Modern neuroscience covers broad fields of scientific research, ranging from molecular biophysics to system physiology and behavior. In order to promote research on core neuroscience themes, we propose to set-up a consortium including highly qualified groups across Europe and in Israel. All these groups are at the forefront of scientific research and technological development, are (or have been) already collaborating on common projects and have a long-time experience in student training. The proposal is that of training young researchers (both ESR and ER) through their own research activity, which will be supervised by the conjoint project board and will be complemented through specific courses and workshops. The trans-disciplinary nature of the research goal is especially suitable to provide a comprehensive training to ESR and ER, who will constitute the next generation of highly competent neuroscientists. Moreover, the contribution of industrial partners will facilitate interactions with other sectors of the society, translational perspectives and training on commercial issues that are usually extraneous to universities and centers for basic research. The core neuroscience theme addresses the functional mechanisms of the olivo-cerebellar system. The olivo-cerebellar system is known to process a huge amount of sensori-motor signals to rapidly control fine movement coordination and to store memories of past procedures. Moreover, in recent years, a role of the cerebellum in cognitive functions has been reported by several groups. Key concepts that emerged recently involve the precise time pattern of spikes and distributed long-lasting synaptic plasticity inside the network. Understanding these aspects provides clues on how the system might work in order to coordinate movements, store procedural memory and elaborate aspects of cognitive functions. Despite decades of intense research, the functional mechanisms of this system remain still largely unclear. We expect, however, that more recently developed approaches including imaging technologies, whole cell recordings in awake behaving animals, and new transgenic mice will lead to a dramatic advance in this research field over the coming years. The first comprehensive model of cerebellar functions, which remains a basic conceptual reference, was inspired by accurate morphological determinations of the number of neurons and synapses but accounted for only very limited knowledge on functional properties of the cerebellar circuitry. In recent years, important achievements have been obtained in individual laboratories that provided insight on segmental aspects of the olivo-cerebellar circuit function. What is still missing is a coordination of efforts to provide an integrated description of cerebellar function. The main motivation to form this consortium is to merge fractured understanding into a comprehensive account, providing the infrastructure for interdisciplinary and interactive training. Cerebellar function is organized in modules including cortical microcomplexes. The investigations of circuit mechanisms can, as a first step, be conceived as addressing one such module. Each module receives two kinds of inputs, one from the mossy fibers and another from the climbing fibers. These inputs converge onto Purkinje cells, which eventually inhibit the deep cerebellar nuclei, representing the sole output of the circuit. A core emerging concept is that timing (the precise pattern of spikes) and plasticity (long-term synaptic plasticity) are distributed network properties and that they are tightly bound together, so that they can influence one each other. Nonetheless, for clarity, several computational sections can be considered.
- Signals coming into the cerebellum through the mossy fibers are processed in the granular layer network. Here, with the intervention of the inhibitory circuits and synaptic plasticity, mossy fiber spikes are transformed into new spatio-temporally organized sequences for further processing in Purkinje cells. Granule cells and Golgi cells have specific electroresponsive properties with resonance in the theta band Moreover, short- and long-term synaptic plasticity is generated by repetitive activity at the mossy fiber – granule cell synapse. The properties of the mossy fibre firing pattern appear to dependent on sensory input: vestibular input is linearly represented through a modulation of a tonic level of firing, typically in the 0-40Hz range, mossy fibers encoding joint angle are also tonically active but have a higher frequency range. While transient stimuli through trigeminal inputs tend to give rise to precise spike bursts of mossy fibre input and granular layer communication toward Purkinje cells, in which the parallel fiber input modulates the generation of simple spikes. The granular layer is capable of sustaining repetitive activity enhancing responses in the theta and probably also in the gamma frequency bands.
- A second input to the cerebellar cortex comes from the inferior olive through the climbing fibers. The inferior olive itself is an oscillator, which can produce theta-frequency patterns influencing Purkinje cells and inhibitory interneurons of the molecular layer. Although much less numerous than parallel fibers, the climbing fibers exert a very powerful effect on the Purkinje cells eliciting the complex spikes. The complex spike has been variously interpreted as a signal carrying either an error or an instruction for generating synaptic plasticity at the parallel fiber – Purkinje cell synapse. More recently however, it has been demonstrated in anesthetized animals that the climbing fiber input may influence the bistable transition of Purkinje cells between up and down states, and various groups within our consortium have now ongoing collaborations to find out to what extent this new phenomenon is functional in awake behaving animals.
- The Purkinje cells have their own processing mechanisms, which also relay on intrinsic electroresponsive properties and synaptic plasticity. Purkinje cells are spontaneously active and their discharge is modulated by the activity coming from the granular layer and the inferior olive. It was recently shown that the molecular layer can sustain synchronous gamma band (30 – 80 Hz) and high-frequency (100-200 Hz) oscillations entraining the Purkinje cells. Purkinje cells were proposed to act as perceptrons and to process spike pauses, and may live in a bistable UP-DOWN state. Therefore, the mechanisms of Purkinje cell processing are all but fully resolved. Purkinje cell synapses are sites of plasticity, including the renown parallel fiber – Purkinje cell LTD and the more recent parallel fiber – Purkinje cell LTP as well as plasticity at climbing fiber and interneuron synapses.
- The cells of deep cerebellar nuclei (DCN) finally convert the activity of microzones into the cerebellar output. The DCN are at a key location within the cerebellar network. All of the afferent pathways to the cerebellar cortex make collateral connections on to neurons of the DCN (both mossy fiber and to a lesser extent climbing fiber), while the main output of the cerebellum is formed by the DCN projection neurons. Moreover, DCN neurons inhibit the IO cells regulating their coupling. IN DCN cells, intrinsic dynamics generate silent pauses and possibly rebound excitation, producing alternating phases of activity. Despite this knowledge, the role of DCN in cerebellar computation is largely unclear or controversial. A classical view was that the DCN simply acts as a "relay station" between cerebellar mossy fiber (MF) input and cerebellar output to premotor areas, either directly (“direct pathway”) or via the cerebellar cortex (“indirect pathway”). This concept has been challenged by hypotheses which suggest the DCN may act as the substrate of motor memory storage. The demonstrations of mechanisms which cause modification of synaptic strength and active membrane properties support this latter viewpoint. In addition to these possible roles in cerebellar memory, it appears likely that the neuronal network within the DCN processes sensory and motor information in “real-time” but very little is known about the computational functions of the DCN, particularly with respect to the role of specific cell types. Indeed, morphological and electrophysiological studies have revealed that the DCN consists of diverse neuronal populations with distinct integrative properties. Thus, one can hypothesize that the synchronous oscillations in the Purkinje cell activities together with plasticity at the mossy fiber – DCN and the Purkinje cell – DCN synapses form the main mechanistic tools to control the activity in the DCN output neurons, and that different sets of neurons in the DCN are sensitive for oscillations at different frequency ranges.
This brief description highlights why timing and plasticity hold the key to understand how specific spike sequences are emitted by a certain cerebellar module. We will therefore address specific unresolved issues regarding the physiology of single neurons and synapses, develop specific techniques to investigate large fields of circuit activity, and generate computational models of the cerebellar network. ER and ESR will typically develop their research and training crossing over the different fields.