The Scientist Whose Name Came Late

A certificate of correction is the most bureaucratic object the patent system produces: a single page, issued by the US Patent and Trademark Office, amending a patent that has already been granted. In 2004 several were issued against patents covering glucagon-like peptide-1, and what they corrected was a list of names. Applications filed by Massachusetts General Hospital in June 1990 had named one inventor, the endocrinologist Joel Habener. The corrections added a second — Svetlana Mojsov, the peptide chemist whose bench work had identified the form of the hormone those patents protected 5. It had taken her a dozen years of legal effort to get that page issued, and it would take nearly two decades more for the rest of the world to catch up with it.

Who discovered GLP-1 has no one-name answer, and the attempt to give it one is part of the story. In 1982 Habener's laboratory reported that the glucagon precursor of the anglerfish carries more than glucagon: a second glucagon-related sequence, arranged in tandem 1. Mojsov, who ran the hospital's peptide synthesis facility from 1983, made the candidate peptides, raised antibodies against them and built the assays that showed which form the intestine actually produces — GLP-1(7-37) 58. In 1987 she, Gordon Weir and Habener showed that this form powerfully stimulates insulin release from the rat pancreas 3. The 2024 Lasker award, shared by Habener, Mojsov and Lotte Bjerre Knudsen, finally named her in its opening line 7. This is the step before the drugs — the companion piece on this site follows Knudsen's work making the hormone last — and it runs opposite to the older hunt for the pituitary-releasing hormones: here the gene came first, and a chemist had to build what the gene predicted before anyone could test it.

From Skopje to the Merrifield laboratory

Mojsov was born in Skopje, then in Yugoslavia and now the capital of North Macedonia. She studied physical chemistry at the University of Belgrade and arrived at the Rockefeller University in New York for graduate work in 1972 5. There she joined the laboratory of Bruce Merrifield, who had invented solid-phase peptide synthesis — building a peptide one amino acid at a time on a solid support — and who would receive the Nobel Prize in Chemistry for it in 1984.

Her doctoral project was glucagon, twenty-nine amino acids long. Several strong groups in academia and industry had tried to make it by solid-phase synthesis and failed; she succeeded, and spent further years as a postdoctoral researcher refining the scheme 56. That matters for everything that followed. In the early 1980s very few people could reliably make pure peptides of the glucagon family to order. She was one of them.

A gene that said too much

The 1982 paper from Habener's group described a DNA copy of the messenger RNA for anglerfish pancreatic preproglucagon. It contained two glucagon-related coding sequences, each flanked by pairs of basic amino acids — lysine and arginine — of the kind at which enzymes cut precursor proteins into finished hormones 1. The glucagon precursor, in other words, appeared to be carrying something else.

A gene sequence, though, is a prediction rather than a hormone. It says what could be cut from a precursor. It does not say what the body actually cuts, in which tissue, or whether the product does anything at all. The predicted glucagon-like peptide ran to thirty-seven amino acids. Whether that full chain was the hormone, or something shorter, was the open question — and it could only be answered by making the candidates and then looking for them in real tissue.

The chemist's question

Mojsov moved to Massachusetts General Hospital in the spring of 1983 and that autumn became director of its peptide synthesis facility, with independent funding 5. Between the autumn of 1983 and the winter of 1984 she synthesised the peptides the precursor could release. Her hypothesis was that the active form was not the full chain but a version missing its first six residues — GLP-1(7-37) 2. By the spring of 1984 she had raised antibodies against each peptide, developed a radioimmunoassay, and devised a chromatographic method able to pick GLP-1(7-37) out of a mixture 5. The Lasker Foundation singles out that separation as the crucial step 7.

The first result came in the Journal of Biological Chemistry in 1986, with Mojsov as first author. Using antisera raised against the synthetic peptides, the team showed that the pancreas and the intestine read the same preproglucagon messenger RNA but cut the precursor differently, releasing distinct and highly specific peptides in each tissue 8. The intestine made GLP-1(7-37). A later historical analysis of the paper attributes all but one of its eleven figures to Mojsov's own bench work 5.

