Become the Whole

Keeping with no stump left to anchor the pattern — a planarian fragment, holding no head and no tail of its own, remaking a complete worm in the right orientation, because the map is carried in its muscle and the builders in its flesh can become any part of it

R947 closed with a salamander that had a stump to work from: a sealed end, a cap, and cells whose position and trade were still given by the limb that remained. This page takes the turn it pointed to. Cut a planarian — in the work below, chiefly Schmidtea mediterranea — into pieces, and a middle piece with neither head nor tail does not restore a part. It becomes a whole animal again, head at the end that faced the old head, tail at the end that faced the old tail. What is kept is not a part, and not a place. What is kept is the pattern, in a piece too small to contain it whole.

First, how small a piece can carry it. The observation is old. T. H. Morgan reported in 1898 that a fragment about 1/279th the size of the intact worm could produce a complete, much smaller planarian. That figure is a reported result for the animals he studied, not a universal minimum for every species and every cut — but it sets the scale of the puzzle honestly: the fragment does not contain a spare head, a spare brain, or a spare anything. It contains ordinary-looking tissue, and somehow that is enough.

Second, the map is in the muscle. A fragment knows which end was nearer the head because positional information is not stored in one organ it might lose. It is written across the body, in position control genes (PCGs) expressed in regional patterns along the animal’s axes — and Witchley and colleagues (2013) found those genes expressed in muscle cells: body-wall, gut and pharynx muscle. Muscle therefore does two jobs here. It moves the worm, and it tells the rest of the worm where it is. After a cut, muscle cells in the remaining piece change which PCGs they express, re-scaling the map to the fragment’s new extent, so a piece that was the middle of a long axis acquires a head end and a tail end of its own. The pattern lives distributed, in a tissue every fragment is bound to contain some of.

Third, the head–tail decision is a switch that can be flipped. Wounding turns on Wnt signalling at wound sites generally. At a posterior-facing wound that signal stays high, acting through Smed-beta-catenin-1, and a tail is made. At an anterior-facing wound, notum — itself switched on by Wnt — is expressed strongly and damps Wnt down, and a head is made instead (Petersen & Reddien, 2011). The experiments that expose the switch are blunt: inhibit Smed-beta-catenin-1 and a wound that owed a tail grows a head, giving a two-headed worm; inhibit notum and an anterior-facing wound grows a tail instead of a head. Heads and tails, here, are not two separate programmes carried in two separate places. They are two outcomes of one signalling balance, read at each wound according to which way the wound faces.

Fourth, builders that, unlike the axolotl’s, did not keep a single trade. Regeneration is built by neoblasts, the worm’s proliferating adult stem cells, scattered through the body — which is why nearly any fragment contains some. And at least a subset of them breaks the rule R947 found in the blastema. Wagner, Wang & Reddien (2011) isolated clonogenic neoblasts (cNeoblasts) that formed colonies containing many differentiated cell types, and transplanted a single cNeoblast into a lethally irradiated worm that had no dividing cells of its own. That one cell repopulated the animal, its descendants replacing host tissue across gut, nerve, muscle, eye and skin, until the rescued worm was, genetically, the donor’s — and it could regenerate in its turn. One caution belongs inside the claim, not under it: this proves a subset of neoblasts is pluripotent. The neoblast population as a whole is mixed, and many of its members are already committed to particular fates. “Every builder can build everything” is more than the experiment shows.

Last, a whole made two ways at once. A fragment does not only grow what it lacks. New tissue is added at the wounds (the blastema route), and the old tissue that remains is remodelled and re-proportioned — Morgan’s morphallaxis — existing organs shrinking or shifting until they fit the smaller animal, so the result is not a large worm’s middle with a head and tail pasted on, but a small, complete, correctly proportioned worm, which then grows with feeding. That double work is the contrast with R947 stated plainly: the axolotl keeps by remaking a lost part from an anchored stump, out of specialists who remember their trade; the planarian keeps by a fragment becoming the whole — a distributed map re-scaled in muscle, a wound-facing switch deciding each end, and at least some builders for whom no trade is final.

A walk from a piece to a worm. Five stages: a fragment far too small to hold a spare animal; the map re-scaled, position control genes re-expressed in the fragment’s muscle; the head–tail switch read at each wound, notum damping Wnt where a head is owed and Wnt left high where a tail is owed; neoblasts building the missing ends, with single clonogenic neoblasts shown able, in the transplant test, to repopulate an entire animal; and morphallaxis re-proportioning the old tissue until piece and new growth are one small whole. Schematic, compressed from the published accounts.

Walk it in order.

Schematic only: the stages compress work across planarian species, fragment sizes and laboratory conditions, so no single cut runs this exact account. Labelled cautions: do not present Morgan’s 1/279 figure as a universal minimum — it is his reported result, and regenerative capacity varies by species, piece and position. Do not say all neoblasts are pluripotent — the single-cell transplant result belongs to clonogenic neoblasts, a subset, and the wider population includes fate-committed cells. Do not call a cNeoblast formally totipotent on this evidence — the asexual donor material in the classic test could not settle the germline question. Do not put the pattern in one place — muscle PCGs, wound signalling and neoblasts each carry a different part of it, and none works without the others. And do not call the animal immortal without qualification: uncut worms still age in their tissues even as turnover replaces them, and a fragment with no viable neoblasts — after lethal irradiation, for instance — regenerates nothing at all.