Dihexa is a lipidized two-residue analog of angiotensin IV, and the mechanism most often attributed to it is potentiation of hepatocyte growth factor signaling at the c-Met receptor tyrosine kinase rather than action at any classical neurotransmitter receptor. That account now rests on unstable ground: the two Journal of Pharmacology and Experimental Therapeutics papers that established the binding measurement and the c-Met dependence were retracted in April 2025 after a Washington State University investigation found falsified or fabricated figure data. Everything published sits at the mechanism and rodent levels, with no controlled human trials and no marketing approval in any jurisdiction.
The mechanism proposed for Dihexa is potentiation of hepatocyte growth factor signaling through c-Met, and the 2012 and 2014 papers that established that mechanism were retracted in April 2025 for falsified or fabricated data.
The formal name carries the entire design: N-hexanoic-Tyr-Ile-(6) aminohexanoic amide is a tyrosine and isoleucine pair with a six-carbon chain fixed to each end. Those caps do two jobs at once, blocking the exopeptidases that chew a short peptide inward from its termini, and pushing the calculated partition coefficient far enough toward lipophilic that the compound is described as passively blood-brain barrier permeable. The permeability claim rests on physicochemical reasoning and functional readouts rather than a published pharmacokinetic dataset, since quantitative brain-to-plasma exposure data in the literature remain thin.
Acylating the amino terminus and amidating the carboxyl terminus strips out the free alpha-amino group and free carboxylate that aminopeptidases and carboxypeptidases require, which is the structural basis for the compound's reported metabolic stability.
Hepatocyte growth factor signaling is a developmental system the adult brain never retired, and in hippocampal and cortical neurons the pathway has been associated with neurite outgrowth, dendritic arborization, synapse formation, and protection against excitotoxic injury. The feature that made it attractive as a drug target is that productive receptor activation depends on the ligand oligomerizing first, a gate a small molecule might in principle influence.
The mechanistic case for this pathway was dependency evidence rather than direct evidence, and the 2014 paper reporting c-Met dependence was retracted in April 2025 for falsified or fabricated figure data.
The proposed interaction is with the ligand, not the receptor, which is why the compound is labeled a potentiator rather than an agonist. The reported picomolar dissociation constant for hepatocyte growth factor came from the 2012 Journal of Pharmacology and Experimental Therapeutics paper retracted in April 2025, with two of its authors held solely responsible for falsified or fabricated figure data.
With both foundational papers retracted for data falsification and the picomolar binding measurement never independently replicated, the ligand-potentiation model stands as an unverified hypothesis rather than a supported mechanism.
Phosphorylation of the activation loop turns carboxy-terminal tyrosines 1349 and 1356 into a docking platform for GRB2, GAB1, SHC, SRC, and the p85 regulatory subunit of PI3K, with GAB1 acting as the scaffold that amplifies and sustains the signal. Two arms carry most of the traffic, and both are mapped far better in epithelial and cancer biology than in the neurons that matter for a cognition claim. In the published work on this compound, the necessity of each arm was inferred from small-molecule inhibitor experiments rather than genetic knockdown.
Because CBL-mediated internalization and phosphatase resetting terminate c-Met signaling on their own schedule, potentiating the ligand is not the same thing as producing a continuously elevated biochemical signal.
Spine counts are where the biochemistry becomes visible under a microscope, and they are also where the evidence is easiest to overread. More spines are not automatically better spines, and a structural count in culture sits a long way from a durable cognitive change in a person.
The reported increases in hippocampal spine density and synaptic marker colocalization at picomolar to nanomolar concentrations rest substantially on the 2014 paper retracted in April 2025, so the quoted concentrations trace to a withdrawn source.
Angiotensin IV has improved rodent memory performance for decades through a binding site first called AT4 and later identified as insulin-regulated aminopeptidase, a zinc metallopeptidase also known as oxytocinase. Truncating and lipidizing that hexapeptide produced compounds whose procognitive activity held or rose while their relationship to enzyme inhibition became less proportional, which is what moved the field toward the growth factor account. That move was a reinterpretation rather than a refutation.
The insulin-regulated aminopeptidase account of angiotensin IV analog activity has been reinterpreted rather than excluded, leaving the mechanism attributed to Dihexa an open research question rather than settled pharmacology.
The evidence base is narrower than the confident tone of most summaries suggests, since the binding claim, the c-Met dependence, the spinogenesis, and the rodent memory results come predominantly from one academic laboratory and its collaborators across a modest number of papers published from roughly 2012 onward, with a patent estate and a small development company attached. Two of those papers were retracted in April 2025 after a Washington State University investigation found falsified or fabricated data in their figures and in a later erratum submission, with two authors held solely responsible; both had carried a notice of concern since 2021. The claims that most outrun the record are those asserting a specific magnitude of human synaptic or cognitive benefit, or describing the mechanism as established rather than proposed.
Settling the mechanism would require independent biophysical confirmation of the ligand interaction, genetic rather than pharmacological removal of the receptor, outside replication of the behavioral work, and adequately powered human studies with pharmacokinetic and safety monitoring.
Any mechanism that amplifies signaling through MET inherits that receptor's oncology history, since amplification, activating mutation, or overexpression of MET is an established driver in subsets of gastric, lung, renal, and hepatic cancers, and most pharmaceutical effort around this receptor has gone into inhibiting it rather than enhancing it. Fairness cuts both ways, because a potentiator's ceiling is set by how much endogenous ligand is present and by the normal negative feedback of receptor internalization and degradation, so the concern is theoretical rather than demonstrated. What keeps it unresolved is an absence of data, not a record of reported harm.
MET is a proto-oncogene whose overactivation drives subsets of gastric, lung, renal, and hepatic cancers, and no chronic toxicology, carcinogenicity assessment, or human safety data has been published for a compound proposed to amplify its signaling.
Educational use only. This article describes what the published scientific and clinical literature reports about Dihexa. It is not medical advice, and it does not recommend, prescribe, or tell anyone to use anything described here. The regulatory status shown at the top of this page reflects what the record showed on the date given there and can change. mdpep.com does not sell any substance described here, does not endorse human use of it, and does not direct anyone to obtain it.
This is not guidance for your situation. Nothing here accounts for your medical history, your current medications, or anything else specific to you, and none of it should be used to make a decision about your own health.
Affiliate disclosure. Some links on this site are affiliate links, and mdpep.com may earn a commission when they are used. That never affects what is written here, it is not an endorsement of any vendor, and it is not a statement that anything described on this page is available to buy.
Every claim here ties to a named primary source and a date, written as plain documentation with nothing for sale. That is how MD PEP covers the whole peptide market.
