其他摘要 | The brain's neural circuits consist of a large number of highly unstable networks. Yet despite the existence of mangy internal and external factors that continuously perturb the balance, our brains employ an array of homeostatic mechanisms that allow neurons or neural circuits to sense how active they are, and when they deviate from a target value, a force must be generated to move neuronal activity back toward this target. This mechanism that maintains phyiological variables within a dynamic range around a `set point' is called homeostasis. Homeostasis is what allows the organism to maintain its relative stability in the face of changing external stimuli, and sleep and neural activity are both important physiological variables in the regulation of homeostasis.
Sleep is essential for the energy recovery of the organism, and to ensure proper functioning, the organism has evolved a mechanism of sleep homeostasis, in which sleep pressure increases during wakefulness and decreases during sleep, maintaining sleep pressure near a set point. Similarly, neural activity increases during wakefulness and decreases during sleep, and it appears that a stable neural activity is associated with sleep homeostasis. Previous studies have confirmed that decreased cortical neural activity in rodents do correlate with slow-wave activity, an electrophysiological marker of sleep homeostasis. However, the decrease in neural activity in the hippocampal region was also found to be modulated by REM sleep, with a sharp decrease in neural activity during REM sleep, pulling down overall activity and thus showing lower neural activity during late sleep. The mechanism for the decrease in neural activity during sleep remains debated. It is worth noting that much of the current evidence comes from rodent studies, and it is not known whether differences in sleep architecture between humans and rodents affect the decrease in neural activity during sleep. It is difficult to record neuronal activity directly in humans, but epileptic patients have intracranially implanted electrodes to record local field potential signals due to therapeutic needs, and its power spectral density (PSD) offset of local field potentials has been shown to be a robust estimate of neural activity, which provides an excellent tool and chance for this study.
In order to investigate whether and by what mechanism human neural activity decrease during sleep, this study performed a series of analyses using intracranial electrodes to record local field potentials during sleep in epileptic patients. We used the offset of the local field potential power spectrum as an indirect measure of neural activity. Study 1 focused on the deep brain region, the hippocampus, which has been found to reduce neural activity during sleep in rodents, and the results of this study further confirm this finding. The results of Study 1 showed that: 1) offset was significantly lower in late sleep (last N2) than in early sleep (first N2), suggesting that neural activity decreases during sleep; 2) offset decreased significantly after REM sleep, and the decrease correlated positively with changes in offset across the night, confirming that human REM sleep is related with the decrease in neural activity during sleep; 3) hippocampal slow wave activity decreased continuously during sleep, and the decrease significantly predicted the change in offset, suggesting that hippocampal slow wave activity is also related with the decrease in neural activity; 4) there was no interaction between REM sleep and slow wave activity on the downregulation of neural activity. Study 2 further explored why there are two independent mechanisms for the decrease in neural activity during sleep, and we hypothesis that this may be due to the fact that the two mechanisms prefer different brain regions. We explored other brain regions excluding the hippocampus and found that: 1) not all brain regions exhibited a decrease in offset during sleep, suggesting that the decrease in neural activity during sleep is local; 2) the decrease in slow-wave activity in prefrontal cortex and entorhinal cortex, which are involved in learning and memory, were most pronounced, whereas the decrease in offset after REM sleep was mainly in the temporal cortex, supporting the hypothesis of the present study. Thus, the results of the present study suggest that human brain neural activity decreases during sleep and that both slow-wave activity and REM sleep are associated with this process. However, the two are not directly related and do not overlap exactly in brain regions. |
修改评论