Christian Hansel - US grants

DESCRIPTION (provided by applicant): Experience-dependent changes in synaptic weight-such as in long-term potentiation (LTP) and long-term depression (LTD)-are at the core of modern theories on memory formation. While LTP is considered to be the main cellular learning correlate in most neural circuits, classic Marr-Albus-Ito theories suggest that, in contrast, cerebellar motor learning is mediated by LTD at parallel fiber (PF) synapses onto Purkinje cells, and a subsequent reduction of the inhibitory Purkinje cell output. Postsynaptic PF-LTP was only recently described (Lev-Ram et al., 2002) and has been suggested to provide a reversal mechanism for LTD (J¿rntell and Hansel, 2006). During the past funding period we demonstrated, however, that potentiation mechanisms play a more active role in cerebellar learning than anticipated. We found that mice with a Purkinje cell-specific knockout of phosphatase PP2B (L7-PP2B)-which does not affect LTD, but prevents LTP and intrinsic plasticity (non-synaptic potentiation)-show impaired cerebellar motor learning (Schonewille et al., 2010). LTP has been described in some detail and its induction was shown to require moderate calcium transients and activation of phosphatases 1, 2A and 2B (Coesmans et al., 2004; Belmeguenai and Hansel, 2005). Intrinsic plasticity, in contrast, remains a poorly understood sibling of LTP and LTD. It has been demonstrated that eyeblink conditioning in rabbits is associated with enhanced Purkinje cell excitability that may result from a modulation of A-type K currents (Schreurs et al., 1998). Moreover, it has been shown that PF tetanization causes changes in Purkinje cell receptive field size (J¿rntell and Ekerot, 2002) that might-as we know now-well result from intrinsic excitability increases that can be co-induced with LTP (Belmeguenai et al., 2010). Finally, genetic blockade of both potentiation mechanisms in L7-PP2B mice impairs motor learning (see above). These studies show that Purkinje cell intrinsic plasticity might provide a crucial component of a cerebellar memory engram. The type of intrinsic plasticity studied here requires-just like LTP-phosphatase activation, and is mediated by a down-regulation of SK2-type K channels, which causes an increase in Purkinje cell spike firing (Belmeguenai et al., 2010; Hosy et al., 2011). Moreover, intrinsic plasticity enhances spine calcium transients and prevents subsequent LTP induction (Belmeguenai et al., 2010). In addition, intrinsic plasticity amplifies dendritic signals in a compartment-specific manner, suggesting that excitability changes can remain locally restricted (Ohtsuki et al., 2012). In this project, we will study how intrinsic and synaptic plasticity may complement each other in cerebellar learning and in generating a memory engram. We will test the hypothesis that a) intrinsic plasticity alters the instructive CF signal that controls the LTD / LTP balance, and thatb) it shortens spike pauses that follow bursts, thus modulating the Purkinje cell output. We will examine motor control and learning in SK2 knockout mice and will use patch-clamp recordings to study intrinsic plasticity properties in vivo. Our goal is to develop a novel theory of cerebellr learning that integrates features of both synaptic and intrinsic plasticity.

