Academic throne succession: from Anne Dejean to Oliver Bischof

Clare Francis recently investigated another set of French scientists, and posted it on PubPeer. One of them is the cancer researcher Anne Dejean- Assemat, Research Director Classe Exceptionnelle at INSERM and professor at the Institut Pasteur in Paris. The other is her German mentee Oliver Bischof, now group leader under Dejean’s leadership. It is very likely Bischof is being groomed to succeed Dejean when she retires. And to be fair, the way this Pasteur lab is run, he is definitely the right man for the job.

This is Dejean’s CV, as presented by her “Nuclear Organization and Oncogenesis” lab at Institut Pasteur:

“Anne Dejean is Research Director at INSERM, Professor at the Institut Pasteur and Head of the Laboratory of Nuclear Organization and Oncogenesis/INSERM U993. She graduated from Pierre et Marie Curie University in Paris and earned her PhD in Pierre Tiollais’ lab at the Institut Pasteur in 1983. Member of EMBO and of the French Academy of Sciences, she has received the Gagna and Van Heck Prize in 2003, the L’Oréal-UNESCO for Women in Science Awards in 2010, the Grand Prix INSERM in 2014 and the Sjöberg Prize in 2018. She was awarded two ERC Advanced Grants, in 2011 and 2018.”

Congratulations to Anne Dejean, Head of the Nuclear Organization and Oncogenesis Unit at the @institutpasteur, awarded an #ERC Advanced Grant. #ERCAdG @ERC_Research @Inserm https://t.co/z47fqHDznW

— Institut Pasteur (@institutpasteur) March 29, 2019

I would like to show you how exactly Professor Dejean achieved so much. And yes, I know European Research Council (ERC) leadership doesn’t believe in research integrity, but still. The scientific community has a right to know who and how receives such exclusive public research funding, which honest scientists see routinely denied. Dejean’s new €2.5mn Advanced Grant “Deconstructing the role of SUMO on chromatin in cell identity and tissue repair” was awarded in 2018, and don’t worry, it will continue for the next 5 years excellently. On this, you can trust ERC. Carlos Lopez-Otin enjoys his ongoing grant despite retractions, mouse murder, admitted data fakery and cessation of research activities, and even proven the fraudster Maria Fousteri was free to use up all of €2mn to her pleasure, after the university notified the ERC.

Now, someone, or a group of someones in Dejean’s lab has been naughty. We cannot know who, but the evidence of their naughtiness is there. Let’s look at this almost 20 year old paper:

F. Lehembre, P. Badenhorst, S. Müller, A. Travers, F. Schweisguth, A. Dejean Covalent modification of the transcriptional repressor tramtrack by the ubiquitin-related protein Smt3 in Drosophila flies Molecular and Cellular Biology (2000) doi: 10.1128/mcb.20.3.1072-1082.2000

Normally, gel lanes do not triplicate like this, unless the authors disagreed with their actual experimental results and decided prove them wrong, using The Photoshop, or in this case, probably the good old scissors and glue. Maybe the issue is serious enough for the journal to overrule the 6-year deadline strategy? After all, the journal MCB already had to correct a Dejean-coauthored paper from INSERM in Paris, for duplicated data (Erker et al MCB 2013).

Or how does this happen, in the same journal, but 5 years later? Something horrible happened to Figure 3C in 2005, this time it was most likely the revolutionary biomedical research tool of Photoshop being applied:

O Bischof, K Nacerddine, A Dejean Human papillomavirus oncoprotein E7 targets the promyelocytic leukemia protein and circumvents cellular senescence via the Rb and p53 tumor suppressor pathways Molecular and Cellular Biology (2005) doi: 10.1128/mcb.25.3.1013-1024.2005

How come 3 out of 4 bands of the top PML western blot look so similar, especially after a rotation? Also the two central E7 bands seem to be mirror images of each other, and it turned out, also the first two p53 bands are strangely similar, and the same applies for the CBP bands below.

How exactly did Dejean and Bischof plan to cure leukaemia with this technology? One year later, they published another paper with similar issues, where some of Bischof et al MCB 2005 data was apparently reused in excitingly new context. Luckily the new journal was Molecular Cell, this Elsevier outlet seems to have an unhealthy obsession of attracting and cultivating manipulated data.

