Ramping time scale synaptic plasticity in the in vivo cerebellum

University Of Chicago, Chicago IL

Investigators

Abstract

Long-term depression (LTD) at cerebellar parallel fiber (PF) to Purkinje cell (PC) synapses has been studied in great detail, because a) it constitutes a major form of cerebellar plasticity that may underlie various forms of motor adaptation and learning, b) it may initiate synaptic pruning and thus shaping of essential connectivity, and c) it provides the perhaps best example of supervised learning due to is dependence on co-activity of the climbing fiber (CF) input. According to the theories of Marr, Albus and Ito, the instructive CF signal in some conditions presents an error that for its correction requires a change in PF synaptic input weights. LTD has been first described in vivo in decerebrate rabbits. Most subsequent studies have been performed in slices, where whole-cell patch-clamp recordings from PCs enabled a characterization of LTD in molecular detail. The goal of this study is to assess the conditions under which LTD is evoked in the cerebellum of intact mice. The motivation to do so is twofold, one is more general, and one specific. First, we have recently determined LTD induction rules in slice, but under more realistic (physiological) recordings conditions, including the use of an accurate extracellular ionic milieu ([Ca2+]o =1.2mM; [Mg2+]o =1.0mM) and near-physiological bath temperature. These ‘updates’ changed LTD rules substantially, and we became aware that recordings in vivo might reveal further differences to the current understanding. Second, a new physiological phenomenon has been observed in vivo that we predict will have a substantial impact on LTD: ramping in the activity of granule cell / PC activity over hundreds of milliseconds up to seconds, which was described during motor planning and in learned behaviors triggered by sensory cues. As a result of prolonged calcium elevation, we predict a shift in plasticity threshold values via a ‘leaky integrator’ mechanism. We will use two-photon measurements of GCaMP7b- encoded calcium transients in PC dendrites (PC-specific promoter L7) to compare plasticity outcomes between classic PF single-pulse or burst / CF protocols and new protocols, in which prolonged PF activity ramps are paired with CF stimulation (aim 1). Here, the PF and CF inputs are electrically stimulated, and plasticity is determined by a dendritic calcium signal measure. We will pair this approach with patch-clamp recordings from PCs in slices, where we mimic ramps by PF stimulation (aim 2). This approach is important to obtain an electrophysiological LTD readout from the somatic patch electrode that complements the imaging results obtained in vivo. Prolonged PF ramps lasting ³ 1s will be ‘interrupted’ by spontaneous CF-evoked complex spikes. There is no current hypothesis of the purpose of these complex spikes, but we do hypothesize that their firing erases prior calcium build-up that results from ramping (‘windshield wiper effect’). Using both in vivo and in vitro imaging approaches, we will study how interfering complex spikes affect PF ramps and plasticity outcomes (aim 3). This study will be the first to assess cerebellar LTD in intact mice with consideration of activity ramping, a motif in temporal input structures that is typical for learned PF input patterns.