Sakae et al, 2008; Tanda et al, 2009; Yamasaki et al, 2008). In contrast, overexpression of CB2r resulted in decreased depressive-like behaviors in acute models (tail suspension and novelty-suppressed feeding test) and in exposure to chronic mild stress, suggesting that pharmacological manipulation of this receptor may be an interesting therapeutic target in depression-related behaviors (Garcia-Gutierrez et al, 2010).

Cognitive dysfunction is one of the three main clusters of symptoms in schizophrenia and rodent models involving susceptibility genes related to schizophrenia manifesting as significant deficits in working memory and spatial learning (Gainetdinov et al, 1999; Ikeda et al, 2008; Yamasaki et al, 2008). Using the step down inhibitory avoidance task, CB2KO mice showed disrupted short- and long-term memory consolidation of the task. Other genetically modified mice, proposed as animal models of behavioral and biochemical alterations implicated in schizophrenia, show deficits in short-term working and spatial memory (Bégou et al, 2008; Gray et al, 2009) and deficits in retention of emotional or spatial memory (Rojas et al, 2007; Tanda et al, 2009). However, not all these mutant mice present memory impairment; for instance, trace amine 1 receptor KO mice show no difference on a working memory task compared with WT mice (Wolinsky et al, 2007).

Impaired sensorimotor gating has been proposed as a common feature of the cognitive dysfunction observed in schizophrenia (Braff et al, 2001). The presence of PPI deficit has thus been considered an important behavioral trait in rodent models of schizophrenia. For instance, a clear attention deficit has been observed in mice with mutations in dopamine or glutamate receptors (Ralph et al, 2001; Wiedholz et al, 2008; Yamashita et al, 2006). In this study, baseline PPI was significantly lower in CB2KO mice than WT mice; however, there is little information about the role of CB2r in the regulation of PPI in mice. Previous studies have reported that blockade of the CB2r fails to alter the response to PPI, but increases the attention deficit induced by MK-801 (Ishiguro et al, 2010). Disruption of PPI in rats is at least partly due to activation of D2r (Swerdlow et al, 1991), suggesting that increased activity at these receptors might also be a substrate for PPI deficits in schizophrenia. In this sense, patients with schizophrenia present significant differences in the acoustic startle reflex, habituation to startle stimuli, and several PPI levels compared with healthy controls (Moriwaki et al, 2009), and antipsychotic treatment with olanzapine, aripiprazole, and risperidone improve PPI, but do not modify the acoustic startle reflex or habituation (Kishi et al, 2010; Wynn et al, 2007). Another study using a longitudinal within-subjects design reported improved results after switching from the conventional antipsychotic zuclopenthixol to long-acting injectable risperidone (Martinez-Gras et al, 2009). On the other hand, the administration of antipsychotics increases PPI or reverses the PPI disruptions induced by psychostimulants in animal studies (Egashira et al, 2005; Gray et al, 2009; Nagai et al, 2006; Powell et al, 2008; Swerdlow et al, 1991; Thomsen et al, 2010). Acute administration of clozapine, risperidone, quetiapine, and haloperidol (Egashira et al, 2005; Powell et al, 2008; Thomsen et al, 2010), and chronic administration of clozapine (Gray et al, 2009) reverse the PPI deficit observed in different genetically manipulated mice. In addition, the administration of haloperidol, clozapine, and risperidone reverses drug-induced PPI deficits (Nagai et al, 2006; Swerdlow et al, 1991). The results of this study are consistent with these reports as chronic oral treatment with the atypical antipsychotic risperidone (at doses commonly used in the treatment of schizophrenia in patients) markedly attenuated the PPI deficits observed in CB2KO mice. In contrast, risperidone tended to disrupt PPI progressively over the course of treatment in WT mice. In particular, the dose of 60 μg/kg markedly decreased PPI after 12 days of treatment (21% at 71 dB). On the other hand, risperidone reduced startle amplitude in both genotypes, being more evident in the CB2KO mice.

The exploration of neurochemical changes in CB2KO mice potentially related with the schizophrenia-like behaviors observed revealed alterations in the expression of dopaminergic, adrenergic, and serotonergic receptor genes. It is important to note that variations in the amount of gene transcript do not necessarily mean that concomitant changes in protein occur. These functional modifications could underlie the attention deficit, at least in part, as treatment with risperidone improved the PPI response and tended to 'normalize' some of the genes that are altered in CB2KO mice. In contrast, treatment with risperidone failed to improve the PPI response in WT mice, but slightly impaired PPI response at the highest dose and longest duration of treatment. In addition, the effects of risperidone induced different alterations in receptor gene expression compared with CB2KO mice.

The genes and the brain regions studied were selected based on two criteria: (1) previously reported alterations in gene expression associated with schizophrenia-like behaviors in genetically modified mice (Willi et al, 2010) and (2) a relation between the main receptor targets and the mechanism of action of risperidone. D2r, 5-HT2Ar, 5-HT2Cr, and α2Cr satisfied these criteria (Abi-Dargham and Laruelle, 2005; McCormick et al, 2010; Meltzer and Huang, 2008).

