Navigation with complex odor dynamics: computational principles and neural circuit implementation in mice

项目摘要 所有的动物都利用感官线索在复杂的环境中找到路，并找到重要的资源， 作为食物或伴侣。从蠕虫等简单生物在厘米尺度上寻找营养丰富的土壤 距离北极熊的鼻子在几公里以外以海豹尸体为食， 利用气味在环境中导航是这方面最古老和最普遍的例子之一 复杂的行为对气味导航能力的兴趣已经持续了几十年， 动物如何完成这一壮举的模型。测试这些模型非常困难，因为 在几乎所有的陆地环境中，气味通过湍流空气流作为波动的羽流传输。这 需要使用简化的气味流动模型或复杂的三维计算模型， 流体动力学模拟在这两种情况下，只能在性能之间建立统计联系 以及动物的行为和神经处理。这种限制排除了 将联合收割机即时神经记录与引导动物行为的感官输入相结合的能力。 这项建议代表了流体动力学、嗅觉系统、 神经科学和神经生理学直接建立用于气味引导导航的算法， 这些算法在小鼠早期嗅觉系统中的神经实现。我们将使用新开发的， 微型气味传感器，以记录气味引导导航期间小鼠鼻子处的气味羽流。通过组合 这些传感器读数与气味流的计算模型，我们将直接测试行为算法 老鼠用气味来导航。为了建立这些算法的神经实现，我们将 进行大规模的神经成像和电生理记录从早期嗅觉系统， 监测鼻子处的气味羽流输入。我们还将使用病毒标记技术来选择性地记录神经元 这些细胞将输出信号发送到特定的下游皮质结构。通过记录神经活动， 嗅觉球细胞与特定的皮质目标，我们将测试气味信息是如何从感觉路由到 支持气味导航的决策区域。最后，我们将联合收割机结合这些层次的分析， 生成一个完整的气味导航模型，将行为算法与神经网络连接起来， 实施.

PROJECT SUMMARY All animals use sensory cues to find their way through complex environments and locate vital resources such as food or mates. From simple organisms such as worms finding nutrient-rich soil at centimeter-scale distances to polar bears following their noses across kilometers to feed upon seal carcasses, the ability to navigate an environment using odors is one of the most evolutionarily ancient and widespread examples of this complex behavior. Interest in the ability to navigate with odors has spanned decades and resulted in numerous models suggesting how animals can accomplish this feat. Testing these models is extremely difficult because in nearly all terrestrial environments odors are transported as fluctuating plumes by turbulent air flow. This necessitates either the use of simplifying models for odor flow or complex three-dimensional computational fluid dynamics simulations. In both cases, only statistical connections can be made between the performance of a simulated searcher and the behavior and neural processing of an animal. This limitation rules out the ability to combine moment-to-moment neural recordings with the sensory input guiding an animal’s behavior. This proposal represents a cross-disciplinary effort between experts in fluid dynamics, olfactory systems neuroscience, and neurophysiology to directly establish the algorithms used for odor-guided navigation and the neural implementation of these algorithms in the early olfactory system of mice. We will use newly developed, miniature odor sensors to record odor plumes at the mouse nose during odor-guided navigation. By combining these sensor readings with computational models of odor flow we will directly test the behavioral algorithms used by mice to navigate with odor plumes. To establish the neural implementation of these algorithms we will perform large-scale neural imaging and electrophysiology recordings from the early olfactory system while monitoring odor plume input at the nose. We will also use viral labeling techniques to selectively record neural activity from cells that send output to specific downstream cortical structures. By recording neural activity from olfactory bulb cells with specific cortical targets we will test how odor information is routed from sensory to decision-making areas to support odor-guided navigation. Finally, we will combine these levels of analysis to generate a complete model of odor-guided navigation that connects behavioral algorithms to neural implementation.