DESCRIPTION (provided by applicant): Plasticity in brain motor systems allows fine adjustments of motor coordination that can develop over long periods of time, but can also occur within minutes. Rapid motor learning rests on synaptic plasticity at excitatory synaptic inputs to cerebellar Purkinje cells. Synaptic gain changes such as long-term potentiation (LTP) and long-term depression (LTD) at parallel fiber (PF) and climbing fiber (CF) synapses onto Purkinje cells provide a cellular basis for cerebellar motor learning. Alcohol is known to interfere with synaptic transmission and plasticity, but acute alcohol effects on cerebellar synaptic plasticity have not been studied so far, although it is well known that acute consequences of alcohol consumption include the impairment of motor coordination and the ability to fine-tune movements. Here, we propose to examine the effects of acute ethanol application on cerebellar synaptic plasticity in rat brain slices using whole-cell patch-clamp recordings as well as microfluorometric imaging techniques. Aside from our preliminary experiments performed in preparation for this application, we have no previous experience in alcohol-related research. However, we have characterized cellular mechanisms underlying cerebellar synaptic plasticity (e.g. Hansel et al., 2001;Jvrntell and Hansel, 2006). This application for financial support from the National Institute on Alcohol Abuse and Alcoholism (NIAAA) is motivated by our hope that we now have the required experience and tools at hand to study how ethanol interferes with cerebellar learning mechanisms. Our first steps into the field of alcohol research will be facilitated by a collaboration with the laboratory of Prof. C.F. Valenzuela (University of New Mexico at Albuquerque, USA), who has long-standing experience in this field. Here, we suggest three specific aims: a) to examine acute ethanol effects on cerebellar synaptic transmission and synaptic plasticity, b) to examine ethanol effects on NMDA receptors, metabotropic glutamate receptors and voltage-dependent calcium channels in Purkinje cells and c) to use calcium imaging techniques, in combination with somato-dendritic patch-clamp techniques, to monitor ethanol effects on the spatio-temporal map of dendritic calcium spikes in Purkinje cell dendrites. These dendritic calcium spikes, which are crucial for cerebellar plasticity, are likely triggered by voltage-gated calcium currents and NMDA receptors. Both have been described as ethanol targets in other types of neurons, and therefore dendritic calcium spikes might be affected by ethanol as well. Our proposal has an innovative and exploratory character, as we plan to use a novel combination of imaging (ultra high-speed calcium imaging) and patch-clamp techniques (triple patching with one somatic and two dendritic locations) to be able to characterize dendritic calcium spike activity, which has not been done so far in alcohol research. We believe that this type of research is important to better understand the cellular basis of acute effects of alcohol consumption. One of these is impaired motor coordination, which contributes to the high rate of deaths caused by alcohol-related traffic accidents. PUBLIC HEALTH RELEVANCE: The cerebellum is a brain structure involved in the fine-adjustment of motor coordination and in motor learning. These processes can be impaired as a consequence of acute alcohol consumption, and yet acute alcohol effects on cerebellar synaptic transmission and on plasticity within cerebellar networks have not been studied so far. Here, we propose to examine alcohol effects on forms of cerebellar synaptic plasticity as well as the cellular causes for such effects using electrophysiological recording techniques and fluorometric calcium imaging techniques in rat brain slices.

Project Summary One of the hallmark features of our brains is their enormous capacity to learn and to adapt to changing environments. Theories of memory storage in neural circuits largely focus on activity-dependent changes in synaptic weights ? e.g. long-term potentiation (LTP) ? as plausible learning correlates. But does our synapto- centric view of the cellular events underlying learning really capture the essence of memory engrams? Recent studies on the formation of engrams, or ?mnemic traces? (a concept introduced by Richard Semon in 1904), suggest that the ultimate step in memory engram participation is the suprathreshold activation of neurons. We thus propose a complementary, neurocentric view, in which the participants in a functionally active engram are at least partly determined by cell-autonomous regulation of the intrinsic excitability of individual neurons. In this view, synaptic plasticity controls the formation of reciprocal connectivity patterns within and between engrams, and thus remains an important factor in circuit plasticity and learning. Recent reports indicate that a substantial percentage of pyramidal cells do not engage (remain silent) or are extremely unreliable when these neural circuits are activated, even in primary sensory cortices upon presentation of appropriate stimuli. Here, we will make use of this phenomenon to provide a proof-of-principle