Semax: ACTH(4-7)PGP Heptapeptide Chemistry, Binding and Preclinical Literature

Semax is a synthetic heptapeptide with the sequence Met-Glu-His-Phe-Pro-Gly-Pro, built from the 4-7 fragment of adrenocorticotropic hormone joined to a C-terminal Pro-Gly-Pro tripeptide. The molecule carries a molecular weight of 813.9 g/mol, the formula C37H51N9O10S and CAS number 80714-61-0, and is catalogued in PubChem as CID 9811102.

The peptide originated in the Russian Academy of Sciences programme on regulatory peptides associated with Myasoedov and colleagues at the Institute of Molecular Genetics, and much of the primary literature was published in Russian-language journals such as Bioorganicheskaia khimiia and Molekuliarnaia biologiia. That publication history matters for anyone assembling a reference set, because a search restricted to English-language titles returns only part of the record. Semax holds no marketing authorization in the United States.

Structurally the molecule is simple: seven residues, a linear backbone, no cysteine and therefore no disulfide constraint, and no lipidation or PEGylation. Its distinguishing chemistry is entirely in the C-terminal glyproline, which is discussed in the next section. The consequence for laboratory work is that Semax behaves as a small, highly polar, water-tractable peptide rather than as a structured biologic.

Semax differs from the native ACTH(4-10) sequence by substitution rather than truncation alone: the C-terminal Arg-Trp-Gly of the parent fragment is replaced with Pro-Gly-Pro. The stated design intent across the literature was to keep the neuroactive N-terminal region of the hormone while shedding the corticotropic portion, producing a peptide that acts in brain tissue without driving adrenal steroidogenesis.

The Pro-Gly-Pro tail is not inert padding. Glyprolines are themselves a studied class of regulatory peptides, and Pro-Gly-Pro has been run as a standalone comparator against Semax in rodent ischemia work, where the two produced overlapping but non-identical transcriptional profiles. Proline-flanked C-termini are also a recognised structural device for slowing exopeptidase attack, which is consistent with what the biodegradation data show.

Those degradation data come from Dolotov and colleagues in 2004, working with plasma membranes of rat forebrain basal nuclei. Dipeptidylaminopeptidases were identified as the main enzymes responsible for breakdown, and the half-life of the peptide in the presence of those membranes exceeded one hour. That figure describes a membrane preparation, not circulating persistence, and the two should not be conflated when reading vendor literature.

A structurally related comparator, ACTH(6-9)PGP, recurs throughout the Russian transcriptome work as a second-generation analogue built on the same glyproline principle from a shorter hormone fragment. Studies that run both peptides side by side are the most informative part of the corpus, because they isolate what the ACTH-derived portion contributes independently of the Pro-Gly-Pro tail.

Semax binding was characterized by Dolotov and colleagues using tritium-labelled peptide against plasma membranes of rat forebrain basal nuclei. The reported dissociation constant was 2.4 plus or minus 1.0 nM with a maximum binding capacity of 33.5 plus or minus 7.9 fmol/mg protein. Binding was time-dependent, specific, reversible, and required calcium ions.

Two features of that result deserve emphasis. The low-nanomolar dissociation constant places the interaction in the range expected of a genuine receptor rather than non-specific membrane partitioning, and the calcium dependence is an unusual constraint that narrows the candidate site. What the work did not do is identify the binding partner molecularly. Reviews commonly discuss melanocortin receptor involvement for ACTH-derived peptides, but published affinity constants at cloned MC3R or MC4R for this specific heptapeptide are not part of the primary record, and the site labelled in the rat forebrain preparation remains formally unassigned.

Downstream of binding, the most quantified endpoint is neurotrophin expression. In rat hippocampus, a single intranasal application at 50 micrograms per kilogram produced a maximal 1.4-fold rise in BDNF protein, a 1.6-fold rise in trkB tyrosine phosphorylation, a 3-fold rise in exon III BDNF mRNA and a 2-fold rise in trkB mRNA. A companion experiment found BDNF protein elevated in basal forebrain three hours after intranasal application at 50 and 250 micrograms per kilogram, with no corresponding change in cerebellum, indicating regional rather than global engagement.

Monoamine measurements add a second axis. Eremin and colleagues reported that Semax raised striatal 5-HIAA by roughly 25 percent in tissue and roughly 180 percent in the extracellular compartment of rodents, and enhanced dopamine release when combined with D-amphetamine. The authors framed the result as a positive modulatory action on the striatal serotonergic system rather than as direct transmitter release.

Semax has been profiled more heavily by RNA sequencing than by classical pharmacology. The dominant paradigm is transient middle cerebral artery occlusion in rats, with differentially expressed genes counted in frontal cortex and subcortical structures at fixed intervals after occlusion. Four separate studies between 2020 and 2025 form a connected series using the same model and the same peptide.

The 2020 work identified 394 differentially expressed genes at greater than 1.5-fold change 24 hours after occlusion in animals receiving the peptide relative to saline, with inflammatory-process genes suppressed and neurotransmission genes activated. A 2023 study moved the sampling point to 4.5 hours and counted 131 differentially expressed genes for Semax against 322 for ACTH(6-9)PGP in dorsolateral frontal cortex, establishing that the two analogues diverge most in immune-related transcripts.

