Fig. 1. A freely-moving delayed non-match-to-position task allows for the examination of prefrontal neural activity during spatial working memory in mice.

a Schematic of the delayed non-match-to-position task. (top) Time course of a single trial. Diagonal parallel slashes represent variable amounts of time between task phases. (bottom) Diagram of a correct trial. Progression through the task starts with the sample phase (left panel), during which one of the two outer front ports lights up (the center port was never used as a sample location). The lit sample port location has a 50% chance of being on either the left or right side, but only a left side example trial is shown here. The mouse pokes its nose into the lit sample port to make the delay port on the opposite side (“back”) of the box available (center panel). Then the mouse pokes and holds its nose in the delay (“back”) port for five seconds, after which it receives a 1 µL reward. This leads to a choice phase (right panel), where both the initial sample port and one of the remaining other front ports lights up, and the mouse is required to poke into the lit port that it has not visited previously. Non-match choices also have a 50% chance of being either of the two non-sample locations on any given trial. A much larger 7 µL reward is dispensed after a correct non-match choice. Created in BioRender. Palumbo, M. (2023) BioRender.com/u81z167. b Mice were implanted with recording electrodes in either MOs (blue), dorsal mPFC (green), or ventral mPFC (pink). Example coronal mouse slice showing the final electrode bundle locations after up to five electrode advancements of ~60 µm (white dots). Across all sessions, we isolated 304 single-units from the MOs, 354 from the dmPFC, and 330 from the vmPFC. Brain slice image credit to the Allen Mouse Brain Atlas. c A one-way ANOVA showed that electrode location did not significantly affect task performance (F(2,70) = 2.71, p = 0.073, n.s. not significant). Each circle represents performance during one session.