Why one biologist says it’s not too late to retract the “arsenic life” paper

David Sanders

An anniversary has prompted this reconsideration of the revolution in biochemistry that wasn’t: the “arsenic bacteria.” Just over 10 years have passed since an infamous Dec. 2, 2010, NASA press conference, which promised the revelation of “an astrobiology finding that will impact the search for evidence of extraterrestrial life.” 

Of course, nothing of the kind occurred. The carefully curated moment was less informative for its scientific value — in effect, nil — than for what it says about how years of attention-seeking and speculation in biology can drive an agenda. Equally concerning, despite the intervening decade in which other researchers debunked the overhyped result, is that the journal involved has yet to retract the article in question, allowing it to live in a zombie state.   

The announcement at the press conference was, to the disappointment of many, the supposed “discovery” of a microbe that could grow on arsenate in the absence of phosphate and incorporate arsenic instead of phosphorus in macromolecules such as nucleic acids and proteins that was being published in ScienceSteven Benner (referring to himself as a curmudgeon) was the only individual at the press conference who talked real sense by undermining the claims.   Mary Voytek, NASA Senior Scientist for Astrobiology (a position she still occupies) employed a Star Trek analogy:

This in our mind is the equivalent of finding that Horta which is a silicon-based life, substituting carbon, which is what we think all life forms are made of, with silica. Now we are talking about an organism that we think we are talking about an organism that, if not replacing all of it, appears to be using another fundamental component of life. The story is not entirely carbon. Nitrogen, phosphorus, and the other essential elements–it is replacing arsenic for phosphorus. This is a huge deal.

The media, while in some cases mentioning the caveats about the research, made this “breakthrough” international front-page news.  In the Wall Street Journal Paul Davies, an Arizona State University (ASU) physicist, admitting pride in his lack of expertise in chemistry, crowed about his role in promoting the postdoctoral researcher, Felisa Wolfe-Simon, who was the main marketer for the phantasm of arsenic replacing phosphorus.  

Davies had been impressed with her “bubbly personality.” and declared victory in “proving” all the well-informed scientists who had earlier expressed doubts about the line of investigation wrong.  He was not troubled by the evident desperation that his protege expressed when she named the arsenate-resistant bacterium GFAJ-1 for “Get Felisa a Job.” Neither was he troubled by the apparent pressure to get a positive result under which he perceived that Wolfe-Simon was operating: “At every step, the experimental results might have shot down her big idea, spelling the probable end of a promising scientific career.”  It should be noted that at no time does Davies mention in his WSJ piece that he is an author on the Science article. 

Michio Kaku, a professor of physics at the City University of New York, wrote, also in the Wall Street Journal, “It means that every biology textbook now has to be revised. Even the very definition of life may have to be changed.”  Kaku and others claimed that the revelations might support the idea that there exists a “shadow biosphere” on earth, an idea not originated by Davies but relentlessly peddled by him for at least fifteen years until today.   

But there is no evidence for a shadow biosphere, and its existence would not solve any outstanding scientific problems. The shadow biosphere is science fiction and resembles an article of religious faith.  While it is true that what is normally referred to as “science fiction” sometimes comes true, in actuality it is in virtually every case merely the technology described in that fiction that manifests as reality.    

It was immediately obvious to the scientific world that the article was rubbish. A number of scientists, including Antoine Danchin, Steven Benner, and I recognized that the most likely explanation for the apparent ability of the bacteria to grow in the presence of arsenate was that they were, in fact, growing on trace phosphate that was present in the growth medium, probably as a contaminant of the arsenate itself.  The supplementary material in the paper itself supported that conclusion. The bacteria were highly resistant to arsenate, although they were by no means unique in this trait.  Indeed, there are even eukaryotes such as nematodes and brine shrimp that live in the arsenate-rich waters of Mono Lake

Others, such as University of British Columbia microbiologist Rosie Redfield, focused upon the inadequacy of the methods used to provide evidence for the incorporation of arsenic into the bacterial DNA.  Redfield disseminated her opinions on her blog — an important moment in the history of online post-publication review that evaded the strictures that journals traditionally placed on such expressions of data analysis.  The authors and staff at NASA, despite having gone outside of normal scientific channels to broadcast their flawed findings, pointedly refused to engage with their critics insisting that responses were only appropriate through the normal scientific channels.

Science held back on publishing the article in the magazine for six months.  The article was accompanied online with a record eight Technical Comments criticizing nearly every aspect of the article.  The authors’ response addressed practically none of the points of the commentators and introduced new distortions of the data in the article.  Comparing the data for the phosphorus content of their media in the response (for example, 4/5/10 -As/–P medium, 3.7 ± 0.8 μM) to the data for the same medium in Supplementary Table 1 of the original article (3.7 ± 0.4 μM), it can be seen that the authors doubled the reported standard deviation in order to reach their conclusions.  They also ignored the analysis of another “batch” of media reported in Supplementary Table 1 of the original article that contradicted the whole premise of the article.

