CRCNS: Understanding Single-Neuron Computation Using Nonlinear Model Optimization

CRCNS：使用非线性模型优化理解单神经元计算

• 批准号: 10612187
• 负责人: FABRIZIO GABBIANI
• 金额: $ 33.57万
• 依托单位: BAYLOR COLLEGE OF MEDICINE
• 依托单位国家: 美国
• 财政年份: 2022
• 资助国家: 美国
• 起止时间: 2022-08-01 至 2027-07-31
• 项目状态: 未结题
• 来源: https://reporter.nih.gov/project-details/10612187
• 关键词: Action Potentials; Address; Agreement; Anatomy; Apical; Axon; Behavior; Behavioral; Biological Models; Biophysical Process; Coupling; Crabs; Data; Dendrites; Detection; Digestion; Drosophila genus; Electrophysiology (science); Environment; Generations; Grasshoppers; Hippocampus (Brain); Hodgkin-Huxley model; Hot Spot; In Vitro; Ion Channel; Language; Membrane Potentials; Methods; Modeling; Motivation; Movement; Neocortex; Neurons; Non-linear Models; Periodicity; Phase; Play; Potassium Channel; Pyramidal Cells; Reporting; Rodent; Role; Spatial Distribution; Specificity; Synapses; Synaptic plasticity; System; Trees; Visualization; base; biophysical analysis; density; detector; information processing; neocortical; neuronal cell body; novel; patch clamp; place fields; simulation

Project Description 1 Motivation and Objectives Why are ion channels localized in subcellular dendritic compartments and is there a tight coupling of the observed localization with neuron function? We argue that this fundamental question [55, 44] can be addressed by studying the biophysical mechanisms of single neuron computation in two model systems where a large amount of electrophysiological and anatomical data is available and has been tied to the functional roles of key neurons. The neurons selected, CA1 hippocampal pyramidal cells and the lobula giant movement detector (LGMD) neuron of grasshoppers, are ideal because we know precisely their role in the emergence of place fields and collision detection, respectively. Furthermore, the involved dendritic ion channels are closely related and can be studied jointly using the common language of compartmental modeling and the Hodgkin-Huxley formalism [97, 56]. Complementarity will allow to draw broader conclusions than by studying either system in isolation. 1.1 Channel Localization and Single Neuron Computation Abundant evidence suggests that ion channels are precisely localized within single neurons. Yet the role of this localization for neuronal information processing remains largely unexplored. The best-known example of zonal channel localization is the axon initial segment, where high densities of Na+ and K+ channels over a short distance play a pivotal role in the generation of action potentials [64]. In dendrites, a variety of conductances are localized in specific dendritic subregions, with densities that often depend on the distance from the spike initiation zone (SIZ) [77, 72]. Channel localization has been studied in specific neuron types such as pyramidal cells of the hip- pocampus and neocortex in rodents through in vitro patch-clamp recordings along the main apical den- drite. These recordings show the presence of Na+ channels that help relay synaptic inputs towards the soma and help action potentials backpropagate in dendrites [99, 66]. Additionally, Ca2+ channel ‘hot spots’ help trigger dendritic spikes favoring non-linear amplification of localized synaptic inputs in layer 5 (L5) neocortical pyramidal cells [63, 73]. In several types of neurons, an increase in HCN channel density away from the SIZ favors consistent synaptic summation across the dendritic tree [71]. Further, a concomitant increase in the density of inactivating K+ channels helps fine tune the role of HCN chan- nels during synaptic integration in hippocampal pyramidal cells [21]. These results have been confirmed through simulations, but their significance for information processing remains elusive. The spatial distribution of channels within dendrites has also been investigated using immunostaining, a method that reveals the localization of ion channels but that is not always in quantitative agreement with electrophysiological methods [67, 70, 40]. In Drosophila, novel methods allows visualization of Na+ channel distributions based on genetically encoded fluorescent markers [90], but the function of ion channels for information processing in single neurons is only beginning to be studied. In the few examples highlighted above, we know little about how constrained ion channel distributions are. This issue has been investigated in the stomatogastric system (STG) of crabs, where a small network of identified neurons generates rhythmic membrane potential oscillations involved in various phases of digestion. In STG neurons, substantial variability in ion channel expression levels has been observed [68, 95, 22]. Simulations confirmed that there exists a large redundancy in neuronal peak conductance levels explaining the STG’s rhythmic behavior [89]. These simulations used point-model neurons lacking dendrites and thus did not address the specificity of dendritic ion channel localization. Thus, little is currently known on the contribution of subcellular ion channel localization to single neuron computation. 1.2 CA1 Pyramidal Cells and Place Fields Behavioral timescale synaptic plasticity (BTSP) was recently reported to underlie the formation of place fields in CA1 pyramidal cells of rodents during spatial exploration of an environment [16]. This plasticity 33

