CoChat
Sign Up
Cited by 114Alzheimer's disease research and treatmentsNeuroinflammation and Neurodegeneration MechanismsDementia and Cognitive Impairment Research

Oligodendrocytes produce amyloid-β and contribute to plaque formation alongside neurons in Alzheimer’s disease model mice

Open access
Sasmita, Andrew Octavian, Depp, Constanze, Nazarenko, Taisiia, Sun, Ting, Siems, Sophie B., Ong, Erinne Cherisse, Nkeh, Yakum B., Böhler, Carolin +27 more·2024·Nature Neuroscience
DOI 10.1038/s41593-024-01730-3PMC PMC11374705

Published version of a preprint

Open this paper in CoChat — free accountOpen PDF

Abstract

Amyloid-β (Aβ) is thought to be neuronally derived in Alzheimer’s disease (AD). However, transcripts of amyloid precursor protein ( APP ) and amyloidogenic enzymes are equally abundant in oligodendrocytes (OLs). By cell-type-specific deletion of Bace1 in a humanized knock-in AD model, APP NLGF , we demonstrate that OLs and neurons contribute to Aβ plaque burden. For rapid plaque seeding, excitatory projection neurons must provide a threshold level of Aβ. Ultimately, our findings are relevant for AD prevention and therapeutic strategies.

AI snapshot

Written from the paper’s full text.

TL;DR

Oligodendrocytes express amyloidogenic machinery and contribute to Aβ plaque burden in a humanized APP knock-in AD mouse model, but excitatory neurons must supply a threshold level of Aβ for rapid plaque seeding. Cell-type-specific Bace1 deletion shows oligodendroglial Aβ accounts for roughly 30% of plaque burden, while neuronal Aβ drives most plaque formation.

Key findings

  • Mouse and human oligodendrocytes express APP, BACE1, PSEN1 and PSEN2 at levels similar to neurons; in human cortical tissue, approximately 50% of gray matter oligodendrocytes expressed considerable APP and BACE1 mRNA in both AD cases and controls.
  • Oligodendrocyte-specific Bace1 knockout in APPNLGF mice (OL-Bace1cKO;AD) reduced plaque burden by approximately 30% versus controls in both sexes, with the greatest decrease in plaque amount and size in the alveus.
  • Excitatory neuron-specific Bace1 knockout (ExN-Bace1cKO;AD) reduced plaque burden by 95–98% versus controls, but soluble Aβ42 remained at 14.925% ± 0.066% in cortex and 27.604% ± 0.072% in white matter, indicating residual non-excitatory-neuron Aβ production.
  • OL-ExN-Bace1cKO;AD mice developed almost no cerebral plaques and showed almost complete loss of Aβ42 compared with ExN-Bace1cKO;AD samples; heterozygous APPNLGF mice developed less than 10% plaque burden rather than 50%, and 12-month-old ExN-Bace1cKO;AD mice showed considerable plaque deposition.

Methods

  • Reanalyzed four mouse and three human single-cell/single-nucleus RNA sequencing datasets for APP, BACE1, PSEN1 and PSEN2 expression using Seurat and SCTransform; validated expression with in situ hybridization in mouse and human tissue.
  • Generated conditional Bace1 knockout APPNLGF knock-in mice using Cnp-Cre for oligodendrocytes and Nex-Cre for dorsal telencephalic excitatory neurons, with littermate controls; analyzed both sexes.
  • Used light sheet microscopy of Congo red-stained hemibrains at 6 months for plaque burden, electrochemiluminescence immunoassay for Aβ38/Aβ40/Aβ42, western blotting, electron microscopy, and in vitro oligodendrocyte cultures; statistics included unpaired two-tailed Student’s t-tests and one-way ANOVA with Tukey multiple comparison tests, with n = 5 per sex for LSM and n = 7 per group for immunoassay.

Limitations

  • The paper states that Bace1 conditional knockout could not be validated by immunocytochemistry using the 3D5 antibody; validation was instead performed by immunoblotting sorted oligodendrocytes.
  • The paper states that blinding was not possible for various imaging analyses because Bace1 conditional knockout visibly affected plaque load, and that no statistical methods were used to pre-determine sample sizes.
  • Evident from the text: Cnp-Cre showed low but nonzero neuronal recombination (cortical 0.756% ± 0.057%; hippocampal 0.468% ± 0.111% tdTomato+ neurons), and human tissue analysis used a small sample (control: one female, two male; AD: two female, two male).

