Peptide Approaches to Alzheimer's Disease
Neurodegenerative Peptides
182 Trials
Active Alzheimer's disease clinical trials in 2025, spanning amyloid-targeting, tau-targeting, neuroprotective, and anti-inflammatory approaches.
Cummings et al., Alzheimer's & Dementia, 2025
Cummings et al., Alzheimer's & Dementia, 2025
View as imageAlzheimer's disease is, at its molecular core, a peptide disease. The amyloid plaques that define the pathology are composed of amyloid-beta (A-beta), a 39-43 amino acid peptide fragment cleaved from amyloid precursor protein. The neurofibrillary tangles are built from hyperphosphorylated tau, a protein whose aggregation is seeded by specific peptide sequences. It makes biological sense that peptide-based therapeutics would target a peptide-driven disease. Yet the Alzheimer's pipeline has been dominated by monoclonal antibodies (lecanemab, donanemab, aducanumab), not peptide drugs. The reasons are practical: peptides face challenges with blood-brain barrier penetration, rapid enzymatic degradation in the CNS, and difficulty achieving sustained brain exposure. Despite these hurdles, at least five distinct peptide strategies are in active development, from amyloid aggregation inhibitors to neuroprotective peptides to GLP-1 receptor agonists being repurposed from diabetes. For the foundational science on amyloid-beta itself, see the pillar article.
Key Takeaways
- Amyloid-beta is a 39-43 amino acid peptide whose aggregation into oligomers and fibrils drives Alzheimer's neurotoxicity; peptide-based approaches target this aggregation at multiple stages
- Beta-sheet breaker peptides (iA-beta-5, D-peptides) disrupt amyloid fibril formation in vitro and reduce plaque burden in transgenic mice, but none have succeeded in human trials
- Davunetide (NAP), an 8-amino-acid neuroprotective peptide derived from ADNP, reduced tau phosphorylation in preclinical models but failed its Phase 2/3 trial in progressive supranuclear palsy
- GLP-1 receptor agonists reduce neuroinflammation and amyloid burden in animal models, but semaglutide did not slow cognitive decline in the EVOKE Phase 3 trials despite biomarker improvements
- Cerebrolysin, a mixture of neurotrophic peptides, showed modest cognitive benefits in multiple Alzheimer's RCTs, though effect sizes were small and regulatory approval outside of limited markets has not been achieved
- Active peptide vaccines (CAD106, UB-311, ACI-24) train the immune system to clear amyloid-beta and are in various stages of clinical testing
Strategy 1: Breaking the Amyloid Cascade with Peptide Inhibitors
Beta-Sheet Breaker Peptides
Amyloid-beta toxicity depends on aggregation. Monomeric A-beta is relatively harmless; the damage comes from oligomers (small clusters of 2-12 peptides) and fibrils (elongated beta-sheet-rich structures that form plaques). Beta-sheet breaker peptides are short synthetic sequences designed to bind the aggregation-prone regions of A-beta and prevent the formation of toxic structures.
The concept originated with Soto and colleagues (1998), who designed a pentapeptide (iA-beta-5) that incorporated proline residues to disrupt the beta-sheet conformation of A-beta. In vitro, iA-beta-5 dissolved preformed amyloid fibrils and prevented new fibril formation. In transgenic mouse models, intracerebral injection reduced amyloid plaque burden.[1]
The challenge was delivery. iA-beta-5 does not cross the blood-brain barrier, degrades rapidly in plasma, and requires direct CNS injection. Subsequent efforts used D-amino acid versions (resistant to proteolysis), cyclic peptides (improved stability), and retro-inverso peptides (mirror-image sequences with enhanced metabolic stability). Willbold and colleagues developed D3, a fully D-amino acid peptide identified by mirror-image phage display, that binds A-beta oligomers with high affinity and reduced plaque burden when administered orally to transgenic mice.[2]
Despite preclinical success, no beta-sheet breaker peptide has completed a successful Phase 2 or Phase 3 trial. The reasons likely include insufficient brain penetration, the complexity of amyloid biology (different aggregation states may require different interventions), and the broader failure of the amyloid-only hypothesis to translate into cognitive benefit.
Peptide-Based Aggregation Modulators
A related approach uses peptides that do not break existing aggregates but prevent their formation by occupying aggregation-prone surfaces on monomeric A-beta. These include peptides derived from the A-beta sequence itself (self-recognition elements) modified to act as "decoys" that bind monomers without extending the aggregation process. Chalifour and colleagues (2003) demonstrated that a pentapeptide mimicking residues 17-21 of A-beta (the central hydrophobic cluster) inhibited fibril formation in vitro by competing for monomer-monomer binding sites.[3]
Strategy 2: Neuroprotective Peptides
Davunetide (NAP)
Davunetide is an 8-amino-acid peptide (NAPVSIPQ) derived from activity-dependent neuroprotective protein (ADNP), an essential protein for brain development. In preclinical models, intranasal davunetide protected neurons against amyloid-beta toxicity, reduced tau hyperphosphorylation, and improved cognitive performance in transgenic Alzheimer's mice.[4]
The mechanism appears to involve microtubule stabilization. Davunetide promotes tubulin assembly and protects microtubule structure, which is relevant because tau pathology in Alzheimer's involves detachment of tau from microtubules, leading to both microtubule destabilization and tau aggregation. By stabilizing microtubules independently of tau, davunetide could theoretically compensate for tau dysfunction.