Deletion of the CB2r increased D2r gene expression in the PFC. This situation could be related with increased dopaminergic tone, thus favoring deficient sensorimotor gating as occurs in DAT KO mice that present a chronic hyperdopaminergic state and deficient sensorimotor gating in the PPI (Ralph et al, 2001). The increased gene expression of this dopamine receptor was significantly reduced after treatment with risperidone, suggesting that the effectiveness of this drug on the PPI deficit could be at least partly related to this reduction in gene expression and its ability to block the receptor. The reduced effect of risperidone treatment on D2r gene expression in the PFC of WT mice could be related with the lack of improvement in %PPI observed in these mice.

CB2KO mice presented significantly reduced 5-HT2Cr gene expression in the DR nucleus. This reduction could be related with a decrease in the inhibitory control of DR serotonergic neurons played by 5-HT2Cr (Quérée et al, 2009), which may result in increased serotonin release in the PFC. Furthermore, deletion of the CB2r significantly reduced 5-HT2Ar gene expression in PFC. Taking into account that PFC activity is modulated by serotonin through 5-HT2Ar (Puig et al, 2010), it can be hypothesized that decreased gene expression of 5-HT2Ar could be a compensatory mechanism induced by persistently increased serotonin release by the DR terminals in the PFC. In CB2KO mice, the effects of risperidone tend to restore 5-HT2Cr gene expression in the DR to the levels observed in WT mice, whereas risperidone treatment tends to reduce gene expression of this serotonergic receptor in the PFC. Typical and atypical antipsychotics show a similar effect of reducing 5-HT2Cr gene expression in the cortex (Buckland et al, 1997; Huang et al, 2006). Therefore, if decreased gene expression of 5-HT2Ar in the PFC reflects a possible compensatory mechanism in response to increased serotonin release from DR terminals, it can be hypothesized that risperidone treatment contributes to this mechanism by reducing 5-HT2Ar gene expression in the PFC even more at the same time that it increases 5-HT2Ar gene expression in the DR. Similarly, clozapine reduces 5-HT2Ar gene expression in the frontal cortex (Burnet et al, 1996). In WT mice risperidone treatment reduces 5-HT2Ar and 5-HT2Cr gene expression in the PFC and DR. In the PFC, CB2KO 5-HT2Ar and 5-HT2Cr were reduced, whereas they were increased in the DR. In addition, risperidone treatment in these CB2KO mice markedly improves the PPI deficit. In summary, CB2KO mice present alterations in the serotonergic pathway from the DR to the PFC that could be responsible, at least in part, for the observed PPI deficit. It seems that the effect of risperidone treatment on this serotonergic pathway could be related to its beneficial effect on PPI.

The α2Cr plays an important role in cognitive processing in the PFC. The activation of the α2Cr by an α2 agonist impairs performance of a spatial delayed alternation task, as seen in α2A KO mice (Franowicz et al, 2002). In this sense, we found increased α2Cr gene expression in both the PFC and LC of CB2KO mice. This alteration could thus be related with the observed PPI deficit. However, results obtained in studies made in two types of α2C mutant mice do not support this explanation. Overexpression of α2Cr results in enhanced PPI response (Sallinen et al, 1998). On the other hand, α2C KO mice show reduced PPI response (Sallinen et al, 1998). Despite this discrepancy, it may be speculated that increased expression of the α2C receptor contributes somewhat to the PPI deficit in CB2KO mice. This contribution could be related to the effects of increased α2Cr expression in the LC. Noradrenaline release in the PFC by LC terminals could be reduced. As risperidone treatment in CB2KO mice reduced α2Cr expression in the LC, this effect could restore noradrenaline release in the PFC, thus contributing to PPI improvement. In WT mice, risperidone did not significantly modify α2Cr gene expression.

In summary, deletion of the cannabinoid CB2r gene produced behavioral alterations that are commonly expressed in preclinical animal models of schizophrenia, namely altered locomotor activity, anxiety-like and depressive-like behaviors, and cognitive deficits including impaired sensorimotor gating. Gene expression studies in the PFC, DR, and LC revealed alterations in different dopaminergic, serotonergic, and noradrenergic receptors. Chronic treatment with the atypical antipsychotic risperidone reduced the PPI deficit, an effect that could be associated with modifications in the biochemical alterations observed in CB2KO mice. These results suggest that CB2r deletion was related to the observed schizophrenia-like behaviors. Pharmacological manipulation of CB2r may be further explored as a potential therapeutic target for the treatment of schizophrenia-related disorders.

Source, Graphs and Figures: Neuropsychopharmacology - Deletion of CB2 Cannabinoid Receptor Induces Schizophrenia-Related Behaviors in Mice
 
Back
Top Bottom