PE-22-28 for Health & Longevity
Evidence Review created on 11/14/2026 using AI4L + evidence-mcp / Opus 5.5
Also known as: PE 22-28, Mini-Spadin, Minispadin, GVSWGLR
Motivation
PE-22-28 (also called mini-spadin) is a laboratory-made chain of seven amino acids copied from a short stretch of a larger peptide that the human body releases when it processes a protein called sortilin. It was built to block a potassium channel that helps keep nerve cells electrically quiet — a target that drew attention after mice bred without that channel behaved differently from ordinary mice in mood tests.
The parent peptide was first described in 2010 by a research group in southern France; the shorter version followed in 2017, meant to be cheaper to produce and longer-lived in the body. Rodent work examined whether a few days of treatment, rather than the weeks ordinary medicines need, was enough to change mood-related behavior. No study of either molecule in people has been published or registered. The compound is nonetheless sold online as a research chemical and discussed in self-experimentation communities, which is how most people encounter it.
This review examines what the published record holds for PE-22-28: the cell and animal evidence behind each claimed effect, what is known and unknown about its safety, its regulatory position, and the practical questions raised by a molecule with no human data.
Benefits - Risks - Protocol - Conclusion
Recommended Reading
Sources giving a high-level overview of PE-22-28 or of TREK-1 (TWIK-related potassium channel 1, the background potassium channel it blocks).
-
Fighting against depression with TREK-1 blockers: Past and future. A focus on spadin - Djillani et al., 2019
The originating group’s own account of the programme that produced PE-22-28, including why spadin was shortened. The same group patented analogs of the fragment, a direct financial interest.
-
TREK-1 in central nervous system diseases - Chen et al., 2026
An independent commentary arguing TREK-1, the channel PE-22-28 blocks, plays opposite roles across stroke, pain, epilepsy, depression and cognition, and weighing whether blockade can be targeted at all.
-
Role for TREK-1 as a polymodal sensor and regulator of cell activity - Winkle et al., 2026
Recent independent synthesis of what opens and closes TREK-1, the channel PE-22-28 blocks — stretch, heat, acidity, lipids — which bears on how a blocker would behave across tissues.
-
Mechanosensitive TREK-1 two-pore-domain potassium (K2P) channels in the cardiovascular system - Wiedmann et al., 2021
Covers TREK-1, the channel PE-22-28 blocks and a K2P channel (one that leaks potassium to keep cells electrically quiet), outside the brain — atrium, ventricle and vasculature — where blockade carries unintended consequences.
Note: no content from any priority expert platform names PE-22-28 or spadin, and none discusses the TREK-1 channel in substantial depth, so four items are listed rather than five. The compound has no human data and no approved product, which is the most likely reason these platforms have not covered it.
Grokipedia
-
Covers discovery from spadin, the amino-acid sequence, channel potency and the rodent behavioral and neurogenesis findings. Useful orientation, but an AI-generated page and no substitute for the primary papers.
Examine
No Examine article exists for PE-22-28. Examine covers supplements and dietary compounds; PE-22-28 is an unapproved synthetic peptide sold for laboratory use, which falls outside that scope.
ConsumerLab
No ConsumerLab article exists for PE-22-28. ConsumerLab tests commercially sold supplements and foods; no consumer product containing this peptide is lawfully marketed, so there is nothing for it to test.
Systematic Reviews
No systematic reviews or meta-analyses for PE-22-28 were found on PubMed as of October 27, 2026.
Neither side of the trade-off is represented: the literature contains no systematic review or meta-analysis of the claimed antidepressant and neurogenic effect, and none of the principal risk — loss of the channel’s protective roles in seizure threshold, pain signalling, cerebral blood flow and cardiac rhythm.
Mechanism of Action
Sortilin (neurotensin receptor-3) is cleaved by furin, releasing a 44-amino-acid propeptide. Spadin is residues 12–28; serum breakdown pointed to a shorter active core, residues 22–28 (Djillani 2017).
Its target, TREK-1 (gene KCNK2), is a two-pore-domain (K2P) background channel: it leaks potassium out and holds membrane voltage below the firing threshold, so blocking it makes neurons easier to excite. Serotonin acting on 5-HT1A autoreceptors opens TREK-1 and damps serotonin-neuron firing; lifting that brake engages the protein kinase A – cyclic AMP response element-binding protein (CREB) – brain-derived neurotrophic factor (BDNF) cascade tied to new-neuron formation, as a 2026 commentary sets out (Chen 2026).
