Alzheimer’s paper retracted over apparent image duplication

A paper touting a cellular protein’s ability to reduce amyloid burden in Alzheimer’s disease models has been retracted. A pair of images in the paper appear to be duplicates, according to the retraction notice, posted 11 September in Neurobiology of Disease.

Eva Carro led the 2009 study as a researcher at the Hospital Universitario 12 de Octubre and the Instituto de Salud de Charles III (ISCIII), a government research organization. The work was funded with grants from ISCIII’s Fund for Research in Health Sciences and the Community of Madrid. This marks the second retraction for Carro.

Concerns about the 2009 paper first emerged a decade after publication, nearly six years before the retraction. The paper was cited 63 times prior to the retraction, according to Clarivate’s Web of Science.

I

n the newly retracted study, the authors used a lentiviral vector to manipulate the levels of a protein called gelsolin in a transgenic mouse model of Alzheimer’s disease; the protein, which regulates the cytoskeleton, is known to bind amyloid beta. They also measured gelsolin in the choroid plexus of postmortem human brains and in the cerebrospinal fluid of living people, comparing participants who have Alzheimer’s disease with healthy controls. The purportedly duplicated images appear in a figure validating the viral vector.

An anonymous commenter on PubPeer pointed out similarities between the two images in a post dated November 2019. Data sleuth Kevin Patrick, who posts on the forum under the alias Actinopolyspora biskrensis, chimed in minutes later with similar concerns.

Carro responded the following month, posting images to PubPeer that she claimed depicted the correct gel bands. These new figures did not answer the sleuths’ questions, Patrick told The Transmitter in an interview. He says the images were unlabeled and that Carro uploaded the same file twice. When he followed up with Carro, she provided labeled figures but no way to validate that they contained the proper images.

Patrick says that journals often do not become aware of issues raised on PubPeer unless the commenters contact the editors. Still, he says he hesitated to alert the journal to his findings until February 2025 because he made his initial comment back when he was fairly new to sleuthing. “I wasn’t trying to run it [up] the flagpole and then be told I was an idiot,” he says.

The eventual retraction notice confirmed the sleuths’ findings and questioned whether the paper’s experiments and methods, including how healthy participants were recruited for lumbar puncture, had received ethical approval. Per the notice, Carro’s answers to the journal editors’ questions about ethics did not satisfy the journal’s concerns. Given that she also did not provide the source data to correct the apparent image duplication, the editors opted to retract the article.

Erwan Bezard, editor-in-chief of Neurobiology of Disease, declined to comment on the retraction in an email to The Transmitter. Carro did not respond to a request for comment, nor did Desiree Antequera, co-first author of the study, who most recently collaborated with Carro on an Analytical Chemistry paper published in August. Teo Vargas, the other first author, could not be reached for comment.

This is not the first paper from Carro and her colleagues to raise suspicions of image manipulation. In 2021, BioMed Research International retracted a 2010 gelsolin-related paper. Journal editors noted the resemblance between two images in the paper and those from the lab’s other work, including the 2009 paper. In May 2020, Molecular Neurobiology issued corrections to two 2018 articles by Carro because of apparent image duplications.