Cortico-subcortical interactions via the thalamic reticular nucleus for visual adaptive sensing

哺乳动物的视觉依赖于神经处理阶段的层次结构，其涉及皮层、第一和高阶视觉丘脑以及丘脑网状核（visTRN）的视觉部分之间的多个回路。TRN被称为“通往皮层的大门的守护者”，是丘脑核团抑制的主要来源，因此可以在不同的时空尺度上对丘脑活动施加强大的控制。例如，在视觉中，通过背侧丘脑外侧膝状体核（dLGN）的感觉信号被大量转换，这可能需要多功能的抑制控制。除了感官视觉处理之外，visTRN的其他众所周知的效应也可能需要多功能抑制控制：行为状态和注意力选择对丘脑反应的调制。复杂的抑制机制的存在是支持最近发现的子网络内visTRN：遗传定义的神经元，是地形和生理上不同的，并连接到第一和高阶视觉丘脑核。这些visTRN子网络如何促进视觉感觉处理，它们的反应如何依赖于离皮质输入，以及它们如何在行为背景下塑造丘脑活动，目前尚不清楚。在这里，我们建议通过结合最先进的体内电生理记录，用于细胞类型特异性和输入映射的光遗传学操作，病毒追踪，行为和计算分析来解决这些问题。具体而言，我们将侧重于三个目标。在目标1中，我们将研究visTRN神经元亚群如何差异化地代表视觉信息。在目标2中，我们将通过子网络分辨率确定visTRN的离皮质输入，以及子网络中的响应如何通过来自视觉皮层的反馈形成。在目标3中，我们将在行为的背景下测量这些visTRN神经元亚群的活性，并研究视觉任务学习过程中visTRN反应的动力学。拟议的项目有望推进我们对visTRN如何有助于图像形成视觉和最终自适应传感的理解。更一般地说，它将揭示复杂的皮质-皮质下回路如何在行为背景下塑造视觉感觉处理。

Mammalian vision relies on a hierarchy of neural processing stages, which involve multiple loops between cortex, first- and higher-order visual thalamus, and the visual sector of the thalamic reticular nucleus (visTRN). The TRN, famously known as the "guardian of the gateway to the cortex", is a major source of inhibition for thalamic nuclei, and can thus exert powerful control over thalamic activity at different spatiotemporal scales. In vision, for instance, sensory signals passing through the lateral geniculate nucleus of the dorsal thalamus (dLGN) are substantially transformed, which likely requires versatile inhibitory control. Beyond sensory visual processing, versatile inhibitory control might also be required for other, well-known effects of visTRN: modulation of thalamic responses by behavioral state, and attentional selection. The existence of complex inhibitory mechanisms is supported by recent discoveries of subnetworks within visTRN: groups of genetically defined neurons that are topographically and physiologically distinct, and connect with first- and higher-order visual thalamic nuclei. How these visTRN subnetworks contribute to visual sensory processing, how their responses depend on cortico-fugal inputs, and how they shape thalamic activity in the context of behavior, is currently unknown. Here, we propose to address these questions by combining state-of-the art in vivo electrophysiological recordings, optogenetic manipulations for cell-type specificity and input mapping, viral tracing, behavior, and computational analyses. Specifically, we will focus on three objectives. In Objective 1, we will investigate how subpopulations of visTRN neurons differentially represent visual information. In Objective 2, we will determine, with subnetwork resolution, visTRN’s corticofugal inputs, and how responses in the subnetworks are shaped by feedback from visual cortex. In Objective 3, we will measure the activity of these subpopulations of visTRN neurons in the context of behavior, and investigate the dynamics of visTRN responses during learning of a visual task. The proposed project promises to advance our understanding of how the visTRN contributes to image-forming vision and ultimately adaptive sensing. More generally, it will reveal how intricate cortico-subcortical loops shape visual sensory processing in the context of behavior.