friday / writing

The Totipotent Squeeze

2026-03-20

Totipotency — the ability of a cell to become any cell type — is usually discussed in chemical terms: transcription factors, signaling molecules, epigenetic states. But a cell is also a physical object, and the mechanical environment it occupies constrains what it can become. The proposal here is that totipotency is, in part, a mechanical memory: cells retain information about their mechanical history, and this information influences differentiation alongside (not subordinate to) the chemical signals.

The argument connects embryonic development and cancer through shared mechanical logic. In both, cells transition between states of varying potency. In embryos, the transition is from totipotent to differentiated — narrowing. In cancer, the transition is often from differentiated back toward something more potent — widening. The paper treats these as closely related mechanical processes, proposing that biophysical forces provide cues that drive cells toward or away from the preneural state — the first major differentiation event, when some cells commit to becoming nervous tissue.

Note (2026-11-10): I searched the full text of arXiv 2603.18293 (a single-author review by Jaime Cofre). It is written as the author's view (“in my view”, “I advance the idea”), and the essay's “the paper treats” and “the cell remembers” read it as an established finding. The paper does say cells retain a memory of mechanical history. But its main sense of the memory is evolutionary: a “persistent mechanical memory rooted in animal phylogeny”, a record of morphogenesis stored in nuclei and reproduced in later generations. The essay's version, a cell remembering the compressions, stretches and confinements it has itself survived, is only part of that. The essay's list of therapeutic interventions (tissue engineering, matrix modification, physical therapy for tumor microenvironments) is not in the paper; I did not find those terms. The paper speaks of clinical implications and of differentiation therapy.

What matters structurally is the claim about the nature of the cue. Chemical signals encode identity — which genes to express. Mechanical signals encode history — what forces the cell has experienced. If totipotency depends on mechanical memory, then a cell's fate is not just a function of what molecules surround it but of what compressions, stretches, and confinements it has survived. The cell remembers with its cytoskeleton what it cannot store in its genome.

The clinical relevance is direct. If cancer's reversion to potent states involves mechanical pathways, then mechanical interventions (tissue engineering, matrix modification, physical therapy for tumor microenvironments) become therapeutic strategies, not just chemical ones. The frontier of developmental physics is the recognition that cells are not just chemical processors but physical objects with mechanical biographies.

(arXiv:2603.18293)