The authors of two Journal of Biological Chemistry (JBC) studies of the molecular underpinnings of hearts whose rhythms have gone awry have retracted the papers, for reasons that are not completely clear.
The two papers are “MicroRNA miR-133 represses HERG K+ channel expression contributing to QT prolongation in diabetic hearts,” published in 2007, and “Down-regulation of miR-1/miR-133 contributes to re-expression of pacemaker channel genes HCN2 and HCN4 in hypertrophic heart,” published in 2008.
This being the JBC, the retraction notices in the August 12, 2011 issue say nothing:
This article has been withdrawn by the authors.
The papers have had some impact: The 2007 article has been cited 126 times, according to Thomson Scientific’s Web of Knowledge, while the 2008 study has been cited 76.
So why were the papers retracted? One of the corresponding authors, the Montreal Heart Institute’s Zhiguo Wang, tells Retraction Watch:
We noticed some mistakes in the Western blot band images shown in these papers. These mistakes do not invalidate our results and conclusions, and we and others have been able to reproduce the data reported in these papers.
That doesn’t sound like the entire story to us. After all, if there were mistakes that “do not invalidate [the] results and conclusions,” why not just issue a correction? We asked some follow-up questions, but haven’t heard back.
Indeed, the response to our questions from Wang’s department chair, Jean-Claude Tardif, suggests there’s a bit more going on:
We are aware and are still investigating.
Wang also has an appointment at the University of Montreal, and is listed as
senior research scholar of the Fonds de Recherche en Sante de Quebec, a ChangJiang scholar professor, and a LongJiang scholar professor of China.
His co-authors include a second corresponding author from the State-Province Key Laboratories of Biomedicine-Pharmaceutics of China and Harbin Medical University, also in China. We have tried reaching him for comment, and will update with anything we hear back.
In the meantime, the PDFs of both papers are both available, with big “withdrawn” stamps that cover some of the Western blots, but perhaps an astute Retraction Watch reader can have a look and figure out what’s wrong with them.
Please see an update on this story.
Hat tip: Uwe Vinkemeier
Well that was easy….
In the v282#17p12363 paper, look at the GAPDH bands in Figure 3A. Now go back to Figure 2C and look at the GAPDH bands. Lanes 4-5-6 are the same. They duplicated the blot.
In the v283#29p20045 paper, Fig 3A right panel GAPDH blot contains the same 2 bands as Fig. 5D.
So, the “mistake” appears to involve using the same GAPDH blot for different experiments. This is so common, it’s not even funny any more!
Yep, totally agree. I was immediately drawn to the GAPDH blots in both papers. I have a feeling there is more duplication in the other blots just using the famous “eyeball test”, but have not had time to really look at them yet. I also agree that it is not even funny any more and JBC is the worst for this kind of stuff. Why are they so arrogant with the retraction notices? They should come out with it. The journal is in a steep decline anyway and i really don’t trust the majority of articles in it anymore. Shame.
I’m no Photoshop expert, but in Vol 283, the GAPDH blot in Fig 2D (lower right of panel) looks completely fabricated. When you zoom in, the pixels just look completely different to the other blots and the shape and density of the bands looks suspicious to me. I could be wrong.
They loved that figure so much, they published it more than twice ! The HCN4 and GAPDH blots in Figure 3a of Luo et al., J Biol Chem 283 (29):20045, 2009 are also the same as those in Figure 4a of Lin et al., Cell Physiol Biochem 23:317-326, 2009. Things like this usually hint at bigger problems. Other dubious data from the Wang lab include:
The Nkx3.1 EMSA in Figure 6c of Lin et al., J Cell Physiol 212:137, 2007 is the same as the Stat3 EMSA in Figure 5a of Gao et al., Mol Pharmacol 70:1621, 2006
The E2F1 ChIP blot in Figure 1d of Lin et al., PloSOne 6(5):e20362, 2011 is the same as the p53 CHiP blot in Figure 3a of Xiao et al., EMBO J 30:524, 2011
Lanes 3-6 of the GAPDH control for the HSP70 blot in Figure 3b of Xu et al., J Cell Sci 120(17): 3045-3052, 2007 are the same as the GAPDH control for the Nkx3.1 blot in Figure 6g of Lin et al., J Cell Physiol, 212:137, 2007
“Oh what a tangled web we weave…”
take a look at the gel images in Figure 3 of the Nature Medicine paper (vol 13, page 489, published in 2007. they look alike!
