ADVANCE Fellows Award: Theta phases of hippocampal place cell firing in REM sleep and waking

Poe等人（Brain Res., 855:176- 188,2000）观察到，在快速眼动睡眠中海马CA1位置细胞的重复相位进动过程中，随着大鼠对相关的轨迹环境越来越熟悉，放电的平均相位相对于海马体θ节律发生了变化。具体而言，大鼠暴露于轨迹的前两天，REM进动期间的平均θ波相位接近θ波节奏的去极化峰。在接下来的几天里，在快速眼动进动期间，平均θ波相移到θ波周期的超极化波谷附近，但在轨道运行期间仍保持在去极化波峰附近。这个项目的目标是研究产生这种相移的神经机制，这种相移发生在快速眼动进动期间。PI是一名研究计算神经科学的应用数学家，他将直接与吉娜·坡教授合作，后者目前正在进行实验，记录活跃清醒和随后的快速眼动睡眠期间海马区CA1的位置细胞放电。对于平均放电相位变化的原因，合作者的工作假设是，据报道，CA1锥体细胞的θ节律驱动中的躯体-树突差异可以根据不同位置的突触输入来不同地调节放电相位。具体来说，θ驱动的调制导致近端树突输入优先在去极化θ波峰附近产生放电，远端树突输入优先在超极化θ波谷产生放电。相移可能是由于远端树突突触输入的增强引起的，远端树突突触输入由内嗅皮层的直接突触通路携带，而近端树突突触输入由Schaffer侧枝和内嗅皮层的三突触环路携带。该项目通过在模拟轨道跑步和快速眼动睡眠的条件下模拟现实CA1锥体细胞模型中的突触后反应，并通过分析实验数据来验证这一假设。一般认为，在特定的时间序列中，海马位置细胞的集合形成了大鼠对相关环境记忆的海马表征。在探索环境后的睡眠期间，位置细胞放电的回放被认为有助于记忆在海马体中的巩固，并将其转移到新皮层的长期储存。实验观察到，当大鼠熟悉相关环境时，在快速眼动睡眠中重放放电时，位置细胞放电的平均θ相位发生偏移，这可能是记忆转移的一个可观察指标，也可能表明海马体在快速眼动睡眠中具有关键功能。作为研究产生这种转变的神经机制的第一步，将进行神经元模拟和数据分析，其结果将确定这种转变是由不同突触输入细胞的变化引起的这一工作假设的限制。研究结果将做出实验可验证的预测，并建议进一步的实验来研究快速眼动睡眠和记忆处理之间的联系。

During replayed phase precession of hippocampal CA1 place cells in REM sleep, Poe et al. (Brain Res., 855:176-180, 2000) observed that over a number of days, as the rat became more familiar with the associated track environment, a shift occurred in the mean phase of firing relative to the hippocampal theta rhythm. Specifically, on the first two days the rat was exposed to the track, mean theta phase during REM precession was near the depolarizing peaks of the theta rhythm. Over following days, the mean theta phase shifted to near the hyperpolarizing troughs of the theta cycle during REM precession but remained near the depolarizing peaks during track running. The goal of this project is to investigate the neural mechanisms generating this phase shift that occurs with familiarity during REM precession. The PI, an applied mathematician whose research is in computational neuroscience, will work directly with Prof. Gina Poe, who is currently conducting experiments to record place cell firing in hippocampal region CA1 during active waking and subsequent REM sleep episodes. The collaborators' working hypothesis for the cause of the shift in mean firing phase is that the reported somato-dendritic differential in the theta rhythm drive to CA1 pyramidal cells can differentially modulate the firing phase in response to synaptic inputs at different locations. Specifically, the theta-driven modulation causes proximal dendritic inputs to preferentially generate firing near the depolarizing theta peaks and distal dendritic inputs to preferentially generate firing at the hyperpolarizing theta troughs. The phase shift would result from the potentiation of distal dendritic synaptic inputs, carried by the direct synaptic pathway from the entorhinal cortex, relative to proximal dendritic synaptic inputs carried by the Schaffer collaterals and tri-synaptic loop from the entorhinal cortex. The project tests this hypothesis through simulations of post-synaptic responses in realistic CA1 pyramidal cell models under conditions mimicking track-running and REM sleep, and through analysis of experimental data. It is generally assumed that an ensemble of hippocampal place cells firing in a specific temporal sequence forms the hippocampal representation of a memory of the associated environment for the rat. Replay of place cell firing during sleep episodes following exploration of an environment is thought to contribute to the consolidation of the memory in the hippocampus and to its transfer to neocortical long-term stores. The experimentally observed shift in mean theta phase of place cell firing that occurs during replayed firing in REM sleep as the rat becomes familiar with the associated environment may be an observable indicator of memory transfer and may also suggest a critical function for REM sleep in the hippocampus. As initial steps of the investigation into the neural mechanisms generating this shift, neuronal simulations and data analysis will be conducted whose results will identify constraints on the working hypothesis that the shift is caused by changes in distinct synaptic inputs to the cell. The results will make experimentally-testable predictions and suggest further experiments to investigate this link between REM sleep and memory processing.