, 2009; Mitelman et al., 2005). To determine whether spatial working memory in mice involved MD-PFC synchrony, we recorded neural activity simultaneously in the MD and

mPFC of mice performing the T maze DNMS task. If MD-PFC synchrony is involved in working memory, it should specifically be modulated during the choice phase of the DNMS T-Maze task, when the mnemonic requirement is high. For example, recent studies have shown that theta-frequency synchrony between the dorsal hippocampus (dHPC) and mPFC is modulated by the DNMS task (Jones and Wilson, 2005; Sigurdsson et al., 2010). These studies found that phase-locking of PFC units to the theta-frequency component of the hippocampal local field potential (LFP) was enhanced during the choice phase of the DNMS task (which requires working memory) compared to the sample phase (which does not). We therefore

Tenofovir examined MD unit phase-locking to mPFC LFPs across multiple frequency ranges in trained animals performing Ku 0059436 the DNMS T-maze task. In saline treated mice, the phase-locking of MD units to beta frequency (13–30 Hz), but not theta (4–12 Hz) or gamma frequency (40–60 Hz) PFC oscillations, was strengthened in the choice phase (two-tailed paired t test, ∗∗p < 0.01) (Figure 5A) suggesting that MD-PFC synchrony in the beta range is selectively modulated by working memory. Looking at individual units, about half of units (17/40; 42.5%) noticeably increased their phase-locking to mPFC beta oscillations during choice phase while phase-locking did not change or decreased in 18/40 (47.5%) and 4/40 (10%) units, respectively (Figure S5A). Strikingly, the increase in MD-PFC beta synchrony was selectively disrupted in CNO-treated MDhM4D mice (repeated ANOVA, task phase × treatment interaction, #p < 0.05) (Figure 5A). This was observed as well in mice that never received CNO injection prior the recording ruling out that these effects could be due to chronic effects of the drug (Figure S5E).

Phase-locking to other frequency ranges was unaffected. Examining individual cells, we observed a significant reduction (Odd ratio = 0.09, p < 0.001) of the percentage of neurons increasing their phase-locking all (2/33; 6.1%) and a significant increase (Odd ratio = 5.5, p < 0.01) of the percentage of cells showing no changes (27/33; 81.8%) in CNO-treated mice compared to controls (Figure S5A). The effects of decreasing MD activity on the sample/choice difference in phase-locking was also confirmed using the pairwise phase consistency measure that controls for spike history effects such as bursting (Figure S5F; Vinck et al., 2012). To rule out the possibility that CNO-treatment affected the quality of unit isolation, we confirmed that this disruption in MD-PFC beta synchronization was not due to differences in unit quality (Figures S5B and S5C).