The decades arc of peptide research — what's changed and what's recurred

Insulin opened a category. Not just a drug, but a proof of concept: that a molecule the body makes in nanogram quantities and uses to regulate physiology across multiple organ systems could be identified, produced outside the body, and administered as a drug. This was not obvious before 1921. After it, pharmacology was never the same.

The decades that followed built on that proof slowly, then explosively.

The 1930s and 1940s added a second category: steroid hormones. Cortisone, partially characterized by the mid-1930s and dramatically introduced to clinical medicine by Philip Hench at the Mayo Clinic in 1948, demonstrated that a second class of endocrine compounds — this time steroid-based rather than peptide-based — could transform the treatment of inflammatory disease. The early thyroid preparations predated insulin by decades — thyroid extract from animal glands had been used in hypothyroidism since the 1890s — but the era from the 1920s through the 1950s established the general principle that hormonal physiology was accessible to pharmacological intervention in ways that would reshape medicine.

Peptides specifically emerged as a distinct pharmacological class through the work of a generation of biochemists who, equipped with improving analytical chemistry, began to systematically characterize the signaling molecules the hypothalamus and pituitary used to control the rest of the endocrine system. The work was painstaking almost beyond comprehension. Andrew Schally and Roger Guillemin — who shared the 1977 Nobel Prize in Physiology or Medicine for this work — independently worked for years to identify the hypothalamic releasing factors: the peptides that the hypothalamus releases in tiny pulses to instruct the pituitary to release its own hormones. To isolate thyrotropin-releasing hormone (TRH), Guillemin's group processed roughly 300,000 sheep hypothalami — about five tons of brain tissue — and Schally's group processed a comparable quantity of pig hypothalami. The final yield was a few milligrams of purified peptide. The structure, when determined, was a three-amino-acid sequence. Three amino acids, from five tons of tissue.

This effort established the hypothalamic-pituitary axis as a cascade of peptide signals: TRH instructing TSH release for thyroid function, GnRH instructing LH and FSH for reproductive function, GHRH instructing growth hormone release, somatostatin inhibiting growth hormone release. The characterization of these releasing factors in the 1960s and 1970s created the conceptual architecture for a generation of synthetic peptide drugs. If you knew the natural peptide signal, you could synthesize it. If you could synthesize it, you could modify it. If you could modify it, you could make it more stable, more potent, longer-acting, or capable of blocking the receptor rather than activating it — an agonist or antagonist, depending on what the clinical need required.

The pharmaceutical translation of this biochemistry proceeded rapidly through the 1970s and 1980s. Leuprolide, a synthetic GnRH agonist developed by Abbott and Takeda and approved by the FDA in 1985 for prostate cancer, demonstrated the principle: a synthetic modification of the natural releasing factor, modified to produce continuous rather than pulsatile receptor stimulation, which paradoxically desensitizes the pituitary and suppresses testosterone production. The same mechanism was subsequently applied to central precocious puberty, endometriosis, and uterine fibroids. This was synthetic peptide pharmacology at its most elegant: using deep knowledge of the natural signaling architecture to produce a modified molecule that exploited the receptor system in a precisely designed direction.

The same era saw two technological shifts that changed what was possible. The first was recombinant DNA technology, which by the early 1980s allowed bacteria and later mammalian cells to produce human peptide and protein sequences at scale. Human insulin produced by engineered bacteria — approved by the FDA in 1982 — was the first recombinant DNA pharmaceutical. Recombinant human growth hormone followed in 1985. These approvals established that the industrialization of peptide and protein drug production was now fundamentally a matter of molecular biology rather than the brute-force tissue extraction that had characterized the Schally and Guillemin era. The second shift was solid-phase peptide synthesis, developed by Bruce Merrifield in the 1960s — work that earned its own Nobel Prize in 1984 — which made it possible to build synthetic peptides one amino acid at a time on a solid support, automating a process that had previously required laborious solution-phase chemistry. Together, these technologies opened the peptide drug pipeline to a scale of development that had been impossible in the tissue-extraction era.

The 1980s and early 1990s were therefore a period of substantial new approvals. Desmopressin, a vasopressin analog for diabetes insipidus and bleeding disorders. Octreotide, a somatostatin analog for acromegaly and carcinoid syndrome. Calcitonin preparations for osteoporosis. Salmon-derived calcitonin, which was oddly more potent in humans than the human version. Erythropoietin, a glycoprotein hormone for anemia in renal failure patients — not a short peptide but part of the same recombinant revolution. The commercial appetite for biological drugs was becoming evident, and the pharmaceutical industry was investing heavily in the receptor-target biology that would identify the next generation of therapeutic targets.

Running parallel to the clinical drug development story was a quieter research stream that would ultimately reshape the field: the study of gut hormones. The gastrointestinal tract produces an extraordinary range of peptide signals — gastrin, secretin, cholecystokinin, motilin, vasoactive intestinal peptide, glucagon-like peptide-1 — that regulate digestion, nutrient absorption, and, as it turned out, appetite and glucose metabolism in ways that reached far beyond the gut. The discovery of GLP-1 by Joel Habener and colleagues in the early 1980s, following from the characterization of the proglucagon gene, was initially understood as incretin physiology: GLP-1 was a gut hormone that enhanced insulin secretion in response to meals. The clinical implications were not immediately obvious. The compound was degraded almost instantly in the bloodstream by the DPP-4 enzyme, making it pharmacologically useless in its natural form. Solving the stability problem would take another two decades.

