The cerebellar neural circuitry is involved in the classical conditioning in both mammals and teleosts. Cerebellum lesions or drug-mediated silencing of cerebellum in rodents or goldfish result in impairments in classical fear conditioning, in which repeating the conditioned stimulus (CS, e.g. light) and the unconditioned fear stimulus (US, e.g. electric shock) leads to bradycardia and freezing in response to the CS. We adopted delayed conditioning paradigm with a light-off as CS and an electric shock as US in zebrafish. We found that zebrafish around 20 days post-fertilization or older were capable to acquire the conditioned bradycardia response. However it is not clear which components in the cerebellar neural circuitry are involved in the classical fear conditioning.
To investigate the involvement of the cerebellum in the classical conditioning in zebrafish, we performed the inhibition of granule cell function. We used a granule cell-specific Gal4 line and an effector line UAS:BoTx-GFP to express botulinum toxin in the granule cells. In the control larvae, the CS-dependent bradycardia immediately recovered after the conditioning. When the activity of the granule cells in the valvula cerebelli and corpus cerebelli (the rostral and major lobes of the zebrafish cerebellum) was silenced, the resultant larvae were capable to acquire conditioned bradycardia. However they displayed a prolonged bradycardia response. The data suggest that the cerebellum is involved in the fear conditioning and at least a portion of the granule cells control the conditioned body escape response. The granule cells function to determine the duration of the CS response or timing of the recovery. To monitor neuronal activities during the classical fear conditioning, we expressed a Ca indicator GCaMP7a by using a combination of HuC:Gal4 and UAS:GCaMP7a lines. We detected CS-dependent change in neuronal activities in the corpus cerebelli: some of the cerebellar neurons quickly responded to the CS, whereas some neurons were slowly activated after the CS. We also found cerebellar neurons whose activity was repressed during the conditioning. Our findings suggest that these cerebellar neurons may be involved in different processes of the fear conditioning. We shall report our progress in understanding the roles of individual component of the cerebellar neural circuits in the classical fear conditioning.