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1038/s41593-024-01730-3

Cites

52 of 52 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

15 of 21 cited sources checked · 5 supported · 0 flagged · accuracy not yet measured

  • Findingown data

    In mouse and human nervous tissue, expression levels of amyloidogenic pathway genes (APP, BACE1, PSEN1 and PSEN2) are similar between neurons and oligodendrocytes.

    “transcripts of amyloid precursor protein (APP) and amyloidogenic enzymes are equally abundant in oligodendrocytes (OLs).”
    • Also stated in: § Main · plain
      “expression levels were similar between neurons and OLs”
    Quote in the paperSubstantial claim6 of 6 sources checked
    • cites [9] Depp, C. et al. Myelin dysfunction drives amyloid-β deposition in models of Alzheimer’s disease. Nature 618, 349–357 (2023). source

      The section “Myelin dysfunction affects APP metabolism” (Fig. 3; page not provided) describes APP and processing enzymes in mouse axonal swellings, not transcript abundance in oligodendrocytes versus neurons. Searches for APP, BACE1, PSEN1 and PSEN2 expression comparisons in mouse and human tissue found no passage establishing the claimed similarity. (LLM, conf=0.76)

      Experimental — accuracy not yet measured
    • cites [10] Ximerakis, M. et al. Single-cell transcriptomic profiling of the aging mouse brain. Nat. Neurosci. 22, 1696–1708 (2019). source

      The Identification of cell types section describes single-cell profiles from mouse brains, including neurons and oligodendrocytes, but does not compare expression of APP, BACE1, PSEN1 or PSEN2 between them. Searches for those genes and amyloid precursor protein found no such comparison in our copy, which is truncated; the claim about human tissue is also unaddressed. (LLM, conf=0.72)

      Experimental — accuracy not yet measured
    • cites [11] Zeisel, A. et al. Molecular architecture of the mouse nervous system. Cell 174, 999–1014 (2018). source

      The Results section on the oligodendrocyte lineage describes cell types and regional expression patterns, but does not compare APP or amyloidogenic-enzyme transcript abundance between oligodendrocytes and neurons. Searches for APP, BACE1, PSEN1, PSEN2, and amyloidogenic enzymes found no passage addressing that comparison in mouse or human tissue. Page numbers were unavailable in our copy. (LLM, conf=0.76)

      Experimental — accuracy not yet measured
    • cites [12] Zhou, Y. et al. Human and mouse single-nucleus transcriptomics reveal TREM2-dependent and TREM2-independent cellular responses in Alzheimer’s disease. Nat. Med. 26, 131–142 (2020). source

      The Results section on AD-reactive oligodendrocytes and the Discussion describe changes in myelination, metabolic-response, C4b and Serpina3n transcripts, but do not compare amyloidogenic-pathway transcript abundance between oligodendrocytes and neurons. Searches for APP, BACE1, PSEN1, PSEN2 and amyloidogenic enzymes found no passage establishing the claimed similarity in mouse or human tissue. No page numbers were available in our copy. (LLM, conf=0.75)

      Experimental — accuracy not yet measured
    • cites [13] Jäkel, S. et al. Altered human oligodendrocyte heterogeneity in multiple sclerosis. Nature 566, 543–547 (2019). source

      The paper’s Abstract (page not provided) describes oligodendrocyte subclusters and their similarities across human and mouse tissue, but does not compare amyloidogenic-gene expression in oligodendrocytes and neurons. Searches for APP, BACE1, PSEN1, PSEN2, and amyloid precursor protein found no passage addressing the claimed equal abundance. (LLM, conf=0.72)

      Experimental — accuracy not yet measured
    • cites [14] Lake, B. B. et al. Integrative single-cell analysis of transcriptional and epigenetic states in the human adult brain. Nat. Biotechnol. 36, 70–80 (2018). source

      The Results section on oligodendrocyte maturation reports myelin and glutamate-receptor gene expression, not a neuron–oligodendrocyte comparison of amyloidogenic genes (page unavailable). Searches for APP, BACE1, PSEN1, PSEN2, and amyloidogenic enzymes found no passage in our copy establishing that their transcript levels are similar in mouse or human nervous tissue. (LLM, conf=0.74)

      Experimental — accuracy not yet measured
  • Findingcitation

    The inferior colliculus, receiving limited cortical input primarily from the auditory cortex, was spared from plaque attenuation in ExN-Bace1cKO;AD mice.

