Ripple-Triggered Stimulation of the Locus Coeruleus during Post-Learning Sleep Impairs Memory Consolidation

Hippocampal ripples, brief high-frequency (150–200 Hz) oscillations occurring during quiet wakefulness or slow wave sleep (SWS), represent simultaneous discharge of a large neuronal population that is synchronized across the entire hippocampus. Learning experience increases frequency of ripple occurrence, which is predictive of memory recall, while ripple suppression impairs hippocampal-dependent learning. Experience-induced replay of neuronal ensembles occurs predominantly during ripples. These observations support the idea that ripples provide a neurophysiological substrate for off-line memory consolidation by facilitating synaptic plasticity within the learning-associated neuronal network. We hypothesized that noradrenaline (NE) release during ripples in subcortical and cortical targets of the Locus Coeruleus (LC) may be beneficial for memory consolidation. Rats implanted with linear electrode arrays for extracellular recording in cortex and hippocampus and a stimulating electrode in LC were trained on a spatial memory task. Neural activity was monitored for 1 h immediately after each learning session. Ripples were detected on-line using a band-pass filtered (150–250 Hz) extracellular voltage signal recorded in the CA1 region of hippocampus by applying a threshold-crossing algorithm. Trains of biphasic electrical pulses (0.4 ms, 0.05 mA) were delivered to LC at each ripple onset. Group 1 received LC stimulation (5 pulses at 20 Hz) that did not produce detectable changes in cortical or hippocampal neural activity. Group 2 received LC stimulation (10–20 pulses at 50–100 Hz) that induced a transient (1–2 s) desynchronization of cortical EEG, during which both thalamocortical sleep spindles and hippocampal ripples were suppressed. Additional control groups included random LC stimulation, stimulation outside of LC, and sham-operated animals. Ripple-triggered LC stimulation produced a spatial memory deficit exclusively in Group 2 rats, when LC stimulation transiently eliminated sleep spindles and ripples. The behavioral performance of none of the other groups differed from intact animals. We conclude that stimulationinduced discharge of LC neurons and concurrent NE release in the projection targets of LC caused a transient brain state change, which was not favorable for off-line hippocampal-cortical communication underlying consolidation of recent memories. The obtained results challenge the original hypothesis on the ripple-coupled NE release for promoting synaptic plasticity within reactivated neuronal assemblies. Yet, the present results further support our recent discovery of a remarkable dichotomy between ripple-associated cortical activation and deactivation of many subcortical regions including thalamus and brain stem neuromodulatory centers (Logothetis, et al., 2012, Nature 491 547–553). Thus, hippocampal ripple events may serve as indicators of a particular brain state that provide low interference for off-line consolidation of the declarative memories. Activation of any competing network during ripples may lead to less efficient consolidation