PE-22-28 inhibits the human channel with a half-maximal inhibitory concentration (IC50, the amount giving half the maximum effect) of 0.12 nanomolar against 40–60 for spadin, and spares the related potassium channels TREK-2, TRAAK, TRESK and TASK-1 at 100 nanomolar and, at 10 micromolar, the cardiac hERG current (which resets heart rhythm) (Djillani 2017). The channel sits mainly in brain, also heart, smooth muscle and pancreas, per the originating group’s review (Djillani 2019). Its distribution is unmeasured: no study reports whether it enters the brain. Being short, it is cleared by blood enzymes, not liver enzymes; serum stability is brief, yet its analogs acted up to 23 hours in mice. No human pharmacokinetic study exists.
The blockade model is contested: in frog egg cells, spadin left basal and activated TREK-1 current unchanged and only blunted arachidonic-acid activation after pre-exposure, implying action away from the pore (Ma 2020).
Historical Context & Evolution
TREK-1 entered depression research in 2006, when a group at the Institut de Pharmacologie Moléculaire et Cellulaire in Valbonne, France reported that mice with the Kcnk2 gene deleted resisted depression-like behavior across five models and showed stronger serotonin neurotransmission (Heurteaux 2006). Two years earlier the same laboratory had documented the channel’s other face: deletion left mice more vulnerable to seizures and to ischemia (loss of blood supply) in brain and spinal cord, and harder to anesthetize with volatile agents (Heurteaux 2004).
The search for a blocker produced spadin in 2010, presented as the first natural antidepressant peptide and a new concept in antidepressant design (Mazella 2010). Its limitation was duration: behavioral effect in mice halved by six hours. Mirror-image, reverse-sequence (retro-inverso) analogs raised affinity five- to sixfold and action duration threefold, to sixteen hours (Veyssiere 2015). Shortening the peptide was the next route, and PE-22-28 was published in 2017 as the shortest fragment retaining channel inhibition, with longer action and lower manufacturing cost (Djillani 2017).
Nothing in this programme has been retracted, and nothing moved into people. Six inventors hold a patent claiming analogs of the fragment, not PE-22-28 itself, assigned to the CNRS and Université Côte d’Azur and running to 2037 (US11220527B2). Twenty years on, no TREK-1 blocker has been tested clinically, and a 2026 independent commentary concluded that the channel’s opposite roles across diseases may themselves constrain drug development (Chen 2026). Use outside the laboratory followed the preclinical papers into research-chemical sales, not a clinical result.
Expected Benefits
Every observation below comes from cells or rodents. No benefit is graded above Speculative because no human outcome of administering PE-22-28 has been measured in any study.
High 🟩 🟩 🟩
No benefit reaches High: that grade requires a human clinical endpoint or a validated clinical surrogate replicated by independent groups, and no human outcome study of PE-22-28 exists.
Medium 🟩 🟩
No benefit reaches Medium: that grade requires the same class of human outcome in at least a single trial or consistent controlled observational data, and none has been reported.
Low 🟩
No benefit reaches Low: the only human data in this field are observational serum concentrations of the body’s own parent propeptide in depressed and post-stroke patients (Devader 2017, Mazella 2025) — a correlation with the endogenous peptide, not an outcome of giving this fragment.
Speculative 🟨
Rapid Change in Depression-Related Behavior
Mice treated for four days showed shorter immobility in the forced swim test and shorter latency to feed in novelty-suppressed feeding, including in a corticosterone-induced model (Djillani 2017). Basis is rodent behavioral screening only.
Hippocampal Neurogenesis and Synapse Formation
Four days of treatment roughly doubled new-cell labeling in mouse hippocampus, and cultured mouse cortical neurons raised the synapse protein PSD-95 (Djillani 2017). Cell-count and protein markers, not human cognition.
Protection and Recovery After Experimental Stroke
In mouse focal ischemia, a dose about a hundredfold lower — which opens rather than blocks the channel — preceded better motor and memory scores (Pietri 2019). One laboratory, animal model only.
Pancreatic Beta-Cell Survival and Insulin Release ⚠️ Conflicted
Sortilin-derived peptides protected beta cells and mini-spadin promoted proliferation (Daziano 2021); spadin raised insulin in mice (Hivelin 2016). High-fat-fed channel-deleted mice showed worse glucose tolerance (Kim 2025). Net: animal and cell only, direction unresolved.
Benefit-Modifying Factors
No factor below has been tested as a modifier of any PE-22-28 effect; each is an inference from the target’s biology, stated as such.
