Sleep shapes more effective cerebro-cerebellar networks performing sequential motor tasks

Sleep is essential for the consolidation of procedural memory. Using fMRI we measured activation while subjects were trained on a sequential finger opposition task. 8 Volunteers participated in the main experiments with retrieval testing taking place 48 hours after initial training. During the first post-learning night subjects either slept regularly or stayed awake. During the second night all subjects had regular sleep. 10 volunteers were examined in supplementary experiments with training at 10:00p.m. and retrieval at 7:00 a.m. The volunteers performed a finger-opposition task with their non-dominant hand as rapidly and accurately as possible. Two different motor sequences were used to allow testing of the same subject in both conditions. Imaging was carried out at 1.5 T and analyzed using SPM2. Initial learning of the sequence was similar in the sleep and the wake condition. Performance rate improved at initial training across retention intervals of sleep. There was no change for wakefulness. Initial training was linked to specific cerebro-cerebellar brain activation performing left hand movements. The before/after×sleep/wake interaction revealed several aspects. Separate analyses of the sleep and the sleep deprivation conditions revealed sleep-dependent decreases in activation in the left medial frontal lobe, the right lateral premotor cortex, the right inferior frontal lobe, the anterior cingulate cortex, the left primary motor cortex, and in inferior primary somatosensory cortex. No brain region showed decrease in response to the post-training wake condition. Post-training sleep led to increases in the right caudate and the left anterior lobe of the cerebellum. Testing of the overnight condition confirmed these results with reduced activation in frontal cortical regions and a distinct decrease in bilateral parietal areas and in the cerebellum.

In onclusion, sleep induces decreases during retrieval as compared with initial learning in various relevant cerebro-cerebellar areas. This pattern was equally revealed in the main study and in supplementary experiments, where retesting immediately followed sleep and sleep deprivation. The supplementary experiments, thus, ensure the importance of sleep on the first post-training night to the enhancement in skill and associated changes in the cortical representation and possibly defines structures like the cerebellum that are predominantly involved in earlier consolidation processes.