Last week, Nature retracted a highly-cited paper published in 1994 because the authors were unable to reproduce the results, according to the retraction notice (1). But the real reason behind the retraction might have more to do with the first author’s scientific misconduct.
1994 paper (2), first author Karel Bezouska and colleagues from Northwick Park Hospital in the UK, the Charles University of Prague, and the Institute of Microbiology of the Czech Academy of Sciences reported that they had identified carbohydrate oligosaccharide ligands that bound strongly to NKR-P1, a membrane protein on natural killer cells. Natural killer cells are part of the immune system. Their function is to identify and mark virally infected cells and tumor cells for death.
Some tumor cells, however, are resistant to natural killer cells, allowing them to evade targeted cell destruction. So when the researchers claimed in their paper that incubating resistant tumor cells with the oligosaccharide ligands they had identified made the cells vulnerable to elimination by natural killer cells, it was considered a breakthrough.
“The findings were indeed considered a big at the time,” lead author Ten Feizi, whose lab Bezouska was working in when he wrote the 1994 paper, said in an email to BioTechniques. Feizi is now a professor at Imperial College London.
Since publication, the paper has been cited 257 times since publication, according to Thomson Reuters Web of Knowledge. “It is most likely that many of the citations were by scientists unaware of the irreproducibility. I say this because I made a point of informing key people in the field about the problem, and between us when refereeing papers citing the article, we discouraged citation,” said Feizi.
Two years later, a correction to the paper was published in Nature (3). In that notice, the researchers said that they were unable to reproduce the results. As a result, the team was re-evaluating the experimental system, which was set up in Prague by three of the study authors: Bezouska, Anna Fiserova, and Miloslav Pospíšil. The team promised to publish the results of their re-evaluation as soon as possible.
But it wasn’t until just last week, 19 years after the original article was published, that the paper was finally retracted. And that retraction came without the signatures of Bezouska, Lawson, Pospíšil (now deceased), and another author who could not be found.
As it turns out, the retraction only came to pass because first author Bezouska was found by an ethics committee at Charles University in January 2013 to have “most likely” committed scientific misconduct in the experiments published in the 1994 Nature paper and in follow-up studies, as reported by Retraction Watch. As a result, Bezouska was terminated from the Charles University and the Institute of Microbiology. In their press releases, the 2 organizations urged his coauthors to review the more than 100 peer-reviewed publications that he had published for invalid results.
Although the irreproducible data from the 1994 paper were mentioned in the report, the most conclusive evidence of Bezouska’s misconduct came from Professor Vladimir Kren and three students from his laboratory at the Institute of Microbiology who, in 2012, suspected someone was manipulating their NKR-P1 carbohydrate binding experiments. So they installed a camera to monitor their samples overnight. The camera caught Bezouska red-handed, showing him manipulating the samples to generate positive results.
“…Bezouska fabricated practically all results dealing with NKR-P1 and CD69 receptors,” said Kren in an email to BioTechniques. CD69 receptors are another surface protein on NK cells. Bezouska claimed to have identified high-affinity ligands for CD69 receptors in a 1995 Biochemical and Biophysical Research Communications paper (4). “At present, we estimate that there [are] some 20 papers contaminated by this misconduct, and these papers have to be corrected or retracted.”
Although Feizi wanted to retract the 1994 paper when the correction was published in 1996, Nature’s editorial policy at the time required that retractions be accompanied by the signature of all authors, according to an email that she wrote to Retraction Watch. Feizi, however, did publish two studies in other journals that re-evaluated the results from the 1994 Nature paper and the 1995 Biochemical and Biophysical Research Communications paper, concluding that all those results were irreproducible (5-6).
In addition, Feizi published a correction in 1999 to another paper that she co-authored with Bezouska in 1994 in the Journal of Biological Chemistry, also on the binding of carbohydrates to NKR-P1 (7-8).
1. Bezouska, K., C.-T. Yuen, J. O/'Brien, R. A. Childs, W. Chai, A. M. Lawson, K. Drbal, A. Fiserova, M. Pospisil, and T. Feizi. 2013. Retraction: Oligosaccharide ligands for NKR-p1 protein activate NK cells and cytotoxicity. Nature advance online publication(July).
2. Bezouska, K., C.-T. Yuen, J. O'Brien, R. A. Childs, W. Chai, A. M. Lawson, K. Drbal, A. Fiserova, M. Posisil, and T. Feizi. 1994. Oligosaccharide ligands for NKR-p1 protein activate NK cells and cytotoxicity. Nature 372(6502):150-157.
3.Bezou[caron]ka, K., C.-T. Yuen, J. O'Brien, R. A. Childs, W. Chai, A. M. Lawson, K. Drbal, A. Fis[caron]erova, M. PospIs[caron]il, and T. Feizi. 1996. Oligosaccharide ligands for NKR-p1 protein activate NK cells and cytotoxicity. Nature 380(6574):559.
4. Bezouska, K., A. Nepovím, O. Horváth, M. Pospísil, J. Hamann, and T. Feizi. 1995. CD 69 antigen of human lymphocytes is a calcium-dependent carbohydrate-binding protein. Biochemical and biophysical research communications 208(1):68-74.
5. Kogelberg, H., A. M. Lawson, F. W. Muskett, R. A. Carruthers, and T. Feizi. 2000. Expression in escherichia coli, folding in vitro, and characterization of the carbohydrate recognition domain of the natural killer cell receptor NKR-P1A. Protein Expression and Purification 20(1):10-20.
6. Childs, R. A., C. Galustian, A. M. Lawson, G. Dougan, K. Benwell, G. Frankel, and T. Feizi. 1999. Recombinant soluble human CD69 dimer produced in escherichia coli: Reevaluation of saccharide binding. Biochemical and Biophysical Research Communications 266(1):19-23.
7. Kogelberg, H., E. Montero, S. Bay, A. M. Lawson, and T. Feizi. 1999. Re-evaluation of monosaccharide binding property of recombinant soluble carbohydrate recognition domain of the natural killer cell receptor NKR-P1A. Journal of Biological Chemistry 274(42):30335.
8. Bezouska, K., G. Vlahas, O. Horváth, G. Jinochová, A. Fiserová, R. Giorda, W. H. Chambers, T. Feizi, and M. Pospísil. 1994. Rat natural killer cell antigen, NKR-p1, related to c-type animal lectins is a carbohydrate-binding protein. Journal of Biological Chemistry 269(24):16945-16952.