-
KCNK2 genotype: In the STARD antidepressant trial, variants in *KCNK2 — the gene encoding the target — were associated with treatment resistance across second- and third-step therapies (Perlis 2008). Whether genotype alters response to a blocker is untested.
-
Baseline propeptide concentration: Serum concentrations of the body’s own sortilin-derived propeptide are lower in major depressive disorder and rise in those who respond to electroconvulsive therapy (Devader 2017, Roulot 2018), which makes baseline a candidate modifier.
-
Sex: In rhesus macaques, female sex was associated with lower cortical TREK-1 protein (Bogdan 2011), and 17β-estradiol potentiates channel activity through the G protein-coupled estrogen receptor (Choudhury 2018). Direction of any effect on blockade is unknown.
-
Pre-existing conditions: A 2026 commentary reports that depression, stroke, epilepsy, pain and cognitive impairment engage the channel in opposing directions (Chen 2026), and type 2 diabetes adds another (Kim 2025), so one action could help one condition while working against another.
-
Age: All published work used young adult mice. Nothing addresses older animals or people, and no age-stratified result exists for either the peptide or the parent propeptide at the older end of the target range.
-
Arachidonic acid and lipid state: Inhibition was measured on channels first opened by arachidonic acid, and blockade appears to need the open state (Djillani 2017, Ma 2020). Membrane lipid status could therefore change the effect.
Potential Risks & Side Effects
No risk is graded above Low. The compound has never been given to a person in a published study, so the evidence is either animal and in-vitro work on the target or human harm attached to the way unregulated peptides are supplied and injected.
High 🟥 🟥 🟥
No risk reaches High: that grade requires a human adverse-event or clinical-endpoint finding replicated by independent groups, and no human exposure to this peptide has been documented at all.
Medium 🟥 🟥
No risk reaches Medium: that grade requires the same class of human outcome in a single trial or consistent controlled observational data, and nothing of that kind exists for this peptide or for any TREK-1 blocker.
Low 🟥
Infection, Contamination and Mislabeled Material
The only human harm evidence attaches to the supply route, not the molecule. Analyses of seized peptide products found fake or non-approved items carrying undeclared substances (Krug 2014), and non-sterile injection of an unregulated peptide caused soft-tissue infection (Neill 2017). Indirect: same practice, different peptides.
Magnitude: Not quantified in available studies. No cohort, registry or poison-center series has measured infection or adulteration rates among people self-administering research peptides, and the single infection report has no comparison group; the product-analysis work reports composition of seizures, not rates of harm in users.
Speculative 🟨
Lower Seizure Threshold ⚠️ Conflicted
Channel-deleted mice are more sensitive to seizures (Heurteaux 2004); spadin at 100 micrograms per kilogram did not worsen chemically induced seizures in mice (Moha Ou Maati 2012). Net: theoretical hazard, unseen at tested doses.
Cardiac Rhythm Disturbance ⚠️ Conflicted
Mice without cardiac TREK-1 show bradycardia (abnormally slow heart rate) and sinus pauses (missed heartbeats) (Unudurthi 2016), yet spadin left pulse and blood pressure unchanged (Moha Ou Maati 2012). Net: unresolved, with no human measurement.
Heightened Pain Sensitivity ⚠️ Conflicted
Channel-deleted mice show mechanical and thermal hypersensitivity (Alloui 2006); spadin did not alter pain thresholds in mice (Moha Ou Maati 2012) but was pain-promoting in rats (García 2021). Net: blockade may heighten pain.
Reversal of Effect With Dose
The same molecule activated the channel at 0.03 and inhibited it at 3 micrograms per kilogram in mice (Pietri 2019). An unintended amount may produce the opposite of what was sought.
Fat Gain and Impaired Glucose Tolerance ⚠️ Conflicted
Blockade increased lipid accumulation in fat cells; channel-deleted mice on a high-fat diet gained fat with worse glucose tolerance (Kim 2025); yet blockade raised insulin (Hivelin 2016). Net: cell and animal only, direction unresolved.
Worse Outcome From Cerebral Ischemia ⚠️ Conflicted
Channel-deleted mice sustain larger ischemic injury, and the channel mediates fatty-acid dilation of brain bypass vessels (Heurteaux 2004, Blondeau 2007); a 2026 commentary reports spadin reduced dead-tissue volume in mice (Chen 2026). Net: direction unresolved.
