Semax Peptide: How It Works in the Brain

Semax is a synthetic seven-amino-acid peptide — Met-Glu-His-Phe-Pro-Gly-Pro — designed in the 1980s at the Institute of Molecular Genetics of the Russian Academy of Sciences by taking a four-residue fragment of adrenocorticotropic hormone (ACTH) and attaching a Pro-Gly-Pro tail to it. Search "how does Semax work" and you will find a great many confident pages about BDNF, neuroprotection and gene expression. Almost none of them attach the two facts that decide what any of those findings mean: which species the experiment ran in, and by which route the peptide was given. This page keeps both attached to every claim, because with them attached the story changes shape.

How does the Semax peptide work at the molecular level?

Semax works primarily by raising brain-derived neurotrophic factor (BDNF) and its receptor trkB in specific brain regions — a finding established in rats, by the intranasal route, and measured in hours rather than weeks.

The structure explains a lot of it. Native ACTH fragments break down quickly. Swapping ACTH residues 8–10 (Arg-Trp-Gly) for a Pro-Gly-Pro tripeptide gives a molecule that resists biodegradation; in animal models that single change extended the duration of effect to roughly 20–24 hours. The same substitution removes the corticotropic activity of the parent hormone, which is why Semax is described throughout the literature as a non-corticotropic ACTH analogue rather than a hormone. (A widely copied claim that Semax was engineered to avoid glucocorticoid receptors is simply a mechanistic error — ACTH acts on melanocortin receptors on the adrenal cortex, not on glucocorticoid receptors.)

The neurotrophic evidence itself is precise. In Dolotov's 2006 study in Brain Research, a single intranasal application of 50 µg/kg in rats produced a maximal 1.4-fold rise in hippocampal 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 (PMID 16996037). The same animals made more conditioned avoidance responses, which is where the "Semax improves learning" line originates — in rats, on a shock-avoidance task.

A companion paper the same year found what a receptor pharmacologist would want next: tritium-labelled Semax binds rat basal-forebrain cell membranes in a time-dependent, specific and reversible way, with a dissociation constant of 2.4 ± 1.0 nM and a maximum binding density of 33.5 ± 7.9 fmol/mg protein, and the binding requires calcium ions (PMID 16635254). Intranasal Semax at 50 and 250 µg/kg raised BDNF in the rat basal forebrain at 3 hours — but not in the cerebellum, and not at 24 hours. Two details usually lost in retelling: the effect is regionally selective, and it is short-lived. A third is that the BDNF appears to come from astrocytes cultured from rat basal forebrain, making the origin glial rather than neuronal.

BDNF is also not the only neurotrophin in the picture, and the time course is stranger than any single figure suggests. Intranasal Semax at 50 µg/kg in rats raises the mRNA for both nerve growth factor (NGF) and BDNF in the hippocampus and the frontal cortex — but biphasically, and in opposite directions to begin with. At 20 minutes expression falls in the hippocampus while it rises in the frontal cortex; both are back to baseline by 40 minutes; both are significantly elevated again by 90 minutes. A mechanism that runs regionally opposite in its first minutes, subsides, and then returns is a more accurate answer to the question than "it raises BDNF" — and it is the part almost every summary drops.

What the gene-expression studies found — and why they disagree

Semax changes the expression of hundreds of genes in the rat brain, but the direction of the change depends on the injury model — which is exactly the nuance most summaries collapse.

Two rat stroke transcriptomes from the same research group point opposite ways on immune and inflammatory genes:

Three conditions, three directions. Vendor summaries routinely merge the first two rows into one sentence and attribute the immunoglobulin/chemokine result to the wrong paper. The defensible reading is narrower and more useful: in rat stroke models, Semax shifts transcription toward the pattern the injury disrupted, and what "toward" means depends on the injury.

Three newer mechanisms, and what each showed

Since 2022 the mechanistic literature has widened beyond BDNF — into calcium signalling, ubiquitination and copper chemistry. All three are preclinical.

Calcium. In rat hippocampal slices, 1 µM Semax significantly increased the frequency of spontaneous intracellular calcium fluctuations in pyramidal-layer cells of the CA1 field (PMID 41171324). The same paper is usually quoted only that far. Its other arm found no significant effect on proton-induced calcium influx in cerebellar granule cells, and the authors concluded that Semax's neuroprotective mechanism appears unrelated to calcium entry through acid-sensing ion channels. Reporting only the positive half inverts the paper's own conclusion.