1987

The next question was physiological. Does the intestinal form do what an incretin should, and tell the pancreas to release insulin? Mojsov took her synthetic peptides to the laboratory of Gordon Weir, where she worked alongside his technician on isolated, perfused rat pancreases 5. The result, published in the Journal of Clinical Investigation in 1987 with Mojsov as first author, was unambiguous: in the presence of glucose, GLP-1(7-37) stimulated insulin secretion at vanishingly low concentrations, of the order of those found in the bloodstream 37.

The same spring a paper in the Proceedings of the National Academy of Sciences, with Daniel Drucker as first author and Mojsov second, reported that the shorter glucagon-like peptide raised cyclic AMP, insulin messenger RNA and insulin release in a cultured rat insulinoma cell line 4. In Copenhagen, Jens Juul Holst's group reported an insulin-releasing effect of GLP-1 that year as well, more modest in size than the Boston findings in rats 5. By the end of 1987 GLP-1(7-37) was an established incretin, and the road to a medicine was open.

Author order on those papers matters to what came later. The fullest published account of the case, by a chemist and a historian of science, argues that her second position on the Drucker paper reflected the relative standing of her laboratory and Habener's rather than the division of the work 5. That is an interpretation, published with evidence, and it should be read as one.

The patents with one name

In June 1990 the hospital filed two patent applications covering GLP-1. They listed Habener as the sole inventor, and Mojsov was not told 5. The patents were granted two years later. What followed was, by the same account, a dozen years of determined legal effort to establish her as a rightful and equal co-inventor. Certificates of correction for the first two patents, and for a related patent granted in 1997, were issued in 2004; a correction to a patent granted in 2005 followed in 2006 5.

The patents matter for a reason beyond pride. Inventorship is the legal record of who conceived an invention, and it is the one document in the whole apparatus of science whose purpose is to answer exactly that question. A name missing from it is missing from the record that institutions, licensees and later historians consult first.

The prizes that went elsewhere

Through the 2000s and 2010s, as GLP-1 drugs moved from diabetes into obesity and became some of the most commercially important medicines of the era, the discovery story settled into a standard version, and Mojsov was largely missing from it. When the 2021 Canada Gairdner International Award recognised the GLP-1 story, it went to Habener, Drucker and Holst. An article in Cell explaining the award initially contained errors that downplayed her role, and was corrected after she alerted the journal's editors 5.

The turn came through the press. Coverage from late 2022 retold her part in the discovery. In 2023 Nature named her among its ten people who mattered in science that year 5, and her own university profiled her under a headline that did the arithmetic: the breakthrough took five years, and getting credit took decades 6. The profile also quotes her on the work itself: "When you design an experiment, do it with your own hands, and actually show that it works, it's fantastic" 6. The 2024 Lasker award then named her, with Habener, for the discovery of GLP-1(7-37) 7.

How credit actually moves

It would be easy to tell this as a story with villains in it, and the documented record does not support one. It supports something more ordinary, and for that reason more worth understanding. Credit in science is not awarded once by an impartial judge. It accumulates, through a series of small routine decisions made by different people for different reasons, each of which looks reasonable on its own.

• Credit follows the head of the laboratory. Findings are reported, reviewed and remembered under the best-known name, a pattern the sociologist Robert Merton called the Matthew effect more than half a century ago.
• Author order encodes standing as well as contribution. A reader decades later sees the order, never the negotiation behind it.
• Chemists are easily filed as a service. Making a peptide to order sounds like a technical step; here the synthesis, the antibodies and the separation method were the discovery.
• Patents are drafted by institutions. Who appears on them depends on who is asked, and on who knows to ask.
• Retellings compound. Each review cites the last, and an omission copied forward becomes the consensus.

Mojsov's case is unusual mainly in its outcome. She pursued the correction for years, through the patent office and through journals, and the record changed 5. Most omissions of this kind are never contested, which is precisely why they are so hard to see.

Where the chemistry picks up

The scientific question at the centre of her work has aged well. Why is the active hormone the version lacking its first six residues? Part of the answer lies in how processing enzymes in the intestinal cells cut the precursor. The other part lies at the new front end of the chain, where dipeptidyl peptidase-4 removes the next two residues and switches the hormone off within minutes. The first cut makes the hormone; the second ends it. Every long-acting drug in the class was designed around that second cut, and the biochemistry of both is where a technical account of GLP-1 has to begin.