Carl read the papersWhat it actually is
Ciliary neurotrophic factor is a real human protein with real neuroprotective effects, and it is unusable as a drug: a plasma half-life of roughly three minutes, poor brain access, and severe side effects reported in people, including appetite loss, skeletal muscle loss, muscle pain and cramps.
A group at the New York State Institute for Basic Research worked down from an 11-residue active fragment of CNTF, called Peptide 6, to a four-residue core, then attached an adamantylated glycine to help it cross the blood-brain barrier. The result is the five-unit Ac-DGGL(A)G-NH2, called Peptide 021 or P021. It is small, given by mouth in the studies, and gets into the brain.
How it’s thought to work
- Inhibits LIF signalling — The parent fragment acts by competitively inhibiting leukaemia inhibitory factor signalling, which the authors link to the formation of neural precursor cells.
- Raises BDNF — The main downstream claim: increased brain-derived neurotrophic factor expression, which supports survival and maturation of new neurons.
- Less tau phosphorylation — Via BDNF, reduced activity of GSK-3 beta, a major tau kinase, which is how a neurotrophic compound ends up with an effect on Alzheimer pathology.
- Neurogenesis and synapses — Reported increases in dentate gyrus neurogenesis and in synaptic markers across several rodent models.
What the research says
Substantial and narrow. Across roughly fifteen years the originating laboratory has published P021 work in mouse models of Alzheimer's disease, in Down syndrome model mice, in aged rats and on cognitive ageing, reporting improved learning and memory, more neurogenesis, increased BDNF and reduced tau hyperphosphorylation. One of the longer studies treated triple transgenic Alzheimer model mice via diet for twelve months. That is a real, sustained research programme. Two caveats carry most of the weight. First, the lead investigator holds patents on P021, is a cofounder and chief scientific officer of the company developing it, and declares as much in the papers. Second, and more informative, a group at the University of Bologna tested it independently in 2024 in a model of CDKL5 deficiency disorder. In cells it worked: proliferation, survival and maturation deficits were corrected. In the live animals it did not. The authors report that chronic treatment failed to increase BDNF levels, did not improve the neuroanatomical defects, and produced limited behavioural benefit. There are no human trials of P021.
- Derived from an active region of human CNTF, cut to a four-residue core with an adamantylated glycine added for brain penetration.
- Reported in rodent work to increase neurogenesis, raise BDNF and reduce tau hyperphosphorylation via GSK-3 beta.
- Twelve-month oral treatment in a triple transgenic Alzheimer mouse model reduced abnormal tau and soluble amyloid beta.
- The originating investigator holds patents and is chief scientific officer of the company developing it, declared in the papers.
- An independent 2024 study in Cdkl5 knockout mice found the in-vitro benefit did not carry into live animals.
- No human trials. No registered clinical trial.
Reported protocols
These are the doses the peptide community actually reports running (Reddit & forums), alongside published research — logged for education, not as a recommendation to use.
| Parameter | Commonly reported range |
|---|---|
| Human dose | Not established. P021 has never been given to a person in a published study, so there is no dose, route or schedule. |
| How it was given in the studies | In diet, for months at a time, in mice and rats, at concentrations chosen per gram of feed. That is a laboratory design, not a human regimen, and nothing sensible converts out of it. |
| Why oral is emphasised | Oral administration and brain penetration were the design goals, since the parent protein fails on both. That is an argument for the molecule's design and not evidence of what an oral dose does in a person. |
| Reported online | There is a modest trail of first-hand posts going back to 2015, by nose or under the skin, most of them around a milligram a day and some considerably higher. They include people who rated it among the best things they have tried and at least one who had a bad time and stopped. None of it is measured, and the people posting are not comparing against anything. |
| Carl's position | A compound this early, with one laboratory behind it and a failed independent in-vivo test, does not get a number invented for it here. |
Common use cases
| Goal | Community protocol notes |
|---|---|
| Following the neurotrophic mimetic field | a clean example of shrinking an unusable protein into something drug-like |
| Understanding the BDNF and GSK-3 beta link | how a growth-factor mimetic ends up affecting tau pathology |
| Reading replication honestly | the 2024 independent study is a model of reporting a result that did not come out as hoped |
Mixing it (reconstitution)
Freeze-dried peptides must be reconstituted with bacteriostatic water before they can be measured. Carl’s calculator turns “mg in the vial + ml of water + target dose” into “units on the syringe” — and tells you how long the vial lasts.
→ Open the Reconstitution Calculator
Injection & handling
P021 was designed for oral administration, so injection is not the question people usually arrive with, and there is no administration guide here either way. The sensible summary: the parent protein, CNTF, caused real problems in people when given peripherally, including appetite loss and skeletal muscle loss, which is part of why a mimetic was built in the first place. P021 has not been given to a human in any published study, so it has no safety profile of its own, favourable or otherwise. The independent 2024 result is the most useful thing to hold on to, because it is the only test of this compound by people with no stake in it, and the in-vivo half of it did not work.
Stacks it appears in
- With Semax or Selank — the cognitive shelf default. Both have more human exposure history than P021, which is not the same as having good trials, and no combination work exists.
- With Dihexa — both are small neurotrophic mimetics aiming at synapses from different directions. One has retracted foundational papers, the other has a failed independent in-vivo test. The pair does not add up to evidence.
- With Epithalon — appears in longevity stacks. Different mechanism, different evidence problem, no interaction data.
Sources
- Li B et al. 2010, FEBS Lett. Neurotrophic peptides incorporating adamantane improve learning and memory, promote neurogenesis and synaptic plasticity in mice (PMID 20600002)
- Kazim SF, Iqbal K 2016, Mol Neurodegener. Neurotrophic factor small-molecule mimetics mediated neuroregeneration and synaptic repair (PMID 27400746, PMC4940708)
- Kazim SF et al. 2014, Neurobiol Dis. Disease modifying effect of chronic oral treatment with a neurotrophic peptidergic compound in a triple transgenic mouse model of Alzheimer's disease (PMID 25046994)
- Bolognin S et al. 2014, Neurobiol Aging. Rescue of cognitive-aging by administration of a neurogenic and/or neurotrophic compound (PMID 24702821)
- Mottolese N et al. 2024, J Neurodev Disord. Effects of a CNTF small-molecule peptide mimetic in an in vitro and in vivo model of CDKL5 deficiency disorder (PMID 39592934)
- Harper JW et al. 1993, Cell. The p21 Cdk-interacting protein Cip1 is a potent inhibitor of G1 cyclin-dependent kinases (PMID 8242751)
- el-Deiry WS et al. 1993, Cell. WAF1, a potential mediator of p53 tumor suppression (PMID 8242752)
- ClinicalTrials.gov search for P021
- WADA 2026 Prohibited List, S0 non-approved substances
- SAHPRA, Peptide Products: Public Information (May 2026)
