Quick answer: Dihexa is an angiotensin IV-derived peptidomimetic studied as a small-molecule research compound. Its proposed HGF/MET mechanism rests partly on a literature chain that now includes retractions and an expression of concern, so it is not settled. Remaining cell and animal studies report model-specific positive and negative findings, but they do not establish human safety or efficacy.
This Dihexa research guide separates three questions that are often compressed into one: what the compound is, what established HGF/MET biology shows, and what Dihexa-specific experiments support. That separation changes the conclusion. Chemical identity is well documented, but direct compound-to-pathway claims carry low confidence. The unaffected preclinical literature is small, method-dependent, and directionally mixed.
Apex-specific research boundary: Apex Laboratory materials are research-grade chemical reagents supplied for in-vitro and preclinical research only. They are not FDA-approved or EMA-authorized pharmaceuticals and are not for human therapeutic use.
What Is Dihexa?
Dihexa, also indexed as PNB-0408, is a defined peptidomimetic that arose from an angiotensin IV analog research program. The common search phrase “Dihexa peptide” is understandable because of that lineage, but it is not the most precise chemical description. PubChem records a small molecule with a specific formula and structure rather than an intact angiotensin peptide.
| Field | Record | Interpretation boundary |
|---|---|---|
| Preferred record name | N-(1-Oxohexyl)-L-tyrosyl-N-(6-amino-6-oxohexyl)-L-isoleucinamide | A defined substance identity, not a use claim. |
| Synonyms | Dihexa; PNB-0408; N-hexanoic-Tyr-Ile-(6)-aminohexanoic amide | Synonyms aid record matching; they do not make related products interchangeable. |
| Molecular formula | C27H44N4O5 | Identity field from PubChem and FDA GSRS. |
| Molecular weight | 504.7 g/mol | PubChem computed value. |
| CAS RN | 1401708-83-5 | Registry mapping, not regulatory approval. |
| PubChem / UNII | CID 129010512 / 9WYX65A5C2 | A UNII identifies a substance; it does not imply FDA review or approval. |
The identity fields above come from PubChem CID 129010512 and the FDA Global Substance Registration System. Neither database establishes a biological result, clinical status, or product specification. A chemical record answers “which substance?”; an experiment and its materials documentation must answer every later question.
Identity firewall. Dihexa, Dihexa acetate, and fosgonimeton/ATH-1017 are not interchangeable evidence objects. FDA’s current compounding discussion names Dihexa acetate, while fosgonimeton is a distinct investigational prodrug associated with a related metabolite identity. Findings, exposure data, formulations, and regulatory status cannot be transferred across those materials or to an Apex research product.
From Angiotensin IV Analogs to Dihexa
The relevant lineage began with work on stabilized analogs of angiotensin IV, a short peptide fragment. Researchers altered that scaffold to study how smaller, more metabolically stable molecules behaved in experimental systems. Dihexa emerged from this program as a compact AngIV-derived peptidomimetic. That history explains the compound’s name and the continued use of peptide language, but it does not allow every AngIV observation to be assigned to Dihexa.
Yamamoto et al. (2010) examined norleual, not Dihexa, and proposed that the AngIV analog could inhibit HGF/c-Met activity in its experimental systems. The current PubMed record is corrected rather than retracted. Its proper role here is historical: it shows how the originating program connected selected AngIV analogs with HGF/MET biology. It does not establish Dihexa’s direction of effect or target engagement, and no norleual potency value transfers to Dihexa.
Wright et al. (2015) summarize the program’s development narrative and proposed model. The review remains indexed and unretracted, but it was written by researchers from the originating program, predates later integrity actions, and is a secondary source. As a narrative review it reports no quantitative endpoint of its own; the figures it recounts belong to primary papers, some later retracted. It can document lineage and how the hypothesis was framed; it cannot independently verify the primary mechanism claims it reviews.
Readers exploring adjacent compounds can use the broader Research Library as a cluster map. The Cerebrolysin evidence guide and Adamax research guide are separate evidence resources, not substitutes or equivalents for Dihexa.
What the Foundational HGF/MET Papers Established
Hepatocyte growth factor, abbreviated HGF, is a ligand for the MET receptor tyrosine kinase. That receptor-level relationship is well established independently of Dihexa. Bottaro et al. (1991) identified the c-met proto-oncogene product as the cell-surface receptor for HGF in Science, immunoblotting a 145 kDa tyrosine-phosphorylated protein induced by HGF in intact target cells and confirming it by cross-linking radiolabelled HGF to proteins of the same size (in-vitro, cultured cells).
