Characterizing in-task corticostriatal circuit operation during habit learning

描述(申请人提供)：纹状体是一种进化上古老的大脑结构，在大脑的任何结构中接受最多的多巴胺输入(41，43)。尽管出现了额叶新皮质区域，但它仍然是大部分基于内隐奖赏的习惯学习的主要结构，这种学习对日常生活至关重要，但可能会变得不适应，导致成瘾、强迫行为和其他障碍(11，16，24，31，32)。根据最近的发现，我们现在认为，背侧纹状体可能存在神经活动模式，这是习惯形成的标志。在背内侧纹状体(DMS)，神经反应在习惯学习的早期出现在任务的决策点(45)。在背外侧纹状体(DLS)，神经反应出现在学习的后期，集中在运动习惯的开始和结束(9，10，27，45)。这些模式与病变研究(6，15，38，49，50，51，53)确定的DMS和DLS的作用是一致的，是很有希望作为习惯形成的神经基础的候选者；然而，这些模式代表的是什么仍不清楚。更重要的是，我们无法了解它们的功能，除非我们考虑它们在皮质纹状体回路中的作用，最终是整个皮质-纹状体-苍白球-丘脑回路，也就是涉及纹状体的延髓大脑回路(1，42)。到目前为止，研究只记录了这些回路中孤立的位置，没有解决不同区域在任务内如何相互作用的问题，也没有专注于探索行为之外的回路的解剖结构。为了弥合这一差距，我们必须通过定义电路中每个点的任务反应如何出现来解构这些电路。在拟议的研究中，我将探讨前面描述的DMS和DLS激活模式是否适用于一系列习惯性任务，并使用一种新的杠杆按压序列任务来阐明它们与运动习惯行为的关系。然后我将阐述这种活动是如何在皮质纹状体环路中出现的。基于已有的知识，即皮质为纹状体提供主要的兴奋性输入(1，42)，投射到DMS和DLS的皮质区域不同(13，35，39)，皮质纹状体的可塑性对于习惯学习是必要的(14，52，53)I假设皮质输入正在驱动纹状体任务反应，皮质纹状体突触强度的塑造与经验产生纹状体任务反应的学习相关变化。为了验证这一假设，我将同时记录皮质部位及其目标纹状体部位的单个单位活动，同时使用皮质输入的局部光遗传抑制来识别哪些纹状体任务反应依赖于这一输入。这种结合的方法将帮助我不仅确定哪些纹状体任务反应是由皮质输入驱动的，而且还将帮助我确定皮质神经元的活动可能如何驱动它们。这些实验将是第一次开始构建行为中皮质纹状体回路功能机制的模型，将任务和学习动力结合在一起。最终，对习惯学习回路的深入理解将是我们长期目标的关键，即找到精确的回路操作，以减轻当习惯变得过于固定时发生的令人衰弱的适应不良行为。

DESCRIPTION (provided by applicant): The striatum is an evolutionarily ancient brain structure that receives the most dopamine input of any structure in the brain (41,43). Despite the emergence of frontal neocortical areas, it remains the primary structure responsible for much of implicit reward-based habit learning that is essential for everyday life but can become maladaptive resulting in addiction, compulsive behaviors, and other disorders (11,16,24,31,32). Based on recent findings, we now believe that there may be neural activity patterns in dorsal striatum that are signatures of habit formation. In dorsomedial striatum (DMS), neural responses are seen early in habit learning during decision points of a task (45). In dorsolateral striatum (DLS), neural responses are seen later in learning and concentrated at the initiation and termination of motor habits (9,10,27,45). These patterns are consistent with the roles of DMS and DLS defined by lesion studies (6,15,38,49,50,51,53) and are promising candidates as the neural bases of habit formation; however, it is still unclear what these patterns represent. More importantly, we cannot understand their function unless we consider their role within the corticostriatal circuit and eventually the entire cortico-striatal-pallidal-thalamic circuit, the pimary brain circuit involving striatum (1,42). So far studies have only recorded from isolated sites within these circuits without addressing the how the different regions interact in- task or have focused on exploring the anatomy of the circuit outside of behavior. In order to bridge this gap, we must deconstruct these circuits by defining how the task responses at each point in the circuit emerge. In the proposed studies, I will address whether the previously described DMS and DLS activation patterns generalize across a range of habitual tasks and clarify their relationship to motor habit behavior using a novel lever press sequence task. I will then address how this activity emerges within the corticostriatal circuits. Based on existing knowledge that cortex provides major excitatory input to striatum (1,42), that the cortical areas projecting to DMS and DLS are different (13,35,39), and that corticostriatal plasticity is necessary for habit learning (14,52,53) I hypothesize that cortical input is driving the striatal task responses and tht sculpting of corticostriatal synaptic strengths with experience produces the learning related changes in striatal task responses. To test this hypothesis I will record single unit activity simultaneously in cortical sites and their target striatal sites while using local optogenetic inhibition of the cortical input to identify which of the striatal task responses are dependent upo this input. This combined approach will help me determine not only which striatal task responses are driven by cortical input, but how the activity of neurons in cortex may be driving them. These experiments would be the first to begin constructing a model of the mechanisms of function of corticostriatal circuits within behavior incorporating task and learning dynamics. Ultimately, developing in-depth understanding of habit learning circuits will be key to our long term goal of finding precise circuit manipulations to alleviate debilitating maladaptive behaviors that occur when habits become too fixed.