Retraction Watch

Tracking retractions as a window into the scientific process

In retraction notice, Bulfone-Paus “declares” data and conclusions confirmed; journals accept six more retractions

with 13 comments

Silvia Bulfone-Paus

There was more news today about papers co-authored by Silvia Bulfone-Paus, whose lab at Research Centre Borstel has been under investigation for scientific misconduct.

The EMBO Journal, which we reported last month had accepted the retraction of a 2005 Bulfone-Paus paper that has been cited 37 times, published the retraction notice for the study today:

Eight of the authors (ZO, LT, UM, PB, CB, DA, RP and SB-P) wish to retract this paper, following an independent formal investigation initiated by the Research Center Borstel into scientific misconduct (see http://www.fz-borstel.de/cms/index.php?id=1). The investigation concluded that multiple figures contained PCR and western blot duplications and possible other manipulations (Figures 2A, 3A, 4A, 5, 7A and 7C, Supplementary Figures S1A, S2A and S2B, unconfirmed: Figure 1C). The above signed declare that Vadim Budagian and Elena Bulanova conducted these experiments and generated the figures. The authors declare that key experiments presented in the majority of these figures were recently reproduced and that the results confirmed the experimental data and the conclusions drawn from them. However, due to these unacceptable irregularities, the listed authors retract this paper in its entirety and regret any adverse consequences that may have resulted from its publication. Vadim Budagian and Elena Bulanova declined to sign the retraction.

As retraction notices go, this is better than average, we must say. It’s certainly better than the many opaque notices we’ve seen. But there’s a line in there that Karin Wiebauer and David Hardman, both of whom brought the notice to our attention, found curious:

The authors declare that key experiments presented in the majority of these figures were recently reproduced and that the results confirmed the experimental data and the conclusions drawn from them.

We see this every now and then, authors retracting a paper but claiming its underlying science is still correct. But Wiebauer and Hardman, who have both been following the Bulfone-Paus case for a while, wondered what exactly this meant. Had the journal editors seen the reproduced data, which had been manipulated in the original paper?

We thought that was a good question, so we asked EMBO Journal editor Bernd Pulverer. He responded:

We did not formally investigate this case at the journal and we have not seen this data, as it does not affect the retraction.

We beg to differ. The journal has just allowed someone whose lab is being investigated for fraud — and who has now had 12 retractions accepted, see below — to declare that data are valid, without having seen it. Sorry, that’s just not right. Let them retract it, and try to submit it elsewhere, subject to peer review, with lots of scrutiny. (We wanted to know if the team was planning to submit it elsewhere, too, and have a message in to Bulfone-Paus. We’ll update with anything we hear back.)

Speaking of those 12 retractions, we’d reported on six accepted retractions last month. Today, Borstel announced that there were six more. They are, with Thomson Scientific Web of Knowledge citation data:

Bulanova, E., Budagian, V., Duitman, E., Orinska, Z., Krause, H., Rückert, R., Reiling, N., and Bulfone-Paus, S. Soluble IL-15Rα is generated by alternative splicing or proteolytic cleavage and forms functional complexes with Il-15. J. Biol. Chem. 282:13167-13179 (2007, cited 37 times)

Budagian, V., Bulanova, E., Orinska, Z., Duitman, E., Brandt, K., Ludwig, A., Hartmann, D., Lemke, G., Saftig, P., and Bulfone-Paus, S. Soluble Axl is generated by ADAM10-dependent cleavage and associates with Gas6 in mouse serum. Mol. Cell. Biol. 25:9324-9339 (2005, cited 32 times)

Bulfone-Paus, S., Bulanova, E., Pohl, T, Budagian, V., Dürkop, H., Rückert, R., Kunzendorf, U., Paus, R., and Krause, H. Death deflected: IL-15 inhibits TNF-α-mediated apoptosis in fibroblasts  by TRAF2 recruitment to the IL-15Rα chain. FASEB J. 13:1575-1585 (1999, cited 118 times)

