Microscale organization and sensory coding in L2_3 of mouse somatosensory cortex

 描述(申请人提供)：大脑皮层的神经回路如何平衡局部回路计算与中远程整合尚不清楚。一个强有力的例子是初级感觉皮层的2/3层(L)，在那里，局部网络接受来自其直接皮质柱的密集神经支配，以及跨柱、远端皮质、神经调制和其他输入。然而，L2/3是否代表啮齿动物的非局部(整合)感觉特征仍不清楚，啮齿动物是皮质回路功能和疾病的主要动物模型。本项目利用双光子群体钙成像、神经生理学和光遗传学研究了小鼠躯体感觉皮质(S1)L2/3的局部和整合感觉特征的神经编码及其柱内和柱间的微观组织。S1的大部分先前工作都集中在非常局部的、单一胡须感觉特征的表示及其在腰椎内局部环路中的基础上。在目标1中，我们证明了尽管S1具有强大的解剖柱状结构，但仍存在高度分布的胡须感受野的盐和胡椒微组织。我们量化了这种组织，跨层测试了它的起源，并测试了自然胡须体验如何影响这种结构，得出了令人惊讶的发现 这种丰富的经验导致形成了一张地形更精确的亚柱状图。在目标2中，我们研究了L2/3中2-晶须序列(最简单的多晶须模式)的神经编码，并测试了一个新的假设，即序列表示如何在列内和列间在空间上组织。在目标3中，我们测试了从听觉和视觉皮质到S1的L2/3的显著皮质输入是否允许L2/3神经元在学习过程中获得跨模式(非胡须)感觉反应。初步数据显示，对于预测奖励的音调，S1L2/3的音调反应具有很强的获得性。我们描述了这一现象，测试了单个S1神经元是否学习编码特定的非胡须刺激或一般的奖励预期，并使用光遗传学来测试这些学习到的跨模式反应是否由来自其他感觉皮质的皮质-皮质投射介导。总体而言，这些实验将揭示L2/3内新的整合功能和组织原则。了解S1的这些特征将有助于未来测试自闭症和精神分裂症小鼠模型的潜在异常，这两种障碍可能反映了大脑皮层局部兴奋性和远程整合之间的失衡。

 DESCRIPTION (provided by applicant): How neural circuits in cerebral cortex balance local circuit computation with intermediate- and long-range integration is unclear. A powerful example is layer (L) 2/3 of primary sensory cortex, where local networks receive dense innervation from their immediate cortical column, as well as cross-columnar, distant cortico- cortical, neuromodulatory and other input. However, it remains unclear whether L2/3 represents nonlocal (integrative) sensory features in rodents, which are dominant animal models for cortical circuit function and disease. This project uses 2-photon population calcium imaging, neurophysiology, and optogenetics to study the neural coding of local and integrative sensory features in L2/3 of mouse somatosensory cortex (S1), and its micro-organization within and across columns. Most prior work in S1 has focused on representation of very local, single-whisker sensory features and their basis in intracolumnar local circuits. In Aim 1, we demonstrate that despite the strong anatomical columnar structure in S1, there is a highly distributed salt-and-pepper micro-organization of whisker receptive fields. We quantify this organization, test its origin across layers, and test how natural whisker experience affects this structure, with the surprising finding that enriched experience causes a more topographically precise subcolumnar map to form. In Aim 2, we study neural coding for 2-whisker sequences (the simplest multi-whisker pattern) in L2/3, and test a novel hypothesis for how sequence representation may be organized spatially within and across columns. In Aim 3, we test whether prominent cortico-cortical input to L2/3 of S1 from auditory and visual cortex allows L2/3 neurons to acquire cross-modal (non-whisker) sensory responses during learning. Preliminary data show robust acquisition of tone responses in L2/3 of S1 for tones that predict reward. We characterize this phenomenon, test whether individual S1 neurons learn to encode specific non-whisker stimuli or the general expectation of reward, and use optogenetics to test whether these learned cross-modal responses are mediated by cortico-cortical projections from other sensory cortices. Overall, these experiments will reveal novel integrative functions and organizational principles within L2/3. Understanding these features in S1 will allow future tests of potential abnormalities in mouse models of autism and schizophrenia, two disorders that may reflect an imbalance between local excitability and long- range integration in cerebral cortex.