The protocol for assessing olfactory working memory capacity in mice

Background Working memory capacity (WMC) is the ability to maintain information over a few seconds. Although it has been extensively studied in healthy subjects and neuropsychiatric patients, few tasks have been developed to measure such changes in rodents. Many procedures have been used to measure WM in rodents, including the radial arm maze, the WM version of the Morris swimming task, and various delayed matching and nonmatching-to-sample tasks. It should be noted, however, that the memory components assessed in these procedures do not include memory capacity. Methods We developed an olfactory working memory capacity (OWMC) paradigm to assess the WMC of 3-month-old 5xFAD mice, a mouse model of Alzheimer's disease. The task is divided into five phases: context adaptation, digging training, rule learning for nonmatching to a single sample odor (NMSS), rule learning for nonmatching to multiple sample odors (NMMS), and capacity testing. Results In the NMSS rule-learning phase, there was no difference between wild-type (WT) mice and 5xFAD mice in the performance correct rate, correct option rate, and correct rejection rate. The WT mice and 5xFAD mice showed similar memory capacity in the NMMS rule-learning phase. After capacity test, we found that the WMC was significantly diminished in 5xFAD mice. As the memory load increased, 5xFAD mice also made significantly more errors than WT mice. Conclusion The OWMC task, based on a nonmatch-to-sample rule, is a sensitive and robust behavioral assay that we validated as a reliable method for measuring WMC and exploring different components of memory in mice.