Altered cortico-striatal crosstalk underlies object recognition memory deficits in the sub-chronic phencyclidine model of schizophrenia

Linking haploinsufficiency of the autism- and schizophrenia-associated gene Cyfip1 with striatal-limbic-cortical network dysfunction and cognitive inflexibility

Impaired behavioural flexibility is a core feature of neuropsychiatric disorders and is associated with underlying dysfunction of fronto-striatal circuitry. Reduced dosage of Cyfip1 is a risk factor for neuropsychiatric disorder, as evidenced by its involvement in the 15q11.2 (BP1-BP2) copy number variant: deletion carriers are haploinsufficient for CYFIP1 and exhibit a two- to four-fold increased risk of schizophrenia, autism and/or intellectual disability. Here, we model the contributions of Cyfip1 to behavioural flexibility and related fronto-striatal neural network function using a recently developed haploinsufficient, heterozygous knockout rat line. Using multi-site local field potential (LFP) recordings during resting state, we show that Cyfip1 heterozygous rats (Cyfip1+/-) harbor disrupted network activity spanning medial prefrontal cortex, hippocampal CA1 and ventral striatum. In particular, Cyfip1+/- rats showed reduced influence of nucleus accumbens and increased dominance of prefrontal and hippocampal inputs, compared to wildtype controls. Adult Cyfip1+/- rats were able to learn a single cue-response association, yet unable to learn a conditional discrimination task that engages fronto-striatal interactions during flexible pairing of different levers and cue combinations. Together, these results implicate Cyfip1 in development or maintenance of cortico-limbic-striatal network integrity, further supporting the hypothesis that alterations in this circuitry contribute to behavioural inflexibility observed in neuropsychiatric diseases including schizophrenia and autism.

Balance Between Projecting Neuronal Populations of the Nucleus Accumbens Controls Social Behavior in Mice

BACKGROUND

Deficient social interactions are a hallmark of major neuropsychiatric disorders, and accumulating evidence points to altered social reward and motivation as key underlying mechanisms of these pathologies. In the present study, we further explored the role of the balance of activity between D1 and D2 receptor-expressing striatal projection neurons (D1R- and D2R-SPNs) in the control of social behavior, challenging the hypothesis that excessive D2R-SPN activity, rather than deficient D1R-SPN activity, compromises social behavior.

METHODS

We selectively ablated D1R- and D2R-SPNs using an inducible diphtheria toxin receptor-mediated cell targeting strategy and assessed social behavior as well as repetitive/perseverative behavior, motor function, and anxiety levels. We tested the effects of optogenetic stimulation of D2R-SPNs in the nucleus accumbens (NAc) and pharmacological compounds repressing D2R-SPN.

RESULTS

Targeted deletion of D1R-SPNs in the NAc blunted social behavior in mice, facilitated motor skill learning, and increased anxiety levels. These behaviors were normalized by pharmacological inhibition of D2R-SPN, which also repressed transcription in the efferent nucleus, the ventral pallidum. Ablation of D1R-SPNs in the dorsal striatum had no impact on social behavior but impaired motor skill learning and decreased anxiety levels. Deletion of D2R-SPNs in the NAc produced motor stereotypies but facilitated social behavior and impaired motor skill learning. We mimicked excessive D2R-SPN activity by optically stimulating D2R-SPNs in the NAc and observed a severe deficit in social interaction that was prevented by D2R-SPN pharmacological inhibition.

CONCLUSIONS

Repressing D2R-SPN activity may represent a promising therapeutic strategy to relieve social deficits in neuropsychiatric disorders.

Drug-induced change in transmitter identity is a shared mechanism generating cognitive deficits

Cognitive deficits are a long-lasting consequence of drug use, yet the convergent mechanism by which classes of drugs with different pharmacological properties cause similar deficits is unclear. We find that both phencyclidine and methamphetamine, despite differing in their targets in the brain, cause the same glutamatergic neurons in the medial prefrontal cortex to gain a GABAergic phenotype and decrease their expression of the vesicular glutamate transporter. Suppressing the drug-induced gain of GABA with RNA-interference prevents the appearance of memory deficits. Stimulation of dopaminergic neurons in the ventral tegmental area is necessary and sufficient to produce this gain of GABA. Drug-induced prefrontal hyperactivity drives this change in transmitter identity. Returning prefrontal activity to baseline, chemogenetically or with clozapine, reverses the change in transmitter phenotype and rescues the associated memory deficits. The results reveal a shared and reversible mechanism that regulates the appearance of cognitive deficits upon exposure to different drugs.

