Davunetide (NAP)

Origin and structure

Davunetide is a chain of eight amino acids, Asn-Ala-Pro-Val-Ser-Ile-Pro-Gln, written NAPVSIPQ in one-letter code and known as NAP after its first three letters (C36H60N10O12, 824.9 g/mol, CAS 211439-12-2). It is an exact fragment of activity-dependent neuroprotective protein (ADNP), a much larger protein whose sequence Illana Gozes and colleagues at Tel Aviv University first reported, in mice, in 1999. They found that the neuropeptide VIP raised its production in astrocytes, the support cells of the brain, and that the NAP fragment protected cultured nerve cells at concentrations below one femtomolar against toxins such as beta-amyloid, the protein that builds up in Alzheimer's disease. Later mouse work showed that ADNP is essential: mice that lack it completely do not form a brain.

Ramot, Tel Aviv University's technology-transfer company, licensed NAP to Allon Therapeutics in Vancouver, of which Gozes was a founder and chief scientific officer. Allon developed it as AL-108, a nasal spray, and AL-208, an intravenous solution, and davunetide became its official generic name. After the failed trial in progressive supranuclear palsy (PSP), Allon was bought by Paladin Labs in 2013 and the trial data went back to Ramot. Coronis Neurosciences then licensed the peptide from Ramot as CP201 and in 2018 obtained the US orphan designation for ADNP syndrome; the licence now lies with ExoNavis Therapeutics in Tel Aviv, where Gozes is vice president of drug development.

How it is thought to work

Davunetide is thought to act on microtubules, the protein tubes that give nerve cells their shape and carry cargo along their long extensions. In laboratory studies NAP bound tubulin, the building block of microtubules, and helped microtubules assemble; later work from the same lab linked this to its Ser-Ile-Pro motif, through which it interacts with the microtubule end-binding proteins EB1 and EB3 (Oz et al., Mol Psychiatry 2014). Tau is a protein that holds microtubules together and clumps into tangles in Alzheimer's disease and in PSP. In preclinical studies davunetide stabilised microtubules and reduced the excess phosphorylation of tau, which is why tau diseases became its main target.

Most of this work comes from Gozes's group and its partners, and all of it was done in cells and animals such as flies and mice. The PSP trial had no direct measure showing that davunetide engaged its target in the human brain, and tau and other markers in the spinal fluid of a subset of participants changed no differently from placebo.

Is davunetide approved?

No. Davunetide has no marketing authorisation anywhere in the world, and research material sold under the name NAP is not a medicine. What it does have is orphan designation for ADNP syndrome in the EU and the US. Orphan designation gives the developer of a medicine for a rare disease regulatory support and incentives; it says nothing about whether the medicine works, and the EMA states that it does not mean a medicine is available or authorised.

• EU: no medicine containing davunetide is authorised. On 23 March 2010 the European Commission designated it an orphan medicine for PSP; that designation was withdrawn at the sponsor's request in April 2013. On 13 December 2024 it was designated an orphan medicine for ADNP syndrome (EU/3/24/3007, sponsor AdRes EU B.V.).
• Netherlands: davunetide is not a registered medicine, so it may not be sold, advertised or supplied for use in people. None of the five patient trials on ClinicalTrials.gov had a Dutch site, and on 28 October 2026 the EU's Clinical Trials Information System (CTIS) listed no trial with davunetide.
• US: not FDA-approved. On 22 February 2018 the FDA granted orphan designation for the treatment of ADNP syndrome; its database lists ExoNavis Therapeutics as the sponsor and the drug as not approved for that use.

What the research shows

By 2013 the developer counted ten clinical trials of davunetide, early studies in healthy volunteers among them (Morimoto et al., Expert Rev Clin Pharmacol 2013). Five patient trials are registered on ClinicalTrials.gov, with 766 participants in all; only the PSP trial was large and long enough to test whether it slows a disease. The main results:

• PSP, phase 2/3 (AL-108-231, NCT01110720; Boxer et al., Lancet Neurol 2014): 313 people with progressive supranuclear palsy at 48 centres in Australia, Canada, France, Germany, the UK and the US were randomised to davunetide nasal spray (157) or placebo (156) for 52 weeks. On both primary endpoints the groups ended up the same: the PSP Rating Scale worsened by a median of 11.8 points in each group (p = 0.41), and daily functioning on the Schwab and England scale declined equally (p = 0.92). Secondary and exploratory endpoints, including MRI brain volumes and spinal-fluid markers, showed no difference either. Allon announced the result on 18 December 2012 and said it would not invest further in davunetide.
• Amnestic mild cognitive impairment, phase 2 (AL-108-211, NCT00422981; Morimoto et al., Dement Geriatr Cogn Disord 2013): 144 adults in the US received a low or a high dose or placebo, in a 2:1 ratio of drug to placebo, for 12 weeks. The trial missed its primary endpoint, a composite memory score; the authors reported signals of efficacy on two individual tests of memory and attention.
• Schizophrenia, phase 2 (NCT00505765; Javitt et al., Schizophr Res 2012): 63 people on antipsychotics received one of two doses or placebo for 12 weeks in a US trial funded by the National Institute of Mental Health. Cognition on the MATRICS test battery did not differ from placebo (p = 0.45); a test of everyday skills (UPSA) favoured davunetide (p = 0.048), and the authors estimated that 45 to 50 people per group would be needed for a firm answer.
• Heart bypass surgery, phase 2 (AL-208, NCT00404014): 234 people undergoing coronary bypass surgery in the US and Canada were enrolled. After an open-label run-in with three doses in 20 people, the main part compared a single intravenous dose of davunetide with placebo. The original analysis found no difference in cognition after surgery, but neither group showed the decline the trial was designed to prevent.
• Tau diseases, pilot (NCT01056965, University of California, San Francisco): 12 people with PSP, corticobasal syndrome or related frontotemporal disorders received davunetide or placebo (2:1) for 12 weeks, mainly to gather safety data. The study was completed in December 2012; we found no posted or published results.
• ADNP syndrome: no human data. In mice carrying an Adnp mutation that mirrors the most common one in patients, NAP corrected some changes in synapses, tau and gene activity (Karmon et al., Biol Psychiatry 2022).

