Several previous studies have demonstrated theta coupling of PFC neurons in working-memory tasks (Siapas et al., 2005, Jones and Wilson, 2005, Hyman et al., 2005 and Benchenane et al., 2010). In addition to PFC, we found that a significant portion of VTA neurons were also phase locked to theta, expanding the realm of theta oscillations to the mesolimbic dopamine system. The anatomical substrate and physiological mechanisms responsible for the theta entrainment GDC-0068 chemical structure of VTA cells remain to be identified. Theta phase-locked PFC neurons can, in principle, convey the theta rhythm to VTA GABAergic neurons (Carr and Sesack, 2000b). An alternative route is

a polysynaptic pathway, including the subiculum, nucleus accumbens, and ventral pallidum. This indirect path has been suggested to carry novelty-induced signals from the hippocampus to the reward neurons in the VTA (Lisman and Grace, 2005). The third possible pathway is the CA3-lateral septum-VTA projection (Luo et al., 2011). In return, VTA neurons can affect theta oscillations by their monosynaptic connections to the septal area (Gaykema and Záborszky, 1996) and the hippocampus (cf. Lisman and Grace, 2005). In support for a role of the dopaminergic system in theta oscillations, transient inactivation of the VTA decreases hippocampal theta power (Yoder and Pang, 2005), and VTA stimulation increases theta burst firing of medial septal

neurons, mediated SAHA HDAC by D1/5 receptors (Fitch et al., 2006). Accordingly, the VTA, with its spontaneously oscillating neurons at 4 Hz, along with the theta pacemaker medial septal area may form an interactive circuit, an ideal substrate for cross-frequency phase coupling between the 4 Hz and theta rhythms. The working-memory component of the task in our experiments was correlated with the sustained power of 4 Hz oscillation and the phase modulation of both gamma power and goal-predicting PFC neurons by the 4 Hz rhythm. Power increase in the 3–8 Hz band near the frontal midline area of the scalp is the dominant EEG pattern during various cognitive tasks in humans, known very as “frontal midline

theta” (fm-theta; Gevins et al., 1997; for a review, see Mitchell et al., 2008 and Sauseng et al., 2010). Two controversial issues of fm-theta have persisted: its specific behavioral correlates and the source of the fm-theta signal. Numerous studies have reported increased power of fm-theta during working-memory tasks (Gevins et al., 1997, Sarnthein et al., 1998, Klimesch et al., 2001 and Onton et al., 2005). Intracranial recordings in patients also demonstrate a correlation between theta power and working memory (Raghavachari et al., 2001 and Canolty et al., 2006). In contrast, other studies emphasize that fm-theta is best correlated with “mental concentration” (Mizuki, 1987, Gevins et al., 1997 and Onton et al.