Temporal patterns of training determines distinct kinetics of structural plasticity, memory formation and decay 学習の時系列パターンによって決まる構造的シナプス可塑性および記憶形成と消失の経時変化
Temporal patterns of training determines distinct kinetics of structural plasticity, memory formation and decay
During learning, memories are formed in the selective population of the neuronal circuits which are then consolidated in order to persist. These memory processes are supported by distinct sub cellular events such as reversible functional modifications of synaptic transmission and persistent structural modifications in size or number of synaptic connections. However, how these synaptic modifications relate to the dynamics of formation and decay of memory in behaving animals remain elusive. Memory formation and its persistence are also known to be sensitive to the temporal features of the stimulus presentation, highlighted in the well-known “spacing effect” observed in a variety of explicit and implicit memory tasks. Training trials with resting intervals between them (spaced training) produce stronger and longer-lasting memory than the same number of trials with no interval (massed training). Despite numerous behavioral and molecular studies of the spacing effect, no conjoint study has demonstrated the underpinning synaptic plasticity in the expression of spacing effect during physiological learning. In invertebrate models, the different phases of memory are well dissected based on the temporal patterns of training. In Drosophila, implicit memory for odor avoidance task formed by massed training decayed over 3 days whereas by spaced training the memory retention was seen at least for a week. Studies using Aplysia model has shown that spaced training produced sensitization of gill withdrawal reflex lasting more than 24 h while massed training failed to induce long-term memory. However, there is no study showing the time course of memory formation and its decay following massed or spaced training in the mammalian brain. Here, we set out to study the temporal evolution and decay of memory and its correlation with synaptic modifications in learning with distinct temporal patterns of training.Adaptation of horizontal optokinetic response (HOKR) is a simple model of cerebellar motor learning to stabilize the visual image of moving surround. A phylogenetically preserved cerebellar lobule, flocculus (Fl) is known to be involved in the control of horizontal eye movement and the adaptation of HOKR. Previously, we have shown that short-term adaptation (STA, within a day) of HOKR in mice is accompanied with a transient reduction in number of AMPA receptors in parallel fiber to Purkinje cell (PF-PC) synapses which is recovered after 24 h. Long-term adaptation (LTA) of HOKR (days to weeks) with repeated 1 h massed training for five days was accompanied with long-lasting reduction of PF-PC synapses in the flocculus (Fl). Thus, HOKR serves as an ideal tool to study the effect of distributed and continues motor practice on kinetics of the memory and its relationship to the modifications in PF-PC synapses.In the present study; I examined adaptation of horizontal optokinetic response by massed or spaced training with varying intervals in mice. Despite similar acquisition of HOKR gain at the end of all training protocols, the retention of HOKR examined at 24 h was significantly resting-interval dependent. Massed training showed significant reduction of gain at 24 h. Also, the shorter intervals of 10 and 20 min produced poor retention of HOKR gain where as 1hour spacing interval showed the highest memory retention at 24 h. These results showed that optimum spacing is critical for efficient memory retention.Further, time course study revealed the distinct kinetics of long-lasting memory. Spaced training with optimum interval of 1 h not only developed the long-lasting memory quickly but also maintained it for about a month. This distinct kinetics of long-lasting memory by spaced training also strongly correlated with reduction of floccular parallel fiber-Purkinje cell (PC) synapses. Shrinkage in PF-PC synapses and PC spine together with reduction in density of AMPA receptors were already detected just at the end of spaced training. These structural changes were completed within 4 h after spaced training accompanied with 50% reduction of PF-PC synapses and PC spines and maintained for at least 2 weeks. Surprisingly, massed training also showed long-lasting memory with corresponding synapse reduction that developed slowly reaching the peak of 50% by 5th day. However, the memory lasted for a week with recovery of PF-PC synapses. Thus, I for the first time revealed a distinct form of long-lasting memory by massed training which is slow to remember and quick to forget as compared to spaced training. Moreover, the distinct memory kinetics was tightly correlated with structural modifications at the synapse. Taken together, the unique temporal profiles of synaptic plasticity may serve as a structural basis for distinct memory processes in learning with and without intervals.