How Semax works: mechanism explained

Semax is a synthetic heptapeptide (seven amino acids)^[1] built from a short fragment of adrenocorticotropic hormone (ACTH) and given a chemical tail to keep it stable long enough to reach the brain. It was developed in the early 1990s at the Institute of Molecular Genetics of the Russian Academy of Sciences, and it remains a licensed prescription medicine in Russia for ischemic stroke recovery and related cognitive indications. Outside Russia, it has no marketing authorisation anywhere, including the EU, UK, or US. Understanding what it does mechanistically starts with understanding what it is built from.

Where the molecule comes from

Full-length ACTH is the pituitary hormone best known for driving cortisol release from the adrenal gland. Earlier research, beginning in the 1970s, established that a short internal stretch of the ACTH chain, roughly positions 4 through 10, carries the hormone's cognitive and behavioural effects without the adrenal-stimulating portion. The Met-Glu-His-Phe (MEHF) fragment at positions 4 through 7 is the active core. Semax takes that ACTH(4-7) tetrapeptide and attaches a Pro-Gly-Pro tripeptide tail at the C-terminus (the chain's tail end). The full sequence is Met-Glu-His-Phe-Pro-Gly-Pro, sometimes written MEHFPGP^[1]. The Pro-Gly-Pro addition slows the enzymatic breakdown that would otherwise degrade the bare ACTH fragment within minutes, giving the molecule enough stability to travel from the nasal cavity to the brain intact.

How it enters the brain

The registered route of administration in Russia is intranasal: a nasal spray that deposits the peptide on the olfactory and trigeminal nerve endings in the nasal cavity. Those nerves run directly to the brain without passing through the liver, so intranasal peptide delivery can bypass first-pass hepatic metabolism, the process that breaks down most oral drugs before they reach systemic circulation. The result is that a relatively small intranasal dose can reach central targets that an oral dose might not.

What it does inside the brain

The most consistently replicated mechanistic finding is elevation of brain-derived neurotrophic factor (BDNF), a growth and survival signalling protein for neurons. Dolotov and colleagues (2006) administered intranasal Semax at 50 and 250 micrograms per kilogram in rats and measured BDNF protein three hours later^[2]. Levels rose in the basal forebrain (the brain region involved in attention and memory) but not in the cerebellum, showing regional specificity rather than a broad diffuse effect. The study also identified specific binding sites for Semax in the basal forebrain with a dissociation constant of approximately 2.4 nanomolar, suggesting the peptide acts on a defined receptor population in that region.

Eremin and colleagues (2005) described a second mechanistic thread: modulation of dopaminergic and serotoninergic (serotonin-using) signalling^[3]. In rats, Semax at 0.15 mg/kg increased the tissue content of 5-HIAA (a serotonin metabolite) in the striatum by approximately 25% within two hours. Semax alone did not alter dopamine concentrations, but when it was given 20 minutes before D-amphetamine, it significantly amplified the stimulant's effect on striatal dopamine release and on locomotor activity. The researchers concluded that Semax exerts a modulatory, rather than directly stimulatory, effect on the dopaminergic system.

A 2025 paper in the British Journal of Pharmacology extended the picture further. Tian and colleagues used RNA sequencing and molecular docking in a female mouse spinal cord injury model to identify the mu-opioid receptor gene Oprm1 as Semax's primary molecular target in that setting^[4]. The proposed pathway runs from Semax binding Oprm1, through regulation of USP18 (a deubiquitinating enzyme), to modulation of the FTO protein involved in RNA modification. This pathway is distinct from the BDNF and monoamine findings and raises the possibility that Semax operates through several partially overlapping mechanisms depending on the tissue and the pathological context.

What it does not do

Because Semax uses only the ACTH(4-7) fragment and not the adrenal-targeting portion of full-length ACTH, it does not bind the melanocortin-2 receptor on the adrenal gland. That means no direct stimulatory effect on cortisol output. The Russian regulatory dossier frames it as a neuroprotective and nootropic (cognition-influencing) agent rather than as an endocrine drug, which fits the preclinical mechanism data. The Russian clinical literature for ischemic stroke recovery, including a 2018 study by Gusev and colleagues in 110 patients, describes gains on motor-function scales and the Barthel index of activities of daily living^[5] after the standard regimen of two 10-day intranasal courses at 6,000 micrograms per day separated by a 20-day interval. That literature is concentrated in a small number of Russian research groups and was not conducted to the multi-centre, placebo-controlled standards the FDA or EMA would require for a marketing authorisation.

The closest analogous drug class

There is no approved Western drug that works by exactly the same mechanism. The closest conceptual relatives are the neurotrophic-factor-modulating drugs being explored for neurodegenerative conditions, BDNF mimetics and tropomyosin receptor kinase B (TrkB) agonists in early clinical development, and the broader class of nootropic peptides that includes semax's Russian-licensed relatives selank (an anxiolytic ACTH-fragment derivative) and cerebrolysin (a hydrolysate of porcine brain proteins also used in Russian and some Eastern European stroke-recovery practice). None of those analogues have Western marketing authorisation either. In pharmacological terms, Semax behaves like a BDNF-modulating, monoamine-sensitising neuropeptide rather than like any drug class currently on the FDA or EMA formulary.

What we do not yet know

The mechanistic work described above is almost entirely preclinical. The BDNF-elevation and monoamine-modulation findings are in rats; the Oprm1 pathway is in female mice. There is no published dose-response curve for any of these endpoints in humans. There is no characterised adverse-event profile built to Western pharmacovigilance standards. The Russian post-marketing record since 2011 has not flagged major acute toxicity at registered intranasal doses, but long-horizon safety data in chronic use, in pregnancy, and in paediatric populations outside specific Russian neonatology protocols are absent from the Western literature. Whether the BDNF elevation seen in rodent basal forebrain translates meaningfully to human brain tissue at the doses used clinically is an open question the published data do not resolve.

For the regulatory picture, including how Semax is classified in the US, UK, EU, and Germany, see the Semax regulation section on the main peptide page.