Rapid eye movement (REM) sleep dysregulation is a symptom of many neuropsychiatric disorders, yet the mechanisms of REM sleep homeostatic regulation are not fully understood. We have shown that, after REM sleep deprivation, the pedunculopontine tegmental nucleus (PPT) plays a critical role in the generation of recovery REM sleep. In this study, we used multidisciplinary techniques to show a causal relationship between brain-derived neurotrophic factor (BDNF)-tropomyosin receptor kinase B (TrkB) signaling in the PPT and the development of REM sleep homeostatic drive. Rats were randomly assigned to conditions of unrestricted sleep or selective REM sleep deprivation (RSD) with PPT microinjections of vehicle control or a dose of a TrkB receptor inhibitor (2, 3, or 4 nmol K252a or 4 nmol ANA-12). On experimental days, rats received PPT microinjections and their sleep-wake physiological signals were recorded for 3 or 6 h, during which selective RSD was performed in the first 3 h. At the end of all 3 h recordings, rats were killed and the PPT was dissected out for BDNF quantification. Our results show that K252a and ANA-12 dose-dependently reduced the homeostatic responses to selective RSD. Specifically, TrkB receptor inhibition reduced REM sleep homeostatic drive and limited REM sleep rebound. There was also a dose-dependent suppression of PPT BDNF up-regulation, and regression analysis revealed a significant positive relationship between REM sleep homeostatic drive and the level of PPT BDNF expression. These data provide the first direct evidence that activation of BDNF-TrkB signaling in the PPT is a critical step for the development of REM sleep homeostatic drive.

Brain derived neurotrophic factor (BDNF) plays a pivotal role in structural plasticity, learning, and memory. Electroencephalogram (EEG) spectral power in the cortex and hippocampus has also been correlated with learning and memory. In this study, we investigated the effect of globally reduced BDNF levels on learning behavior and EEG power via BDNF heterozygous (KO) rats. We employed several behavioral tests that are thought to depend on cortical and hippocampal plasticity to varying degrees: novel object recognition, a test that is reliant on a variety of cognitive systems; contextual fear, which is highly hippocampal-dependent; and cued fear, which has been shown to be amygdala-dependent. We also examined the effects of BDNF reduction on cortical and hippocampal EEG spectral power via chronically implanted electrodes in the motor cortex and dorsal hippocampus. We found that BDNF KO rats were impaired in novelty recognition and fear memory retention, while hippocampal EEG power was decreased in slow waves and increased in fast waves. Interestingly, our results, for the first time, show sexual dimorphism in each of our tests. These results support the hypothesis that BDNF drives both cognitive plasticity and coordinates EEG activity patterns, potentially serving as a link between the two.

Sleep plays an important role in memory consolidation through the facilitation of neuronal plasticity; however, how sleep accomplishes this remains to be completely understood. It has previously been demonstrated that neural oscillations are an intrinsic mechanism by which the brain precisely controls neural ensembles. Inter-regional synchronization of these oscillations is also known to facilitate long-range communication and long-term potentiation (LTP). In the present study, we investigated how the characteristic rhythms found in local field potentials (LFPs) during non-REM and REM sleep play a role in emotional memory consolidation. Chronically implanted bipolar electrodes in the lateral amygdala (LA), dorsal and ventral hippocampus (DH, VH), and the infra-limbic (IL), and pre-limbic (PL) prefrontal cortex were used to record LFPs across sleep-wake activity following each day of a Pavlovian cued fear conditioning paradigm. This resulted in three principle findings: (1) theta rhythms during REM sleep are highly synchronized between regions; (2) the extent of inter-regional synchronization during REM and non-REM sleep is altered by FC and EX; (3) the mean phase difference of synchronization between the LA and VH during REM sleep predicts changes in freezing after cued fear extinction. These results both oppose a currently proposed model of sleep-dependent memory consolidation and provide a novel finding which suggests that the role of REM sleep theta rhythms in memory consolidation may rely more on the relative phase-shift between neural oscillations, rather than the extent of phase synchronization.