项目描述 1动机和目标 为什么离子通道位于亚细胞树突状区室， 观察到的神经元功能定位？我们认为，这个基本问题[55，44]可以是 通过研究两个模型系统中单神经元计算的生物物理机制来解决 其中大量的电生理学和解剖学数据是可用的，并且已经与 关键神经元的功能。所选神经元、海马CA 1区锥体细胞和小叶 蚱蜢的巨型运动检测器（LGMD）神经元是理想的，因为我们确切地知道它们的作用 在出现的地点字段和碰撞检测，分别。此外，参与的树突状细胞 离子通道是密切相关的，可以共同研究使用共同的语言， 模型和Hodgkin-Huxley形式主义[97，56]。互补性将允许更广泛地 而不是孤立地研究任何一个系统。 1.1通道定位与单神经元计算 大量的证据表明，离子通道精确地定位在单个神经元内。然而， 神经元信息处理的这种定位仍然在很大程度上未被探索。最著名的 带状通道定位的例子是轴突起始段，其中高密度的Na+和K+ 短距离通道在动作电位的产生中起着关键作用[64]。在树突中， 各种电导分布在特定的枝晶子区域，其密度通常取决于 距离尖峰起始区（SIZ）[77，72]。 通道定位已经在特定的神经元类型中进行了研究，例如髋关节的锥体细胞。 通过体外膜片钳记录沿着主要顶窝， drite。这些记录显示了Na+通道的存在，这些通道有助于将突触输入传递到突触。 索马和帮助动作电位在树突中反向传播[99，66]。此外，Ca 2+通道“热” 斑点有助于触发树突棘波，有利于层中局部突触输入的非线性放大 5（L5）新皮质锥体细胞[63，73]。在几种类型的神经元中，HCN通道的增加 远离SIZ的密度有利于树突树上一致的突触总和[71]。此外，本发明还 伴随的K+通道失活密度的增加有助于微调HCN改变的作用， 海马锥体细胞突触整合过程中的神经元[21]。这些结果得到了证实。 但它们对信息处理的意义仍然难以捉摸。 树突内通道的空间分布也已经使用免疫染色进行了研究， 一种揭示离子通道定位的方法，但并不总是定量一致 电生理学方法[67，70，40]。在果蝇中，新的方法允许可视化 Na+通道分布基于遗传编码的荧光标记[90]，但离子通道的功能 单神经元中的信息处理通道才刚刚开始研究。 在上面强调的几个例子中，我们对受约束的离子通道分布 是.这个问题已在螃蟹的胃肠道系统（STG）中进行了研究，其中有一个小型网络 识别出的艾德神经元产生节律性的膜电位振荡，参与了 消化.在STG神经元中，已经观察到离子通道表达水平的实质性变化 [68，95，22]。模拟证实，神经元峰值电导存在较大的冗余 解释STG的节律行为的水平[89]。这些模拟使用了点模型神经元， 树突，因此没有解决树突状离子通道定位的特异性。因此， 目前已知亚细胞离子通道定位对单神经元计算的贡献。 1.2 CA 1金字塔单元格和放置字段 行为时间尺度突触可塑性（BTSP）是位置形成的基础 在空间探索环境期间啮齿动物的CA 1锥体细胞中的场[16]。这种可塑性 33