Numerous human imaging studies have revealed an absolute or relative metabolic hypofunction within the prefrontal cortex, thalamus and temporal lobes of schizophrenic patients. The former deficit correlates with cognitive deficits and negative symptoms, whereas the latter correlates with positive symptomologies. There is also general consensus that schizophrenia is associated with decreased parvalbumin expression in the prefrontal cortex. Since the drug phencyclidine can induce a psychosis resembling schizophrenia in humans, we have examined whether repeated phencyclidine (PCP) treatment to rats could produce similar metabolic and neurochemical deficits to those occurring in schizophrenia and whether these deficits could be modulated by antipsychotic drugs. We demonstrate here that chronic intermittent exposure to PCP (2.58 mg kg À1 i.p.) elicits a metabolic hypofunction, as demonstrated by reductions in the rates of glucose utilization, within the prefrontal cortex, reticular nucleus of thalamus and auditory system, key structures displaying similar changes in schizophrenia. Moreover, chronic PCP treatment according to this regime also decreases parvalbumin mRNA expression in the rat prefrontal cortex and reticular nucleus of the thalamus. Chronic coadministration of haloperidol (1 mg kg À1 day À1 ) or clozapine (20 mg kg À1 day À1) with PCP did not modulate PCP-induced reductions in metabolic activity in the rat prefrontal cortex, but reversed deficits in the structures of the auditory system. Clozapine, but not haloperidol, reversed PCP-induced decreases in parvalbumin expression in prefrontal cortex GABAergic interneurons, whereas both drugs reversed the deficits in the reticular nucleus of the thalamus. These data provide important new information, which strengthen the validity of chronic PCP as a useful animal model of schizophrenia, when administered according to this protocol. Furthermore, we propose that reversal of PCP-induced reductions in parvalbumin expression in the prefrontal cortex may be a potential marker of atypical antipsychotic activity in relation to amelioration of cognitive deficits and negative symptoms of schizophrenia.

Although our knowledge of the pathophysiology of schizophrenia has increased, treatments for this devastating illness remain inadequate. Here, we critically assess rodent models and behavioural end points used in schizophrenia drug discovery and discuss why these have not led to improved treatments. We provide a perspective on how new models, based on recent advances in the understanding of the genetics and neural circuitry underlying schizophrenia, can bridge the translational gap and lead to the development of more effective drugs. We conclude that previous serendipitous approaches should be replaced with rational strategies for drug discovery in integrated preclinical and clinical programmes. Validation of drug targets in disease-based models that are integrated with translationally relevant end point assessments will reduce the current attrition rate in schizophrenia drug discovery and ultimately lead to therapies that tackle the disease process.

No abstract