Semax vs Selank: Two Russian Heptapeptides With the Same Tail, Different Parents, and Very Different Amounts of Accessible Evidence

Semax vs Selank: sequences, research, forms and differences

Semax and Selank are discussed together because they share a tail (Pro-Gly-Pro) and a country of origin, but they come from different parent molecules and have been studied for different things. Semax is an ACTH(4-7) fragment investigated for BDNF expression, attention and ischemic stroke; Selank is a tuftsin analog investigated for anxiety and GABAergic signaling. This comparison sets the sequences side by side, then works through what the accessible evidence supports for each. PrymaLab supplies Semax 5mg, Semax 11mg, Selank 5mg and Selank 11mg vials, a Semax 10mg Nasal Spray and a Selank 10mg Nasal Spray, and preloaded pens for research use only.

Michael Phelps

Founder & Peptide Research Specialist, PrymaLab

Research reference · Last updated August 19, 2026 · ~18 min read

The sequences, side by side

Start here, because almost every comparison of these two skips the structures and goes straight to described effects, which makes the relationship impossible to see.

Semax is Met-Glu-His-Phe-Pro-Gly-Pro, written MEHFPGP.^[1] The first four residues correspond to ACTH(4-10), a fragment of adrenocorticotropic hormone.

Selank is Thr-Lys-Pro-Arg-Pro-Gly-Pro. The first four residues, TKPR, are tuftsin, a tetrapeptide that comes from the heavy chain of immunoglobulin G.

So the comparison is not between two versions of one idea. It is between an ACTH derivative and an immunoglobulin derivative that happen to have been finished the same way.

The Pro-Gly-Pro tail, and why both have one

This is the part I find genuinely interesting, and it took me a while to notice that it was the same three residues on both.

Short peptides are cleared fast. Aminopeptidases and carboxypeptidases chew inward from both ends, and a four-residue fragment in circulation has a very short life. Any attempt to turn a short bioactive fragment into something usable has to solve that problem first.

The solution used on both of these is a Pro-Gly-Pro extension on the C-terminus.^[1] Proline is unusual among amino acids because its side chain loops back to the backbone nitrogen, which constrains the local geometry and makes proline-containing bonds poor substrates for many peptidases. Sticking prolines at the exposed end obstructs the enzymes that would otherwise start there.

The same modification appears on both compounds because both came out of the same Russian design programme, and it is a recognisable signature of that work rather than a coincidence.

There is a wrinkle that makes it more than a stabilising trick. Pro-Gly-Pro has been reported to carry pharmacological activity of its own, independent of whatever it is attached to.^[1] That complicates interpretation of any result from either compound. If the tail is not inert, then a study of Semax is a study of an ACTH fragment plus PGP, and part of what you measure could belong to the tail.

Semax, and the BDNF numbers

Semax research centres on neurotrophic signalling, neural plasticity, neuroplasticity, and brain health, specifically focusing on brain-derived neurotrophic factor.

The figure that gets quoted comes from rodent work: a single administration produced roughly a 1.4-fold increase in BDNF protein and a 3-fold increase in BDNF mRNA.^[1]

Those two numbers are worth reading against each other rather than separately, and almost nobody does. Transcription more than tripled while protein went up by 40 percent. That gap says the response is limited somewhere downstream of transcription, at translation or at protein turnover, and it means quoting the mRNA figure alone would overstate what happened by roughly a factor of two.

I would treat the protein number as the meaningful one and the mRNA number as a mechanistic observation about where the compound acts. A source that leads with "3-fold increase in BDNF" without saying which measurement is doing something misleading.

Semax has also been reported to affect nerve growth factor (NGF) and other neurotrophins, demonstrating neuroprotective potential against oxidative stress, and the applied research in Russia has concentrated on acute neurological injury, particularly ischemic stroke recovery, and on optic nerve conditions.^[1]

PrymaLab supplies Semax as 5mg and 11mg reference material, and in nasal spray and autoinjector formats, all for laboratory research use only.

Neurotrophin genes in cultured glial cells

The BDNF work on Semax began in cell culture. Shadrina and colleagues had observed Semax affecting neuronal cell survival in vitro and proposed that neurotrophic factors such as NGF and BDNF, which govern the proliferation, differentiation and survival of neurons and glia, carried the effect (Shadrina et al. 2001). To test the idea they cultured glial cells from the basal forebrain of newborn rats, added Semax and measured NGF and BDNF gene expression over time (Shadrina et al. 2001).

