Mechanosensory feature extraction for directed motor control

摘要 这项提议涉及两个基本问题： (1)神经元如何提取与运动控制相关的机械感觉刺激的特征？ (2)中央回路如何在机械感觉刺激和行为之间建立灵活的联系？ 这些问题与人类健康有关，因为感觉加工和感觉-运动整合在 许多神经和精神疾病。然而，感觉加工和感觉-运动整合并不完全 在细胞机制水平上理解--即在神经连接、细胞生理学和突触水平上 生理学。这种机制解释水平对于理解为什么与疾病相关的基因产生 特有的表型。这对开发更好的治疗方法也很重要。作为一种模型系统 对这些大脑功能的机械学洞察，这个项目将专注于果蝇最大的机械感觉器官 (约翰斯顿的器官)和这个器官下游的回路和行为。约翰斯顿的器官神经元(Jons)编码 远端触角节段的偏转。这些偏转可能是由于物体接触天线、风、 姿势变化，或声音。因此，从本质上讲，约翰斯顿的器官具有一系列功能--躯体感觉、前庭、 而且是听觉的。不同的乔恩刺激引起不同的行为。这些行为是可变的，并且依赖于上下文(不 刻板印象的行为模式)，因此我们可以使用这个系统来研究感觉-运动耦合中的灵活性。我们的首要目标是 来确定乔恩是如何编码机械刺激的。为了检验乔恩斯高度专门化的假设 关于天线偏转的时空特征，我们将结合体内钙成像、电生理、 和电压成像。其次，我们将使用活体全细胞记录来检验中枢神经元 Jons的突触后能通过其固有的电带通提取天线振动的特定频率 过滤特性。第三，我们将在体内进行全细胞记录，以研究机械感觉信号如何 在三阶神经元水平上进行编码，以及这些信号是如何传递到运动控制中心的。我们 假设风和声音刺激将由很大程度上不同的神经通道编码。第四，我们将结合 用同步行为测量的全细胞记录来确定来自乔恩斯的机械感觉提示 以上下文相关的方式控制行走方向。我们假设航向线索和背景线索将 汇聚在投射到腹神经索的下行运动控制神经元的水平。作为一个整体，这项工作 将为机械感觉处理和机械感觉中发生的神经计算提供新的见解- 运动集成，以及执行这些计算的细胞机制。

Summary This proposal addresses two fundamental questions: (1) How do neurons extract features of mechanosensory stimuli that are relevant for motor control? (2) How do central circuits create a flexible linkage between mechanosensory stimuli and behavior? These questions are relevant to human health because sensory processing and sensory-motor integration are disrupted in many neurological and psychiatric disorders. However, sensory processing and sensory-motor integration are not fully understood at the level of cellular mechanisms – i.e., at the level of neural connectivity, cellular physiology, and synaptic physiology. This level of mechanistic explanation is important to understanding why disease-linked genes produce their characteristic phenotypes. It is also important to developing better therapeutics. As a model system for gaining mechanistic insight into these brain functions, this project will focus on the largest mechanosensory organ in Drosophila (Johnston's organ) and the circuits and behaviors downstream from this organ. Johnston's organ neurons (JONs) encode deflections of the distal antennal segment. These deflections can result from an object touching the antenna, wind, postural changes, or sound. In essence, therefore, Johnston's organ has a range of functions – somatosensory, vestibular, and auditory. Different JON stimuli elicit different behaviors. These behaviors are variable and context-dependent (not stereotyped action patterns) and so we can use this system to study flexibility in sensory-motor coupling. Our first aim is to determine how JONs encode mechanical stimuli. To test the hypothesis that JONs are highly specialized for specific spatiotemporal features of antennal deflections, we will use a combination of in vivo calcium imaging, electrophysiology, and voltage imaging. Second, we will use in vivo whole cell recordings to test the hypothesis that central neurons postsynaptic to JONs can extract specific frequencies of antennal vibrations by virtue of their intrinsic electrical bandpass filtering characteristics. Third, we will perform in vivo whole cell recordings to investigate how mechanosensory signals are encoded at the level of third-order neurons, and how these signals are relayed to motor control centers. We hypothesize that wind and sound stimuli will be encoded by largely distinct neural channels. Fourth, we will combine whole-cell recording with simultaneous behavioral measurements to determine how mechanosensory cues from JONs steer walking direction in a context-dependent manner. We hypothesize that heading direction cues and context cues will converge at the level of descending motor control neurons that project to the ventral nerve cord. As a whole, this work will provide new insights into the neural computations that occur in mechanosensory processing and mechanosensory- motor integration, as well as the cellular mechanisms that implement these computations.