Intranasal delivery was a practical advantage: davunetide reached the brain via the olfactory pathway, bypassing the blood-brain barrier. Phase 2 studies in mild cognitive impairment showed trends toward cognitive improvement on some measures. However, the pivotal Phase 2/3 trial was conducted in progressive supranuclear palsy (PSP), a tauopathy chosen because it progresses faster than Alzheimer's, allowing shorter trials. Davunetide failed to show benefit on functional or cognitive endpoints in PSP, and clinical development was discontinued. For a detailed history of this peptide, see NAP Peptide (Davunetide): A Neuroprotective Peptide That Reached Clinical Trials.
Cerebrolysin
Cerebrolysin is not a single peptide but a standardized mixture of low-molecular-weight neurotrophic peptides and amino acids derived from porcine brain tissue. It contains fragments analogous to brain-derived neurotrophic factor (BDNF), ciliary neurotrophic factor (CNTF), and glial cell-derived neurotrophic factor (GDNF). For a detailed look at how this peptide mixture works, see How Cerebrolysin Works: Neurotrophic Factor Activity Explained.
Multiple randomized controlled trials have tested intravenous cerebrolysin in mild-to-moderate Alzheimer's disease. A Cochrane-style meta-analysis found modest but statistically significant improvements on the ADAS-cog (Alzheimer's Disease Assessment Scale, cognitive subscale) and CGI (Clinical Global Impression) scales. Effect sizes were small (2-3 points on ADAS-cog), and the clinical meaningfulness of these improvements is debated.[5]
Cerebrolysin is approved for Alzheimer's disease in several countries (Austria, China, South Korea) but not in the United States, where the FDA has not accepted the existing trial data as sufficient for approval. The peptide mixture faces the additional challenge of standardization: batch-to-batch consistency of a complex biological product is harder to guarantee than for a single synthetic peptide.
Strategy 3: GLP-1 Receptor Agonists
The repurposing of GLP-1 receptor agonists for Alzheimer's disease is one of the most-watched stories in neurology. GLP-1 receptors are expressed in the hippocampus, cortex, and other brain regions affected by Alzheimer's. Preclinical evidence shows that GLP-1 agonists reduce neuroinflammation, decrease amyloid plaque burden, protect synapses, and improve memory in transgenic Alzheimer's mice.
Liraglutide was the first GLP-1 agonist tested in an Alzheimer's trial. The ELAD trial (Evaluating Liraglutide in Alzheimer's Disease) randomized 204 participants with mild Alzheimer's to liraglutide or placebo for 12 months. Liraglutide slowed the decline in cerebral glucose metabolism measured by FDG-PET but did not produce significant differences on cognitive endpoints.[6]
The larger EVOKE and EVOKE+ trials tested oral semaglutide in early Alzheimer's disease. Despite being adequately powered and well-designed, semaglutide did not meet its primary endpoint of slowing cognitive decline on the CDR-SB (Clinical Dementia Rating Sum of Boxes). Some biomarker improvements were observed, but the disconnect between biomarkers and clinical outcomes echoed the broader challenge in Alzheimer's drug development: modifying disease biology does not always translate to detectable cognitive benefit within trial timescales.[7]
The EVOKE failure does not necessarily close the door on GLP-1 agonists for neurodegeneration. The trial tested oral semaglutide, which achieves lower brain concentrations than injectable formulations. Whether higher brain exposure (via injectable delivery or more brain-penetrant analogs) would produce different results is being debated. Other GLP-1/GIP dual agonists with improved BBB penetration are in preclinical development for neurodegenerative indications.
The GLP-1 agonist story in Alzheimer's also connects to broader epidemiological observations. Large real-world cohort studies of type 2 diabetes patients found that those taking GLP-1 receptor agonists had a 30-50% lower incidence of Alzheimer's diagnosis compared to matched controls on other diabetes medications. A 2025 propensity-matched cohort study confirmed this association, though the direction of causality remains uncertain: GLP-1 agonists may directly protect neurons, or they may reduce Alzheimer's risk indirectly through improved metabolic health, weight loss, and cardiovascular protection. The discrepancy between epidemiological signals and the EVOKE trial results suggests that any neuroprotective effect may be preventive (working before significant pathology accumulates) rather than therapeutic (reversing existing damage).
