Semax

BDNF / NGF / TrkB upregulation (flagship proposed mechanism)

A single intranasal 50 ug/kg dose in rats raised hippocampal BDNF protein ~1.4-fold and trkB phosphorylation ~1.6-fold, with parallel mRNA increases. A specific, reversible 3H-Semax binding site (KD ~2.4 nM) was reported in rat basal forebrain but has never been molecularly identified or replicated. Effect sizes are modest and, in the same lab, neurotrophin changes were region- and time-dependent (multidirectional).

Dopaminergic modulation (permissive, not direct)

In rat striatal microdialysis, Semax alone did not alter dopamine or its metabolites; it amplified amphetamine-stimulated dopamine release. No D1/D2/D3 receptor binding data exist. The amphetamine potentiation was flagged by the FDA as an unaddressed abuse-potential signal.

Serotonergic turnover

Striatal 5-HIAA (the serotonin metabolite) rose after intraperitoneal Semax in rats, consistent with increased serotonin turnover rather than increased tone. No serotonin-receptor or SERT data. Single primary study.

Melanocortin heritage: non-melanocortin in practice

The melanocortin pharmacophore is His-Phe-Arg-Trp. Semax retains His-Phe but lacks Arg8 and Trp9 entirely, so on structural grounds it cannot satisfy the minimal agonist requirement for any melanocortin receptor and cannot be corticotropic. The structural inference is strong, but no MC1R-MC5R binding or functional assay has ever been run on Semax, so this is inferred, not directly measured.

Anti-inflammatory / immunomodulation in ischemic brain (time-dependent)

In rat MCAO, more than half of Semax-altered cortical genes at 24 h were immune-response genes, with decreased Il1b, Il6, Ccl3, Cxcl2. Important caveat: at an early timepoint (4.5 h) the same lab found Semax increased Il1b and Ccl3, so the direction reverses over time and a blanket "anti-inflammatory" label is not supported as stated.

Intranasal delivery and the Pro-Gly-Pro (PGP) metabolite

Using C-terminally labeled Semax in rats, roughly 0.093% of the dose reached the brain per gram at 2 min, about 80% intact, with Pro-Gly-Pro prevailing as the metabolite. PGP is independently bioactive and reproduces much of Semax neurotrophic and anti-excitotoxic activity, so a meaningful fraction of the effect may be glyproline pharmacology rather than MEHF pharmacology. Semax may be a prodrug system rather than a simple stabilized fragment.

Cholinergic support (in vitro)

Semax increased cholinergic neuron survival and choline-acetyltransferase activity in rat basal-forebrain cultures without affecting GABAergic neurons or glia. One of the better-controlled studies and one of the few with Western co-authorship, but in vitro only.

Copper(II) chelation hypothesis

The free N-terminal Met-Glu-His motif is a plausible high-affinity Cu(II) binder (ATCUN-like), so some neuroprotection may be metal buffering rather than receptor signaling. This is the only mechanistic hypothesis generated outside Russia and is largely ignored by the Russian literature.

Semax, an analogue of ACTH(4-10), increases BDNF and trkB phosphorylation in rat hippocampus (2006)

Dolotov O.V., Karpenko E.A., Inozemtseva L.S., et al. · Brain Research 1117:54-60

Participants

Rats (single intranasal 50 ug/kg dose)

Methodology

Measured hippocampal BDNF protein and mRNA, trkB phosphorylation, and conditioned-avoidance behavior after acute intranasal Semax.

Result

Hippocampal BDNF rose ~1.4-fold and trkB phosphorylation ~1.6-fold, with improved conditioned avoidance.

Honest read

Internationally published and real, but a modest, rat-only effect. The companion nanomolar binding site has never been molecularly identified or replicated in ~20 years, and other work from the same lab reports bidirectional, region- and time-dependent neurotrophin changes. No human BDNF/NGF/Trk endpoint has ever been measured.

Semax in the treatment of acute ischemic stroke (foundational trial) (1997)

Gusev E.I., Skvortsova V.I., Miasoedov N.F., et al. · Zh Nevrol Psikhiatr Im S S Korsakova 97(6):26-34

Participants

30 treated vs 80 "conventional therapy" controls

Methodology

Intranasal Semax added to standard care; non-randomized, non-blinded, no placebo, allocation 30:80.

Result

Claimed faster recovery of neurological function.

Honest read

The 30:80 allocation is itself a non-randomization marker, and "analogous" controls are the most bias-prone design in stroke research. No effect size, confidence interval, or p-value in the abstract; two inventors are co-authors; Russian-language, abstract-only, never independently replicated.

Meta-analysis of Semax for ischemic stroke (2018)

Shmonin A.A., et al. · Vestnik vosstanovitelnoy meditsiny 17(2):81-88

Participants

3 studies pooled (n=181)

Methodology

Random-effects pooling of NIHSS change; home-made quality scale; not PubMed-indexed.

Result

Small pooled NIHSS benefit in moderate/severe (not mild) stroke, with very high heterogeneity (I2=88%).

Honest read

The authors own quality tables record that almost nothing pooled was randomized or blinded (only the unpublished 2011 booklet scored for double-blinding), comparators are "conventional therapy" rather than placebo, there is no publication-bias assessment, and the search is stale. The authors themselves conclude a proper multicenter double-blind trial is still needed and note that zero Semax trials are registered.

Synthetic ACTH analogue Semax displays nootropic-like activity in humans (1996)

Kaplan A.Ya., et al. · Neuroscience Research Communications 19(2):115-123

Participants

Healthy adults (n not reported)

Methodology

EEG changes after hyperventilation plus a claim about "operator work efficiency" at intranasal 0.25 to 1.0 mg. English-language but not MEDLINE-indexed; abstract only.

Result

Reported EEG changes resembling nootropics and a claimed long-term (20 to 24 h) benefit on operator work efficiency.

Honest read

The abstract contains no sample size, no randomization, no blinding, no placebo, no test names, no effect sizes, and no statistics, only a dose range. A co-inventor is an author. This single non-MEDLINE abstract is the entire evidentiary basis for the healthy-human nootropic claim, and it has never been replicated in ~30 years.

Effects of Semax on the default mode network (resting-state fMRI) (2018)

Lebedeva I.S., et al. · Bulletin of Experimental Biology and Medicine 165(5):653-656

Participants

24 healthy volunteers (14 Semax / 10 placebo)

Methodology

Randomized, placebo-controlled resting-state fMRI at baseline, 5, and 20 minutes.

Result

Greater volume of the rostral (medial frontal) default-mode subcomponent versus placebo.

Honest read

The best-designed human Semax study, and it measures no behavior. This is pharmacodynamic imaging with no cognitive, memory, attention, or reaction-time outcome, so it cannot support a cognitive-enhancement claim (the authors did not claim it does). The unequal 14/10 allocation is unexplained and it was not pre-registered.

Semax in motor neuron disease (the mis-tagged "RCT") (2007)

Serdiuk A.V., et al. · Zh Nevrol Psikhiatr Im S S Korsakova 107(4):29-39

Participants

n=27

Methodology

Open-label, uncontrolled, "sequential groups"; the only Semax record carrying PubMed's Randomized Controlled Trial tag, which is incorrect.

Result

Semax did not influence the course of disease; only a subjective quality-of-life subscale improved.

Honest read

Read honestly, the highest-"quality"-labeled Semax study in PubMed is a negative uncontrolled study. Citing it as an RCT supporting Semax inverts the actual result.