Author says ‘misguided efforts for the ideal western blot led to the withdrawal of these studies’

The Journal of Biological Chemistry has retracted two papers by a group from the University of Toronto over what the leader of the research says were “misguided efforts” by a co-author to make the perfect Western blot. 

The retractions are among a batch of seven recent removals by the journal for image issues, some of which were flagged on PubPeer. 

Two of the papers, which date back to 2004, come from one group, two from another and two from a third. The seventh is a stand-alone for a group with a history of problematic articles. 

All of the articles are marked “Withdrawn” rather than “Retracted,” implying that the authors were responsible for the move, although some of the articles have been flagged on PubPeer. 

Two of the six retractions are for papers from a group led by Anurag Tandon, a neuroscientist at the University of Toronto. One, from 2006, was titled “Cytosolic proteins regulate α-synuclein dissociation from presynaptic membranes.” and has been cited 38 times. It was first flagged on PubPeer in 2016. According to the notice: 

This article has been withdrawn by Sabine Wislet-Gendebien, Cheryl D’Souza, Toshitaka Kawarai, Peter St George-Hyslop, David Westaway, Paul Fraser, and Anurag Tandon. The α-synuclein and GAPDH bands at 30 min were superimposed on the earlier time points in Fig. 1B. The GAPDH bands in Fig. 4B were reused in Fig. 7A.

The other involves the 2011 article “Effect of Ser-129 phosphorylation on interaction of α-synuclein with synaptic and cellular membranes”:

This article has been withdrawn by Naomi Visanji, Sabine Wislet-Gendebien, Loren Oschipok, Isabelle Aubert, Paul Fraser, and Anurag Tandon. Gang Zhang was not reachable. The pS129 immunoblot in Fig. 1 was used for both membrane and cytosol fractions. In Fig. 4B, the pS129 blots from representative WT and A30P tissues are the same.

The paper has been cited 39 times.

Tandon told us: 

I was contacted by the data integrity officer from JBC in March of this year requesting data from several old papers from my lab. I had no idea at the time that there was any issue with the data. As it turned out, it seems that subsequent work by my former postdoc Sabine Wislet-Gendebien after she had left my lab had been flagged on PubPeer for various issues. That presumably triggered an examination of all her earlier work, and hence, of the papers from my group. It is regrettable that her misguided efforts for the ideal western blot led to the withdrawal of these studies that involved other trainees and colleagues, none of whom deserve to be tarnished by this. It is all the more frustrating that the original data that was modified was acceptable to begin with and also replicated in later studies by other members of the lab. We offered to provide corrections for the figures, but JBC editors were not amenable to that possibility for these papers.

The 2004 article, “A novel human phosphatidylethanolamine-binding protein resists tumor necrosis factor α-induced apoptosis by inhibiting mitogen-activated protein kinase pathway activation and phosphatidylethanolamine externalization,” came from a group led by Xuetao Cao, a prominent Chinese immmunologist whose work has been investigated for image manipulation — problems first identified by data sleuth Elisabeth Bik. Cao now has at least five previous retractions and three expressions of concern, by our count. 

Bik raised questions about the 2004 paper on PubPeer last November. On September 25, the JBC issued the following statement: 

This article has been withdrawn by the authors except for Hongzhe Li, who could not be reached. In Fig. 2A, lanes 3 and 4 of the hPEBP4 panel and lanes 3, 4, and 6 of the actin panel are the same. In the hPEBP4 panel of Fig. 2B, the last two lanes are the same. In the actin panel for the same figure, lanes 3 and 4 are duplicates and lanes 6 and 7 are duplicates. In Fig. 2C, lanes 3 and 4 of the actin panel are duplicates. Lanes 8 and 9 of the same panel are duplicates. Lanes 1 and 3 of the myc immunoblot in Fig. 3C are the same. The last lane of the actin immunoblot in Fig. 4A was reused in the first lane of the actin immunoblot in Fig. 4B. The first and last lanes of the Raf-1 immunoblot in Fig. 4C are the same. A portion of the FACS plot shown in the top middle panel of Fig. 5A was reused in the bottom left panel. The lower phospho-JNK band in lanes 5 and 12 in Fig. 6A are the same. Lanes 6 and 7 of the actin immunoblot were reused in lanes 10 and 11.