Oh dear!!! Here we go again.
The peer-review process file for the EMBO J paper mentioned above is public, and is available here:
http://www.nature.com/emboj/journal/v30/n3/extref/emboj2010347s2.pdf
The control CHiP assay in Figure 3a (identical to the figure later published in the PLoSOne article) was a late addition to the EMBO J paper, after it was specifically requested by the reviewers and editor. Questions were raised about the labeling of this figure and about inappropriate adjustment of contrast to downplay the intensity of one of the bands.
Less than 2 months separate the formal acceptance of the EMBO J article and submission of the PLoSOne paper, and all but one of the authors on the former are also on the latter… I can’t believe someone didn’t catch this.
No retraction notice yet for either of these papers, but I’d bet it’s only a matter of time.
Is it usual to have peer-review process files online, or has EMBO J put this one up because the paper was retracted? As an incredibly junior researcher (barely started second year of PhD) it’s useful to see the types of comments that appear – ideally on papers that WEREN’T later retracted!
All EMBO J papers are now published with the review process files, unless the authors choose not to make these public, which is rare. I think it will become more widespread over the next few years.
(this is probably a side issue, but might become important)
Having just looked at the review process files, I have no idea how that paper was published in light of the type of comments raised. There is so much doubt in the reviewer comments and even the editors start to question the authors AFTER acceptance. It is just not right. Reviewer 1 raised a lot of very good points (particularly with the WB and ChIP experiments) but then at the end was essentially forced to concede and stated: “I won’t object any longer”. I have been in this position myself – when you raise technical concerns, you are pushed aside until you eventually have to give up. In this case, it just looks as if the editors were going to publish the paper no matter what. Looks like that attitude may come back to bite them in the arse.
Well done EMBO J for the transparency here, truly, but next time when a reviewer tells you almost overtly (in public) that they don’t trust the article, pay more attention.
did you guys notice that the authors had another major publication in Nature Medicine? The stories are quite similar. If the data from JBC are fabricated, how could those in nature medicine be right?
I would suggest an email to the editor of Nat Med to investigate their paper.
it’s this one:
http://www.nature.com/nm/journal/v13/n4/abs/nm1569.html
The muscle-specific microRNA miR-1 regulates cardiac arrhythmogenic potential by targeting GJA1 and KCNJ2
take a look at fig 3!!
I don’t see what’s especially noticeable about figure 3 — is it similar to a previously published figure? Or do you think that the quantitations are too similar to each other?
Yeh I don’t see the issue with Fig 3. I think ang is trying to say that the two blots are the same, maybe one is over-exposed and labeled as a different blot. If you look closely though, the band shapes are too different.
Enlarge the left panel of the western blots in figure 3: it looks a bit strange. this blot might have been through manipulation.
Well spotted, ang! There are some lines which seem to straight. Could also be some unusual compression artefact. Did anybody notify Nature Medicine? I’d really like to see the unprocessed blot.
This story just made the national news in Canada: http://www.cbc.ca/news/technology/story/2011/08/10/heart-medical-studies-retraction-montreal.html
To quote:
“Wang said he and his co-authors withdrew the papers because of a mixup of images used to illustrate the data. The rest of the information in the articles was accurate, he said.
The conclusions in the paper were correct and the lab results have been reproduced, Wang said.”
That’s quite the mixup.
Right… what are those people thinking? So many stories out there are not reproducible. For me, it takes only one fishy figure to disregard the rest of the manuscript, or, if this happens more than once, the rest of the lab’s results. If a group has to retract a paper because of data duplication, who will ever again believe what’s in their publications? Not only the damage done to science at large and other groups in the field can be big, but the damage to the PI’s and the first author’s reputations is always huge.
Personally, I find that I fundamentally disbelieve these excuses of getting the images mixed up because of the usual difficulty getting them in the first place.
When you get a publishable image, sometimes at the end of a long series of experiments and experimental failures, it’s such a big deal that it’s inconceivable one could end up putting it in the wrong paper. Are we supposed to believe these labs are so awesome that they generate entire atlases of publishable images and it’s no wonder that sometimes they pick up the wrong one? Yeah, right.
“…only one fishy figure to disregard the rest of the manuscript, or, if this happens more than once, the rest of the lab’s results.”
Wow. That’s harsh, tk. Lets back up and step away from the case at hand for a moment.
