The Circadian Clock
A loop of genes and proteins that builds up, switches itself off, breaks down and starts again, roughly every twenty-four hours. Light does not drive it. Light resets its phase.
Almost everything that looks like a daily rhythm in biology could be the sun showing through. This one is not. Put a mouse, a fly, a plant or a person in constant conditions with no time cue at all and the rhythm keeps running, on a period that is close to a day but not exactly a day, which is the giveaway that the period is being produced inside rather than copied from outside. That makes it a clean example of a structure that produces its own order. It is also a clean example of one that cannot make its own parts, which is where the reading stops.
Every figure is sourced in the list at the bottom, with the date it was read. This is a reading of published work. There is no simulation on this page, and nothing here is medical advice.
The loop
The mammalian version is a transcription and translation feedback loop. Two activator proteins switch on the genes for two repressor proteins, the repressors take hours to accumulate and move back into the nucleus, and once there they shut off the activators that made them. Then they degrade, the block lifts, and the cycle starts again.
- The activators (CLOCK and BMAL1 in mammals), which switch on the clock genes.
- The clock genes (the Period and Cryptochrome families), transcribed under that activation.
- The repressor proteins they encode, which accumulate in the cytoplasm.
- A delay, built from transcription, translation, phosphorylation and transport back into the nucleus. Negative feedback carrying a delay of this kind is among the conditions under which a loop sustains oscillation rather than settling to a steady level. Delay by itself does not guarantee it.
- Degradation, which clears the repressors and releases the activators for the next round.
activators switch on the genes → repressor proteins accumulate → repressors shut the activators off → repressors degrade → activators switch on the genes
Supplied from outside the loopthe cell that transcribes and translates, the amino acids and nucleotides the proteins are built from, and the energy that runs all of it. The loop sets the timing. It does not build its own components.
The work that identified the genes and this mechanism in fruit flies received the 2017 Nobel Prize in Physiology or Medicine.
Why this counts as self-produced order
If the rhythm were the day showing through, it would stop when the day was taken away, and it would run at exactly twenty-four hours while it lasted. It does neither.
The evidence below is a ladder of removals. Take away the day, then the organism, then the tissue, then the genes. The oscillation is still there at the bottom.
It keeps running in constant conditionswith no light cycle, no temperature cycle and no clock, the rhythm continues for weeks.
And it runs at its own period, not the day'sunder a protocol designed to remove the influence of the light schedule, Czeisler and colleagues report that "the intrinsic period of the human circadian pacemaker averages 24.18 hours in both age groups, with a tight distribution consistent with other species." Not twenty-four. That eleven-minute excess is the fingerprint of an internally generated period.
Down to single cellsWelsh and colleagues dissociated neurons from the rat suprachiasmatic nucleus and recorded them alone. Of the cells followed for at least three days, 53 percent, 17 of 32, kept statistically significant circadian firing rhythms, with periods of 24.36 hours plus or minus 1.20, ranging from 21.25 to 26.25 hours. Each cell ran on its own period, out of phase with its neighbours, which is what you would expect from many separate oscillators rather than one broadcast signal.
And it survives the loss of the genes entirelyin cyanobacteria the timing persists with no transcription or translation at all. Nakajima and colleagues put three purified proteins, KaiA, KaiB and KaiC, in a tube with ATP, and the phosphorylation of KaiC oscillated on a stable circadian period. Their abstract records that the cycle "persists even without transcription or translation" and that "the period of the in vitro oscillation was stable despite temperature change." A clock made of three proteins and an energy supply, keeping time in a test tube.
The basin
In nonlinear dynamics this object already has a name. The circadian oscillator is conventionally modeled as a stable limit cycle: trajectories from a range of admissible starting states converge on the same recurring cycle, differing in phase rather than in whether they cycle at all. The framework did not discover the cycle. What it adds is a name for the region those starting states occupy and a reason to care about where that region ends.
What settlesthe same near-24-hour cycle in the concentrations of the clock proteins, reached from a wide range of starting concentrations. The alternative outcome within the same system is flatlining, meaning the concentrations settle to a steady level and stay there.
Where the region ends is not settled herebreaking the loop is not evidence about that boundary, and the distinction is worth being exact about. A basin boundary separates starting states that return to the cycle from starting states that do not, within one set of governing equations. Knocking out a clock gene changes the equations. It moves the system to a different rule rather than moving the state to a different part of the same landscape, so it says nothing about where the region of admissible starting states ends. No experiment on this page locates that boundary.
Return after a pushshift the light schedule and the clock does not jump to the new phase. It re-entrains over successive cycles, and the period it returns to is its own rather than the new schedule's.
