A year and a half ago, a cryptic NASA press release focused attention on what it promised would be new evidence in the hunt for life on alien worlds. NASA and the journal involved (Science) sat on the results for days as rumors built. Eventually, the paper debuted: NASA scientists had found evidence that life on Earth could rely on alternate chemistry, one that replaced the phosphorus used in many biomolecules for a chemical relative that is usually toxic: arsenic.
Controversy did not end there. Researchers quickly identified a number of holes in the initial analysis, both logical and experimental. Less than six months later, Science published a series of responses to the original paper that raised significant questions about its accuracy. Now, the topic is back in the pages of the same journal. Two labs have obtained the original arsenic-tolerant bacteria and shown that some of the original paper's conclusions are completely off-base.
The bacteria themselves were originally obtained from California's Mono Lake, which naturally contains high levels of arsenic. Over multiple generations in the lab, scientists forced them to evolve further, gradually ramping up the arsenic concentrations while lowering the amount of phosphorus available to the organisms. After enough generations, the resulting bacteria reportedly had some rather exceptional properties: they could get by without having any phosphorus added to their growth media, but only if arsenic were added. And, possibly as part of that adaptation, the arsenic started appearing with biomolecules that normally contain phosphorus (although, as our initial coverage noted, the data didn't indicate how the arsenic was being used).
The responses suggested a couple of problems with the work. On purely theoretical grounds, we know that arsenate compounds should spontaneously fall apart in water. This makes their apparent presence in biomolecules difficult to accept. And, on purely practical grounds, it turns out to be really difficult to get rid of all the phosphorus from the other lab chemicals scientists typically rely on. That suggests these bacteria weren't really doing without the stuff.
The new papers, from groups in Zurich and Vancouver, nail down a lot of loose ends. The material suggests that the initial results were a mix of artifact and not pushing the experiments sufficiently far.
First, the inability to grow without phosphate. The Zurich group shows that these organisms can survive down to remarkably low amounts of phosphorus—the equivalent of adding 0.000162 grams for every liter of liquid growth medium (for those of you hung up on English units, that's "almost nothing"). But if you use ultrapure materials that drop the phosphorus levels down to below what the team could detect (somewhere less than 20 percent of that level), the bacteria won't grow. They still need phosphate.
The original paper reported that when phosphate starved, the organisms needed arsenic in their media to grow at all. But the Vancouver group found that wasn't the case. They suggest this was an artifact: the bacteria also had a requirement for a specific amino acid, something that was not reported in the original paper. They suspect that the arsenic solution used in the original experiments had been contaminated with the amino acid, explaining its apparent requirement.
So the bacteria don't appear to require arsenic. But, just as clearly, they can tolerate extremely high levels of the substance, and there was evidence that it is incorporated into biomolecules, including DNA. The new papers tackle this as well. The Zurich group used a technique called mass spectrometry, in which every chemical in the cell is separated based on how much an individual molecule weighs. The results can be compared to a database of known weights to identify most molecules very specifically. In this case, the authors updated their database to include molecules with phosphate replaced by arsenate.
The results largely came back blank. Some arsenate was associated with sugars, but this apparently formed spontaneously, without the need for any biochemical activity. If you let the standard lab bacteria E. coli grow in a medium containing some arsenic, they formed all of these compounds as well. This suggests they have nothing to do with a newly evolved tolerance to arsenic.
The Vancouver group focused on the reported association of arsenic with DNA, a phosphate-rich molecule. That makes DNA a great candidate for detecting even a rare substitution of arsenate for phosphate, a change that should make the DNA sensitive to breaking down in a simple solution of water. But, even after extended storage in water, DNA obtained from these bacteria was as stable as that obtained from the same strain grown without arsenic. Based on the groups' measurements, that means that less than one-in-25,000 phosphates could possibly be swapped out.
Given all this data, it's hard to argue with the Vancouver team's conclusion: "The end result is that the fundamental biopolymers conserved across all forms of life remain, in terms of chemical backbone, invariant."
But an argument may be brewing. MSNBC's Alan Boyle has been communicating with the team behind the original findings, and they suggest there's a paper in the works that backs up their initial findings. Although the two new publications seem like the final word on the topic, more appears to be forthcoming.
Arsenic and open science
The researcher who led the Vancouver group, Rosie Redfield, deserves special mention in this saga. Redfield has embraced blogging as part of the scientific process, and was one of the first researchers to publicly raise doubt about the original arsenic results. Her doubts caught the attention of science journalists and led to some of the first skeptical coverage. Once she obtained the bacteria and started working on them, she put results up on her blog as they came in. As she prepared her paper for publication, she also put draft versions up on the arXiv. That is common in physics and astronomy, but almost unheard of in biology.
Not only is Redfield pushing her fellow biologists for greater openness, she has apparently pushed the journals as well. Science normally distributes papers that will be released in a given week to the press on Sunday night, but asks the press to hold them for a Thursday embargo. Redfield was scheduled to speak at an evolution meeting last night, and let the editors of Science know that her arsenic work would be the topic (and, naturally, she blogged about letting them know, too). In response, Science made these two papers available without restrictions last night.
Had Science been equally responsible with its embargo of the initial results from NASA, some of the hype that developed in response to the uncertainty might never have become a problem in the first place.
Science, 2012. DOI: 10.1126/science.1218455, 10.1126/science.1219861 (About DOIs).
reader comments
31