'GLP-3' Is Not a Hormone: What People Mean When They Say It

I spent part of this week reading about a hormone that does not exist, which in this corner of drug development is more normal than it sounds. If you typed GLP-3 into a search engine, here is the short answer: there is no such hormone, no such receptor, and no recognized class of drugs by that name. It is a nickname.

The nickname mostly points at retatrutide, an experimental weekly injection that a review on hormone-receptor drugs for chronic kidney disease describes, accurately and without excitement, as “a triple GLP-1/GIP/glucagon receptor agonist.” It hits three receptors. GLP-1 is famous, so a drug that works on three pathways gets called the three version (the arithmetic is doing a lot of work here). That arithmetic is wrong, but it is wrong in an instructive way, and following it shows who gets paid for what in the business of naming molecules.

What the “1” in GLP-1 actually counts

GLP-1 is glucagon-like peptide-1, a hormone made in the gut. A structural paper that imaged the activated GLP-1 receptor with cryo-EM summarizes its job as having “essential roles in regulating insulin secretion, carbohydrate metabolism, and appetite.” The “1” is part of the hormone’s name. It is not a tally of receptors, pathways or anything a drug does, and a drug does not move up the numbers by doing more.

The receptors involved are cousins. Pharmacologists call them class B1 GPCRs, which are polypeptide hormone receptors with a large outer section called the N-terminal extracellular domain. Binding is described by the “two-step binding model,” in which residues near the hormone’s tail end first “interact with the N-terminal ECD of the receptor,” and only then does the front end move inward to “transmit signals.” So the hormone gets acquainted at the door before it is invited in. Very proper.

The family resemblance is, I guess, what makes it plausible to ask one peptide to stand in for several hormones at once. That is the design bet behind tirzepatide (two receptors) and retatrutide (three). None of it creates a new hormone. It creates a molecule that some receptors mistake for their own.

How three receptors turned into GLP-3

Here is the chain, one link at a time.

The first link is the drug companies, which get paid for molecules that beat the last molecule. A review of GLP-1 receptor biology lists dual and triple agonists as one of “the three major axes of next-generation incretin drug development,” alongside biased signaling and pills. Adding receptors is a product strategy, and the trial numbers have rewarded it so far. Tirzepatide, the dual GIP/GLP-1 agonist, produced mean weight reductions of 19.5% and 20.9% at its 10 mg and 15 mg doses in the SURMOUNT-1 phase 3 trial. Retatrutide, in the phase 3 TRIUMPH-1 trial, produced mean weight loss of 28.3% at 80 weeks on the 12 mg dose, 25.9% on 9 mg and 19.0% on 4 mg. AJMC reported the 12 mg figure as “a mean weight loss of 70.3 lb (28.3%) at 80 weeks, from a baseline average body weight of 248.5 lb.” (How the trial counted people who stopped treatment matters a lot here, and the site’s careful read of those results handles it.)

The second link is the official names, which are not built for talking about at dinner. “Retatrutide” is a word most people mistype at least once (I have managed it several times in this article alone). So writers reach for nicknames, and the field has precedent: a cardiology review of tirzepatide notes that its “unique dual mechanism” earned it “the nickname ‘twincretin.’” That one at least used a real category, the incretins, and added a real count.

The third link is the public, which already knows one term. GLP-1 became shorthand for a whole drug category, the way a brand name becomes a verb. Once that happens, a drug that does more than GLP-1 needs a bigger number, the way a sequel needs one (nobody went to see Toy Story 3 thinking it was about three toys). Naming a drug by its receptor count this way is like a landlord numbering apartments by how many bedrooms they have: the label tells you something true about the unit, and it will reliably send the mail carrier to the wrong door.

The fourth link is search engines and the websites that live off them. People type what they have heard, so pages appear that answer what people typed. Somewhere an editor says, “Can we get a GLP-3 explainer up by Friday, the traffic is just sitting there.” This article is, in some sense, that explainer. Lol.

And here is where the chain arrives: a drug class that does not exist now has explainers, comparison pages and search volume, which is most of what a drug class has in public life. Nobody planned this. Each link did the locally sensible thing.

Why glucagon makes the nickname stranger

The third receptor retatrutide hits is the glucagon receptor. GLP-1 means glucagon-like peptide. So the nickname takes a drug that activates actual glucagon receptors and names it after a hormone defined by resembling glucagon, then adds a number as though the imitation had a sequel. It is a strange compliment.

The glucagon piece is also not decorative. A review of glucagon signaling notes that mazdutide, survodutide and retatrutide all “incorporate GLP-1 receptor-activating moieties to counteract glucagon’s hyperglycaemic effects.” Glucagon raises blood sugar, which is an awkward property for a metabolic drug, so the GLP-1 component partly exists to cancel it out. Another review argues that GLP-1/glucagon agonism adds energy expenditure, by stimulating fat oxidation and glucose production, rather than relying mainly on appetite. Why a weight-loss drug would want glucagon at all is its own long story, told elsewhere on this site.

So the receptor that does the most distinctive work in the molecule is the one the nickname leaves out. Typical, really.

Does GLP-3 even mean one drug?

Not really. Retatrutide is not the only triple. UBT251, a triple GLP-1/GIP/glucagon agonist linked to Novo Nordisk, produced weight loss of 13.6% to 19.7% versus 2.0% for placebo in a phase 2 trial of 205 adults in China over 24 weeks. By the nickname’s arithmetic, that is also GLP-3. A label that fits every three-receptor molecule is a category, and the explainer comparing retatrutide with semaglutide and tirzepatide says flatly that the label “does not correspond to any recognized scientific drug class.” So it is a category that the people who define categories say is not a category. Fine for search, I suppose, but dumb for anything else.

Now a scenario, clearly made up. Suppose a company adds amylin, the satiety hormone already paired with GLP-1 in CagriSema and amycretin, to a triple agonist. The internet calls it GLP-4. A rival bolts on a fifth receptor and press releases start leading with the number. By the end of the decade there is a GLP-9 that activates receptors for appetite, insulin, liver fat, bone, sleep and possibly whether you finish a television series, and analysts are asking on earnings calls why the pipeline is “stuck at seven.” Pharmacologists keep writing “triple agonist” and “quadruple agonist” in their papers, which nobody searches for.

The real version is less silly but has the same structure. Receptor count is easy to say and easy to compare. Whether a given receptor helps, and how much, is not easy to say. GIP is the cleanest example: tirzepatide activates the GIP receptor, while Amgen’s maridebart cafraglutide blocks it, and both seem to add weight loss to GLP-1 agonism. A paper examining that puzzle calls it “a central unresolved scientific question in incretin pharmacology.” A naming scheme that counts receptors cannot even tell agonism from antagonism.

So what should people call it?

The accurate name is a triple GLP-1/GIP/glucagon receptor agonist, which no one will ever say out loud at a barbecue. Whether retatrutide reaches the market, and under what brand name, is not settled by the trial results so far; TRIUMPH-2, in adults with obesity and type 2 diabetes, reported smaller losses (18.8% at 12 mg versus 5.1% on placebo at week 80) and the regulatory path is still being walked.

One reading is that GLP-3 is just an error, a misread number that stuck because it was convenient. The other reading is that it is, oddly, the most useful name the drug has, because it encodes the one fact the public actually wants (this is the next one, and it does more) in five characters. Those readings point in opposite directions and both seem true. For instance, a person who says GLP-3 at a pharmacy counter and a chemist who winces at it probably mean the exact same molecule.

I spent the week reading about a hormone that does not exist, and by the end of it this article had become one more small reason that it sort of does.