Two later PNAS papers refined the pathway background. Hartmann et al. (1992) mapped HGF/scatter-factor activity to the N-terminus plus the first two kringle domains: a single Arg-494-to-Gln substitution at the protease cleavage site left the molecule largely inactive, while the separately expressed heavy chain still bound c-Met, drove receptor autophosphorylation and scattered epithelial cells without inducing mitogenesis (in-vitro, transiently expressed constructs). Gherardi et al. (2003) mapped the MET ectodomain, placing the HGF/SF- and heparin-binding site across the alpha-chain (residues 25-307) and the first 212 residues of the beta-chain, and resolved a stable HGF/SF-heparin-MET complex of 1:1:1 stoichiometry (in-vitro, recombinant ectodomain). Both are important receptor papers, but structural mapping studies rather than treatment experiments: each reports no quantitative endpoint of the kind an intervention study produces. Neither studied Dihexa.
PNAS distinction: A fresh PubMed review found no Dihexa-specific PNAS paper. Foundational PNAS studies mapped HGF-MET binding and receptor structure, but the compound-specific mechanism belongs to a separate and less secure evidence chain.
This distinction prevents a common category error. Established ligand-receptor biology can make a proposed compound mechanism biologically plausible, but it cannot demonstrate that the compound binds the ligand, engages the receptor, or causes a particular downstream result. Those are separate experimental claims that require unaffected Dihexa-specific evidence.
Why the Dihexa HGF/MET Mechanism Remains Proposed
The careful answer to “What is the Dihexa mechanism?” is that HGF/MET involvement has been hypothesized, not securely established. The direct sequence often repeated online—compound binding to HGF, facilitation of MET signaling, then a synaptogenic outcome—depends heavily on an early literature chain now affected by retractions and an expression of concern. Repeating that sequence without its current status would turn a contested model into a fact.
The diagram therefore uses dashed arrows between Dihexa, proposed HGF modulation, and proposed MET involvement. It deliberately omits an affinity value, docking pose, solid binding symbol, and direct activation arrow because the unaffected source set does not establish them. The solid lane is reserved for the separate HGF-to-MET relationship demonstrated by foundational receptor research.
A later APP/PS1 mouse study reported PI3K/AKT-associated findings, but downstream-pathway association does not close the missing target-engagement step. PI3K/AKT can sit downstream of many inputs, and the experiment did not independently demonstrate direct HGF binding or MET engagement by Dihexa. The finding should remain attached to its model, measured endpoints, and intervention design.
Proposed and materially compromised
Identity and HGF/MET pathway background are high-confidence. The direct Dihexa-to-HGF/MET mechanism is low-confidence because central direct and adjacent records were affected. An unaffected independent binding or target-engagement study would be needed to restore that missing step.
What Unaffected Dihexa Studies Report
The remaining literature does not form one steadily strengthening proof chain. It spans different organisms, tissues, endpoints, research groups, and multi-component protocols. A useful evidence map must therefore show directness, direction, provenance, and present integrity status—not merely count papers.
