Insulin: The First Peptide Medicine

By Priya Nandakumar · 30 June 2026 · 7 min read
A computer generated image of a spiral of bubbles

A molecule that changed the clinic , and the chemistry textbook

Ninety-nine amino acids. Two chains. One disulphide bridge holding the whole thing together. Insulin is not a complicated peptide by modern standards, but in 1921 it was essentially a miracle , and the story of how it got from a dog's pancreas in Toronto to a Schedule 4 prescription product on the Australian Register of Therapeutic Goods is one of the clearest illustrations of what peptide medicine actually is.

I've been turning this story over in my mind for a few months now, partly because insulin keeps coming up whenever I write about newer GLP-1 receptor agonists , molecules that are structurally cousins to insulin and that sit in the same regulatory and pharmacokinetic neighbourhood. Understanding insulin well is, I'd argue, the single best entry point into understanding the whole class.

What insulin actually is, structurally

Insulin is a peptide hormone produced in the beta cells of the pancreatic islets of Langerhans. It consists of an A-chain (21 amino acids) and a B-chain (30 amino acids), linked by two interchain disulphide bonds, with a third disulphide bond internal to the A-chain. Those bonds are not decorative; they hold the three-dimensional shape that allows the molecule to bind its receptor.

The peptide bonds linking each amino acid residue are what make insulin a peptide rather than a small molecule drug. That distinction matters enormously for delivery. Peptide bonds are hydrolysed in the gastrointestinal tract, digestive enzymes do exactly what they're designed to do and break the molecule down before it can reach systemic circulation. That's why insulin cannot be taken orally in a standard formulation. It requires subcutaneous injection, intravenous infusion, or, in more recent inhaled formulations; absorption across the alveolar membrane.

The pharmacokinetics of subcutaneous insulin are worth pausing on. After subcutaneous injection, insulin is absorbed into local capillaries and lymphatics, with a time-to-peak that varies depending on formulation. Short-acting analogues are engineered to disrupt the tendency of insulin hexamers to self-associate, speeding up monomer absorption. Long-acting analogues introduce modifications, substitutions at specific amino acid positions, or fatty acid chains that anchor the molecule to albumin, to slow absorption and flatten the concentration-time curve. Same basic molecule, radically different PK profile. That's pharmacokinetics doing real clinical work.

The 1921 discovery and why it still matters

Frederick Banting and Charles Best, working in J.J.R. Macleod's laboratory at the University of Toronto, isolated the pancreatic extract they called "isletin" in the summer of 1921. James Collip then purified it sufficiently for human use. Leonard Thompson, a 14-year-old patient, received the first human injection in January 1922. The initial attempt caused a severe reaction; Collip's extract was impure, but the refined injection two weeks later worked. Thompson survived. Banting and Macleod were awarded the Nobel Prize in Physiology or Medicine in 1923, though the credit distribution remains contested to this day. Honestly, that particular dispute has all the drama of a true-crime episode without the crime.

The reason this history matters for a pharmacology audience is what happened next: the Toronto team sold the patent to the University of Toronto for one dollar each, explicitly so that no single commercial interest could monopolise access. That decision shaped how the medicine was scaled and distributed. Within a year, pharmaceutical manufacturers on multiple continents were producing bovine and porcine insulin. The molecule went from bench to global supply faster than almost anything before it.

From animal extract to recombinant production

Bovine and porcine insulins dominated clinical use for six decades. They differ from human insulin by three and one amino acid residues, respectively. That difference is small, small enough to be clinically workable for most patients; but large enough to trigger immunogenic responses in some.

The shift came in 1982, when the US FDA approved the first recombinant human insulin, produced by inserting the human insulin gene into Escherichia coli. Recombinant DNA technology meant the amino acid sequence was now exactly human. Australian regulatory history followed: recombinant human insulins and analogues are now listed on the Australian Register of Therapeutic Goods (ARTG), and their manufacture is subject to Good Manufacturing Practice requirements enforced by the TGA.

In Australia, all insulin formulations are regulated as prescription-only medicines under Schedule 4 of the Poisons Standard, meaning supply requires a valid prescription from an authorised medical practitioner. The Schedule 4 classification governs how pharmacies dispense these medicines, how they're advertised, and what representations can be made about them. Insulin is not and has never been available over the counter in the way paracetamol is. The regulatory framework exists for clear safety reasons: the dose-response relationship is narrow and the consequences of error are serious.

It's worth noting that the TGA's framework for peptide medicines sits within a broader system that includes the ARTG, the Special Access Scheme, and the Authorised Prescriber pathway for unapproved medicines. Insulin products on the ARTG are approved; the SAS and AP pathways exist for medicines not yet listed, and come with their own obligations and scrutiny.