Wolfe-Simon and her colleagues made numerous pronouncements about future research and articles concerning GFAJ-1.  While working in the laboratory of John Tainer at the Lawrence Berkeley National Laboratory, Wolfe-Simon implausibly suggested, according to Chemical & Engineering News, that determining the three-dimensional X-ray crystallographic structure of the GFAJ-1 ribosome from cells grown in the presence of arsenate was the way forward.  She also maintained counterfactually that her original paper never actually claimed that arsenate was being incorporated in GFAJ-1’s DNA, but that others had jumped to that conclusion.

Tainer said of a preprint from Redfield refuting the original Science article,”I would call it more of an attack than a paper. It aims to shut the door on additional research.” 

Long after the scientific community had eviscerated the arsenic-bacteria claims, segments of the media continued to act as if there was something of value in the chimeric vision.  Despite being warned about the pseudoscientific content, the TED Conferences allowed Wolfe-Simon to speak in 2011.  Time declared her one of the 100 most influential people in the world.  She also was featured in Glamour giving advice on how to introduce oneself to “rock-star scientists,” beginning by informing them how “she admired their work.”  

Some members of the media continued to maintain  an avid interest in the affair.  In an enterprising move, Dan Vergano of USA Today obtained the reviews of the manuscript by the Science editor, Caroline Ash, and the three peer reviewers, along with the responses of the authors.  They reveal that with the exception of Reviewer #2, the editor and the reviewers were completely unequipped for the task at hand.  The light shone upon the inadequacies of peer review illuminated for many the shortcomings of peer review and reinforced the burgeoning post-publication-review movement that arose from the arsenate-bacteria fiasco.

It emerged from the authors’ response to the reviews obtained by Vergano that additional largely irrelevant and poor-quality data (as well as three authors) were attached to the paper with seemingly little oversight.  This incident illustrates how peer review can sometimes be circumvented by inclusion of material after the review.  Often there is an additional round of evaluation, but reviewers are frequently too otherwise occupied to give it their full attention.  Also, the large number of authors diffused ultimate responsibility for the article.  As often happens, everyone could take credit for a high-profile article, but no one would take responsibility for the fact that it was so flawed.

Tom Clynes (Popular Science) wrote an article about Wolfe-Simon that, while largely balanced, misrepresented the problems with the science and the peer review.  “Yet there’s no indication that the paper’s reviewers were asleep at the switch,” Clynes stated.  He’s clearly wrong.  He went on to proclaim, “What made the level of criticism so extraordinary is that the paper, in itself, is not so flawed that it should not have been published. The argument was compelling, the conclusions were measured, the data was thorough, and the paper made it through the same peer-review process as other articles in Science.”  Utter nonsense.

Clynes describes the filming of an upcoming episode of PBS’ Nova by the director Oliver Twinch.  I have been watching Nova since its inception and it was also the one television program that I routinely watched with my children.  I was appalled that Nova was going to be contaminated by the arsenic-bacteria nonsense in a show titled “Finding Life Beyond Earth: Are We Alone?”

I contacted Paula Apsell, the executive producer of Nova, to encourage her to have the fallacious material removed.  After another expert was consulted, the program appeared to have been altered.  There is a reference to life in “lakes full of poisonous arsenic” (1:17) and images of Mono Lake, but the damage was limited.  There are a couple of out-of-focus video clips of  Wolfe-Simon, but there is also what appears to be an abrupt cut with a transition to microbial cave life.  Interestingly, a still image of Wolfe-Simon among many others was shown at the end of broadcast episodes for several years. 

The Science article was thoroughly refuted by later published studies although, of course, there were no valid significant data that needed to be refuted in the original article.  It was ranked the 68th biggest failure of the decade by The Verge. 

The episode revealed the limits of peer review, the failure of institutions, the immense truth-distorting power of fame-seeking and careerism, and the emerging clout of electronic post-publication analysis. The NASA press conference and its media coverage, in particular, undermined public faith in science. It has been used as a case study for teaching students about scientific discourse and skepticism about the scientific literature

None of the protagonists, authors, editors, reviewers, institutions associated with this shameful press conference and article has expressed regret or admitted error.  Very few members of the media who breathlessly touted this mirage have done so, either. For misrepresenting its own data, and for sheer implausibility, the article should have been retracted a decade ago. But it’s not too late, as recent retractions of decades-old papers have shown. Retracting it today — which I argued for on this site eight years ago — would still be the right move.

David A. Sanders is an associate professor of biological sciences at Purdue University.

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 team@retractionwatch.com.

