Dihexa

Dosing from the Literature

Published for the brain: animal doses only: 1.44 and 2.88 mg/kg a day for three months in mice (Sun et al., 2021) and 2–4 mg/kg in rats after nerve repair (Weiss et al., 2021). Not published: any human dose; no trial of dihexa exists.

The rows below are the doses the published animal studies gave. No human dose has been published.

What It Is

Dihexa (developmental code PNB-0408; chemical name N-hexanoic-Tyr-Ile-(6) aminohexanoic amide) is a synthetic hexapeptide-like molecule — formally a peptidomimetic — derived from a two-decade structure-activity-relationship campaign on the C-terminal end of angiotensin IV. It was developed at Washington State University (WSU) by Joseph W. Harding, Ph.D., Jay Wright, Ph.D., and collaborators working on the procognitive properties of angiotensin IV analogs. Its key engineered properties are oral bioavailability, blood-brain-barrier permeability, and metabolic stability — characteristics that earlier angiotensin IV peptides lacked.

The dihexa headline that made it famous in the optimization community is the claim that it is "approximately 10 million times more potent than BDNF" at promoting new synapse formation. That claim originates from in vitro hippocampal-culture data showing measurable spinogenesis at picomolar concentrations of dihexa, compared with the nanomolar concentrations conventionally needed for BDNF. The seven-orders-of-magnitude potency comparison is a specific in vitro readout — not a demonstration of in vivo equivalence — and the foundational papers behind it have a complicated and now-compromised provenance.

Dihexa was licensed from WSU in 2013 to a company originally called M3 Biotechnology (later renamed Athira Pharma). M3 / Athira did not advance dihexa itself into clinical trials. Instead it advanced a related phosphate prodrug called fosgonimeton (development codes NDX-1017 and ATH-1017), positioned as a clinically optimized HGF (a tissue-repair signal)/c-Met (the tissue-repair receptor) positive modulator. Fosgonimeton ran a Phase 1 trial in healthy volunteers and Alzheimer's subjects, the ACT-AD Phase 2 trial (mild-to-moderate AD on standard-of-care AChEIs), the SHAPE Phase 2 trial in Parkinson's disease dementia and dementia with Lewy bodies, and the LIFT-AD Phase 2/3 trial in mild-to-moderate AD. The LIFT-AD trial reported topline results in late 2024 and did not meet its primary endpoint (Global Statistical Test). Athira's stock collapsed; the program was reframed as a "broad neuroprotective mechanism with biomarker signal" rather than a Phase 3 success.

The research record has an integrity problem. In September 2021 the Journal of Pharmacology and Experimental Therapeutics put expressions of concern on four 2011–2014 papers from the Harding and Wright labs at Washington State University, for possible image manipulation; Leen Kawas, then chief executive of Athira, the company spun out to develop the work, co-wrote all four (Retraction Watch, September 24, 2021). In October 2021 an Athira internal investigation found “that Dr. Kawas altered images in her 2011 doctoral dissertation and in at least four research papers that she co-authored while a graduate student at WSU, published from 2011 to 2014” (Retraction Watch, January 7, 2025). In January 2025 Athira agreed to pay about $4 million to settle a whistleblower’s False Claims Act allegations over NIH grant applications that cited the papers (Retraction Watch, January 7, 2025), and in April 2025 the 2014 Benoist et al. paper that set out dihexa’s HGF/c-Met mechanism was retracted (PMID 40312093).

Mechanism of Action

The proposed mechanism — based on the WSU papers, with the caveat that the central mechanism paper (Benoist 2014) has been retracted — is hepatocyte (liver cell) growth factor (HGF) / c-Met receptor positive modulation. The path from angiotensin IV to dihexa to HGF/c-Met evolved through a two-decade research arc.

