Arsenic: Still Poisonous
In December 2010, NASA announced a momentous discovery: bacteria that could use arsenic instead of phosphorus in its genetic material, a direct substitution in the DNA. This would be a huge deal if true, rewriting much of what we know about basic biology. People were hoping for alien life forms, but this would be nearly as important, if much less glitzy.
We already knew that bacteria could use arsenic in their metabolism; the ability to use it in their DNA would mean that they didn’t need phosphorus at all. The most abundant elements in living organisms are: carbon, hydrogen, nitrogen, oxygen, phosphorus, and sulfur. All are believed to be essential for life. Substituting arsenic for phosphorus would alter that basic principle.
Scientists were immediately skeptical of the claims of Dr. Wolfe-Simon and her colleagues. Many insightful critiques were published online, but the authors of the original paper stated that they would only answer peer-reviewed rebuttals. Our own Dr. Athena Andreatis discussed the arsenic findings on her blog, and for Science in My Fiction.
Such a tremendous claim requires immaculate science and immaculate reporting, and neither were apparent.
In June 2011, the original paper finally saw print in Science (it had been available online since December). The reason for the delay: eight additional peer-reviewed technical comments on the original paper, and a response to those comments by the original authors.
I read all of them carefully with the intent of writing a summary to accompany my earlier essays about the arsenic bacteria, but never did. The short version of what I would have said: The eight comments pointed out several errors in methods and analysis. Some of the most major problems were described by several of the comments. I’m not the right kind of biologist to appreciate the nuances of molecular technique, but these descriptions of failings in the research were convincing.
Dr. Wolfe-Simon’s response boiled down to, “We did so do it right.” There wasn’t a substantive response to any of the problems raised.
The story isn’t over, but perhaps close. Dr. Rosie Redfield, one of the most outspoken critics of the original study and author of one of the Science comments, has tried to replicate the arsenic study with stricter methods and failed. Dr. Redfield and her colleagues found no arsenic in any of the bacterial DNA.
The original study appears to be a case of how not to do science. Based on interviews, Dr. Wolfe-Simon and her collaborators set out to demonstrate that this particular bacterium could substitute arsenic for phosphorus, and did not consider alternative scenarios or design methods that might clearly disprove their hypothesis. Their results were rushed into public without adequate peer review, but with much fanfare from NASA.
But since then, science has proceeded exactly as it should. Other scientists have raised issues in clear, technical fashion, and have tried to replicate these controversial results with more appropriate methods.
Although the key bits are or will be peer reviewed, much of the discussion has occurred openly on the internet, at least among the critics. None of the authors on the original paper have spoken up publicly, to the best of my knowledge, and have only responded to the comments that appeared in Science.
This has been a fascinating story to follow, but for the way the research was presented and followed up on than for the findings themselves. I try to end these articles with some consideration of story ideas. Here, I think the story comes from the people involved, rather than the science itself. (As, arguably, all the best science fiction does.) The discovery could have been anything, but the misinterpretation, critique and defense are what make it so interesting.
Edit (1 Feb. 2012): Dr. Redfield has now posted her mansucript on Arxiv for public comment. I have not read it yet, but wanted to let you know.



Sarah — I wanted to write about this but could not (grant deadlines, etc), so I’m very glad you did this excellent update.
Dr. Wolfe-Simon continues the fig-leaf response by saying that she will only comment on Dr. Redfield’s definitive disproof after it has been peer-reviewed. As you point out, her rebuttal to the (peer-reviewed) criticisms was “Did too!” plus alleging that she had been hounded out of science for being a rebel (!) Among her meticulous experiments, Dr. Redfield did the simple but decisive cesium chloride density gradient coupled with mass spec that shows zero arsenic incorporated into the DNA backbone.
Thanks, Athena. Like you, I wanted to write a more detailed piece when the original technical comments came out, but it fell through the cracks. When Dr. Redfield started talking about her work, I knew it was time to say something, even if it wasn’t the detailed analysis I’d been hoping to find time to produce.
I’m helping another author worldbuild and he wants an inhabitable volcanic planet. Just so we’re clear, life that uses a lot of sulfur in its biochemistry is perfectly fine, isn’t it?
Yes, Paul — many archaea use it as one of their energy staples and it’s an integral part of bacterial and eukaryotic biochemistry as well. If he wants sulfur-based DNA, that’s a different story: it would substitute for oxygen (above it in the periodic table, just as phosphorus is above arsenic) and equivalent stability issues emerge.
I’m not surprised that Redfield was unable to reproduce the original results, but deep down I’m a bit disappointed. I think arsenic-based life would be pretty cool.
I’ve seen some people arguing that this paper doesn’t provide any evidence that the original conclusions were wrong, since the methodology was different. It seems like wishful thinking.
Also, with the massive media circus around the original paper, there’s going to be some embarrassment if the original results turn out to be completely wrong.
The methodology was identical — the only difference was that Redfield did it correctly. For example, she bothered to purify the DNA before subjecting it to mass spec analysis. Putting fingers in ears and going Lalalalala won’t change that. And the original conclusions were totally unwarranted even by the original results.