We have an update on the case of a University of Utah lab that retracted two Cell Metabolism papers last month after a fired lab technician disposed of two lab notebooks without permission. The team has now corrected two other papers, one in Blood and the other in the Journal of Clinical Investigation.
Here’s the notice from Blood:
Correction to De Domenico et al. 110 (10): 3780
De Domenico I, Vaughn MB, Yoon D, Kushner JP, Ward DM, Kaplan J. Zebrafish as a model for defining the functional impact of mammalian ferroportin mutations. Blood. 2007;110(10):3780–3783.
On page 3781 of the 15 November 2007 issue, there is an error in Figure 1. The panel depicting Fpn-GFP N144H in Figure 1B was inadvertently duplicated from another panel in the original published figure. The authors have replaced the panel with the correct image. The findings of the paper have not been affected by the error. The authors apologize to the editors and readers for this mistake. The corrected figure is shown.
Figure 1
Expression of mutant Fpn in zebrafish affects hemoglobinization of erythrocytes. Zebrafish embryos were injected with wild-type or mutant Fpn-GFP. At 48 hpf, the embryos were homogenized (A), and Fpn-GFP levels were assayed by Western blot analysis or (B) stained with o-dianisidine to detect hemoglobinized cells (brown color denoted by the arrow). The figures are representative of 6 different experiments in which 100 embryos were injected with each construct. The survival rate (n = 600) was 69.6% for embryos injected with wild-type constructs, 50.8% for embryos injected with H32R constructs, 64.7% for embryos injected with N144H constructs, and 48.2% for embryos injected with N174I constructs. (C) Embryos were injected with wild-type or H32R constructs with or without coinjection of iron-dextran. The survival rate was 70% for wild-type embryos injected with or without iron-dextran, 50% for wild-type embryos injected with H32R without iron-dextran, and 62% for wild-type embryos injected with iron-dextran.
And here’s the one from the JCI:
Hepcidin mediates transcriptional changes that modulate acute cytokine-induced inflammatory responses in mice
Ivana De Domenico, Tian Y. Zhang, Curry L. Koening, Ryan W. Branch, Nyall London, Eric Lo, Raymond A. Daynes, James P. Kushner, Dean Li, Diane M. Ward and Jerry Kaplan
Published June 1, 2012
Original citation: J. Clin. Invest. 2010;120(7): 2395–2405. doi:10.1172/JCI42011.
Citation for this corrigendum: J. Clin. Invest. 2012;122(6):2326. doi:10.1172/JCI63977.
The anti-Fpn and anti-Jak2 rows in Figure 2A were reprinted from I. De Domenico et al. (ref. 10) without attribution. They represent the same rows as were present in Figure 3B of that manuscript.
The authors regret the error.
Of note: Reference 10 of the JCI paper is a PNAS paper that the team has already corrected, as we reported last month. According to the PNAS correction notice, Figure 3b was
replaced due to errors in preparing the figure for publication.
So it’s unclear which version the figures in the JCI paper came from. It’s also unclear whether the same dismissed technician involved in the Cell Metabolism retractions was involved in these figure errors.
We’ve been trying since Monday to contact senior author Jerry Kaplan for comment on those issues, as well as to find out whether there are any more corrections or retractions planned, and will update with anything we find out.
Hat tip: Commenter Anonymous
I don’t buy it (the “inadvertently duplicated” argument). Take the 2 images that are allegedly duplicated and crank up the contrast – it’s the same fish, but NOT the same image. Someone didn’t accidentally paste in the same image twice, they pasted 2 different images of the same fish. The deception was further up the production chain!
You are right. Hence why I think the initial blame on lab technician for their first two retractions was not entirely the technician’s fault if they didn’t even originally credit the technician.
why to blame lab technician, when credit goes to first author and corresponding author? Without Supervisors’ involvement, these malpractices are impossible. That is where science is in truble. Supervisor propose hypothesis and ask to show this and that. Successful first authors make data for supervisors and left over innocent goos scsientists become whistle blower!