The withdrawing authors do not agree with some of the Journal’s analyses and state that their analysis shows differences in some of the concerned bands that exclude reuse of the same bands. The withdrawing authors state that they repeated all of the relevant experiments and that the results are the same in their hands. The withdrawing authors state that the issues described above do not affect the interpretation of the data or conclusions of this work. This article describes the novel molecule hPEBP4 and characterizes its apoptosis-resistant function in breast cancer. The withdrawing authors state that its function and mechanism have been confirmed in breast cancer and other types of cancer by other independent studies (Garcia, R., et al. (2009) EMBO Rep. 10, 278–284; Wang, S., et al. (2017) Biomed. Pharmacother. 90, 659–664; Huang, R., et al. (2019) Mol. Carcinog. 58, 135–143; and Yu, G., et al. (2013) Tumor Biol. 34, 91–98). The withdrawing authors thus stand by the overall conclusion of the study.

The paper has been cited 84 times, according to Clarivate Analytics’ Web of Science.

The other withdrawal for Cao’s group involves this 2006 article, titled “hPEBP4 resists TRAIL-induced apoptosis of human prostate cancer cells by activating Akt and deactivating ERK1/2 pathways.” Per the notice:

This article has been withdrawn by the authors except for Hongzhe Li, who could not be reached. In Fig. 2C, lanes 7 and 14 of the p-ERK1/2 immunoblot are the same. In Fig. 2F, the upper p-ERK1/2 bands in lanes 3 and 4 are the same. Portions of the actin immunoblot from Fig. 2C are reused in Fig. 2, E and F. There are several duplicated features in the FACS plots shown in Fig. 3 (A and B). Portions of the Procaspase 3 immunoblot in Fig. 4B were reused in Procaspase 3 and BID immunoblots in Fig. 4D. The withdrawing authors do not agree with the Journal’s analysis of Fig. 2G and Fig. 4 (C and E). The withdrawing authors state that magnification of the bands shows differences that exclude duplication. The withdrawing authors state that they repeated all of the relevant experiments and that the results are the same in their hands. The withdrawing authors state that the issues described above do not affect the interpretation of the data or conclusions of this work. This article characterizes the apoptosis-resistant function, in prostate cancer, of the novel molecule hPEBP4 identified by the authors from breast cancer. The withdrawing authors state that its function and mechanism have been confirmed in prostate cancer and other types of cancer by other independent studies (Li, W., et al. (2016) Biomed. Pharmacother. 81, 1–6; Huang, R., et al. (2019) Mol. Carcinog. 58, 135–143; Zhang, D., et al. (2016) Tumor Biol. 37, 1699–1705; Huang et al. (2016) J. Neurooncol. 127, 235–242; and Liu, H., et al. (2012) Tumor Biol. 33, 267–273). The withdrawing authors thus stand by the overall conclusion of the study. 

The paper has been cited 46 times.

Of the last two papers, one comes from a group of pediatrics specialists in Philadelphia led by Drexel’s Francesco De Luca. The article, “Nuclear factor-κB p65 facilitates longitudinal bone growth by inducing growth plate chondrocyte proliferation and differentiation and by preventing apoptosis,” was published in 2007 and has been cited 46 times. According to the notice: 

