A guiding assumption in neurobiology has been that storage of information in the brain involves persistent, use-dependent alterations in neuronal responsivity. In recent years, considerable effort has been expended to investigate examples of this phenomenon. What has emerged is a picture of neuronal information storage in which many diverse and overlapping mechanisms co-exist to endow neurons with varying computational and mnemonic properties. As such, it is crucial to investigate a broad range of neuronal mechanisms for information storage. One very useful, but largely unexploited model system for this endeavor is cerebellar long- term depression (LTD). In cerebellar LTD, co-activation of parallel fiber and climbing fiber inputs to a Purkinje neuron (PN) induces a persistent depression of the parallel fiber-PN synapse. This phenomenon is thought to underlie several forms of motor learning. Using cultured mouse PNs, a preparation has been developed in which iontophoretic glutamate pulses and PN depolarization may be substituted for parallel fiber and climbing fiber stimulation, respectively. This depression is measured as a reduction of the glutamate-induced current as measured with a perforated-patch electrode attached to the PN soma. This preparation has several advantages that further the analysis of information storage in the single neuron and has allowed for experiments that have shed light on the mechanisms of LTD induction, specifically the findings that activation of both AMPA and metabotropic receptors is necessary for this process, that activation of protein kinase C is required, and that LTD is expressed specifically as a decrease in AMPA receptor-mediated current. I propose four major lines of work to continue to investigate the cellular mechanisms of cerebellar LTD. First, are some simple parametric studies that will seek to determine some of the basic phenomenology of LTD induction that remains unexamined (additivity, influence of prior activity, induction thresholds). Second is a further investigation the role of second messenger systems in LTD induction, particularly those activated by free fatty acids and cyclic nucleotides. A third line of work will seek to elucidate the effects on LTD of extrinsic noradrenergic modulatory inputs. fourth is the use of ultra- high resolution optical imaging techniques to determine the fine structure of Ca gradients in the PN as induced by depolarization and glutamate application. This work is specifically focused upon the role of the dendritic spines and the different spatial consequences of Ca influx via voltage-gated channels, ligand-gated channels, and liberation from intracellular stores. Understanding these processes may prove critical to determining how LTD is induced in a manner specific to a single dendritic input location, a property that is critical in conferring computational power upon the PN which receives ~200,000 separate synaptic contacts from parallel fibers. These investigations of basic neuronal mechanisms have clinical relevance not only to cerebellar motor disorders, but to disorders of learning, memory and information storage generally.
