14.3 Oscillatory Behaviour: When Stability Becomes Dynamic

Negative feedback is often associated with stability. Whenever a biological variable deviates from its desired level, negative feedback acts to restore the original state. It therefore seems natural to expect that negative feedback always produces steady behaviour.

Surprisingly, this is not always the case.

Many biological systems exhibit regular oscillations instead of converging to a constant value. Rather than remaining fixed, their activity rises and falls in a rhythmic and highly reproducible manner. Oscillatory behaviour is one of the most widespread dynamic phenomena in biology and occurs across many levels of biological organization.

Oscillations are everywhere in biology

Perhaps the best-known biological oscillator is the circadian clock, which synchronizes physiology and behaviour with the 24-hour day-night cycle. However, rhythmic behaviour is by no means restricted to circadian biology. Oscillations also occur during cell-cycle progression, embryonic development, calcium signalling, hormone secretion, neuronal activity, and gene regulation.

Although these systems perform very different biological functions, they often share remarkably similar regulatory architectures. Many of them are built around feedback loops that repeatedly activate and inhibit their own activity.

This observation raises an important biological question:

How can a stabilizing feedback loop generate continuous oscillations instead of simply restoring equilibrium?

To answer this question, we consider one of the simplest possible regulatory systems.

A simple model of gene regulation

Imagine a gene that is continuously transcribed into messenger RNA. The transcript is translated into a protein, and this protein inhibits further transcription of the same gene.

This simple regulatory circuit consists of only two components—a transcript and its corresponding protein—but already contains a negative feedback loop (Figure XX).

Initially, transcript levels increase because transcription is active. As more transcript becomes available, protein synthesis also increases. The accumulating protein then inhibits transcription, causing transcript abundance to decline. Because less transcript is now available for translation, protein production eventually decreases as well. Once protein levels have fallen sufficiently, inhibition is relieved, allowing transcription to resume.

The system therefore repeats the same sequence of events continuously.

Instead of approaching a constant steady state, transcript and protein concentrations oscillate over time.

Oscillations require delayed responses

The crucial ingredient underlying these oscillations is time delay.

If the inhibitory protein acted instantaneously, every increase in transcription would immediately be counteracted. The system would rapidly settle into a stable equilibrium.

In reality, however, biological regulation is never instantaneous. After transcription, messenger RNA must first be translated into protein. Newly synthesized proteins may require folding, transport, or post-translational modification before becoming active. These intermediate processes introduce a delay between the production of the transcript and the inhibitory effect of the protein.

During this delay, transcription continues, allowing transcript abundance to rise beyond its eventual equilibrium level. When inhibition finally becomes effective, it overshoots in the opposite direction, reducing transcript abundance too strongly. The same process then repeats in reverse, generating sustained oscillations.

Delayed negative feedback is therefore one of the simplest mechanisms capable of producing rhythmic biological behaviour.

Perturbing an oscillating system

Oscillatory systems often respond in unintuitive ways to perturbations.

Suppose we experimentally remove part of the protein at a particular moment during the oscillation. One might expect the system simply to replace the missing protein and return to its original behaviour.

Whether this actually happens depends critically on when the perturbation occurs.

Removing protein during one phase of the oscillation may have little long-term effect, whereas the same perturbation applied at another phase can produce a completely different response. The outcome therefore depends not only on the magnitude of the perturbation but also on the current state of the system.

This observation reveals an important limitation of conventional time-series plots. Although they show how individual variables change over time, they often make it difficult to understand why identical perturbations produce different outcomes.

To answer these questions, we require a new way of representing dynamic systems.

Beyond time-series plots

So far, we have described biological dynamics by plotting individual variables against time. While this representation is intuitive, it hides an important aspect of system behaviour: the relationships between the variables themselves.

A more informative representation treats the complete biological system as a point moving through an abstract space whose coordinates are the system’s state variables. Instead of following transcript and protein concentrations separately, we follow the evolution of the entire system.

This representation, known as state space, will become one of the central concepts of dynamical systems theory.

Key concepts

Negative feedback does not always produce a steady state.
Many biological systems exhibit sustained oscillatory behaviour.
Oscillations commonly arise from delayed negative feedback.
The response of an oscillatory system depends on its current state.
Understanding oscillations requires a representation that goes beyond conventional time-series plots.

Summary

Oscillatory behaviour is a hallmark of many biological systems. Surprisingly, such behaviour can emerge from relatively simple negative feedback loops when regulatory delays are present. These oscillations illustrate that dynamic biological systems cannot always be understood from static network diagrams or simple time-series plots alone. Instead, we require representations that capture the complete state of the system and its evolution over time.

Self-check questions

Why can negative feedback produce oscillations instead of a stable equilibrium?
What role do biological time delays play in generating oscillatory behaviour?
Why are oscillations common in biological regulation?
Why can the same perturbation have different effects depending on when it is applied?
Why do oscillations motivate the introduction of state-space representations?