Cell types as the conserved unit of animal evolution

The standard textbook story of animal evolution is told in the language of body plans. Bilaterians have body axes; chordates have notochords; vertebrates have spines; mammals have hair. Phyla are defined by structural features. The deep evolutionary tree is read as a sequence of body-plan innovations.

A growing body of single-cell evidence suggests this story is told at the wrong resolution. The features that are conserved across animal evolution are not primarily body plans — they are cell types. Body plans are recombinations of a partially conserved cell-type toolkit. Once you make the shift, several puzzles in animal evolution become much less puzzling.

The classical view

The classical evolutionary view treats organisms as integrated systems in which selection acts on whole-organism phenotypes. The unit of evolutionary change is the trait — a wing, an eye, a digestive tract — and the unit of conservation is the trait's developmental program, often summarized by the network of genes that build it.

This view has produced enormous results. It is the basis of the modern synthesis, of Hox-pattern reasoning, of comparative anatomy. It is not wrong. But it has a known limitation: explaining the deep evolutionary persistence of structural features requires positing the persistence of large gene regulatory networks across enormous evolutionary distances, and the mechanisms by which such networks persist are not always clear.

The problem is most acute for animal phyla that diverged 600 to 700 million years ago. The fossil record at that depth is thin, and the molecular evidence has had to do most of the work. The classical view asks us to believe that complex regulatory networks were assembled before the bilaterian split and have been conserved, with modifications, ever since.

The cell-type view

The cell-type view starts from a different observation. When you compare cell-type atlases of animals separated by hundreds of millions of years, you find clear correspondences at the cell-type level even where the structures around them are non-homologous. The cell types — defined by their transcriptional state and regulatory program — persist; the structures recombine.

This implies a different unit of conservation. What was passed forward from the last common ancestor of bilaterians is not a body plan. It is a cell-type toolkit — a partially conserved set of regulatory modules, each instantiating a recognizable cell type, that descendant lineages have repeatedly reassembled into different structural configurations.

A neuron, an epithelial secretory cell, a muscle cell — each is a transcriptionally coherent identity that recurs across animal phyla, even where the organs containing them have evolved independently multiple times. The persistence of the cell type, not the persistence of the organ, is what selection preserves.

What this resolves

Several persistent puzzles in animal evolution become more tractable under the cell-type view.

Repeated evolution of complex structures. Eyes have evolved independently dozens of times across animal lineages. Under the classical view, this requires the independent assembly of complex developmental programs each time — a difficult thing to explain. Under the cell-type view, the constituent cell types (photoreceptors, pigment cells) are already in the toolkit; building an eye is a matter of assembly, not invention. The same applies to wings (insect, bird, bat), to limbs, to circulatory systems.

Cambrian explosion. The rapid appearance of disparate body plans in the Cambrian becomes less anomalous if the underlying cell-type toolkit was already substantially in place. Body plans diversified rapidly because the building blocks — the cell types — already existed and could be combined into novel architectures with comparatively modest regulatory modifications.

Deep regulatory conservation without structural conservation. The persistence of master regulators like Pax6 and Hox genes across animal lineages, even where the structures they pattern have changed, makes more sense if these regulators control cell-type identity rather than structural identity. Pax6 specifies a class of cell across animal evolution; the structures those cells participate in (compound eye, camera eye) are downstream and lineage-specific.

Functional convergence at the cell level. Cnidarians have neurons that are functionally indistinguishable from bilaterian neurons in many respects, despite the absence of a centralized nervous system. Cell-type homology offers a clean explanation: the neuron as a cell type predates the centralized nervous system, and was inherited from a common ancestor.

What this does not resolve

The cell-type view does not abolish the classical story. It refines it.

Body plans still matter. The integration of cell types into structures is not arbitrary; selection acts on whole organisms, not on cell-type catalogs. A toolkit explains what could exist; selection explains what does.

Many cell types are themselves products of relatively recent evolution. The mammalian neocortex contains cell types not found in other lineages. The vertebrate immune system contains specialized cells with no clear pre-vertebrate analogues. The cell-type toolkit is large but not closed; it has expanded over deep time.

Some structures have evolved without obvious cell-type homologues. Plants are the clearest example — their cell-type toolkit is largely independent of the animal one, having evolved separately from the unicellular ancestor. The cell-type view of animal evolution is genuinely an animal-evolution story; it does not generalize trivially to other domains of life.

Implications for the textbook

If the cell-type view holds up — and the evidence is increasingly strong that it does — undergraduate biology curricula will need structural revision. The story of animal evolution would be reframed around cell-type origin and elaboration, with body plans introduced as recombinations rather than as primary objects.

This is not a small change. It would shift the conceptual core of comparative animal biology from anatomy to cellular identity. It would also bring the field into closer alignment with how working evolutionary biologists actually think — many of whom have long regarded the body-plan-centric story as a useful pedagogical fiction more than a defensible primary account.

The textbook revision will lag the literature, as it usually does. The literature has already made the move.