Effects of Aripiprazole and Haloperidol on Mesolimbic System Functioning

The advent of neuroimaging has opened a new window into the human brain. Here, we propose to investigate if the activation signature of a traditional antipsychotic agent is dissociable from a new drug with a unique pharmacological profile that differs from both typical and the more recent atypical antipsychotics. The introduction of aripiprazole is of interest because while clinically it has all the features of an atypical antipsychotic (antipsychotic effect with very low motor side effects) it differs from all other antipsychotics in that it is a partial D2 receptor agonist.

Antipsychotic drugs can be classified as either typical or atypical. Typical antipsychotics, like haloperidol, are dopamine antagonists which are clinically efficacious when they occupy 60-70% of the striatal dopamine D2 receptors, and extra pyramidal side (EPS) effects begin to show when D2 receptor occupancy (D2RO) is >80%. Atypicals, like the typicals, are efficacious; however, they are unique insofar as they have a low affinity for dopamine receptors, target other neurotransmitter systems (in particular the serotonergic system), and show low motor side effects. While most atypicals are dopamine antagonists, a new class of antipsychotics characterized as partial agonists have emerged. Aripiprazole is one of these novel partial agonists. It has the ability to act as an antagonist in the presence of high concentrations of dopamine and as an agonist in low concentrations. Animal models can be used to predict both the clinical efficacy of antipsychotics as well as to predict their propensity to induce EPS. What follows is a comparison of the behavior of the typical antagonist haloperidol and the novel partial antagonist aripiprazole in these models. Both haloperidol and aripiprazole show dose-dependent striatal D2 receptor occupancy (D2RO) (Natesan et al., 2006). In two measures of clinical efficacy, conditioned avoidance response (CAR) and amphetamine induced locomotion (AIL), haloperidol and aripiprazole behave differently. Haloperidol's ED50 for the inhibition of CAR and AIL corresponds to \~60% D2RO. Aripiprazole, however, shows dissociation between function and occupancy. While aripiprazole is similar to haloperidol in that it blocks AIL at \~60% D2RO, it differs in that a 23-fold higher dose, corresponding to an 86% D2RO is required to inhibit CAR. It has been suggested by Natesan et al., 2006 that this functional dissociation occurs as a result of aripiprazole's partial agonist property. If aripiprazole has approximately 20% in vivo intrinsic activity, this could raise the threshold for CAR inhibition from the \~60% D2RO seen in haloperidol to the \~85% D2RO seen in aripiprazole. A similar result to that in the CAR model is found in another measure of clinical efficacy - FOS expression in the shell of the nucleus accumbens (NAcc). While haloperidol induced FOS at a dose corresponding to \~60% D2RO, an \~80% D2RO was needed for aripiprazole. Despite the >80% D2RO required for aripiprazole to be efficacious in the CAR model, and induce FOS in the NAcc shell, catalepsy was not present. This is contrary to what is seen in the typicals, including haloperidol. It is also surprising, as aripiprazole D2ROs >80% induce FOS in the core of the NAcc and the dorsolateral striatum (a marker for catalepsy and extrapyramidal symptoms).

It is well established that animals treated with low doses of antipsychotic drugs fail to acquire or perform avoidance responses to a conditioned stimulus (CS) that is associated to an aversive unconditioned stimulus (US), whereas their escape responses to that stimulus are not affected in aversive avoidance paradigms. This selective disruption of avoidance, but not escape responses, is a characteristic pharmacologic effect of all antipsychotics, including newer atypical ones. This feature has been effectively used to differentiate antipsychotic medication from other classes of psychotropic drugs to predict their clinical potencies and to identify potential drugs to be used as antipsychotics.

More appetitive paradigms where animals have to work for rewarding stimuli (food, water, sex, brain stimulation) show a dramatic decrement in performance/pursuit when administered antipsychotics. This deficit can be reversed with dopamine agonists such as amphetamine and thus suggesting symmetrical effects of dopamine agonists and antagonists in these kinds of paradigms. While typical antipsychotics like haloperidol reduce impact of both appetitive and aversive CSs but not the US at doses producing >60% D2 occupancy, it has been reported that aripiprazole was effective at doses that gave rise to occupancies of >85%. This has to be examined in humans and functional Magnetic Resonance Imaging (fMRI) might be a tool specific and sensitive enough.

fMRI is now a standard tool in human brain imaging and provides a good spatial resolution at the millimeter level with a second level temporal resolution. Event-related fMRI, wherein brain responses are recorded in response to specific events, are optimally suited for studying the neurobiology of phenomena such as classical conditioning in humans. In a previous event-related fMRI study from our group, using Pavlovian aversive conditioning, the ventral striatum was reliably and directly activated in three experiments. These results have been replicated in healthy controls, whereas patients with schizophrenia have shown a different activation pattern. The ventral striatum was activated in the anticipation of aversive events regardless of whether there was an opportunity to avoid the aversive stimulus (similar to the animal models described above) or not. Thus, our data suggested that the ventral striatum, a crucial element of the dopaminergic mesolimbic "reward" system, is directly activated in anticipation of aversive stimuli. Similar to other conditioning studies we also found robust activations of the anterior insula and anterior cingulate, both parts of the greater limbic cortex and projects to the ventral striatum. The anterior insula has been proposed to play a role in processing emotional relevant contexts such as disgust and pain whereas the anterior cingulate is involved in assessing motivational content of internal and external stimuli and regulating context dependent behavior. Both structures have extensive connections with the amygdala which is the substrate most commonly associated with fear.

The field of pharmacological fMRI is emerging and some early studies have demonstrated the proof of principle that drugs can influence activations. The dopamine system has been the focus of several fMRI studies, most probably as a result of its innervations of key cortical and sub-cortical regions implicated in motor and neurocognitive functions, in addition to its implication in a range of psychiatric and neurological conditions. Early pharmacological fMRI studies suggest that this is a tool sensitive enough to pick up pharmacologically modulated activations in the brain. Pharmacological fMRI has a potential for testing transmitter models of disorder, predicting treatment response and supporting the development of novel compounds in neuropsychiatry.

The main objective of this study is to investigate if the activation signature of a typical antipsychotic agent is dissociable from a new drug with a pharmacological profile that differs from both typical and the atypical antipsychotics since it is a potent partial D2 agonist. For this purpose we will employ fMRI paradigms based on findings from previous research on animals and humans described above.