A role for mesencephalic dopamine in activation: commentary on Berridge (2006)
A role for mesencephalic dopamine in activation: commentary on Berridge (2006)
复制标题
中脑多巴胺在激活中的作用：Berridge 评论 (2006)
作者：
T. Robbins;B. Everitt
关于中枢多巴胺（DA）系统的功能的想法似乎在过去十年中有了相当大的发展。虽然基于经典神经精神药理学方法的理论争论已经变得更加复杂和完善，但主要的新概念可能来自（1）电生理学观察，即腹侧被盖区细胞的快速相位放电似乎模拟了与巴甫洛夫或时间差异学习模型相关的错误预测信号（Schultz 2002），以及（2）这些阶段性反应与相同DA系统的紧张性作用模式的相对贡献（后藤和Grace 2005）。这两个经验上的进步也与理论上的完善相匹配;例如，在人类认知神经科学领域，对DA在强化学习和“错误预测”学习中的作用的研究已经变得非常流行（参见，例如，Montague et al. 2004; Frank and O 'Reilly 2006），而紧张性-阶段性的区别已经成为建模精神病理学的主要新结构，包括基因组方法（Bilder et al. 2004）。Berridge的文章的主要主旨之一是对DA在强化学习中的作用进行了批判性分析，我们特别欣赏他试图从学习的角度区分他的叙述，例如，通过他对连续条件刺激（CS）的实验。在我们对DA在习惯学习中的作用的推测方面，我们被慷慨地分配了一个与Berridge的“激励敏感化”假说相竞争的观点。然而，我们的立场是一个更普遍的立场，在许多方面与Berridge的立场是一致的，尽管基于非常不同的证据。他对提升DA系统的“动机”作用的讨论在许多方面让我们想起了我们在之前几篇文章中所接受的概念（Robbins and Everitt 1982，1987，1992，1995; Robbins et al. 1989）。因此，我们认为中央DA系统的功能可以用“能量”结构来解释（即，解释行为输出的活力和频率）的“激活”。这种激活状态在行为（和认知）输出的调节中特别重要，必须与影响皮层处理效率的唤醒概念区分开来。正如我们在1992年对当时已有的大量经验数据的回顾中所假设的那样，（有时被混淆为“行为唤醒”）是由许多相关的状态或刺激引起的，包括食物剥夺，“压力”，精神兴奋剂药物，厌恶性刺激，如夹尾和脚电击，新奇感，CS，包括食欲事件的预测因子，如食物，以及厌恶性事件（Robbins and Everitt 1992）。这些刺激和状态的范围，顺便说一句，远远超出了中脑DA系统简单地响应错误预测信号的假设，特别是激活刺激的类别还包括某些条件下的新刺激（参见图11）。Bardo等人，1990年）。然而，我们承认，关于中脑DA神经元对不同状态和刺激的相对敏感性存在争议-通常是由于用于索引这些变化的方法的差异引起的，例如，电生理学，其对阶段性精神药理学更敏感（2007）191：433-437 DOI 10.1007/s 00213 -006-0528-7
Ideas about the functions of the central dopamine (DA) system may seem to have evolved quite considerably in the last decade. While the theoretical debate based on classic neuropsychopharmacological approaches has become much more sophisticated and refined, perhaps the major new concepts have derived from (1) electrophysiological observations that fast phasic firing of cells in the ventral tegmental area appear to model an error prediction signal relevant to Pavlovian or temporal difference learning models (Schultz 2002), and from (2) the relative contributions of such phasic responses with the tonic mode of action of the same DA systems (Goto and Grace 2005). Both of these empirical advances were also matched by theoretical refinement; for example, in the domain of human cognitive neuroscience, the study of the role of DA in reinforcement learning and “error prediction” learning has become highly fashionable (see, e.g., Montague et al. 2004; Frank and O’Reilly 2006) and the tonic–phasic distinction has been a major new construct for modeling psychopathology, including genomic approaches (Bilder et al. 2004). One of the main thrusts of Berridge’s article is a critical analysis of the role of DA in reinforcement learning, and we particularly appreciate his attempt to differentiate his account from a learning standpoint, for example, by his experiments with serial conditioned stimuli (CSs). We are graciously assigned one of the competing perspectives to Berridge’s “incentive sensitization” hypothesis in terms of our speculations about the role of DA in habit learning. However, our position is a much more general one that in many ways is in harmony with that of Berridge, although on the basis of very different evidence. His discussion of a “motivational” role for the ascending DA system reminds us in many ways of notions that we have entertained in several previous articles (Robbins and Everitt 1982, 1987, 1992, 1995; Robbins et al. 1989). Thus, we suggested that functions of the central DA systems could be explained in terms of an “energetic” construct (i.e., one that accounts for the vigor and frequency of behavioral output) of “activation.” This activational state, which is particularly important in the modulation of behavioral (and cognitive) output, has to be distinguished from concepts of arousal that affect the efficiency of cortical processing. As posited in our 1992 review of the considerable empirical data already then available, activation (sometimes confusingly called “behavioral arousal”) is induced by many related states or stimuli, including food deprivation, “stress,” psychomotor stimulant drugs, aversive stimuli such as tail-pinch and foot-shock, novelty, CSs, including predictors of appetitive events such as food, and also of aversive events (Robbins and Everitt 1992). The range of these stimuli and states, incidentally goes far beyond the hypothesis that the midbrain DA system responds simply to error prediction signals, especially as the class of activating stimuli also includes novel stimuli under certain conditions (c.f. Bardo et al. 1990). However, we acknowledge that there is controversy about the relative sensitivity of the midbrain DA neurons to different states and stimuli—often arising from differences in the methods for indexing such changes, for example, electrophysiological, which are more sensitive to phasic Psychopharmacology (2007) 191:433–437 DOI 10.1007/s00213-006-0528-7