• Angiotensin IV (AngIV) parent — The Harding group’s 2015 review summarizes rodent work linking AngIV, a fragment of angiotensin II, acting at a site called the AT4 receptor, to learning and memory, and notes that the receptor’s identity is disputed: Albiston and colleagues identified it as insulin-regulated aminopeptidase (IRAP), while the Harding group linked the AngIV/AT4 system to the HGF/c-Met receptor (Wright et al., 2015). That link is the basis of dihexa’s proposed mechanism.
• Modification rationale — Native AngIV has very poor pharmacokinetics: rapid enzymatic degradation, no oral bioavailability, no blood-brain barrier penetration. McCoy et al. 2013 (J Pharmacol Exp Ther; PMID 23055539) screened N- and C-terminally modified Nle-Tyr-Ile analogs for metabolic stability, reversal of scopolamine-induced memory deficits and hippocampal synaptogenesis; the paper carries an expression of concern (J Pharmacol Exp Ther, 2021), and the HGF link came in the 2014 paper, since retracted. The N-hexanoic acyl group plus the 6-aminohexanoic-amide tail produced a metabolically stabilized, orally active, BBB-permeant compound — dihexa.
• HGF binding and c-Met phosphorylation (retracted paper) — Benoist et al. 2014 (J Pharmacol Exp Ther; PMID 25187433; RETRACTED April 2025) reported that dihexa binds with high affinity to HGF and that, in the presence of subthreshold HGF concentrations, dihexa augments HGF-mediated c-Met phosphorylation, downstream signaling, and HGF-dependent cell scattering. The proposal: dihexa is a positive allosteric modulator of HGF/c-Met signaling, not an HGF mimetic and not a direct c-Met agonist. The retraction of this paper is the single most important caveat to this mechanism.
• Synaptogenesis claim (picomolar potency) — The “10 million times more potent than BDNF” line in community writing traces to the WSU work on hippocampal culture; the papers read for this page do not state that comparison, and the one that reported dihexa's spinogenesis and synaptogenesis has been retracted (Benoist et al., 2014, retracted in 2025). No paper read here compares dihexa with BDNF (PubMed, searched October 4, 2026).
• HGF/c-Met biology in CNS — Independent of the dihexa controversy, HGF and its receptor c-Met are expressed throughout the CNS. HGF promotes neuronal survival, neurite outgrowth, synaptogenesis, and possibly neurogenesis. The HGF system has been independently identified by other groups as a plausible therapeutic target in Alzheimer's, Parkinson's, ALS, and stroke. This larger HGF biology is not dependent on the dihexa-specific findings; it provides at least some independent rationale for the therapeutic hypothesis.
• Downstream signaling — In APP/PS1 Alzheimer's mice given dihexa for three months, the PI3K/AKT pathway was activated, inflammation markers fell, and a PI3K inhibitor reversed the anti-inflammatory and anti-apoptotic effects (Sun et al., 2021; PMID 34827486). That group is independent of the WSU laboratory.
• What we do not know — Whether dihexa actually engages HGF/c-Met in vivo at doses humans use, whether it crosses the human BBB at physiologically relevant concentrations, and whether picomolar in vitro potency translates to any clinically meaningful synaptogenesis in the human brain. These questions are unanswered, and the largest clinical attempt to answer them (Athira's fosgonimeton) failed.

What the Research Shows

Dihexa research falls into three buckets: the original WSU preclinical campaign (now partially compromised), independent preclinical work, and the indirect human evidence from fosgonimeton (Athira's clinical analog).

