Retraction Reaction

Recently there was some hullabaloo in the small corner of the scientific world that I inhabit with the decision by Science Magazine to retract an article published in 2010 that claimed to have discovered arsenic-based life. Or at least those were the headlines at the time, the reality was that these researchers had claimed to discover a microorganism, a bacterium named Halomonas to be more specific, that could use arsenic in place of phosphorus as a building block of life. This seems like a wild idea, especially for any Agatha Christie fans, since arsenic was her poison of choice, and poison was her favorite means of offing someone to get the mystery started, or keep it going. No blood spatter for Agatha!

One (among several) of the things that make arsenic toxic to humans, and all animals, and most plants is that it can substitute for phosphorus in things like adenosine triphosphate (ATP) the energy currency of our cells. When it does, ATP no longer works as ATP, and that’s a short recipe for a long rest underground. Nonetheless, there are microbes that can use forms of arsenic like arsenate, not for ATP, but as a substitute for oxygen to respire and live, there are also microbes that can oxidize arsenite (the product of arsenate reduction) to gain energy for life. Amazing stuff, and shown quite conclusively to the point where there is a well characterized suite of genes responsible for these reactions.  So clearly arsenic is not toxic to these microbes. In fact, many microbes carry genes that allow them to detoxify arsenic (a different process from using arsenic for growth); this tells us that arsenic is relatively common, at least in trace quantities, in many environments.

So back to the arsenic-based life story. The main claim in the 2010 Science paper was that the newly discovered microbe could use arsenic in place of phosphorus in its deoxyribonucleic acid or DNA. The picture might be a bit clearer if we called it deoxyribonucleic phosphoacid, because a phosphoric molecule with acidic properties is a structural component of the nucleotides that compose DNA as shown.

Arsenic (As, atomic number 33) turns out to be just above phosphorus (P, atomic number 15) in the periodic table and shares some of its properties, hence its ability to substitute for P.  What if in some places where arsenic is more abundant, or better yet, on another world from ours, life was framed around using arsenic instead of phosphorous as a fundamental building block of life. This was certainly one of the hopes of this group of researchers, showing an alternative biochemistry to what we know to be true for all life on Earth…or is it? Heady stuff.

When it came out, the paper by Wolfe-Simon et al claiming that they had discovered this bug that could use arsenic instead of phosphorus as an essential part of its DNA (now it would be deoxynucleic arsenoacid) was big news. NASA, whose funding supported part of this work, held a press conference announcing it as a major finding in the possible search for extraterrestrial life. Despite being published in a prestigious journal known for rigorous peer review, there was almost immediately skepticism, mostly from biochemists (none of the authors was a card-carrying biochemist). These folks questioned technical aspects of how effectively arsenic could really replace phosphorus, and methodological aspects of how sure the researchers could be that they really had eliminated every bit of phosphorus from the medium used to grow these bacteria. This latter concern keeps me up some nights, since it can surprisingly difficult to be sure you’ve scrubbed every last atom of an element out of an experiment.

Initially the criticisms were written, but within a few months experimental work by other teams showed in detail how the original authors may have been led astray and/or mis-interpreted their sensational findings. The arsenic paper’s authors rebutted the written criticisms of their work, but also never published any follow-up work to demonstrate that indeed arsenic did replace phosphorus.

What’s also somewhat remarkable is that Ron Oremland’s lab, where the discovery was made (or was claimed to have been made), provided the microbial culture of Halomonas to these other labs that then used it to rebut the original claim. While the fine print in most journals says that when you publish a paper you are willing to make materials, in this case a bacterial culture, available to others, the real world is rife with stories about how hard it can be to get your hands on a competitor’s materials.  

Wolfe-Simon, the lead author, who was a postdoc at the time in Oremland’s lab moved to another lab before leaving science. Interestingly, she was recently the subject of a New York Times article (published prior to the papers retraction by Science). The NYT article goes into detail about the fact that publication of the arsenic paper coincided with the beginning of ‘trial by internet’. This ‘trial’ did not go well for Wolfe-Simon or her co-author’s, since it included personal attacks against her, unrelated to the science. Subsequently, ‘trial by internet’ has become a pervasive, cruel, and entirely unproductive part of our culture.  Seriously, if you can name five new ideas that have been born by the ‘trial by internet’ crowd that have real, atomic substance, I’ll buy you a cookie; if you can name one, I’ll buy you two cookies!   

So, should Science have unilaterally retracted the arsenic paper? Personally, I don’t think so. Primary reasons for retractions by journals as opposed to authors, is in cases of fraud or plagarizing the work of others. No one has accused Wolfe-Simon or colleagues of fraud, indeed as mentioned above, Oremland provided the Halomonas strain to the labs that published scientific rebuttals.  In terms of retraction, cases where original claims are shown to be dubious by follow-on studies do shade into a gray area, reflecting many of the 'gray areas' in real life, and are perhaps most valuable for this very insight.

Authors may also ask for their papers to be retracted. This usually happens when the lab itself finds errors in the work that undermine the major findings of the paper. These can range from discovering a piece of equipment wasn’t properly calibrated, to realizing a lab member was making up, or manipulating data to make it fit the hypothesis. Should the authors have retracted the arsenic paper when it was shown the mostly likely explanation for their results was contamination of the medium with phosphorus, perhaps so, but they chose not to.

It always hard to judge the motivations of others, actually, it’s really easy, but what makes it hard is that you often end up being wrong. I don’t know Wolfe-Simon, but do know several of the other authors on the arsenic paper (sadly, Ron Oremland passed away in 2021), and consider them all first-rate scientists. The Oremland lab led a number of important discoveries about how microbes can use arsenic for carrying out important metabolic functions that I described above. It’s reasonable that there was some, and possibly quite a bit, of bias towards finding arsenic-based life by this research group. Therein lies a challenge for us all, not letting our biases outweigh our ability to consider all alternatives.

I’ll end by pointing out that some of the funding for arsenic work was provided by the NASA Astrobiology program, which is focused on understanding the potential for life to exist outside of Earth. It also funds studies on how life originated on Earth, and what might be learned from studying life in ‘analog environments’ on Earth that might approximate conditions on on another planet, for example a lake with very high concentrations of arsenic.  As a funded field of study, astrobiology largely arose from the purported discovery in 1996 of life forms preserved on a Martian meteorite in a paper published in Science. At the time, that paper was hailed as the most conclusive evidence for ‘life on Mars’, with a distinguished author list, and multiple lines of evidence that the Allen Hills meteorite harbored the remains of Martian microbes. There was also a NASA press conference touting the discovery. This generated considerable excitement and controversy in the scientific world. Over the next few years each of those lines of evidence for 'life on Mars' was shown to be false in a series of follow-on papers by other researchers. This all played out in the scientific literature, the main stream press, and the internet, but was before the days of social media. This paper is still available at Science, or the page proof is available on the internet.