FROM THE PERSPECTIVE OF A SLEEP RESEARCHER ALL THE ACTION SEEMS TO OCCUR DURING SLEEP WHILE WAKING IS THIS OTHER, UNAPPRECIATED AND undiffer-

126 FROM THE PERSPECTIVE OF A SLEEP RESEARCHER ALL THE ACTION SEEMS TO OCCUR DURING SLEEP WHILE WAKING IS THIS OTHER, UNAPPRECIATED AND undifferentiated state that simply fills the time between periods of sleep. Sleep comes in two flavors: REM sleep and non-REM (NREM) sleep. One defining characteristic of the latter is the presence of delta oscillations (1-4Hz) in the EEG. The prevalence and amplitude of these oscillations, quantified mathematically as EEG delta power, are critically dependent on the time we are awake before we fall asleep, as well as the time passed since sleep onset. Twenty-five years ago this observation led Alexander Borbély to propose that delta power is an EEG correlate of a homeostatically regulated sleep-dependent process, which he coined ‘Process S’.1 The consistency with which EEG delta power tracks the sleepwake distribution under various experimental conditions has long given sleep researchers the impression that, where the sleep homeostat is concerned, wakefulness is a homogeneous state. An earlier, clear indication that it is not all that simple and that differences in waking ‘quality’ do affect EEG delta power during subsequent NREM sleep, was observed by Meerlo et al. in rats experiencing social defeat.2 After such challenge EEG delta power was higher than could be predicted on the proceeding time-spentawake. In mice we identified differences in the rate by which wakefulness increased EEG delta power among inbred strains3 and between male and females.4 With Meerlo, we argued that these differences might be related to differences in waking quality and found some support for that in the waking EEG; a slower build-up rate of Process S seemed to be accompanied with a lower activity in theta frequency range (5-10Hz).4 On the other hand, the rate with which EEG delta power decreases during NREM sleep did not seem to vary with sex or genetic background3,4 suggesting that NREM sleep, compared to wakefulness, is the more homogenous state. In this issue of SLEEP, Huber et al. report that the increase in EEG delta power observed under acute dark (DD) compared to normal light-dark (LD) conditions could, to a large extent, be attributed to the DD-induced increase in exploratory behavior.5 Under DD, exploratory behavior markedly increased at the expense of quiet waking leaving total time-spent-awake unaltered, again demonstrating that besides wake duration, wake quality also matters. The relationship between increased time-spent-exploring and increased EEG delta power during DD was observed both under baseline and sleep deprivation (SD) conditions although the magnitude of this relationship seemed more important during the latter—that is, the ~50% extra time-spent-exploring during baseline was associated with a modest ~10% increase in delta power, while the ~18% increase in exploration during SD was associated with a ~30% increase in delta power. Similarly, the SD-to-baseline increase in exploratory behavior seemed to have a larger impact on delta power when rats were kept in the dark. Thus, the efficacy of exploratory behavior in raising delta power seems to depend on light and experimental condition; i.e., also the quality of exploratory behavior seems to matter. It might be of interest to verify whether the prevalence and amplitude of EEG theta oscillations, characteristic of exploratory behavior in rodents, could be used to index its quality. Finally, in contrast to the observations in mice, also the decrease of EEG delta power during NREM sleep seemed to depend on the lighting condition; delta power during recovery sleep remained higher than baseline much longer under DD than under LD. Given that the decrease of Process S is thought to be proportional to the level of EEG delta power and that NREM sleep duration was increased in the subjective dark period, one would expect values to decrease faster during the DD recovery period. Alternatively, rats might have spent more of their waking time exploring also during this period. The results of this study were put in the context of the hypotheses that sleep is a local and use-dependent phenomenon subserving synaptic homeostasis.6,7 The results of Huber et al. discussed above, provide additional support for the ‘use-dependent’ aspect of these hypotheses, which predicts that different waking behaviors will impact the homeostatic need for sleep according to their respective levels of cortical activation. Thus, exploratory behavior, presenting the more active state, will be followed by a higher level of EEG delta power compared to quiet wakefulness. Tononi and Cirelli propose that cortical activation induces synaptic potentiation, which becomes increasingly important over the course of a waking period and that total synaptic strength reached at sleep onset determines the level of subsequent EEG delta power.6 Cortical activation is accompanied by increases in the transcription of genes encoding proteins involved in neuronal plasticity. From this, it follows that increased time-spent-exploring should be accompanied by a larger increase in these transcripts. Huber et al. tested this conjecture for four such genes: BDNF, Arc, Homer (1a-c), and NGFI-A.5 The expression of all four genes increased with SD and for all four the relative increase was higher when SD took place in the dark. However, only for BDNF was the expression modulated by time-spent-exploring and predicted subsequent EEG delta power. Other studies also reported on a positive relationship between BDNF expression and EEG delta power.8,9 The Quality of Waking and Process S