Peptides That Became Medicines — And the Many That Did Not
Insulin was once an experimental extract that had never been given to a person. Around eighty peptide medicines are now approved worldwide. Here is how they got there — and why most candidates never do.
Why This Page Exists
There is a habit of thinking about peptides in two bins: the real medicines your doctor prescribes, and the experimental compounds discussed on the internet. The bins feel like different worlds made of different stuff.
They are not. Every peptide in the first bin was once in the second. Insulin was an extract of ground-up dog pancreas that had never been given to a human being. Semaglutide was a laboratory modification of a hormone found in the venom of a lizard. The distinction between the bins is not the nature of the molecules. It is whether anyone completed the work.
The bins are more like a queue than a wall. Everything approved passed through the same sequence of trials and review set out below, and everything unapproved is either still in the queue, was never entered into it, or was removed from it. Knowing which of those three applies to a given compound tells you far more than the bin it currently sits in.
A Century of Crossings
Insulin
Banting and Best extracted it from animal pancreas; a fourteen-year-old boy dying of type 1 diabetes recovered. The first peptide medicine, and still the clearest demonstration of what the class can do. The patent was sold for one dollar.
Sequenced and synthesised
Frederick Sanger determined insulin's full amino acid sequence — the first protein ever sequenced — proving peptides had definite, readable structures rather than being vague biological substances.
Synthesis becomes practical
Merrifield's solid-phase method made building peptides in a laboratory realistic, opening the door to designed analogues rather than only extracted material.
Analogues arrive
Desmopressin, leuprolide and others: naturally occurring peptides deliberately modified to last longer or act more selectively than the originals.
Recombinant production
Human insulin grown in bacteria replaced animal-extracted insulin, removing both the supply ceiling and the immune problems of using another species' protein.
Exenatide
Derived from a peptide in Gila monster venom, which happened to resist the enzyme that destroys the human equivalent within minutes. The first GLP-1 medicine, and the ancestor of everything that followed.
Semaglutide
Engineered for a much longer duration, then shown in large trials to affect weight as well as glucose. Became one of the most prescribed medicines in the world.
Tirzepatide
Acts at two receptors rather than one — the first dual agonist approved, and evidence that the design problem was still yielding to work.
Two threads run through the whole sequence, and both are worth carrying to any compound you read about. The first is that nature supplied the starting point almost every time — the interesting molecules were found, not invented. The second is that nature's version was almost never good enough on its own: human GLP-1 survives about two minutes in the bloodstream, which is useless as a medicine. Every crossing above involved taking something biology had already built and re-engineering the parts that made it impractical.
One Crossing, Close Up
The exenatide story is worth telling properly, because it shows how contingent these journeys are — and how little the outcome resembles a straight line from idea to product.
GLP-1 was known to be an excellent target long before anything could be done with it. It prompted insulin release when glucose was high and stopped when it was not — close to ideal. It was also destroyed by an enzyme within about two minutes of entering the bloodstream. A perfect message written in disappearing ink.
The solution came from an entirely unrelated observation. A researcher studying the venom of the Gila monster — a large desert lizard that eats rarely and then enormously — found a peptide closely resembling human GLP-1 but with a small structural difference at exactly the point the human enzyme attacks. The lizard's version simply did not present the target the enzyme was looking for. The same message, written in permanent ink.
That accident became exenatide, and exenatide established the pathway that liraglutide, semaglutide and tirzepatide were all built on. The entire GLP-1 category — now among the most consequential drug classes in modern medicine — traces back to somebody looking closely at lizard venom for reasons that had nothing to do with diabetes.
The lesson is not that unusual sources yield miracles. It is that the gap between an interesting molecule and a usable medicine is almost always an engineering problem — duration, stability, selectivity, delivery — and that solving it takes decades and enormous expense. The biology was understood in the 1980s. The medicine that made it useful arrived in 2005.
The Ones That Did Not Make It
A page listing only the successes would be dishonest, and it would be dishonest in a specific and recognisable way. Any field can look inevitable if you only count the survivors.
Consider what a hall of fame leaves out. Walk through one and you will conclude the sport produces champions reliably, because nobody builds a wing for the people who trained just as hard and did not make it. The room is not lying. It is simply not a random sample. Every list of peptide successes has the same shape — and the missing wing is very much larger than the one on display.
Most peptide candidates fail, and the reasons are worth knowing because they recur constantly in compounds discussed online today.
By far the most common ending. A mouse is not a small human, and effects that are robust in rodents routinely vanish in Phase II trials.
Many excellent targets were abandoned because nobody could engineer a version that survived long enough to be practical. GLP-1 nearly ended here.
Receptor families are related. A compound that hits the intended receptor and three of its relatives produces effects nobody wanted.
Not a scientific failure at all. Trials cost enormous sums, and promising compounds are shelved when funding stops or a company changes direction.
The last one deserves emphasis, because it is the most misused. A compound that was shelved for commercial reasons is often presented as suppressed — hidden knowledge that big pharmaceutical companies did not want you to have. The mundane explanation is nearly always the true one: the molecule was off-patent, or the market was small, or the company reorganised. That is a real and frustrating problem with how medicine gets funded. It is not evidence the compound worked.
What This Means For What You Read
Three things follow, and they are the reason this page sits in the library at all.
First, unapproved does not mean unpromising. Every approved peptide spent years in exactly the state most compounds in this library are in now. Dismissing a compound for lacking approval mistakes a stage for a verdict.
Second, and pulling hard in the opposite direction: promising does not mean it will work. Most compounds that looked exactly this good at exactly this stage did not make it. The history above is a survivor list. For every entry there are many more that had an equally compelling mechanism and an equally enthusiastic write-up.
Third, holding both at once is the whole skill. "This might be significant and probably will not be" is an uncomfortable position that resolves neither into excitement nor dismissal, which is precisely why almost nobody writing about peptides commercially adopts it. It also happens to be the accurate one.
To be plain about what this page is not: nothing in this history suggests any particular unapproved compound is on its way to becoming a medicine. Nobody knows that about any of them. The history is offered because it is genuinely interesting and because it makes the categories legible — not as an argument for any compound in it.
Frequently Asked Questions
Compiled by the Southern Pacific Labs research team. Last updated August 2026.
Published References
- Banting FG, Best CH, Collip JB, et al. “Pancreatic Extracts in the Treatment of Diabetes Mellitus.” Canadian Medical Association journal, 1922. PubMed
- Eng J, Kleinman WA, Singh L, et al. “Isolation and characterization of exendin-4, an exendin-3 analogue, from Heloderma suspectum venom. Further evidence for an exendin receptor on dispersed acini from guinea pig pancreas.” The Journal of biological chemistry, 1992. PubMed
- Wang L, Wang N, Zhang W, et al. “Therapeutic peptides: current applications and future directions.” Signal transduction and targeted therapy, 2022. PubMed
Related Research Guides
More in Foundations
Judging which compounds have the evidence to go further.
Read guide →FoundationsThe wider scientific story around these crossings.
Read guide →MetabolicWhere the Gila monster discovery ultimately led.
Read guide →