Dihexa

Dihexa is a small molecule derived from angiotensin IV, developed at Washington State University with the specific design goal of creating an orally active, blood-brain-barrier-permeant, metabolically stabilized compound built on a minimal three-amino-acid core previously shown, in earlier and separate research, to retain procognitive activity from the larger angiotensin IV peptide. The foundational paper naming and characterizing dihexa, published in 2013, reported that oral dosing reversed scopolamine-induced memory deficits and improved cognition in aged rats, alongside increased hippocampal synapse formation. That paper has not itself been retracted, but it carries an unresolved 2021 notice of concern tied to the same investigation described below, and its specific behavioral findings have not been independently replicated by an unaffiliated laboratory in a separate publication. Dihexa has never been administered to a human being in any published clinical study, and no human safety, dosing, or pharmacokinetic data exist for it under any circumstance.

The single most important fact governing how to read the dihexa literature is that its entire proposed mechanistic explanation, that it works by binding hepatocyte growth factor and activating the c-Met receptor tyrosine kinase, rests on two papers that were formally retracted by the Journal of Pharmacology and Experimental Therapeutics in April 2025. Those retractions followed a 2021 notice of concern and a Washington State University investigation that found a co-author on the papers, Leen Kawas, had altered image data in her 2011 doctoral dissertation and in multiple co-authored papers published between 2011 and 2014. Kawas later became chief executive officer of Athira Pharma, a company that had licensed dihexa-related technology from the university; she resigned from that position in 2021 after the investigation findings became public, and in 2022 Athira Pharma agreed to pay more than 4 million dollars to resolve federal allegations that it had failed to properly disclose the misconduct investigation to investors, according to contemporaneous reporting from the Seattle Times, GeekWire, Fierce Biotech, and Chemistry World. This is not a case of thin or ambiguous evidence; it is a case in which the specific supporting evidence for a mechanistic claim has been formally withdrawn by an academic journal following an institutional finding of data falsification, and it should be described in exactly those terms rather than softened into a general statement about limited research.

Separate from the misconduct findings, the underlying premise that the angiotensin IV (AT4) receptor is c-Met is itself a minority position in the field. An independent line of angiotensin IV research instead identifies the AT4 receptor as insulin-regulated aminopeptidase (IRAP), an enzyme rather than a growth-factor receptor, and has developed IRAP-targeted inhibitors as an unrelated class of cognitive enhancers with their own separate evidence base. This means dihexa proposed molecular target carries genuine, independent scientific controversy that predates and is distinct from the specific retraction findings. On a more encouraging note, a 2021 study from an unaffiliated research group in China reported that dihexa improved cognition and reduced inflammatory markers in an Alzheimer mouse model through PI3K/AKT pathway activation, an independent data point suggesting some behavioral effect of the compound in a rodent disease model even without relying on the retracted c-Met mechanism, though this single study from a single laboratory has not been further replicated.

A chemically distinct but related compound from the same original Washington State University research program, fosgonimeton (marketed under the development code ATH-1017), was carried forward by Athira Pharma into human testing and reached a phase 2/3 randomized, placebo-controlled trial (LIFT-AD, n=315 adults with mild-to-moderate Alzheimer disease). Reported in September 2024, that trial did not meet its primary endpoint or key secondary endpoints, though it found a statistically significant reduction in a blood biomarker of Alzheimer pathology in the active-treatment group. This result concerns a different molecule than dihexa itself and should not be read as a direct verdict on dihexa, but it is the closest thing to human outcome data connected to this research program, and it shows that even the most clinically advanced compound to emerge from it has not yet demonstrated a clinical benefit in a controlled human trial. Dihexa itself remains, and has always remained, an unregulated research chemical with no human trial history, no established human dose, and a scientific foundation now under substantial, independently documented doubt.

Mechanism of Action

Design Rationale and the Retraction of Its Core Mechanistic Claim

Dihexa was designed at Washington State University as a chemically stabilized derivative of a minimal three-amino-acid fragment (Nle-Tyr-Ile) of angiotensin IV, previously shown in earlier, separate research to retain procognitive activity when the full angiotensin IV peptide could not, due to its own rapid degradation and poor barrier permeability. The specific chemical modifications used to build dihexa were intended to increase hydrophobicity and reduce hydrogen bonding, improving both oral absorption and blood-brain barrier penetration relative to the parent tripeptide. This part of the story, the medicinal-chemistry rationale for stabilizing an angiotensin IV fragment, is not in dispute. What is now in serious and formally documented doubt is the specific claim, made in two papers from the same laboratory, that dihexa produces its effects by binding hepatocyte growth factor and activating the c-Met receptor tyrosine kinase. Both of those papers were retracted by the Journal of Pharmacology and Experimental Therapeutics in April 2025, following a 2021 notice of concern and a Washington State University investigation that found a co-author, Leen Kawas, had altered image data in her doctoral dissertation and in multiple papers published between 2011 and 2014. Readers should treat the HGF/c-Met mechanism as withdrawn from the reliable literature, not as an established or even a currently supported explanation for how dihexa works.

The Foundational Behavioral Paper and Its Unresolved Status

The paper that first named dihexa and reported its procognitive effects in rats (McCoy et al., 2013) has not itself been retracted. It reported that oral dihexa reversed scopolamine-induced memory deficits in the Morris water maze, improved spatial memory in aged rats, and increased hippocampal synapse formation. This paper carries an unresolved 2021 notice of concern connected to the same investigation that led to the two retractions described above, meaning its findings remain published but under an active, publicly disclosed cloud of uncertainty. As of this writing, no laboratory unaffiliated with the original Washington State University group has published an independent replication of these specific rat behavioral results.

