# RETRACTED ARTICLE: A specific amyloid-β protein assembly in the brain impairs memory

> **Reliability concern:** This paper was retracted on 2024-06-24. See notice: 10.1038/s41586-024-07691-8.

- **DOI:** [10.1038/nature04533](https://doi.org/10.1038/nature04533)
- **Rating:** Sound (3.7 out of 5)
- **Verdict:** A field-shaping oligomer study linking Aβ*56 to memory impairment, constrained by indirect endogenous causality.
- **Framework version:** v5.5
- **Pipeline version:** v0.4
- **Evaluation date:** 2026-08-07
- **Evaluation page:** https://www.nabu.science/paper/10.1038%2Fnature04533

## Key claims

1. memory deficits in middle-aged Tg2576 mice are caused by the extracellular accumulation of a 56-kDa soluble amyloid-β assembly, which we term Ab*56
   - Support: Tentative
   - Type: Primary finding
   - Basis: mouse evidence is correlational, with causality inferred indirectly from rat infusion
   - Location: Abstract; Results, Figs. 1, 3, 4
2. Ab*56 purified from the brains of impaired Tg2576 mice disrupts memory when administered to young rats
   - Support: Supported
   - Type: Secondary
   - Basis: rat probe-trial impairment with acquisition and visible-platform controls
   - Location: Abstract; Results, Fig. 4b–e
3. the 56-kDa assembly correlating substantially better than the 40-kDa assembly
   - Support: Supported
   - Type: Secondary
   - Basis: reported inverse correlations favor 56-kDa but 40-kDa also correlates
   - Location: Results, Fig. 3g–h
4. Ab*56 impairs memory independently of plaques or neuronal loss
   - Support: Supported
   - Type: Secondary
   - Basis: timing before abundant plaques and model context support plaque-independent impairment within Tg2576
   - Location: Abstract; Introduction; Results, Fig. 1
5. trimers are the fundamental Ab assembly unit in vivo
   - Support: Tentative
   - Type: Secondary
   - Basis: inferred from denaturation and migration patterns rather than direct structural determination
   - Location: Results, Fig. 2a–c

## Strengths

- **Specific candidate for soluble Aβ toxicity** The article moves from a broad soluble-oligomer hypothesis to a named 56-kDa assembly with temporal emergence at the memory-decline window and stronger correlation with impairment than the 40-kDa band. This gives the field a concrete candidate rather than a general class of species.
  - Dimension: contribution
  - Location: Results, Figs. 1c and 3g–h
- **Multi-method biochemical characterization strategy** Aβ*56 is assessed with multiple antibodies, denaturation behavior, size-exclusion chromatography, silver staining, and mass spectrometry as summarized in the main text. These methods reduce the likelihood that the 56-kDa band is simply APP, an APP cleavage product, or an SDS-PAGE artifact.
  - Dimension: methodological rigour
  - Location: Results, Fig. 2; Results, Fig. 4a
- **Behavioral controls distinguish retention impairment** The rat experiment reports no effect on hidden-platform acquisition or visible-platform performance, while the probe test shows impaired spatial bias after Aβ*56 infusion. The ten-day retest without infusion is presented as evidence against lasting debilitation.
  - Dimension: methodological rigour
  - Location: Results, Fig. 4b–e; Supplementary Fig. 6 described in Results

## Limitations

- **Endogenous causality remains indirectly tested** The paper’s strongest claim is that Aβ*56 causes memory decline in middle-aged Tg2576 mice, but the Tg2576 evidence is temporal and correlational. The causal experiment shows that exogenous purified Aβ*56 can impair memory in rats, not that endogenous Aβ*56 is necessary or sufficient in Tg2576 mice.
  - Dimension: contribution
  - Location: Abstract; Results before and after Fig. 4
- **Infusion dose lacks physiological benchmark** The rat study infuses 10 µl of 0.85 mM Aβ*56, but the article does not benchmark this exposure against endogenous Aβ*56 concentrations in Tg2576 brain. That gap limits interpretation of how closely the rat experiment models the endogenous mouse condition.
  - Dimension: methodological rigour
  - Location: Methods, Behavioral tests; Results, Fig. 4
- **Animal ethics statement is absent** The supplied text describes transgenic mouse testing, rat cannula implantation, and intracerebroventricular infusion, but it does not include an explicit animal ethics approval statement. This is recorded as a conduct note rather than a severe reliability concern, given the publication era and absence of a visible protocol violation.
  - Dimension: methodological rigour
  - Location: Methods, Behavioral tests; Author Information

## Reliability checks

Concerns (4 of 4 checks).

- **Internal coherence:** Noted. The experimental narrative is internally coherent, moving from age-course detection and correlations to purified Aβ*56 infusion. The main internal-coherence concern is claim proportionality: the paper’s causal wording about Aβ*56 as the cause or principal cause of Tg2576 memory decline is stronger than the mouse correlation plus rat infusion design directly proves.
- **Research conduct:** Noted. The supplied text describes animal behavioral testing, cannula implantation, and intracerebroventricular infusion but does not include an explicit animal ethics approval statement. Competing financial interests are disclosed as present, but the details are linked externally and are treated as an assessability gap rather than a conduct concern.
- **Reference integrity:** Clean. 27 of 27 checkable references verified
  - 28 references in manuscript
  - 1 have no canonical index record — counted, but not index-checkable
  - 3 references confirmed by manual review
  - References not verified automatically are resolved manually within 24 hours.

## Post-publication record

- **Status:** Concern
- **Summary:** This paper was retracted on 2024-06-24. See notice: 10.1038/s41586-024-07691-8.
- **Last checked:** 2026-10-17
