The Gears of the Sleep Clock

Is replenishment of resources the key to the sleep/wake cycle?

By Allan Pack

Artwork by Michael Morgenstern

man coming off his night shift gets into his car. He knows it's the most dangerous part of his day, a time when his body aches for sleep. He struggles to stay awake while driving home. He's tried coffee. He's tried driving with the windows open, or cranking the air conditioning up high. He's tried keeping a Vicks menthol inhaler on hand. So every day he makes a pit stop at the local gas station and buys a bag of ice, and as he turns onto the highway, applies it to different parts of his anatomy to try to keep himself awake and driving safely.

This shift worker's story made an impression on me when I first heard it over 10 years ago. At the time, there...

Since the early 1990s, researchers had known that a major circadian clock was located in the brain, specifically in the suprachiasmatic nucleus (SCN) of the hypothalamus. (To prove it, researchers restored circadian cycles in mice whose SCN was damaged by transplanting the SCN of another mouse.1) But it wasn't until the mid- to late 90s that the genes involved in the circadian clock in mammals were discovered. (Genes that made up the clock in Drosophila were described some years earlier.) The discovery of mammalian clock genes revealed an unexpected aspect of circadian rhythms. Not only were clock genes expressed by the neurons of the SCN but they were present in other neurons and many tissue types throughout the body that exhibited their own cyclic patterns of gene expression. It is now clear that there is cyclical oscillation across the day of many genes in all tissues.

Related Articles

2 We sampled both the cerebral cortex, a region whose function is affected by sleep and wake, and the hypothalamus, a region known to regulate sleep and wake.

When we grouped all of the genes that were turned on during sleep, we saw a pattern emerge. We mapped genes that increased expression during sleep into functional categories by determining which classes of genes were over-represented. Genes that turned on during sleep are involved in synthesis of key molecules as well as intracellular transport. During sleep, the brain seems to be building the complex factories of molecules that produce things like protein, steroids (including cholesterol), as well as heme. There were also a number of genes involved in synthesis of vesicles that are needed to release the neurotransmitter chemical signals. Cholesterol also plays a role in signaling as a major component of lipid rafts in cellular membranes. Lipid rafts are important in bringing neurotransmitters, receptors and other signaling molecules close together, thereby altering the efficiency of signaling. Genes for all steps in cholesterol synthesis were upregulated during sleep, as was the relevant transcriptional factor that controls expression of these genes.

The categories of genes that are turned on suggest that a major purpose of sleep is to rebuild the molecules that are essential for cell function. This is compatible with the concept of temporal metabolic cycles that are found even in yeast and in plants such as Arabidopsis. While many researchers assume that the metabolic cycle is driven by the circadian clock in mammalian species, our data would lead to a different conclusion, i.e., at least in the brain, it is coupled to behavioral states—sleep and wakefulness. It seems that when an animal is awake, energy resources are used to maintain the electricity, i.e., support neuronal firing. However, in sleep, with neuronal firing reduced, energy is now available to rebuild key molecules and be ready for the wear and tear of the next period of wakefulness. This concept is compatible with the energy hypothesis of sleep/wake control that posits that energy resources in brain are depleted during wakefulness and replenished during sleep.

While many researchers assume that the metabolic cycle is driven by the circadian clock in mammalian species, our data would lead to a different conclusion.

In contrast, in mice that were forced to stay awake, we saw an increase in heat-shock proteins and molecular chaperones. These genes are usually turned on at times of cellular stress, i.e., when protein folding in the endoplasmic reticulum (ER) does not function properly. When stressed, the cell turns on a protective mechanism, the unfolded protein response, which protects the ER from the toxic effects of misfolded proteins that tend to aggregate together. The unfolded protein response leads to reduced protein translation and an increase in chaperones that assist in escorting misfolded proteins to degradation. A biochemist in our group—Nirinjini Naidoo—has shown that all aspects of the unfolded protein response occur after 6 hours of sleep deprivation in the mouse brain.3 Recent studies have shown that cellular stress may be a factor in neurodegenerative disorders, although we haven't yet established a direct connection with this disorder and sleep.

Sleeping Genes: We wanted to understand which genes were turned on during sleep and sleep deprivation in mice that were kept on the same circadian cycle. We sacrificed mice that slept (blue) and mice that stayed awake (yellow), every 3 hours for 4 time points and compared the gene expression at time zero. Over 2,000 genes were turned on or off during sleep. Many of these belong to a broad category of proteins that synthesize large molecules, suggesting that sleep is needed to rebuild the molecules that are essential for cell function.

The proteins upregulated to protect the ER during sleep deprivation are not restricted to the brain. A group led by Mehdi Tafti in Lausanne, Switzerland has shown recently that the liver also upregulates Hspa5, a key protein in the unfolded protein response in sleep-deprived mice. While our study focused on brain tissue, the picture that is beginning to emerge is that sleep deprivation—in addition to circadian mechanisms—not only affects molecular mechanisms in the brain, but also those of organs throughout the body. While the clocks in peripheral organs receive signals from the master clock in the SCN, the peripheral clocks can shift their temporal pattern more rapidly than the SCN clock. Some peripheral clocks can be retrained independently of the slow 1-hour-per-day shift of the SCN clock. For example, some clocks shift quickly, up to 12 hours per day, in response to an altered feeding schedule, although we have yet to characterize the molecular drivers for this. The temporal patterns of gene expression in organs such as liver, fat, heart, and lung are likely determined by a coordinated interaction of the local and brain clocks as well as the consequences of behaviors such as temporal feeding pattern and sleep and wakefulness. Interestingly, changing a feeding schedule can also produce a temporal shift in sleep and wake, an action that may involve the dorsomedial hypothalamus rather than the SCN.