    “the inferior colliculus, receiving limited cortical input (Extended Data Fig. 8c) primarily from the auditory cortex19, was spared from plaque attenuation in ExN-Bace1cKO;AD mice.”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites [19] Blackwell, J. M., Lesicko, A. M., Rao, W., De Biasi, M. & Geffen, M. N. Auditory cortex shapes sound responses in the inferior colliculus. eLife 9, e51890 (2020). source

      The Introduction (page unavailable) states that the auditory cortex sends feedback to the inferior colliculus and that the colliculus receives inputs from auditory-cortex neurons. It does not establish that auditory cortex is the primary source of the colliculus’s cortical input. Citation 19 is attached to that anatomical clause, not to the separate plaque-attenuation finding. (LLM, conf=0.78)

      Experimental — accuracy not yet measured
  • Premisecitation

    Aβ fibrillation and plaque formation follow sigmoidal growth kinetics.

    “Aβ fibrillation and plaque formation follow sigmoidal growth kinetics20,21”
    Quote in the paperSubstantial claim2 of 2 sources checked
    • cites [20] Lee, C. C., Nayak, A., Sethuraman, A., Belfort, G. & McRae, G. J. A three-stage kinetic model of amyloid fibrillation. Biophys. J. 92, 3448–3458 (2007). source

      The In vitro fibrillation kinetics section describes sigmoidal behavior in amyloid fibrillation and reports that the model fits Aβ-40 fibrillation data. It identifies Aβ-40 as a protein fragment that aggregates into plaques, but does not show that plaque formation itself follows sigmoidal kinetics; the Conclusions likewise concern fibril formation, not plaque growth. Page numbers were not available in our copy. (LLM, conf=0.86)

      Experimental — accuracy not yet measured
    • cites [21] Burgold, S., Filser, S., Dorostkar, M. M., Schmidt, B. & Herms, J. In vivo imaging reveals sigmoidal growth kinetic of β-amyloid plaques. Acta Neuropathol. Commun. 2, 30 (2014). source
  • Premisecitation

    A threshold level of Aβ accumulation is essential for plaque seeding to occur.

    “a threshold level of Aβ accumulation is essential for plaque seeding to occur.”
    Quote in the paperSubstantial claim2 of 2 sources checked
    • cites [20] Lee, C. C., Nayak, A., Sethuraman, A., Belfort, G. & McRae, G. J. A three-stage kinetic model of amyloid fibrillation. Biophys. J. 92, 3448–3458 (2007). source

      The 'Effects of seeding' section says fibril growth requires sufficient nuclei, and 'In vitro fibrillation kinetics' applies the model to Aβ-40. Neither section establishes that Aβ must accumulate to a threshold before plaque seeding can occur; the seeding example concerns insulin fibrils. Page numbers are unavailable in our copy. (LLM, conf=0.82)

      Experimental — accuracy not yet measured
    • cites [21] Burgold, S., Filser, S., Dorostkar, M. M., Schmidt, B. & Herms, J. In vivo imaging reveals sigmoidal growth kinetic of β-amyloid plaques. Acta Neuropathol. Commun. 2, 30 (2014). source
  • Findingown data

    The residual amount of white matter Aβ42 was higher in ExN-Bace1cKO;AD brains (27.604% ± 0.072%).

    “the residual amount of WM Aβ42 was higher in ExN-Bace1cKO;AD brains (27.604% ± 0.072%).”
    Quote in the paperSubstantial claim
  • Classified as probably not a substantial claim. 13
  • May combine two claims. 10

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductcitation

    No statistical methods were used to pre-determine sample sizes, but sample sizes for primary experiments (that is, quantitative LSM of cKOs and immunoassay) are similar to those shown in the previous publication.

