Research Summary
Sascha du Lac is using a range of scientific techniques to understand how experience alters neural circuits that are involved in learning.
Our research is focused on the neuronal mechanisms of learning. We study a simple form of motor learning in a well-defined behavior, the vestibulo-ocular reflex (VOR), which generates eye movements that enable clear vision during self-motion. Adaptive changes in the performance of the VOR (i.e., motor learning) occur whenever vision is compromised persistently during head movements. The neural circuitry for the VOR is sufficiently well understood to enable us to link specific cellular and molecular mechanisms with their consequences for oculomotor performance. Using a variety of techniques—including mouse molecular genetic techniques, electrophysiology in vivo and in vitro, and quantitative behavioral analyses—our laboratory studies how experience-dependent changes in intrinsic neuronal excitability and synaptic transmission operate within a defined neural circuit to mediate motor learning.
Motor Learning over Multiple Timescales
The advantage of studying learning in eye movements is that behavioral performance can be manipulated and measured with quantitative precision. We study the VOR by rotating mice on a turntable and measuring eye movements with an infrared video camera. The performance of the VOR is excellent in mice; in the presence of vision, eye movements compensate perfectly for head motion. VOR performance is quantified as the ratio of eye and head speed and is near unity across a wide range of stimulus conditions. Learning is expressed as adaptive increases or decreases in VOR gain. To induce learning, mice are challenged with discrepant motion of the head and visual surround.
Over the past several years, we and others have established that the VOR in mice is plastic over multiple timescales and that there are short-term, intermediate-term, and long-term forms of motor memory. For example, brief exposure to discrepant visual and vestibular experience induces gain changes that decay over the course of an hour, while multiple days of experience produce persistent, consolidated gain changes. We are exploring the mechanisms that underlie motor learning over different timescales with electrophysiological and behavioral analyses in wild-type and genetically modified mice.
Candidate Cellular Mechanisms of Learning: Plasticity of Intrinsic Neuronal Excitability
Although experience-dependent changes in synapses are widely assumed to underlie learning and memory, plasticity of intrinsic neuronal excitability is likely to play an important role. Ion channels that govern the relationship between synaptic currents and postsynaptic firing can be regulated dynamically by phosphorylation, membrane trafficking, and gene expression. Recording from brain slices, we have discovered a novel form of intrinsic plasticity in neurons in brainstem vestibular nucleus neurons that mediate the VOR. Vestibular nucleus neurons fire spontaneously at high rates in vivo and in vitro. Brief periods of synaptic inhibition or membrane hyperpolarization produce persistent increases in both spontaneous firing rate and the gain of vestibular neuronal firing responses to intracellularly injected current. This firing rate potentiation differs from previously described forms of cellular plasticity in a number of ways: it is induced by decreases in intracellular calcium and is expressed as reductions in BK-type calcium-dependent potassium currents.
Firing rate potentiation involves a novel twist on the activity dependence of calcium/calmodulin-dependent protein kinase II (CaMKII). In previous studies, CaMKII has been described as a biochemical switch that is turned on by the brief increases in intracellular calcium that trigger synaptic plasticity and that is maintained in the active state by autophosphorylation. In contrast, in spontaneously firing vestibular nucleus neurons, CaMKII is switched on at baseline and appears to be switched off by the decreases in intracellular calcium that trigger firing rate potentiation. We are examining both the mechanisms that regulate CaMKII activity in vestibular nucleus neurons and the downstream targets of CaMKII.
Identifying and Manipulating Cells Involved in Learning
To understand the consequences of firing rate potentiation or other forms of cellular plasticity on the VOR, we must identify and manipulate the neurons that are critical for learning and memory storage. The cerebellum is required for all forms of motor learning. Purkinje cells, the only output neurons of the cerebellum, exert their effects on behavioral performance and learning via inhibitory synapses onto neurons in the deep cerebellar and vestibular nuclei. Current evidence indicates that Purkinje cells are required to induce motor learning but that motor memories are stored in the deep nuclei and vestibular nuclei. To identify the subset of vestibular nucleus neurons that are responsible for VOR plasticity, we used the Purkinje cell selective promoter L7 to generate a line of transgenic mice that express green fluorescent protein (GFP) in Purkinje cell dendrites, axons, and synaptic terminals. Unexpectedly, we found that a very small subset of vestibular nucleus neurons—1 percent of the total population—are completely surrounded by Purkinje cell terminals. These cerebellar target neurons (CTNs) have unusual intrinsic electrophysiological properties, including high spontaneous firing rates in vitro and exceptionally powerful postinhibitory rebound firing, that are unique within the vestibular nucleus but similar to functionally equivalent neurons in the deep cerebellar nuclei.
Because CTNs are critical for motor learning and memory storage in the VOR, we are investing considerable effort in identifying mechanisms of plasticity in CTNs as well as experience-dependent changes in their intrinsic and synaptic properties after the induction of different forms of motor learning. To establish firm links between particular cellular mechanisms of plasticity and their consequences for learning and memory storage requires manipulating specific mechanisms in the relevant population of neurons in the behaving animal. Consequently, we are searching for genes expressed selectively in cerebellar target neurons that can be used to manipulate signaling pathways involved in firing rate potentiation and other forms of cellular plasticity.
