Thymosin and the Gland It Was Named For

For most of the twentieth century the thymus was the organ nobody could account for. It sits behind the breastbone, is at its largest in childhood and withers steadily thereafter, and for a long stretch of medical history the leading opinion was that it did very little at all. Then came the experiments showing that removing it from a newborn animal wrecked immunity in a way nothing else explained. Something came out of the thymus. Nobody knew what. So a group of biochemists did the only thing available to them: they took calf thymus tissue, ground it up, put the soluble material through a purification sequence and tested what survived for biological activity. In 1966 Allan Goldstein and colleagues reported a preparation that stimulated lymphocytes, and they called it thymosin 1. That is the entire reason the peptide is called thymosin. It came out of a thymus, it did something to the immune system, and the suffix was the one you gave a hormone. It was a sensible name. It has been wrong for roughly forty years.

An organ in search of a function

It is difficult now to recover how opaque the thymus was. The immune system had been described in outline, but the machinery that produced and instructed lymphocytes was guesswork. The thymus was implicated by subtraction: take it away early enough and the animal cannot mount the responses it should — leave it in and it withers by adulthood anyway, as though it had finished a job nobody could name.

The hypothesis that organised the field was endocrine. If the thymus mattered but was not itself the site of the action, it must be issuing instructions — secreting something that travelled and told lymphocytes elsewhere what to become. That was the standard shape of a 1960s explanation, with strong precedent in insulin, adrenaline and the pituitary hormones. Find the gland, find the secretion, name the secretion after the gland.

So the search was for a thymic hormone, and the assay was built to find one. The 1966 paper is titled for exactly that — a thymic lymphocytopoietic factor, meaning something from the thymus that makes lymphocytes 1. The name thymosin sits in that title in brackets, a proposed label for whatever the active principle turned out to be. That bracket is the whole problem in miniature: the name was assigned before the molecule was known.

What a fraction actually was

Here is the part modern readers tend to skate over, and it is the crux of everything that followed. Thymosin was not a molecule. It was a fraction — and to see why that matters you have to know what biochemistry could and could not do before high-resolution separation existed.

Start with kilograms of calf thymus. Homogenise it. Heat it, because the thing you want survives heating and much of what you do not want does not. Add salt until proteins begin falling out of solution, and collect what precipitates in a particular range — that is one cut. Add cold acetone and collect what drops out of that. Run the residue down a column packed with a gel that separates molecules by size, and collect the eluate in sequence: tube one, tube two, tube three. Pool the tubes that behave alike. Number the pools. Then ask each pool the only question you can actually answer, which is not "what is in here" but "does this one do the thing".

The fifth of those pooled cuts was where the activity concentrated, and so it was called fraction 5. That is genuinely all the name means. It is not a chemical description; it is a position in a queue. Thymosin fraction 5 was the fifth thing off a particular purification sequence performed on bovine thymus, and it was interesting because when you gave it to a lymphocyte assay, the assay responded.

This was not sloppiness. It was the honest limit of the method. You cannot name a molecule you cannot see, and in the 1960s seeing one meant sequencing it by hand, a residue at a time, at a cost that made it a project rather than a step. What you could do was purify towards an activity. The activity was the thing being tracked; the molecule was an inference. Whole careers of excellent biochemistry were conducted this way. But the method carries a characteristic hazard — an operational definition dressed as a chemical one. Thymosin was defined as whatever was in the tube that made the assay move.

And what was in the tube was not one thing. Fraction 5 was eventually resolved into a substantial number of distinct peptides, sorted into families by their behaviour in an electric field — an α series for the more basic, a β series for the acidic. The β-thymosins turned out to be a family in their own right, small acidic peptides distributed widely across animal species, often with more than one member in the same organism 5. A name used in the singular, as though it denoted a hormone, was already covering a crowd.

The sequence, and the discovery that undid the name

In 1981 the complete amino acid sequence of one of those components was published: bovine thymosin β4, 43 residues, fully specified 2. Careful work, and it converted an entry in a fraction into a defined molecule for the first time. Notice what the paper calls it in its own title — a thymic hormone. The framing the extract had been prepared to confirm was still attached to the molecule at the moment the molecule was finally pinned down. The sequence was new information. The interpretation around it was inherited.

That interpretation did not survive the following decade. As antibodies and probes became available and people looked for the peptide in tissues other than the one it had been extracted from, they kept finding it — not in traces, and not in a pattern suggesting a gland secreting into blood. Thymosin β4 is one of the most abundant peptides in the cytoplasm of a great many cell types across the body 35. Its concentration inside a cell can be startling. It is not a rare messenger. It is stock.

And its job is structural bookkeeping. Actin, the protein that gives cells their shape and drives their movement, exists in two states: free monomers drifting in the cytoplasm, and long filaments assembled from those monomers. A cell that needs to crawl, divide or close a wound must convert one into the other very quickly, which means keeping a large reserve of monomer that will not spontaneously polymerise while it waits. Thymosin β4 holds that reserve. It binds monomeric actin one to one and keeps it out of filaments until it is wanted 3.