Ubiquitination and the µ-opioid receptor gene. In female C57BL/6 mice given a spinal-cord impact injury at T9–T10, plus a PC12 cell model of neuroinflammation, Semax modulated lysosomal membrane permeabilisation and ubiquitination, and acted on the µ-opioid receptor gene Oprm1 to promote deubiquitination and functional recovery (PMID 40692165). This is a mouse and cell-line finding published in 2025; no human spinal-cord-injury data exists.

Copper and amyloid. In artificial phospholipid membrane models — no animals, no people — Semax formed a stable complex with Cu²⁺, prevented formation of amyloid-beta:Cu²⁺ complexes, and interfered with fibrillogenesis (PMID 35080861). A separate 2025 paper found Semax strips Cu(II) from Cu(II)-amyloid species, redox-silences the complex and was cytoprotective for SH-SY5Y cells. Interesting chemistry; it is not evidence about anyone's brain.

The monoamine picture deserves one correction of its own. In rodents, Semax raised striatal 5-HIAA (a serotonin metabolite) by 25% at 2 hours, with extracellular 5-HIAA climbing to 180% within 1–4 hours after 0.15 mg/kg intraperitoneally — but Semax alone did not change dopamine. It only amplified dopamine release when given 20 minutes before D-amphetamine (PMID 16362768). Semax is a serotonergic modulator that can potentiate a stimulant, not a dopamine-releasing agent.

So how does Semax work in people?

The human record is small, almost entirely Russian-language, and contains no randomised placebo-controlled trial with a clinical endpoint. Twelve studies could be located, roughly 500 participants in total. Here are the four with published numbers:

Two things follow. First, the largest study is a non-randomised 110-patient stroke cohort in which BDNF correlated with functional recovery — a correlation in a stroke population, not a demonstration that a healthy person's cognition improves. Second, the two placebo-controlled studies are imaging pilots in healthy volunteers; their endpoints are fMRI connectivity, not memory, mood or work performance. The Alzheimer's Drug Discovery Foundation's independent review of Semax reaches the same place: on clinical research supporting a dementia or cognitive-aging benefit, its verdict is the single word "None", and it records no Semax research underway in the United States.

You will also see a human dose range of 250–1,000 µg/kg quoted for attention and short-term memory. That figure traces back through a review to a Russian-language pilot in 11 shift workers that the ADDF itself reported being unable to access — and the 1997 Russian review of the programme gives the human intranasal range as 0.015–0.050 mg/kg, five to twenty times lower. Treat the higher figure as unverified.

Route matters more than almost anyone says

Nearly all of the human evidence for Semax comes from intranasal administration, and the registered Russian product is intranasal drops only — so evidence gathered by that route should not be quietly transferred to any other.

Vendor pages circulate a confident-sounding "subcutaneous research protocol" for Semax. No published human study of subcutaneous Semax could be located, and no animal one either; the widely repeated 6,000 µg/day figure attached to it is in fact the intranasal regimen from the 110-patient stroke study above, and even there it was two courses, not one.

Route is not a delivery detail. A rodent head-to-head found Semax more potent for learning improvement intranasally than intraperitoneally, while analgesic activity appeared only after intraperitoneal dosing. The route changed which effects appeared, not just how strong they were.

What the safety record shows

Semax has a reassuring reputation and a thin evidence base, and both statements are true at once.

The most frequently reported adverse events in human use, per the Kolomin 2013 review summarised in the ADDF's Cognitive Vitality report, are discoloration of the nasal cavity in about 10% of patients and increased blood glucose in patients with diabetes in about 7.4%. The second of those is prescribing-relevant and routinely omitted from consumer write-ups. Beyond that: no consistent reports of severe toxicity, organ damage or dependence appear in the published literature, and the 36-patient glaucoma study above recorded no adverse effects over one month.

Three limits belong next to that:

1. The "15-year safety study" does not exist as described. The source everyone cites (PMID 9173745) is a 1997 Russian-language narrative review of a 15-year research programme, whose "in no case produced negative side actions" line is an author's summary statement in a review abstract — not a controlled safety dataset, and now 29 years old.
2. Long-term side effects are unknown. No human dataset extends meaningfully beyond about a month of continuous use.
3. There is no safety data in pregnancy, in lactation, or in anyone under 18.

