Davunetide

A synthetic eight-amino-acid fragment (NAPVSIPQ) of activity-dependent neuroprotective protein, developed as an intranasal microtubule-stabilizing tau therapy, whose 313-patient phase 2/3 trial in progressive supranuclear palsy was negative on every endpoint.

Also referenced as: NAP, NAPVSIPQ, AL-108, AL-208, CP-201, Davunetide acetate

Also appears in: Cognitive

COAPublic product evidenceSearch the public certificate ledger for this compoundNo exact compound records are currently indexed under this profile name.→

---

What is Davunetide?

Davunetide is a synthetic octapeptide, Asn-Ala-Pro-Val-Ser-Ile-Pro-Gln — written as NAPVSIPQ and abbreviated NAP after its first three residues. The sequence is the shortest neuroprotective fragment identified within activity-dependent neuroprotective protein (ADNP), a much larger protein of 1,102 amino acids that was cloned and sequenced at Tel Aviv University in the late 1990s (Bassan et al., Journal of Neurochemistry, 1999;72(3):1283–1293). The fragment sits at residues 354–361 of the human ADNP sequence.

Allon Therapeutics developed it under two formulation codes that are still used interchangeably in the literature: AL-108 for the intranasal spray and AL-208 for the intravenous solution. Allon became insolvent and was acquired by Paladin Labs in July 2013, after the pivotal trial read out; the peptide was later licensed from Tel Aviv University to ExoNavis Therapeutics in 2021 and carries the additional code CP-201 in that program.

It appears in the research-chemical and research-peptide market almost entirely on the strength of its mechanism story — microtubule stabilization and tau protection — rather than on any positive clinical result. Coverage of ADNP syndrome, a rare neurodevelopmental disorder caused by mutations in the parent protein, has renewed interest in the fragment independently of the failed neurodegeneration program.

How it works

• Microtubule end-binding protein interaction through the SxIP motif. Sequence analysis identified the Ser-Ile-Pro (SIP) motif inside NAP as an EB1-interacting SxIP site, and NAP was shown to bind members of the microtubule end-binding protein family. NAP increased PSD-95 expression in dendritic spines, an effect abolished by EB3 silencing, while EB1 or EB3 — but not EB2 — silencing blocked NAP-mediated cell protection (Oz et al., Molecular Psychiatry, 2014;19(10):1115–1124).
• Enhanced tau–microtubule association. Follow-up work from the same group reported that ADNP and NAP markedly increase the interaction between end-binding proteins and Tau, framed as a route to protection against tauopathy (Ivashko-Pachima et al., Molecular Psychiatry, 2017;22(9):1335–1344).
• Reduced tau hyperphosphorylation in rodent models. NAP reduced tau hyperphosphorylation and improved learning in an ADNP-haploinsufficient transgenic mouse (Vulih-Shultzman et al., Journal of Pharmacology and Experimental Therapeutics, 2007;323(2):438–449), and reduced tau pathology with improved cognitive function in a mouse model of Alzheimer’s disease (Matsuoka et al., Journal of Pharmacology and Experimental Therapeutics, 2008;325(1):146–153).
• Pleiotropy beyond the cytoskeleton. More recent work reports that NAP penetrates cell nuclei, which the authors offer as an explanation for effects that a purely cytoplasmic microtubule mechanism does not cover (Ganaiem et al., Cells, 2023;12(18)).

A caveat that applies to the whole list: the great majority of this mechanistic literature originates from the laboratory that discovered ADNP and NAP. The molecular story is internally consistent and has been developed over twenty-five years, but it has not been carried forward by a comparable volume of independent groups, and it did not predict the clinical outcomes below.

Research status

Davunetide has a substantial controlled human record for a peptide of this kind — roughly 750 participants across four randomized trials — and that record is dominated by missed endpoints. Five davunetide studies are registered on ClinicalTrials.gov in total, and all five are completed.

Amnestic mild cognitive impairment. AL-108-211 (ClinicalTrials.gov NCT00422981) randomized 144 subjects 2:1 to drug or placebo across ascending intranasal dose levels for 12 weeks. Its registered primary endpoint was the change from baseline to week 12 on a composite memory variable, and the trial did not meet it: the analysis “failed to detect a statistically significant difference between the treatment groups” on the composite cognitive memory score, with signals reported on two individual tests of memory and attention (Morimoto et al., Dementia and Geriatric Cognitive Disorders, 2013;35(5–6):325–336). This study is the origin of the widely repeated claim that davunetide improves memory in prodromal Alzheimer’s disease; the claim rests on components of a study whose pre-specified composite was negative.

Schizophrenia. A multicenter, double-blind trial (NCT00505765, run by an academic sponsor rather than Allon) gave 63 outpatients intranasal davunetide at 5 or 30 mg per day, or placebo, for 12 weeks. The registered primary endpoint was the MATRICS Consensus Cognitive Battery, and it missed — no significant difference from placebo (p=0.45). The UCSD Performance-based Skills Assessment, a functional-capacity measure registered as a secondary endpoint, did reach significance (p=0.048). The authors concluded that groups of at least 45–50 subjects would be needed to obtain significant effects on both measures (Javitt et al., Schizophrenia Research, 2012;136(1–3):25–31). An exploratory magnetic resonance spectroscopy substudy in 18 of those patients reported small changes in prefrontal NAA/creatine and choline/creatine, most of them at trend level (Jarskog et al., Neuropsychopharmacology, 2013;38(7):1245–1252). Cognition, the endpoint the program was built around, was null.