Insulin's receptor mechanism, simplified

The insulin receptor is a receptor tyrosine kinase; a transmembrane protein with two alpha subunits (extracellular, binding site) and two beta subunits (intracellular, kinase domain). When insulin binds the alpha subunits, it triggers a conformational change that activates the kinase domains on the intracellular side. Those kinase domains autophosphorylate and then phosphorylate downstream signalling proteins, initiating a cascade that (among other effects) drives translocation of GLUT4 glucose transporters to the cell membrane in muscle and adipose tissue.

The receptor mechanism is what distinguishes insulin from the GLP-1 receptor agonist class, which act on a G-protein coupled receptor. Different receptor superfamily, different intracellular signalling logic, though both classes are peptides, both require injection, and both sit in the broader landscape of endocrine pharmacology.

I spent a weekend recently going back through the original 1969 Dorothy Hodgkin crystallography paper on insulin's structure, she was the first to map it by X-ray diffraction, which is a sentence that should make every pharmacologist briefly stop and appreciate how hard that was before computational tools existed. She did it at the Physical Chemistry Laboratory at Oxford. The structures she published are still referenced in the insulin analogue literature today.

A note on unregulated peptide products

Insulin's regulatory story is, in my view, the model for how peptide medicines should be handled; rigorous preclinical work, controlled trials, approved manufacture, scheduled supply. The contrast with some corners of the peptide market is stark.

A number of peptides circulating in fitness and anti-ageing contexts are not on the ARTG and have not been approved by the TGA. Supplying unapproved therapeutic goods in Australia is illegal under the Therapeutic Goods Act 1989, and the TGA has been actively enforcing against compounders and suppliers who distribute such products without approval. This is not a grey area, the TGA's published enforcement actions make the position clear. The existence of a molecule and the existence of legal, approved access to that molecule are two very different things.

Insulin's century-long trajectory from extracted pancreatic tissue to recombinant analogue to Schedule 4 listed product is a reminder that the regulatory pathway exists because peptide chemistry is powerful, the dose margins can be narrow, and the consequences of uncontrolled supply are real.

Sources

, Priya Nandakumar, Pharmacology writer; peptides & PK

Common questions

Is insulin a peptide or a protein?
Technically, both terms are defensible. Insulin is a polypeptide — 51 amino acid residues across two chains — which places it at the upper end of what most chemists would call a peptide and the lower end of what most biologists would call a small protein. For regulatory and pharmaceutical classification purposes, it is generally categorised as a peptide hormone.
Why can't insulin be taken as a tablet?
Insulin is broken down by digestive enzymes in the gastrointestinal tract before it can be absorbed into systemic circulation. The peptide bonds that make up the molecule are hydrolysed under GI conditions, so the intact structure never reaches the bloodstream. This is why insulin requires subcutaneous injection, intravenous infusion, or specialised inhalation delivery. Oral delivery of peptide drugs is an active area of pharmaceutical research, but no standard oral insulin formulation is currently approved in Australia.
What schedule is insulin under in Australia?
Insulin formulations listed on the Australian Register of Therapeutic Goods (ARTG) are regulated as Schedule 4 (prescription-only) medicines under the Poisons Standard. They require a valid prescription from an authorised medical practitioner and are dispensed by pharmacies under the conditions that apply to all Schedule 4 medicines.
How is recombinant insulin different from animal-derived insulin?
Recombinant human insulin is produced by inserting the human insulin gene into a host organism — typically Escherichia coli or yeast — which then synthesises insulin with an identical amino acid sequence to endogenous human insulin. Earlier animal-derived insulins (bovine and porcine) differ from the human sequence by three and one amino acid residues respectively. Recombinant production also allows for greater consistency, scalability, and the ability to engineer specific amino acid substitutions to alter the molecule's pharmacokinetic profile.
What is the difference between insulin and GLP-1 receptor agonists?
Both are injectable peptide-based medicines, but they act on entirely different receptor systems. Insulin acts on the insulin receptor, a receptor tyrosine kinase, to drive glucose uptake into cells. GLP-1 receptor agonists act on the glucagon-like peptide-1 receptor, a G-protein coupled receptor, which stimulates insulin secretion (among other effects) in a glucose-dependent manner. They are distinct classes with different mechanisms, different structures, and different regulatory histories.

Related reading

About the author
PN
Priya Nandakumar
Pharmacology writer — peptides & PK · Sydney, NSW

Pharmacology is basically my love language. I write the peptide and pharmacokinetics material and try to keep the molecules from putting you to sleep. I run long, cook far too much Tamil food, and I am deep in a true-crime podcast hole.

MSc Pharmacology

More from Priya Nandakumar