Budagian, V., Bulanova, E., Brovko, L., Orinska, Z., Fayad, R., Paus, R., and Bulfone-Paus, S. Signaling through P2X7 receptor in human T cells involves p56LCK, MAP kinases, and transcription factors AP-1 and NF-kB. J. Biol. Chem. 278:1549-1560 (2003, cited 67 times)

Budagian, V., Bulanova, E., Orinska, Z., Ludwig, A., Rose-John, S., Saftig, P., Borden, E., and Bulfone-Paus, S. Natural soluble IL-15Rα is generated by cleavage that involves the tumor necrosis factor-α-converting enzyme (TACE/ADAM17). J. Biol. Chem. 279:40368-40375 (2004, cited 47 times)

Budagian, V., Bulanova, E., Orinska, Z., Pohl, T., Borden, E., Silverman, R., and Bulfone-Paus, S. Reverse signaling through membrane-bound Interleukin-15. J. Biol. Chem. 279:42192-42201 (2004, cited 31 times)

We should also note that while there is no retraction notice text yet for a previously accepted retraction, in the Scandinavian Journal of Immunology, the paper’s abstract and PDF are both marked as retracted. The PubMed abstract doesn’t say it’s been retracted, either.

There will be more news on all of this, we’re sure. Twelve retractions is a lot, although Naoki Mori is up to 14.

Written by Ivan Oransky

February 2nd, 2011 at 5:55 pm

Comments
  • e-Patient Dave February 2, 2011 at 6:03 pm

    Thanks – the point on claiming the results were replicated is a good teachable. I’m a case in point: I myself presumed that the editors’ permission to print it signaled acceptance of the statement.

    Is there a typo (a missing negative) in your penultimate paragraph? I’m guessing it should be “while there is *no* retraction notice text.”

    • ivanoransky February 2, 2011 at 6:06 pm

      Indeed there is a missing word, thanks. Fixed.

  • ph Starck February 2, 2011 at 6:39 pm

    In my opinion it is crucial to note that in this list there is one paper that has nothing to do with Budagian and Bulanova:

    “Bulfone-Paus, S., Bulanova, E., Pohl, T, Budagian, V., Dürkop, H., Rückert, R., Kunzendorf, U., Paus, R., and Krause, H. Death deflected: IL-15 inhibits TNF-α-mediated apoptosis in fibroblasts by TRAF2 recruitment to the IL-15Rα chain”

    In my opinion we are going to see this more and more: if you were used to sloppy and fraudulent science as a grad student or postdoc, it is not surprising that you will attract and finally hire someone like you once you start your own group.
    Depressing.

    • ph Starck February 2, 2011 at 6:42 pm

      sorry, I meant Budagian and Bulanova as first authors, wrote too fast! She is actually first and corresponding author on that paper, not student, but it is still impossible that she did not know what was going on.

      • ivanoransky February 2, 2011 at 6:54 pm

        Thanks for clarifying.

    • ivanoransky February 2, 2011 at 6:43 pm

      Thanks for the comment. Can you explain what you mean? Budagian and Bulanova are fourth and second authors, respectively, of this paper.

  • ph Starck February 3, 2011 at 5:12 am

    So let me write this down in a clearer way.
    To me there are two main problems connected to the 1999 paper published in the Faseb J.:

    1. Dr. Bulfone-Paus participated actively to the falsification. She is first and corresponding author, therefore her role in the fraud in this case could be extremely relevant. As far as I know, in the investigation the entire fault was on Budagian and Bulanova.
    2. Dr. Bulfone-Paus had no idea of what was going on, because she did not produce the data, that were still made up by Budagian and Bulanova. She simply put her name as first so that she could get her habilitation.

    In both cases, we can envision a serious misconduct.

    In my first comment I got confused because I did not know exactly what it is a “habilitation”, so I thought she was still at a postdoc level.
    Still, I stand by my point about people performing sloppy science as student and keeping up with such behaviour as group leaders.

