Stability of synaptic memory engrams

记忆被认为是大脑中持久的物理变化。在小鼠模型中的几项研究已经确定了在记忆形成过程中显示出增加的神经元活动的各个脑区域的神经元亚群，并表明操纵这些细胞可以诱导人工检索或丢失存储的记忆。这表明记忆的储存和提取是由特定的神经元群介导的，因此被认为是细胞记忆印迹。然而，可以在不同的空间尺度上识别记忆痕迹，从系统到突触水平。根据Hebbian假设，具有高度相关活动模式的神经元之间的连接被加强，而活动模式弱相关的神经元之间的连接被抑制甚至丢失。因此，共同活动神经元之间的突触强度的增加增加了在编码期间发生的神经活动的相同时空模式将在稍后的时间在检索期间再次发生的可能性。这一假说表明，共同激活的神经元之间的突触强化形成了记忆的神经基质，或突触印迹。最近，由于绿色荧光蛋白跨突触伙伴重建（mGRASP）技术，标记记忆印迹神经元之间的突触并研究其特性成为可能。该技术允许用GFP表达标记表达突触前和突触后mGRASP组分的神经元之间的突触。将该系统与已有的基于Immediate-Early基因标记神经元的方法相结合，最终为在突触水平上可视化记忆印迹提供了具体的可能性。在这个项目中，我建议在小鼠中使用纵向双光子脑深部光学成像来研究CA 1中结构性突触印迹的稳定性。首先，我们将通过使用mGRASP标记结构CA 1突触印迹以突出CA 3和CA 1神经元之间的突触，并通过使用立即早期基因Arc的启动子将mGRASP前和后构建体靶向CA 3和CA 1印迹神经元。第二，我们将采用光学成像技术来跟踪CA 1突触恐惧痕迹在基线、消退和强化下的长期稳定性。这将使我们能够第一次研究持续学习对结构性突触痕迹的影响。

Memories are thought to be encoded as enduring physi¬cal changes in the brain. Several studies in murine models have identified subpopulations of neurons throughout various brain regions showing increased neuronal activity during memory formation and shown that manipulation of these cells can induce either artificial retrieval or loss of stored memories. This demonstrates that memory storage and retrieval are mediated by specific populations of neurons which are thus believed to be cellular engrams. Engrams, however, can be identified at different spatial scales, ranging from the systems to the synaptic levels. According to the Hebbian postulate connections between neurons with highly correlated activity patterns are strengthened while connections between neurons whose activity patterns are weakly correlated are depressed or even lost. Increase in synaptic strength between co-active neurons thus increases the likelihood that the same spatial-temporal pattern of neural activity that occurred dur¬ing encoding will occur again at a later time, during retrieval. This hypothesis suggests that synaptic strengthening between coactivated neurons forms the neural substrate of memory, or a synaptic engram. It has recently become possible to label synapses between engram neurons and study their properties, thanks to the Green fluorescent protein Reconstitution Across Synaptic Partners (mGRASP) technique. This technique allows to label with GFP expression synapses between neurons expressing the presynaptic and the postsynaptic mGRASP components. Combining this system with established methods to label engram neurons based on Immediate-Early genes, finally offers the concrete possibility to visualize the engram at the synaptic level. In this project I propose to investigate the stability of structural synaptic engrams in the CA1 by using longitudinal two-photon deep-brain optical imaging in mice. First, We will label structural CA1 synaptic engrams by using mGRASP to highlight synapses between CA3 and CA1 neurons and by using the promoter of the immediate early gene Arc to target the pre- and post-mGRASP constructs to CA3 and CA1 engram neurons. Second, we will employ optical imaging to track the long-term stability of CA1 synaptic fear engrams under baseline, extinction and reinforcement. This will enable us to investigate for the first time the effects of continued learning on structural synaptic engrams.