For most of pharmaceutical history, peptides had a geography problem. They existed on a hidden cul-de-sac between two huge medicinal estates. Small-molecule drugs were compact, stable and easy to turn into pills. Antibodies were enormous, exquisitely targeted and increasingly valuable.
Peptides sat in the middle; chains of amino acids capable of recognizing biological targets with remarkable precision, but generally too fragile for the stomach’s chemical cauldron and too large to easily slip through cell membranes. So we injected them.
Insulin made that bargain famous more than a century ago and the history of insulin is really the first chapter of the peptide economy. Modern GLP-1 drugs have made it spectacularly profitable. Then something changed; not with the biology, but with the engineering.
The Wall Street Journal pointed past the wellness-clinic fascination with injectable “peptides” and toward a much more consequential story: companies learning how to make peptide medicines as pills. It sounds like a formulation improvement, but it’s not. It is more akin to figuring out how to land airplanes on water; bringing an entire category of new destinations within reach.
Consider psoriasis. For years, some of the most effective targeted medicines for serious inflammatory disease have been powerful antibodies that usually must be injected. Protagonist Therapeutics built a peptide-discovery platform around the idea that a much smaller molecule might deliver biologic-like precision while behaving more like a conventional drug.
That work produced icotrokinra, an oral peptide that blocks the IL-23 receptor. In March 2026, Johnson & Johnson announced FDA approval of ICOTYDEfor moderate-to-severe plaque psoriasis. A targeted peptide had crossed one of the field’s most formidable borders from injection to pill.
Two months later, the same idea appeared in a completely different disease. PCSK9 is a proven cholesterol target, but the blockbuster drugs that inhibit it have historically been injectable antibodies. Merck developed enlicitide, a macrocyclic peptide engineered into a ring-shaped structure, and in July 2026 the FDA approved LIPFENDRA, the first approved oral PCSK9 inhibitor. The important word is not cholesterol. It is oral.
Peptides are becoming a programmable middle layer between traditional drug classes. Small molecules can be wonderfully convenient, but many biological surfaces are simply too broad or complicated to grab effectively. Antibodies can bind those surfaces beautifully, but they are large, expensive to manufacture and generally can’t be swallowed.
Macrocyclic and engineered peptides merge these worlds: enough surface area and structural complexity to engage difficult targets, yet enough chemical tunability to improve stability, permeability and, in selected cases, oral delivery. Nature Chemical Biology has already demonstrated cyclic-peptide discovery workflows that simultaneously screen for target activity and permeability, exactly the combination that used to make oral peptides feel contradictory.
And that changes where the IP lives. A conventional view says the asset is the molecule. Patent the sequence, the composition, or the therapeutic use. But the emerging peptide stack is much deeper. The valuable IP can also sit in the screening library that found the molecule, the chemical modifications that keep it from being chewed apart by enzymes, the macrocyclization strategy that locks it into the right shape, the formulation that helps it cross a membrane, the process that manufactures it at scale, and the data that teaches the next design which molecular compromises actually work.
Merck offers a particularly good example. Getting a complex macrocyclic peptide to work in a tablet is only half the problem. You also have to make millions of doses. In May, Merck scientists described a large-scale biocatalytic synthesis process for enlicitide using a tailored suite of enzymes. That is a reminder that the defensible asset may not be one patent sitting alone on the shelf. It may be an innovation catalog; molecule claims, process claims, platform know-how, analytical methods, manufacturing trade secrets and regulatory data all protecting the same commercial outcome.
The financial market has noticed. In 2024, Novartis agreed to a peptide-drug-conjugate collaboration with PeptiDream that carried up to $2.89 billion in potential value. In 2025, AstraZeneca signed a macrocyclic-peptide discovery agreement with Syneron Bio valued at up to $3.475 billion. Those numbers are not proof that every peptide platform will succeed. They are evidence that pharmaceutical giants can see peptide discovery itself, (the engine, not just the output), as licensable IP.
Artificial intelligence adds another layer. Peptide chemistry has always involved a brutal combinatorial problem - change one amino acid, close the ring differently, hide one hydrogen bond, add a non-natural residue, and the molecule may bind better while becoming impossible to absorb. Or, it may absorb beautifully and stop binding. Recent peptide research increasingly combines chemical synthesis with computational modeling and AI-guided design. Nature Reviews Chemistry described 2025 as a convergence of peptide chemistry and computational modeling. The molecule is becoming partially computational before it ever becomes physical.
That creates a fascinating ownership question. If the competitive advantage comes from a model trained on thousands or even millions of failed peptide designs, what is the real asset? The successful molecule? The training data? The screening protocol? The negative results nobody publishes? The algorithm that predicts permeability? The manufacturing recipe that makes the winning structure economical? The answer, increasingly, is ‘all of the above’.
There is also a caution hiding inside. “Peptides” now describes two very different markets. One is a regulated pharmaceutical race producing approved medicines and carefully characterized manufacturing processes. The other includes wellness and online sellers offering unapproved injectable compounds. In August 2026, FDA warning letters targeted sellers of products including semaglutide, tirzepatide, retatrutide and other peptides that the agency said were unapproved new drugs. The scientific opportunity is real, but so is the regulatory boundary.
Ironically, regulation may make the IP story even more interesting. In July 2026 the FDA issued 17 revised draft product-specific guidances for generic peptide products. As peptide medicines mature, innovators will face the same question every successful technology eventually faces; what remains proprietary when the original molecule gets competition? Expect the answer to move outward toward delivery, combinations, next-generation sequences, manufacturing efficiency, platform technology and the accumulated know-how required to make difficult peptides behave like easy drugs.
For a century, peptides were valuable despite their limitations. Now those limitations are becoming engineering problems. And engineering problems create IP. The next pharma land grab may not be over a single miracle molecule. It may be over the rules for designing an entire class of molecules that can reach antibody targets and that patients can swallow.
