Dihexa Cognitive Enhancement: Preclinical Evidence

Reviewed by PeptideJournal.org Research Team | Last updated: February 2026

Dihexa made headlines when its developers at Washington State University announced it was ten million times more potent than BDNF at promoting new synaptic connections in cell culture. That single claim — published in a 2012 press release alongside a peer-reviewed paper — transformed an obscure angiotensin IV analog into one of the most discussed experimental nootropics on the internet.

But the story of dihexa is more complicated than the hype suggests. Key papers from the original research group were retracted after a university investigation found falsified data. The clinical-stage derivative, fosgonimeton, failed its Phase 2/3 Alzheimer's trial in 2024. And the same molecular pathway that gives dihexa its brain-boosting potential — HGF/c-Met signaling — is one of the most well-characterized cancer-driving pathways in oncology.

This article walks through what the preclinical research actually shows, where the evidence holds up, where it has fallen apart, and what any of it means for the future of cognitive peptide research.

Table of Contents

What Is Dihexa?
The HGF/c-Met Mechanism
Key Preclinical Studies
The Potency Claim: Seven Orders of Magnitude
The Retraction Problem
Independent Replication: The APP/PS1 Mouse Study
Beyond Cognition: Otoprotection Research
Fosgonimeton: From Lab Bench to Failed Clinical Trial
Safety Concerns and Cancer Risk
How Dihexa Compares to Other Nootropic Peptides
The Bottom Line
References

What Is Dihexa?

Dihexa (N-hexanoic-Tyr-Ile-(6) aminohexanoic amide, also called PNB-0408) is a small synthetic peptide developed by Dr. Joseph Harding and Dr. Jay Wright at Washington State University in Pullman, Washington. It emerged from years of research on angiotensin IV (AngIV), a fragment of the angiotensin hormone system that was long dismissed as biologically inert.

Starting in the 1990s, Harding's lab discovered that AngIV and its analogs could improve learning and memory in animal models. The problem: AngIV broke down quickly in the body and could not cross the blood-brain barrier. After years of structural modifications, they identified that the core procognitive activity resided in just three amino acids — Nle-Tyr-Ile — and that adding hexanoic acid caps to both ends produced a compound that was stable, orally active, and blood-brain barrier permeable. That compound was dihexa.

In rat pharmacokinetic studies, dihexa demonstrated a remarkably long half-life: 12.68 days after intravenous administration and 8.83 days after intraperitoneal injection (McCoy et al., 2013).

The HGF/c-Met Mechanism

Unlike most nootropic compounds that work through neurotransmitter modulation, dihexa operates through the hepatocyte growth factor (HGF) signaling system. HGF is a large protein involved in cell growth, tissue repair, and organ regeneration throughout the body. In the brain, HGF signaling through its receptor (c-Met) plays roles in neurogenesis, dendritic branching, synapse formation, and neuroprotection.

Here is how the mechanism is proposed to work:

Binding to HGF. Dihexa binds to endogenous HGF with high affinity (Kd = 65 pM), specifically at the "hinge" region that controls HGF's ability to dimerize.
Forming an active complex. In the presence of subthreshold levels of HGF — concentrations too low to activate signaling on their own — dihexa synergizes with HGF to form an active dimer capable of engaging the c-Met receptor.
Activating downstream signaling. Once the HGF-dihexa complex binds c-Met, it triggers phosphorylation cascades including PI3K/AKT and MAPK/ERK pathways, which promote cell survival, growth, and synaptic plasticity.
Driving synaptogenesis. The end result is increased dendritic spine formation and new synapse creation — the physical infrastructure of learning and memory.

This mechanism is distinct from semax (which works primarily through BDNF upregulation and melanocortin receptors) and selank (which modulates GABA and monoamine systems). It is also different from cerebrolysin, which contains a mixture of neurotrophic peptides rather than targeting a single pathway.

For a broader comparison, see our Dihexa vs Semax vs Selank nootropic ranking.

Key Preclinical Studies

McCoy et al., 2013 — The Foundational Study

The paper that introduced dihexa to the scientific community was published in the Journal of Pharmacology and Experimental Therapeutics in January 2013 (PMID: 23055539).

What they tested: Researchers evaluated whether dihexa could reverse cognitive deficits in two standard rodent models — scopolamine-induced amnesia in young rats and natural age-related cognitive decline in older rats.

Results in the scopolamine model: Scopolamine blocks acetylcholine receptors and impairs spatial memory. Oral dihexa reversed this impairment on the Morris water maze — a standard test where rats must find a hidden platform in a pool of opaque water. The effect was robust and statistically significant.