• Foundational dihexa papers (WSU, partly compromised) — McCoy 2013 (PMID 23055539) — orally active dihexa improved performance in scopolamine-amnesia and aged-rat cognitive models; received an expression of concern in 2021. Benoist 2014 (PMID 25187433) — HGF/c-Met mechanism paper; retracted April 2025. Wright & Harding 2015 reviews (PMID 25649658, 25455861) — synthesize the mechanistic story; not retracted but rest on the now-questioned underlying data.
• Preclinical — hair cell preservation — Uribe et al. 2015 (Front Cell Neurosci; PMID 25674052) — in larval zebrafish, 1 µM dihexa protected lateral-line hair cells from neomycin and gentamicin, and an HGF antagonist and Akt, TOR and MEK inhibitors each reduced the protection. Not independent of the WSU group: Kawas and Harding are co-authors.
• Independent preclinical — APP/PS1 Alzheimer's mouse model — Sun et al. 2021 (Brain Sci; PMID 34827486) — dihexa at 1.44 or 2.88 mg/kg a day for three months restored Morris water maze performance, raised neuronal counts and the synaptic protein SYP, and lowered IL-1β and TNF-α, through PI3K/AKT. Independent of the Harding/Kawas group.
• Stem cell / hepatic differentiation — Siller et al. 2015 (Stem Cell Reports; PMID 25937370) describes a growth-factor-free, small-molecule protocol that turns human pluripotent stem cells into hepatocyte-like cells; its abstract does not name dihexa.
• Peripheral nerve repair — In a rat sciatic nerve transection-repair model, dihexa (2–4 mg/kg body weight) was delivered locally by hydrogel, systemically, or into the gastrocnemius muscle across ten treatment groups followed to a 16-week endpoint. Walking-footprint motor grades improved significantly at 8–16 weeks with mesenchymal stem cells plus G-CSF, or mesenchymal stem cells plus dihexa — not with dihexa alone; the authors describe all three as candidate adjunct therapies (Weiss et al., Ann Med Surg 2021; PMID 34703584). Independent group.
• Fosgonimeton Phase 1 (NCT03298672) — 88 subjects: healthy young men, healthy elderly volunteers and 11 people with Alzheimer's disease. Safe and well tolerated, dose-proportional pharmacokinetics with no accumulation over 9 days; sustained gamma-power induction on qEEG, and in the Alzheimer's group a significant shift toward normal ERP P300 latency against placebo (Hua et al., 2022).
• ACT-AD Phase 2 (NCT04491006, fosgonimeton) — Mild-to-moderate AD on standard-of-care acetylcholinesterase inhibitors. Did not meet primary endpoint (ERP P300 latency change) when used adjunctively. Monotherapy subgroup analysis suggested benefit (ADAS-Cog11 change of −3.3 points; ERP P300 latency change of −28 ms). Athira reframed the trial design and pursued LIFT-AD.
• SHAPE Phase 2 (NCT04831281, fosgonimeton) — 28 subjects with Parkinson's disease dementia or dementia with Lewy bodies. Originally designed for ~75 subjects; truncated to 28 to redirect resources to LIFT-AD. Exploratory positive cognitive signal at 40 mg dose; the 70 mg arm showed inconsistent results with high dropout.
• LIFT-AD Phase 2/3 (NCT04488419, fosgonimeton) — the headline failure — The trial did not achieve its primary or secondary endpoints: the Global Statistical Test differed by −0.08 (p=0.70) at week 26 on 40 mg subcutaneous fosgonimeton against placebo, in a primary analysis of 287 people with mild-to-moderate Alzheimer's not taking acetylcholinesterase inhibitors (549 dosed in all). Only plasma p-tau217 changed nominally; serious adverse events were 4.2% against 6.9% on placebo (Porsteinsson et al., 2025).

Human Data

There is no published human trial of dihexa itself. All clinical evidence comes from fosgonimeton (Athira's prodrug analog), which is structurally related but pharmacokinetically distinct.

• NCT03298672 — Fosgonimeton Phase 1 — 88 subjects: healthy young men, healthy elderly volunteers and 11 people with Alzheimer's. Safe and well tolerated, with gamma-power induction on qEEG (Hua et al., 2022).
• NCT04491006 — ACT-AD Phase 2 (fosgonimeton) — 77 patients with mild-to-moderate AD. Failed primary endpoint (ERP P300 latency) in adjunct-therapy population; monotherapy subgroup signal was not powered to be statistically definitive but informed LIFT-AD design.
• NCT04831281 — SHAPE Phase 2 (fosgonimeton in PDD/DLB) — Truncated at 28 of 75 planned subjects. Exploratory positive cognitive signal at 40 mg.
• NCT04488419 — LIFT-AD Phase 2/3 (fosgonimeton) — 554 enrolled, 26-week double-blind, 40 mg fosgonimeton against placebo. Did not meet its primary or secondary endpoints; the Global Statistical Test differed by −0.08 (p=0.70) (Porsteinsson et al., 2025; ClinicalTrials.gov, searched October 4, 2026).
• NCT04886063 — open-label extension (fosgonimeton) — Phase 2/3, 423 enrolled, terminated (ClinicalTrials.gov, searched October 4, 2026).
• None of these studied dihexa itself — Fosgonimeton is a phosphate-modified prodrug with different oral bioavailability, plasma protein binding, BBB permeability, and metabolic profile. Translating fosgonimeton clinical results to dihexa is not straightforward in either direction. Negative fosgonimeton results do not strictly disprove dihexa, but they are the most direct human translation of the HGF/c-Met cognitive hypothesis to date.
• No dihexa trial is registered — ClinicalTrials.gov returns no study of dihexa, and PubMed no human study of it (both searched October 4, 2026).

No study has given dihexa to a person. The closest human translation, fosgonimeton, missed its primary endpoint.

