GLP-1 was not discovered in lizard venom. The peptide found in Gila monster venom is exendin-4, the product of a separate gene expressed in the lizard's salivary gland — a different molecule from GLP-1 altogether. The real story is better than the myth: a venom peptide from an unrelated gene turned out to fit the mammalian GLP-1 receptor well enough to become the first medicine in the class. The rest of the history runs across forty years of gut physiology, gene sequencing and molecular biology, and the credit for the central discovery is still being argued in the literature.
Exendin-4: a 39-amino-acid peptide isolated in 1992 from the venom of the Gila monster. It is roughly 50% homologous to human GLP-1 and binds the mammalian GLP-1 receptor, but it is encoded by a different gene, in a different tissue, in a different animal. It is not lizard GLP-1.
What was known before anyone had the molecule
From the 1960s it was known in humans that swallowed glucose produces a far larger insulin response than the same glucose infused into a vein. Two 1964 papers reported the observation. Perley and Kipnis then quantified it in 1967: across normal-weight, obese, diabetic and non-diabetic subjects, the insulin response to intravenous glucose was only 30 to 40% of the response to oral glucose, and studies in subjects with diverted portal circulation placed the mechanism in the intestinal tract (incretin-effect observations).
That effect got a name — the incretin effect — long before anyone knew which molecule produced it. By 1986 Nauck and colleagues had measured it in people with type 2 diabetes and found it roughly halved: the incretin contribution to the total insulin response was 72.8% in control subjects and 36.0% in patients with type 2 diabetes (incretin-effect observations).
The first incretin hormone to be isolated was not GLP-1 at all. GIP was isolated by John Brown in the 1970s (Drucker, Habener & Holst 2017). GLP-1 arrived through a different route entirely — not by chasing the physiology, but by reading a gene.
How GLP-1 itself was found, and in what species
Almost all of the foundational GLP-1 work was done in fish, rats, pigs and cultured cells, with a single small human study at the end of it — and the mechanism it established does carry over to humans.
In the early 1980s the Habener laboratory worked out proglucagon sequences from cDNAs and genes isolated from anglerfish, and Graeme Bell and others identified the corresponding hamster, bovine and human sequences (Drucker, Habener & Holst 2017). Proglucagon turned out to be a long precursor protein with several peptides embedded in it, glucagon among them.
The mid-1980s work established that the same gene yields different peptides in different tissues. In pancreatic and intestinal tissue, Mojsov and colleagues showed in 1986 that preproglucagon mRNAs are identical in the two organs while the peptides liberated from them differ, because the processing after translation is cell-specific. In cell lines, Drucker, Mojsov and Habener showed the precursor being cut into glucagon, glicentin, GLP-1(1-37), GLP-1(7-37) and other fragments (discovery papers).
Then 1987, and three results that between them made a drug class possible. In pig tissue, Holst and colleagues isolated a peptide from intestinal mucosa that dose-dependently and potently increased insulin secretion from the isolated perfused pig pancreas. In rat tissue, Mojsov, Weir and Habener showed that GLP-1(7-37) stimulated insulin secretion at concentrations as low as 5 × 10⁻¹¹ M, while the longer GLP-1(1-37) had no effect at all even at ten thousand times that concentration. And in December 1987, in the first human demonstration, Kreymann, Williams, Ghatei and Bloom infused GLP-1 7-36 into seven volunteers at a rate mimicking the concentration seen after a meal: insulin rose significantly while glucose and glucagon fell, and they concluded it was a physiological incretin more powerful than GIP (discovery papers).
The rat result is the hinge of the whole field. Which fragment is active — the truncated one, not the full-length one — is the fact everything downstream depends on.
Who discovered GLP-1 is genuinely contested
The two most authoritative records name different people, and that disagreement is the accurate account.
The 2024 Lasker~DeBakey Clinical Medical Research Award was given for the discovery and development of GLP-1-based drugs for obesity. Its laureates were Joel Habener of Massachusetts General Hospital, Svetlana Mojsov of The Rockefeller University, and Lotte Bjerre Knudsen of Novo Nordisk. Daniel Drucker and Jens Juul Holst are not laureates on that award (contested attribution).
The most-cited historical review of the field is by Drucker, Habener and Holst, published in 2017. Mojsov is not an author of it, and the review describes her as working "in the Habener laboratory" — which is precisely the framing others contest (Drucker, Habener & Holst 2017). Mojsov has since published her own account of the discovery in JAMA, in 2024 and again in 2025 (contested attribution).
That a co-discoverer felt the need to publish her own account of a discovery in a general medical journal, decades after the fact, is itself the documentary evidence that the attribution is disputed. What the record supports is a contest over credit among several groups working across more than a decade — not a single eureka, and not an allegation about anyone's conduct.
The lizard: what the 1992 paper actually reported
The 1992 discovery paper is animal and test-tube work from beginning to end, and it reports nothing at all about GLP-1, insulin, glucose, appetite or weight.