Semax and Selank: the same trick, two starting molecules

Semax and Selank are structurally parallel: each takes a short fragment of a larger human protein and stabilises it with the same C-terminal Pro-Gly-Pro tail. That shared tail is the actual family resemblance, and it is the part most comparisons miss.

The only direct human head-to-head is the 52-participant resting-state fMRI study above, which compared Semax, Selank and placebo — by injection, with an imaging endpoint. It tells you the two molecules do measurably different things to brain connectivity within 20 minutes. It does not rank them for any outcome a person would notice.

Where Semax stands with regulators

Semax is not FDA-approved for any indication, in any form — there is no NDA, no ANDA and no approved US labelling.

It is, however, a genuine registered pharmaceutical in Russia, approved by the Russian Ministry of Health in the 1990s (1994 is the year usually cited) and developed at the Institute of Molecular Genetics, Russian Academy of Sciences. The registered dosage forms are intranasal drops at 0.1% and 1%: the 0.1% strength for cognitive disorders, asthenia and periods of high cognitive load; the 1% strength for acute-phase ischaemic stroke and optic-nerve atrophy. Outside Russia and the CIS it is approved nowhere, including by the EMA.

On 24 July 2026, the FDA's Pharmacy Compounding Advisory Committee reviewed semax for the 503A bulk drug substances list and voted 8 yes – 5 no – 1 abstention to recommend it, against FDA staff's own recommendation. At the same meeting the committee evaluated six other peptides. In its briefing materials FDA noted that the nominated uses it assessed were cerebral ischemia, migraine and trigeminal neuralgia; that it did not evaluate "ADHD" because it could find no supporting literature; and that it did not evaluate "nootropic" because the term has no ICD-10 code and no professional-society treatment guideline. FDA also stated that a foreign regulatory approval issued by an authority not operating under FDA's standards for clinical evidence, manufacturing oversight and product quality does not itself constitute an adequate demonstration of safety and efficacy for this purpose.

Across all seven peptides reviewed at that meeting, FDA's objections ran along three lines: that the substances are not well characterised; that there is little or no human evidence of effectiveness for the proposed routes, which are mostly injectable; and that human safety data is insufficient, including unassessed immunogenicity risk. The middle objection is the same observation this article arrives at independently from the literature — whatever human record Semax has was built intranasally, and it does not transfer to a needle by assumption.

Advisory-committee votes are non-binding, and the next step is notice-and-comment rulemaking. That is the status as of this writing — and none of it changes how Semax reaches a patient in the United States today: it remains available as a compounded medication that a licensed provider can prescribe. Whether it is appropriate for you is a decision made between you and your doctor, not by an advisory vote.

One note for anyone subject to drug testing: Semax's status under anti-doping rules is genuinely contested, and it is not named on the WADA Prohibited List. If you compete under a testing programme, assume risk and check with your federation before anything else.

What this means if you are considering Semax

Semax offered through a telehealth service in the United States is a compounded medication — prepared by a licensed pharmacy for an individual patient rather than manufactured to an approved label. It is not FDA-approved, and unlike compounded semaglutide there is no US counterpart carrying the same active ingredient to measure it against. That distinction is not a technicality; it is the whole regulatory frame around it.

At Promise, a licensed provider reviews every request, and not everyone qualifies — a provider can decline on medical eligibility, and does. The questions worth bringing to that visit are the ones this evidence base raises: whether your goal matches anything that has actually been measured, whether you have diabetes or another reason the blood-glucose signal matters, and what the plan is for reassessing after a defined period rather than continuing indefinitely into territory no dataset covers. The Semax / Selank page shows what is currently available and what a visit involves.

Questions worth bringing to the visit

Semax offered here is dispensed as a compounded preparation, and it is not FDA-approved. That is precisely why the decision is not yours to make alone: a licensed provider reads your history against the request and can decline it.

If this page has left you with sharper questions rather than a conclusion, that is the right condition to start a visit in — bring them, including the ones about what has never been measured.