Results in aged rats: Dihexa also improved performance in naturally aged rats, though with more variability because not all aged rats develop cognitive decline. Improvements were statistically significant on most test days but less consistent than in the scopolamine model.

Synaptogenesis data: In hippocampal neuron cultures treated for five days, dihexa nearly tripled the number of dendritic spines compared to untreated controls. Immunocytochemistry confirmed these new spines contained proper synaptic machinery, and electrophysiology recordings showed increased frequency of miniature excitatory postsynaptic currents (mEPSCs) — evidence that the new connections were functional.

HGF dependence: When researchers delivered an HGF antagonist (called "Hinge") directly into the brain ventricles, it blocked the procognitive effects of orally delivered dihexa, confirming the HGF pathway was required.

Benoist et al., 2011 — Earlier Synaptogenesis Work

A 2011 study from the same lab (PMID: 21719467) examined C-terminal truncated Nle1-AngIV analogs — the parent compounds from which dihexa was derived. Treatment with Nle1-AngIV produced a 103% increase in average dendritic spine numbers compared to controls (32.4 vs. 16.0 spines per segment). This paper established the synaptogenic potential of the angiotensin IV analog family.

Benoist et al., 2014 — The HGF/c-Met Mechanistic Paper (Now Retracted)

Published in JPET in 2014 (PMID: 25187433), this paper provided the most detailed mechanistic evidence linking dihexa's effects to the HGF/c-Met system. It showed that dihexa bound HGF with high affinity, promoted c-Met phosphorylation, and that both an HGF antagonist and c-Met shRNA blocked dihexa-induced synaptogenesis. This paper was retracted — a critical issue discussed in detail below.

The Potency Claim: Seven Orders of Magnitude

The most widely repeated claim about dihexa is that it is "10 million times more potent than BDNF." This number comes from the Harding lab's own press release accompanying the McCoy et al. 2013 publication, where they stated dihexa was "seven orders of magnitude more powerful than BDNF" in cell culture assays measuring new neuronal connections.

A few important clarifications:

This was an in vitro comparison. The assay measured synapse formation in cultured neurons, not cognitive performance in living animals.
The comparison used effective concentrations, not doses. Dihexa drives spinogenesis at picomolar concentrations (10^-12 M), while BDNF requires nanomolar concentrations (10^-9 M) for similar effects. The "seven orders of magnitude" reflects this concentration difference.
BDNF and dihexa work through entirely different mechanisms. Comparing their potency is like comparing the effective dose of aspirin to caffeine — they operate on different targets.
The number does not mean dihexa is "ten million times better for your brain." It means less compound is needed to trigger synapse formation in a dish.

This distinction matters because the potency claim has been used in marketing contexts in ways that dramatically overstate what the data supports.

The Retraction Problem

Any honest assessment of the dihexa evidence must reckon with the retraction of key papers from the Harding lab.

What happened: In 2021, JPET issued expressions of concern for four papers from the Harding and Wright labs, all co-authored by Leen Kawas during her PhD work at WSU. The university investigated and found that figures and data in multiple papers were "falsified and/or fabricated." Kawas and Harding were found "solely responsible."

Which papers were affected:

Benoist et al. (2014) — the HGF/c-Met mechanistic paper — was formally retracted
A 2012 paper on HGF dimerization domain mimics was also retracted
At least two additional papers received expressions of concern

Why this matters: The Benoist 2014 paper was the primary evidence linking dihexa's cognitive effects to HGF/c-Met-dependent synaptogenesis. Its retraction does not prove the mechanism is wrong, but it removes the most detailed experimental support for it.

The Athira connection: Kawas became CEO of Athira Pharma (originally M3 Biotechnology), the company founded to commercialize dihexa. She was ousted in October 2021 over the data integrity allegations. WSU's dihexa patent incorporates images from the affected papers.

What remains intact: The McCoy et al. 2013 paper — containing the core behavioral data — has not been retracted. Neither has the 2011 Benoist paper on Nle1-AngIV synaptogenesis. The APP/PS1 mouse study (discussed below) was conducted by an independent group in China.

Independent Replication: The APP/PS1 Mouse Study

In 2021, a research group from China published what is currently the most important independent study on dihexa's cognitive effects, using the APP/PS1 transgenic mouse model of Alzheimer's disease (PMC8615599).

APP/PS1 mice carry human gene mutations that produce amyloid plaques, neuroinflammation, and progressive cognitive decline — features that mirror key aspects of Alzheimer's pathology.