Immune Response to a Synthetic Peptide
The Food and Drug Administration (FDA) cites aggregation and peptide impurities as immunogenicity grounds for placing another short injected peptide, GHRP-2, in compounding category 2 (FDA category 2 list). None exists for this peptide.
Risk-Modifying Factors
As with the benefits, no factor below has been tested against a PE-22-28 outcome; each is inferred from the target’s biology and stated as inference.
-
KCNK2 variants: A mutation in the channel’s ion-selecting pore that left it overactive caused right ventricular outflow tract tachycardia (abnormally fast heart rate) in a patient (Decher 2017), showing that channel activity sets cardiac risk in humans.
-
Circulating anti-TREK-1 autoantibodies: Activating autoantibodies against the channel were found in 7 of 14 patients with short-coupled ventricular fibrillation (a lethal heart rhythm) (Li 2024). Whether a blocker interacts with that state is unexamined.
-
Baseline propeptide concentration: Because the endogenous propeptide occupies the same system and its serum concentration varies with depression state and recovery (Mazella 2025), baseline level is a plausible determinant of how much added blockade does.
-
Sex: Estradiol potentiates channel activity through the G protein-coupled estrogen receptor (Choudhury 2018), so hormonal status could shift the balance between blockade and baseline activity. No sex-stratified safety data exist for the peptide.
-
Pre-existing conditions: A 2026 commentary reports channel activity protective in epilepsy and pain models, contested in ischemia (Chen 2026); sinus node disease (Unudurthi 2016) and diabetes (Kim 2025) add others. Each is a candidate route to harm.
-
Age: No published work used aged animals, and nothing addresses people at the older end of the target range, where seizure threshold, cerebral collateral flow and sinus node reserve are all lower to begin with.
Key Interactions & Contraindications
Documented interactions:
- None identified. No human interaction or pharmacokinetic study of PE-22-28 has been published, and no pairing has been tested in a person either with or without a resulting interaction.
Theoretical interactions (inferred from mechanism, not studied in humans):
-
Selective serotonin reuptake inhibitors (SSRIs — drugs that raise serotonin, e.g., fluoxetine, sertraline): Fluoxetine inhibits the same channel (IC50 19 micromolar) and its metabolite norfluoxetine more potently (9 micromolar) (Kennard 2005). Inferred additive blockade; consequence is excess serotonergic activity. Dose separation cannot avoid it.
-
Methylxanthines (caffeine, theophylline): Both inhibit the human channel through the cyclic AMP – protein kinase A pathway, which contributed to their pro-convulsant reputation (Harinath 2005). Inferred additive blockade; clinical consequence would be a lower seizure threshold. Mitigation would mean limiting intake, not timing it.
-
Riluzole: Activates the channel (Duprat 2000), the opposite action. Inferred mutual antagonism; clinical consequence is loss of either drug’s intended effect rather than toxicity. No dose adjustment can resolve directly opposing actions on one target.
-
Mood stabilizers (lithium, valproate, carbamazepine): These activate the same channel (Kim 2017), the opposite action. Inferred mutual antagonism; clinical consequence is loss of either agent’s intended effect, which matters most where the peptide is taken for mood. No timing separation resolves it.
-
Volatile anesthetics (sevoflurane, isoflurane, halothane): Channel-deleted mice were resistant (Heurteaux 2004); a later study using different knockout alleles found none (Spencer 2023). Inferred reduced sensitivity, consequence intraoperative awareness or higher agent requirement, is unresolved; the mitigating action is disclosure to an anesthetist.
-
Polyunsaturated fatty acids (the omega-3 fats EPA and DHA — eicosapentaenoic and docosahexaenoic acid — from fish oil, and α-linolenic acid): These open the channel and dilate cerebral arteries through it (Blondeau 2007). Inferred mutual opposition; consequence is blunting of the peptide’s action.
-
Quercetin: Opens the channel, the opposite action, and spadin blocked both its neuroprotective and its anti-manic effect in mice (Ren 2021). Inferred mutual opposition; consequence is loss of either agent’s intended effect. No timing separation resolves it.
-
Insulin secretagogues (drugs that make the pancreas release insulin — sulfonylureas such as glipizide, meglitinides such as repaglinide): Channel blockade raised glucose-stimulated insulin in mice (Hivelin 2016). Inferred additive insulin release; consequence is hypoglycemia (low blood sugar). Mitigation requires glucose monitoring, which no protocol defines.
