PT  - JOURNAL ARTICLE
AU  - Elinor Tzvi
AU  - Christoph Zimmermann
AU  - Richard Bey
AU  - Thomas F. Müente
AU  - Matthias Nitschke
AU  - Ulrike M. Kräemer
TI  - Cerebellar degeneration affects cortico-cortical connectivity in motor learning networks
AID  - 10.1101/136820
DP  - 2017 Jan 01
TA  - bioRxiv
PG  - 136820
4099  - http://biorxiv.org/content/early/2017/05/11/136820.short
4100  - http://biorxiv.org/content/early/2017/05/11/136820.full
AB  - AcknowledgementsWe would like to thank Susanne Schellbach and Christian Erdmann for assisting with data acquisition, Steffan Frässle for helpful advice on dynamic causal modelling, and Matthias Liebrand for helpful discussions on this work. This study was supported by internal funding of the University of Lübeck. UMK and TFM are supported by the DFG.The cerebellum plays an important role in motor learning as part of a cortico-striato-cerebellar network. Patients with cerebellar degeneration typically show impairments in different aspects of motor learning, including implicit motor sequence learning. How cerebellar dysfunction affects interactions in this cortico-striato-cerebellar network is poorly understood. The present study investigated the effect of cerebellar degeneration on activity in causal interactions between cortical and subcortical regions involved in motor learning. We found that cerebellar patients showed learning-related increase in activity in two regions known to be involved in learning and memory, namely parahippocampal cortex and cerebellar Crus I. The cerebellar activity increase was observed in non-learners of the patient group whereas learners showed an activity decrease. Dynamic causal modelling analysis revealed that modulation of M1 to cerebellum and putamen to cerebellum connections were significantly more negative for sequence compared to random blocks in controls, replicating our previous results, and did not differ in patients. In addition, a separate analysis revealed a similar effect in connections from SMA and PMC to M1 bilaterally. Again, neural network changes were associated with learning performance in patients. Specifically, learners showed a negative modulation from right SMA to right M1 that was similar to controls, whereas this effect was close to zero in non-learners. These results highlight the role of cerebellum in motor learning and demonstrate the functional role cerebellum plays as part of the cortico-striato-cerebellar network.