Both genes responded within half an hour. The largest change came 30 minutes after Semax was added, when BDNF mRNA stood at eight times the control level and NGF mRNA at five times (Shadrina et al. 2001). The 2006 rat study discussed above, which shares an author with this one, found a threefold rise in exon III BDNF mRNA but a 1.4-fold rise in BDNF protein in the hippocampus (Dolotov et al. 2006). Read together, the two papers describe a large, fast transcriptional response in isolated glia and a smaller change in protein in the intact brain. The culture study measured mRNA only, so it cannot say how much of its eightfold signal would become BDNF protein, and the 2006 abstract reports BDNF and trkB but not NGF. The NGF effect therefore rests, among the papers cited here, on cultured cells alone.

Selank, and its unlikely parent

Selank is an anxiolytic in its research profile, often discussed alongside models of generalized anxiety disorder, emotional stability, and stress management for overstimulated individuals, and described as providing anxiety relief without psychostimulant effects, sedation, drowsiness, or risk of dependence that characterises benzodiazepines.^[1]

The odd part is where it comes from. Tuftsin is an immunological peptide, cleaved from the Fc region of immunoglobulin G, and its known activity is stimulating phagocytosis in macrophages and neutrophils. An anxiolytic derived from an antibody fragment is not an obvious design path.

Reported mechanism of action theories include effects on GABAergic, dopaminergic, and serotonergic signalling and on the enzymes that degrade endogenous enkephalins, alongside potential antidepressant properties and effects on BDNF expression that overlap with Semax territory. I am going to be straight that I find the mechanistic picture for Selank considerably muddier than for Semax, and I do not think that is only my reading. A compound that touches various neurotransmitter systems like GABA, serotonin, enkephalin metabolism and neurotrophins is either genuinely pleiotropic or has been studied by measuring many things and reporting whatever moved.

PrymaLab supplies Selank as 5mg and 11mg reference material, and in nasal spray and autoinjector formats, all for laboratory research use only.

The tuftsin question

Selank's parent deserves more than the sentence it usually gets, because the lineage raises a question about the compound that I have not seen addressed.

Tuftsin is Thr-Lys-Pro-Arg, four residues cleaved from the Fc region of immunoglobulin G heavy chain. It was characterised as an immunostimulant: it binds receptors on macrophages and neutrophils and promotes phagocytosis. That is its documented biology and it has nothing obvious to do with anxiety.

So Selank is an immunomodulatory tetrapeptide with a stabilising tail attached, studied and approved for anxiolytic effects. Either the anxiolytic activity emerges from the modification, or tuftsin itself has central actions that its immunological characterisation missed, or the two effects are connected through something like neuroimmune signalling.

The third possibility is the interesting one and it is not far-fetched. Immune signalling, emotional regulation in regions like the amygdala, and mood are connected through well-documented routes, and cytokines influence behaviour. A peptide acting on macrophage receptors having downstream behavioural effects under conditions of chronic stress to support stress resilience is not absurd.

What I cannot tell you is which of those explanations the Russian literature supports, or whether it addresses the question at all. This is the clearest example of the access problem described in the next section: the question is obvious, the answer probably exists somewhere, and I cannot get to it.

It also means that when a supplier page describes Selank's mechanism as GABAergic or serotonergic, that description may be downstream of something else entirely, and it is presented with more confidence than the lineage justifies.

The evidence problem nobody states

Every page comparing these two mentions that both are approved in Russia, usually as a credential. It is a real fact and it deserves more scrutiny than it gets.

Approval means a regulator reviewed a dossier and permitted marketing. For a Western researcher trying to evaluate these compounds within peptide therapy, the useful question is whether that dossier can be examined, and largely it cannot. The supporting trial literature is predominantly Russian-language, much of it published in journals that are not indexed in the databases most people search, and a good deal of it predates the era when trial registration and data sharing became normal expectations.

That is not an accusation that the work is bad. It is a statement that it is hard to check, and those are different things which get conflated in both directions. Sceptics treat inaccessibility as evidence of weakness. Advocates treat approval as though it carried the same information as an FDA approval whose review documents are downloadable. Neither is right, especially when trying to evaluate clinical safety or potential side effects such as headache.

What it means practically: for Semax and Selank you are relying on secondary summaries more heavily than for a compound whose primary literature you can read. I would weight claims about these two accordingly, and I would be especially cautious about specific numerical claims that circulate without a traceable source.

The amidate variants, and what they change

Search around either compound for long enough and you will hit N-Acetyl Semax Amidate and N-Acetyl Selank Amidate, usually abbreviated NASA and NA-Selank. They are worth understanding because the modification follows the same logic as the PGP tail and applies it to the other end of the molecule.

Two changes are made. N-terminal acetylation caps the free amino group at the front of the peptide with an acetyl group. C-terminal amidation converts the terminal carboxylic acid to an amide.

Both changes attack the same problem the PGP tail attacks, from different directions. Aminopeptidases require a free N-terminal amino group to grip, and acetylation removes it. Carboxypeptidases require a free C-terminal carboxyl, and amidation removes that. The base peptide already has proline residues obstructing the C-terminus; the amidate versions add chemical protection on top of geometric protection.

