CRCNS: Active Sensing and Odor Processing in the Olfactory Bulb

描述（由申请人提供）：感觉是一个活跃的过程，涉及动物与环境的动态相互作用，这导致感觉神经元和中枢处理阶段的神经活动的动态模式，最终构成感知。该项目将研究清醒动物的感觉系统功能，利用嗅觉系统了解刺激采样的动态、嗅觉系统感觉输入的时间结构之间的关系，以及这两个因素如何影响嗅球（哺乳动物嗅觉系统信息处理的第一阶段）的处理。时间动态活动模式是许多不同层次嗅觉处理的一个突出特征，慢（100 - 500毫秒）和快（10 - 50毫秒）时间尺度的动态活动是许多气味编码模型的核心。在所有这些水平上，气味诱发活动的大部分动态都是围绕呼吸循环构建的，呼吸循环控制着气味进入感觉神经元本身。然而，感官输入的时间结构如何影响气味表征和信息处理尚不清楚。此外，在动物的行为中，呼吸本身是高度动态的，动物在气味取样（即嗅探）过程中主动改变个体嗅探的频率、幅度和形状。迄今为止，很少有研究调查了这些取样行为的变化——以及嗅觉驱动的感官输入——对高阶嗅觉处理的影响。与此同时，最近的研究指出嗅球肾小球层的电路在通过二尖瓣细胞形成球输出方面的重要性。这些新描述的回路尚未被纳入生物物理现实的嗅球功能计算模型中。该项目将结合实验和计算方法来解决两个核心问题：

DESCRIPTION (provided by applicant): Sensation is an active process involving dynamic interactions of the animal with its environment, which leads to dynamic patterns of neural activity in sensory neurons and in the central processing stages that ultimately underlie perception. The proposed project will investigate sensory system function in the awake animal, using the olfactory system to understand the relationship between the dynamics of stimulus sampling, the temporal structure of sensory inputs to the olfactory system, and how these two factors shape processing in the olfactory bulb - the first stage of information processing in the mammalian olfactory system. Temporally dynamic activity patterns are a prominent feature of many different levels of olfactory processing, and dynamic activity at both slow (100 - 500 msec) and fast (10 - 50 ms) timescales is central to many models of odor coding. Much of the dynamics of odorant-evoked activity - at all of these levels - is structured around the respiratory cycle, which controls the access of odorant to the sensory neurons themselves. However, it remains unclear how the temporal structure of sensory input shapes odor representations and information processing. In addition, in the behaving animal respiration itself is highly dynamic, with animals actively changing the frequency, amplitude, and shape of individual sniffs during odor sampling (i.e., sniffing). Very few studies to date have examined the consequences of these changes in sampling behavior - and in the sniff-driven sensory input - for higher-order olfactory processing. At the same time, recent studies have pointed to the importance of circuitry in the glomerular layer of the olfactory bulb in shaping bulb output via mitral cells. These newly-described circuits have yet to be incorporated into a biophysically-realistic, computational model of olfactory bulb function. This project will use a combination of experimental and computational approaches to address two central questions: 1. How does the glomerular circuitry transform realistic patterns of sensory input into patterns of mitral cell output? 2. What are the consequences of the olfactory bulb transformation for the representation of odorants by mitral cell populations? Intellectual Merit: The project is a joint experimental and computational effort which uses spatiotemporal patterns of sensory neuron activity recorded from awake, behaving rodents as inputs to computational models of the olfactory bulb network. The project will also use electrophysiological and behavioral measures to test model predictions. This approach is unique and significant in that it uses - for the first time - natural sensory inputs to a model of olfactory bulb function, and also because it incorporates - for the first time - a realistic model of the circuitry around the olfactory bulb glomerulus, many features of which have only recently been described. The experiments are designed to provide a picture of how odor information is transformed at the first stage of synaptic processing and how this transformation is actively shaped by the animal's own sampling behavior. Broader Impacts: The proposed work should lead to important insights into how sensation can be actively modulated by sensory acquisition behavior. A potentially important biomedical impact of this work is the development of improved prosthetic devices - for example, artificial limbs which use sensor-driven proprioceptive information to help control movement. Another potential impact is the improved design of sensor devices for detecting analytes in a complex environment. The general concept of using naturalistic sensory information as inputs to model circuits is an idea that could also lead to important breakthroughs in understanding sensory processing in other modalities. Generation of the first circuit model of the recently-described glomerular network in the olfactory bulb will be an important shared resource for others in the field who wish to characterize computations at the first few synapses in the olfactory system.