August 22, 2014. The SIRS 2014 plenary session on Monday, April 7, titled "Behavioral and Imaging Translational Paradigms in Drug Development," provided a dialogue between preclinical and clinical schizophrenia researchers. The well-attended session focused on three aspects of schizophrenia symptoms: psychosis, motivation, and social cognition. For each topic, audience members would hear from a pair of researchers, one preclinical and one clinical, explained session Chair Holly Moore, Columbia University, New York City, in her introduction.
Anthony Grace, University of Pittsburgh, Pennsylvania, reviewed data suggesting that in both schizophrenia and the MAM model of the illness there is hyperactivity in the ventral hippocampus, perhaps resulting from decreased interneuron function (see SRF related conference report). This hyperactivity may result in dopamine system hyperactivity, he said. The result would be an activation of the system in response to both salient and non-salient stimuli, making it difficult for patients to distinguish relevant from irrelevant information, a process termed aberrant salience.
Inactivation of the ventral hippocampus in the MAM model restores general dopamine system function, suggesting that this circuit is a potential drug target that would be more effective than the canonical modulation of dopamine receptor stimulation. However, a number of such drugs that showed promise in preclinical studies have failed in schizophrenia trials, said Grace, hypothesizing that the exposure to years of antipsychotic drugs in patients may be to blame. This treatment is likely to alter the system substantially, he said, resulting in dopamine supersensitivity rather than just increased dopamine neuron firing.
Grace's clinical partner was Graham Murray, University of Cambridge, UK, who discussed how associative learning, the process by which one learns what features of the environment are connected and important, is abnormal in schizophrenia. Theories of associative learning center around prediction error, the mismatch between expectation and outcome. Murray reviewed evidence of midbrain, striatal, and cortical prediction abnormalities in psychosis. Patients show overactivation to expected reward receipt but underactivation in response to unexpected reward, he said. Evidence from some studies suggests that these signals correlate with the severity of psychotic signals, Murray added.
When healthy subjects are given methamphetamine, they develop mild psychotic disturbances as well as impaired striatal reward prediction error signaling and frontal cortical signaling of expected value. The amount of disruption predicts the severity of psychosis symptoms, suggesting that dopaminergic reward processing abnormalities underlie the pathogenesis of psychosis, concluded Murray.
In the next section of the plenary, Jim Gold of the University of Maryland, Baltimore, also discussed the reward system, focusing on negative symptoms in schizophrenia as a target for translation research. There are many aspects of the reward system that could be awry in the illness, so it's challenging to know just which ones are affected, he stated. In general, behavioral studies have shown impairments in rapid feedback-driven learning and in translating value representation to action, and also revealed an exaggerated cost of effort. The result is that "the costs of action loom large, but the benefits are imprecise" in schizophrenia, Gold concluded.
The major nodes in the reward system can be modeled in animals, but there are a number of challenges, he continued. For example, it's only possible to model select features of schizophrenia rather than the illness as a whole. The aberrant salience thought to underlie psychosis cannot be simultaneously modeled with the reduced salience characteristic of negative symptoms. Additional challenges include the fact that humans may fail at a task for other reasons than animals do, and the fact that reinforcement learning tasks involve other cognitive systems that may be impaired in schizophrenia, said Gold.
Pierre Trifilieff, University of Bordeaux, France, described the role of dopamine in the mesoaccumbens pathway—originating in the ventral tegmental area and projecting to the ventral striatum and prefrontal cortex—that mediates motivation. He focused on the dopamine D2 receptor in this pathway, which PET studies have shown exhibits altered availability in schizophrenia and other psychiatric disorders, suggesting it may underlie the motivational deficits of schizophrenia. For example, levels of trait motivation and striatal D2 receptors are strongly correlated in attention deficit hyperactivity disorder. In fact, Trifilieff and colleagues showed that overexpression of the D2 receptor in the ventral striatum improved motivation in the mouse (Trifilieff et al., 2013).
Michael Green, University of California, Los Angeles, discussed issues surrounding the translation of social cognition and social neuroscience for schizophrenia clinical trials. Social cognition is closely linked to functional outcome in the illness, but better measures of social neuroscience are needed, he said. Using the example of light, which is both a wave and a particle, Green explained that social cognition can be identified by both domains and levels, and both aspects need to be kept in mind simultaneously when selecting measures for clinical trials. Two domains have been identified: nonverbal social cues and high-level inferences of social information. Both within and across domains, social cognition also varies from low to high levels. The lower-level aspects of social cognition (the identification of social cues and mirroring/experience sharing) interact and work together with higher levels that require more thought (evoked when cues don't match) but are represented in different areas of the brain.
Green discussed five paradigms that measure social cognition, highlighting two lessons learned: just because a test is expected to be impaired in schizophrenia doesn't mean it will be, and that tests that work well in college students don't necessarily translate well in schizophrenia patients. He concluded with a discussion of empathic accuracy as an example of a new social processing construct for schizophrenia research.
In the final talk of the plenary, Karen Bales, University of California, Davis, discussed behavioral tests that can be used in animals to probe social behavior relevant to schizophrenia. Social symptoms of the illness can be broken down into a number of more measurable constructs such as anxiety and accurate assessment of social cues such as empathy and facial expression recognition. She called for the use of well-validated paradigms with construct validity and reproducibility in the development of drugs targeting social deficits. Two examples of these types of constructs are social interaction and eye gaze, both nominated by the Cognitive Neuroscience Treatment Research to Improve Cognition in Schizophrenia (CNTRICS) initiative, she said.
Bales also discussed the use of the peptide hormone oxytocin as a positive control for social paradigms. For example, chronic intranasal oxytocin treatment improves partner preference formation with a partner in voles (Bales et al, 2013). However, while titi monkeys show wide expression of the oxytocin receptor throughout the brain, rhesus monkeys have very limited expression. Human receptor levels have not been measured, she noted, suggesting the need for further study of the oxytocin system as well as the effects of chronic administration of the hormone.—Allison A. Curley.