Neuronal Control of Interactions Among Behaviors

描述（由申请人提供）：我们建议研究简单神经系统用于做出选择行为决定的细胞特性和系统级机制。在过去的27年里，我的实验室帮助描述了药用水蛭神经系统中负责四种不同行为的神经元回路：局部弯曲，全身缩短，游泳和爬行。此外，我们还发现了一些参与激活和终止这些行为的神经元。近年来，我们帮助开发了新一代的电压敏感染料，这些染料非常敏感，我们现在可以同时记录许多神经元的电活动。随着CCD摄像机灵敏度的沿着提高，以及速度更快、内存更大的计算机，我们现在可以实现直到最近还只是一个美好的希望：在水蛭神经系统中所有相关神经元的行为决策过程中进行记录。 我们之前的电生理数据表明，决策神经元并不专门用于启动单一行为，而是帮助启动两个或更多个行为，其中一些行为相互不相容（例如，缩短和游泳）。这些发现表明，决策是由组合代码做出的，其中一组决策神经元决定产生哪种行为。我们将确定这个代码的性质，并测试每个神经元在产生最终决策中的重要性。特别是，我们将测试是否所有的决策者都有同等的影响力，或者是否相反，是否决定是在neuons之间加权，以及是否决策是分层的，与一般的决定之前，更具体的。此外，我们将使用非线性动力学的工具来帮助我们理解代码，并捕捉决策过程的基本特征。此外，我们将利用我们对决策机制的了解来确定当动物学习一项复杂任务时，神经系统中发生了哪些变化，将一种反应转变为另一种反应。 水蛭的决策过程中的各种相互作用已经在更复杂的神经系统中得到了假设，包括我们自己的神经系统。由于水蛭神经系统相对简单，而且非常容易获得，而且由于上述技术的进步，我们现在能够逐个神经元地测试这些想法。

DESCRIPTION (provided by applicant): We propose to investigate the cellular properties and systems-level mechanisms used by a simple nervous system to make decisions about selecting behaviors. In the past 27 years, my laboratory has helped to characterize the neuronal circuits in the nervous system of the medicinal leech that are responsible for four different behaviors: local bending, whole-body shortening, swimming, and crawling. In addition, we have found a number of neurons that are involved in activating and terminating these behaviors. In recent years, we have helped to develop a new generation of voltage-sensitive dyes that are so much more sensitive that we can now record the electrical activity of a many neurons at once. With parallel improvements in the sensitivity of CCD cameras along with faster computers with huge memories, we can now realize what was until very recently just a fond hope: to record from all the relevant neurons in a leech nervous system as the animal makes decisions among behaviors. Our previous electrophysiological data indicated that decision-making neurons are not dedicated to the initiation of a single behavior, but instead help to initiate two or more behaviors, some of them mutually incompatible (e.g., shortening and swimming). These kinds of findings indicate that decisions are made by a combinatorial code, in which a cluster of decision-making neurons determines which behavior is produced. We will determine the nature of this code, and will test the importance of each of the neurons in producing the ultimate decision. In particular, we will test whether all decision-makers have equal influence or whether instead, whether the decision is weighted among neuons, and whether decisions are made hierarchically, with general decisions being made before more specific ones. In addition, we will use the tools of nonlinear dynamics to help us to understand the code, and to capture the essential features of the decision-making process. Also, we will use our knowledge of the mechanisms of decision-making to determine where in the nervous system changes take place that turn one response into another when an animal learns a complex task. The kinds of interactions being posited for decision-making in the leech have been postulated in more complex nervous systems, including our own. Because the leech nervous system is relatively simple and very accessible, and because of the technological advances described above, we are now in a position to test these ideas on a neuron-by-neuron basis.