Strategy 4: Peptide Vaccines for Amyloid
Rather than administering therapeutic peptides directly, active immunization uses peptide fragments to train the immune system to produce antibodies that clear amyloid-beta. This approach was pioneered by the AN-1792 vaccine trial, which used full-length A-beta 1-42. The trial was halted when 6% of participants developed meningoencephalitis, an autoimmune inflammatory response driven by T-cell activation against the peptide antigen.
Second-generation vaccines were redesigned to target only the B-cell epitope (the N-terminal fragment of A-beta, residues 1-6 to 1-15) while avoiding T-cell epitopes that triggered inflammation:
CAD106 conjugates the A-beta 1-6 fragment to a virus-like particle carrier. Phase 2 trials showed robust anti-A-beta antibody responses with acceptable safety. The Generation Study tested CAD106 in presymptomatic carriers of autosomal dominant Alzheimer's mutations but was discontinued due to strategic reprioritization rather than safety concerns.
UB-311 uses A-beta 1-14 peptides conjugated to a synthetic T-helper epitope. Phase 2 results showed anti-A-beta antibody responses comparable to monoclonal antibody treatments. Phase 2 cognitive endpoints showed encouraging trends.
ACI-24 uses A-beta 1-15 in a liposomal formulation designed to specifically target amyloid-beta in its beta-sheet conformation. This conformational selectivity could spare monomeric A-beta (which may have normal physiological functions) while targeting toxic aggregates.
For a deeper analysis of this approach, see Peptide Vaccines for Alzheimer's: Targeting Amyloid with Immunotherapy.
Strategy 5: Tau-Targeting Peptides
While amyloid-beta has received the most attention, tau pathology correlates more closely with cognitive decline. Peptide approaches to tau include:
Tau aggregation inhibitors that target the hexapeptide motifs (PHF6: VQIVYK and PHF6*: VQIINK) responsible for tau self-assembly. Short peptides incorporating proline substitutions at these motifs can act as caps that bind to tau and prevent fibril elongation, conceptually similar to beta-sheet breaker peptides for amyloid-beta.
Microtubule-stabilizing peptides that compensate for the loss of tau-microtubule binding. Davunetide operated on this principle, though its clinical failure in PSP dampened enthusiasm for this approach.
Tau immunotherapy with peptide-based vaccines is in early development. AADvac1, developed by Axon Neuroscience, uses a synthetic tau peptide spanning residues 294-305 (a truncated tau fragment found in neurofibrillary tangles) conjugated to KLH carrier protein. Phase 2 trials showed that the vaccine produced anti-tau antibodies and reduced blood neurofilament light chain (NfL), a marker of neurodegeneration, in a subgroup analysis. Phase 2 testing in mild cognitive impairment is underway with results expected through 2026.[8]
The Blood-Brain Barrier Problem
All peptide-based Alzheimer's approaches face the central delivery challenge: getting enough peptide across the blood-brain barrier to achieve therapeutic concentrations in the brain. Solutions being developed include:
- Intranasal delivery: exploited by davunetide and being explored for other neuroprotective peptides
- Cell-penetrating peptide conjugation: attaching brain-penetrant sequences (TAT, penetratin) to therapeutic peptides
- Receptor-mediated transcytosis: conjugating peptides to ligands that bind transferrin or insulin receptors on brain endothelial cells
- Nanoparticle encapsulation: lipid nanoparticles and polymeric carriers that enhance BBB penetration
A 2025 review noted that recent advances in cyclization, PEGylation, and lipid nanoparticle incorporation have substantially improved the pharmacokinetic profile of candidate peptides, though no peptide therapeutic for Alzheimer's has yet achieved the sustained brain exposure that monoclonal antibodies like lecanemab achieve through direct CSF entry via ARIA-related mechanisms.[9]
The relationship between Huntington's disease peptide research and Alzheimer's peptide work is worth noting: both diseases involve toxic protein aggregation, and lessons from one field increasingly inform the other. Similarly, Parkinson's disease peptide approaches targeting alpha-synuclein aggregation face many of the same delivery and efficacy challenges. The broader field of neurotrophic peptides provides context for how growth factor-based strategies may complement the disease-specific approaches described here, and cerebrolysin's specific Alzheimer's evidence is covered in a dedicated article.
The Bottom Line
Multiple peptide strategies are being developed for Alzheimer's disease, targeting amyloid-beta aggregation, tau pathology, neuroinflammation, and neuronal protection. Beta-sheet breaker peptides and aggregation inhibitors have shown preclinical promise but have not produced clinical benefit. Davunetide failed in a tauopathy trial. Cerebrolysin shows modest cognitive effects in some markets. GLP-1 agonists were highly anticipated but semaglutide failed to slow cognitive decline in the EVOKE trials. Active peptide vaccines (CAD106, UB-311, ACI-24, AADvac1) remain in clinical testing with mixed but ongoing signals. The blood-brain barrier continues to be the primary obstacle to peptide therapeutics for neurodegeneration.