| Evidence layer | Reported observation, as indexed | Current interpretation | Source |
|---|---|---|---|
| 2015 zebrafish lateral-line model | Dihexa protected lateral-line hair cells against neomycin and gentamicin in larval zebrafish (in-vivo, non-mammalian), with 1 µM the optimal protective concentration. Protection was attenuated by the HGF antagonist 6-AH and partially attenuated by Akt, TOR and MEK inhibitors, while pretreatment did not reduce uptake of fluorescently tagged gentamicin. | Direct experiment, but same-program provenance and dependence on the affected mechanism framework limit confidence. It does not establish human relevance or definitive MET engagement. | Uribe et al. (2015), PMID 25674052 |
| 2015 and 2022 hepatic differentiation systems | Dihexa is one component of growth-factor-free small-molecule protocols for human pluripotent stem cells (in-vitro); in the 2022 protocol it is one third of the VDF cocktail (vitamin C, Dihexa, forskolin) used at the hepatic-specification step. Neither record reports a quantitative endpoint isolating Dihexa. | Supports a broader in-vitro research-use record. It does not isolate Dihexa’s contribution or support a CNS mechanism. | Siller et al. (2015), PMID 25937370; Pan et al. (2022), PMID 35410439 |
| 2018 non-human systematic review | Screened 450 records and included n=32 experimental animal studies of AngIV and angiotensin-(1-7). Eight of nine cognitive-deficit studies reported that AngIV analogs (Nle1-AngIV, Dihexa, LVV-hemorphin-7) improved spatial working memory or passive avoidance; seven of 11 normal-animal studies reported benefit for Ang IV itself on passive or conditioned avoidance and object recognition, not for its analogs (non-human models only). | Useful for evidence-base scope, but it predates the later integrity actions and cannot rehabilitate affected primary evidence. | Ho & Nation (2018), PMID 29733881 |
| 2021 APP/PS1 mouse study | In APP/PS1 transgenic mice (in-vivo), oral Dihexa was reported to raise tissue AngIV, restore Morris water maze spatial learning, increase neuronal cell counts and SYP expression on Nissl staining, reduce astrocyte and microglial activation, lower IL-1 beta and TNF-alpha while raising IL-10, and activate PI3K/AKT; the PI3K inhibitor wortmannin reversed the anti-inflammatory and anti-apoptotic effects. The indexed record states direction rather than magnitude: it reports no quantitative endpoint, so no effect size is quoted. | Core positive, model-specific evidence from a separate research team. It does not independently show direct HGF binding, MET engagement, or a human outcome. | Sun et al. (2021), PMID 34827486 |
| 2024 3-NP rat study | n=40 male Wistar rats (in-vivo) randomized to vehicle, 3-nitropropionic acid, or 3-NP plus PNB-0408, with body weight, motor function and cognition measured over 5 weeks before histopathology. 3-NP reduced weight gain and produced spatial-learning and motor deficits; PNB-0408 did not protect against them. | Core negative, model-specific evidence. It limits broad generalization without proving inactivity in every experimental system. | Wells et al. (2024), PMID 38489193 |
The 2021 positive result is model-specific
Sun and colleagues studied APP/PS1 mice, a transgenic model with defined experimental endpoints. Their paper reported directional changes across behavioral and molecular measures and linked selected observations with PI3K/AKT signaling. This is a legitimate positive entry in the evidence map. Its strongest conclusion is that the reported changes occurred under that study’s conditions—not that Dihexa has a settled receptor mechanism or a demonstrated effect in people.
The 2024 negative result carries equal weight
Wells and colleagues evaluated PNB-0408 in a 3-NP rat model and reported no protection against the measured model deficits. Different species, injury paradigms, endpoints, exposure designs, and study methods can produce different observations. The negative result does not erase the APP/PS1 findings, but it directly prevents those findings from being presented as a general compound effect.
Putting the two papers beside each other is more informative than choosing one. The pair shows that the unaffected animal evidence is not consistent across models. It also identifies the next scientific need: preregistered, independently reproduced studies that test clearly defined target engagement and comparable outcomes across more than one experimental system.
Why Retractions Change the Evidence Assessment
A central early paper used to support Dihexa’s proposed HGF/MET mechanism was retracted in 2025. A separate early development report carries an expression of concern, and two related HGF/Met-modifier papers from the same research program were also retracted. This guide excludes those originals from positive support and treats HGF/MET involvement as proposed rather than established.
The current record status is documented by the 2025 PubMed-indexed notice from Benoist et al. (2025) for the central Dihexa mechanism paper and two notices from Kawas et al. (2025) and Kawas et al. (2025) for related HGF/Met-modifier records. Each notice reports no quantitative result of its own: it records that a named 2011, 2012 or 2014 article was withdrawn, removing its figures from the usable evidence base. These notices establish publication status. They do not create new biological evidence.
A review article cannot restore a retracted experiment simply by repeating its conclusion. Reviews inherit the strengths and weaknesses of their underlying sources, especially when written before an integrity action. Later independent studies can contribute new model-specific observations, but unless they retest the disputed molecular step, they do not validate it indirectly.
Status is part of the evidence, not a footnote
Affected originals are excluded from affirmative support. Unaffected reviews are limited to lineage. Foundational receptor papers remain pathway background. Later studies support only their own models and endpoints. This approach follows Apex’s research and review methodology.
