Testing a computational model of the role of cortical circuits in auditory hallucinations in people with psychosis

People with lived experience of psychosis (PWLEP) identified auditory hallucinations (AH) as a top research priority. Computational modelling of functional neuroimaging data in PWLEP with chronic AH indicates that disinhibition of pyramidal neurons could underlie AH. However, this has not been tested at first presentation of AH. We will address this key gap in knowledge. Multiple lines of evidence indicate that reduced activity of parvalbumin-expressing interneurons (PVI) contributes to the disinhibition. We will test this mechanism using a drug that augments PVI activity in PWLEP and AH. We will probe mechanism further using chemogenetic mouse models to alter firing in cortical circuits in a cell-type and regionally specific manner and measure the effect on neural oscillatory activity relevant to the studies in PWLEP, and test effects of drugs to target oscillatory abnormalities during neurodevelopment and in a genetic model. This will determine the contribution of pyramidal, PVI and other neuronal sub-types to the loss of inhibition and other alterations seen in PWLEP and the potential for preventive interventions. These data in PWLEP and mice will be used to refine the computational model of AH, and test whether a network model can explain the lived experience of AH.