Authors reused images in three papers, concludes journal probe

Biologists are retracting three papers after the journal concluded they contain reused images, designed to represent different experiments. The authors stand by the conclusions, some of which they say have been “extensively validated.”

The Journal of Biological Chemistry used image analysis software to evaluate the images, first published at least a decade ago. Unfortunately, the raw data behind the problematic images were not available. The authors have also corrected a fourth paper in another journal, and wrote on PubPeer that they are working with journals to address concerns in three more.

The papers share two authors: Mireia Duñach at the Autonomous University of Barcelona, and Antonio García de Herreros at the Institut Hospital del Mar d’Investigacions Mèdiques. A representative of the Institut Hospital del Mar d’Investigacions Mèdiques told us it is looking into Garcia de Herreros’s work.

We’ll start with “β-Catenin N- and C-terminal tails modulate the coordinated binding of adherens junction proteins to β-catenin,” which has been cited 45 times since it was published in 2002, according to Thomson Reuters Web of Science. The retraction notice says:

This article has been withdrawn by the authors. Errors were identified in several figures. Evaluation by the Journal with image analysis software determined that images were reused to represent different experimental conditions in the GST immunoblots in Fig. 2, A and C, and the N-tail immunoblots in Fig. 3B. The raw data are no longer available to validate the information. The authors have expressed the opinion that none of these errors affect the final conclusions of this article that, according to them, have been validated.

Next, here’s the retraction notice for “Regulation of β-catenin structure and activity by tyrosine phosphorylation:”

This article has been withdrawn by the authors. Errors were identified in several figures. Evaluation by the Journal with image analysis software determined that images were reused to represent different experimental conditions in the β-catenin immunoblot in Fig. 2, the β-catenin immunoblots in Fig. 5, the E-cadherin immunoblot in Fig. 6A, the Tcf-4 immunoblot in Fig. 6B, and the β-catenin immunoblot in Fig. 7B. The raw data are no longer available to validate the information. The authors have expressed the opinion that none of these errors affect the final conclusions of this article that, according to them, have been extensively validated during these 15 years.

Published in 2001, the paper has been cited 178 times.

Here’s the retraction notice for “Presenilin-1 interacts with plakoglobin and enhances plakoglobin-Tcf-4 association. Implications for the regulation of beta-catenin/Tcf-4-dependent transcription:”

This article has been withdrawn by the authors. Errors were identified in several figures. Evaluation by the Journal with image analysis software determined that images were reused to represent different experimental conditions in the β-catenin, PS1, and E-cadherin immunoblots in Fig. 2A, the PS1 immunoblots in Fig. 4, B and D, the PS1 and GST immunoblots in Fig. 4E, the Tcf-4 and β-catenin immunoblots in Fig. 6A, the PS1 immunoblot in Fig. 6C, and the GST immunoblots in Fig. 9A. The raw data are no longer available to validate the information. The authors have expressed the opinion that none of these errors affect the final conclusions of this article that, according to them, have been extensively validated.

The 2006 paper has been cited 12 times.

All three retracted papers were questioned on PubPeer, starting in December 2015.

Kaoru Sakabe, Manager of Publishing Issues at JBC told us:

In these three cases, and all other withdrawal cases, the corresponding author made the request.  The request for the withdrawals came in on April 2016.

Sakabe declined to provide further information, noting that “the details of individual cases are kept in strict confidence.”

The authors have also issued a correction in the Journal of Cell Science for “possible duplications and/or splices” in “Upon Wnt stimulation, Rac1 activation requires Rac1 and Vav2 binding to p120-catenin:”

In Fig. 2B, panels labelled as the actin input and p120-catenin IP:p120-catenin panels were inadvertently assembled to show the same blots. Concerns were also raised about possible duplications and/or splices in Figs 4A, 4E, 5D, 6A and 6C. The authors were unable to obtain the original data in order to address these concerns and therefore repeated the experiments. The new data are shown in the figures below and have been verified by the corresponding author’s institute as supporting the original conclusions of the study. There are no changes to the figure legends, which are accurate.

