The interaction of Purkinje cell and Inhibitory Interneuron plasticity during classical conditioning

Abstract

Event Abstract Back to Event The interaction of Purkinje cell and Inhibitory Interneuron plasticity during classical conditioning Decades of experimental work using the the eye-blink conditioning paradigm points to the Cerebellar Cortex (CC) as the locus of the motor conditioning engram. It is in the cerebellar cortex where a memory trace is formed that allows both to associate Conditioning (CS) and Unconditioned Stimulus (US) and to elicit an adaptively timed Conditioned Response (CR). Single cell recordings show that the evolution of the firing rate of a Purkinje (PU) cell during the CS-US period provides a substrate for such learning (Jirenhed et al., 2007). Classical computational theories of the Cerebellum (like the Marr-Albus-Ito model) predicted that the mechanism responsible for motor learning was plasticity at the parallel fiber to PU (pf-PU) synapse. However in vitro and in vivo studies have reported multiple loci of plasticity in the CC among them plasticity at the parallel fibers to Inhibitory Interneuron (pf-II) synapse (Jörntell, & Ekerot, 2002). This site of plasticity is of particular interest because it is inversely related to the synaptic dynamics of the pf-PU synapse: whenever the pf-PU synapse undergoes LTD the pf-II undergoes LTP, and vice versa. Hence, this suggests that the engram is not restricted to the parallel fiber Purkinje cell synapse only. A particular challenge for models that exclusive rely on the plasticity at pf-PU synapses is that they have difficulty to explain results that have shown that increasing CS intensity shortens the latency of already acquired CRs (Svensson et al, 1997). We hypothesize that the inclusion of the active role of the Inhibitory Interneurons (II) in learning can account for this CS intensity effect. We have implemented a biologically constrained model aimed at reproducing the learning process of a single Purkinje cell, in such a way that it will respond to presentations of a CS with an adaptively timed pause. In previous work we demonstrated that having plasticity at the pf-PU was sufficient to achieve this goal provided the cortico-nuclear-olivary negative feedback loop was included (Verschure and Mintz, 2000; Hofstötter et al., 2002). Now, we have extended this model and also included plasticity at the pf-II synapse in order to test its computational implications. Our results confirm that potentiation of the II can account for the CS intensity effect. Our model also predicts that if these two plasticity processes coexist it should be possible to identify different components in the dynamics of the acquisition/extinction processes at the level of single Purkinje cells. Although two components have already been founded in behavioral data (Ivarsson & Svensson, 2000) a confirmation of this prediction could only be obtained by analyzing the responses of single Purkinje cells in such experiments.