Retraction Watch

Tracking retractions as a window into the scientific process

Diabetes researcher logged 1 retraction, 3 correx, after PubPeer comments

with 5 comments

Kathrin Maelder

Kathrin Maedler

A journal has retracted a paper by a leading diabetes researcher — who has also issued three corrections — after questions about her research were raised on PubPeer.

Kathrin Maedler — who works at the University of Bremen in Germany — corrected another paper in 2014. All of the notices are dated from 2015, and all describe issues with figures.

The ongoing comments have led Maedler to carefully look through her original data, according to a statement she emailed us:

Mistakes have indeed occurred by reusing loading controls from different experiments, which was not apparent to me at the time of publication. We have immediately put the appropriate actions in place to avoid such in the future. Several replicate experiments were provided to the respective journals, which fully confirm scientifically correct data, and we could correct the wrongly placed loading controls in our publications. Journals investigated all provided full blots of the experiments, which led to correction of the data.

I am thankful to the journal editors to have had the opportunity to correct the mistakes publicly and these corrections are in no way altering interpretation of our results and the conclusions of the original papers.

She affirmed that the conclusions in her papers with notices remain valid — including in the newly retracted paper.

Maedler told us that she initially requested a correction for a 2011 Journal of Biological Chemistry paper, but the journal retracted it with this notice, citing duplicated images:

This article has been retracted by the publisher. Fig. 1E contained images that had been reused from the following references (Glas, R., Sauter, N. S., Schulthess, F. T., Shu, L., Oberholzer, J., and Maedler, K. (2009) Purinergic P2X7 receptors regulate secretion of interleukin-1 receptor antagonist and β cell function and survival. Diabetologia52, 1579–1588 and Maedler, K., Schumann, D. M., Sauter, N., Ellingsgaard, H., Bosco, D., Baertschiger, R., Iwakura, Y., Oberholzer, J., Wollheim, C. B., Gauthier, B. R., and Donath, M. Y. (2006) Low concentration of interleukin-1β induces FLICE-inhibitory protein-mediated β-cell proliferation in human pancreatic islets. Diabetes55, 2713–2722), and these reused images represented different experimental conditions.

Neutralizing interleukin 1β (IL-1β) induces β-cell survival by maintaining PDX1 protein nuclear localization” has been cited 17 times, according to Thomson Reuters Web of Science.

According to Maedler’s statement, the duplications were “mistakes”:

Errors in the images in Fig.1E occurred in the Histone H3-band and the Tubulin-band. Both bands do originally not belong as house keeping controls to this experiment from 2007, but were duplicated from another experiment. I have not detected the mistakes myself at the time of submission and even thereafter. In light of erroneously misplaced loading controls and because I was not confident with loading controls for the published experiment from Fig.1E panel, we provided three independent replica to JBC, which show similar findings and support our published data, and asked for a correction of Fig.1E from JBC…Since then, many independent labs reproduced our data upon our publication and PDX1 shuttling and reduction in diabetes was confirmed in numerous in vivo and in vitro models…

That wasn’t enough for the journal, apparently, Maedler noted:

Despite presenting full blots of multiple parallel experiments as well as presentation of the same scientific result using multiple model systems, which fully confirm our published data, the correction was not accepted by JBC and the paper was retracted.

Kaoru Sakabe, the publishing manager of JBC, could not comment on this specific case, but told us that:

…corrections are deemed appropriate by the journal’s editors when after an internal investigation, it is determined that the error does not affect the results or conclusions of the work.

Maedler said that she feels she is being targeted based on the response she received following the JBC retraction:

Only minutes after the appearance of the retraction on the JBC website, the retraction has been posted as comments on PubPeer, LabJournal and Retraction Watch and were copied-pasted again and again. Obviously the anonymous commenters are constantly following our public events. The fact that the images were wrong in its presentation but are fully correct scientifically raises the question of its detection. Thus, allegations raised against me suspect, that they are rather part of a targeted violent campaign against my person, since they in no case uncover wrong published data, especially since many independent laboratories have replicated the data.

Maedler added that the university where she completed her PhD has not found evidence of misconduct:

Various papers from my own doctoral thesis were also accused. The University of Zurich has initiated investigation of the work conducted there and has already finished its investigation with the conclusion, that there is no reasonable suspicion for any misconduct and not a single case reported indicates any data falsification. The case was closed.

We ran that statement by a spokesperson for the University of Zurich, who told us:

We can confirm what Prof. Mädler said. The Executive Board of the University of Zurich has determined that no allegations of scientific misconduct against Prof. Kathrin Mädler can be sustained. The proceedings on suspicion of academic dishonesty have been concluded.

