Retraction Watch

Tracking retractions as a window into the scientific process

Second retraction stemming from Cardiff investigations appears

with 6 comments

cancer research 1113A second retraction of a paper by a Cardiff University researcher found to have committed misconduct has appeared.

In April, a Cardiff investigation found that Rossen Donev, a former researcher at the university, had manipulated images in four different papers. Donev, who was at the University of Swansea until August, according to his LinkedIn profile, and is now director of Biomed Consult Ltd., had already retracted a Journal of Immunology paper in late 2011 after a different Cardiff investigation.

The investigation cleared co-author and dean Paul Morgan of misconduct. Morgan resigned from Cardiff in August, but “he categorically denied that his decision had anything to do with the misconduct investigation,” according to Times Higher Education.

Here’s the new notice, from Cancer Research:

The article titled “p53 Regulates Cellular Resistance to Complement Lysis through Enhanced Expression of CD59,” which was published in the February 15, 2006, issue of Cancer Research (1), is being retracted at the request of the authors following the release of the conclusions of an internal investigation panel established by Cardiff University to examine allegations of research misconduct in the preparation of the manuscript. The panel found evidence of splicing or pasting affecting Figs. 1B and 4A, without indication that this had been done. The panel feels that although these image manipulations cannot be characterized as “fabrication” because there is no reason to doubt the validity of the underlying science in the article, they represent unacceptable practice when submitting a manuscript for publication. Dr. R.M. Donev accepted full responsibility for these actions. The panel concluded that none of the other coauthors of this manuscript knew, or had reason to suspect, that the data presented in the manuscript had been manipulated by Dr. R.M. Donev.

Four of the five authors agreed to this Retraction. Attempts to contact the first author, Rossen M. Donev, were unsuccessful.

Duncan S. Cole, Baalasubramanian Sivasankar, Timothy R. Hughes, B. Paul Morgan

Department of Medical Biochemistry and Immunology, School of Medicine, Cardiff University, Cardiff, United Kingdom.

The study has been cited 22 times, according to Thomson Scientific’s Web of Knowledge.

Please see an update on this post.

Hat tip: Rolf Degen

Written by Ivan Oransky

November 11th, 2013 at 9:30 am

Comments
  • aceil November 11, 2013 at 10:25 am

    An examination of studies which cite a flawed publication should follow each retraction. Those who claim to have reproduced similar results in verification of the validity of questionable findings should face scrutiny.

  • david hardman November 11, 2013 at 11:19 am

    J Biol Chem. 2003 Feb 14;278(7):5214-26. Epub 2002 Nov 14.
    Recruitment of heterogeneous nuclear ribonucleoprotein A1 in vivo to the LMP/TAP region of the major histocompatibility complex.
    Donev R, Horton R, Beck S, Doneva T, Vatcheva R, Bowen WR, Sheer D.

    Source
    Human Cytogenetics Laboratory, Cancer Research, UK London Research Institute, Lincoln’s Inn Fields Laboratories, 44 Lincoln’s Inn Fields, London WC2A 3PX, United Kingdom.

    PMID:12435746

    Figure 5.
    http://www.jbc.org/content/278/7/5214/F6.large.jpg

    Please compare the MAR2 AHB band in figure 5C with the MAR2 AHB band in figure 5A (the adjacent bands are different).

    Please compare the MAR1 AHB band and MAR4 AHB band.

    Figure 6.
    http://www.jbc.org/content/278/7/5214/F7.large.jpg

    Figure 6A. As well as the vertical splicing between the sections (some may say that this is legitimate), there are also horizontal changes in background. For example in the left beta-actin section the bands seem to be sitting above a horizontal change in background and with darker than background signal around them. hnRNP-A1 section. The bands seem to have darker than background areas around them.
    rechnRNP-A1 section. The bands seem to have darker than background areas around them.