What Remains Unestablished
The most useful limitations are specific enough to shape the next experiment. “More research is needed” does not say what evidence is missing or why the current record cannot answer the question.
| Unresolved question | Why the existing record does not close it |
|---|---|
| Direct binding or target engagement | The clean HGF/MET papers do not study Dihexa, while the central direct compound-mechanism chain is affected. |
| Independent MET confirmation | The 2021 mouse study reports PI3K/AKT-associated findings but does not independently establish the proposed HGF/MET step. |
| Cross-model consistency | The positive APP/PS1 mouse result and negative 3-NP rat result point in different directions under different designs. |
| Compound-isolated effects | Several in-vitro and nerve-repair records use multiple small molecules, cells, or combination interventions. |
| Reconciled AngIV target model | IRAP/AT4 and HGF/MET interpretations sit within a broader, contested research history; analog results are not automatically transferable. |
| Human safety and efficacy | The source set is preclinical and contains no basis for a human conclusion, use protocol, or therapeutic claim. |
| Long-term compound-specific risk | Pathway relevance alone cannot establish a compound-specific risk or a finding of safety. |
ClinicalTrials.gov searches for exact Dihexa and PNB-0408 names returned no direct study records in the July 21, 2026 evidence check. FDA GSRS lists ATH-1001 among Dihexa’s substance names, while the NCT04488419 protocol describes fosgonimeton (ATH-1017) as a distinct water-soluble prodrug converted in plasma to ATH-1001. Those trials evaluate the prodrug and its pharmaceutical exposure context, not direct administration of the Apex research compound; formulation, exposure, and outcome data must not be transferred between them.
These gaps also explain why this page contains no dosing, preparation, administration, or outcome guidance. A preclinical signal in one model is not a validated human-use instruction. The relevant research-grade regulatory context keeps chemical-reagent documentation separate from pharmaceutical evidence and clinical decision-making.
Identity and Analytical Documentation to Inspect
A literature review and a product record answer different questions. Before an in-vitro or preclinical experiment, the researcher should be able to connect the material in hand to a named lot, a declared chemical form, and analytical records whose methods and limitations are visible.
- Identity: confirm the intended chemical name, synonym, molecular formula, molecular weight, counterion or salt form where applicable, and lot-to-document match.
- Certificate traceability: verify the lot, issue date, testing laboratory, methods, results, and document identifiers. The guide to read a peptide certificate of analysis explains how these fields fit together.
- Chromatographic context: inspect method, detector, integration, specified peaks, and system suitability. HPLC purity testing and its limits explains why area percentage does not by itself establish molecular identity, total content, sterility, or biological activity.
- Complementary identity evidence: look for an identity-appropriate orthogonal method, such as mass spectrometry, rather than asking one chromatogram to answer every analytical question.
- Experimental fit: confirm that the supplied form and documentation match the planned model, controls, solvent system, concentration range, and institutional SOP.
Commercial product records must be evaluated independently from this guide. Current lot, price, availability, and specification facts require same-day first-party documentation. Nothing in this article asserts that a current lot has a particular purity, inventory status, or analytical result.
Sources and Scope
- PubChem: Dihexa, CID 129010512 – chemical name, formula, computed molecular weight, synonyms, and identifiers.
- FDA GSRS: UNII 9WYX65A5C2 – substance identity and the explicit boundary that a UNII does not imply regulatory review or approval.
- FDA: Certain Bulk Drug Substances for Use in Compounding – source-scoped context for the named Dihexa acetate form; not transferred to other forms or products.
- Foundational HGF/MET biology: Bottaro et al. (1991), Hartmann et al. (1992), and Gherardi et al. (2003); explicitly not Dihexa experiments.
- Precursor and lineage context: Yamamoto et al. (2010) and Wright et al. (2015); not independent mechanism confirmation.
- Qualified model, protocol, and evidence-landscape context: Uribe et al. (2015), Siller et al. (2015), Ho & Nation (2018), and Pan et al. (2022).
- Central unaffected model-specific evidence: Sun et al. (2021) and Wells et al. (2024).
- Publication-status notices, not biological evidence: Kawas et al. (2025), Benoist et al. (2025), and Kawas et al. (2025).
Every PMID cited in the deployable article was freshly matched through NCBI E-utilities to its title, first author, year, journal, and current retraction flag. Concern-marked and retracted originals were not used as affirmative evidence. Source roles are narrow by design: identity sources establish identity, receptor papers establish pathway background, notices establish publication status, and experiments support only their stated models and endpoints.
Research Use Disclaimer
This article is provided for educational and research reference purposes only. Dihexa and all products sold by Apex Laboratory are intended exclusively for in-vitro laboratory research use and are not for human consumption. Researchers should consult the primary peer-reviewed literature cited throughout this article for detailed methodological protocols, experimental designs, and complete data sets.