-
Other peptides stacked in self-experimentation (e.g., Selank, Semax): Inferred from shared serum peptidase clearance and overlapping action on mood and neurotrophic signalling; consequence is an unpredictable net effect, with neither benefit nor harm attributable to one agent (Hailu 2026).
Populations who should avoid PE-22-28:
-
Anyone with epilepsy or a prior unprovoked seizure: Inferred from increased seizure sensitivity in channel-deleted mice and from the pro-convulsant profile of other channel inhibitors (Heurteaux 2004); no human data.
-
Anyone with prior ischemic stroke or transient ischemic attack: Inferred from loss of channel-mediated widening of brain bypass vessels and greater ischemic injury in channel-deleted mice (Blondeau 2007); no human data.
-
Anyone with sinus node disease (weak natural pacemaker), bradycardia (resting heart rate below 50 beats per minute) or unexplained syncope (fainting): Inferred from bradycardia and sinus pause in cardiac channel-deleted mice (Unudurthi 2016); no human data.
-
Anyone with an inherited channel disorder (channelopathy) of this gene or with short-coupled or idiopathic ventricular fibrillation: Inferred from the human gain-of-function tachycardia phenotype and the activating-autoantibody finding, which together establish that channel activity sets arrhythmic risk in people (Decher 2017, Li 2024).
-
Anyone pregnant, breastfeeding or seeking conception: Inferred from the mechanism — the target channel is expressed in fetal brain and smooth muscle, including uterine, and sets excitability there — together with the absence of reproductive, developmental or lactation data in any species; no human data.
-
Anyone with a chronic pain condition: Inferred from mechanical and thermal hypersensitivity and amplified inflammatory pain in channel-deleted mice (Alloui 2006); no human data.
Risk Mitigation Strategies
Parameters below that carry no citation reflect common practice rather than any established protocol; no risk mitigation strategy for this peptide has been studied, and none has been validated in a person.
-
Treating the absence of human data as the principal risk: No toxicology, reproductive or pharmacokinetic study of this peptide was found in any species, so no monitoring plan can detect a harm whose form is unknown. This mitigates misplaced confidence.
-
Avoiding the compound in the populations above: Each exclusion addresses an inferred harm — seizure, stroke extension, sinus arrest, lethal heart rhythm, fetal exposure, pain — and avoidance is the only measure available where no threshold or antidote is defined.
-
Baseline and interval electrocardiogram if used: Addresses inferred bradycardia and sinus-pause risk from cardiac channel loss (Unudurthi 2016). No source gives an interval; common practice repeats it at 1 and 4 weeks, stopping on new bradycardia or pause.
-
Glucose monitoring alongside any insulin secretagogue: Addresses inferred additive insulin release (Hivelin 2016). Glucose below 70 mg/dL defines level 1 hypoglycemia requiring action, per the American Diabetes Association, whose members deliver the care it recommends (ADA Standards of Care).
-
Single-agent use rather than stacking: A narrative review of gray-market peptide use reports that combining poorly characterized peptides removes attribution of any adverse event to one agent (Hailu 2026). This mitigates unassignable harm.
-
Independent identity and purity testing of any material: Addresses the documented risk of a wrong or undeclared substance in seized peptide products (Krug 2014). A supplier certificate of analysis does not establish sterility or batch identity.
-
Sterile single-use technique for any injection: Addresses soft-tissue and mycobacterial (slow-growing bacteria) infection documented after non-sterile injection of an unregulated peptide product (Neill 2017), which is a hazard of the practice independent of the molecule.
-
Disclosure to clinicians, especially before anesthesia: Addresses the disputed resistance to volatile anesthetics (Heurteaux 2004) and the tendency of unregulated peptide use to go unreported, which a narrative review of that market reports leaves symptoms misattributed (Hailu 2026).
-
Reporting any adverse event to the national safety-monitoring system: Addresses the surveillance gap that a narrative review of this market says leaves harms from gray-market peptides undetected (Hailu 2026). In the United States this is the FDA MedWatch programme.
Therapeutic Protocol
No human regimen exists for PE-22-28. There is no approved product, no clinical guideline, no trial regimen in a person, and no named practitioner protocol traceable to a published source. Parameters below that carry no citation reflect common practice in the preclinical literature rather than any human protocol.
-
Published regimen, acute: A single intraperitoneal (into the abdominal cavity) dose of 3.0–4.0 micrograms per kilogram in mice, the amount used in the forced swim test (Djillani 2017). A rodent laboratory dose; no validated human equivalent exists.
