人们往往对带有强烈情绪色彩的事件记忆深刻、难以忘怀，尤其是那些痛苦、恐怖的回忆。负性情绪对记忆的不良影响可能导致一些情绪障碍和疾病，如恐惧症、抑郁症以及创伤后应激障碍等。因此，探究情绪效应影响记忆加工的神经机制将有助于我们更全面地了解情绪障碍的形成和发展。已有研究证明，情绪标记使得相关事件更好地被记忆。但目前大多数的研究都集中在情绪标记对个体记忆的影响上，无论是情绪直接标记的事件，还是情绪所泛化到的相关事件。实际上，由于相关事件在大脑中的重叠表征形式，新旧相关事件的记忆会通过重激活的方式整合为一个整体的记忆网络。然而，情绪标记在记忆编码阶段是如何追溯到已有相关事件并影响其记忆整合，以及记忆巩固过程是如何调控情绪标记对记忆整合的影响，对此我们还知之甚少。 本文通过三个实验研究对上述问题进行了阐释。研究一采用行为实验，探讨了情绪标记对相关事件记忆整合的影响。被试首先学习72个面孔-物品配对（即预学习阶段），然后只呈现预学习阶段中出现的面孔（重叠线索），并伴有恐怖或中性的声音（即情绪标记阶段）。最后，被试分别对面孔、面孔-物品配对、面孔-声音配对进行再认记忆测试。结果发现，情绪标记对面孔-物品的关联记忆有增强作用，但是对面孔的个体记忆没有显著影响。研究二沿用了研究一的实验设计，应用事件相关的功能磁共振脑成像和同步皮肤电记录技术，探究了情绪标记对相关事件记忆整合产生增强作用的认知神经机制。结果发现，情绪标记增强了海马重激活和海马-杏仁核的功能联结。此外，海马重激活与海马-杏仁核功能联结强度都与随后关联记忆的表现成正相关。研究三在研究一的基础上增加了两组被试，分别在学习后3小时和24小时后进行记忆测试，以探究情绪标记对记忆整合的增强作用是否是稳定且持久的。结果发现，3小时和24小时后测的记忆表现呈现出与立即后测相反的效果，情绪标记下面孔-物品的关联记忆反而低于没有情绪标记的关联记忆。 上述结果说明，情绪标记可以回溯性地增强过去相关中性事件的记忆整合。这可能是由于情绪标记下增强的海马-杏仁核功能联结引发了更高的海马重激活，从而促进了相关记忆的整合，并表现为相关事件关联记忆的增强。此外，情绪标记对记忆整合强化作用的消失揭示了记忆巩固的调节作用，这可能与巩固过程中对负性记忆的主动遗忘与抑制机制有关。这些研究成果为探索相关情绪障碍的形成机制及其有效预防、干预和矫治的方法手段提供了重要的科学依据。

Our memories for events with high arousal emotion are often vivid and long lasting relative to neutral memories, especially those aversive or traumatic events. The maladaptive effect of negative emotion on memory can result in some mood disorders, such as phobia, depression and PTSD (post-traumatic stress disorder). And therefore researches about neural mechanisms of emotional effect on memory processing are very necessary and helpful to understand the formation and development of these mood disorders thoroughly. It is generally known that neutral events can be better remembered if tagged with emotion during memory processing. While most of the existing researches are focusing on the emotional effect on item memory for events directly tagged with emotion or related events where emotion generalizes to. Actually, due to the overlapping representations for related events in our brains, memories of related initial and subsequent experiences would integrate into a whole mnemonic network through reactivation. However, the neurobiological underpinnings of how emotional tagging traces back to and affects memory integration for past related events, and how consolidation changes the effect of emotional tagging on memory integration for related events remain poorly understood. This research includes three studies. In the study 1, we investigated how emotional tagging affected the integration of related events into episodic memory with behavioral experiment. Participants first learnt a set of 72 face-object pairs (i.e., pre-learning phase), and then were reminded of pre-learned memories using overlapping faces as cues in concurrent concert with an aversive (vs. neutral) voice (i.e., emotion-tagging phase). Thereafter, participants performed subsequent memory tests separately for faces, face-object associations and face-voice associations. We found that emotional tagging enhanced associative memory for face-object pairs, but not item memory for faces. In the study 2, using event-related fMRI with concurrent skin conductance recording, we further investigated neurobiological processes underlying the enhancement effect of emotional tagging on memory integration for past related events. We found that hippocampal reactivation was higher in emotion-tagging (vs. neutral) condition with greater hippocampus-amygdala functional coupling. Moreover, both hippocampal reactivation and hippocampus-amygdala functional connectivity were positively related with subsequent associative memory performance. In the study 3, we investigated whether this enhancement effect of emotional tagging on memory integration was lasting and stable, and how consolidation modulated this effect with a set of behavioral experiments. Based on study 1, two more groups of participants were added in study 3, performing memory tests 3h or 24h after encoding. We found that the enhancement effect of emotional tagging on memory integration in immediate memory test disappeared over time. In 3-hour and 24-hour delayed memory tests, face-object associative memory with emotional tagging (emotion condition) was weaker than it without emotional tagging (neutral condition). Our results demonstrate that emotional tagging can retrospectively enhance memory integration for related neural events, likely through reactivation of hippocampal pattern coupling with amygdala emotional systems. Moreover, such enhancement effect disappears in delayed retrieval, which indicates a vital role of consolidation in modulating the effect of emotional tagging on related memory integration. These findings have positive significance in learning more about the formation mechanism of mood disorders, and provide important evidence for their effective prevention, intervention and treatment.