The GLP-1 story is, in retrospect, a case study in how long the interval between discovery and clinical impact can be. The biology was described in the early 1980s. The first GLP-1 receptor agonist — exenatide, derived from a peptide found in the saliva of the Gila monster lizard and found to be DPP-4-resistant — was approved for type 2 diabetes in 2005. Liraglutide, a fatty-acid modified human GLP-1 analog engineered for once-daily dosing, was approved in 2010. The recognition that GLP-1 receptor agonists produced substantial weight loss — not merely incidental to improved glycemia but as a direct effect of the receptor system's role in appetite regulation — came through this period. Semaglutide, approved for diabetes in 2017 and for obesity in 2021, completed the trajectory from a gut hormone characterized in a university biochemistry laboratory in 1983 to a drug generating tens of billions of dollars annually in sales and dramatically reshaping how medicine thinks about obesity treatment. The arc from discovery to blockbuster was approximately 38 years.

The 2020s have introduced a new technological layer that may compress future arcs significantly: computational peptide design. Machine learning tools — particularly protein structure prediction algorithms like AlphaFold and the generative design tools that followed — have made it possible to design novel peptide sequences that bind specific receptor targets with high affinity and selectivity, without requiring the prior isolation of a natural molecule doing the job. The dual GLP-1/GIP agonist tirzepatide, approved in 2022, was developed through conventional medicinal chemistry but represents the dual-agonist concept that computational screening can now explore at scale. Triple agonists engaging GLP-1, GIP, and glucagon receptors simultaneously are in late-stage clinical development. Oral peptide delivery — historically impractical because peptides are destroyed in the GI tract — is being solved through absorption enhancers, structural modifications, and novel formulation approaches; oral semaglutide (Rybelsus) was approved in 2019, and a higher-dose oral formulation is now in trials for obesity. The delivery barrier that constrained peptide pharmacology for decades is being methodically dismantled.

Throughout this long arc, several themes recur with enough consistency to be worth naming. First: the interval between biochemical discovery and clinical application is routinely measured in decades, not years. Insulin was a partial exception — the discovery-to-patient timeline was less than two years, which is essentially unprecedented and attributable to the urgency of a uniformly fatal disease and extraordinary scientific talent concentrated in one place. GLP-1 was more typical: thirty-eight years from discovery to major clinical impact. Many of the peptide systems currently being explored in research and wellness contexts have been characterized biochemically for decades and remain far from the clinical translation that would establish their utility definitively.

Second: commercial incentives shape which compounds get developed, and they shape the direction of development within a compound. Semaglutide is studied for every possible indication partly because Novo Nordisk can capture the returns on that investment. Compounds that are short, unpatentable, off-patent, or discovered in research traditions without large commercial sponsors — the Russian neuropeptides, the tissue-derived peptides, many growth hormone secretagogues — accumulate research evidence without the investment that would convert that evidence into regulated clinical products. This is not a story about bad science. It is a story about the economics of drug development, in which the bottleneck is not knowledge but capital allocation.

Third: consumer interest in peptide compounds routinely runs ahead of the evidence base. This has been true since at least the growth hormone era of the 1980s and 1990s, when recombinant GH became a sought-after performance compound well before any trial had established benefit for healthy aging adults. It has been true for every subsequent category of research peptide that has crossed from clinical research into consumer or gray-market use. The biology is often genuinely interesting. The research is often suggestive. The translation to human clinical outcomes in heterogeneous populations is often not established. The consumer market, operating outside the evidentiary framework that governs drug approval, reaches conclusions that the science has not yet reached.

Fourth: the regulatory framework is always catching up. The Dietary Supplement Health and Education Act of 1994, the Drug Quality and Security Act of 2013, the ongoing FDA rulemaking on compounding, the pending regulatory questions about specific peptides — these are all instances of a regulatory apparatus trying to keep pace with a pharmaceutical innovation environment that generates new compounds, new delivery methods, and new market categories faster than regulatory frameworks can absorb them.

What the long arc of peptide research ultimately teaches is that biology becoming therapy is a process that rewards patience and resists shortcuts. Insulin required a summer. GLP-1 required four decades. The compounds being researched today in cell culture and animal models, and being used by patients outside clinical frameworks, occupy an intermediate state in a process whose outcome is genuinely unknown. Some will become major drugs. Most will not. The ones that will aren't necessarily the most popular in current consumer channels. Predicting which findings from the research pipeline will translate to clinical impact has proven, across a century of peptide pharmacology, to be a task that consistently humbles even the most informed observers. What can be said is that the pipeline has never been richer, the tools have never been more powerful, and the category that Banting and Best opened with a summer in Toronto continues to expand in directions they could not have imagined.