Science finally retracts the 2010 “arsenic life” paper by Felisa Wolfe-Simon et al.

Fifteen years ago (!) a group of authors headed by Felisa Wolfe-Simon published a paper in Science heard ’round the world, a paper that garnered huge publicity. (If you can’t get it by clicking the second link, since the paper has largely been vanished, try this archived link.)

Why the problem? I explained this in February of this year:

A remarkable discovery appeared in the journal Science in 2010.  Felisa Wolfe-Simon and her colleagues reported finding, in California’s salty Mono Lake, a bacterium that could [in the lab] substitute arsenic for phosphorus in its metabolism.  This was stunning, as phosphorus was thought to be an essential constituent of many biological macromolecules, including proteins and DNA—the latter using phosphorus as part of its backbone.  (The bacterium was, by the way, named GFAJ-1, standing for “Give Felisa a job,” as she was apparently looking for a permanent academic position.)

At any rate, this was huge news, and implied, to many, including hype-promoting journalists, that if life could thrive on arsenic, perhaps the chances of life on other planets was higher than we thought. Wolfe-Simon herself implied that perhaps there was a “shadow biosphere,” on Earth, including organisms that we didn’t know of because their biochemistry was so different from that of life we knew.

The publicity attending this discovery was huge: NASA held a press conference in which Simon was the only one of the dozen authors to appear. Simon also gave a TED talk on this subject, and in 2011 Time Magazine named her one of “Time’s 100 people,” supposedly the most influential group in the world.

The problem, which emerged pretty rapidly, is that this discovery was wrong. The research was sloppy, the reviewers apparently didn’t have the proper expertise to review the paper, and researchers who did have the expertise began pointing out the discovery’s flaws, first online and then in a series of eight critiques published in Science.

You can see my four posts on this paper here, including the original publication plush simultaneous pushback, along with a more recent New York Times article that more or less whitewashed the paper by painting Wolfe-Simon as a victim (see Greg Mayer’s take in the same post). I wonder if the NYT will finally give a more balanced view of the controversy in which Wolfe-Simon fought like a honey badger for the accuracy of the paper’s results. She was far from being a victim.

Click below to read the editorial retraction, but I’ve reproduced it below:

Here’s what Thorp said:

 The Research Article “A bacterium that can grow by using arsenic instead of phosphorus” by F. Wolfe-Simon et al. (1) has been the subject of discussion and critique since its online publication in 2010. In 2011, Science published the print version of the paper accompanied by eight Technical Comments (2–9) and a Technical Response from Wolfe-Simon et al. (10), along with an Editor’s Note (11). In 2012, Science published two papers that failed to reproduce the finding that the GFAJ-1 bacterium can grow using arsenic instead of phosphorus (12, 13). 

Science did not retract the paper in 2012 because at that time, Retractions were reserved for the Editor-in-Chief to alert readers about data manipulation or for authors to provide information about postpublication issues. Our decision then was based on the editors’ view that there was no deliberate fraud or misconduct on the part of the authors. We maintain this view, but Science’s standards for retracting papers have expanded. If the editors determine that a paper’s reported experiments do not support its key conclusions, even if no fraud or manipulation occurred, a Retraction is considered appropriate. 

Over the years, Science has continued to receive media inquiries about the Wolfe-Simon Research Article, highlighting the extent to which the paper is still part of scientific discussions. On the basis of the 2011 Technical Comments and 2012 papers, Science has decided that this Research Article meets the criteria for retraction by today’s standards. Therefore, we are retracting the paper. Author Ronald S. Oremland is deceased. Author Peter K. Weber disagrees with the Retraction. Authors Felisa Wolfe-Simon, Jodi Switzer Blum, Thomas R. Kulp, Gwyneth W. Gordon, Shelley (Hoeft) McCann, Jennifer Pett-Ridge, John F. Stolz, Samuel M. Webb, Paul C. W. Davies, and Ariel D. Anbar also disagree with the decision to retract and have posted an online letter  explaining their objections, available at https://www.science.org/doi/10.1126/science.adu5488#elettersSection, explaining their objections. 

H. Holden Thorp, Editor-in-Chief, Science 

Well, now, Thorp brings up an interesting point: should papers be retracted if their data don’t support the conclusions? In my view, no.  There are tons of papers published in which authors make claims that go beyond the data but in which there is no fraud.  A famous example, and one which I criticized in two papers in Evolution, was Sewall Wright’s famous “shifting balance” theory of evolution.  By combining several real phenomena like genetic drift with speculations about possible phenomena like adaptive peak shifts, Wright confected a grand overarching theory of evolution” (SBT) that was unsupported by the data. But most evolutionists bought the SBT and it appeared in all the textbooks.

But our criticisms were on the mark.  Even Jim Crow, Wright’s colleague and friend at Wisconsin, admitted that we (Michael Turelli, Nick Barton and I) were right. Likewise, many of Theodosius Dobzhansky’s papers have conclusions unsupported by the data.  But they stand unretracted in the literature. And, indeed, even though the conclusions may have been overstated, the data are still valuable, and should not be retracted.  Retraction should be reserved for papers containing fraud or misconduct, not papers whose data don’t support the conclusions, for those papers have still gone through peer review and they have data that might be useful. The main culprits should be the reviewers who let shoddy data into the literature.