Maedler noted that in 2013, while at her current post at the University of Bremen:

A previous accusation against me was filed internally, based on a symposium presentation. In multiple meetings I presented all original data together with multiple repetitions, which strangely none of the accusing colleagues were interested in. At the end, both ombudspersons of the university stopped the case and made clear that no misconduct was performed.

A current ombudsperson at the University of Bremen to verified this statement, and told us:

The university is currently investigating all information concerning Karthrin Maedler’s work…previous accusations were evaluated as being no misconduct.

Here are the papers that have been corrected, all of which in 2015. First, we’re copying the notice for a PLOS ONE paper, “Deletion of the Mitochondrial Flavoprotein Apoptosis Inducing Factor (AIF) Induces β-Cell Apoptosis and Impairs β-Cell Mass:”

Panel 1 in Fig. 1A is incorrect. Please view a correct version of Fig. 1 here.

The 2015 paper has not yet been cited.

Here’s the correction for “Decreased TCF7L2 protein levels in type 2 diabetes mellitus correlate with downregulation of GIP- and GLP-1 receptors and impaired beta-cell function” in Human Molecular Genetics, also issued in 2015:

In this article, Figure 2F was incorrect. The correct panel is shown below. The authors sincerely apologise for this error.

F1.small

The 2009 paper has been cited 119 times.

And, finally, here’s the correction notice for “Purinergic P2X7 receptors regulate secretion of interleukin-1 receptor antagonist and beta cell function and survival,” in Diabetologia:

The authors have discovered that the incorrect β-actin lane was included in the right-hand panel of the western blot shown in Fig. 2j. Because they no longer have access to the β-actin blot source file from 2006, the blot is corrected here using lanes from one of the other two replicates performed at the time of the original experiment. The replicate of the experiment shows the same results as the originally published blots and the correction does not affect the results or conclusion of the paper. The figure legend has been corrected here to reflect the fact that the culture time was from 30 min to 48 h, not 24 h as stated originally.

https://static-content.springer.com/image/art%3A10.1007%2Fs00125-015-3681-x/MediaObjects/125_2015_3681_Figa_HTML.gif

(j) The rise in IL-1Ra was confirmed by a time-dependent experiment using isolated human islets exposed to 11.1 and 33.3 mmol/l glucose and 0.5 mmol/l palmitate (NEFA) for 30 min to 48 h. Western blot analysis was performed for IL-1Ra and β-actin as loading control on the same blot. Representative blots from three experiments from three donors are shown

That 2009 paper has been cited 46 times.

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.

Comments
  • Feo Takahari July 4, 2016 at 6:15 pm

    Mistakes have occurred, eh? She’s phrasing it like they’re earthquakes or tornadoes, something that doesn’t have an individual cause and can’t be traced back to a person. Who made these mistakes–her, or some other part of her nebulous “we”?

  • Gary July 5, 2016 at 3:39 am

    “the blot is corrected here using lanes from one of the other two replicates performed at the time of the original experiment”
    It does seem that the experiment was performed in triplicate – which suggests that using the wrong loading control was a genuine mistake…

  • fernandopessoa July 5, 2016 at 6:40 am

    J Clin Invest. 2002 Sep;110(6):851-60.
    Glucose-induced beta cell production of IL-1beta contributes to glucotoxicity in human pancreatic islets.
    Maedler K1, Sergeev P, Ris F, Oberholzer J, Joller-Jemelka HI, Spinas GA, Kaiser N, Halban PA, Donath MY.
    Author information
    1Division of Endocrinology and Diabetes, University Hospital, Zurich, Switzerland.
    https://pubpeer.com/publications/479EF98E32BC94B688B66F5A5329D2#fb53772

    Figure 4F J Clin Invest. 2002 Sep;110(6):851-60 and figure 4B PNAS 99,8236-8241.
    http://i.imgur.com/iKfxRDA.jpg

    For reference PNAS 99,8236-8241.
    https://pubpeer.com/publications/4CC5CD97AFCDE13689DA648926DAB7#fb53771

  • fernandopessoa July 5, 2016 at 11:43 am

    Proc Natl Acad Sci U S A. 2004 May 25;101(21):8138-43. Epub 2004 May 12.
    Leptin modulates beta cell expression of IL-1 receptor antagonist and release of IL-1beta in human islets.
    Maedler K1, Sergeev P, Ehses JA, Mathe Z, Bosco D, Berney T, Dayer JM, Reinecke M, Halban PA, Donath MY.
    Author information
    1Division of Endocrinology and Diabetes, University Hospital, CH8091 Zurich, Switzerland.

    Figure 2B.
    http://imgur.com/6ZFwU0K

    Figure 4A.
    http://imgur.com/T2bJoC0

  • Bibby July 5, 2016 at 4:44 pm

    I think Kathrin Maedler should be commended for such action. Mistakes were made, lessons learned, and responsibility was taken.

  • Post a comment

    Threaded commenting powered by interconnect/it code.