Author Correction: APOE4 impairs myelination via cholesterol dysregulation in oligodendrocytes

Joel Blanchard, Leyla Anne Akay, José Dávila-Velderrain, Djuna von Maydell, Hansruedi Mathys, Shawn M. Davidson, Audrey H. Effenberger, Chih‐Yu Chen, Kristal Maner-Smith, Ihab Hajjar, Eric A. Ortlund, Michael Bula, Emre Agbas, Ayesha P. Ng, Xueqiao Jiang, Martin C. Kahn, Cristina Blanco‐Duque, Nicolas Lavoie, Liwang Liu, Ricardo Reyes‐Chilpa, Yuan-Ta Lin, Tak Ko, Lea R’Bibo, William T. Ralvenius, David A. Bennett, Hugh P. Cam, Manolis Kellis, Li‐Huei TsaiNature 2024 · DOI · OpenAlex

This document is an Author Correction published in Nature (December 2024) regarding a high-profile 2022 study on the genetic drivers of Alzheimer’s disease. The correction clarifies the provenance of human single-nucleus RNA sequencing (snRNA-seq) data, noting that a subset of the samples used in the original analysis was reprocessed from a previously published dataset rather than being entirely novel.

TL;DR

This formal amendment corrects the Methods section of a study investigating how the APOE4 risk variant disrupts cholesterol metabolism in oligodendrocytes. It clarifies that 8 of the 32 human brain samples used for transcriptomic analysis were reprocessed from a 2019 study to ensure transparency in data sourcing and experimental overlap.

What Problem Does This Paper Address?

The original paper (published in November 2022) explores the mechanistic link between the APOE4 allele—the strongest genetic risk factor for late-onset Alzheimer’s—and impaired myelination in the brain. Specifically, it tackles how APOE4 causes cholesterol to be sequestered within oligodendrocytes (the cells responsible for insulating neurons with myelin), thereby starving the brain of the lipids necessary for healthy white matter maintenance.

This specific Correction addresses a transparency gap in the original publication's "Methods" section. In large-scale "omic" studies, researchers often combine newly generated data with existing datasets to increase statistical power. However, the original text failed to explicitly state that a portion of the human snRNA-seq data was shared with a prior landmark study. This correction ensures the scientific record accurately reflects which data were newly generated versus reprocessed.

Methods

The study utilized two primary techniques for transcriptomic profiling of human brain tissue:

  • Isolation of nuclei from frozen post-mortem brain tissue: A standard but delicate procedure to extract genetic material from preserved human samples.
  • Droplet-based snRNA-seq: A high-throughput method (likely using platforms like 10x Genomics) to sequence the RNA of individual cell nuclei, allowing researchers to identify cell-type-specific changes in gene expression.

The correction specifies that for these methods, the authors have now added references to account for the "partial overlap" of samples. The 8 reprocessed samples were integrated and re-analyzed alongside the 24 newly generated samples to ensure a consistent computational pipeline.

Datasets

The study's human transcriptomic analysis relied on a total of 32 samples:

  • Mathys et al. (2019): 8 samples were taken from this previous study (Nature 570, 332–337). These samples were reprocessed using the current study's specific bioinformatics pipeline.
  • Newly Generated Data: 24 samples were unique to the 2022 study.
  • Reference: The primary external dataset is cited as Mathys, H. et al., "Single-cell transcriptomic analysis of Alzheimer’s disease."

Key Insights and Ideas

  • Data Provenance and Transparency: The primary "aha moment" of this correction is the reminder that in the era of "Big Data" biology, clear documentation of data reuse is critical. Reprocessing old data with new algorithms or alongside new samples is a valid and powerful approach, but it must be cited to avoid the appearance of "double-counting" samples in meta-analyses.
  • Consistency in Reprocessing: By reprocessing the 8 older samples "together with the newly generated data," the authors minimized batch effects, ensuring that the biological signals (cholesterol dysregulation) were not artifacts of different sequencing runs or software versions.

Key Results

The correction itself does not change the biological conclusions of the original paper—that APOE4 impairs myelination via cholesterol dysregulation. Instead, the quantitative result of this amendment is the clarification of the sample size:

  • Total Human snRNA-seq N: 32 samples.
  • Data Split: 25% (8/32) reprocessed from 2019; 75% (24/32) newly generated.
  • Status: The Methods sections in the HTML and PDF versions of the original article have been updated to reflect this clarification.

Limitations and Future Work

Not discussed in the paper. (As this is a formal correction notice, it focuses strictly on the data provenance error and does not propose new experimental directions or acknowledge further biological limitations beyond the scope of the correction).

Summary updated

BibTeX
@inproceedings{blanchard2024author,
  author = {Joel Blanchard and Leyla Anne Akay and José Dávila-Velderrain and Djuna von Maydell and Hansruedi Mathys and Shawn M. Davidson and Audrey H. Effenberger and Chih‐Yu Chen and Kristal Maner-Smith and Ihab Hajjar and Eric A. Ortlund and Michael Bula and Emre Agbas and Ayesha P. Ng and Xueqiao Jiang and Martin C. Kahn and Cristina Blanco‐Duque and Nicolas Lavoie and Liwang Liu and Ricardo Reyes‐Chilpa and Yuan-Ta Lin and Tak Ko and Lea R’Bibo and William T. Ralvenius and David A. Bennett and Hugh P. Cam and Manolis Kellis and Li‐Huei Tsai},
  title = {{Author Correction: APOE4 impairs myelination via cholesterol dysregulation in oligodendrocytes}},
  booktitle = {Nature},
  year = {2024},
  doi = {10.1038/s41586-024-08434-5},
}
Figure