Cognitive impairment after coronary artery bypass grafting. A phase 2 study of the intravenous AL-208 formulation enrolled 234 patients (NCT00404014) and completed in June 2008 without a primary efficacy publication at the time; a re-analysis appeared seventeen years later (Gozes et al., Translational Psychiatry, 2025;15(1):454).

Progressive supranuclear palsy — the pivotal failure. AL-108-231 (NCT01110720) randomized 313 patients 1:1 — 157 to davunetide, 156 to placebo — to intranasal davunetide 30 mg twice daily or placebo for 52 weeks at 48 centers across six countries, with co-primary endpoints of change from baseline in the PSP Rating Scale and the Schwab and England Activities of Daily Living scale. The groups were indistinguishable: median PSPRS change 11.8 in both arms (p=0.41) and SEADL change −0.20 in both arms (p=0.92). The two pre-specified secondary endpoints — clinical global impression of change and ventricular volume on volumetric MRI — showed no group differences either. Serious adverse events were balanced at 54 per group, with 11 deaths on davunetide and 10 on placebo; nasal events were more frequent on drug (epistaxis 12% vs 8%, rhinorrhea 10% vs 5%, nasal discomfort 10% vs under 1%). The authors’ conclusion was unambiguous — davunetide is not an effective treatment for PSP (Boxer et al., The Lancet Neurology, 2014;13(7):676–685). An accompanying commentary used the result to ask when neuroprotective drugs should move from mice to men (Tariot, The Lancet Neurology, 2014;13(7):641–643). A separate 12-patient phase 1 pilot in predicted tauopathies (NCT01056965) had run in parallel. Clinical development in neurodegeneration stopped after this readout.

Post-hoc re-analyses. Beginning in 2023, the originating laboratory published sex-stratified re-analyses of three of the older trials: the PSP study (Gozes et al., Translational Psychiatry, 2023;13(1):319), the aMCI study (Gozes et al., Translational Psychiatry, 2024;14(1):412), and the CABG study (Gozes et al., Translational Psychiatry, 2025;15(1):454). The claimed direction is not consistent across them: the PSP paper reports efficacy in women, the CABG paper reports neuroprotection and cognitive benefit in men, and the aMCI paper splits its findings across different tests in each sex. These are retrospective subgroup analyses of trials that were negative on their pre-specified endpoints, performed by the compound’s discoverers, who disclose an international patent application covering sex-dependent use of davunetide — licensed to the company now developing it — with themselves as inventors. They are hypothesis-generating and are not a substitute for a prospective trial.

Current status. Davunetide holds an EU orphan designation for ADNP syndrome granted 13 December 2024 (EU/3/24/3007); an earlier EU designation for PSP (EU/3/10/728, granted 23 March 2010) was withdrawn from the Community Register in April 2013 at the sponsor’s request, and US orphan-drug and rare-pediatric-disease designations for ADNP syndrome have been reported. A phase 3 study in approximately 97 children with ADNP mutations is described by the sponsor as having begun in October 2024, but no matching record appears in ClinicalTrials.gov as of August 2026 — the only davunetide entries in the registry remain the five completed studies above. No regulatory authority has approved davunetide for any indication.

Common dosage forms

• Lyophilized powder in vials, typically at milligram scale and frequently supplied as the acetate salt of davunetide (free peptide CAS 211439-12-2), sold as a research reagent for laboratory use only.
• Intranasal solution or spray, the AL-108 presentation, which is the route used in every davunetide trial except the perioperative bypass-surgery study.
• Sterile solution for intravenous infusion, the AL-208 presentation, used only in that single perioperative study.
• Reconstituted vials intended for injection have no counterpart in the clinical record; nothing in the trial program was administered subcutaneously.

This section describes formats only and is not dosing guidance.

Key considerations

• The pivotal trial was negative across the board. The PSP study was a full-size, 52-week, multinational phase 2/3 trial and it separated from placebo on nothing — neither co-primary, nor either of the two pre-specified secondary endpoints. Material that cites davunetide’s “phase 3” pedigree without that outcome is describing how far the program got, not what it showed. Development in neurodegeneration ended there.
• The two “positive” trials did not meet the endpoints they are cited for. The mild-cognitive-impairment study missed its composite memory primary, and the schizophrenia study missed the cognitive primary that motivated it. What remains is a secondary functional-capacity result in 63 patients and post-hoc components of a failed composite — not demonstrated cognitive benefit.
• Evidence is concentrated in a single laboratory. Discovery, mechanism, and the 2023–2025 sex-stratified reinterpretations of the negative trials all trace to the same group, which holds patent rights over the sex-dependent use those re-analyses propose. Independent replication of the clinical claims does not exist, and subgroup findings extracted from trials that already failed carry the usual multiplicity problems — sharpened here by the fact that the benefiting sex is not the same one from trial to trial.
• Route and tolerability. Intranasal dosing was well tolerated overall, with nasal irritation, rhinorrhea, and epistaxis the consistent drug-related findings; mortality and serious adverse events were balanced against placebo in the PSP trial. There is no human safety database for injected davunetide beyond one intravenous perioperative study, and none at all for repeated parenteral use.
• Naming is a live source of error. “NAP” is a heavily overloaded abbreviation in the biomedical literature, so searches on it return large volumes of unrelated material; the same compound also appears as AL-108, AL-208, CP-201, and NAPVSIPQ. Davunetide (the eight-residue fragment) should not be conflated with ADNP (the full parent protein) or with ADNP syndrome (the disorder). Orphan designation in either the US or EU is an incentive status granted for rare-disease development and is not evidence of efficacy or a form of approval.