“My lab tech ate my notebooks!” yeah…
Junk Science wrote: “In addition it looks like the Anti-GFP blot (the two cut and pasted bands to the right) of Figure 4A is very similar to the two first lanes of the Anti-Tubulin blot in Figure 1C, however with longer exposure time in Figure 4A.”
You are right. Bands are similar … but one blot was labeled “anti-GFP” whereas the other was labeled “anti-tubulin”.
do you think those papers are not trustworthy? what would be the next step? too many of these cases are distrubing..
You can find the original Blood picture of the zebrafish here:
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2077323/figure/F1/
One fish is fuzzy compared to the first one. It cannot be a copy-paste picture.
I am wondering if it was made fuzzy to hide it?
If you look at the retracted papers there are some that look like image manipulation too. For example the western blot in figure 1C in ‘The Role of Ubiquitination in Hepcidin- Independent and Hepcidin-Dependent Degradation of Ferroportin’. The lower FT blot looks to be an inversion of the upper PM blot, which could only be achieved by image manipulation and not by accidentally scanning your blot the wrong way round (this would create a mirror image)…. Yet again the western blot!
@Anon: You are absolutely correct! We could probably find more if we just keep digging deeper, but hopefully we don’t have to and the person(s) behind these anti-scientific shortcuts in this lab get the proper punishment.
Here is a paper just published in Cell Metabolism. Authors demonstrate that hypothesis from De Domenico are WRONG! This paper is the PNAS (De Domenico 2009) mega-corrected some weeks ago. In the correction De Domenico & Kaplan wrote: “These errors do not affect the conclusions of the article.”
No comment.
Molecular Mechanism of Hepcidin-Mediated Ferroportin Internalization Requires Ferroportin Lysines, Not Tyrosines or JAK-STAT.
Cell Metab. 2012 Jun 6;15(6):905-17
http://www.ncbi.nlm.nih.gov/pubmed?term=22682226
@Anon
Yes absolutely. There is an inversion in figure 1C. It’s like that in all papers!
@Junk Science
“and the person(s) behind these anti-scientific shortcuts in this lab get the proper punishment.”
I’m not sure. Wait and see.
@Ressci Integrity
“do you think those papers are not trustworthy?”
Here you can have an answer:
Molecular Mechanism of Hepcidin-Mediated Ferroportin Internalization Requires Ferroportin Lysines, Not Tyrosines or JAK-STAT.
Cell Metab. 2012 Jun 6;15(6):905-17
http://www.ncbi.nlm.nih.gov/pubmed?term=22682226
This paper demonstrates that some hypothesis from De Domenico/Kaplan’s group are wrong.
For me there is a correlation between a wrong hypothesis and wrong datas. And I am sure that after several years, scientist involve in this manipulation finish to believe that this hypothesis is true. They are so sure that it becomes normal to do a flip/inversion in a western blot instead to do the REAL experiment. It is like a vicious circle.
@iron-man (are you in the iron scientific field and/or do you like Black Sabbath?)
Good job iron-man! The Cell Metab. 2012 Jun 6;15(6):905-17 paper is really taking the bull by the horns. Hopefully, there will be more editors/journals that start to publish papers that questions the prevailing model.
@Junk Science
“are you in the iron scientific field?”
Yep I am interested by this topic.
It seems that this bad Domenico’s story ends. No more correction. No more retraction.
Read also the following link:
http://www.retractionwatch.com/2013/07/01/scientists-whose-papers-were-retracted-and-corrected-leave-university-of-utah/
And this one:
http://www.retractionwatch.com/2013/08/05/university-of-utah-finds-former-faculty-member-guilty-of-misconduct-because-of-reckless-disregard/
http://www.sltrib.com/sltrib/news/56676417-78/lab-kaplan-papers-research.html.csp