O Bischof, K Schwamborn, N Martin, A Werner, C Sustmann, R Grosschedl, A Dejean The E3 SUMO ligase PIASy is a regulator of cellular senescence and apoptosis Molecular Cell (2006) doi: 10.1016/j.molcel.2006.05.016

Now, what do we have here? A set of 3 bands from Figure 2C of MCB 2005 which used to be p53, but then somehow became Rb, a very different protein. One of this trio of jumping bands grabbed another friend and cloned themselves again to become yet another protein, Cyclin D1 in Figure 4E of Molecular Cell. It is not helpful that the sample legend is different for every single instance. And since we are discussing the magical permutation of p53 into Rb protein, let’s look at Figure 4A:

The band even got hyper-phosphorylated while transforming itself from p53 to pppRb. There are more gems in that 2006 paper, have a look at Figure 6H:

Can it be that top western blot panel shows the same band all over, copy-pasted with an occasional flip? Was someone taking the proverbial PIAS here? Also the Figure 6F has what Institut Pasteur experts will probably call “an inadvertent duplication”.

Now, Clare Francis did report all these and other issues to Molecular Cell editors, but it is rather like writing to Philip Morris to complain about someone smoking in a restaurant. This Cell Press journal chose to do nothing at all about fraudulent papers by former CNRS president Anne Peyroche, and rejected retraction requests in the case of Maria Fousteri.

"We work at @institutpasteur on SUMO and chromatin, a work which opens avenues to target a druggable pathway in clinical settings such as regenerative medicine and #cancer treatment", Anne Dejean. #Pasteur130Ans pic.twitter.com/8o3LMhN8YB

— Institut Pasteur (@institutpasteur) May 17, 2019

Before the first author of this Photoshoppian masterpiece joined the Dejean lab in Paris, the German Oliver Bischof did a brief postdoc stint at the lab of Farzin Farzaneh at King’s College London. There was much to learn also: Farzaneh was recently investigated for data manipulation, but no research misconduct was found, only “poor research practices” which required Errata in 5 journals.

Between Farzaneh and Dejean postdoc, Bischof did a stint at the Berkeley lab of Judith Campisi, a star researcher in the field of cellular senescence. This collaboration moved the research area ahead with this important contribution:

O Bischof, S Galande, F Farzaneh, T Kohwi-Shigematsu, J Campisi Selective cleavage of BLM, the bloom syndrome protein, during apoptotic cell death The Journal of biological chemistry (2001) doi: 10.1074/jbc.m006462200

Now this cannot have happened by incident, no matter how drunk The Bischof might have been, because the upper and lower set of bands are supposed to have been on one solid gel. No way the upper 20/46 bands can accidentally become lower 20 set, flipped.

The shenanigans apparently continued in Paris. This last-author paper from Institut Pasteur Bischof proudly shows among his selected publications:

M Ogrunc, RI Martinez-Zamudio, P Ben Sadoun, G Dore, H Schwerer, P Pasero, JM Lemaitre, Anne Dejean, O Bischof USP1 Regulates Cellular Senescence by Controlling Genomic Integrity Cell Reports (2016) doi: 10.1016/j.celrep.2016.04.033

Now I should immediately declare that the first author Müge Ogrunc and I worked for several years together at the same DNA damage research small lab at IFOM in Milan, Italy. I tried to contact Müge about this copy-pasted and flipped CHK-1 western blot, but received so far no feedback. But then again, my former colleague Müge joined the Dejean-Bischof lab at Pasteur only for two years, 2015-2017, while the data integrity issues go way back there.

Dejean is now 62 and will probably retire in a couple of years, maybe when her new ERC grant ends. With Bischof, she prepared a worthy successor to take over her artisan craft business, which now started to conquer new sources of grant cash: ageing or senescence research. The general questionable research ethics in Dejean lab shine through in examples like this:

N Martin, K Schwamborn, V Schreiber, A Werner, C Guillier, XD Zhang, O Bischof, J Seeler, A Dejean PARP-1 transcriptional activity is regulated by sumoylation upon heat shock The EMBO Journal (2009) doi: 10.1038/emboj.2009.279

Now the paper is 10 years old, and the French and German authors might declare that they missed previous guidelines in biomedical publishing where gel splicing should be avoided, and where not possible, clearly indicated, because the guidelines were in English. But thing is, in Figure 4B they did provide a clear black vertical line to indicate splicing in one case, but they chose not to do this in other cases. Why? Because you cannot compare the signal of a + vs – treatment on a spliced gel. In brief, the authors bullshitted the EMBO editor, the peer reviewers and the scientific community while knowing perfectly well that what they did was not right.

