Computational modeling investigates the neural mechanisms underlying working memory updating impairments in...

Depression is a severe psychiatric disorder characterized not only by persistent emotional disturbances but also by significant cognitive impairments. Especially in working memory updating tasks, individuals with depression show difficulty effectively updating negative emotional information, leading to repeated processing and consolidation of the negative content. However, the neural mechanisms underlying this updating impairment remain insufficiently understood. Existing research suggests that this phenomenon may be associated with abnormal neuron-glia regulation in the prefrontal cortex and valence-dependent emotional processing. In this study, we constructed an emotional working memory updating model based on the theory of short-term synaptic plasticity, incorporating glial activity and emotional valence dynamics into the model framework. The modeling results indicate that dysregulated glial activity reduces the efficiency of glutamate clearance, thereby strengthening the representation of memory items. Emotional valence further amplifies this effect, and the magnitude of this enhancement is negatively correlated with the valence value, resulting in negative memories being more stable in storage and more resistant to updating than positive memories. Moreover, neural dynamical analyses revealed a "dual-influence" mechanism, providing a plausible theoretical explanation for the observed updating impairment. Thus, this study establishes a novel theoretical framework for understanding the impairment of working memory updating in depression, highlights the application potential of computational models in psychiatric research, and provides a theoretical basis for clinical diagnosis and treatment strategies.