PE-22-28 Dosage: What the One Published Study Actually Measured

PE-22-28 at a glance

What it is

A synthetic 7-amino-acid peptide corresponding to residues 22-28 of spadin, which is itself PE 12-28, a fragment of the 44-amino-acid sortilin propeptide (PE). PE-22-28 was not designed from scratch — it was identified by studying what spadin degrades into in blood, then testing which of those shorter sequences still worked

Researched for

Antidepressant-like activity via selective blockade of the TREK-1 (KCNK2) potassium channel, plus hippocampal neurogenesis and cortical synaptogenesis — all in mice and in cell lines

Commonly reported range

50-200 mcg per administration is what circulates in community and vendor material. No human dose exists. The published rodent doses are 3.0 mcg/kg by intraperitoneal injection or 1 mg/kg by oral gavage in the sub-chronic behavioural experiments

Route reported

Community material describes subcutaneous and sometimes intranasal use. The published work used intraperitoneal injection and oral gavage in mice — neither of which is what the market describes

Reported frequency

Once daily in community protocols. The published behavioural work used 4-day sub-chronic treatment

Reported cycle

8-16 weeks in community protocols, with nothing published at any duration in humans. The longest published treatment is 4 days

Plasma half-life

Not established for PE-22-28 in any species. The frequently quoted '23 hours' is NOT a half-life and is NOT PE-22-28 — it is a half-time of behavioural EFFECT measured for biotinylated-G/A-PE 22-28, a doubly modified analog, at 40 mcg/kg. The source paper explicitly contrasts a short in vitro serum half-life with a long duration of effect

Regulatory status

Not an approved medicine in any jurisdiction and never tested in a human trial — a PubMed filter for clinical trials of spadin returns nothing. Not approved for human use, which is the category WADA prohibits under S0. Research use only

Reported ranges from research/community — examples, not recommendations.

What it is / mechanism

PE-22-28 blocks TREK-1, a two-pore-domain background potassium channel encoded by KCNK2, and the reason anyone cares about that channel is genetic rather than pharmacological. Deleting kcnk2 in mice produces a depression-resistant phenotype across five different models of depression, with enhanced serotonergic neurotransmission and increased hippocampal neurogenesis. That result is what sent the group looking for a TREK-1 blocker in the first place; the peptide came after the target. The naming chain is worth following because it explains the number in the name. Sortilin's propeptide, PE, is 44 amino acids long. Spadin is PE 12-28, a 17-residue fragment of it, and blocks TREK-1 with an IC50 around 40-60 nM — against roughly 6 micromolar for fluoxetine at the same channel, which is the comparison the original work uses to argue that this is a genuinely selective mechanism rather than an incidental one. PE-22-28 is residues 22-28 of spadin: seven amino acids. It was found by asking what happens to spadin in serum. Incubating spadin with mouse serum at 37 degrees made almost all of it disappear within 30 minutes, leaving two products identified by mass spectrometry as PE 14-25 and PE 12-27. Screening shortened sequences for TREK-1 blockade then produced the shortest one that still worked, and that is PE-22-28. The potency result is the striking one: an IC50 of 0.12 nM at human TREK-1 in patch-clamp on hTREK-1/HEK cells, against spadin's 40-60 nM. Roughly a few hundred-fold more potent than the peptide it came from. One finding from the same paper deserves care, because it is easy to state backwards. Modifying PE-22-28 at its N- or C-terminus maintained or abolished TREK-1 channel ACTIVITY without affecting its AFFINITY. Binding and blocking are separable here — a modified analog can still bind the channel and no longer block it. Any claim that a modified version is simply 'a better PE-22-28' has to say which of those two properties it means.