Outcome	Result
Spatial learning (Morris water maze)	Restored to near-normal levels
Neuronal cell count (Nissl staining)	Significantly increased vs. untreated APP/PS1 mice
Synaptophysin (SYP) protein	Increased expression, indicating preserved synapses
Astrocyte/microglia activation	Significantly reduced
Pro-inflammatory cytokines (IL-1B, TNF-alpha)	Markedly reduced
Anti-inflammatory cytokine (IL-10)	Increased
PI3K/AKT pathway	Activated in treated mice

The researchers also used wortmannin, a PI3K inhibitor, to confirm that blocking PI3K/AKT reversed dihexa's anti-inflammatory and anti-apoptotic effects — providing mechanistic validation separate from the retracted HGF/c-Met paper, though PI3K/AKT is a known downstream target of c-Met.

This study matters because it was conducted independently of the Harding lab, used a more clinically relevant disease model, and demonstrated both cognitive and anti-inflammatory benefits. It is not a full replication of the retracted mechanistic work, but it is the strongest independent evidence available.

Beyond Cognition: Otoprotection Research

A 2015 study published in Frontiers in Cellular Neuroscience (PMC4309183) tested dihexa in a completely different context: protecting sensory hair cells from aminoglycoside antibiotic damage.

Aminoglycoside antibiotics (gentamicin, neomycin) carry a well-known risk of permanent hearing loss. Using larval zebrafish — whose lateral line hair cells are homologous to mammalian inner ear cells — researchers found:

1 uM dihexa conferred robust protection against both neomycin and gentamicin toxicity
Dihexa did not block aminoglycoside entry into hair cells, meaning protection came from intracellular signaling rather than drug exclusion
The HGF antagonist 6-AH attenuated dihexa's protective effects
Protection required activation of AKT, TOR, and MEK signaling pathways
Dihexa did not protect against cisplatin-induced ototoxicity

This study provides independent evidence that dihexa activates HGF-dependent signaling in a non-neuronal context, partially supporting the proposed mechanism of action even after the retraction of the 2014 mechanistic paper.

Fosgonimeton: From Lab Bench to Failed Clinical Trial

The most direct test of whether dihexa's mechanism translates to humans came through fosgonimeton (ATH-1017), a phosphate prodrug of dihexa developed by Athira Pharma. Fosgonimeton converts to the active dihexa molecule after administration.

Clinical Trial Timeline

Trial	Phase	Population	Year	Outcome
Phase 1	Safety/PK	Healthy volunteers	2021-2022	Safe; well-tolerated
ACT-AD	Phase 2	Mild-moderate AD	2022	Mixed; P300 improvement of -28ms; ADAS-Cog11 improvement of -3.3 points
SHAPE	Phase 2	Parkinson's disease dementia/Lewy body dementia	2022-2023	Primary endpoint not met; post-hoc subgroup showed -7.2 point ADAS-Cog13 improvement (n=5)
LIFT-AD	Phase 2/3	Mild-moderate AD (~315 patients)	2024	Failed. Primary and key secondary endpoints did not reach statistical significance vs. placebo

The LIFT-AD failure was decisive. Neither the composite Global Statistical Test (combining cognition and function measures) nor its individual components reached significance at 26 weeks. Athira attributed part of the failure to a lack of placebo-group decline over the study period, but the result effectively ended fosgonimeton's Alzheimer's program.

By late 2024, Athira had laid off approximately 70% of its workforce. The company paused fosgonimeton development entirely and pivoted toward ATH-1105 (a next-generation oral HGF modulator for ALS) and a licensed breast cancer drug.

For researchers studying peptides for Alzheimer's disease, the fosgonimeton story illustrates a recurring pattern: compounds that show striking results in rodent models of neurodegeneration often fail to translate to meaningful clinical benefit in human trials.

Safety Concerns and Cancer Risk

The most discussed safety issue with dihexa is theoretical but grounded in well-established biology.

The Oncogenic Pathway Problem

HGF/c-Met signaling is not just a neurotrophic pathway. It is one of the most thoroughly characterized oncogenic signaling axes in cancer biology:

HGF and c-Met are overexpressed in many human cancers and correlate with tumor progression, metastasis, and poor prognosis.
The FDA has approved c-Met inhibitors (capmatinib, tepotinib) specifically to treat cancers driven by excessive c-Met signaling.
A 2025 review in Frontiers in Oncology demonstrated that sustained MET overexpression induces malignant transformation in human osteoblast models, producing tumors in immunodeficient mice.

In other words, pharmaceutical companies are spending billions to develop drugs that block the very same pathway that dihexa activates.

