Cell biologist Jacob Hanna, the highly cited stem cell researcher currently at the Weizmann Institute of Science, has posted a long erratum for a 2005 paper in Blood for “inadvertent mistakes,” among other issues; soon after, Hanna’s team issued another erratum for a 2009 Cell Stem Cell paper.
There’s more to tell: Last month, commenters on PubPeer noticed that images from at least 10 of the research papers Hanna coauthored in seven journals — that commenters had posted on the image hosting website Imgur and linked to on PubPeer — had been deleted.
Imgur did not confirm whether these specific images had been deleted, but told Retraction Watch:
There are a few possibilities–(1) the original Imgur poster deleted the images and/or (2) the images were flagged by other Imgurians as in violation of Imgur’s TOS [Terms of Service] and/or Community Rules.
Imgur declined to clarify which occurred in this case.
PubPeer confirmed that they did not remove the images themselves, but instead reposted about 32 images from backups stored on PubPeer servers. “We confirm that we did not remove them,” they told Retraction Watch:
We are currently hosting backups of the images on the PubPeer servers. We will continue to host the images until we receive complaints (so far we have not received any). We will take any DMCA [Digital Millennium Copyright Act] takedown requests seriously, but unlike Imgur, we will probably be able to look into the requests on a case by case basis (we note that filing false DMCA requests is illegal and those who do so can be sued and held liable).
PubPeer said that they will continue to allow users to use Imgur as a third party image hosting site. For now, it:
is a convenient solution for us for several reasons but may not be a long-term solution. The backups exist because we (as well as at least one user of the site) have anticipated that images could be removed from third party websites for various reasons. We can allow users to upload images to PubPeer and will do so if/when we are convinced it is necessary. For now the third party solution works well for us.
Hanna’s work has been plagued by scrutiny since late 2014, when posters on PubPeer noticed issues such as signs of duplications in several figures in his papers. In April of this year, the Journal of Clinical Investigation retracted his 2004 paper on the activity of natural killer (nk) cells in immune response.
Neither Hanna nor the two corresponding authors on both of the latest corrected papers responded to requests for comment.
The first article, “Functional aberrant expression of CCR2 receptor on chronically activated NK cells in patients with TAP-2 deficiency,” published in Blood, has been cited 18 times, according to Thomson Scientific’s Web of Knowledge. From the April erratum notice (which we believe could be considered a “mega-correction”):
In the 15 November 2005 issue, there are errors in Figures 2, 4, 5, and 7. In Figure 2A on page 3469, the authors would like to explicitly clarify that the representative fluorescence-activated cell sorting (FACS) staining for unactivated NK cells were presented in a previous article from the authors’ laboratory,1 as they were representative of 3 to 23 independent replicates performed as indicated in the figure legend. Importantly, the CANK data presented were obtained and analyzed simultaneously with unactivated NK cells shown in Figure 2A. FACS plots for CXCR4 and CCR2 on CANK cells were inadvertently duplicated in Figure 2A. The corrected Figure 2A is shown, featuring the validated CXCR4 staining on CANK cells.
In Figure 4 on page 3470 and Figure 7C on page 3471, inadvertent duplication in the assembly of FACS panels occurred. In Figure 4, duplication occurred in unactivated T cells from patient A and C samples, unactivated NK cells from the control, and mother samples. In Figure 7C, duplication occurred between controls and patients’ samples. Because the authors could not track down the original FACS data files for these stainings, the experiments were repeated with all ethical and experimental approvals in place. The corrected Figures 4 and 7C are shown (patient A staining was not replaced in Figure 7C).
In Figure 5 on page 3470, the wrong gel images belonging to a different experiment were inadvertently used and misassembled. To resolve this, the authors repeated this experiment with all ethical and experimental approvals in place. The corrected Figure 5 is shown.
These inadvertent mistakes do not change the conclusions made in the article. The authors apologize for these mistakes.
A spokesperson for Blood said:
we do not comment on retracted articles and hope that the explanation provided on the Blood website is self-explanatory.
The second corrected article, “Metastable Pluripotent States in NOD-Mouse-Derived ESCs,” has been cited 183 times, according to Thomson Scientific’s Web of Knowledge. From the erratum in Cell Stem Cell, posted in May:
Readers have brought to our attention that some of the figures presented in this paper contain duplicated images that are also included in other figures in the paper. We have examined the original data associated with the study and have determined that we made errors during the assembly of two of the supplemental figures. The affected panels are as follows.
(1) In the original paper, the Sox2 gel shown in Figure S9B is the same as the Ubi-Klf4 gel in Figure 2D. This duplication occurred because we used the wrong gel for Sox2 expression in Figure S9B, and instead used a gel from Figure 2D (Ubi-Klf4). The rest of the panels in Figure S9 are accurate and match our original gels. The corrected Figure S9 shown as follows replaces the Sox2 gel in panel (B) to represent the correct original gel run at that time. All of the other panels remain the same as presented previously.
(2) In the original paper, the fluorescence images shown in Figure S10A are the same as images in Figures S3A and 4E. This image duplication occurred because we used fluorescence images to assemble Figure S10A that in fact represent the different experiments shown in Figures S3A and 4E. The images shown originally in Figure S10A are incorrect. In the corrected version of Figure S10A shown as follows we have replaced the fluorescence images in panel (A) with the correct original data. The rest of the data remain the same as presented originally. Having examined the original data associated with this paper carefully, we are confident that these errors in figure assembly do not affect the conclusions drawn or the validity of the underlying research. We are grateful to the readers who brought these unfortunate mistakes to our attention, and we apologize to the community for any confusion they have caused.
A spokesperson for Cell Stem Cell told Retraction Watch:
Cell Stem Cell is aware of the issues raised about Dr. Hanna’s work and the 2009 paper is currently under further investigation.
He declined to elaborate further on the nature of the investigation.
Like Retraction Watch? Consider supporting our growth. You can also follow us on Twitter, like us on Facebook, add us to your RSS reader, and sign up on our homepage for an email every time there’s a new post. Click here to review our Comments Policy.