Reconstitution & Storage

Nothing to reconstitute: dihexa is taken by mouth. The animal studies dissolved it for injection or gavage — in DMSO, PEG300, Tween-80 and saline in the mouse study (Sun et al., 2021) and in a hydrogel at the repair site in the rat nerve study (Weiss et al., 2021). No label or guideline gives a strength, storage condition or shelf life for a product sold as dihexa.

→ Calculator — vial-to-syringe math

Side Effects & Risks

No study has recorded side effects of dihexa in people (PubMed and ClinicalTrials.gov, searched October 4, 2026). In the fosgonimeton trials — a different molecule — serious adverse events were 4.2% on the drug and 6.9% on placebo, and 14.2% of participants stopped for adverse events (Porsteinsson et al., 2025; Hua et al., 2022).

• The c-Met question — c-Met is dysregulated in many human cancers. Whether a compound that acts on HGF and c-Met matters for that has not been studied: PubMed holds no study of dihexa and cancer (searched October 4, 2026), and the paper that described dihexa's effect on c-Met has been retracted (Benoist et al., 2014).
• WADA status — Not named on the Prohibited List, but prohibited at all times under S0, non-approved substances, because no government has approved it for human therapeutic use (WADA Prohibited List 2026).
• Long-term safety — No study of any duration has given dihexa to a person; the longest fosgonimeton trial ran 26 weeks (Porsteinsson et al., 2025).

Bloodwork & Monitoring

No trial, label or guideline sets monitoring for dihexa, and no human has been studied (ClinicalTrials.gov and PubMed, searched October 4, 2026). The animal studies measured Morris water maze performance, neuronal counts, the synaptic protein SYP and inflammatory markers (Sun et al., 2021), and walking foot-print grades after nerve repair (Weiss et al., 2021). The fosgonimeton trials measured ADAS-Cog11, ADCS-ADL23, qEEG, ERP P300 latency and plasma neurofilament light and p-tau217 (Hua et al., 2022; Porsteinsson et al., 2025).

Commonly Stacked With

No study has tested dihexa with anything (PubMed, searched October 4, 2026). In rats, one group gave it with mesenchymal stem cells after nerve repair: motor grades improved with the pair, not with dihexa alone (Weiss et al., 2021).

→ Calculator — vial-to-syringe math →

Stacking this? Read the receipt →

Legal Status

Related Compounds

Other cognitive-signal peptides weighed against dihexa:

