Influence of GABA on reinforcement learning in individuals with current and remitted depression

 描述(由申请人提供)：严重抑郁障碍(MDD)是导致全球疾病负担的主要原因之一，其特点是复发率高。尽管进行了几十年的研究，但研究结果并未转化为可用于测试新的治疗干预措施的可靠的MDD模型。这种适度的进展可能部分源于MDD的异质性，以及动物模型无法探测MDD的关键症状，包括抑郁情绪、内疚或沉思。在强化学习(RL)理论的背景下，快感缺乏、情绪低落、无法做出决定和过度内疚可能是由于对奖惩的学习不足以及未能相应地更新行为。与这一理论一致，MDD个体表现出奖励减少而惩罚学习增加的现象。值得注意的是，缓解后奖赏反应的降低可能会持续，而惩罚处理在症状减轻时会正常化，这表明异常的奖赏而不是惩罚，学习可能代表了MDD的一个特征标记。尽管有这些有希望的线索，但这些功能障碍背后的神经生物学机制仍然未知。重要的是，动物研究表明腹侧被盖区(VTA)的多巴胺(DA)传递和来自外侧缰核(LHb)的伽马氨基丁酸(GABA)信号以及这些系统之间的相互作用影响RL。这些相互作用调节DA的释放和基底节(BG)的活动，从而帮助有机体选择或避免某个动作。尽管动物研究为这些相互作用提供了令人信服的证据，但这些过程尚未在MDD中进行测试。这项拟议的研究的目标是填补这一关键空白，并使我们离开发MDD中RL功能障碍的机制模型更近一步。为了实现这一目标，我们将整合一种创新的多模式分子和功能神经成像方法来测试60名受试者(20名HC，20名MDD和20名RMDD)的RL。为此，将使用一种新的多体素MRS技术测量BG中的基线GABA水平，然后在参与者完成社交RL任务的同时进行fMRI会议。这将使我们能够研究MDD学习障碍的神经相关性以及GABA对RL的影响。我们假设控制组在BG中表现出奖赏学习信号，而在LHB中表现出惩罚信号。与我们提出的MDD是一种RL障碍的假设一致，我们假设MDD将分别在BG和LHb显示钝化的奖赏和惩罚信号。MDD的这种异常模式将分别与BG和LHb中强化的惩罚和奖励信号有关。此外，通过对学习障碍的状态/特质效应的分析，我们假设学习缺陷在rMDD中是奖励性的，而不是惩罚性的。最后，我们假设基线GABA水平将预测所有组的RL。

 DESCRIPTION (provided by applicant): Major Depressive Disorder (MDD) is one of the leading causes of disease burden worldwide and is characterized by a high relapse rates. In spite of decades of research, findings have not translated into a reliable model of MDD that can be used to test novel therapeutic interventions. This modest progress might partially stem from the heterogeneous nature of MDD and the fact that animal models cannot probe key symptoms of MDD, including depressed mood, guilt or rumination. Within the context of reinforcement learning (RL) theories, anhedonia, depressed mood, inability to make decisions and excessive guilt may be explained by deficits in learning about rewards and punishments and a failure to update behavior accordingly. Consistent with this theory, MDD individuals show reduced reward and increased punishment learning. Of note, reduced reward responsiveness might persist after remission, whereas punishment processing normalizes when symptoms abate, suggesting that abnormal reward, but not punishment, learning might represent a trait marker of MDD. In spite of these promising leads, the neurobiological mechanisms underlying these dysfunctions remain unknown. Critically, animal studies have shown that dopamine (DA) transmission in the ventral tegmental area (VTA) and gamma-aminobutyric acid (GABA) signals originating from the lateral habenula (LHb), as well as interactions between these systems, influences RL. These interactions modulate DA release and basal ganglia (BG) activity, which helps organisms to choose or avoid an action. Despite compelling evidence from animal studies for these interactions, these processes have not been tested in MDD. The goal of the proposed research is to fill this critical gap and take us one-step closer to developing a mechanistic model of RL dysfunction in MDD. To achieve this goal, we will integrate an innovative multi-modal molecular and functional neuroimaging approach to test RL in 60 subjects (20 HC, 20 MDD and 20 rMDD). To this end, baseline GABA levels in the BG will be measured using a novel multi-voxel MRS technique, followed by an fMRI session while participants complete a social RL task. This will allow us to investigate the neural correlates of learning deficits in MDD and the influence of GABA on RL. We hypothesize that controls will exhibit reward learning signals in the BG and punishment signals in the LHb. Consistent with our proposal that MDD is an RL disorder, we hypothesize that MDD will show blunted reward and punishment signals in the BG and LHb, respectively. This abnormal pattern in MDD will be linked to enhanced punishment and reward signals in the BG and LHb, respectively. In addition, parsing state/trait effects of MDD, we hypothesize that reward and not punishment learning deficit will be observed in rMDD. Lastly, we hypothesize that baseline GABA levels will predict RL in all groups.