And, as Thorp noted above, the authors were given a chance to reply to the retraction. Here’s their response (references can be seen at the links):

Ariel Anbar
Paul Davies
Gwyneth Gordon
Tom Kulp
Shelley (Hoeft) McCann
Jennifer Pett-Ridge
John Stolz
Jodi Switzer Blum
Samuel Webb
Felisa Wolfe-Simon 

In 2011, we published a manuscript in Science proposing that an extremophilic microbe isolated from arsenic-rich Mono Lake, CA (GFAJ-1) utilized arsenic (As) in place of phosphorus (P) in its biochemistry (1). This provocative hypothesis was based on our interpretation of data from growth experiments and compositional studies of key biomolecules. The high-profile nature of the publication’s release focused attention on its interpretations of As-substitution in DNA, drawing widespread attention, criticism, and follow-up research (e.g., 2-5, but see also 6). In the years since, an alternative interpretation emerged that GFAJ-1 is an extremely As-resistant bacterium that remains dependent on P but uses several novel tactics to grow under P-starved conditions (2, 3). 

Nearly 15 years later, the editors of Science have retracted our publication. We do not support this retraction. While our work could have been written and discussed more carefully, we stand by the data as reported. These data were peer-reviewed, openly debated in the literature, and stimulated productive research. 

The editors’ basis for retraction is that the “paper’s reported experiments do not support its key conclusions”. No misconduct or error is alleged. This represents a major shift from the standards Science adhered to in the past, which aligned with those of the Committee on Publication Ethics (COPE). COPE guidelines state that “Retraction might be warranted if there is clear evidence of major errors, data fabrication, or falsification that compromise the reliability of the research findings” (7). In going beyond COPE, the editors of Science explain that “standards for retracting papers have expanded”. 

We disagree with this standard, which extends beyond matters of research integrity. Disputes about the conclusions of papers, including how well they are supported by the available evidence, are a normal part of the process of science. Scientific understanding evolves through that process, often unexpectedly, sometimes over decades. Claims should be made, tested, challenged, and ultimately judged on the scientific merits by the scientific community itself. 

I agree with the authors’ point here, though they are a bit disingenuous, not admitting that a passel of other scientists could not reproduce their data, and therefore they no longer stand by it. Instead, they dug in their heels.  Their point is right but they are not being fully honest about what happened. My solution: Science can put a disclaimer beside with the original paper saying that the results could not be reproduced, but no—no retraction.

But wait! There’s more! Thorp, a man for whom I don’t have a huge amount of respect, had to have the last word, and published the attached along with Valda Vinson, the executive editor of Science.  Again, click on the headline to read, but I’ve put Vinson and Thorp’s response to the authors’ response below:

The text of Vinson and Thorp:

On 2 December 2010, Science published online the Report “A bacterium that can grow by using arsenic Instead of phosphorus” by F. Wolfe-Simon et al., which caused a media sensation and a firestorm in the scientific community that has continued for years. Today, Science is retracting the paper.

The Wolfe-Simon et al. Report described a bacterium, GFAJ-1, that purportedly not only grew in the presence of arsenate, which is normally highly toxic, but grew “by using” the arsenic atoms and incorporating them into nucleic acids. The US National Aeronautics and Space Administration (NASA), which contributed authors to the paper and provided funding, held a press conference announcing the findings as proof of “arsenic life,” a breakthrough in astrobiology. The scientific community immediately expressed skepticism, raising serious questions about both the plausibility of arsenic-containing nucleic acids and the way the experiments were conducted.

Science was flooded with commentary on the problems with the paper and held off publishing it in print. Eventually, the paper was published in the 3 June 2011 issue, along with eight Technical Comments, a Technical Response from the authors, and a note from Editor-in-Chief Bruce Alberts explaining the decision and timing. The authors agreed to make the bacterium available, and in July 2012, Science published two papers showing that the bacterium was resistant to arsenate and did not incorporate it into biomolecules. One of the Technical Comments had pointed out that the nucleic acids that were analyzed were not sufficiently purified before the acquisition of spectra that suggested the presence of arsenic. Given the evidence that the results were based on contamination, Science believes that the key conclusion of the paper is based on flawed data.

At no point has there been any discussion or suggestion at Science of research misconduct or fraud by any of the authors. Nonetheless, the response of many in the scientific community, especially on social media, went beyond technical criticism and instead verbally abused the authors, especially the first author. Science emphatically rejects and condemns all ad hominem attacks that have been directed toward the authors.

In his 2011 Editor’s Note, Alberts explained that the publication of the Technical Comments was “only a step in a much longer process.” We are ending that process today by retracting the paper. We have made this decision after an extensive set of deliberations and discussions among the editors. We have consulted with the Committee on Publication Ethics (COPE) and are confident that our decision aligns with their guidelines for research integrity.

Over the years since the paper was published, and especially in the past 5 years, as research integrity has become an even more important topic, Science has moved to retract papers more frequently for reasons other than fraud and misconduct. In this case, a number of factors led to the publication of a paper with seriously flawed content, including the peer review process and editorial decisions that we made. With this retraction—and with all retractions and corrections—we acknowledge and take responsibility for the role that we played in the paper’s publication.

The authors of the paper all disagree with our decision, as noted in the Retraction. All but one of the authors have signed an eLetter explaining their disagreement. In their eLetter, the authors state that Science “went beyond COPE” in our decision. We disagree. COPE guidelines have for some time allowed for editorial retraction due to honest error or naïve mistakes.

Despite our disagreement with the authors, we hope this decision brings the story to a close. We also hope that the scientific community will engage graciously and with professionalism in this resolution.

Valda Vinson is the Executive Editor of the Science journals. vvinson@aaas.org

H. Holden Thorp is the Editor-in-Chief of the Science journals. hthorp@aaas.org

This would have sufficed as the explanation for the original retraction, even though there should have been no retraction. Instead, what’s right above could have been the editors’ first word, not their last: an explanation of the problems with the paper. The editors simply had to have the last word, publishing two notes instead of one.