demonstration of a role of changes in membrane excitability (?intrinsic plasticity?) in engram formation. We will test the hypothesis that intrinsic plasticity activates previously silent (?dormant?) or unreliable neurons and integrates them into reliable engrams, thus providing a mechanism to dynamically regulate engram composition. We propose that activation of dormant or unreliable neurons constitutes a memory trace in cortical circuits (?intrinsic theory? of memory), by enhancing the capacity for input pattern representation, by increasing the engram activation probability, or by promoting engram stability. This hypothesis will be tested using whole-cell patch-clamp recordings from L2/3 pyramidal cells in the primary somatosensory cortex (S1; barrel cortex) of awake mice, which will be paired with two-photon imaging of GCaMP6s-encoded population activity. We plan to enhance intrinsic excitability by two methods: a) repeated injection of depolarizing currents through the patch pipette (non-synaptic activation; test for the intrinsic nature of this type of plasticity when combined with blockade of synaptic transmission; note that ?intrinsic? refers to the expression phase, but that under physiological conditions synaptic activity will be needed for induction), or b) deflection of select groups of whiskers at active whisking frequency (10-20Hz). We will not only monitor intrinsic excitability, but will test whether neurons become responsive to whisker stimulation (activation of dormant neurons). To assess neuronal integration into engrams, we will image activity patterns in populations of 100-200 neurons. Finally, we will examine whether cholinergic signaling ? through downregulation of SK2- type K+ channels ? facilitates intrinsic plasticity. The R21 mechanism is appropriate, because we are at an early stage of exploring and developing critical tests for the intrinsic hypothesis of learning presented here.

The architecture of brain circuits plays an essential role in understanding their function. The discovery of principal concepts in physiology and neurocomputation critically depends on the availability of precise morphological knowledge. A prominent example supporting this notion is the cerebellum, a brain circuit that is involved in motor (and non-motor) control and adaptation. At the center of cerebellar circuits are Purkinje cells, projection neurons that receive thousands of excitatory synapses, which convey sensory information needed for proper motor control. A hallmark feature of these unique neurons is that they have a massive dendrite that is innervated, in the adult brain, by one climbing fiber input. The climbing fiber plays a crucial role in theories of cerebellar supervised learning, and within this conceptual framework provides an instructive signal (here an error signal) guiding plasticity at parallel fiber synapses. More recent studies have, in addition, pointed out roles in reward signaling and noted that climbing fiber-evoked complex spikes occur in response to a variety of sensory stimuli and may also carry motor command signals. These activity-dependent complex spikes are observed in addition to spontaneous complex spikes that occur at an average frequency of about 1Hz. Thus, while we do not know what the exact functions of climbing fiber signaling are, it is obvious that this unusual input provides a core element of cerebellar cortical circuits and to some degree will define its operations. Our discovery ? presented here as pilot data ? that a subgroup of the entire adult Purkinje cell population (~15%) has two climbing fiber inputs instead of just one, that two inputs are almost exclusively observed in Purkinje cells with two primary dendrites (either separately exiting from the soma, or separating in close proximity to it) and that within this subgroup 20-25% of Purkinje cells show two climbing fiber inputs, leads to the question whether there is a functional significance of this double innervation. Is persistent double climbing fiber innervation of adult Purkinje cells a bug or a feature? In this exploratory study, we plan to examine the basic phenomenon further and to assess whether two separate climbing fiber inputs may signal independently. If the answer to our ?bug or feature? question is ?feature?, we will examine consequences for cerebellar function (e.g. behavioral learning tests) in subsequent studies that we will seek separate funding for. The first aim uses patch-clamp recordings from mice in combination with confocal microscopy in vitro to test the hypothesis that the persistence of multiple climbing fibers is not random, but occurs in Purkinje cells with two primary dendrites. We will also inject tracer dyes into the inferior olive to stain climbing fibers and anatomically confirm the existence of two climbing fiber inputs. The second aim will add calcium imaging in vitro to examine whether a double climbing fiber input leads to functionally separate, dendritic calcium signaling domains. The third aim uses GCaMP6f-based two-photon imaging from Purkinje cells in awake mice to assess whether under spontaneous conditions, or with sensory stimulation, two climbing fiber inputs may operate independently.