The "two-headed" latent inhibition model of schizophrenia: modeling positive and negative symptoms and their treatment

Abstract

Rationale: Latent inhibition (LI), namely, poorer performance on a learning task involving a previously pre-exposed non-reinforced stimulus, is disrupted in the rat by the dopamine (DA) releaser amphetamine which produces and exacerbates psychotic (positive) symptoms, and this is reversed by treatment with typical and atypical antipsychotic drugs (APDs) which on their own potentiate LI. These phenomena are paralleled by disrupted LI in normal amphetamine-treated humans, in high schizotypal humans, and in schizophrenia patients in the acute stages of the disorder, as well as by potentiated LI in normal humans treated with APDs. Consequently, disrupted LI is considered to provide an animal model of positive symptoms of schizophrenia with face, construct and predictive validity.

Objectives: To review most of the rodent data on the neural substrates of LI as well as on the effects of APDs on this phenomenon with an attempt to interpret and integrate these data within the framework of the switching model of LI; to show that there are two distinct LI models, disrupted and abnormally persistent LI; to relate these findings to the clinical condition.

Results: The nucleus accumbens (NAC) and its DA innervation form a crucial component of the neural circuitry of LI, and are involved at the conditioning stage. There is a clear functional differentiation between the NAC shell and core subregions whereby damage to the shell disrupts LI and damage to the core renders LI abnormally persistent under conditions that disrupt LI in normal rats. The effects of shell and core lesions parallel those produced by lesions to the major sources of input to the NAC: entorhinal cortex lesion, like shell lesion, disrupts LI, whereas hippocampal lesion, like core lesion, produces persistent LI with changes in context, and basolateral amygdala (BLA) lesion, like core lesion, produces persistent LI with extended conditioning. Systemically induced blockade of glutamatergic as well as DA transmission produce persistent LI via effects exerted at the conditioning stage, whereas enhancement of DA transmission disrupts LI via effects at the conditioning stage. Serotonergic manipulations can disrupt or potentiate LI via effects at the pre-exposure stage. Both typical and atypical APDs potentiate LI via effects at conditioning whereas atypical APDs in addition disrupt LI via effects at pre-exposure. Schizophrenia patients can exhibit disrupted or normal LI as a function of the state of the disorder (acute versus chronic), as well as persistent LI.

Conclusions: Different drug and lesion manipulations produce two poles of abnormality in LI, namely, disrupted LI under conditions which lead to LI in normal rats, and abnormally persistent LI under conditions which disrupt it in normal rats. Disrupted and persistent LI are differentially responsive to APDs, with the former reversed by both typical and atypical APDs and the latter selectively reversed by atypical APDs. It is suggested that this "two-headed LI model" mimics two extremes of deficient cognitive switching seen in schizophrenia, excessive and retarded switching between associations, mediated by dysfunction of different brain circuitries, and can serve to model positive symptoms of schizophrenia and typical antipsychotic action, as well as negative symptoms of schizophrenia and atypical antipsychotic action.