-
Published four-day regimen: 3.0 micrograms per kilogram intraperitoneally once daily for four days, the schedule under which neurogenesis and behavioral change were measured in mice; 1 mg/kg by oral tube feeding remained active in the same test (Djillani 2017).
-
Competing regimen, different aim: For stroke protection, mini-spadin was given at 0.03 micrograms per kilogram daily for seven days to open the channel, then at 3 micrograms per kilogram to block it (Pietri 2019).
-
Unsourced figures in circulation: Vendor pages and forums give fixed microgram amounts by nasal or subcutaneous route. None is traceable to a study, and no published work on this peptide used either route, so bioavailability by them is unknown.
-
Route: All published work used intraperitoneal injection, with one oral gavage comparison (Djillani 2017). Intraperitoneal injection is a laboratory animal route and is not used in people outside specific clinical procedures.
-
Half-life and duration of action: Serum half-life in vitro is short, but behavioral effect in mice reached 23 hours for the analogs, against 6 hours for spadin (7 in the same paper’s abstract) (Djillani 2017). No human measurement exists.
-
Single versus split dosing: Every published schedule used one administration per day, and the roughly day-long effect in mice is consistent with that. No split-dose comparison has been run, in animals or people.
-
Time of day: Not studied. No published experiment compared morning with evening administration, and the rodent work does not report dosing time, so no timing claim can be sourced.
-
Genetic factors: KCNK2 variants track antidepressant treatment resistance in humans (Perlis 2008). No pharmacogenetic dosing rule follows from this, and no genotype-guided protocol has been tested.
-
Sex differences: Published efficacy work used male mice; estradiol potentiates the channel through the G protein-coupled estrogen receptor (Choudhury 2018). No sex-specific dosing or response data exist.
-
Age: Only young adult animals were studied. Nothing addresses older animals or older adults, where the channel’s protective roles in cerebral flow and cardiac rhythm matter more.
-
Baseline biomarkers: Serum concentration of the endogenous propeptide varies with depression state and with treatment response (Devader 2017), making it the only candidate baseline marker — untested as a dosing input.
-
Pre-existing conditions: A 2026 commentary reports the channel to be protective in epilepsy and pain models, contested in ischemia, and permissive in depression models (Chen 2026), so the condition present determines whether blockade is the wanted direction at all.
Discontinuation & Cycling
-
Intended duration: Unknown. Every published schedule ran four days in mice, or a few weeks in the stroke model (Djillani 2017, Pietri 2019). Nothing establishes whether use would be short-term or continuous in a person.
-
Withdrawal effects: None reported, and none looked for. No study of the peptide, in any species, included a post-treatment observation period designed to detect rebound, dependence or symptom return after stopping.
-
Tapering: Not applicable to anything studied: the published courses were short and were stopped outright. No taper schedule has been described for this peptide or for spadin.
-
Rebound of the target: Chronic blockade of a background potassium channel could in principle alter its surface expression, since the channel’s membrane level is regulated by its partner protein sortilin (Moreno 2018). Untested after discontinuation.
-
Cycling for efficacy: No tolerance appeared over three weeks of spadin or four weeks of continuous analog release (Moha Ou Maati 2012, Veyssiere 2015). Nothing ran longer, and nothing tested this peptide, so cycling rests on no data.
Sourcing and Quality
-
No pharmaceutical-grade source exists: PE-22-28 is not an approved drug anywhere, so no licensed manufacturer, prescribing information or pharmacopeial monograph exists for it. Material reaching consumers is laboratory reagent or, as a narrative review of this market describes it, gray-market product (Hailu 2026).
-
Compounding is not a lawful route: A section 503A pharmacy may use a bulk substance only with a monograph in the US drug standards (USP or NF), as a component of an approved drug, or on the FDA 503A bulks list. PE-22-28 meets none.
-
Research suppliers state the exclusion explicitly: Catalogue material is sold at ≥98% purity for laboratory use and labeled “not for human or veterinary use” (Cayman Chemical product page). Buying reagent-grade peptide does not convert it into a medicine.
-
What a certificate of analysis does and does not show: It documents identity and purity of a tested sample. A narrative review of this market notes it does not establish sterility, bacterial-toxin content, or that the vial shipped matches the batch tested (Hailu 2026).
-
Documented failures in this market: Seized peptide products have proved fake or non-approved, carrying undeclared substances, an unidentified protein, or modified analogs of the labeled peptide (Krug 2014, Krug 2018).