Anxiety and cognitive-related effects of Δ 9-tetrahydrocannabinol (THC) are differentially mediated through distinct GSK-3 vs. Akt-mTOR pathways in the nucleus accumbens of male rats

RATIONALE

Δ⁹-tetrahydrocannabinol (THC) is the primary psychoactive compound in cannabis and is responsible for cannabis-related neuropsychiatric side effects, including abnormal affective processing, cognitive and sensory filtering deficits and memory impairments. A critical neural region linked to the psychotropic effects of THC is the nucleus accumbens shell (NASh), an integrative mesocorticolimbic structure that sends and receives inputs from multiple brain areas known to be dysregulated in various disorders, including schizophrenia and anxiety-related disorders. Considerable evidence demonstrates functional differences between posterior vs. anterior NASh sub-regions in the processing of affective and cognitive behaviours influenced by THC. Nevertheless, the neuroanatomical regions and local molecular pathways responsible for these psychotropic effects are not currently understood.

OBJECTIVES

The objectives of this study were to characterize the effects of intra-accumbens THC in the anterior vs. posterior regions of the NASh during emotional memory formation, sensorimotor gating and anxiety-related behaviours.

METHODS

We performed an integrative series of translational behavioural pharmacological studies examining anxiety, sensorimotor gating and fear-related associative memory formation combined with regionally specific molecular signalling analyses in male Sprague Dawley rats.

RESULTS

We report that THC in the posterior NASh causes distortions in emotional salience attribution, impaired sensory filtering and memory retention and heightened anxiety, through a glycogen-synthase-kinase-3 (GSK-3)-β-catenin dependent signalling pathway. In contrast, THC in the anterior NASh produces anxiolytic effects via modulation of protein kinase B (Akt) phosphorylation states.

CONCLUSIONS

These findings reveal critical new insights into the neuroanatomical and molecular mechanisms associated with the differential neuropsychiatric side effects of THC in dissociable nucleus accumbens sub-regions.

Facilitating mGluR4 activity reverses the long-term deleterious consequences of chronic morphine exposure in male mice

Understanding the neurobiological underpinnings of abstinence from drugs of abuse is critical to allow better recovery and ensure relapse prevention in addicted subjects. By comparing the long-term transcriptional consequences of morphine and cocaine exposure, we identified the metabotropic glutamate receptor subtype 4 (mGluR4) as a promising pharmacological target in morphine abstinence. We evaluated the behavioral and molecular effects of facilitating mGluR4 activity in abstinent mice. Transcriptional regulation of marker genes of medium spiny neurons (MSNs) allowed best discriminating between 4-week morphine and cocaine abstinence in the nucleus accumbens (NAc). Among these markers, Grm4, encoding mGluR4, displayed down-regulated expression in the caudate putamen and NAc of morphine, but not cocaine, abstinent mice. Chronic administration of the mGluR4 positive allosteric modulator (PAM) VU0155041 (2.5 and 5 mg/kg) rescued social behavior, normalized stereotypies and anxiety and blunted locomotor sensitization in morphine abstinent mice. This treatment improved social preference but increased stereotypies in cocaine abstinent mice. Finally, the beneficial behavioral effects of VU0155041 treatment in morphine abstinent mice were correlated with restored expression of key MSN and neural activity marker genes in the NAc. This study reports that chronic administration of the mGluR4 PAM VU0155041 relieves long-term deleterious consequences of morphine exposure. It illustrates the neurobiological differences between opiate and psychostimulant abstinence and points to pharmacological repression of excessive activity of D2-MSNs in the NAc as a promising therapeutic lever in drug addiction.

Sustained NMDA receptor hypofunction impairs brain-derived neurotropic factor signalling in the PFC, but not in the hippocampus, and disturbs PFC-dependent cognition in mice

BACKGROUND

Cognitive deficits profoundly impact on the quality of life of patients with schizophrenia. Alterations in brain derived neurotrophic factor (BDNF) signalling, which regulates synaptic function through the activation of full-length tropomyosin-related kinase B receptors (TrkB-FL), are implicated in the aetiology of schizophrenia, as is N-methyl-D-aspartate receptor (NMDA-R) hypofunction. However, whether NMDA-R hypofunction contributes to the disrupted BDNF signalling seen in patients remains unknown.

