부산대학교 도서관

The functions of sleep are diverse and still poorly understood, but a strong effect on cognition is evident. An impressive number of studies suggest that particularly deep sleep, with characteristic cortical slow-wave activity, mediates some important beneficial effects on learning and memory, such as memory consolidation and integration. In the light of this, it is surprising how little we know about the specific rules of synaptic plasticity associated with characteristic activity patterns of different sleep stages. Therefore, using whole-cell recordings from single or synaptically coupled principal cells and two-photon Ca2+ imaging of dendritic spines, I explored how the ongoing network state might promote activity-dependent synaptic plasticity in an in vitro model of the medial entorhinal cortex. This experimental setup allowed precise control over Up-Down state oscillations (cellular membrane potential fluctuations associated with slow-wave activity) - a methodological advantage that is difficult to achieve in vivo. I found that evoking subthreshold synaptic inputs during the Up state phase of cortical slow-wave activity induced N-methyl-D-aspartate receptor-dependent synaptic weakening. In fact, the spontaneous, intrinsically generated recurrent network activity that underlies cortical Up states was able to depress the very inputs that help maintain it. These findings are in agreement with the synaptic homeostasis hypothesis of sleep, a proposal for which descriptions of clear molecular and cellular mechanisms have been missing. Next, I investigated spike-timing dependent plasticity during Up state periods. I found that input-correlated postsynaptic spiking can prevent synaptic weakening. This suggests that while subthreshold synaptic inputs become continuously weaker during slow-wave activity, correlated inputs become relatively more dominant - a process that could be related to memory consolidation. Finally, I investigated Ca2+ signalling in dendritic spines during Up- Down state oscillations using a novel multi-photon microscope based on the remotefocusing technology. These experiments identified a biochemical signature that could drive the observed plasticity rules.