-
Form and salt: Catalogued material is the acetate salt of the heptapeptide (formula weight 773.9), supplied as a solid at ≥98% purity (Cayman Chemical product page). Peptide content differs from gross vial weight, so labeled milligrams overstate deliverable peptide.
-
Storage: The catalogue product is stored at −20 °C (Cayman Chemical product page); common practice keeps a short peptide dry and dark, and cold once reconstituted. Degradation of a mishandled peptide is not visible on inspection.
Practical Considerations
-
Time to effect: Unknown in humans. In mice, a single dose altered behavior within hours and four days changed neurogenesis markers (Djillani 2017). No human time course has been measured.
-
Regulatory status: Not approved by the FDA or any comparable regulator, and absent from both FDA category 2 tables for substances nominated under sections 503A or 503B (FDA category 2 list) — absence reflects that no one nominated it for compounding, not a safety clearance.
-
No clinical trial to join: A ClinicalTrials.gov search for the peptide, for spadin and for the target returned no interventional study of the compound. There is no investigational route to supervised use.
-
Most common pitfall — treating preclinical potency as human efficacy: Sub-nanomolar channel affinity and rapid rodent effects are frequently quoted as if they were clinical results. Reviewers of unregulated peptide use note that technical language is readily mistaken for clinical authority (Hailu 2026).
-
Second pitfall — assuming more is better: The molecule reverses direction with amount, activating the channel at roughly a hundredth of the inhibiting dose (Pietri 2019). Misestimating the amount can invert the intended pharmacology.
-
Cost and payer incentive: Reagent vials are inexpensive, so cost is not a barrier. No insurer or health system has any stake, since nothing is reimbursable; the patent holders are the only party with a financial interest in its adoption (US11220527B2).
-
Competing established options are cheap: Generic antidepressants that themselves inhibit this channel at high concentrations (Kennard 2005) are inexpensive and extensively studied, which weakens the commercial case for developing a patented peptide and helps explain the stalled programme.
Interaction with Foundational Habits
-
Sleep: Direction is unknown; the mechanism cuts both ways. Spadin identifies the channel in thalamocortical relay neurons, whose burst firing underlies sleep rhythms (Bista 2012), and channel-deleted mice may resist volatile anesthetics (Heurteaux 2004). No sleep measurement exists for this peptide in any species; self-reported lethargy in community accounts is uncontrolled anecdote.
-
Nutrition: Direct and opposing. Polyunsaturated fatty acids — the omega-3 fats EPA, DHA and α-linolenic acid — open the same channel (Blondeau 2007), and inhibition was measured on channels first opened by arachidonic acid (Djillani 2017). High fish oil intake plausibly blunts blockade; caffeine inhibits the channel and is additive (Harinath 2005).
-
Exercise: No direct evidence; interaction is indirect at most. A 2026 review reports the channel to be mechanosensitive, opened by membrane stretch and warmth (Winkle 2026), so exercise-related stretch and temperature could oppose blockade. No training study has examined the peptide, and no timing around workouts is supported.
-
Stress management: Indirect, through the same axis the animal work used. Channel-deleted mice showed substantially smaller stress-induced corticosterone rises (Heurteaux 2006), and the peptide was tested in a corticosterone-induced depression model (Djillani 2017). Whether blockade alters the human stress response is unmeasured.
Monitoring Protocol & Defining Success
No monitoring protocol for PE-22-28 has been published, because no study has given it to a person. What follows is the set of measurements the target’s biology makes relevant, not a validated plan.
Before any use, a baseline would fix the values later readings are compared against: a resting electrocardiogram for rhythm and rate, given the bradycardia and sinus-pause phenotype of cardiac channel loss; fasting glucose and glycated hemoglobin, given the insulin-secretion and body-fat findings; and a structured mood score, since mood is the claimed effect.
For ongoing measurement, no cadence rests on evidence. A schedule matching the short rodent courses would repeat the rhythm and glucose measures at 1 week and 4 weeks and the mood score weekly, with any new bradycardia, pause, seizure, hypoglycemia or neurological symptom treated as a reason to stop.
