What it is
Amyloid-beta is a short peptide cut from a long protein. The parent is the amyloid-beta precursor protein, APP — 770 residues in its longest form, sitting in the membrane of nerve cells (UniProt P05067). Two enzymes cut it in sequence: beta-secretase makes the first cut, gamma-secretase the second. What falls out between them is amyloid-beta, mainly in two lengths:
| Form | Precursor residues | Length |
|---|---|---|
| Aβ40 | 672–711 | 40 |
| Aβ42 | 672–713 | 42 |
The full Aβ42 sequence is DAEFRHDSGYEVHHQKLVFFAEDVGSNKGAIIGLMVGGVVIA. The peptide-chain entry explains how a sequence like this is read.
How it was found
In 1984 Glenner and Wong purified a protein from the amyloid deposits in the blood vessels of Alzheimer's brains and sequenced it. A computer search found no match with any protein sequenced up to then (Biochemical and Biophysical Research Communications, PMID 6375662). They suggested it might come from a unique precursor in the blood and might offer a diagnostic test.
What it does
Its normal job is not settled. UniProt annotates amyloid-beta peptides as metal chelators — binding copper, zinc and iron — that bind lipoproteins in spinal fluid and plasma, and annotates Aβ42 as the more effective reductant of the two forms. Some of these annotations come from laboratory studies.
What is not in doubt is the clumping: amyloid-beta is the main constituent of the amyloid plaques in Alzheimer's disease. Whether those plaques drive the disease or accompany it is debated; the antibody trials below test it directly in people.
| What is described | How firm |
|---|---|
| Main protein of Alzheimer's amyloid | Biochemistry of human brain tissue (1984 onward) |
| Normal role in metal and lipoprotein binding | UniProt annotation, partly in vitro |
| Removing it slows decline | One large trial: a moderate difference, with side effects |
The antibodies aimed at it
The approved Alzheimer's drugs connected with amyloid-beta are antibodies — large proteins — not peptides. FDA's records list three: aducanumab (Aduhelm, 2021-06-07), lecanemab (Leqembi, 2023-01-06) and donanemab (Kisunla, 2024-07-02).
The largest published result is lecanemab's Clarity AD trial (New England Journal of Medicine, 2023, PMID 36449413; funded by Eisai and Biogen):
| 18-month outcome, 1,795 people with early Alzheimer's | Lecanemab | Placebo |
|---|---|---|
| Change in CDR-SB (0–18; higher = worse) | +1.21 | +1.66 |
| Brain amyloid (PET substudy, 698 people) | −59.1 centiloids vs placebo | — |
| Infusion-related reactions | 26.4% | — |
| Brain swelling or effusion on scans (ARIA-E) | 12.6% | — |
The authors described moderately less decline with adverse events, and called for longer trials. These medicines are prescribed and monitored by specialists; this entry records the trial, not a recommendation.
Where it stands
Amyloid-beta is a target, not a treatment, and no product contains it (openFDA, 2026-10-04). PubMed returns 91,174 records for amyloid beta, 515 tagged as randomised controlled trials.
Related reading
Another peptide that forms amyloid — in the pancreas rather than the brain — is amylin. For the compounds marketed on cognition claims, see peptides for focus and mood, and for one brain-peptide preparation with its own trial record, cerebrolysin.