What John Eng and colleagues did was isolate a 39-amino-acid peptide, which they named exendin-4, from the venom of the Gila monster, Heloderma suspectum. The search was prompted by an earlier finding in a different species: exendin-3, identified in the venom of Heloderma horridum, the beaded lizard. Exendin-4 differs from exendin-3 by two amino acid substitutions and is otherwise identical. The pharmacology in the paper is a cAMP response in dispersed pancreatic cells from guinea pigs, beginning at 100 pM and plateauing at 10 nM, blocked by an exendin receptor antagonist (Eng 1992).
That is the whole of it. Two lizard species, one new peptide, and a receptor response in guinea-pig tissue. The comparison in the paper is between exendin-4 and exendin-3, not between exendin-4 and human GLP-1. Everything the general reader associates with this story came later and from other laboratories.
Why "GLP-1 was discovered in lizard venom" is wrong
Because exendin-4 and GLP-1 are different peptides, encoded by different genes, expressed in different tissues.
In lizard tissue, Chen and Drucker showed in 1997 that exendin mRNA transcripts were expressed in the salivary gland but not in pancreas or intestine, while the lizard's own proglucagon mRNA was not detectable in the salivary gland at all. Their conclusion, verbatim: GLP-1 and exendin-4 "represent related yet distinct peptides encoded by different genes" (Chen & Drucker 1997).
In lizard tissue plus rat and human blots, Pohl and Wank showed in 1998 that exendin-4 is approximately 50% homologous to GLP-1 and interacts with the mammalian GLP-1 receptor with equal or higher affinity and efficacy. They also went looking for a mammalian version of exendin-4 across a range of rat and human tissues and found none (Pohl & Wank 1998).
So the accurate sentence is the more interesting one. A venom peptide, produced by a gene humans do not have, in an organ that has nothing to do with the gut, happens to fit the mammalian receptor at least as well as the human hormone does. That coincidence — not a discovery of GLP-1 in a lizard — is what turned into the first drug in the class.
Approval, jurisdiction by jurisdiction
Approval dates differ by country, and none of them travels. A bare year is close to meaningless without the jurisdiction attached.
In the United States, FDA's own records give the original approval dates for the class (FDA): exenatide on 28 April 2005, the first GLP-1 receptor agonist approved there; liraglutide on 25 January 2010 and a second liraglutide product on 23 December 2014; semaglutide injection on 5 December 2017, an oral semaglutide tablet on 20 September 2019, and a further semaglutide injection product on 4 June 2021; tirzepatide on 13 May 2022, and a second tirzepatide product on 8 November 2023.
In the European Union, exenatide was authorised on 20 November 2006 — about eighteen months behind the United States — with a registered indication of type 2 diabetes (EMA).
In Singapore, the register records different events again. Semaglutide injectable products were registered from 21 April 2021 and an oral form from 1 June 2021, with further injectable products in February and March 2023; tirzepatide injectable products were registered from 1 March 2023. Every one of them is classified Prescription Only (HSA register). Registration is also not the same as launch — a product can be registered here well before it is available.
One point of framing is worth stating plainly, because it is often reversed. The first medicine in this class was registered for glycaemic control in type 2 diabetes in both the United States and the European Union (EMA). Weight management came later, and on different products. The class spent most of its clinical history as diabetes medicine, which is why the weight-management evidence base is younger than the class itself — and why what is registered for what, in which country, is a question worth checking rather than assuming. For Singapore, that is covered in GLP-1 medication in Singapore: what is legal and licensed.
Where the class goes next — one agent approved overseas, two still in clinical trials — is in what is actually coming.
Common questions
Was GLP-1 discovered in lizard venom?
No. The peptide isolated from Gila monster venom is exendin-4, and it is encoded by a separate gene expressed in the lizard's salivary gland rather than being the lizard's version of GLP-1 (Chen & Drucker 1997). GLP-1 itself was identified through work on the proglucagon gene in the mid-1980s (discovery papers).
What is exendin-4?
A 39-amino-acid peptide isolated in 1992 from the venom of the Gila monster, Heloderma suspectum (Eng 1992). It is approximately 50% homologous to human GLP-1 and binds the mammalian GLP-1 receptor with equal or higher affinity and efficacy, which is how a lizard peptide could become a human medicine — and probing rat and human tissues found no mammalian equivalent of it (Pohl & Wank 1998).
Who discovered GLP-1?
The credit is genuinely contested. The 2024 Lasker~DeBakey Clinical Medical Research Award named Joel Habener, Svetlana Mojsov and Lotte Bjerre Knudsen, and did not include Daniel Drucker or Jens Juul Holst; the most-cited historical review of the field was written by Drucker, Habener and Holst without Mojsov; and Mojsov has since published her own account of the discovery in JAMA (contested attribution).
When was the first GLP-1 medication approved?
Exenatide was approved in the United States on 28 April 2005 (FDA) and authorised in the European Union on 20 November 2006 (EMA). Singapore's own registration events are different again — semaglutide products from April 2021 and tirzepatide from March 2023 (HSA register). Every date belongs to its jurisdiction.
Was the first GLP-1 medicine a weight-loss drug?
No. Exenatide was registered for type 2 diabetes in both the United States and the European Union (EMA). Weight management came later, and on different products.