Neuroscience of Reward-Related Learning and Memory in Depression

描述（由申请人提供）：重度抑郁症（MDD）是一个昂贵的公共卫生问题，并且MDD的诊断异质性使治疗复杂化。一种方法是研究内表型，这是重度抑郁症的关键方面，可能涉及离散神经回路的功能障碍。快感缺乏（快乐丧失）是一种很有前景的内表型，但这一MDD核心症状的神经认知机制尚不清楚。目前的应用将检验刺激-奖励和行动-奖励学习失败导致快感缺乏的假设。在为期两年的K99阶段，申请人将追求四个目标。首先，为了发展关于MDD如何影响强化学习的定量假设，他将从Michael Frank博士（K99的共同导师）那里学习计算建模。在Frank博士和Diego Pizzagalli博士（K99导师）的指导下，K99研究基金将支持从执行奖励的巴甫洛夫条件反射任务的对照中收集fMRI数据。这将为R00阶段的MDD受试者的研究奠定基础，同时也提供了将用于测试强化学习的时间差分算法的有价值的数据。为了学习R00阶段的额外技能，申请人还将完成Frank博士提供的一学期的“计算认知神经科学”课程。其次，Nicholas Lange博士将训练申请人进行弥散张量成像分析，以探测涉及记忆和奖励处理的大脑区域的结构完整性和连通性，这些区域可能在MDD中退化。第三，申请人将在诊断性访谈方面进行重点培训，这将在他过渡到独立并开始领导一个专注于患者导向研究的实验室时非常宝贵。第四，根据Pizzagalli博士和Frank博士的意见，申请人将进行一次有效的求职谈话，并进行教师求职，以便建立一个专注于重度抑郁症中与奖励相关的学习和记忆的实验室。在独立阶段，将对对照组和重度抑郁症受试者进行3项与奖励相关的学习和记忆的功能磁共振成像（fMRI）研究。第一项研究将集中在腹侧纹状体，并将使用巴甫洛夫条件反射来检查重度抑郁症对线索奖励偶然性学习的影响。第二项研究将关注背纹状体，并将使用工具条件反射来检查MDD的行动-奖励学习。第三项研究将涉及刺激-奖励关联的显式编码，随后是两个时间点的延迟回忆，以研究重度抑郁症如何影响海马-纹状体在编码和巩固奖励信息过程中的相互作用。最后，这些研究的MRI数据将被汇总，以确定快感缺乏症是否反映了纹状体、海马体和前额皮质区域之间的弱功能或结构连接，这些区域先前与奖励处理的不同方面有关。将严谨的范例、计算模型和尖端的连通性分析结合起来，有可能显著地促进对MDD病理生理学的理解。

DESCRIPTION (provided by applicant): Major Depressive Disorder (MDD) is a costly public health problem, and the diagnostic heterogeneity of MDD complicates treatment. One approach is to study endophenotypes, key facets of MDD that may involve dysfunction in discrete neural circuits. Anhedonia (loss of pleasure) is a promising endophenotype, but the neurocognitive mechanisms underlying this core symptom of MDD are unclear. The current application would test the hypothesis that failures of stimulus-reward and action-reward learning contribute to anhedonia. During the two-year K99 phase, the applicant will pursue four aims. First, in order to develop quantitative hypotheses about how MDD affects reinforcement learning, he wil learn computational modeling from Dr. Michael Frank (K99 co-mentor). With guidance from Dr. Frank and Dr. Diego Pizzagalli (K99 mentor), the K99 research funds will support collection of fMRI data from controls performing a rewarded Pavlovian conditioning task. This will lay the foundation for a study with MDD subjects in the R00 phase, while also providing valuable data that will be used to test temporal difference algorithms of reinforcement learning. To learn additional skills for the R00 phase, the applicant will also complete a semester-long "Computational Cognitive Neuroscience" course offered by Dr. Frank. Second, Dr. Nicholas Lange wil train the applicant to conduct diffusion tensor imaging analyses in order to probe the structural integrity and connectivity of brain regions implicated in memory and reward processing, and that may be degraded in MDD. Third, the applicant will pursue focused training in diagnostic interviewing, which will be invaluable when he transitions to independence and begins directing a laboratory focused on patient-oriented research. Fourth, with input from Dr. Pizzagalli and Dr. Frank, the applicant will develop an effective job talk and conduct a faculty jo search in order to establish a laboratory focused on reward-related learning and memory in MDD. During the independent phase, three functional magnetic resonance imaging (fMRI) studies of reward- related learning and memory in controls and MDD subjects will be conducted. The first study will focus on the ventral striatum and will use Pavlovian conditioning to examine effects of MDD on cue-reward contingency learning. The second study will focus on the dorsal striatum and will use instrumental conditioning to examine action-reward learning in MDD. The third study will involve explicit encoding of stimulus-reward associations, followed by delayed recall at two time-points, to investigate how MDD affects hippocampal-striatal interactions during encoding and consolidation of rewarding information. Finally, the MRI data from these studies will be pooled to determine whether anhedonia reflects weak functional or structural connections among the striatum, hippocampus, and regions of prefrontal cortex previously implicated in different facets of reward processing. The proposed combination of rigorous paradigms, computational models, and cutting-edge connectivity analyses has the potential to significantly advance understanding of the pathophysiology of MDD.