12 thoughts on “Why one biologist says it’s not too late to retract the “arsenic life” paper”

  1. Sorry, do you have a Twitter mob and 1000 signature petition?

    If expert evisceration is all you have, expect it to stay up. That is the usual experience of “data thugs”.

  2. Frankly I haven’t given this garbage much thought after 8 technical comments. Even a firing squad needs only 5 riflemen…

  3. I’m completely against fraud, but a tenured faculty member moralizing against it to try to help prevent it from occurring seems disingenuous. The reason why fraud occurs, much of the time, is because the people doing it have relatively crappy jobs in system (academic research) that dualizes the labor force between crappy (temporary and low paying graduate student or post-doc jobs) and good (tenured faculty) jobs. I think its OK if a member of this low strata moralizes here, but the tenured faculty just need to keep their opinions to themselves–they really cannot understand the pressures that the lower strata may be under to do this kind of stuff.

    Usually tenured faculty love the dualized system of academic research that has created the problem. The system needs to be torn down, and data generators and their managers (faculty) need to have similar job protections and salary, and then this problem will go away. I really cannot see any faculty member being for this, so they need to sleep in the bed they made for themselves, IMO.

  4. I’m not an expert in any experimental science (being a pure mathematician by background) but I would have thought that 10 years was ample time for the acid test (see what I did there?) which is reproducibility. Have these experiments been repeated and if so, what were the results? This question isn’t addressed, so I suppose that the answer is that no-one has attempted to do so. Until then, this blog post represents the author’s opinion: an opinion which is vehemently argued, shared by some (and not by others), maybe even correct, but an opinion nevertheless.

    1. Ah, I see I was too hasty. The sentence “The Science article was thoroughly refuted by later published studies although, of course, there were no valid significant data that needed to be refuted in the original article.” links to a study that repeated the experiment and got a different result. But surely, that’s all one needs? The original paper reported a result, a second paper gave a different result. It would have been more helpful to say so explicitly …

  5. You sound really anti-science, a person that doesn’t believe that 97% of scientist are able to reach the same conclusion, a person that doesn’t believe in journalists and that scientists egos, although inflated , exist to help us all.

    What is the matter in. little marketing and some bootlicking journalists that will forget you the moment a new brilliant discovery is put in front to them?

  6. You seem to have an issue with the tenure system in academia, and its other systemic ills. But by itself it doesn’t create crap science. The paper in question is the result of uncritical view of their work by people way out of their depth/competence. You attacking Dr Sanders for holding a tenured position is at best irrelevant. Dismissing opinions of all tenured faculty is a bit too sweeping a generalization, but I see that you started a war with China in another set of comments already, so for you this must seem normal.

  7. The paper was wrong. The results were over-hyped, and the referees probably should not have let it get published in a premier journal. However, none of these are reasons to retract a paper against the wishes of the authors.

    No one criticizing the work has provided any evidence (or even, so far as I am aware, claimed) that the errors in Wolfe-Simon’s work were the result of dishonesty. And that is the crux. That a paper turns out to be mistaken is simply not justification for retracting it, unless the author(s) themselves feel that it should be removed from the published literature. I don’t think that Felisa Wolfe-Simon is a good scientist or that her results are right, but that judgement (no matter how widely shared) are not reasons for retractions.

  8. This paper is not a mistake……ask any chemist how stable esters based on arsenate (as opposed to phosphate) will be in aqueous solution.

    A 1st year chemistry undergraduate could tell you that the idea is nonsensical. The whole thing was about generating a high impact paper in order to establish a career and is an absolute disgrace. I was stunned that it’s not been retracted!

    Big clue in the name of the bacteria….GFAJ…..Give Felicia a Job.

    Sad thing is…..this is actually a very successful career strategy!

  9. I was not aware of this arsenic “finding” but in my experience with peer review the way the system is supposed to work is like this: the manuscript in question is sent to peers who work in a very similar or the same field. These scientists carefully read the content of the manuscript (not just the abstract or the author list), look at the hypotheses and the experiments that were performed to prove or refute the hypotheses. The reviewers then think hard about the claims in the manuscript and the proof provided and say: well, that looks interesting but did you think of this and that and have you performed such and such experiments to exclude such and such errors and to provide further support for your claims. The editor sends the reviewers’ comments back to the corresponding author and tells him or her: this looks interesting but you really should perform these additional experiments. Because, exceptional claims require exceptional proof. So the authors perform the additional experiments, realize that their claims are untenable and presto another paper that never sees the light of day. But maybe I’m a bit deluded as far as the reality of the peer review is concerned.

  10. A response from Felisa-Wolfe Simon:


    copy paste of tweet:

    My paper hasn’t been retracted because there’s no reason to retract it. This is a big lie that continues to be promoted by the misogynist
    Move on people, there’s nothing to see here. #arseniclife



Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.