Next Steps

References

1. McCoy AT, Benoist CC, Wright JW, Kawas LH, Bule-Ghogare JM, Zhu M, Appleyard SM, Wayman GA, Harding JW. Evaluation of metabolically stabilized angiotensin IV analogs as procognitive/antidementia agents. J Pharmacol Exp Ther. 2013 Jan;344(1):141-54. PMID: 23055539. PMCID: PMC3533412. (Foundational dihexa paper. Expression of concern: J Pharmacol Exp Ther. 2021;378(3):313. PMID: 34551989.)
2. Benoist CC, Kawas LH, Zhu M, Tyson KA, Stillmaker L, Appleyard SM, Wright JW, Wayman GA, Harding JW. The procognitive and synaptogenic effects of angiotensin IV-derived peptides are dependent on activation of the hepatocyte growth factor/c-Met system. J Pharmacol Exp Ther. 2014 Nov;351(2):390-402. PMID: 25187433. PMCID: PMC4201273. RETRACTED: J Pharmacol Exp Ther. 2025;392(4):103567. PMID: 40312093. following investigation into image manipulation.
3. Wright JW, Harding JW. The brain hepatocyte growth factor/c-Met receptor system: A new target for the treatment of Alzheimer's disease. J Alzheimers Dis. 2015;45(4):985-1000. PMID: 25649658. (Mechanistic review.)
4. Wright JW, Kawas LH, Harding JW. The development of small molecule angiotensin IV analogs to treat Alzheimer's and Parkinson's diseases. Prog Neurobiol. 2015 Feb;125:26-46. PMID: 25455861. (Comprehensive Wright/Harding review.)
5. Uribe PM, Kawas LH, Harding JW, Coffin AB. Hepatocyte growth factor mimetic protects lateral line hair cells from aminoglycoside exposure. Front Cell Neurosci. 2015 Jan 28;9:3. PMID: 25674052. PMCID: PMC4309183. (Independent dihexa hair-cell preservation study.)
6. Siller R, Greenhough S, Naumovska E, Sullivan GJ. Small-molecule-driven hepatocyte differentiation of human pluripotent stem cells. Stem Cell Reports. 2015 May 12;4(5):939-52. PMID: 25937370.
7. Sun X, Deng Y, Liang J, Lin Y, Song J, Zhao S, Zhuang G, Jia Z. AngIV-Analog Dihexa Rescues Cognitive Impairment and Recovers Memory in the APP/PS1 Mouse via the PI3K/AKT Signaling Pathway. Brain Sci. 2021 Oct 28;11(11):1421. PMID: 34827486. (Independent Chinese group; APP/PS1 model.)
8. Weiss JB, Phillips CJ, Malin EW, Gorantla VS, Withée CE, Doyle A. Stem cell, Granulocyte-Colony Stimulating Factor and/or Dihexa to promote limb function recovery in a rat sciatic nerve damage-repair model: Experimental animal studies. Ann Med Surg (Lond). 2021;71:102917. PMID: 34703584.
9. Hua X, Church K, Walker W, L'Hostis P, Viardot G, Danjou P, Hendrix S, Moebius HJ. Safety, Tolerability, Pharmacokinetics, and Pharmacodynamics of the Positive Modulator of HGF/MET, Fosgonimeton, in Healthy Volunteers and Subjects with Alzheimer's Disease: Randomized, Placebo-Controlled, Double-Blind, Phase I Clinical Trial. J Alzheimers Dis. 2022;86(3):1399-1413. PMID: 35180125. (Fosgonimeton Phase 1 — NCT03298672.)
10. Johnston JL, Reda SM, Setti SE, Taylor RW, Berthiaume AA, Walker WE, Wu W, Moebius HJ, Church KJ. Fosgonimeton, a Novel Positive Modulator of the HGF/MET System, Promotes Neurotrophic and Procognitive Effects in Models of Dementia. Neurotherapeutics. 2023 Mar;20(2):431-451. PMID: 36538176.
11. Porsteinsson AP, Sabbagh M, Tariot PN, Church KJ, et al. Fosgonimeton in mild-to-moderate Alzheimer's disease. J Alzheimers Dis Rep. 2025;9:25424823251405817. PMID: 41393340. (LIFT-AD, NCT04488419: primary and secondary endpoints not met.)
12. Athira Pharma, Inc. Topline Results from ACT-AD Phase 2 Proof of Concept Study of Fosgonimeton in Mild-to-Moderate Alzheimer's Disease. Press release, 2022. (NCT04491006.)
13. Athira Pharma, Inc. Results from SHAPE Phase 2 Clinical Trial of Fosgonimeton in Parkinson's Disease Dementia and Dementia with Lewy Bodies. Press release, December 2023. (NCT04831281.)
14. Retraction Watch. Four papers by Athira CEO earn expressions of concern. September 24, 2021. retractionwatch.com/2021/09/24/four-papers-by-athira-ceo-earn-expressions-of-concern/. Read October 11, 2026.
15. Retraction Watch. Biotech company agrees to pay four million dollars to settle data falsification allegations. January 7, 2025. retractionwatch.com/2025/01/07/biotech-company-athira-settlement-data-falsification-allegations/. Read October 11, 2026.
16. U.S. Department of Justice. Athira Pharma Agrees to Pay Multi-Million-Dollar Sum to Settle False Claims Act Allegations. Press release, January 2025.
17. U.S. Patent — WSU dihexa patent (WO2008/156865) — exclusively licensed to Athira Pharma.
18. FDA. Bulk Drug Substances Nominated for Use in Compounding — 503A and 503B Categories. FDA.gov. Updated 2025–2026.
19. World Anti-Doping Agency. Prohibited List 2026 (in effect January 1, 2026). S0, Non-approved substances. wada-ama.org/sites/default/files/2025-09/2026list_en_final_clean_september_2025.pdf. Read October 3, 2026.

Related Research

• Enclomiphene CitrateSelective Estrogen Receptor Modulator (SERM)
• ZuclomipheneSERM (Selective Estrogen Receptor Modulator)
• BromantaneActoprotector / Adamantane Derivative
• CerebrolysinNeurotrophic Brain Bioregulator
• FDA TrackerDihexa is one of 12 peptides under FDA Pharmacy Compounding Advisory Committee review for the 503A Bulks List.
• FGLNCAM-Derived FGF Receptor Agonist; paired with Dihexa in forum posts
• NoopeptDipeptide-Derived Nootropic (Omberacetam)