The 2012 paper has been cited 11 times. The corresponding author is Gabriela Valls, then a postdoc at Universitat Autònoma de Barcelona. Her LinkedIn profile says she is currently at Novartis; we could not find current contact information.

Several more papers by Garcia de Herreros and Duñach have been questioned on PubPeer. In the thread for “Gamma-secretase-dependent and -independent effects of presenilin1 on beta-cateninTcf-4 transcriptional activity,” Duñach has posted a comment saying the authors plan to submit a correction to PLOS ONE:

The duplicated band indicated in Fig 3 is actually the same: it corresponds to the analysis of the same protein (plakoglobin) in the same cells (MEFs wild-type for PS1). The three bands of plakoglobin analysis in panels A and B were initially presented in a single figure that was split in two for clarity. The actin loading control of panel 1C (24 h) was duplicated by mistake at 48 h; we have reproduced this figure and it will submitted to the editor for correction.

The 2008 paper has been cited 13 times. A spokesperson for PLOS confirmed that the authors are in communication with the journal.

Peers have also raised concerns about figures in “Multivesicular GSK3 Sequestration upon Wnt Signaling Is Controlled by p120-Catenin/Cadherin Interaction with LRP5/6,” published in 2014 by Molecular Cell. Here, too, Duñach responded:

We have submitted to the Editor original data corresponding to all the panels of these five figures that corroborate their conclusions. We are ready to published new versions of these figures if the Editor requested us to do it.

The paper has been cited 28 times. A spokesperson for Cell Press, the journal’s publisher, told us that they are looking into concerns about the paper raised by a reader.

Finally, a peer raised concerns about a duplicated blot in “p120 Catenin-associated Fer and Fyn tyrosine kinases regulate beta-catenin Tyr-142 phosphorylation and beta-catenin-alpha-catenin Interaction” published in Molecular and Cellular Biology. Duñach wrote on PubPeer:

We have sent to the Editor a response to your concerns. In any case, we are willing to reproduce any result published in this article; all the results presented in these figures and their conclusions have been validated in other articles by our group and others.

The 2003 paper has been cited 208 times.

The director of the Institut Hospital del Mar d’Investigacions Mèdiques told us:

We are aware of retractions concerning Dr. Antonio Garcia-Herreros work. Internally this issue is managed by our Good Scientific Practices and Scientific Misconduct Committee. We are expecting the report of the committee to proceed with corrective actions.

In 2014, Duñach received an Academia Award from the Catalan Institution for Research and Advanced Studies.

We have reached out to Garcia de Herreros, Duñach, and to the Office of Research at the Autonomous University of Barcelona. We will update this post with anything else we learn.

Update, June 3rd, 5:30 PM : 

We have come across a correction from February for García de Herreros and Duñach from the Journal of Cell Science.

Here’s the notice, for “Wnt controls the transcriptional activity of Kaiso through CK1ε-dependent phosphorylation of p120-catenin:”

Some of the blots presented in Fig. 4A were inadvertently duplicated in Fig. 6D. The blots in Fig. 4A are correct as presented. The correct Fig. 6 is presented below. There are no changes to the figure legend, which is accurate. This error does not affect the conclusions of the study.

That paper, published in 2011, has been cited 21 times.

Update, June 20th, 2:19 PM eastern: 

Iván Martínez Flores, the Vicegerent de Recerca (that translates to “assistant research manager”) at the Autonomous University of Barcelona told us that the school:

is aware of the fact that Professor Duñach [has] had three papers retracted.

We started our internal procedures with the relevant authorities to analyse the situation. Once we have the report, we will implement any correction measure deemed necessary.

Like Retraction Watch? Consider making a tax-deductible contribution to support our growth. You can also follow us on Twitter, like us on Facebook, add us to your RSS reader, sign up on our homepage for an email every time there’s a new post, or subscribe to our new daily digest. Click here to review our Comments Policy. For a sneak peek at what we’re working on, click here.