In vitro classical conditioning of the turtle eyeblink reflex: approaching cellular mechanisms of acquisition.

The classically conditioned eyeblink reflex is the best studied model for understanding the neural mechanisms that underlie learning and memory. Here, data from an in vitro model of the conditioned eyeblink reflex are summarized with the aim of shedding some light on potential cellular mechanisms that may underlie eyeblink classical conditioning. An isolated brainstem-cerebellum preparation from turtles was developed in which to study the synaptic circuitry of pathways involving the cerebellum, red nucleus and brainstem nuclei. A neural correlate of an eyeblink response recorded in the abducens nerve can be conditioned entirely in vitro by pairing trigeminal and auditory nerve stimulation. Conditioned abducens nerve responses (CRs) are not generated or sustained by unpaired stimuli and their long latencies, on the order of hundreds of milliseconds, support the interpretation that the CRs are not unconditioned responses. Ablation experiments show that CRs can be generated in brainstem preparations lacking a cerebellum or the medulla. However, the timing of the CRs are disrupted by removal of the cerebellar circuitry. Thus, a highly reduced in vitro brainstem preparation demonstrates acquisition of CRs but poor timing features. Recent experiments have focused on elucidating cellular mechanisms for CR acquisition in the brainstem blink circuitry. These studies show that NMDA-mediated synaptic mechanisms are required to generate CRs and that the level of conditioning is associated with the upregulation of GluR4-containing AMPA receptors in the abducens motor nuclei. Data from immunocytochemistry and physiological experiments using the calcium/calmodulin-dependent protein kinase II (CaMKII) inhibitor KN-93 suggest that CaMKII does not have a key role in mediating the induction or expression of abducens nerve CRs. It is hypothesized that GluR4-containing AMPA receptors in the abducens motor nuclei are targeted to auditory nerve synapses by an NMDA receptor-dependent process to strengthen the CS input during conditioning which results in the generation of CRs. Future studies will examine the synaptic localization of GluR4 and potential signal transduction pathways involved in in vitro conditioning. Moreover, the role feedback loops through the cerebellum and their role in CR timing will be a key issue to address using this preparation.