Circuit principles of demotivation in the decision to switch behaviors

项目摘要：决定开始一种新的行为通常需要终止正在进行的 行为。这意味着动物产生的许多驱动状态不仅影响神经回路 他们直接相关的行为，以及许多其他行为的基础。我的实验室已经证明 雄性果蝇的交配行为受动机控制，在现场可能会被抛弃 激发竞争动力的刺激--取决于竞争动力的相对强度。这些 行为是由多巴胺能神经元驱动的，这是人类动机控制的许多共同特征之一。 我提供的初步数据表明，多个相互竞争的驱动程序协同整合，导致男性 果蝇提前终止交配，如果不受干扰，交配将持续约23分钟。这种整合 发生在一组八个男性特有的GABA能交配去动力神经元(CDN)上，这些神经元导致 当刺激超过阈值时立即终止，并可长期集成不同的输入 时间尺度。在交配的前5分钟，即使是最严重的威胁也不能刺激CDN和 因此不会导致终止，但随着交配的进展，CDN变得更容易被不同的 抑制刺激，逐渐允许终止，以响应越来越弱的输入。我向大家介绍一份 行为特异性去动力神经元竞争输入协同整合的计算模型， 演示此电路逻辑如何提高行为稳定性或灵活性，具体取决于 个人的优势充分补充了动力状态。我还提出了一个新的假设，即行为是如何-- 随着行为目标的实现，特定的去动力神经元增加了它们的敏感度。这些 实验发现将很少研究的去动机这个话题置于行为决策的中心 我们的计算工作提出了几个新颖且可检验的假设。这笔赠款的主要目标是i) 了解来自竞争对手驱动器的信息是如何处理并传递给特定行为的 去动机回路；以及ii)了解这种信息是如何与人的动机状态相结合的 正在进行的行为，以决定是否切换行为。这项工作将建立一个新的、一线的模式 多个动机之间的高阶相互作用系统，强烈表明原则和 我们得出的模型将为理解人类的动机和决策提供一个框架。 项目相关性：这个项目探索了一个基本但未被充分研究的激励调节原则： 随着目标的实现和环境的变化，士气低落。动机失调是药物的核心 上瘾、抑郁、强迫症，以及许多其他行为和情绪障碍。健壮的 这项建议中的行为和精确操作将把神经元活动与行为联系起来，并允许直接的 因果关系的归属。收集的数据将用于扩展我们的电路和计算模型，这些模型 可以在动物和行为中概括。

Project Summary: The decision to commence a new behavior often requires termination of the ongoing behavior. This implies that the many drive states produced by an animal impact not only the neural circuits underlying their directly associated behaviors, but also those of many other behaviors. My lab has shown that the mating behaviors of male Drosophila are under motivational control and may be abandoned in the presence of stimuli evoking competing drives—depending on the relative intensities of the contending drives. These behaviors are motivated by dopaminergic neurons, one of many features shared with human motivational control. I present preliminary data demonstrating that multiple competing drives integrate synergistically to cause male Drosophila to prematurely terminate copulations that would last ~23 minutes if left undisturbed. This integration occurs at a set of eight male-specific, GABAergic Copulation Demotivating Neurons (CDNs) that cause immediate termination when stimulated beyond a threshold and can integrate diverse inputs over long timescales. During the first 5 minutes of mating, even the most severe threats cannot stimulate the CDNs and therefore do not cause termination; but as the mating progresses the CDNs become more accessible to diverse demotivating stimuli, gradually permitting termination in response to weaker and weaker inputs. I present a computational model for synergistic integration of competing inputs at behavior-specific demotivating neurons, demonstrating how this circuit logic can promote either behavioral stability, or flexibility, depending on the individual strengths of the full complement of drive states. I also propose a novel hypothesis for how behavior- specific demotivating neurons increase their sensitivity as the goals of the behavior are achieved. These experimental findings place the rarely studied topic of demotivation at the center of behavioral decision making and our computational work suggests several novel and testable hypotheses. The main goals of this grant are i) to understand how information from competing drives is processed and delivered to behavior-specific demotivation circuitry; and ii) to understand how this information is integrated with the motivational state of the ongoing behavior to decide whether or not to switch behaviors. This work will establish a new, front-line model system for high-order interactions between multiple motivations, with strong indications that the principles and models we derive will provide a framework for understanding motivations and decision making in humans. Project Relevance: This project explores a fundamental but understudied principle of motivational regulation: demotivation as goals are achieved and circumstances change. Dysregulation of motivation is central to drug addiction, depression, compulsive disorders, among many other behavior and mood disorders. The robust behaviors and precise manipulations in this proposal will relate neuronal activity to behavior and allow a direct attribution of causality. The data collected will be used to extend our circuit and computational models that are generalizable across animals and behaviors.