Interactions between piriform cortex and amygdala in a model of associative learning in the mammalian olfactory system

描述（由申请人提供）：生物体的嗅觉系统必须在其环境中创造气味的感官表征，并使用这些来驱动适当的行为。要做到这一点，嗅觉系统必须学会将气味身份与其行为后果联系起来。最近的实验进展表明梨状皮质和杏仁核是气味学习的场所。梨状皮质中分布的神经元集合形成气味的表征，而杏仁核中的神经元集合则介导行为。为了学会在气味刺激下做出适当的行为，嗅觉系统必须重新组织梨状皮质-杏仁核通路中的连接，以便气味呈现导致适当的杏仁核整体激活。需要实验观察和生物约束模型的相互作用来指导未来的实验，并在这个系统中建立对联想学习的理论理解。通过实验合作者的数据，我们将开发一个梨状皮质反应模型，以确定其连接的随机性和结构如何导致气味刺激的可靠表征。然后，我们将扩展这个模型，包括杏仁核的投射，并确定这些投射的可塑性如何导致气味和食欲或厌恶行为反应之间的关联。将开发实验范例来测试和验证模型预测。这种理论/实验方法将导致在这个系统中的联想学习的机械理解。这是一个令人兴奋的前景，因为大多数感觉系统模型关注的是它们对输入的反应的结构，而没有扩展到这些反应如何指导学习行为。这些结果将作为系统和行为神经科学之间的联系而广泛相关。他们还将推进我们对感官知觉和情绪行为之间关系的理解，与涉及情绪和药物滥用的疾病相关。

DESCRIPTION (provided by applicant): An organism's olfactory system must create sensory representations of odors in its environment and use these to drive appropriate behaviors. To do so, the olfactory system must learn to associate odor identities and their behavioral consequences. Recent experimental advances have implicated the piriform cortex and the amygdala as sites of odor learning. Distributed ensembles of neurons in piriform cortex form representations of odors, while ensembles in amygdala mediate behavior. To learn to perform appropriate behaviors given an odor stimulus, the olfactory system must reorganize connections in the piriform cortex-amygdala pathway so that odor presentation leads to activation of the appropriate amygdalar ensemble. Interplay of experimental observations and biologically constrained models is needed to guide future experiments and develop a theoretical understanding of associative learning in this system. With data from experimental collaborators, we will develop a model of piriform cortex responses to determine how randomness and structure in its connectivity leads to reliable representations of odor stimuli. We will then exten this model to include projections to amygdala and identify how plasticity in these projections can lead to associations between odor and appetitive or aversive behavioral responses. Experimental paradigms will be developed to test and verify model predictions. This theoretical/experimental approach will lead to a mechanistic understanding of associative learning in this system. This is an exciting prospect, as most models of sensory systems focus on the structure of their re- sponses to inputs, without extension to how these responses guide learned behavior. The results will be broadly relevant as a link between systems and behavioral neuroscience. They will also advance our understanding of the relationship between sensory perception and emotional behavior, with relevance for disorders involving emotion and substance abuse.