    • david hardman November 11, 2013 at 11:34 am

      Mol Psychiatry. 2007 Jul;12(7):681-90. Epub 2007 Mar 13.
      A role for SC35 and hnRNPA1 in the determination of amyloid precursor protein isoforms.
      Donev R, Newall A, Thome J, Sheer D.

      Source
      Human Cytogenetics Laboratory, Cancer Research UK London Research Institute, Lincoln’s Inn Fields Laboratories, London, UK.

      PMID: 17353911

      Figure 3.
      http://www.ncbi.nlm.nih.gov/core/lw/2.0/html/tileshop_pmc/tileshop_pmc_inline.html?title=Click%20on%20image%20to%20zoom&p=PMC3&id=2684093_ukmss-4677-f0003.jpg

      Please compare the background in panels 3c, 3d and 3e. For example, in the top to the left of the main white band there is an area that looks a bit like Wales. In the SW of this area you can see something resembling Pembrokeshire, and in the NW something like the Llyn peninsula sticking out. Also notice to the right of the 200 marker lane (right lane) a diagonal streak. This diagonal streak appears in the panels in figures 3b, 3c, 3d and 3e.

      This Wales-like feature is in figures 3c, 3d, and 3e, yet the bands to the left of the marker bands in the lower part of the panel are different.

      Please compare the marker bands in the right lane in figures 3b, 3c, 3d and 3e.

      Please compare the marker bands in the panel in figure 3a with a horizontal mirror image of the marker bands in figures 3b, 3c, 3d and 3e.

  • ferniglab November 11, 2013 at 12:34 pm

    Cardiff are not pointing the finger, they probably do not wish to leave themselves open to litigation – there is no First Amendment in the UK. At least they seem to be “doing the right thing” here, which is to apply the same standard to their research as they apply to their undergraduate students.

    • david hardman November 11, 2013 at 4:12 pm

      Mol Immunol. 2008 Jan;45(2):534-42. Epub 2007 Jun 26.
      The mouse complement regulator CD59b is significantly expressed only in testis and plays roles in sperm acrosome activation and motility.
      Donev RM, Sivasankar B, Mizuno M, Morgan BP.

      Source
      Department of Medical Biochemistry and Immunology, School of Medicine, Cardiff University, Cardiff CF14 4XN, UK.

      PMID: 17597212

      This paper may have been discussed before.

      One thing which caught my eye.

      Please compare band lane 1 of CD59b panel (testis) figure 4 Mol Immunol. 2008 Jan;45(2):534-42.
      http://www.ncbi.nlm.nih.gov/core/lw/2.0/html/tileshop_pmc/tileshop_pmc_inline.html?title=Click%20on%20image%20to%20zoom&p=PMC3&id=1995235_gr4.jpg

      with band lane ii figure 1B (EL4 cells) J Immunol. 2004 Sep 15;173(6):3684-92.
      http://www.jimmunol.org/content/173/6/3684/F1.large.jpg

      Please compare the backgrounds in the panels.

      Please note how the dot in lane iii (level with the bottom of the band in lane ii)
      figure 1B J Immunol. 2004 Sep 15;173(6):3684-92
      becomes three dots in lanes 3 and 4 of CD59b panel figure 4 Mol Immunol. 2008 Jan;45(2):534-42.

      For reference:

      J Immunol. 2004 Sep 15;173(6):3684-92.
      CD59a is the primary regulator of membrane attack complex assembly in the mouse.
      Baalasubramanian S, Harris CL, Donev RM, Mizuno M, Omidvar N, Song WC, Morgan BP.
      Source

      Complement Biology Group, Department of Medical Biochemistry and Immunology, University of Wales College of Medicine, Cardiff, United Kingdom.

      PMID:15356114

  • david hardman January 5, 2014 at 10:25 am

    4th retraction.

    There is a retraction notice for Mol Immunol. 2008 Jan;45(2):534-42 in

    Molecular Immunology Volume 58, Issue 1, March 2014, Pages 150.

    http://www.sciencedirect.com/science/article/pii/S0161589013005750

  • Post a comment

    Threaded commenting powered by interconnect/it code.