A group of genetics researchers in Italy has lost a 2014 paper in PLOS ONE for a range of image problems and a glaring conflict of interest.
The article, titled “Neuronal differentiation dictates estrogen-dependent survival and ERK1/2 kinetic by means of caveolin-1,” came from a team led by Luca Colucci-D’Amato, of the Second University of Naples.
The retraction notice in PLOS ONE lays out a raft of issues with the paper, for example:
In Figure 1A, there appears to be an image splicing line between lanes 2 and 3 of the β-actin blot.
…
In Figure 4C, the P-ERK1/2 and ERK1/2 panels do not include the same numbers of lanes, and similarities were noted between data shown in lanes 1 and lane 7 of the P-ERK1/2 blot. There also appear to be vertical discontinuities in Figure 4C, P-ERK1/2 panel after lanes 2 and 6.
…
For Figure 1, the authors noted that the A1 and mesPC data were obtained using different gels/blots, and that the corresponding experimental and β-actin data were obtained from the same blot in each case.
According to the notice, the researchers provided raw data for Figure 4, which didn’t really help:
These data clarified that lanes were rearranged in preparing the ERK1/2 panel in Figure 4A and the P-ERK1/2 panels in Figure 4A and 4C, and that lane 7 in each of these panels should have been labelled as exposed to both estradiol (E2) and ICI 182–780. For the ERK1/2 panel of Figure 4C, the E2 + ICI 182–780 data are in the third lane, an additional timepoint (20 min) was included in error as the fourth lane, and lanes 5–8 are the data corresponding to the lane 3–6 labels. The raw data indicated that the wrong control data (control without ethanol) were included in lane 1 of the P-ERK1/2 panels of Figure 4A and 4C. The authors commented that similarities between lanes 1 and 7 of the P-ERK1/2 panel in Figure 4C may reflect similarities in expression levels in the two samples, or could be due to an error in the figure preparation. Per our editorial assessment the pixel data in these lanes are more similar than would be expected for different data, and the E2 + ICI 182–780 data on the raw blot image did not appear to match the data in the published figure. Hence, concerns about the lane 7 P-ERK1/2 data in Figure 4C remain unresolved.
So, too, with their efforts to clear up questions about Figure 5.
Although the authors stated in the article that they had provided all their data, that wasn’t the case:
The underlying data were not provided with this article, although the Data Availability statement reads, “All relevant data are within the paper and its Supporting Information files.” The authors commented that the underlying data are not available for Figures 1, 2A, 5, S1, S2, S3, and S6. For Figure 3, quantitative data are available but the underlying blot images are not. Underlying data supporting other results reported in the article are available from the authors, as are FACS data to support the “data not shown” statement in the Results section.
In light of the unresolved concerns about Figures 4 and 5 as well as the unavailability of some data needed to support this article’s results, the PLOS ONE Editors retract this article.
New processes check for conflicts
Then there was this:
It was noted after publication that the Academic Editor who handled this article’s peer review was affiliated with the same institution as one of the authors. We regret that this was not identified and addressed prior to the article’s publication.
All authors disagreed with the retraction.
We asked Colucci-D’Amato why, exactly, he and his colleagues disagreed with the retraction but haven’t heard back.
Meanwhile, a PLOS spokesperson told us:
We have strengthened our requirements for submissions reporting blots and gels and since July 2019 we require authors to submit original, uncropped and minimally adjusted images supporting all blot and gel results reported in an article’s figures and supporting information files.
With regard to competing interests, we ask our Academic Editors to recuse themselves from situations where potential competing interests may arise. Since this article was published, we have also implemented processes to routinely check for potential competing interests between handling editors and authors. We regret that in this case this was not identified and addressed prior to the article’s publication.
The spokesperson added:
Those updates were among several changes we have implemented in recent years in an effort to proactively address issues that arise in regard to submitted and published work. They were not made in response to any one case.
Update, 1600 UTC, 12/5/19: The authors have provided a response to PLOS ONE.
Like Retraction Watch? You can make a tax-deductible contribution to support our work, follow us on Twitter, like us on Facebook, add us to your RSS reader, or subscribe to our daily digest. If you find a retraction that’s not in our database, you can let us know here. For comments or feedback, email us at [email protected].