Dihexa vs P21.

Dihexa (N-hexanoic-Tyr-Ile-(6)-aminohexanoic amide; also referenced as PNB-0408) is a small, orally active peptidomimetic derived from angiotensin IV (Ang IV), a C-terminal fragment of the renin-angiotensin system. It was designed at Washington State University by chemically modifying Ang IV to increase lipophilicity and metabolic stability, yielding a brain-penetrant molecule investigated as a procognitive / antidementia research agent. It is a research chemical, not a drug; it has never been an approved medicine.

P21 (also written P021) is a small synthetic peptide derivative modeled on a biologically active region of ciliary neurotrophic factor (CNTF). It was designed by researchers at the New York State Institute for Basic Research to be an orally active, blood-brain-barrier-penetrant neurotrophic compound. It has been investigated in animal models of Alzheimer's disease, Down syndrome and other neurodegenerative and neurodevelopmental conditions. It is an early-stage research compound that has not been tested in humans.

Dihexa was developed from Ang IV, whose procognitive effects were originally studied at the AT4/IRAP site, but its proposed primary mechanism is potentiation of hepatocyte growth factor (HGF) signaling through its receptor tyrosine kinase c-Met, which drives dendritic spine formation (spinogenesis) and synaptogenesis in hippocampal neurons. Activating HGF/c-Met is thought to remodel synaptic connectivity rather than act as a classical neurotransmitter. Important caveat: the central biochemical evidence that dihexa works by binding HGF and augmenting HGF/c-Met activation came from a paper that has since been retracted for data fabrication, so the molecular mechanism should be treated as unproven. The downstream behavioral phenotype (improved spatial learning in rodents) was reported in separate, non-retracted work.

P21 is reported to act partly by inhibiting the leukemia inhibitory factor (LIF) signaling pathway and by increasing transcription of brain-derived neurotrophic factor (BDNF). Elevated BDNF is proposed to enhance neurogenesis and synaptic plasticity and to reduce the activity of GSK-3 beta, an enzyme that drives abnormal tau phosphorylation. Through this pathway it is hypothesized to have a disease-modifying effect on tau-related pathology. These mechanisms are drawn from cell-culture and rodent studies.

There is NO human data on dihexa: no completed clinical trials, no published human pharmacokinetics or safety. The non-retracted foundational study (McCoy et al., J Pharmacol Exp Ther 2013, PMID 23055539) reported that orally administered dihexa reversed scopolamine-induced learning deficits and improved Morris water maze performance in aged (24-month) rats, and induced spinogenesis in cultured hippocampal neurons at picomolar concentrations. Later angiotensin-IV-analog work continued in disease models (e.g., a 2024 study of an Ang IV analog in a 3-nitropropionic-acid Huntington's-like rat model, PMID 38489193). Crucially, the most-cited mechanistic paper tying dihexa's cognitive effects to HGF/c-Met (Benoist et al., J Pharmacol Exp Ther 2014, PMID 25187433) was RETRACTED in 2025 (retraction notice PMID 40312093) after a Washington State University investigation found falsified/fabricated figure data; the widely repeated "thousands of times more potent than BDNF" claim derives from this now-discredited line of work and should not be cited as established fact. The honest summary: rodent behavioral evidence exists, but the headline mechanistic and potency claims rest partly on retracted literature, and human efficacy is entirely unestablished.