| Biomarker | Optimal Functional Range | Why Measure It? | Context/Notes |
|---|---|---|---|
| Resting heart rate and rhythm (electrocardiogram) | 60–100 beats per minute (standard reference range) | Safety check: new bradycardia or sinus pause would stop use | Inferred risk from cardiac channel-deleted mice (Unudurthi 2016); spadin left pulse unchanged in mice (Moha Ou Maati 2012). No human threshold is defined for this peptide |
| Fasting plasma glucose | 70–99 mg/dL (standard reference range) | Safety check: hypoglycemia would stop use | Channel blockade raised glucose-stimulated insulin in mice (Hivelin 2016). Requires an 8-hour fast; draw with the lipid-independent morning panel |
| Glycated hemoglobin (HbA1c) | 4.0–5.6% (standard reference range) | Marker the intervention is expected to change, in either direction | HbA1c is average blood sugar over about three months. Beta-cell protection would predict improvement (Daziano 2021); channel-deleted mice on a high-fat diet worsened (Kim 2025). No fasting needed; lags by weeks |
| Structured depression score (e.g., PHQ-9) | No established target — track change from the individual’s own baseline | Marker the intervention is expected to change: mood is the claimed effect | PHQ-9 is a nine-item depression questionnaire; its score bands are severity categories, not a reference range. Administer at the same time of day; rodent behavioral tests have no human equivalent endpoint |
Qualitative markers worth tracking alongside the table:
- Mood and motivation, recorded daily rather than recalled
- Energy and daytime sleepiness, the most commonly reported subjective complaint in uncontrolled community accounts
- Sleep onset and continuity
- Pain threshold and any new or worsened pain, given the animal hypersensitivity findings
- Cognitive clarity and working memory in ordinary tasks
- Injection-site appearance, where any injection is used
Emerging Research
-
No registered trial of the compound: A ClinicalTrials.gov search for PE-22-28, spadin, mini-spadin and the target returned no interventional study of the peptide. Until one is registered, nothing in the clinical pipeline can change the evidence grade above Speculative.
-
Anti-TREK-1 autoantibodies in ventricular fibrillation (TRACK-VF, NCT06943365): Observational, 300 participants, recruiting, primary completion July 2027. A positive result strengthens the channel-activity-sets-arrhythmic-risk argument underlying this review’s cardiac caution; a null result weakens it.
-
Human propeptide as a biomarker: Serum concentrations of the endogenous parent peptide tracked depression scores across five years after stroke (Mazella 2025). Positive replication would supply the first human handle on this system; a null result removes the main translational bridge.
-
Binding-site identification: A 2026 commentary notes that most channel modulators’ binding sites remain unidentified, one mapped inhibitor aside, which blocks precise targeting (Chen 2026). Mapping them would make tissue-selective blockade conceivable; failure keeps whole-body effects unavoidable.
-
Evidence that could weaken the case — mechanism: Spadin did not inhibit basal or activated channel current in oocytes and only blunted lipid activation after pre-exposure (Ma 2020). Independent replication of this on PE-22-28 would undercut the stated mechanism.
-
Evidence that could weaken the case — metabolism: Channel blockade increased adipocyte lipid accumulation, and channel-deleted mice gained more fat with worse glucose tolerance (Kim 2025). Extension to chronic dosing would turn a speculative metabolic risk into a measured one.
-
Beta-cell direction: The originating group reports protection of beta cells and proliferation with mini-spadin (Daziano 2021). Independent replication would open a metabolic indication; failure to replicate would leave the metabolic claims resting on one laboratory.
-
Gray-market surveillance: Reviewers call for laboratory testing of online peptide products and better adverse-event capture (Hailu 2026). Such work would quantify the supply-route risk that is currently the only documented human harm.
Conclusion
PE-22-28 is a seven-amino-acid fragment of a peptide the body makes itself, built in one French laboratory to block a potassium channel that keeps nerve cells quiet. Everything known about what it does comes from cells and from mice: less passive behavior in standard rodent mood tests, more newly formed cells in a memory structure of the brain, protection of insulin-producing cells, and, at a far smaller amount, protection after an experimental stroke. None of these observations has been repeated in a person. No study in people has been published or registered, no amount has been established for human use, and no safety or toxicity work has been reported in any species. Almost the whole record comes from the group that patented its close variants.
That absence is the central fact. The channel does several jobs at once — it also raises the threshold for seizures, limits pain signalling, helps blood reach tissue around a blocked brain vessel, and steadies the heart’s own rhythm — so the action that changed behavior in mice carries plausible costs nobody has measured. The molecule also reverses direction with amount, opening the channel at very small quantities and closing it at larger ones.
Supply is a separate problem: nothing approved exists, no pharmacy may lawfully prepare it, and seized products in this market have proved fake or held a substance other than the label’s. Measured against a long healthy life, the compound stands as a laboratory hypothesis rather than a usable one.