Our research has shown that the act of sleeping has biological functions that supersede the role and control of the circadian clock. While the circadian clock does impact the timing and the drive for sleep, there are other drivers—such as the depletion of cerebral energy and the molecular mechanisms leading to sleepiness. It is clear that there are many factors affecting sleep.

The shift worker is an excellent example of what happens when these processes are disrupted. Their lives are impacted by the effects of shifted clocks, shifted eating schedules, and sleep deprivation on top of it all. Their sleep/wake cycle and clock mechanisms are desynchronized. This may account for the medical problems that have been demonstrated in shift workers, including an increased risk for heart disease and some forms of cancer, such as breast and prostate.

Since we can't study gene expression in the human brain, my group has also been working on developing other animal models of sleep that allow us to look across phyla for the genetic functions and regulators of sleep that are most conserved. In mammals, sleep is assessed by using surface electrodes on the head to detect brain waves using the electroencephalogram. This is not feasible in the important model systems for study of genetic/molecular mechanisms—the worm (C. elegans), the fruit fly (Drosophila), and the zebra fish (Danio rerio).

To catch a fly sleeping: In order to get a better measure of fly sleep, we developed a quantitative videotaped measuring method. First we took images of flies in their tubes ever 5 seconds, the subtracted each image from the one before using computer software. When flies were sleeping, each successive 5-second frame would cancel itself out. When the fly started to move, we'd see a white area where the fly had been and a black area in the spot the fly moved into—thus quantifying fly activity and sleep.

Sleep in these organisms is characterized by behavioral criteria that include not moving, reduced or absent response to sensory stimulation, and the animal's ability to recover sleep when sleep deprived. One of my colleagues—David Raizen—developed a technique to record sleep in C. elegans.4 By creating a digital video of the worm, and overlaying every black-and-white frame over the last, he could create a quantifiable map of movement and stillness. When the animal was still, the shape of the worm would cancel itself out with every successive frame, leaving a blank image, whereas every movement resulted in a black shape in the area the worm moved into and white spots where it moved from (see graphic “To catch a fly sleeping” above). Using this approach, he has demonstrated that there are four periods of sleep during development just prior to the worm's four moults in what is called lethargus.

Another colleague—John Zimmerman—has applied similar digital video imaging to assess sleep and wake in the fruit fly. We, along with other laboratories, are elucidating which regulatory molecular mechanisms for sleep/wake control are conserved throughout the evolutionary tree. Early evidence is that there are a number of key conserved pathways across species.5 For example, epidermal growth factor (EGF) promotes sleep in C. elegans, Drosophila, rodents, and rabbits.

Our research has shown that the act of sleeping has biological functions that supersede the role and control of the circadian clock.

Thus, moving forward, our goal is to identify the molecular mechanisms regulating sleep and wake. We and others have started, but much remains to be done. Since sleep and wake are controlled by complex neuronal circuits, we need to understand what genes are changing expression within these neuronal groups and how modifying gene expression in specific types of brain cells alters sleep and wake. As described above, however, sleep is not just for the brain, and understanding the role of sleep and wake in metabolic processes throughout the body is an area that deserves further study.

Ultimately, these findings need to be translated to humans. We know that the behavioral impact of sleep deprivation is highly variable between individuals and we have recently shown that variations in the impact are highly heritable. Sleep disorders such as insomnia, obstructive sleep apnea, and narcolepsy are also heritable. Elucidating the gene variants that make some individuals resistant to the effects of sleep deprivation while others suffer tremendously will provide new therapeutic targets for intervention.

As the field of molecular sleep research shifted almost toward the molecular mechanism of the circadian clock, I and others have started to take another look at an older idea: that the need for sleep is greater than the need to stay on the appropriate circadian schedule.

How to catch a fly sleeping: John Zimmerman at the Center for Sleep and Respiratory Neurobiology explains a new technique for determining when a fly is sleeping or awake - a prerequisite for fly-based sleep research.

Have a comment? Email us at References

1. G.J. Boer et al., "Cellular requirements of suprachiasmatic nucleus transplants for restoration of circadian rhythm," Chronobiol Int, 15:551–66,1998.
2. M. Mackiewicz et al., "Macromolecule biosynthesis: a key function of sleep," Physiol Genomics, 31:441–57, 2007.
3. N. Naidoo, et al., "Aging impairs the unfolded protein response to sleep deprivation and leads to pro-apoptotic signaling," J Neurosci, 28:6539-48, 2008.
4. D.M. Raizen, et al., "Lethargus is a C. elegans sleep-like state," Nature, 451:569-72, 2008.
5. J.E. Zimmerman et al., "Conservation of sleep: insights from non-mammalian model systems," Trends Neurosci, 31:371–76, 2008.

Interested in reading more?

Magaizne Cover

Become a Member of

Receive full access to digital editions of The Scientist, as well as TS Digest, feature stories, more than 35 years of archives, and much more!
Already a member?