    “No statistical methods were used to pre-determine sample sizes, but sample sizes for primary experiments (that is, quantitative LSM of cKOs and immunoassay) are similar to those shown in our previous publication9.”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [9] Depp, C. et al. Myelin dysfunction drives amyloid-β deposition in models of Alzheimer’s disease. Nature 618, 349–357 (2023). source

      The cited paper’s Methods describe LSM procedures, and its results give sample sizes for some experiments, but the retrieved passages do not say whether statistical methods were used to predetermine sample sizes. Nor do they establish that the citing paper’s quantitative LSM and immunoassay sample sizes are similar to those in this paper. Searches for the sample-size statement, cKOs, LSM and immunoassay did not resolve the comparison. (LLM, conf=0.75)

      Experimental — accuracy not yet measured

  • Classified as probably not a substantial claim.

    Premisecitation

    Myelin alterations have been implicated in Alzheimer's disease.

    “myelin alterations have been implicated in AD7–9”
    Quote in the paperSubstantial claim3 of 3 sources checked
    • cites [7] Chen, J. F. et al. Enhancing myelin renewal reverses cognitive dysfunction in a murine model of Alzheimer’s disease. Neuron 109, 2292–2307 (2021). source
    • cites [8] Mathys, H. et al. Single-cell transcriptomic analysis of Alzheimer’s disease. Nature 570, 332–337 (2019). source
    • cites [9] Depp, C. et al. Myelin dysfunction drives amyloid-β deposition in models of Alzheimer’s disease. Nature 618, 349–357 (2023). source

  • May combine two claims.

    Findingown data

    ExN-Bace1cKO;AD brains were almost devoid of insoluble Aβ, but a moderate amount (14.925% ± 0.066%) of soluble Aβ42 was detected in ExN-Bace1cKO;AD cortical tissue.

    “ExN-Bace1cKO;AD brains were almost devoid of insoluble Aβ, but a moderate amount (14.925% ± 0.066%) of soluble Aβ42 was detected in ExN-Bace1cKO;AD cortical tissue.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Cell-type-specific losses of BACE1 diminished β-CTFs in the cKOs and restored FL-APP to nearly baseline WT amounts.

    “Cell-type-specific losses of BACE1 diminished β C-terminal fragments (β-CTFs) in the cKOs and restored FL-APP to nearly baseline WT amounts.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Levels of presenilin-1 (PSEN1) remained unchanged.

    “Levels of presenilin-1 (PSEN1) remained unchanged.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    In ExN-Bace1cKO;AD mice, plaque sizes were smaller and microgliosis was markedly reduced.

    “Accordingly, plaque sizes were smaller, and microgliosis was markedly reduced”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductnone given

    No animals or data points were excluded from this study.

    “No animals or data points were excluded from this study.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductown data

    cKO of Bace1 did not result in any changes to neuronal nor axonal abundance.

    “cKO of Bace1 did not result in any changes to neuronal nor axonal abundance”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductnone given

    Selected images were randomized using a Fiji filename-randomizer plugin, and counting was done using the Fiji CellCounter plugin.

    “Selected images were randomized using a Fiji filename-randomizer plugin, and counting was done using the Fiji CellCounter plugin.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductreasoning

    Due to the visible effects that Bace1 cKO has on the plaque load, blinding was not possible for various imaging analyses.

    “Due to the visible effects that Bace1 cKO has on the plaque load, blinding was not possible for various imaging analyses.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    In all assays performed, two technical replicates of samples and calibrators were included.

    “In all assays performed, two technical replicates of samples and calibrators were included.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    Experimenters were blinded in the analysis of electron microscopy data.

    “Experimenters were blinded in the analysis of electron microscopy data.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductnone given

    Most experimental cohorts were defined by genotype, and littermate controls were analyzed.

    “Most experimental cohorts were defined by genotype, and littermate controls were analyzed.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    Immunocytochemistry using the 3D5 antibody failed to validate Bace1 cKO.

    “we failed to validate Bace1 cKO via immunocytochemistry using the 3D5 antibody”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductnone given

    Both sexes were used and indicated in the respective figures.

    “Both sexes were used and indicated in the respective figures.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductown data

    In Bace1 KO tissue, 3D5 labeling loss was observed only in mossy fibers, with no somatic BACE1 observed.

