We study neuronal hyperactivity from both its physiological and its pathological sides. Using repeated optogenetic stimulation (REPOPS) that models electroconvulsive therapy, we showed that repetitive neuronal activation rewrites the state of neurons in the long term via nuclear reprogramming, together with antidepressant-like behavioural changes (Nat Commun 2026). We now study why repeated neuronal hyperactivity has antidepressant effects, and how neuronal hyperactivity contributes to the pathophysiology of psychiatric disorders, using transcriptomics and epigenomics (supported by JST FOREST and JSPS KAKENHI).
We have consistently analysed public gene-expression data from human post-mortem brains and disease models across studies, to find molecular changes shared by psychiatric and neurological disorders: immature-like gene expression in the hippocampus and prefrontal cortex of patients with alcoholism (Sci Rep 2017), gene-expression changes induced by neuronal hyperexcitation that are shared by multiple neuropsychiatric disorders (Commun Biol 2019), and gene expression linked to brain pH across disorders (Front Psychiatry 2023). Along this line we develop methods that score gene expression against external reference signatures; RFind-sc extends this approach to single cells (bioRxiv 2026; see Software & Data).
We are interested in how information is encoded by populations of neurons, especially in the dentate gyrus of both normal and disease-model mice. With calcium imaging in freely moving mice and machine learning, we showed that multiple types of navigational information — position, speed, and direction of motion — are independently encoded in the population activity of dentate gyrus neurons, and that decoding of some of this information is selectively reduced in αCaMKII heterozygous knockout mice, which show deficits in spatial memory (PNAS 2022).