Hippocampal cortical interactions and the extraction of knowledge from episodic memory

我的研究项目侧重于记忆的神经计算基础以及神经元群体之间的动态相互作用，这些相互作用使大脑能够创建其自适应模型。在过去的 35 年里，我在理解中枢突触可塑性机制、空间信息处理、皮质-海马相互作用和记忆巩固以及系统神经科学研究的显着先进技术方面做出了重大贡献。在这项工作的过程中，我为 33 名研究生和 37 名博士后提供了出色的培训，他们中的大多数人都在主要大学从事研究工作。在接下来的 5 年里，我的总体计划（包括其他支持）将为至少 4 名本科生、10 名研究生和 7 名博士后提供培训，因为我们将继续研究海马体向新皮质的反向投射对该结构中的神经编码和神经可塑性的影响。了解这种影响是理解为什么在缺乏功能性海马体的情况下，新皮质无法形成新记忆或有效地从经验中提取一般知识的关键。它也是理解计算知识获取和表示模式的关键，这种模式可能对信息和计算机技术产生深远的影响。********“海马体 (HC) 对新皮质 (NC) 有何作用？”这个问题在三个具体的子主题下得到解决：1）HC 如何影响 NC 神经元的实际编码特性？特别是，某些 NC 区域中对象编码和行为的明显空间调制是否继承自 HC 中的“位置编码”？ NC 编码统计数据是否会随着大脑完善其世界模型而发生变化，HC 是否能够实现这种变化？ 2) 在休息和慢波睡眠期间在 NC 中观察到的记忆痕迹的重新激活在多大程度上取决于 HC 输出？与恢复或自发检索的旧记忆相比，新记忆的依赖性是否不同？对于新记忆，HC 本身是 NC 重新激活所必需的，还是仅仅需要在广泛区域同步皮质重新激活？在记忆重新激活过程中，HC 流出最强烈地激活 NC 的哪些部分？随着记忆逐渐巩固，NC 和 HC 记忆重新激活之间的相对时间是否会发生变化（即 HC 是否最初领先 HC，随后跟随它）？ 3）HC对于NC中与学习相关的树突棘可塑性是必需的吗？ HC 是否有助于调节对 NC 神经可塑性至关重要的 NC 神经营养因子的表达？******** 在具有完整或失活 HC 功能的动物中，使用多神经元记录的各种组合、使用电生理学和光学成像方法、新皮质电压记录、体内 NC 树突棘的结构和功能成像以及神经营养因子的免疫化学定量来解决这些问题。******

My research program focusses on the neuro-computational basis of memory and the dynamic interactions among neuronal populations that lead to the brain's ability to create an adaptive model of its. Over the past 35 years, I have made major contributions to understanding central synaptic plasticity mechanisms, spatial information processing, cortico-hippocampal interactions and memory consolidation, and significantly advanced technologies for systems neuroscience investigation. In the course of this work I have provided outstanding training to 33 graduate students, and 37 postdocs, most of whom are pursuing research careers at major universities. Over the next 5 years, my overall programme (including other support) will provide training for at least 4 undergrads, 10 graduates and 7 postdocs, as we continue to study the impact of hippocampal back projections to the neocortex on neural coding and neural plasticity in that structure. Understanding this impact is key to understanding why, in the absence of a functional hippocampus, the neocortex is unable to form new memories or to extract general knowledge from experience efficiently. It is also key to understanding a mode of computational knowledge acquisition and representation that may have far reaching impacts on information and computer technology.******* "What does the hippocampus (HC) do for neocortex (NC)?" This question is addressed under three specific subtopics: 1) How does HC affect the actual coding properties of NC neurons? In particular, is the apparent spatial modulation of object encoding and behavior in some NC regions inherited from "place coding" in HC? Do NC coding statistics change as the brain refines its model of the world, and does HC enable such changes? 2) To what degree does the reactivation of memory traces observed in NC during rest and slow-wave-sleep depend on HC output? Is the dependence different for new memories compared to reinstated or spontaneously retrieved old memories? For new memories, is HC required for NC reactivation per se, or rather is it merely required for synchronizing cortical reactivation over broad areas? Which parts of NC are most strongly activated by HC outflow during memory reactivation? Does the relative timing between NC and HC memory reactivation change as memories gradually become consolidated (i.e., does HC initially lead HC and subsequently follow it)? 3) Is HC necessary for learning-associated dendritic spine plasticity in NC? Does HC contribute to the regulation of expression of NC neurotrophins which are crucial to neuroplasticity in NC?******* These questions are addressed using various combinations of multi-neuron recording, using electrophysiological and optical imaging methods, neocortical voltage recording, structural and functional imaging of NC dendritic spines in vivo, and immunochemical quantification of neurotrophic factors, in animals with intact or inactivated HC function.******