So the accounting came out like this. A peptide named for a gland, with a suffix implying a circulating hormone, isolated in a programme designed to explain immunity, turned out to be a bulk intracellular component of the cytoskeleton's housekeeping. Not secreted. Not thymus-specific. Not, in the ordinary sense, a hormone at all.

Why the name stuck anyway, and what it cost

Nobody renamed it, and nobody was ever going to. There is no authority that can. A molecule's name is load-bearing infrastructure across thousands of papers, databases and reagent catalogues, and the cost of changing it falls on everyone who has to reconcile the old literature with the new label. The cost of leaving it alone is paid diffusely, by readers, in small increments, forever. Institutions reliably choose the second.

The trouble is that this particular name is not inert. Plenty of misleading names are harmless because they assert nothing — they are sounds attached to a structure. Thymosin asserts twice over. It names an organ, inviting you to believe the molecule belongs to that organ, and it takes the morphology of a hormone, inviting you to believe it circulates and signals. A reader who has never opened the primary literature picks up both implications free, from the word alone, and neither is true.

That inheritance is still visible in how the compound is described commercially, where the immune-hormonal flavour of the name colours the framing decades after the cell biology settled the question. Findings get corrected in print; a correction is a publishable act with a mechanism behind it. Names are not, because there is no equivalent mechanism and nobody whose job it is. Error in the record is self-repairing. Error in the vocabulary is not.

The second naming problem, which makes the first one worse

There is another layer, and it deserves stating plainly, because it is both genuinely funny and a genuine practical problem. The actin-binding activity of thymosin β4 does not require the full 43-residue chain. It concentrates in a short internal stretch of the sequence, and a short peptide is far cheaper to synthesise than a long one. Material circulating under the designation TB-500 is generally not full-length thymosin β4 but a shorter construct built around that active motif. TB-500 is a commercial label rather than a sequence identifier: it has no standardised definition in the chemical literature, and two suppliers using the name are not thereby selling the same molecule.

Follow the chain of reference down. TB-500 is named after a fragment of thymosin β4. Thymosin β4 is named after a fraction called thymosin. The fraction was named after the thymus, on a hypothesis about immune signalling that the molecule itself later contradicted. Every link in that chain points at something the compound is not.

The consequence is not academic. When a study reports an effect of thymosin β4, it has tested a defined, full-length, 43-residue peptide. Whether that result transfers to a shorter construct is an open question rather than a formality — peptide activity is frequently length-dependent, and dissection work on this molecule has found different reported activities localising to different parts of the sequence. Applying a full-length finding to a fragment takes an inferential step, and in most writing about this compound it is taken silently.

What the repair literature actually shows

None of this makes the molecule uninteresting. The reason anyone outside cell biology has heard of it is a body of work reporting that a peptide whose day job is holding actin monomers in reserve also appears to accelerate tissue repair — the idea caught neatly in one review's description of an actin-sequestering protein that moonlights 3. Reviews of the animal work describe effects across dermal wound models, corneal injury and cardiac injury 4.

The mechanistic story is at least coherent: repair is largely a migration problem. Cells must move into a damaged area, and movement is actin assembly, so a molecule governing the size of the monomer pool is plausibly placed to influence how fast that happens. Coherent, though, is a low bar, and the evidence needs stating exactly.

• The repair findings are essentially all preclinical — rodent injury models and cell culture, not human clinical endpoints.
• Consistency across several different injury models within one species is meaningful, but it is evidence about that species.
• The studies were conducted with full-length thymosin β4, which is not what the designation TB-500 generally denotes.
• Cardiac repair results in rodents have a long institutional history of failing to transfer to human outcomes.
• A peptide that promotes cell migration and vessel formation acts on processes that are not unconditionally desirable, and long-term human safety data does not exist.

Names outlive the hypotheses that made them

The useful generalisation is about what a name is doing. Some names describe where a thing was found, and those age gracefully, because where something was found does not stop being true. Some names assert what a thing is for, and those age badly, because purpose is precisely the claim most likely to be revised. Thymosin managed both at once: it recorded an origin and, through its form, asserted a function. The origin was accurate and irrelevant. The assertion was the part everyone kept.

It is not an isolated case. BPC — body protection compound — states a conclusion about a molecule's purpose that predates any conclusion being available, and that name now does a great deal of persuasive work on readers who have never examined what supports it. The pattern repeats across the peptide field, which is unusually rich in compounds christened while their function was still a hopeful guess.

So a reasonable reading habit: when a compound's name makes a claim about what it does, treat the claim as an artefact of the date of discovery rather than a finding. It tells you what the people who first purified the substance were hoping to find, and which assay they ran while they hoped. It tells you almost nothing about what the molecule turned out to be — because by the time that was known, the name had been printed on too many things to change.