Abstract
Rationale Working memory deficits are a core feature of psychiatric disorders such as schizophrenia and bipolar disorder, yet effective pharmacological strategies remain elusive. The trial unique non-matching to location (TUNL) task offers improved face validity compared to traditional maze-based tasks; however, no systemically administered compound has been shown to enhance performance in this task in naïve mice. Furthermore, the neural circuits underlying working memory impairments remain incompletely understood. Functional ultrasound (fUS) imaging was used alongside behavioral testing to provide circuit-level context for pharmacological effects on working memory. Objectives This study had two primary aims: (i) to assess whether an optimized TUNL task can detect both impairments and improvements in working memory in mice, and (ii) to examine behavioral and circuit-level effects of the GlyT1 inhibitor bitopertin. Methods We used an optimized TUNL task protocol to assess the effects of pharmacological modulation of glutamatergic and cholinergic signaling on task performance. fUS imaging was used to investigate associated neural circuit-level changes. Results Scopolamine impaired task performance to chance levels, whereas the GlyT1 inhibitor bitopertin improved task performance and rescued deficits induced by the NMDA receptor antagonist MK-801. Performance was evaluated across different cognitive loads by varying the spatial distance between sample and non-match cues. fUS imaging revealed that bitopertin prevented ketamine-induced increases in cerebral blood volume within key working memory regions, including the prefrontal cortex, the dorsal hippocampal regions and the dorsal thalamus. Conclusion In summary, the TUNL task is sensitive to both impairments and enhancements of working memory in mice, even in the absence of pharmacological impairment. These findings identify GlyT1 inhibition as a pharmacological strategy capable of improving working memory performance and normalizing NMDA receptor antagonist-induced network activity.