There is a second effect that gets less attention. Amidation removes a negative charge from the C-terminus, which changes the molecule's overall charge distribution. For a compound intended to cross mucosal membranes, charge is one of the properties that determines whether it gets anywhere, so the modification is not purely about protease resistance.

What I have not been able to establish is whether the amidate versions have their own body of published research or whether the evidence is assumed to transfer from the parent compounds. My impression is that most of it transfers by assumption. Acetylation and amidation are conservative modifications, but they are modifications, and a compound with a capped N-terminus is not identical to one without.

PrymaLab does not currently stock either amidate form. If we add them, the honest framing on the product page will be that they are stability-modified analogues of compounds with Russian clinical history, rather than compounds with clinical history of their own.

Both are usually studied intranasally

One practical thing these two share beyond the tail: the Russian clinical application of both has been predominantly intranasal, and that is not incidental to how they are discussed.

Semax at 813 Da and Selank at roughly 751 Da are small by peptide standards, which puts them in the size range where paracellular passage through the nasal epithelium is plausible rather than aspirational. That distinguishes them from most of the nasal spray catalogue, where the compounds are considerably larger and the route is more of a hope than a mechanism.

The nose-to-brain question is also more relevant for these two than for a peripherally acting compound, because both are studied for central effects. Olfactory and trigeminal pathways offer a route toward the central nervous system that does not require crossing the blood-brain barrier, and for a compound whose target is in the brain that pathway is the whole argument for the route.

The caveats from the intranasal delivery reference apply in full. The olfactory region is a small fraction of total nasal surface area in humans, ordinary spray devices deposit most of their volume lower down, and the fraction of an applied dose reaching the CNS by that route in humans has been contested for two decades. Being the right size to permeate is necessary and not sufficient, and local tolerance factors like nasal irritation must still be considered.

How much Semax reaches the brain, and what it becomes

One study has put numbers on the nasal route for Semax itself. Shevchenko and colleagues prepared Semax labelled with tritium at its C-terminal proline, with a molar radioactivity of 56 Ci/mmol, and followed the label in the brain and blood of outbred white rats after intranasal application (Shevchenko et al. 2006). Two minutes after application, 0.093% of the total radioactivity introduced was found per gram of brain tissue, and 80% of that brain radioactivity belonged to intact Semax, the rest to its metabolites (Shevchenko et al. 2006).

The same study found that Semax is degraded quickly. Across the biological samples the tripeptide Pro-Gly-Pro was the predominant species relative to the total of Semax and its metabolites (Shevchenko et al. 2006). Two points follow for reading the wider literature. A small fraction of labelled peptide was detectable in brain tissue within minutes, which is evidence that the nasal route delivers some intact Semax to the rat brain, not a measure of how much reaches the brain in other species or from other devices. And because the C-terminal Pro-Gly-Pro outlasts the parent peptide, effects measured hours after Semax may involve its fragments as well as the heptapeptide, which bears on the question raised above about whether the tail is inert. The paper appeared in Russian, in Bioorganicheskaya Khimiya, and no replication is among the references cited here.

Where they sit in 2026

On 23 and 24 July 2026 the Pharmacy Compounding Advisory Committee reviewed seven peptides for the 503A affirmative list. One of these two was included and the other was not.

Semax passed 8 to 5 with one abstention on 24 July.^[2] Epitalon passed 7 to 5 the same day. DSIP, often studied for sleep quality and submitted as emideltide, was rejected 6 to 7.^[2]

Selank was not among the seven reviewed, so it has no PCAC position either way.

Three things follow. The votes are advisory and not binding on the FDA.^[2] None of the recommended compounds is legal to compound today, because rulemaking has to run and the HHS Secretary must approve any addition.^[2] And an 8 to 5 vote with an abstention is a divided committee deciding narrowly rather than an endorsement.

For laboratory reference material this changes nothing directly, since 503A governs pharmacy compounding rather than research supply. It does mean Semax is now a compound under active regulatory attention for compounding pharmacies and Selank is not, which is a difference between them that did not exist a year ago.

One has a storage problem the other does not

This falls out of the sequences and it is the kind of practical difference that never appears in a nootropic comparison.

Semax has methionine at position 1. Methionine oxidation to methionine sulfoxide is the most common peptide degradation reaction, and a methionine sitting at the exposed N-terminus is about as accessible as it gets. The oxidised product weighs 16 daltons more than the parent and is otherwise intact, so a mass check that is not looking for it can miss it.

Selank contains no methionine, no cysteine and no asparagine. Threonine, lysine, proline, arginine, proline, glycine, proline. Oxidation has no obvious target and deamidation has no obvious target. Two of the four major degradation routes are largely unavailable to it.

