Intriguingly, this representation is linked to input specifically from the vHPC. Numerous reports have demonstrated synchrony between mPFC units and ongoing oscillations in its inputs, particularly the hippocampus (Adhikari et al., 2010a, Jones and Wilson, 2005, Siapas et al., 2005, Sigurdsson et al., 2010 and Taxidis Autophagy Compound Library et al., 2010). Here, we show similar synchrony between mPFC units and ongoing theta-frequency oscillations in the ventral, but not dorsal HPC, consistent with the known roles of these subregions in EPM behavior (Kjelstrup et al., 2002). Moreover, we demonstrate that units that synchronize with the vHPC

have stronger task-related firing patterns. This effect of synchrony on EPM representations suggests that paradigm-related activity in the mPFC is at least facilitated by input from the vHPC. Consistent with this idea, firing in anticipation of a reward in mPFC units is abolished after vHPC lesions (Burton et al., 2009). Here we demonstrate that mPFC representations and open-arm avoidance

are inversely correlated. Animals with mPFC units with strong representations of open versus closed arms are those that fail to avoid the open arms. At the very least, these data argue that the representation present in the mPFC is not used to guide avoidance behavior in avoidant animals; there is no evidence that such a representation exists in these mice. The role of the mPFC representation in the behavior of animals that fail to avoid the open arms is less clear; the time course of unit HIF pathway firing during arm transitions allows for the possibility that such representations help guide choice behavior during exploration.

A causal relationship between the single-unit representation and exploratory behavior is also suggested by the inconsistent effects of mPFC inactivation on EPM behavior in rodents. Some studies report anxiolytic effects (Deacon et al., 2003, Lacroix et al., 2000, Shah et al., 2004, Shah and Treit, 2003, Shah and Treit, 2004 and Stern et al., 2010), while others report anxiogenic or no effects (Klein et al., 2010, Lisboa et al., 2010 and Sullivan and Gratton, 2002). Consistent with our findings, for studies that reported anxiolytic effects of silencing or lesioning the mPFC were those in which the control group showed relatively low levels of anxiety (Figure S3). mPFC inactivation, therefore, appears to reduce open arm exploration only in those animals that would be expected to have robust mPFC representations. Reconciling the current data with our previous findings presents something of a challenge. We have previously shown that increased theta-frequency synchrony between the vHPC and mPFC is associated with increased open arm avoidance (Adhikari et al., 2010b). The current data demonstrate that mPFC neurons that represent safety versus aversiveness are preferentially synchronized to the vHPC.