Morphogenesis and form  ·  Article

Morphostasis and robustness of morphogenesis to noise

One model has cells compare a remembered target signal with the current one and stop growing when the two match.3 Later simulations test how noise and evolution shape this ability. Hartl and colleagues found that lineages which efficiently solved morphogenesis typically also evolved to solve morphostasis efficiently.8 Pio-Lopez and colleagues (2025) suggest that development and morphostasis still diverge.9

Earliest held
2012, Lobo D, Beane WS, Levin M
Most discussed in
Evolutionary Implications of Self-Assembling…, 2024
In the library
38 passages in 15 works
Rewritten
2026-10-27
01

Noise and its meanings

The word noise is used in more than one sense. Pai and colleagues (2012) used developmental noise for fluctuating asymmetry, simple left-right differences that arise from noise. They separated this from biased left-right patterning, which is coordinated with the other body axes.1 A different use comes from imaging. Dolenšek and colleagues (2015) reported that membrane potential recorded from a few cells or a small stretch of plasmalemma is exceedingly noisy in time.2 That is noise in the measurement, not in the embryo.

02

Target memory and reinforcement

Bessonov and colleagues (2015) proposed a regeneration model based on cell memory. Each cell stores the total signal it receives. When the signal changes, the cell makes a growth-stimulating substance in proportion to the difference from the stored value.3 Growth therefore ends when the old signal is restored. The authors presented this as one answer to an open question about morphostasis. Newman (2019) took a more guarded view of bioelectricity. He wrote that it has thus far not been demonstrated to be a primary form-generating or patterning mechanism, but that voltage gradients reinforce patterns and restore them if perturbed.4 He also named canalizing evolution as a stabilizer against mutation and developmental noise.5

03

Simulated noise and competency

Pio-Lopez and colleagues (2022) simulated cells that infer their identities. When some cells held too high an expectation about identity, differentiation was incomplete, and the authors attributed the resulting disorder of inference to developmental noise.6 Reducing signal concentration and sensitivity in two over-precise cells cured the defect in the simulation and removed the noise.7 Hartl, Risi and Levin (2024) evolved cellular automata at several noise levels. Lineages that solved morphogenesis efficiently typically also solved morphostasis efficiently.8

04

Maintenance after development

Pio-Lopez, Hartl and Levin (2025) argued that existing models neglected the lack of anatomical morphostasis after development. In their simulations, aging effects appeared in long runs even when the model parameters were not perturbed. They concluded that development and morphostasis still diverge, although the tasks are close.9 They suggested that providing the tissue with a regenerative goal could reverse this decline. Their evidence comes from simulation only.

SourcesEach quotation was checked word for word against the passage it opens.
  1. a distinct phenomenon from fluctuating asymmetry, which involves simple differences between the left and right sides derived from developmental noisePai VP, Vandenberg LN, Blackiston D, Levin M, 2012 · Neurally Derived Tissues in Xenopus laevis Embryos Exhibit a Consistent… · open at passage 3
  2. The MP signal from area covering few cells or a small part of plasmalemma (square, Figure 6A,B) is exceedingly noisy in the time domainDolenšek J, Špelič D, Klemen MS, Žalik B, Gosak M, Rupnik…, 2015 · Membrane Potential and Calcium Dynamics in Beta Cells from Mouse Pancreas… · open at passage 30
  3. If the total signal changes and becomes u, the cell produces some substance with the rate proportional to u* − u. This substance stimulates the proliferation of new cells.Bessonov N, Levin M, Morozova N, Reinberg N, Tosenberger A…, 2015 · On a model of pattern regeneration based on cell memory · open at passage 54
  4. bioelectric establishment of voltage gradients across tissues serves to reinforce developmental patterns and restore them if perturbed during embryogenesis or woundingNewman SA, 2019 · Inherency of Form and Function in Animal Development and Evolution · open at passage 71
  5. canalizing evolution that serves to stabilize the phenotype of an evolutionary lineage against mutation and developmental noise once it has become ecologically establishedNewman SA, 2019 · Inherency of Form and Function in Animal Development and Evolution · open at passage 71
  6. There is an important developmental noise, which leads to a disorder of inference for the other cells in terms of differentiation.Pio-Lopez L, Kuchling F, Tung A, Pezzulo G, Levin M, 2022 · Active inference, morphogenesis, and computational psychiatry · open at passage 87
  7. After the simulated biomedical intervention, as we can observe in Figure 8, the developmental defect is cured and all cells reach their location with the appropriate differentiation.Pio-Lopez L, Kuchling F, Tung A, Pezzulo G, Levin M, 2022 · Active inference, morphogenesis, and computational psychiatry · open at passage 91
  8. lineages that efficiently evolved to solve morphogenesis in our experiments also (typically) evolve to solve morphostasis efficientlyHartl B, Risi S, Levin M, 2024 · Evolutionary Implications of Self-Assembling Cybernetic Materials with… · open at passage 57
  9. These results suggest that development and morphostasis, while close in terms of morphological trajectories and tasks, still significantly diverge.Pio-Lopez L, Hartl B, Levin M, 2025 · Aging as a Loss of Goal-Directedness: An Evolutionary Simulation and Analysis… · open at passage 98
Linked ideas
The Bessonov model supplies a concrete mechanism for stopping growth when the stored target signal is matched.
Morphostasis is the maintenance of a target morphology once reached.
Newman names canalizing evolution as a way to stabilize phenotype against noise, alongside bioelectric reinforcement.
Left-right asymmetryin tension with
Fluctuating asymmetry from noise is distinguished from consistently biased left-right patterning.
Holding a form against disturbance extends the older idea of holding a physiological variable steady.
The 2025 simulations tie aging to loss of a goal and rejuvenation to restoring a regenerative one.
Evolvabilityrelated to
Evolution shapes morphostasis: Pio-Lopez says evolution prioritized development over long-term morphostasis, so aging follows without regenerative goals; Newman says canalizing evolution stabilizes phenotypes against developmental perturbation.
Passages present NCAs as the model of morphogenesis, growing tissue and maintaining it against cell-state errors; morphostasis is the maintenance phase, and aging is modeled as its failure in an NCA.
The noise results concern how cells keep a target pattern reliable when decisions are imperfect.
Where it is discussedPassages matching morphostasis, developmental noise, correct cell state, denoising
2024Evolutionary Implications of Self-Assembling Cybernetic Materials with… · Hartl B, Risi S, Levin M13
2015Membrane Potential and Calcium Dynamics in Beta Cells from Mouse Pancreas… · Dolenšek J, Špelič D, Klemen MS, Žalik…6
2025Aging as a Loss of Goal-Directedness: An Evolutionary Simulation and Analysis… · Pio-Lopez L, Hartl B, Levin M5
2022Active inference, morphogenesis, and computational psychiatry · Pio-Lopez L, Kuchling F, Tung A…2
2012Modeling planarian regeneration: a primer for reverse-engineering the worm · Lobo D, Beane WS, Levin M2
2026Atavistic Genetic Expression Dissociation (AGED) During Aging… · Pio-Lopez L, Levin M1
2012Neurally Derived Tissues in Xenopus laevis Embryos Exhibit a Consistent… · Pai VP, Vandenberg LN, Blackiston D…1
2019Inherency of Form and Function in Animal Development and Evolution · Newman SA1
2022When left does not seem right: epigenetic and bioelectric differences between… · Masuelli S, Real S, Campoy E, Branham…1
2019Modeling somatic computation with non-neural bioelectric networks · Manicka S, Levin M1