Both are proline-rich at the C-terminus, which is the point of the PGP tail, and both are therefore reasonably resistant to enzymatic degradation. Chemical degradation in storage is a separate matter from enzymatic degradation in circulation, and on the chemical side these two are not equivalent.

The practical version: if you are comparing certificates of analysis for these two, ask specifically whether the Semax assay resolves the sulfoxide. For Selank that question matters much less. Fuller treatment of the underlying chemistry is in the peptide storage and stability reference.

Side by side

What research has not established

No published head-to-head comparison places Semax and Selank against each other under matched conditions. Every comparison available, including this one, assembles separate literatures with separate designs and endpoints.

No study I can find runs Pro-Gly-Pro alone as a control arm, which leaves the contribution of the shared tail unresolved for both compounds.

No Western regulatory review has assessed either compound, and the Russian evidence base is difficult to examine independently for the reasons above.

For Selank specifically, the mechanistic picture is reported across GABAergic, serotonergic, enkephalinergic and neurotrophic systems, and I have not seen a convincing account of which of those is primary. That is a real gap and not a rhetorical hedge.

What is reasonably established: the sequences, the design logic behind the PGP tail, the direction and rough magnitude of the Semax BDNF effect in rodents, the Russian approval status of both, and the PCAC outcome. Those are checkable. Most of what surrounds them in this category is not.

If the two were compared directly

Most content on these two compounds describes them. Very little of it says what an experiment separating them would look like, which is odd given that both are sold as research material.

Three design problems stand out, and none of them is exotic.

The endpoint problem. Semax and Selank are studied against different outcomes, so a head-to-head needs an endpoint both could plausibly move. BDNF expression is the obvious candidate, since Semax has direct data there and Selank has been reported to affect it too. Anxiety-like behaviour in a rodent model is the other, and it favours Selank by construction. Picking either one loads the result before you start, which may be why nobody has published the comparison.

The control problem. Covered earlier, and it applies to both compounds equally. If Pro-Gly-Pro has activity of its own, then a two-arm study comparing Semax against Selank cannot distinguish core effects from tail effects, because both arms contain the tail. A useful design needs four arms: Semax, Selank, PGP alone, and vehicle. I have not found that done and it is the first thing I would fund.

The route problem. Both have Russian clinical histories built on intranasal administration, and most benchtop work uses parenteral routes like subcutaneous injection or direct application to cultured cells. Those are not interchangeable, and a comparison run one way does not automatically inform the other.

None of this makes the compounds uninteresting. It means the confident rankings circulating online rest on cross-study inference rather than on anyone having run the comparison, and cross-study inference between two literatures with different endpoints, different models and different languages is weak evidence however many times it gets repeated.

How These Are Characterised

Both are a synthetic peptide (specifically a heptapeptide) and both are confirmed by reversed-phase HPLC for purity and mass spectrometry for identity.

Two things are worth specifying. For Semax, ask whether the method resolves methionine sulfoxide, for the reasons above. For both, ask whether the certificate states the counterion, usually acetate or trifluoroacetate, because it affects the salt-corrected peptide content and therefore what a stated milligram figure actually contains. That second point applies across the catalogue and is one of the more common sources of confusion when comparing suppliers on price per milligram.

At PrymaLab, every lot is tested in PrymaLab's own laboratory (RP-HPLC purity, LC-MS identity), and certificates can be checked at https://prymalab.net/testing/verify/. No lot-specific figures are asserted in this general reference.

Frequently asked questions

What is the difference between Semax and Selank?

Different parent molecules and different research areas. Semax is an ACTH fragment studied for neurotrophic signalling. Selank is a tuftsin derivative studied for anxiolytic effects. They share a design feature but not a mechanism.

Why do Semax and Selank both end in Pro-Gly-Pro?

Both were built by attaching a Pro-Gly-Pro tail to a short bioactive fragment to resist enzymatic degradation. Proline-rich sequences are poor peptidase substrates. The tail has also been reported to carry activity of its own.

What does Semax do to BDNF?

A single rodent administration produced about a 1.4-fold increase in BDNF protein and a 3-fold increase in BDNF mRNA. The gap between those figures points to a limit downstream of transcription.

Are Semax and Selank approved anywhere?

Both are approved in Russia and neither by the FDA or EMA. The supporting literature is largely Russian-language and difficult to verify independently.

Did the FDA review Semax or Selank in 2026?

Semax only, recommended 8 to 5 with one abstention by PCAC on 24 July 2026. Selank was not reviewed. The vote is advisory and rulemaking has not completed.

Which one is more stable in storage?

Selank, on sequence grounds. Semax has methionine at position 1 and an oxidation route. Selank has no methionine, cysteine or asparagine.

Are Semax and Selank approved for human use?

No. Material supplied here is laboratory reference material for research use only.

Related reading

• Cerebrolysin: composition, trials and FDA status
• Oxytocin neuronal circuits: OXTR, reward and fear

References