Researched effects

Everything demonstrated for this peptide was demonstrated in mice or in cultured cells. There is no human evidence of any kind, and no clinical trial of spadin or any spadin analog is indexed in PubMed. In the forced swimming test, mice treated with spadin analogs showed significantly reduced immobility time. The analogs remained active after sub-chronic treatment whether given by intraperitoneal injection at 3.0 mcg/kg or by oral gavage at 1 mg/kg — an oral route working at all is a genuinely unusual result for a seven-residue peptide and is one of the paper's more interesting findings. In the novelty-suppressed feeding test, a 4-day sub-chronic course of PE-22-28 significantly reduced the latency to eat. The 4-day timescale is the point the original spadin work leaned on: conventional antidepressants take three to four weeks to show effect, and this mechanism was proposed to act in days. Neurogenesis was induced after only 4 days of treatment, measured by BrdU labelling in hippocampus, with the most prominent effect from G/A-PE 22-28 rather than from PE-22-28 itself. In cultured mouse cortical neurons, PE-22-28 and its derivatives increased PSD-95 expression, which is the paper's synaptogenesis measure. The duration-of-effect experiment is the one most often misreported, including previously on this page. Half-times of behavioural effect were approximately 14 hours for G/A-PE 22-28 at 3.2 mcg/kg, 17 hours for biotinylated-G/A-PE 22-28 at 4.0 mcg/kg, 21 hours for G/A-PE 22-28 at 32 mcg/kg, and 23 hours for biotinylated-G/A-PE 22-28 at 40 mcg/kg. Every one of those figures belongs to a MODIFIED analog. The 23-hour number that circulates as 'the half-life of PE-22-28' is the longest-lasting doubly modified derivative at the highest dose tested, and it is a half-time of effect in a behavioural assay, not a plasma measurement.

Dosage — reported ranges (overview)

There is no human dose for PE-22-28, and the rodent figures that do exist are smaller and more specific than the versions in circulation. The published sub-chronic behavioural experiments administered spadin analogs at 3.0 mcg/kg intraperitoneally or 1 mg/kg by oral gavage. A separate experiment, measuring how long the effect lasted rather than whether it occurred, used 3.2 and 32 mcg/kg of G/A-PE 22-28 and 4.0 and 40 mcg/kg of the biotinylated analog. This page previously reported 'roughly 3.2-4.0 mcg/kg intraperitoneally' as the dosing figure, which merged the two lowest doses from the duration experiment into a range that appears nowhere in the paper and attributed it to the behavioural work. That is corrected here: the behavioural dose is 3.0 mcg/kg ip, and 3.2 and 4.0 mcg/kg are two separate doses of two different modified analogs in a different experiment. The community figures — 50 to 200 mcg per administration, daily, over cycles of 8 to 16 weeks — have no study behind them at all. It is worth seeing how far they sit from the published work. A 3.0 mcg/kg mouse dose does not translate to a fixed human microgram amount by any principled route, and the longest treatment period anywhere in the literature is four days, against community cycles measured in months. The frequency convention has the same problem as the half-life. Once-daily dosing is usually justified by the 23-hour figure, and that figure describes a doubly modified analog in a behavioural assay rather than PE-22-28 in plasma. None of this is a protocol or a recommendation. It is a description of what was administered, to mice, in the one study that exists.

A printable protocol sheet with a reconstitution reference and an injection log comes with All-Access Lifetime.

Reconstitution — bac-water math

PE-22-28 is supplied lyophilised, most often in 5 mg and 10 mg vials, and dosed in micrograms. That combination makes the dilution choice matter more here than for almost anything else in this library. Concentration is the vial's contents divided by the millilitres of bacteriostatic water added. Units to draw = dose in mcg divided by concentration in mcg/mL, times 100, because 100 units is 1 mL on a U-100 insulin syringe. Work it on a 5 mg vial and the problem is immediate. Five milligrams is 5,000 mcg, so 1 mL gives 5,000 mcg/mL — and a 100 mcg example is then 100 / 5000 x 100 = 2 units. Two units is not a measurement; it is a guess on most insulin syringes, where the smallest reliable graduation is one unit and the meniscus covers a good part of it. The direction that helps is more diluent, not less. At 5 mL the same vial is 1,000 mcg/mL and the same 100 mcg becomes 10 units, which is readable. This is the opposite of the milligram-dosed compounds in this library, where over-dilution pushes a dose past the syringe — here under-dilution pushes it below the scale. This is concentration math. It is not a dose, and the microgram figures used in the table are the community-reported ones described above, which have no study behind them.

