How peptides became a drug category — from insulin to GLP-1, one hundred years of peptide pharmacology

That injection is the founding moment of peptide pharmacology. Not because it was the first time researchers had noticed that the pancreas produced something regulating glucose — that observation had been circling European laboratories for decades — but because it was the first time a peptide was purified well enough, delivered precisely enough, and administered to a human with a clear enough effect to establish that peptide molecules could be medicine. Everything that followed in the next hundred years — the entire arc from bovine insulin extracts to semaglutide to the research peptides being studied today — is the elaboration of that moment.

The first two decades after Thompson's injection were defined by the challenge of supply. Insulin was extracted from the pancreatic tissue of cattle and pigs — a process that required enormous quantities of abattoir material and produced variable, immunogenic preparations. Bovine and porcine insulin differ from human insulin by several amino acids; the immune system of long-term users eventually recognizes these differences and produces antibodies, sometimes severe allergic responses, sometimes resistance that required escalating doses. Eli Lilly and Company began commercial production in 1923, and the global market for insulin grew rapidly, but the limitations of animal extraction — both in supply and in immunogenicity — established a ceiling that would take sixty years to crack.

That crack came in 1982, with the FDA approval of Humulin, the first recombinant human insulin, developed by Genentech and licensed to Eli Lilly. Recombinant DNA technology had given scientists the ability to insert the gene for human insulin into Escherichia coli bacteria and produce the protein at industrial scale. The significance extended far beyond insulin. This was the proof of concept that mammalian proteins of therapeutic interest — molecules that the body itself was using as signals, regulators, and effectors — could be manufactured with precision, at scale, in a bioreactor. Peptide pharmacology was no longer limited by what could be extracted from animal tissue. It was limited only by what could be sequenced and expressed. The implications were not immediately obvious to the broader medical world, but the biotechnology industry understood them within a decade.

The growth hormone story runs parallel to insulin and illuminates a different chapter of peptide pharmacology's complications. By the 1950s, researchers had identified growth hormone as a pituitary peptide responsible for stimulating growth and tissue repair. Children with GH deficiency were short, and the idea of treating them with exogenous GH had obvious appeal. But there was no recombinant technology yet, and GH could only be obtained one way: from pituitary glands extracted from human cadavers at autopsy. The National Hormone and Pituitary Program collected hundreds of thousands of glands through the 1960s and 1970s, pooled them, extracted the GH, and distributed it to pediatric endocrinologists. Children grew. The program appeared to be a straightforward medical success.

In 1985, the FDA halted the program. Several recipients of cadaveric GH had developed Creutzfeldt-Jakob disease — a fatal prion disease caused by misfolded proteins that had been present in the pituitary extracts from donors who were unknowingly incubating the condition. The contamination was undetectable by any technology available at the time. Eventually, more than two hundred people who had received cadaveric GH developed CJD and died. The human pituitary extraction program ended permanently. Within the year, recombinant GH — produced through the same technology that had just been used for insulin — received FDA approval. The tragedy accelerated the transition that recombinant biotechnology was already making inevitable.

Through the 1970s and into the 1980s, the biochemical mapping of hypothalamic-pituitary signaling produced a cascade of peptide discoveries that would become the basis for major pharmaceutical classes. Roger Guillemin and Andrew Schally — who shared the Nobel Prize in Physiology or Medicine in 1977 — spent years isolating the hypothalamic releasing factors that govern pituitary hormone secretion. GHRH — growth hormone-releasing hormone — stimulates GH secretion. Somatostatin inhibits it, along with a range of other gut hormones. GnRH — gonadotropin-releasing hormone — governs the secretion of LH and FSH, which govern reproductive function. Each discovery was an engineering template: if you understood the releasing factor, you could build analogs that either mimicked or blocked it, with pharmaceutical applications that turned out to be extensive.

GnRH analogs were among the first to reach broad clinical use. The principle is counterintuitive: continuous GnRH stimulation, rather than the pulsatile stimulation of normal physiology, actually suppresses gonadotropin secretion — the pituitary's receptors downregulate in response to constant rather than episodic input. Leuprolide, a GnRH agonist approved in 1985, works precisely by delivering this continuous signal, suppressing testosterone production in hormone-sensitive prostate cancer. It is now also used in endometriosis, precocious puberty, and uterine fibroids. Somatostatin analogs — octreotide, approved in 1988 — provided a longer-acting version of somatostatin's inhibitory activity, with clinical applications in acromegaly, carcinoid syndrome, and variceal bleeding. These were the first generation of synthetic peptide drugs: molecules derived from natural signaling peptides, structurally modified for pharmacological purposes, approved through rigorous clinical trials for specific indications.

The GLP-1 story begins in the late 1980s and moves at a pace that, in retrospect, has the texture of the century's most important medical development in metabolic disease. Joel Habener's laboratory at Massachusetts General Hospital identified glucagon-like peptide 1 as a product of the glucagon gene — a related but distinct molecule that, unlike glucagon, stimulated insulin secretion and reduced glucagon release in a glucose-dependent manner. The glucose-dependence was critical: GLP-1 only stimulated insulin when blood glucose was elevated. A drug that worked this way would not cause hypoglycemia, the most dangerous side effect of existing diabetes treatments. The pharmacological problem was that native GLP-1 is degraded by the enzyme DPP-4 within minutes of secretion. The therapeutic window was too narrow to exploit.

