, 2006; Thompson and Best, 1989; Crizotinib order Ylinen et al., 1995), but intermittent periods of reduced CA1 ensemble firing rates of 1 Hz and below may also occur during other states of network activity (Thompson and Best, 1989). On the other hand, during theta rhythmic ensemble activity that occurs during explorative behavior and rapid-eye-movement sleep (O’Keefe and Nadel, 1978), CA1 pyramidal neurons repetitively discharge single or bursts of action potentials at frequencies of 4–12 Hz (Csicsvari et al., 1999; Ranck, 1973). Under these conditions, our data predict a strong attenuation of

proximal inhibition already after the first few theta cycles, and a concomitantly higher contribution of weak dendritic spikes and EPSPs to action potential output. A further corollary of the theta-induced decrease of inhibition would be a degraded output precision, since in addition to strong dendritic spikes, now temporally more variable EPSPs and weak dendritic spikes contribute BMS-777607 datasheet to action potential output. This observation implies that during theta activity of CA1 ensembles more branches can efficiently contribute to output. The increase in output probability caused by a decrease in recurrent inhibition

is consistent with the elevated firing frequency of the CA1 ensemble during exploration observed in vivo (Thompson and Best, 1989). Our data suggest that the perforant path input is less strongly regulated by recurrent inhibition, when compared to Schaffer collateral input. However, repetitive CA1 pyramidal neuron activity at theta frequency resulted in a mildly augmented local inhibition observed with two-photon calcium imaging. Therefore, the proper integration of both, perforant path and Schaffer collateral excitatory inputs during theta activity, which is thought to be critically relevant for long-term spatial memory consolidation for (Remondes and

Schuman, 2002, 2004), may depend on the differential recruitment of recurrent interneurons (Pouille and Scanziani, 2004). Taken together, the input to output coupling of CA1 pyramidal neurons is controlled by an activity-dependent regulation of recurrent inhibition, which may be observed during different network activities in vivo. Furthermore we identify a mechanism, by which correlated input on dendritic branches could resist inhibitory control, which is widely independent of the network activity patterns: In both, periods of sparse and periods of higher rhythmic network activity, the generation and plasticity of dendritic spikes may serve to reliably couple the information conveyed by the activity of a presynaptic cell assembly to precise CA1 action potential output. Male Wistar rats (P21–P28, Charles River, Germany) were deeply anesthetized with an injection of ketamine (100 mg/kg, Pfizer, Germany) and xylazine (15 mg/kg; Bayer, Leverkusen, Germany) and then decapitated.