    “We only observed a loss of 3D5 labeling in the mossy fibers in Bace1 KO (Extended Data Fig. 5c), yet no somatic BACE1 was observed”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    Statistical tests were chosen based on tests for normality.

    “Statistical tests were chosen based on tests for normality.”
    Quote in the paperNot a substantial claim
17 of 17

  • The paper relies on this without support.

    Bace1 is the rate-limiting enzyme in Aβ generation.

    “conditionally knock out Bace1 (Bace1 cKO), the rate-limiting enzyme in Aβ generation”

  • The paper relies on this without support.

    Oligodendrocyte lineage cells are abundantly present in both gray matter and white matter.

    “Because OL lineage cells are abundantly present in both gray matter and white matter (WM)”
2 of 2

Cites

Flagged first, then by citations

6 of the 58 references couldn’t be matched to a known paper.

Cited by

Sorted by citations

Related

Authors
Sasmita, A. O., Depp, C., Nazarenko, T., Sun, T., Siems, S. B., Ong, E. C., Nkeh, Y. B., Böhler, C., Yu, X., Bues, B., Evangelista, L., Mao, S., Morgado, B., Wu, Z., Ruhwedel, T., Subramanian, S., Börensen, F., Overhoff, K., Spieth, L., Berghoff, S. A., Sadleir, K. R., Vassar, R., Eggert, S., Goebbels, S., Saito, T., Saido, T., Saher, G., Möbius, W., Castelo-Branco, G., Klafki, H.-W., Wirths, O., Wiltfang, J., Jäkel, S., Yan, R., & Nave, K.-A.
Affiliations
  • Max Planck Institute for Multidisciplinary Sciences (DE)
  • Karolinska Institutet (SE)
  • University College Cork (IE)
  • LMU Klinikum (DE)
  • Northwestern University (US)
  • RIKEN Center for Brain Science (JP)
  • Universitätsmedizin Göttingen (DE)
  • German Center for Neurodegenerative Diseases (DE)
  • UConn Health (US)
Corresponding author
  • Sasmita, Andrew Octavian
  • Depp, Constanze
  • Nave, Klaus-Armin
ORCID
Venue
Nature Neuroscience
Publisher
Springer Science and Business Media LLC
Published
2024-08-05
Volume
27
Issue
9
Pages
1668-1674
ISSN
  • 1097-6256 (ISSN-L)
  • 1546-1726
DOI
10.1038/s41593-024-01730-3
PMC
PMC11374705
OpenAlex
W4401339790
Cited by
114
References
58
Citations by year
2026: 42 · 2025: 56 · 2024: 16
FWCI
27.24
Type
journal article
Language
en
Open access
Licence
Locations
  • Nature Neuroscience (publishedVersion, cc-by)
  • bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion)
  • PubMed (publishedVersion)
  • PubMed Central (submittedVersion, cc-by)
  • Open access LMU (Ludwid Maxmilian's Universitat Munchen) (submittedVersion, cc-by)
  • DZNE Pub (submittedVersion, cc-by)
  • GoeScholar The Publication Server of the Georg-August-Universität Göttingen (Georg-August-Universität Göttingen) (submittedVersion, cc-by)
Funders
  • Deutsche Forschungsgemeinschaft (TRR-274)
  • Max-Planck-Gesellschaft
Topics
  • Health Sciences › Medicine › Physiology › Alzheimer's disease research and treatments
  • Life Sciences › Neuroscience › Neurology › Neuroinflammation and Neurodegeneration Mechanisms
  • Health Sciences › Medicine › Psychiatry and Mental health › Dementia and Cognitive Impairment Research
Keywords
oligodendrocytes, amyloid-beta, Alzheimer's disease, amyloid plaques, amyloid precursor protein, BACE1, humanized mice
MeSH terms
Alzheimer Disease, Animals, Disease Models, Animal, Humans, Mice, Transgenic, Neurons, Oligodendroglia, Amyloid beta-Peptides, Aspartic Acid Endopeptidases, Amyloid beta-Protein Precursor, Mice, Amyloid Precursor Protein Secretases, Plaque, Amyloid

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-11.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-11.

Cited sources are checked by gpt-6-sol.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review