Regulation and plasticity of dendritic compartmentalization features

锥体神经元是大脑中主要的兴奋细胞，它们在精心设计的树突上接收成千上万的突触输入。近几十年来，人们在概述锥体细胞树突的综合特性方面取得了显著进展，现在人们已经充分认识到，“树突计算”的一个关键方面涉及大量非线性机制，这些机制最终控制突触输入如何转化为尖峰活动，从而有力地调节神经回路中的信息传递。最近在体内获得的一些研究结果进一步证实了锥体细胞计算的一个核心组成部分涉及突触活动中时空相关性的解码和传递。这些见解支持了一个日益增长的共识，即神经元由于其树突，是高度划分的信息处理器。树突的兴奋性由质膜中的离子通道控制，在无源电缆特性的约束下起作用。电压门控钠通道（VCSCs）和/或电压门控钙通道（VGCCs）的激活参与了树突突的产生，而钾通道的存在调节了这些事件的关键方面。尽管树突支持的计算操作本身在本质上是高度动态的，但在给定神经元中指导这些操作的集体规则在很大程度上（尽管不是唯一的）被概念化为静态的。然而，神经调节剂可以阶段性地调节大量的活性离子电导。因此，尽管其重要性得到公认，但神经调节系统在动态和双向调节整个大脑树突综合特征的规则方面的潜在作用在很大程度上仍未得到实验的探索。这种功能的揭示将为这些调节递质在调节大脑信息处理过程中所起的作用提供关键的见解。当前研究计划的总体目标是：1)通过结合细胞电生理学、双光子成像和释放神经递质的笼化类似物，识别和表征树突操作目录和树突功能的区隔特征，并确定这组操作如何受到神经传递的动态调节；2)为了理解树突事件是如何参与一种新型突触可塑性的，我们最近开始研究那些间隔数百毫秒的事件。这种新颖的可塑性为现代学习理论中众所周知的时间信用分配问题提供了一个有趣的解决方案。我们将用复杂的和最先进的建模方法来研究我们发现的计算意义。

Pyramidal neurons are the principal excitatory cells in the brain and they receive thousands of synaptic inputs onto their elaborate dendritic arbors. Remarkable progress has been made in recent decades outlining the integrative properties of pyramidal cell dendrites, and it is now well appreciated that one key aspect of `dendritic computation' involves a wealth on non-linear mechanisms that ultimately controls how synaptic inputs are transformed into spiking activity, therefore powerfully modulating information transfer in neural circuits. A number of recent obtained in vivo further gives credence to the idea that a central component of pyramidal cell computation involves the decoding and transmission of spatiotemporal correlations in synaptic activity. These insights support a growing consensus that neurons, by virtue of their dendrites, are highly compartmentalized information processors. Dendritic excitability is governed by ion channels in the plasma membrane, acting within the constraints of passive cable properties. Activation of voltage-gated sodium channels (VCSCs) and/or voltage-gated calcium channels (VGCCs) are involved in dendritic spike generation, while the presence of potassium channels regulates key aspects of these events. Whereas the computational operations supported by dendrites are themselves highly dynamic in nature, the collective rules guiding these operations in a given neuron have historically largely (though not exclusively) been conceptualized as being static. However, neuromodulators can phasicaly modulate a vast array of active ionic conductances. As such, despite its well-recognized importance, the potential role of neuromodulator systems to dynamically and bi-directionally regulate the rules governing integrative features of dendrites throughout the brain remains largely unexplored experimentally. The disclosure of such a function would provide critical insights into the role played by these modulatory transmitters in regulating information processing in the brain. The overall objective of the current research program is :1) to identify and characterize the catalogue of dendritic operation and compartmentalization features of dendrite function by using a combination of cellular electrophysiology, two-photon (2P) imaging and uncaging of caged analogs of neurotransmitters and to identify how this set of operations is dynamically regulated by neurotransmission; 2) to understand how dendritic events are involved in a novel form of synaptic plasticity we have recently begun to investigate that binds events separated by hundreds of milliseconds. This novel plasticity offers an intriguing solution to the well-known temporal credit assignement problem of modern learning theories. We will investigate with sophisticated and state-of-the-art modeling approaches the computational meanings of our findings.