Modeling the Molecular Networks that Underlie the Formation and Consolidation of Memory

项目摘要/摘要 这项提议将检验这样一个假设，即学习和记忆可以通过使用计算来改善 设计的训练方案优化了激酶级联和转录因子之间的相互作用 参与了长时记忆(LTM)的诱导。将使用三个模型系统：长期敏化， 恐惧的条件反射和消退，以及抑制性回避学习。这一假设是基于我们之前的 研究表明，通过计算设计的方案最大限度地增加了蛋白质之间的活动重叠 激酶A(PKA)和丝裂原活化蛋白激酶(MAPK)亚型称为细胞外信号调节 激酶(ERK)增强长时程突触易化(LTF)和LTM敏化，以及获得 和消除恐惧的学习。这项提议有两个关键的创新方面。首先，我们利用了一种新颖的、多方面的 加强学习和提高不同类型记忆提取的纪律策略。药理作用 许多人一直在进行改善学习和记忆、挽救记忆缺陷的干预 几十年，但这些方法依赖于反复试验，并且高度不特定。相比之下，我们的战略是 将生物现实的计算模型与经验方法相结合，使我们能够 有效和系统地探索不同类型长时突触背后的分子过程 可塑性，并预测个人训练方案，以优化学习和记忆。第二，据我们所知， 我们的团队是第一个开发出描述婴儿期发病机制的计算模型的人 学习和明显的快速遗忘与婴儿期健忘症相关。我们的模拟表明， 调节幼年动物的激酶和转录因子的基础活性有助于快速遗忘 婴儿期的记忆。需要通过模拟和活体实验验证的具体假设包括：目标1)LTF 通过新的计算设计的训练方案，致敏的LTM可以延长到7天； 目的2)基于杏仁核激酶动力学的计算机设计的方案可以增强 条件性恐惧记忆的获得和消失；以及目标3)明显的婴儿快速遗忘 在抑制回避范式中观察到的记忆可以通过计算设计的方案来克服 基于海马区蛋白激酶的动态变化。我们认为，这些预测与同时发生的 经验性测试，将为增强学习和改善记忆的有效策略提供原则证明 取回。我们的研究可能对旨在促进记忆形成的干预措施具有临床意义 与人类认知障碍相关的一系列精神障碍以及改善 焦虑相关障碍患者的基于消退的治疗。

PROJECT SUMMARY/ABSTRACT This proposal will test the hypothesis that learning and memory can be improved by using computationally designed training protocols that optimize the interactions among kinase cascades and transcription factors involved in the induction of long-term memory (LTM). Three model systems will be used: long-term sensitization, fear conditioning and extinction, and inhibitory avoidance learning. This hypothesis is based on our previous work demonstrating that computationally designed protocols maximizing the overlap of activities between protein kinase A (PKA) and the mitogen-activated protein kinase (MAPK) isoform termed extracellular signal-regulated kinase (ERK) enhance long-term synaptic facilitation (LTF) and LTM for sensitization, as well as the acquisition and extinction of fear learning. This proposal has two key innovative aspects. First, we utilize a novel, multi- disciplinary strategy to enhance learning and improve different types of memory retrieval. Pharmacological interventions to improve learning and memory, and rescue memory deficits, have been ongoing for many decades, but these approaches rely on trial-and-error and are highly nonspecific. In contrast, the strategy we have developed, combining biologically realistic computational models with empirical approaches, enables us to efficiently and systematically explore the molecular processes that underlie different types of long-term synaptic plasticity, and predict individual training protocols to optimize learning and memory. Second, to our knowledge, our groups are the first to develop a computational model describing the possible mechanism underlying infantile learning and the apparent rapid forgetting associated with infantile amnesia. Our simulations suggest that altered regulation of basal activities of kinases and transcription factors in infant animals contributes to fast forgetting of infantile memory. Specific hypotheses to be tested by simulation and in vivo experiments include: Aim 1) LTF and LTM for sensitization can be prolonged up to 7 days by novel computationally designed training protocols; Aim 2) Computationally designed protocols based on the dynamics of amygdala kinases can enhance the acquisition and extinction of conditioned fear memories; and Aim 3) The apparent rapid forgetting of infantile memory observed in an inhibitory avoidance paradigm can be overcome by computationally designed protocols based on the dynamics of hippocampal kinases. We believe that these predictions, combined with concurrent empirical tests, will provide a proof of principle for an efficient strategy to enhance learning and improve memory retrieval. Our study may have clinical relevance for interventions aiming at facilitating memory formation in a series of psychiatric disorders associated with cognitive impairment in humans, as well as for improving extinction-based therapies in patients suffering from anxiety-related disorders.