AIMS

The purpose of this study was to characterise BDNF signalling and function in a preclinical rodent model relevant to schizophrenia induced by prolonged NMDA-R hypofunction.

METHODS

Using the subchronic phencyclidine (PCP) model, we performed electrophysiology approaches, molecular characterisation and behavioural analysis.

RESULTS

The data showed that prolonged NMDA-R antagonism, induced by subchronic PCP treatment, impairs long-term potentiation (LTP) and the facilitatory effect of BDNF upon LTP in the medial prefrontal cortex (PFC) of adult mice. Additionally, TrkB-FL receptor expression is decreased in the PFC of these animals. By contrast, these changes were not present in the hippocampus of PCP-treated mice. Moreover, BDNF levels were not altered in the hippocampus or PFC of PCP-treated mice. Interestingly, these observations are paralleled by impaired performance in PFC-dependent cognitive tests in mice treated with PCP.

CONCLUSIONS

Overall, these data suggest that NMDA-R hypofunction induces dysfunctional BDNF signalling in the PFC, but not in the hippocampus, which may contribute to the PFC-dependent cognitive deficits seen in the subchronic PCP model. Additionally, these data suggest that targeting BDNF signalling may be a mechanism to improve PFC-dependent cognitive dysfunction in schizophrenia.

Modulation of stimulated dopamine release in rat nucleus accumbens shell by GABA in vitro: Effect of sub-chronic phencyclidine pretreatment

Dopamine signaling in nucleus accumbens (NAc) is modulated by γ-aminobutyric acid (GABA), acting through GABA-A and GABA-B receptors: dysregulation of GABAergic control of dopamine function may be important in behavioral deficits in schizophrenia. We investigated the effect of GABA-A (muscimol) and GABA-B (baclofen) receptor agonists on electrically stimulated dopamine release. Furthermore, we explored whether drug-induced changes were disrupted by pretreatment with phencyclidine, which provides a well-validated model of schizophrenia. Using brain slices from female rats, fast-scan cyclic voltammetry was used to measure electrically stimulated dopamine release in NAc shell. Both muscimol and baclofen caused concentration-dependent attenuation of evoked dopamine release: neither effect was changed by dihydro-β-erythroidine, a nicotinic acetylcholine receptor antagonist, or the α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA)-type glutamate receptor antagonist, 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX), precluding indirect mechanisms using these transmitter systems in the GABAergic actions. In slices taken from rats pretreated with phencyclidine, the attenuation of evoked dopamine release by baclofen was abolished, but the attenuation by muscimol was unaffected. Since phencyclidine pretreatment was followed by drug-free washout period of at least a week, the drug was not present during recording. Therefore, disruption of GABA-B modulation of dopamine is due to long-term functional changes resulting from the treatment, rather than transient changes due to the drug's presence at test. This enduring dysregulation of GABA-B modulation of accumbal dopamine release provides a plausible mechanism through which GABA dysfunction influences accumbal dopamine leading to behavioral changes seen in schizophrenia and may provide a route for novel therapeutic strategies to treat the condition.

Phosphodiesterase type 1 inhibition alters medial prefrontal cortical activity during goal-driven behaviour and partially reverses neurophysiological deficits in the rat phencyclidine model of schizophrenia

Positive modulation of cAMP signalling by phosphodiesterase (PDE) inhibitors has recently been explored as a potential target for the reversal of cognitive and behavioural deficits implicating the corticoaccumbal circuit. Previous studies show that PDE type 1 isoform B (PDE1B) inhibition may improve memory function in rodent models; however, the contribution of PDE1B inhibition to impulsivity, attentional and motivational functions as well as its neurophysiological effects have not been investigated. To address this, we recorded single unit activity in medial prefrontal cortex (mPFC) and nucleus accumbens (NAc) in Lister Hooded rats treated with the PDE1B inhibitor Lu AF64386 and tested in the 5-choice serial reaction time task (5-CSRTT). We also asked whether PDE1B inhibition modulates neurophysiological deficits produced by subchronic phencyclidine (PCP) treatment, a rat pharmacological model of schizophrenia. Lu AF64386 significantly affected behavioural parameters consistent with a reduction in goal-directed behaviour, however without affecting accuracy. Additionally, it reduced mPFC neuronal activity. Pre-treatment with PCP did not affect behavioural parameters, however it significantly disrupted overall neuronal firing while increasing phasic responses to reward-predicting cues and disrupting mPFC-NAc cross-talk. The latter two effects were reversed by Lu AF64386. These findings suggest PDE1B inhibition may be beneficial in disorders implicating a dysfunction of the mPFC-NAc network.