Later reanalyses and ADNP syndrome

Since 2023 Gozes's group has reanalysed the old trials separately for men and women. In the PSP trial, women declined faster than men, and women on davunetide declined more slowly than women on placebo on the daily-living scale from week 39 (p = 0.008); a 2026 analysis with a shorter, 10-item version of the PSP Rating Scale also favoured women (Shapira et al., Mol Psychiatry 2026). A 2024 reanalysis of the mild cognitive impairment trial reported dose-dependent gains on a visual memory test in men, and a gain over placebo on a digit-span test in women at the higher dose. A 2025 reanalysis of the bypass trial reported better verbal memory scores in men, based on 70 men on davunetide and 64 on placebo, against only 16 and 13 women.

These are post-hoc findings: the trials were not designed to compare men and women, the subgroups are small, many outcomes were tested, and none of the effects has been confirmed in a new trial. Gozes and co-authors are named as inventors on a patent application for sex-dependent use of davunetide, which is licensed to ExoNavis.

ADNP syndrome, also called Helsmoortel-Van der Aa syndrome, was described in 2014 by geneticists at the University of Antwerp, with colleagues from Radboud University Medical Center in Nijmegen and other centres among the co-authors. It is caused by new (de novo) mutations in the ADNP gene and brings intellectual disability and autism; the first paper, on ten patients, estimated that ADNP is mutated in at least 0.17% of people with autism. The case for davunetide is that it is the protective fragment of the very protein that is faulty in these children. By 28 October 2026 no davunetide trial in ADNP syndrome was listed on ClinicalTrials.gov or in CTIS, and no data in children had been published.

Side effects and safety

In the trials davunetide was generally well tolerated. The best data come from the PSP trial, in which people used the nasal spray twice daily for up to a year. Serious adverse events were equally frequent (54 in each group), as were deaths (11 on davunetide, 10 on placebo), and ECGs and laboratory values did not differ. Nasal side effects were more common: nosebleeds in 12% against 8% on placebo, a runny nose in 10% against 5%, and nasal discomfort in 10% against under 1%. Side effects led 21 people on davunetide to stop treatment, against 12 on placebo, mostly because of nosebleeds or a blocked nose.

In the 12-week schizophrenia trial the investigators saw no significant side effects. In the bypass trial, where patients received one intravenous dose during heart surgery, the reported adverse events were the usual complications of the operation and were not more frequent on davunetide. Three patients died, two of them on davunetide; the investigator judged one of those deaths possibly related to the drug, but an independent safety review did not agree.

According to the developers, repeated-dose studies in rats and dogs found no adverse effects at concentrations about 500 times the biologically active dose. What is missing: no study lasted longer than a year, no data in children or in pregnancy have been published, and every published human study gave davunetide as a nasal spray or an intravenous infusion, never as an injection under the skin. Research-grade NAP is not the trial formulation, so its identity and purity rest on its certificate of analysis.

Frequently asked questions

Is davunetide the same as NAP peptide?

Yes. NAP is the laboratory name of the eight-amino-acid peptide NAPVSIPQ, and davunetide is its official generic name. AL-108 was the developer's code for the nasal spray, AL-208 for the intravenous form, and CP201 a later code for the same peptide. Material sold online as NAP is research material, not the formulation used in the trials.

Did davunetide work in progressive supranuclear palsy?

No. In the phase 2/3 trial of 313 patients, a year of davunetide nasal spray changed neither the PSP Rating Scale nor daily functioning compared with placebo, and none of the secondary endpoints moved. Later reanalyses by the group that discovered the peptide report a benefit in women, but those are post-hoc subgroup findings that have not been tested in a new trial.

Does davunetide improve memory?

That has not been shown. In 144 people with mild cognitive impairment it missed its main memory endpoint, with signals on only two individual tests, and in schizophrenia it did not improve cognition on the main test battery. No published trial has tested it as a memory aid in healthy people; claims that it is a nootropic rest on mouse studies.

Is davunetide a treatment for ADNP syndrome?

Not yet. It has orphan designation for ADNP syndrome in the US (2018) and the EU (2024), which supports development but is not an approval. The evidence so far comes from mice with Adnp mutations, and by 28 October 2026 no trial in children with ADNP syndrome was registered on ClinicalTrials.gov or in the EU trial register CTIS.

Is davunetide legal in the Netherlands?

Davunetide is not a registered medicine in the Netherlands or anywhere else in the EU, so it may not be sold, advertised or supplied for use in people, and its EU orphan designation does not change that. Trading it as research material is a different status, which our guide on research-use-only products in the Netherlands explains. For athletes it is prohibited however it was obtained.

Is davunetide banned in sport?

Yes, at all times. It is not named on the WADA Prohibited List, but class S0 bans any pharmacological substance that no other class covers and that has no current approval for human therapeutic use, explicitly including drugs whose development was discontinued. The 2027 list, in force from 1 January 2027, also names peptides as an example. The Dutch Dopingautoriteit applies the same list.

Storage and handling

Store according to product-label instructions and batch handling notes. Lyophilized research peptides are commonly kept dry, protected from light, and handled with clean laboratory technique.

References

Checked against these sources on 28 October 2026.

Related peptides