This article has been withdrawn by Shufang Wu, Geoffrey Rezvani, and Francesco De Luca. Doris Fadoju could not be reached. Fig. 3F was reused from Wu, S., et al. (2006) J. Biol. Chem. 281, 51250–5127. Lanes 2 and 3 of Fig. 4A are the same. The actin immunoblots in Figs. 5D and 8F are the same. The β-actin panel from Fig. 5C was reused in Fig. 7A. Portions of the actin immunoblot in Fig. 5E were reused in Fig. 5F and Fig. S7. The first lane of the BMP-2 immunoblot in Fig. 7E was reused in Fig. 7F. Additionally, a portion of the actin immunoblot in Fig. 7E was reused in Fig. 7F. Portions of Fig. S5 were reused in Wu, S., et al. (2008) J. Biol. Chem. 283, 34037–34044. The withdrawing authors affirm that all the experimental results and conclusions reported in this manuscript accurately reflect the experimental work carried out in their laboratory. However, given the issues listed above, the withdrawing authors wish to withdraw this article to preserve the full accuracy of the published scientific literature.

The retraction is the second for De Luca in the JBC, which also removed a 2013 paper titled “ Increased expression of fibroblast growth factor 21 (FGF21) during chronic undernutrition causes growth hormone insensitivity in chondrocytes by inducing leptin receptor overlapping transcript (LEPROT) and leptin receptor overlapping transcript-like 1 (LEPROTL1) expression.”

Per the notice, issued September 11: 

This article has been withdrawn by Shufang Wu, Alexei Kharitonenkov, Julie Dam, Ralf Jockers, and Francesco De Luca. Tal Grunwald could not be reached. In Fig. 1C, lanes 1 and 2 of the actin immunoblot are the same as lanes 3 and 4. In Fig. 2C, lanes 1 and 4 of the LEPROT immunoblot are the same, and lanes 1 and 3 of the LEPROTL1 immunoblot are the same. There is also a splice in the LEPROTL1 immunoblot.

The withdrawing authors affirm that all the experimental results and conclusions reported in this manuscript accurately reflect the experimental work carried out in their laboratory. However, given the issues listed above, the withdrawing authors wish to withdraw this article to preserve the full accuracy of the published scientific literature.

The paper has been cited 29 times.

Finally, the JBC has retracted a 2011 article by Mamta Chawla-Sarkar and colleagues, titled “Active participation of cellular chaperone Hsp90 in regulating the function of rotavirus nonstructural protein 3 (NSP3)” and cited 24 times.” According to the notice:

This article has been withdrawn by the authors. The Journal determined that there are similar repetitive features in the background area of Figs. 4B, 5B, 6B, and 7A and Fig. S4, but the authors do not agree to the raised concerns. The authors state that they provided uncropped scans of the autoradiograms from the corresponding and/or replicate experiments done at same time of the original experiments. The authors state that due to the tropical climate, the quality of the autoradiograms may have partially deteriorated as the work was done more than 10 years ago, but still they could not visualize any repetitive elements in the background. However, the authors could not provide the original, uncropped electronic copies of the scans that were used for preparation of the published images. The authors state that all available hard copy autoradiograms present in the laboratory records were reviewed and verified favorably by the institutional review committee and a third-party forensic laboratory. The authors state that they had provided the cropped tiff files from the final assembled figures used during manuscript submission as well as figures downloaded from the JBC website to the forensic laboratory. Copies of both of the reports were shared with the Journal. The authors affirm that all the experiments were performed appropriately and the concerns raised do not affect the accuracy of the results or the conclusions of the scientific work. However, in the interest of maintaining the accuracy in the published scientific literature, the authors wish to withdraw this article but plan to publish a new version of the article corroborating the findings of this work in the future.

Chawla-Sarkar, of the National Institute of Cholera and Enteric Diseases in India, now has two retractions and three expressions of concern in the JBC, by our count, as well as corrections in two other journals.

Like Retraction Watch? You can make a tax-deductible contribution to support our work, follow us on Twitter, like us on Facebook, add us to your RSS reader, or subscribe to our daily digest. If you find a retraction that’s not in our database, you can let us know here. For comments or feedback, email us at [email protected].