An Independent Replication Using a Different Proposed Pathway

A 2021 study from an unaffiliated research group in China reported that dihexa improved cognitive performance and reduced neuroinflammatory markers in the APP/PS1 mouse model of Alzheimer disease, attributing the effect to activation of the PI3K/AKT signaling pathway rather than to the retracted HGF/c-Met mechanism. This is a genuinely independent data point, from a laboratory with no connection to the Washington State University group or the data-integrity findings, and it provides some support for the idea that dihexa produces a measurable behavioral effect in a rodent Alzheimer model. It does not, however, validate the specific c-Met mechanism that has been withdrawn, and it remains a single study from a single laboratory that has not itself been further replicated.

A Contested Receptor Target

Independent of the misconduct findings, the underlying premise that the angiotensin IV (AT4) receptor is c-Met is a minority position within the broader angiotensin IV research field. A separate, non-retracted line of research identifies the AT4 receptor instead as insulin-regulated aminopeptidase (IRAP), a zinc-dependent aminopeptidase enzyme rather than a growth-factor receptor, and this research program has developed IRAP-selective inhibitors as an independently evidenced, mechanistically distinct class of cognitive enhancers. This means that even setting aside the specific retracted papers, dihexa proposed molecular target has never been the uncontested consensus view in its own field.

Clinical Evidence

Human Evidence Gap: No Human Trial Has Ever Been Conducted

Dihexa has never been administered to a human participant in any published clinical study of any phase. No human pharmacokinetic, safety, tolerability, dosing, or efficacy data exist for the compound under any circumstance. This is a more absolute statement of the evidence gap than applies to most other research-only peptides in this queue, several of which have at least a small human pilot or open-label study; for dihexa, the evidence stops entirely at the animal level, and the animal-level mechanistic literature underlying it has itself been substantially compromised by the retractions described above. Any human use of dihexa, whether self-administered or otherwise, would represent exposure to a compound with no documented human safety record in the accessible scientific literature.

Fosgonimeton (ATH-1017): A Related Compound That Did Reach Human Trials

A chemically distinct but related compound developed from the same original Washington State University angiotensin IV research program, fosgonimeton (development code ATH-1017), was licensed and advanced by Athira Pharma into human clinical testing. The phase 2/3 LIFT-AD trial randomized 315 adults with mild-to-moderate Alzheimer disease to once-daily subcutaneous fosgonimeton or placebo. Reported in September 2024, the trial did not meet its primary endpoint, a global statistical test combining cognitive (ADAS-Cog11) and functional (ADCS-ADL23) measures, nor its key secondary endpoints, though the active-treatment group showed a statistically significant reduction in plasma p-tau217, a blood biomarker associated with Alzheimer disease pathology. This trial concerns a different molecule than dihexa itself, and its result should not be interpreted as a direct verdict on dihexa. It is nonetheless the most clinically advanced human outcome connected to this overall research program, and it demonstrates that even the compound that did reach human testing has not yet shown a clinical benefit on its primary measures.

Regulatory Status

Dihexa has no regulatory status of any kind: it is not FDA-approved, not approved by any other national regulator, and not in any registered human clinical trial as of this writing. It is sold exclusively as an unregulated research chemical, explicitly labeled for research use only by vendors, with no pharmaceutical manufacturing oversight, no verified identity or purity testing comparable to a regulated drug product, and no oversight of the claims made about it in marketing material. Its underlying scientific literature includes two formally retracted papers and one paper carrying an unresolved notice of concern, a regulatory and evidentiary position that is meaningfully more compromised than a typical unapproved research peptide with simply thin or exclusively preclinical data.

The Research Misconduct Finding and Its Direct Relevance to This Page

This section exists because the research-misconduct finding is not incidental background but directly determines which claims about dihexa can currently be trusted. A Washington State University investigation, reported contemporaneously by the Seattle Times, GeekWire, Fierce Biotech, and Chemistry World, found that Leen Kawas, a graduate student in the laboratory that developed dihexa and later chief executive officer of Athira Pharma, had altered image data in her 2011 doctoral dissertation and in multiple co-authored papers published between 2011 and 2014. Kawas resigned as Athira Pharma chief executive officer in 2021 after the findings became public, and in 2022 Athira Pharma agreed to pay more than 4 million dollars to resolve federal allegations that the company had failed to properly disclose the misconduct investigation to investors. The Journal of Pharmacology and Experimental Therapeutics issued notices of concern for four affected papers in 2021 and formally retracted two of them, the papers specifically describing the hepatocyte growth factor/c-Met mechanism, in April 2025. A reader encountering a confident description of dihexa mechanism of action should understand that the paper making that specific claim has been withdrawn by its publisher following a documented finding of altered data, and should weight any such description accordingly.

Dosing Guidance

There is no scientifically established human dose for dihexa, because it has never been tested in a human in any published study. Doses reported by research-chemical vendors or online community sources are not derived from a peer-reviewed human study and have no verified relationship to a safe or effective human dose. Animal dosing in the surviving (non-retracted) literature used oral administration in rodent memory-impairment models; extrapolating a human dose from rodent data, especially for a compound whose core mechanistic paper has been retracted, is not a scientifically defensible practice, and no dosing guidance beyond this caveat can responsibly be offered.