The Molecular Clock: How a Gene Feedback Loop Keeps a 24-Hour Beat

A clock made of genes turning themselves off

Every cell in the body carries its own roughly 24-hour clock, built not from any dedicated timekeeping organ but from a feedback loop of gene expression — a transcription-translation feedback loop (TTFL). The core mammalian loop centres on two pairs of genes: BMAL1 and CLOCK form a complex that activates transcription of the Period (PER) and Cryptochrome (CRY) genes; PER and CRY proteins accumulate, dimerise, and travel back into the nucleus, where they inhibit their own activators, BMAL1/CLOCK — shutting off their own further production. As PER and CRY are gradually degraded, the inhibition lifts, BMAL1/CLOCK activity recovers, and the whole cycle restarts. The delay built into transcription, translation, protein accumulation and nuclear re-entry is what stretches one full loop out to roughly a day rather than completing in minutes.

The Goodwin oscillator as a minimal model

Brian Goodwin's 1965 model strips this down to the smallest set of equations that still oscillates: an mRNA (X), a protein it's translated into (Y), and a final regulatory product (Z) that loops back to inhibit X's own transcription, with the inhibition modelled by a Hill function — a saturating, sigmoidal term that captures cooperative, switch-like repression rather than a simple linear one:

dX/dt = k₁ / (1 + (Z/K)ⁿ) − k₂·X      // transcription, repressed by Z
dY/dt = k₃·X − k₄·Y                    // translation
dZ/dt = k₅·Y − k₆·Z                    // feedback repressor accumulates, decays

n is the Hill coefficient, describing how sharply, non-linearly the repressor switches transcription off as its concentration rises. A key mathematical result about the classical three-variable Goodwin model is that it only sustains stable oscillations — a genuine limit cycle rather than everything settling to a fixed steady state — when the Hill coefficient exceeds a threshold, n greater than 8 for the plain three-variable form, a level of cooperativity that is higher than most individual biochemical binding steps plausibly provide on their own. Real circadian networks reach oscillation with much softer nonlinearity than that because the biology adds extra delay stages and multiple interlocking feedback loops beyond the minimal Goodwin skeleton, effectively achieving the same qualitative "delay + negative feedback" recipe through more, gentler steps rather than one very sharp one.

Delay is what makes it oscillate at all

A negative feedback loop with no delay just settles smoothly to a fixed equilibrium — pure negative feedback is a stabiliser, not an oscillator. What turns the loop into a sustained oscillation is the combination of negative feedback and a built-in time lag between cause and effect: by the time the repressor Z has built up enough to shut X off, X's downstream effects (Y, then Z) are still rising from X's earlier, higher output, so the system overshoots, then undershoots as it corrects, and the correction itself lags too. Chain enough delayed steps (transcription → translation → nuclear transport → dimerisation → repression) together and the system doesn't damp back to equilibrium — it rings, indefinitely, at a period set by the sum of those delays.

Why the period lands near 24 hours specifically

The Goodwin equations, on their own, oscillate at whatever period the chosen rate constants produce — there is nothing in the bare mathematics that pins the period to exactly 24 hours. The real circadian loop's period is tuned by the actual biochemical rate constants (transcription and translation speeds, protein stability, degradation rates) that evolution has set for PER, CRY, BMAL1 and their many auxiliary regulators, and it is then fine-tuned every day by light input through the SCN — the same entrainment machinery that also governs how jet lag resolves. A cell's clock left running with no light input at all (in constant darkness, in a lab) keeps ticking at its own close-to-24-hour free-running period rather than stopping, which is itself strong evidence that the oscillation is a genuine self-sustaining property of the feedback loop and not merely a response driven by an external day-night signal.

From one cell to a synchronised body

Every cell runs this loop more or less independently, so without coordination, individual cells would drift out of phase with each other. The SCN's roughly 20,000 neurons couple to each other electrically and chemically, which synchronises their individual noisy molecular clocks into one coherent tissue-level rhythm, and the SCN then broadcasts that consensus phase to peripheral clocks throughout the body via hormonal and neural signals — which is exactly the multi-cell, multi-tissue system that a transmeridian flight throws out of alignment, and that the light-and-melatonin protocols used against jet lag are trying to re-synchronise from the top down.

Frequently asked questions

Try it live

Everything above runs in your browser — open Molecular Clock and change the parameters while it is running. Nothing is installed, nothing is uploaded, the whole model lives in one tab.