    Sorry for the confusion generated by previous comments!

  • Prof. Dr. Joerg Zwirner February 3, 2011 at 6:10 am

    Good point, ph Starck. Bulfone-Paus is in deep trouble!
    EMBO Journal editor Bernd Pulverer has done science a disservice. Without good cause he provided platform to people (shall we still call them scientists?) who gambled away their credibility. Key experiments reproduced and confirmed? No proof? No evidence? Come on! Bulfone-Paus is not worthy of our trust anymore. Sadly enough, EMBO J appears to be part of the problem not the solution.

  • BTC February 3, 2011 at 12:53 pm

    It seems that with the statement, “The data have been reproduced,” the retraction notice is, now, more of an afterthought. To be completely honest, the retraction statement itself would be transparent but for that statement above.

    Like I said, the statement sort of nullifies the retraction in my opinion. Sure, somebody made stuff up, it got published, we found out it was fake, but we repeated the fake stuff anyway.

  • RM February 3, 2011 at 1:07 pm

    The problem with such non-unanimous retraction statements is that they suggest that the blame lies with the scientist(s) that did not sign it. But that is unfortunately not true, and there is even some selection pressure against the opposite, as the other authors have a clear benefit in loading the blame onto mere postdocs/PhDs instead of onto the group leader; the latter might also discredit their own papers that they performed with her.

  • Bernard Soares February 7, 2011 at 11:15 am

    A process of elimination.

    I see that the retraction notice for EMBO J. 24:4260-4270 (2005) states:

    “The investigation concluded that multiple figures contained PCR and western blot duplications and possible other manipulations (Figures 2A, 3A, 4A, 5, 7A and 7C, Supplementary Figures S1A, S2A and S2B, unconfirmed: Figure 1C).”
    This leaves figures 8 and 9 “without duplication other manipulations”. Figure 9 is an imaginary scheme with no data in it. Therefore we are just left with figure 8 which contains data and is unblemished according to the investigation.

    The question is: which of the people who requested a retraction of this paper did the work for figure 8, that is giving them the benefit of the doubt assuming they were not responsible for the blemished figures? Perhaps the 2 people blamed did the work for figure 8 too. So what did the 8 people do to be authors?

    8 people, ZO, LT, UM, PB, CB, DA, RP and SB-P requested the retraction. Surely not all of them were needed to make just one figure? Are they not breaking the overriding rules for scientific authorship as found at http://www.icmje.org ?

  • clare francis February 19, 2011 at 2:17 pm

    If nearly all the figures contain “manipulations” according to the EMBO J retraction notice how can the statement:

    “key experiments presented in the majority of these figures were recently reproduced and that the results confirmed the experimental data and the conclusions drawn from them”

    be true?

    It just doesn’t seem to be logical (even if the editor of the EMBO Jouranl had bothered to look at the purported “reproduced” results, which he hadn’t). Just how can you “reproduce” manipulations without the “reproduced” results being manipulations themselves? If the original figures were “manipulations” you should not be able to “reproduce” them unless you resorted to manipulating the data by definition. I think the editor-in-chief of the EMBO Journal, Bernd Pulverer, needs to go back to the dictionary (a standard one, not his own) and find another word to describe what has happened. He is rewriting the dictionary.

    • BTC February 20, 2011 at 3:32 pm

      You could be able to reproduce fudged results if the ‘fudging’ happened to predict the right result. In other words, whoever manipulated the images just so happened to predict ‘correctly’ what would happen, and this prediction was repeated.

      It’s common; repeating an experiment to prove scientific misconduct doesn’t really prove that the scientific misconduct did NOT take place initially because the initial fudged data could have accurately predicted the outcome of the repeat.

      But yes, I do agree that this retraction statement is poorly worded and a bit partial.

  • Post a comment

    Threaded commenting powered by interconnect/it code.