So, in your mind, Ron Breslow’s ~468 papers not published with Monica Mehta are bunk? Bertram Batlogg’s work on superconductivity before Schon also garbage?
Some of the commenters on this blog seem to think that anyone who’s retracted a paper needs to wear a scarlet “R” for “retraction” in order to appear in public. You should hope this never happens to you, tk. The world isn’t black and white. Sometimes the in-house independent reproduction experiments are sabotaged. Sometimes scientists get duped. It can work either way in student/PI relationships. Should people make stuff up and try to dupe others? – No, but it happens. Should the dupes have better scrutinized the data? -Yes, but all relationships (even those between scientists – including students/PI’s and authors/editors) involve some level of trust.
Before you discard the rest of some lab’s results, why not look to see if any of the results have been independently reproduced? Or at least learn more about the case at hand – is a “bad apple” to blame or does it look like the PI encouraging/perpetrating bad science?
One-two, I also think my statement sounds harsh. But I’ve come to this through negative experience. Trying to reproduce results and not succeeding (with positive controls being positive) just makes me tired. Personally, I follow the rule I stated above. Other lab members sometimes do not, and from this I know that my predictions are true in most of the cases. I guess it would be very hard for R-marked people if everybody acts like this, but most scientists do not. For me, this rule just saves me a lot of time and effort, both of which I have limited supply.
Also, to me it is not clear which course of action yields better (more economic) results for science and society. You could be right with the point that giving PIs who published fraudulent work a second chance is superior, but my comment is my personal opinion. I’m still pipetting and I cannot afford to waste time on unreproducible data. And you are very right, I very much hope that I never have to wear the R.
One-Two: I agree with you that we probably need to take a step back from this, but those of us who are regulars to this blog know that there is usually not this type of smoke without a raging fire somewhere. However, I’m tired of this whole “if somebody else has reproduced my results then my data is valid” idea. Of course, my personal favorite on this site is the good old “we retract the paper, but the conclusions are still valid because so-and-so has replicated the results” comment. What a load of rubbish!
I could literally sit here on my computer this weekend and write an awesomely fabricated paper in my area of interest (with figures) that I am 99% sure would be replicated. There are two issues with the replication argument: 1) If you do not maintain the integrity of scientific research in your publications, whether that be by duplication or outright fraud, your results have NO relevance in my opinion and should be deleted from the public record. If you think it is OK to re-use figures in multiple papers, what else are you OK with? and, 2) Just because it’s replicated, doesn’t mean it’s right. A replicated false positive (or negative) is still a false positive. Throw into the mix publication bias, maybe caused by retracted paper itself, and the replication argument is weakened in my opinion.
Also, your argument about being duped is ridiculous. I’m assuming you’re a PI from your post. If you get duped as a corresponding author, it’s your fault. Period. End of story. You take the fall. I can’t even be bothered to type out why this is the case as you should know this.
So, this makes it sound like Wang initiated the withdrawals. This of course begs the question whether he will also be initiating withdrawals at the other journals as discovered above by Figure Sleuth (nice work BTW) ?
From looking at a few of the older papers, there are some odd similiarities in some DNA digest blots from his early work, but the images are low-resolution presumably because the PDFs were created from scanned-in paper copies, so I couldn’t really tell if they were duplicated. At least the modern use of hi res images in PDFs makes it easier to spot this type of thing!
I think he has initiated the retractions (if that is indeed the case) in an attempt to gain control of what will soon become a very bad situation for him. As Figure Sleuth pointed out, there are duplications in numerous figures across journals and I am convinced some of the figures are completely fabricated, but that is more difficult to prove. The domino effect is inevitable at this point. Prior to this blog, he may have got away with one or two retractions, but now I fear it will be more widespread. If you read the EMBO J review file, these guys are definitely up to some dodgy stuff.
What a shame! Following the instruction of ang, I found there do have some artificial traces with the boundary between lane 1 and lane two in the left panel of fig3. I can not believe there are so many scientists involved in this kind of disgrace. The PIs could not be innocent during the frauds. They should have taken the responsibility for the quality and accuracy of all of the presented data as well as the figures before before they submitted the papers and enjoy the benefits brought by the publication of their paper!
This guy went back to China as the founder director of cardiac institute in Harbin Medical University ! He was also selected as 1000 talent scholar. What a Shame!
Dajia100– do you have a source or link for this information about him?
http://www.icm-mhi.org/files//pdf/Communique-EN-2011/Press_release_MHI_Research_Centre.pdf