References

1. Dolotov OV, et al. Semax, an analog of ACTH(4-10) with cognitive effects, regulates BDNF and trkB expression in the rat hippocampus. Brain Research. 2006;1117(1):54-60. DOI 10.1016/j.brainres.2006.07.108. Read the study
2. Dolotov OV, et al. Semax, an analogue of adrenocorticotropin (4-10), binds specifically and increases levels of brain-derived neurotrophic factor protein in rat basal forebrain. Journal of Neurochemistry. 2006;97 Suppl 1:82-86. DOI 10.1111/j.1471-4159.2006.03658.x. Read the study
3. Medvedeva EV, et al. The peptide semax affects the expression of genes related to the immune and vascular systems in rat brain focal ischemia: genome-wide transcriptional analysis. BMC Genomics. 2014;15:228. DOI 10.1186/1471-2164-15-228. Read the study
4. Filippenkov IB, et al. Novel insights into the protective properties of ACTH(4-7)PGP (Semax) peptide at the transcriptome level following cerebral ischaemia–reperfusion in rats. Genes. 2020;11(6):681. DOI 10.3390/genes11060681. Read the study
5. Biochemistry (Moscow). 2024;89:1643-1656 — transcriptome effects of Semax and ACTH(6-9)PGP in normal, non-ischemic rat frontal cortex.
6. Kolbaev SN, Sharonova IN, Skrebitsky VG. The effect of peptide Semax, an ACTH(4-10) analogue, on intracellular calcium dynamics in rat brain neurons. Bulletin of Experimental Biology and Medicine. 2025;179(4):416-420. Read the study
7. Liu R, Chen Y, Huang H, et al. Semax peptide targets the μ opioid receptor gene Oprm1 to promote deubiquitination and functional recovery after spinal cord injury in female mice. British Journal of Pharmacology. 2025;182(22):5489-5516. Read the study
8. Sciacca MFM, Naletova I, Giuffrida ML, Attanasio F. Semax, a synthetic regulatory peptide, affects copper-induced Abeta aggregation and amyloid formation in artificial membrane models. ACS Chemical Neuroscience. 2022;13(4):486-496. DOI 10.1021/acschemneuro.1c00707. Read the study
9. Bioinorganic Chemistry and Applications. 2025:4226220 — Semax strips Cu(II) from Cu(II)-amyloid species and reduces copper-catalysed oxidative stress in SH-SY5Y cells.
10. Radchenko AI, Kuzubova EV, Apostol AA, et al. The potential of the peptide drug Semax and its derivative for correcting pathological impairments in the animal model of Alzheimer's disease. Acta Naturae. 2025;17(4):110-120. DOI 10.32607/actanaturae.27808. Read the study
11. Eremin, et al. Semax, an ACTH(4-10) analogue with nootropic properties, activates dopaminergic and serotoninergic brain systems in rodents. Neurochemical Research. 2005;30:1493-1500. Read the study
12. Gusev, et al. [The efficacy of semax in the treatment of patients at different stages of ischemic stroke]. Zhurnal Nevrologii i Psikhiatrii imeni S.S. Korsakova. 2018;118:61-68. Read the study
13. Functional connectomic approach to studying Selank and Semax effects. Doklady Biological Sciences. 2020;490:9-11. Read the study
14. Lebedeva, et al. 2018 — resting-state fMRI pilot of intranasal Semax in 24 healthy volunteers. Read the study
15. Ashmarin IP, Nezavibat'ko VN, Miasoedov NF, et al. [A nootropic adrenocorticotropin analog 4-10-semax (15 years experience in its design and study)]. Zhurnal Vysshei Nervnoi Deiatelnosti imeni I.P. Pavlova. 1997;47(2):420-430. Read the study
16. Strakhov VV, Popova AA, Fedorov VN. The results of Semax neuroprotective efficacy investigation. Ophthalmology Reports. 2014;7(4).
17. Tait J, Karkoutly MY, Rakay D, Aquino E. Semax for Parkinson's neuroprotection: a qualitative review of preclinical evidence. UTRGV School of Medicine Research Colloquium, 2025 (poster). ScholarWorks @ UTRGV.
18. Alzheimer's Drug Discovery Foundation. Cognitive Vitality report: Semax.

Three of these are weaker than their reputation, and the article labels them as such where it uses them. Reference 15 is a Russian-language narrative review written by the group that originated Semax — a programme retrospective, not a longitudinal safety study. Reference 17 is an unrefereed poster by medical students. Reference 16 is not indexed in PubMed. They appear here because the article discusses them, not because they carry weight.

Important information

This article is for general education and is not medical advice, diagnosis or treatment. It does not replace a conversation with a licensed clinician who knows your history. Compounded medications discussed here are not FDA-approved, and nothing on this page is a claim that any product prevents, treats or cures any disease.