The Moscow Institute That Made Peptides for the Mind

On 24 July 2026, the second day of a two-day meeting, the US Food and Drug Administration's Pharmacy Compounding Advisory Committee turned to a peptide seven amino acids long. Its sponsors had nominated it for cerebral ischaemia, migraine and trigeminal neuralgia, and the committee's task was to advise whether it should join the list of bulk substances American pharmacies may compound 6. The committee voted to recommend it — a recommendation, not a decision, and not binding on the agency 7. The peptide was Semax, and it had been designed more than forty years earlier at the Institute of Molecular Genetics in Moscow.

Semax and its sibling Selank came out of that institute in the Soviet 1980s, from a programme associated with the chemist Nikolai Myasoedov and the physiologist Igor Ashmarin. The idea behind them was imported: that fragments of pituitary hormones can act on the brain without acting as hormones. The fix that made the idea usable was their own — a three-amino-acid tail that protected short, fragile peptides from the enzymes that destroyed them in minutes. The parallel with the Zagreb story told elsewhere on this site is real: another literature with one centre, built largely by one research culture around one unifying theory. But the Moscow story turns on something Zagreb never had — its compounds became registered medicines at home, and the science of why they worked came afterwards. Why that literature is so hard to read from outside is told in a separate piece; this one is about how the molecules were made.

An idea from Utrecht

The story starts outside Russia. At the Rudolf Magnus Institute in Utrecht, the Dutch pharmacologist David de Wied spent the 1960s and 1970s showing that hormones of the pituitary and hypothalamus do more than run the endocrine system. Pieces of them — fragments of adrenocorticotropic hormone, of vasopressin, of related peptides — could affect motivation, attention, learning and memory in animals while lacking the hormone's classical endocrine effects. In a 1977 review he framed the parent hormones as precursors of smaller, behaviourally active fragments 1.

One fragment above all carried that argument: the stretch of adrenocorticotropic hormone running from residue 4 to residue 10. It influenced avoidance learning in rats without stimulating the adrenal glands. It was, in principle, a way of acting on the brain with a piece of a hormone rather than the whole. It also had a flaw that made it close to useless as a medicine. A short, unprotected peptide is a meal for the peptidases of blood and tissue, and fragments of this kind lasted only minutes.

Two academicians and an institute

The Institute of Molecular Genetics of the Academy of Sciences was founded in Moscow at the start of 1978, and it would later be folded into the Kurchatov Institute's national research centre, in 2020. Its peptide work is tied to two men from different disciplines.

Igor Ashmarin, born in Leningrad in 1925, was a biochemist and physiologist who from 1986 headed the Department of Human and Animal Physiology at Moscow State University and built there a school devoted to the methodical study of regulatory peptides. His work on analogues of the adrenocorticotropin fragment and of tuftsin led, with colleagues, to Semax and Selank 8. His theoretical contribution was an insistence that regulatory peptides form an interlocking network — he called it a functional continuum — rather than a set of separate signals. He died in Moscow in 2007.

Nikolai Myasoedov, a bioorganic chemist born in Kursk in 1936, supplied the other half: the chemistry of making physiologically active substances, and the institutional base. He became a full member of the Russian Academy of Sciences in 2003, served as a deputy director of the Institute of Molecular Genetics, and in 2021 the Academy marked his eighty-fifth birthday 9. His name runs through the institute's later peptide papers, from studies of Semax in the rat hippocampus to the genome-wide work of the 2010s 35.

Three amino acids as a shield

The Moscow answer to de Wied's flaw was a tail. Take a short active fragment, and add to its end the tripeptide proline–glycine–proline, a sequence peptidases handle poorly. Semax is the result of applying that to the adrenocorticotropin fragment: methionine, glutamic acid, histidine, phenylalanine, then proline, glycine, proline — seven residues in all 5. The developers' own retrospective, published in 1997, counted fifteen years of design and study behind it, which places the start of the work in the early 1980s 2.

Selank shows that the tail was not a one-off but a method. The institute took tuftsin — a four-residue peptide cut from the heavy chain of immunoglobulin G, with its own role in the immune system — and fitted the same proline–glycine–proline end to it, to make a stable analogue 4. One stabilising module, two unrelated messages. Where Semax pointed at the brain through a hormone fragment, Selank pointed at it through an immune peptide, and it went on to clinical use in Russia as an anxiolytic.

Medicines first, mechanisms after

Here the Moscow story departs from the sequence a Western reader expects. In the model that governs drug development in the United States and Europe, a mechanism is proposed, tested in cells and animals, and only then taken into the controlled human trials a regulator will accept. Semax and Selank reached clinical use in Russia on the evidence of the Soviet and Russian system, and a large part of the molecular work on how they act was done afterwards, in the 2000s and 2010s.

The institute's own papers say so with some candour. A 2014 genome-wide study of Semax in a rat model of focal brain ischaemia opens by noting that the peptide had proved efficient in the therapy of stroke, while the molecular mechanisms underlying its action remained obscure 5. A 2006 study reported that Semax raised levels of brain-derived neurotrophic factor and activated its receptor in the rat hippocampus 3. For Selank, the same era produced studies of how it shifted the expression of inflammation-related genes in the mouse spleen 4. These are rodent and cell findings, and they should be read as such. What they show about the programme is its order of work: a registered medicine first, and then a long effort to explain it.

The parallel with Zagreb, and where it breaks

Set this beside the Zagreb programme behind BPC-157 and the likeness is plain. In both, a single research culture produced most of what is known. In both, one broad theory tied the findings together — vascular repair and nitric oxide in Zagreb, a network of regulatory peptides in Moscow. In both, independent replication from outside is thin relative to the size of the literature, and the standing question is the same: what is evidence worth when nearly all of it comes from one place?

The differences matter more. Zagreb followed one compound into many tissues; Moscow followed one design principle into several compounds. Zagreb's peptide never became a registered medicine anywhere, so its literature stayed preclinical; Moscow's went through a national regulator and into clinical use, so its literature includes clinical work, much of it in Russian and designed to domestic conventions. And Zagreb's absence of trials is a story about who would pay. Moscow's is a story about which regulator has looked.

A committee reads the file

Which brings the story back to July 2026. The Pharmacy Compounding Advisory Committee met to consider seven peptides that had never been approved by the agency, Semax among them 6. It recommended six, Semax included, and rejected one 7. The recommendation goes to the Department of Health and Human Services, which must act before pharmacies can compound from the substance; the committee's vote binds nobody 7. It is worth being exact about what inclusion would mean. A place on that list would make Semax a substance a pharmacy could compound on prescription. It would not make it an approved medicine, and it would not mean the agency had judged it effective.

Still, something happened that had not happened before. An American body with regulatory standing sat down, in public, with a file built largely in Moscow over forty years, and argued about it. However the agency responds, the literature with one centre acquired, for two days, a second set of readers.

Where the chemistry takes over

What the story leaves are technical questions, and they are good ones. How long does a peptide with a proline–glycine–proline tail actually survive compared with its parent, and in which tissue? Does the tail merely protect the message, or — as some of the institute's own work suggests — carry activity of its own? And how would a seven-residue peptide delivered through the nose reach the brain in amounts that matter? Those belong to the pharmacology of these molecules, and that is where a technical article on Semax and Selank has to begin.