What the Safety Data Actually Shows

Short-duration animal safety studies conducted for patent purposes reported no apparent toxicity or neoplastic induction. The patent literature notes this is "unsurprising since oncogenesis requires multiple mutations including both oncogene induction and tumor suppressor attenuation." However:

No long-term carcinogenicity studies have ever been conducted
No chronic exposure studies exist in any species
Maximum tolerated dose has not been established
Human safety data is limited to fosgonimeton trials, which ran a maximum of 26 weeks

The Alzheimer's Drug Discovery Foundation concluded that dihexa could "theoretically promote tumorigenesis and cancer" and that no studies have evaluated long-term safety or tumor-promoting potential.

Regulatory Status

Dihexa has never been approved by any regulatory body for any indication. It is banned by WADA and not approved as a dietary supplement. Any human use outside of clinical trials is unsupervised self-experimentation with an uncharacterized compound.

How Dihexa Compares to Other Nootropic Peptides

For readers exploring peptides for cognitive enhancement or peptides for memory, here is how dihexa compares:

Peptide	Mechanism	Human Trial Data	Safety Profile	Evidence Strength
Dihexa	HGF/c-Met synaptogenesis	No direct trials (fosgonimeton failed Phase 2/3)	Unknown long-term; theoretical cancer risk	Weakened by retractions
Semax	BDNF upregulation, melanocortin	Approved in Russia; limited Western RCTs	Generally well-tolerated	Moderate (regional approval)
Selank	GABA modulation, enkephalin	Approved in Russia; limited Western data	Good short-term profile	Moderate (regional approval)
Cerebrolysin	Multi-target neurotrophic mix	Multiple RCTs in stroke/dementia	Well-characterized	Strongest in class
P21	CNTF-derived neurogenesis	Preclinical only	Unknown	Early-stage

Dihexa has the most dramatic cell culture data but the weakest overall evidence package once you account for the retractions, absence of human data, and the fosgonimeton failure. Cerebrolysin holds the strongest position with multiple randomized controlled trials.

The Bottom Line

Dihexa occupies a strange position in neuroscience. The basic science is genuinely interesting: a small, orally bioavailable peptide that crosses the blood-brain barrier and drives synapse formation through the HGF/c-Met pathway at remarkably low concentrations. Independent replication in APP/PS1 mice and otoprotection models partially supports the mechanism.

But the evidence has serious problems. The most detailed mechanistic paper was retracted for data fabrication. The clinical-stage derivative failed its Phase 2/3 Alzheimer's trial. No long-term safety data exists, and the compound activates a well-known oncogenic pathway.

For researchers, dihexa remains a useful pharmacological tool for studying HGF/c-Met signaling in the nervous system. The pathway itself is real and important — Athira continues developing next-generation HGF modulators (ATH-1105) for ALS.

For anyone considering dihexa as a cognitive enhancer, the math does not work in your favor. You would be using a compound whose key mechanistic data was fabricated, whose clinical derivative failed, whose cancer risk profile is completely uncharacterized, and whose long-term effects in any species are unknown. That is not a calculated risk. It is a blind one.

The peptide research field has better-characterized options for cognitive enhancement with more transparent safety profiles. Start there.

References

McCoy AT, Benoist CC, Wright JW, et al. Evaluation of metabolically stabilized angiotensin IV analogs as procognitive/antidementia agents. J Pharmacol Exp Ther. 2013;344(1):141-154. PubMed | PMC3533412
Benoist CC, Wright JW, Zhu M, et al. Facilitation of hippocampal synaptogenesis and spatial memory by C-terminal truncated Nle1-angiotensin IV analogs. J Pharmacol Exp Ther. 2011;339(1):35-44. PubMed | PMC3186286
Benoist CC, Kawas LH, Zhu M, et al. 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;351(2):390-402. [RETRACTED] PubMed
Wang J, Zhang Z, Chen J, et al. AngIV-analog dihexa rescues cognitive impairment and recovers memory in the APP/PS1 mouse via the PI3K/AKT signaling pathway. Brain Sci. 2021;11(11):1487. PubMed | PMC8615599
Bhatt KA, Bhatt SM, Li X, et al. Hepatocyte growth factor mimetic protects lateral line hair cells from aminoglycoside exposure. Front Cell Neurosci. 2015;9:3. PubMed | PMC4309183
Athira Pharma. Topline results from Phase 2/3 LIFT-AD clinical trial of fosgonimeton for mild-to-moderate Alzheimer's disease. September 2024. Press Release
Retraction Watch. Four papers by Athira CEO earn expressions of concern. September 2021. Article
JPET. Retraction: The procognitive and synaptogenic effects of angiotensin IV-derived peptides are dependent on activation of the hepatocyte growth factor/c-Met system. Retraction Notice
Alzheimer's Drug Discovery Foundation. Cognitive Vitality Report: Dihexa. ADDF Report
Athira Pharma. Phase 2 SHAPE trial results. 2024. Press Release