The evidence for P21 is entirely preclinical and comes largely from a single research group. In triple-transgenic Alzheimer's (3xTg-AD) mice, chronic oral P021 reduced tau hyperphosphorylation and rescued neurogenesis, synaptic markers and cognition (Kazim et al., Neurobiology of Disease 2014). A later study reported that early P021 treatment prevented dendritic and synaptic deficits and cognitive impairment in the same model (Baazaoui and Iqbal, Alzheimer's Research and Therapy 2017). Subsequent work from the same laboratory extended the approach to treatment begun in early postnatal development, again in a transgenic rodent model, and reported prevention of Alzheimer-like behavior and synaptic dysfunction (Journal of Alzheimer's Disease, 2021). A later review by the same investigators frames the compound as a therapeutic opportunity to be tested rather than an established treatment (Biomolecules, 2022). The design of this body of work sets clear limits on what it can support. These are rodent experiments in genetically engineered models that reproduce selected features of human Alzheimer's pathology and that have a long record of poor translation to the clinic. They were conducted by the originating institution rather than by independent replicating laboratories, and the published reports are academic studies, not blinded, multi-site, pre-registered confirmatory trials. The outcomes are surrogate measures: phosphorylated tau, synaptic protein density, neurogenesis markers, and rodent behavioral tasks. None is a clinical endpoint, and treatment durations span weeks to months of animal life rather than years of human disease. What is unknown is more substantial than what has been shown. No human clinical trial of P21 has been completed or registered. There is therefore no Phase 1 dataset, no human pharmacokinetic or bioavailability profile despite the oral-activity claim, no characterized exposure range, and no published formal toxicology package. The contrast with the class it is compared against is stark: approved central nervous system drugs for Alzheimer's disease have moved through sequential Phase 1, 2 and 3 programs enrolling thousands of participants with adjudicated cognitive and functional endpoints, and even candidates that failed later usually established a human tolerability and exposure baseline in Phase 1. P21 has not reached that first step. Positive rodent findings do not establish efficacy or safety in people.

No human safety data exist; there is no published clinical adverse-event profile, and dihexa is sold only as a research chemical of variable purity. The principal theoretical concern follows directly from its proposed mechanism: c-Met is a validated oncology target because aberrant HGF/c-Met signaling promotes tumor proliferation, invasion, angiogenesis, and metastasis, so a chronic c-Met potentiator raises an unresolved oncogenic-risk question that no human study has addressed. Its reported long circulating half-life and high lipophilicity also mean tissue accumulation and off-target effects are poorly characterized. Because foundational mechanistic data were retracted for fabrication, even the basic biology underlying any risk assessment is uncertain.

There are no human safety data for P21; all safety information comes from short- to medium-term rodent studies, where it was reported to be tolerated. Long-term effects, appropriate exposure and human toxicology are unknown. No published Phase 1 study has characterized adverse events, dose-limiting toxicity, immunogenicity, or interactions with other medicines in people, and there is no public repeat-dose toxicology, reproductive toxicity, or carcinogenicity dataset of the kind regulators expect before first-in-human testing. Because the proposed mechanism involves raising brain-derived neurotrophic factor and modulating leukemia inhibitory factor signaling, pathways that influence cell growth, survival and inflammation in many tissues, chronic systemic effects cannot be excluded on the basis of rodent behavioral studies alone. Material sold as a research chemical is not quality-controlled for purity or identity. Products of this type are produced outside pharmaceutical good manufacturing practice oversight, are not reliably tested batch by batch for peptide content, related-substance impurities, endotoxin, or residual synthesis reagents, and are labeled for laboratory use rather than administration. A purchaser has no practical way to verify what a vial contains, and the sterility of any reconstituted preparation is unverified. The absence of documented adverse events should not be read as evidence of safety. It reflects the absence of any human exposure under systematic observation, not a record of uneventful use: with no clinical monitoring, no adverse-event reporting channel and no registry, harms would simply go unrecorded. It is not a medicine and is not intended for human use.

Dihexa is investigational/research-use-only: it is not approved by the FDA or any major regulator for any indication, has no DailyMed monograph, and has not completed human clinical trials. It is not a dietary supplement and is widely sold only as a "research chemical." It is not a WADA-listed named substance, though non-approved experimental compounds are broadly disallowed in sport.

P21 is not approved by the FDA or any other regulatory authority and has no approved medical use. It exists only as a preclinical research compound. This content is educational only and contains no dosing guidance.