Large-scale, neuronal ensemble recordings in motor cortex of the behaving marmoset

摘要 该项目旨在表征运动皮层（M1）活动的时空组织在多个 空间尺度与无约束的绒猴上肢运动有关， 行为学行为该项目有两个目标：1）统计评估单一的性质和稳定性 神经元和合奏水平的运动代表在M1的列和面积的空间尺度，和2） 使用我们的实验数据来开发M1的3D补丁的网络模型， 关于M1中运动表征的实验可检验的预测。我们将联合收割机 大规模神经记录的补充技术：1）无线，高密度多电极阵列 和2）钙荧光成像-而普通绒猴（Callithrix jacchus） 自然觅食行为微电极阵列技术的进步允许同时 从行为动物的数百个神经元的电生理记录。然而，由于大规模的... 电极距离（>= 400微米），在亚柱状水平的许多微电路活动是未分辨的。在 相比之下，钙荧光成像提供了密集和同时记录钙荧光的机会。 一个皮层柱内数百个神经元的尖峰活动。这种密集的大规模成像技术 对于彼此紧邻的神经元的分辨率，这增加了它们被 突触连接。我们将使用一个微型荧光显微镜连接到头骨， 手臂和手的头部自由、不受约束的运动。此外，通过在其中添加棱镜透镜 在显微镜下，我们将能够对从层2/3到层5的跨层神经元进行成像。同时使用 技术，我们将表征单神经元编码特性，网络动力学和功能 皮质柱内部和之间的连接。通过连接空间尺度，我们将能够 皮层微电路水平和整个皮层区域水平之间的差异。我们还将研究 运动编码的时空组织随着运动技能的获得而改变。独特而重要的 该项目的特点是使用自然和不受约束的觅食任务，包括捕获猎物 这将不需要操作性条件反射，并将提供更丰富的行为，以建立更准确的 编码模型我们还将建立一个大规模的运动皮层的网络模拟， 记录的数据，以了解连接如何与单个神经元的调谐特性相关。该模型将 然后让我们研究什么样的突触规则导致了所观察到的时空模式的变化 与运动学习有关。最终，网络动力学、计算和编码的原理 从运动皮层推导出的理论可能更普遍地适用于其他新皮层区域。这项研究也可能 已经将相关性应用到脑机接口技术中。

Abstract This project seeks to characterize the spatio-temporal organization of motor cortical (M1) activity at multiple spatial scales associated with upper limb movements of unrestrained marmoset monkeys performing ethological behaviors. The project has two goals: 1) To statistically evaluate the nature and stability of single neuron and ensemble-level motor representations in M1 at the columnar and areal spatial scales, and 2) To use our experimental data to develop a network model of a 3D patch of M1 capable of generating experimentally testable predictions about the movement representations in M1. We will combine two complementary technologies for large-scale neural recording: 1) wireless, high density multi-electrode arrays and 2) calcium fluorescence imaging - while common marmoset monkeys (Callithrix jacchus) perform naturalistic foraging behaviors. Advances in microelectrode array technology have permitted simultaneous electrophysiological recordings from hundreds of neurons in behaving animals. However, given the large inter- electrode distance (>=400 microns), much of the microcircuit activity at the subcolumnar level is unresolved. In contrast, calcium fluorescence imaging provides the opportunity to densely and simultaneously record the spiking activity of hundreds of neurons within a single cortical column. This dense, large-scale imaging allows for the resolution of neurons immediately adjacent to one another which increases the likelihood that they are synaptically connected. We will use a miniature fluorescence microscope attached to the skull which allows for head-free, unconstrained movements of the arm and hand. Moreover, by adding a prism lens to the microscope, we will be able to image neurons across lamina from layer 2/3 through layer 5. Using both technologies, we will characterize single neuron encoding properties, network dynamics, and functional connectivity within and between cortical columns. By bridging spatial scales, we will be able to interpolate between the cortical microcircuit level and the level of a whole cortical area. We will also investigate how the spatio-temporal organization of movement coding changes with motor skill acquisition. A unique and important feature of this project will be the use of natural and unconstrained foraging tasks that involve prey capture which will not require operant conditioning and will provide richer behaviors in order to build more accurate encoding models. We will also build large-scale network simulations of a patch of motor cortex constrained by the recorded data to understand how connectivity relates to tuning properties of single neurons. The model will then allow us to investigate what synaptic rules result in the observed changes in spatiotemporal patterning associated with motor learning. Ultimately, the principles of network dynamics, computation, and encoding deduced from the motor cortex may apply more generally to other neocortical areas. This research may also have applied relevance to brain-machine interface technology.