[📝Communiqué] Félicitations à Anne Dejean-Assémat et Hugues de Thé, lauréats du prix #Sjöberg 2018 décerné par l’Académie royale suédoise pour leurs recherches sur la leucémie aiguë promyélocytaire, une des formes les plus agressives du cancer du sang pic.twitter.com/UNBtoqb5Wg

— Frédérique Vidal (@VidalFrederique) February 13, 2018

Or how about another paper from same time, with same key authors:

N Martin, K Schwamborn, H Urlaub, B Gan, JL Guan, A Dejean Spatial interplay between PIASy and FIP200 in the regulation of signal transduction and transcriptional activity Molecular and Cellular Biology (2008) doi: 10.1128/mcb.01210-07

Now it might appear like some sloppy but otherwise innocent gel splicing, but it is not. The horizontal splice edges are the giveaway that something more sinister than mere excision of irrelevant gel lanes took place. The PIAS panel in Figure 2A is apparently a collage of various gel bands slapped on in Photoshop, the same thing happened to the PIAS panel of Figure 3C. There however, the bands were not just stuck on randomly: they were precisely arranged in Photoshop on a ladder, to indicate the transgenic vs endogenous PIAS protein. Basically, another PIAS-taking exercise by Anne Dejean and her lab, and yet another reason for the publisher American Society for Microbiology (ASM) to investigate these older works in their journal MCB.

The following is a collaborative paper from the Institut Pasteur, maybe Dejean can check if the problematic figures were made in her lab.

CA Renard, C Labalette, C Armengol, D Cougot, Y Wei, S Cairo, P Pineau, C Neuveut, A De Reyniès, A Dejean, C Perret, MA Buendia Tbx3 is a downstream target of the Wnt/beta-catenin pathway and a critical mediator of beta-catenin survival functions in liver cancer Cancer Research (2007) doi: 10.1158/0008-5472.can-06-2344

Not just some splicing, a b-cat band was most obviously cloned in Figure 1B. Even Nature, where Dejean published quite a lot, was not safe. That journal however replied to Clare Francis and announced to look into the case. On its own, there is just a duplicated SUMO gel:

D Ribet, M Hamon, E Gouin, MA Nahori, F Impens, H Neyret-Kahn, K Gevaert, J Vandekerckhove, A Dejean, P Cossart Listeria monocytogenes impairs SUMOylation for efficient infection Nature (2010) doi: 10.1038/nature08963

But considering what else went on and still goes on in the Dejean and now Dejean-Bischof lab? Well, the worst things ended up strategically in Molecular Cell. We are dealing with real professionals here.

INSERM is already well aware of the PubPeer evidence. Clare Francis also informed Institut Pasteur, and its president Stewart Cole replied to him:

“As is our practice in such matters I have instructed the Committee for Scientific Integrity to conduct an enquiry into the allegations. This will take some time. The PIs concerned have been informed.”

Now I would not get overexcited here. Institut Pasteur previously investigated another affair, relating to their collaboration with the present rector of Karolinska Institutet, and found no data manipulation despite authors admitting a possible gel lane triplication, while the alleged original data did not match.

My prediction: conclusions not affected, original data unfortunately eaten by some goats.

Update 25.05.2021

Bischof has 4 retractions now, 3 with Dejean (Bischof et al MCB 2005, Bischof et al Mol Cell 2006, Ogrunc et al Cell Reports 2016) and one from his postdoc period with Judith Campisi (Bischof et al JBC 2001). Today, I received this email from Christophe d’Enfert, scientific director of Institut Pasteur:

“An enquiry into the alleged case of scientific misconduct by Dr. Oliver Bischof, a CNRS scientist, and Prof. Anne Dejean was launched in the fall of 2019 and recently concluded. Prof. Anne Dejean’s unit at Institut Pasteur is affiliated with Inserm. This enquiry was thus conducted jointly by the scientific integrity officers of the CNRS, Inserm and the Institut Pasteur. Several of the publications linked to the allegation were retracted prior to completion of the enquiry (PMID: 33626350, PMID: 33338405, PMID: 33277404, PMID: 33106371).

The team in charge of the enquiry presented a written report to the heads of the three organizations concerned. The evidence and conclusions indicate that during the preparation of many of the publications O. Bischof, but not A. Dejean, had engaged in fraudulent practices, in violation of the rules of honesty and scientific integrity.  After due consultation with the President of the CNRS, Prof Cole, President of the Institut Pasteur, asked O. Bischof to leave the Institut Pasteur without delay. This is why the group of Oliver Bischof is not listed anymore.“

---