Sleep has been implicated as playing an important role in memory consolidation. Considerable indirect evidence suggests a role for sleep in hippocampus-dependent memory in rodents. Hippocampal "place cells" that were active during maze running were more likely to be activated during subsequent sleep, a phenomenon known as neural replay. In addition to hippocampal neural replay, it has been asserted that the hippocampus and cortex communicate during sleep by means of hippocampal generated high frequency burst patterns, which are temporally correlated with spindles in the medial prefrontal cortex during slow-wave sleep. Despite this evidence not a single study specifically demonstrates improved memory after sleep in rodents, although several studies demonstrate deficits in memory due to sleep deprivation.

The most commonly used approach to examine behavioral effects of sleep on memory consolidation has been the sleep deprivation method. Sleep deprivation studies typically find that handling the animals during the sleep phase (i.e. depriving sleep) leads to memory deficits. Although this may be suggestive that sleep is important for memory, there may be other explanations for the impairment, such as stress-related deficits from the constant handling. In Chapter 1, we reexamine the effects of sleep deprivation (i.e. gentle handling) on rodent Pavlovian fear conditioning. We found that the deprivation method itself (i.e., gentle handling) induced deficits independent of sleep suggesting that the sleep deprivation method might be a problematic way to examine the role of sleep in memory consolidation. In Chapter 2, we detailed a naturalistic method to examine whether Pavlovian fear conditioning is enhanced after a sleep phase, as compared with an equivalent passage of an awake phase. We found that sleep selectively enhanced hippocampus-dependent memory in mice. In Chapter 3, we investigated the relationship between pharmacologically induced sleep and Pavlovian fear conditioning. We found that therapeutic doses of zolpidem (0.01-0.5 mg/kg) did not affect memory acquisition or consolidation. Additionally, we found that 8 mg/kg of zolpidem induced deep sleep. Next, we explored the effects of pharmacologically induced deep sleep and memory consolidation. We found that a drug-induced "mouse nap," an episode of deep sleep during the awake phase, selectively enhanced contextual memory.

Next, we wanted to expand our findings to human memory. In Chapter 4, we utilized a nap paradigm to examine how sleep affected memory compared to caffeine and an awake-placebo control. We found that sleep improved consolidation of verbal memory compared to both caffeine and the control. Lastly, in Chapter 5, we took advantage of the nap paradigm to study how sleep integrates new information with prior experiences. We found that REM sleep, compared to NREM sleep and quiet wake, improved problem solving by assimilating new information with past experience to create a richer network for future use. Both the rodent and human evidence provide strong support for the critical role of sleep in stabilizing and reorganizing new information.