Different roles of Rac1 in the acquisition and extinction of methamphetamine-associated contextual memory in the nucleus accumbens

Rationale: Repeated methamphetamine (METH) exposure induces long-term cognitive deficits and pathological drug-associated memory that can be disrupted by manipulating memory reconsolidation and extinction. The nucleus accumbens (NAc) is the key region of the brain reward system and predominantly consists of two subtypes of medium spiny neurons (MSNs) based on the expression of D1 or D2 dopamine receptors (D1-MSNs or D2-MSNs). Spine structural plasticity in the NAc is critical for the acquisition, reconsolidation and extinction of drug-associated memory. However, the molecular mechanisms underlying METH-associated memory and spine remodelling in each type of MSNs in the NAc remain unknown. Here, we explored whether Rac1 in the NAc mediates METH-associated contextual memory and spine remodelling.

Methods: Pharmacological and genetic manipulations of Rac1 were used to investigate its role during the acquisition, reconsolidation and extinction of METH-associated contextual memory. Recombinant adeno-associated viruses expressing mCherry under the control of the dopamine D1 receptor gene promoter (Drd1-mCherry) or dopamine D2 receptor gene promoter (Drd2-mCherry) were used to specifically label D1-MSNs or D2-MSNs.

Results: Using viral-mediated gene transfer, we demonstrated that decreased Rac1 activity was required for the acquisition of METH-associated contextual memory and the METH-induced increase in thin spine density, whereas increased Rac1 signalling was important for the extinction of METH-associated contextual memory and the related elimination of thin spines. Moreover, the increase of dendritic spines was both found in D1-MSNs and D2-MSNs during the acquisition process, but extinction training selectively decreased the spine density in D1-MSNs. Interestingly, Rac1 was responsible for METH-induced spine plasticity in D1-MSNs but not in D2-MSNs. Additionally, we found that microinjection of a Rac1 inhibitor or activator into the NAc was not sufficient to disrupt reconsolidation, and the pharmacological activation of Rac1 in the NAc facilitated the extinction of METH-associated contextual memory. Regarding cognitive memory, decreased Rac1 activity improved the METH-induced impairment in object recognition memory.

Conclusion: Our findings indicate that Rac1 plays opposing roles in the acquisition and extinction of METH-associated contextual memory and reveal the cell-specific role of Rac1 in METH-associated spine remodelling, suggesting that Rac1 is a potential therapeutic target for reducing relapse in METH addiction and remediating METH-induced recognition memory impairment.

Cross-talk between guanidinoacetate neurotoxicity, memory and possible neuroprotective role of creatine

Guanidinoacetate Methyltransferase deficiency is an inborn error of metabolism that results in decreased creatine and increased guanidinoacetate (GAA) levels. Patients present neurological symptoms whose mechanisms are unclear. We investigated the effects of an intrastriatal administration of 10 μM of GAA (0.02 nmol/striatum) on energy metabolism, redox state, inflammation, glutamate homeostasis, and activities/immunocontents of acetylcholinesterase and Na⁺,K⁺-ATPase, as well as on memory acquisition. The neuroprotective role of creatine was also investigated. Male Wistar rats were pretreated with creatine (50 mg/kg) or saline for 7 days underwenting stereotactic surgery. Forty-eight hours after surgery, the animals (then sixty-days-old) were divided into groups: Control, GAA, GAA + Creatine, and Creatine. Experiments were performed 30 min after intrastriatal infusion. GAA decreased SDH, complexes II and IV activities, and ATP levels, but had no effect on mitochondrial mass/membrane potential. Creatine totally prevented SDH and complex II, and partially prevented COX and ATP alterations. GAA increased dichlorofluorescein levels and decreased superoxide dismutase and catalase activities. Creatine only prevented catalase and dichlorofluorescein alterations. GAA increased cytokines, nitrites levels and acetylcholinesterase activity, but not its immunocontent. Creatine prevented such effects, except nitrite levels. GAA decreased glutamate uptake, but had no effect on the immunocontent of its transporters. GAA decreased Na⁺,K⁺-ATPase activity and increased the immunocontent of its α3 subunit. The performance on the novel object recognition task was also impaired. Creatine partially prevented the changes in glutamate uptake and Na⁺,K⁺-ATPase activity, and completely prevented the memory impairment. This study helps to elucidate the protective effects of creatine against the damage caused by GAA.