The solution came from an unexpected source. John Eng, working at the Veterans Affairs Medical Center in the Bronx in 1992, identified a GLP-1-like peptide in the venom of the Gila monster lizard — exendin-4 — that was structurally similar enough to human GLP-1 to activate the same receptor but resistant to DPP-4 degradation, giving it a much longer half-life. Exenatide, the synthetic version of exendin-4, received FDA approval in 2005 for type 2 diabetes. Liraglutide, a human GLP-1 analog modified for once-daily dosing, followed in 2010. The clinical effects went beyond blood sugar: patients on GLP-1 agonists lost substantial body weight, a finding that had not been the primary endpoint but rapidly became the most commercially significant one. The obesity indication was developed over subsequent years, with liraglutide approved for chronic weight management in 2014.

Semaglutide — modified for once-weekly dosing — arrived in 2017 for diabetes and 2021 for obesity under the Wegovy brand. The SUSTAIN and STEP trial programs documented weight loss of fifteen to twenty percent of body weight, cardiovascular risk reduction, and in subsequent trials, reductions in heart failure hospitalization and progression of kidney disease. Tirzepatide, approved in 2022, added GIP agonism to GLP-1 agonism — a dual mechanism producing even greater weight loss, with some trials showing averages exceeding twenty percent. The pipeline includes triple agonists, amylin combinations, and oral formulations that, if approved, would transform accessibility. In one hundred years, peptide pharmacology had moved from a boy in a Toronto hospital to molecules reshaping the global epidemiology of obesity.

Running in parallel to this pharmaceutical mainstream was a different and less visible tradition. From the 1970s onward, the performance and longevity subcultures — bodybuilding, later anti-aging medicine — developed a parallel practice of using peptide compounds outside of pharmaceutical approval and clinical trials. IGF-1, GH, and their analogs circulated in bodybuilding communities in the 1980s. The development of synthetic growth hormone-releasing peptides — GHRPs — in the 1980s and 1990s by Cyril Bowers and colleagues at Tulane University produced compounds like GHRP-2 and GHRP-6 that were studied as pharmaceutical candidates before being largely abandoned by industry in favor of orally available small-molecule analogs. They never left the research and performance communities. CJC-1295, ipamorelin, and hexarelin circulated outside regulatory channels, their synthesis driven by research chemical suppliers, their use driven by practitioners working outside — or at the edges of — standard clinical practice.

The Soviet and Russian research tradition represents a separate genealogy that intersects awkwardly with Western pharmacology. Beginning in the 1960s and continuing through the 1980s and 1990s, Vladimir Khavinson and colleagues at what became the Saint Petersburg Institute of Bioregulation and Gerontology developed a class of short peptides — dipeptides and tripeptides — derived from organ extracts, with claimed effects on organ-specific function, aging-related decline, and longevity. Epithalamin from the pineal gland, thymogen from the thymus, Cortexin from cortical tissue. Some of these compounds — including the thymus-derived thymogen and the pineal peptide epitalon — have been studied in Russian and Ukrainian clinical trials for immune function, aging markers, and longevity endpoints, with some results suggesting effect sizes that Western researchers would find extraordinary and that have not been independently replicated. Whether this represents genuinely undiscovered pharmacology, artifact of research methodology, or something in between is a question that Western science has not prioritized studying.

BPC-157, a synthetic peptide derived from a sequence found in gastric juice, emerged from Croatian research in the 1990s, primarily from the laboratory of Predrag Sikiric at the University of Zagreb. The preclinical literature on BPC-157 is extensive — hundreds of animal studies in wound healing, gut protection, tendon and bone repair, and neurological contexts — and the compound has circulated widely in performance and recovery communities based on this preclinical data. As of this writing, BPC-157 has not completed a Phase II clinical trial for any indication, and the FDA has flagged compounded BPC-157 as raising regulatory concerns. The preclinical evidence is real; the clinical translation remains incomplete. This gap — extensive preclinical data, absent clinical trials — is characteristic of much of the research peptide space and reflects the economic reality that small peptides without patent protection are difficult to fund through the multi-billion-dollar clinical trial process.

The current regulatory landscape is defined by this tension. The FDA-approved peptide drugs — the insulins, the GLP-1 agonists, the GnRH analogs, the somatostatin analogs, the oxytocin preparations, the vasopressin analogs — have completed the clinical trial process and occupy the pharmaceutical mainstream. They are prescribed, reimbursed, and monitored. The compounded and research peptides — BPC-157, TB-500, Semax, Selank, epithalon, MOTS-c, and many others — exist in a space that the regulatory framework has not fully resolved. Some can be compounded by licensed pharmacies for individual patients under prescriber direction; others have been placed on FDA lists of compounds that raise safety concerns for compounding; still others circulate as research chemicals not intended for human use in any legal framing. Patients and practitioners navigate this landscape without the guidance of clinical trial data, relying on preclinical research, practitioner experience, and a community knowledge base that is real but imperfectly curated.

The trajectory from here involves, most likely, three parallel developments. First, the continued pharmaceutical development of peptide drugs for new indications — with the GLP-1 space now expanding into cardiovascular, renal, metabolic, and neurological disease, and new mechanisms emerging from the same incretin biology. Second, a gradual scientific engagement with some of the compounds that have been circulating in the research peptide space — as their mechanisms become better understood and as some practitioners accumulate enough observational data to motivate more formal study. Third, regulatory clarification that will likely restrict some compounding practices while providing more defined pathways for others.

One hundred years of peptide pharmacology teaches something that is easy to miss in the specifics: the history is a history of natural molecules, evolved over billions of years of biology, being borrowed and modified by medicine. Insulin was not invented. GLP-1 was not designed from scratch. They were discovered — identified as signals that biology was already using, extracted or synthesized, purified and characterized, and then given to people who were missing them or who could benefit from augmenting them. The pharmacological revolution in peptides is, in an important sense, a translation project: learning the language that the body has been speaking since long before medicine existed, and finding ways to speak it more precisely when the original voice has gone quiet.