Phase moves, period does notthis is the sharpest thing the experiments show, and the two words are not interchangeable. Phase is a position on the cycle. Period is a property of the rule that generates the cycle. A well-timed pulse of light moves the first by hours and leaves the second where it was. The intrinsic period is set by the loop's own chemistry, and it holds nearly constant even through a temperature change, a property called temperature compensation. It is not fixed for all time. Change the chemistry, by mutation or by altering the enzymes that tag the repressors, and the period changes with it. That is the distinction rather than a problem for it: a perturbation that changes the period has changed the rule, and a pulse of light that moves phase has changed the state. A structure whose state can be shoved a long way while its generating rule stays put is the signature this page is looking for.
Light is a resetter, not a driverthis is the distinction the page exists to make. A driven rhythm stops when the drive stops. This one continues in constant darkness and merely gets re-timed by light.
The critical region
Properly timed light does something that no amount of badly timed light does. Winfree showed in Drosophila that a stimulus of critical strength at a critical phase could stop the rhythm. Jewett, Kronauer and Czeisler tested the same idea in humans: critically timed bright light markedly attenuated the endogenous circadian amplitude, and in some cases produced an apparent loss of rhythmicity, which is what is expected near the singularity of an oscillator.
A phase singularity is not the same object as the boundary of a basin, and this page does not treat them as one. At a singularity the amplitude has been driven toward zero and phase is no longer defined. A basin boundary separates starting states that return to the cycle from states that do not. The two can coincide and they need not, and nothing in these experiments settles which is the case here.
What the experiments do establish is narrower and still worth having. A small, well-aimed disturbance at one moment does something that a large disturbance at another moment cannot, which means the system's response depends on where it is in its cycle and not only on the size of the push. That is the behavior of a state on an attractor rather than a reading on a dial. Whether a driven amplitude collapse should count as a boundary event is an open question for this framework, not a finding from the laboratory. The rhythm recovers afterward, so on the evidence it is a knock rather than a loss.
What would show this reading wrong
A rhythm that stops immediately when the light cycle is removed is a driven rhythm, and the reading on this page would not apply to it.
An extract that made its own clock proteins, regenerating from raw supply the components its continuation requires, would break the not-sovereign half of the classification. Running on an outside energy supply would not, and it is worth being exact about why. Everything alive runs on delivered free energy, and this framework's own note on sovereignty says the energy cut "may never be used on its own to deny Recursion Lock." The question it asks is organizational rather than energetic: can the loop remake what it is made of?
Honest limits
It is a reading of published work. Nothing here was measured or simulated on this site, and the panel for this card does not exist yet.
The loop is drawn simply. The real mechanism includes several interlocking loops, kinases that set the pace by tagging the repressors, and tissue-level coupling between cells that the single-cell picture leaves out.
The figures quoted come from particular protocols and particular species. The human period is a laboratory measurement under forced desynchrony, and the single-cell numbers are from dissociated rat neurons.
Nothing here is medical advice, and no claim is made about anybody's sleep.
Sources
Czeisler, C. A., Duffy, J. F., Shanahan, T. L., Brown, E. N., Mitchell, J. F., Rimmer, D. W., Ronda, J. M., Silva, E. J., Allan, J. S., Emens, J. S., Dijk, D. J., & Kronauer, R. E. (1999). Stability, Precision, and Near-24-Hour Period of the Human Circadian Pacemaker. Science, 284(5423), 2177–2181. Source of the 24.18-hour figure. doi.org/10.1126/science.284.5423.2177
Welsh, D. K., Logothetis, D. E., Meister, M., & Reppert, S. M. (1995). Individual neurons dissociated from rat suprachiasmatic nucleus express independently phased circadian firing rhythms. Neuron, 14(4), 697–706. Source of the single-cell periods. doi.org/10.1016/0896-6273(95)90214-7
Nakajima, M., Imai, K., Ito, H., Nishiwaki, T., Murayama, Y., Iwasaki, H., Oyama, T., & Kondo, T. (2005). Reconstitution of circadian oscillation of cyanobacterial KaiC phosphorylation in vitro. Science, 308(5720), 414–415. Source of the three-protein clock and the temperature quotation. doi.org/10.1126/science.1108451
Jewett, M. E., Kronauer, R. E., & Czeisler, C. A. (1991). Light-induced suppression of endogenous circadian amplitude in humans. Nature, 350(6313), 59–62. Source of the amplitude suppression and the singularity reference to Winfree. doi.org/10.1038/350059a0
The 2017 Nobel Prize in Physiology or Medicine was awarded to Jeffrey C. Hall, Michael Rosbash and Michael W. Young for discoveries of molecular mechanisms controlling the circadian rhythm. nobelprize.org/prizes/medicine/2017/summary
Related on this site: Finding a Basin for the tests used here, and What Is an Attractlet for the classification this card sits against.