Digital behavioral phenotyping and multi-region electrophysiology to determine behavioral and neural network changes underlying the stress response in mice

摘要： 慢性心理压力会引发并加剧严重抑郁障碍(MDD)和其他许多疾病 精神疾病-导致睡眠、饮食习惯、成瘾行为、活动水平、昼夜节律的变化 节奏、情绪和其他领域。啮齿动物的应激反应具有许多行为和生理特征。 与人类的变化。慢性压力对大脑也有广泛的影响。但在我们的国家中存在重大差距 关于整合的行为和生理以及相应的大脑回路变化的知识 有长期的压力。先前的工作已经发现了许多行为和生理压力的表型，但我们缺乏一种 这些变量如何随着时间的推移共同演变的凝聚力。我们的第一个目标是描绘压力的共同进化。 应激小鼠与非应激小鼠的反应元件。我们将通过在一个 慢性不可预测应激(CUS)范式与我们新的自然主义观察系统中的对照 “数字家庭”。这个系统允许我们在几周内同时监测50多项行为测量。 小鼠将在这些家庭中生活8周：基线2周，CUS 4周，恢复2周。一个 该目标的探索性元素是使用机器学习来确定一个连贯的小鼠应激生物标记物 未来的定量研究。我们的下一个目标是确定慢性应激的电生理特征。它是 众所周知，慢性应激会改变大脑回路、突触结构和神经调节平衡。众所周知， 行为是由大脑网络的电生理状态控制的，这些网络既操作又操作 在当地区域内并通过协调的多区域传播。因此，我们的目标是研究 区域内和区域间的电生理学。我们关注的是内侧前额叶皮质，即腹侧 海马体和下缘内侧前额叶皮质与慢性应激密切相关。我们会 将Tetrode阵列植入这些区域，将如上所述记录8周以上。在目标2中，我们将确定 慢性应激对区域内尖峰电位的影响，包括尖峰频率变异性和兴奋性。 抑制性平衡。在这个目标的第二部分，我们将使用机器学习来应用于更广泛的内部- 确定CUS和CUS之间潜在的更完整差异集的区域动力学度量 控制小鼠。在我们的最终目标中，我们将评估这三个区域之间的跨区域协调。我们将测试 假设区域之间的成对耦合将以与MDD一致的方式被改变，通过 同时使用尖峰和LFP测量耦合。同样，我们将在我们的大型数据库中使用机器学习方法 数据集，以检测在我们的假设驱动的测试中未揭示的进一步的区域间动态。这种混合体 行为和电生理学是为了产生对慢性应激的新理解。我们还有一个 创建用于未来分析的大数据集、面向未来的全面完善的数字家庭系统的长期愿景 研究，着眼于开发基于自然电生理回路功能的干预措施。

ABSTRACT: Chronic psychological stress triggers and exacerbates major depressive disorder (MDD) and many other psychiatric conditions – causing changes in sleep, eating habits, addictive behaviors, activity levels, circadian rhythms, mood and other domains. The rodent stress response shares many behavioral and physiologic alterations with that of humans. Chronic stress also has broad effects on the brain. But major gaps exist in our knowledge in regard to the integrated behavior and physiology as well as the corresponding brain circuit changes with chronic stress. Prior work has found many behavioral and physiologic phenotypes of stress, but we lack a cohesive sense of how these variables co-evolve over time. Our first aim is to delineate this co-evolution of stress response elements in stressed versus unstressed mice. We will accomplish this by examining mice under a chronic unpredictable stress (CUS) paradigm versus controls in our new naturalistic observation system the “Digital Homecage”. This system allows us to monitor over 50 behavioral measures simultaneously over weeks. Mice will live in these homecages for 8 weeks: 2 weeks baseline, 4 weeks CUS and 2 weeks of recovery. An exploratory element of that aim is to use machine learning to determine a coherent mouse stress biomarker for future quantitative studies. Our next goal is to determine electrophysiologic signatures of chronic stress. It is known that chronic stress alters brain circuit synaptic structure and neuromodulatory balance. It is known that the behavior is controlled by the electrophysiologic state of brain networks and that those networks operate both locally within regions and via coordinated multi-regional transmission. Therefore, we aim to study changes in electrophysiology both within and across regions. We focus on the medial prefrontal cortex, the ventral hippocampus and infralimbic medial prefrontal cortex given their strong involvement in chronic stress. We will implant tetrode arrays into these regions and will record over 8 weeks as above. In Aim 2, we will determine the effects of chronic stress on within-region spiking tendencies including spike rate variability and excitatory- inhibitory balance. In a second part of this aim we will use machine learning applied to a wider variety of within- region dynamical measures to determine a potentially more complete set of differences between CUS and control mice. In our final Aim, we will assess cross-regional coordination between these 3 regions. We will test the hypothesis that pairwise coupling between regions will be altered in a manner consistent with MDD by measuring coupling using both spiking and LFP. Again, we will then use machine learning methods on our large dataset to detect further inter-regional dynamics un-revealed in our hypothesis-driven testing. This mixture of behavior and electrophysiology is done to generate new understanding about chronic stress. We also have a long-term vision of creating large dataset for future analysis, a fully-refined Digital Homecage system for future studies, with an eye towards developing interventions based on natural electrophysiologic circuit function.