The 2024 study is the most numerically explicit. Ischemia alone produced 3,774 differentially expressed genes at 24 hours in penumbra-associated frontal cortex. Under Semax that count was 1,539, and under ACTH(6-9)PGP it was 2,066. For 1,171 genes tied to immune and neurosignaling pathways, both peptides reduced the expression distortion caused by ischemia, and only 32 genes separated the two peptides from each other at that time point. The authors also noted that the pattern of peptide action depends on how much time has elapsed since occlusion, which is why the 4.5-hour and 24-hour datasets cannot be read interchangeably.

A 2025 continuation extended the analysis to brain regions differing in the severity of ischemic damage, comparing striatum against frontal cortex, and reported that the peptides reduced ischemia-driven profile disturbance across close to two thousand genes in frontal cortex associated with neurotransmitter and inflammatory response. Protein-level work published in 2021 in the same model corroborated the transcript picture, describing upregulated active CREB in subcortical structures alongside downregulated MMP-9, c-Fos and active JNK in adjacent frontoparietal cortex.

Semax attracted four distinct 2025 publications, and notably none of them repeated the ischemia transcriptome paradigm. They span electrophysiology in rat brain slices, a mouse spinal cord injury model with a molecular target assignment, a transgenic amyloid mouse line, and cell-free coordination chemistry. Together they widen the mechanistic surface considerably beyond the Russian stroke corpus.

Kolbaev, Sharonova and Skrebitsky examined intracellular calcium dynamics directly. At 1 micromolar the peptide raised the frequency of spontaneous intracellular calcium fluctuations in pyramidal-layer cells of the hippocampal CA1 field, while producing no significant effect on proton-stimulated calcium entry in cerebellar granule cells. The authors read the second result as evidence that acid-sensing ion channel attenuation is not the primary neuroprotective route, which is a useful negative finding in a literature dominated by positive associations.

Liu and colleagues, publishing in the British Journal of Pharmacology, assigned a specific molecular target. Working in female C57BL/6 mice with T9-T10 spinal cord impact injury and in PC12 cells under neuroinflammatory challenge, they reported that the peptide acts through the mu opioid receptor gene Oprm1, regulating USP18 and thereby the deubiquitination of the m6A demethylase FTO. Measured endpoints included suppressed lysosomal membrane permeabilization-linked pyroptosis and reduced oxidative stress markers.

Radchenko and colleagues, writing in Acta Naturae, worked in APPswe/PS1dE9 transgenic mice using open field, novel object recognition and Barnes maze paradigms, and reported that both the peptide and a derivative reduced the number of amyloid inclusions in cortex and hippocampus alongside changes in the behavioural measures. Separately, Tomasello and colleagues characterized the molecule as a copper chelator: it extracted Cu(II) from Cu(II)-amyloid-beta species, altered the redox cycling of the complex, lowered associated reactive oxygen species production, and protected SH-SY5Y cells against oxidative challenge. That copper work follows a 2022 study in which the peptide reduced fibril formation in artificial membrane models specifically under copper-containing conditions.

Semax appears in the literature under at least four designations, and a reference search on the trade name alone will return an incomplete set. The peptide is written as Semax, as ACTH(4-7)PGP, as an ACTH(4-10) analogue, and by its one-letter sequence MEHFPGP. Russian-language primary papers frequently use the bracketed fragment notation rather than the name.

Salt form is the second issue. Catalog listings for CAS 80714-61-0 include both the free peptide and a trifluoroacetate salt, and the two differ in formula weight and in residual counter-ion content even though vendors index them against the same registry number. Documentation that records only the CAS number therefore under-specifies the material, and a certificate of analysis naming the salt is the item worth retaining.

A third distinction is the comparator peptides that appear alongside Semax in the same papers. Pro-Gly-Pro on its own and ACTH(6-9)PGP are separate chemical entities with their own transcriptional signatures, and findings reported for those molecules are regularly cited loosely as though they applied to the heptapeptide. When reading secondary summaries, checking which of the three peptides an experiment actually used resolves a large share of apparent contradictions in the record.

Semax handling figures come from chemical-catalog specifications rather than from peer-reviewed measurement, and should be read as vendor specifications. One catalog listing for CAS 80714-61-0 specifies purity of 95 percent or higher, storage of the solid at minus 20 degrees Celsius, and solubility of 1 mg/mL in dimethylformamide, 5 mg/mL in dimethyl sulfoxide and 10 mg/mL in phosphate-buffered saline at pH 7.2.

The peer-reviewed number most relevant to bench work is the biodegradation result. Against rat forebrain basal nuclei plasma membranes the peptide showed a half-life exceeding one hour, with dipeptidylaminopeptidases identified as the principal degrading enzymes. That is a membrane-preparation figure and describes enzymatic susceptibility in that system only; it is not a circulating half-life, and secondary sources frequently blur the two.

Because the sequence carries an N-terminal methionine, oxidation of the thioether sulfur is the identity-relevant degradation route to watch analytically, and it shifts molecular weight in a way reversed-phase HPLC coupled to mass spectrometry will resolve. The peptide has no cysteine, so disulfide scrambling, which complicates handling of many research peptides, does not apply here. Purity and identity are conventionally established by reversed-phase HPLC with mass-spectrometric confirmation against the 813.9 g/mol free-peptide value.

One documentation point recurs. Catalogs index both the free peptide and the trifluoroacetate salt against the same CAS number, and the two differ in measured formula weight. Recording the salt form from the certificate of analysis, rather than inferring it from the registry number, is what keeps gravimetric records internally consistent across lots.