№ 160 · biology

The body clock is inside the cell

Your body keeps daily time, and the timing comes from a loop of genes and proteins. Yi, Díaz, D'Souza and Buhr describe how that loop works, and Patton and Hastings review how it is organised across the body, from a small region of the hypothalamus, the suprachiasmatic nucleus (SCN), out to the other tissues.

What the clock is made of

Yi and colleagues call the core timekeeping genes clock genes. The main loop involves about ten of them, including Clock, Bmal1, Period (Per1, Per2, Per3) and Cryptochrome (Cry1, Cry2). Their proteins form a transcriptional/translational feedback loop, or TTFL: transcription makes RNA, translation makes protein, and the protein returns to switch transcription off.

The loop, step by step

In the morning, Clock and Bmal1 are transcribed. Their proteins join into a CLOCK:BMAL1 complex, which binds DNA at E-box sites and switches on its targets, including the Per and Cry genes. PER and CRY proteins build up outside the nucleus, peaking in the late afternoon or evening. At night the PER:CRY complex enters the nucleus and blocks CLOCK:BMAL1, so transcription falls. Patton and Hastings estimate that making PER and CRY and moving them into the nucleus takes roughly 12 hours, and breaking them down takes roughly 12 more, which gives a period near 24 hours.

Predict first

If you removed the suprachiasmatic nucleus from an animal, what would happen to the clocks in its other tissues?

Check your answer

They keep ticking, but not in step. Patton and Hastings write that without the SCN the network quickly falls out of coherence: the local clocks drift away from one another and their rhythms weaken.

Clocks beyond the SCN

Yi and colleagues report that nearly all mammalian tissues contain working clocks that run on this loop inside single cells, and that in the majority of cell types examined the molecular clock keeps cycling even in isolation. Patton and Hastings describe heart, lung, liver, skin and parts of the brain producing rhythms by themselves when cultured outside the body. The SCN is small, about 10,000 neurons alongside about 3,000 glial cells, yet it is not the main driver of these clocks. Patton and Hastings cast it as their synchronizer and coordinator: it is the one part of the network wired to timing cues from the retina, and it matches the network's phase to solar time through neural, endocrine and behavioural routes.

What sets the pace

The loop's speed depends largely on how stable PER and CRY are, because their levels must fall before a new cycle can start. Yi and colleagues describe the tau mutant hamster, the first mammal found to carry a circadian mutation, whose free-running rhythm lasted about 20 hours rather than the 24 hours of normal hamsters. The mutation lies in casein kinase I epsilon, an enzyme that phosphorylates PER proteins and marks them for degradation. Patton and Hastings note that making PER less stable can shorten the mouse SCN clock's period by up to 4 hours, while making CRY more stable can lengthen it by 3 hours.

What is still argued

The repression step is not settled. Yi and colleagues write that CRY binds CLOCK:BMAL1 directly, while the role of PER is less clear; PER does not bind the complex on DNA but seems to help pull CLOCK:BMAL1 off it. They also note that a nuclear CRY:PER assembly has been found as large as 1.9 MDa. Patton and Hastings add two open questions: how the night-time rise in glutamate outside SCN cells, which likely comes from astrocytes, inhibits SCN neurons, and how those neurons pass circadian time to the astrocytes.

In short

The unit of circadian time is the cell. A delayed feedback loop of clock genes makes a roughly 24-hour rhythm, and the SCN coordinates those cell clocks rather than generating them. As of 2023, Patton and Hastings still list how SCN neurons and astrocytes signal to each other as an open question.

Interactive Drag the slider to change how long PER and CRY take to build up and to be cleared (both together), and watch the loop's period change.

A schematic of one feedback loop: the relative level of PER and CRY rises over the build-up delay, then falls over an equal clearance delay, and the period is the sum of the two. The default, about 12 hours for each and about 24 hours in total, is the figure Patton and Hastings give. Every other slider setting, and the straight-line rise and fall, is illustrative: the slider moves both delays together to show that in this sketch longer delays give a longer period; it is not a measured relationship.

Where this comes from

  1. The molecular clockwork of mammalian cells (Semin Cell Dev Biol 126:87-96) linked only, not reproduced
    Yi, Díaz, D'Souza, Buhr · 2021
    pmc.ncbi.nlm.nih.gov/articles/PMC8481349/
  2. The Mammalian Circadian Time-Keeping System (J Huntingtons Dis 12(2):91-104) linked only, not reproduced
    Patton, Hastings · 2023
    pmc.ncbi.nlm.nih.gov/articles/PMC7614869/