This is concentration math, not a dose recommendation.

Injection / administration basics

The published route and the marketed route are not the same, which is unusual enough to state plainly. Every in vivo experiment in the defining paper used intraperitoneal injection or oral gavage in mice. Intraperitoneal injection is a laboratory route in rodents and has no human equivalent in this context. Oral gavage is the more interesting of the two, because the analogs remained active in the forced swim test when given at 1 mg/kg by that route — a seven-residue peptide retaining activity orally is a real finding, and it is the one part of this compound's profile that the market tends not to mention at all. What community material describes instead is subcutaneous injection, sometimes intranasal. Neither route appears in the published work for this peptide, so there is no study establishing a site, a volume, an interval or a bioavailability for either. The practical constraint is the one in the reconstitution section: at microgram doses from a 5 or 10 mg vial, the draw can fall below what a U-100 syringe can resolve unless the vial is diluted well past the volume that feels natural. Route and schedule are reported here as published study parameters in mice. Nothing on this page is an administration instruction.

Half-life & frequency rationale

No plasma half-life has been established for PE-22-28 in any species, and the figure usually quoted in its place is a different quantity measured on a different molecule. The 23 hours that circulates as this peptide's half-life is a half-time of behavioural EFFECT — how long the reduction in forced-swim immobility persisted — and it belongs to biotinylated-G/A-PE 22-28 at 40 mcg/kg, a doubly modified analog at the highest dose in that experiment. The full series is 14 hours for G/A-PE 22-28 at 3.2 mcg/kg, 17 hours for the biotinylated analog at 4.0 mcg/kg, 21 hours for G/A-PE 22-28 at 32 mcg/kg, and 23 hours for the biotinylated analog at 40 mcg/kg. PE-22-28 itself does not have a figure in that series. The source paper is explicit that these two things come apart, and says so in its own words: despite a relatively short in vitro serum half-life, the in vivo antidepressant efficacy lasted for almost 24 hours. A molecule can be gone from serum long before its effect is. Treating the effect duration as a half-life and then using it to justify a dosing interval reverses the paper's own point. What is measured for the parent is that spadin's antidepressant activity in mice disappears beyond 7 hours after acute intraperitoneal administration, and that almost all spadin disappears from mouse serum within 30 minutes of incubation at 37 degrees. If you see a plasma half-life quoted for PE-22-28, the useful question is which species, which assay, and whether the molecule measured carried a modification.

Stacking — overview

Nothing has been combined with PE-22-28 in any published study, so every stack suggestion for this compound is invention rather than extrapolation. The one comparison the literature does support is with conventional antidepressants, and it is a mechanistic contrast rather than a head-to-head. Fluoxetine blocks TREK-1 with an IC50 around 6 micromolar; spadin does so at 40-60 nM and PE-22-28 at 0.12 nM. The original argument for this class is that an SSRI touches this channel only incidentally at concentrations far above those it is used at, and that hitting it deliberately produces an antidepressant-like effect in days rather than weeks in mice. That argument has never been tested in a person.

For how combinations are grouped by research context, the named blends, and why a pre-mixed blend vial cannot be calculated from its total milligrams, see the peptide stacks guide.

Storage & handling

• Lyophilized (powder): store sealed and refrigerated at 2-8°C, protected from light and moisture; long-term storage is commonly at -20°C or below. Do not use past intended research shelf life.
• Reconstituted (liquid): refrigerate at 2-8°C, protect from light, and avoid freeze-thaw cycles and vigorous shaking.
• Peptide stability in serum is not the same thing as vial stability, and this compound's literature makes the point vividly: almost all of its parent peptide disappeared from mouse serum within 30 minutes at 37°C. That says nothing about how a sealed refrigerated vial behaves, and the two are routinely confused.
• Label the vial with the reconstitution date and the resulting concentration in mcg/mL. When a correct draw is between 2 and 20 units depending on dilution, the concentration is the only number that makes the arithmetic recoverable later.

Guide FAQ

Quick answers about guide scope, access, and educational use context.