A New Game for a New Year

Mono Lake. Source.

by Elio

Let's come up with ways to tell definitively whether or not the nucleic acids in the Mono Lake bug GFAJ-1  have substituted a sugar-arsenate backbone for the usual sugar-phosphate one, and if so, to what extent. We suspect that everyone will have a different notion. Please share yours by posting a comment.

Comments

[christoph weigel]

i can hardly believe that Wolfe-Simon et al. did not simply put an x-ray film on their gel (Fig.2A). their nucleic acid prep. was apparently performed rapidly and carefully enough to have distinct bands retained for 16/23S RNA, which could also have been eluted to perform conventional 2D thin-layer chromatography to show 73As-incorporation into the RNA, even at low rate.

sorry, but to me GFAJ-1 (Halomonadaceae) seems not particularly interesting. adaptation to 'adverse' conditions is just what we know all bugs are perfect in, also the Gammaproteobacteria, a fairly novel branch. at least i cannot see a "profound evolutionary and geochemical significance". i found the selenocystein story by the Böck lab in the early 90es far more intriguing (and mind-opening).

[Erika Szymanski]

Crystallography, anyone?

[qetzal]

I can't believe they didn't run their supposedly 73As-labeled DNA on a gel and then either cut out the major genomic DNA band & determine radioactivity, or better - done autoradiography. That would show fairly definitively whether 73As was truly associated with the DNA (though it still wouldn't prove substitution of As for P in the backbone).

(Note that this should work even if only a small fraction of P was replaced with As. In contrast, I think a Meselson-Stahl type experiment would require a very high percentage replacement.)

Instead, the authors tried to argue that the 11% of 73As in the aqueous fraction after cell lysis and organic extraction represented DNA incorporation. But we know there will be many other organic-insoluble molecules in that fraction.

I also found it curious that in the Materials and Methods, they say they determined 73As in all fractions including the EtOH-precipitated DNA/RNA pellet. However, Table 2 only reports the percent 73As recovered in the final aqueous fraction (parenthetically labeled "DNA/RNA"). It sure would have been nice to show how much of that 11% comes down during EtOH precipitation.

[Ted K. Raab]

When I first saw the headlines, I imagined that the first experiment would be to precipitate the DNA with EtOH, and do a simple infrared absorption experiment. Fairly certain that arseno-diesters will show peaks at lower frequency than phospho-diesters.

[barry goldman]

Can we synthesise arsenic based nucleotides? for RNA or DNA? Can we set up a minimal system for polymerizing them and see at what point it jams?

have we synthesised some small arsenic based peices of RNA or DNA and explored their stability etc.. under different conditions?

[EF!]

We all know which experiments need to be performed in order to confirm that As substitutes DNA in GFAJ-1...simply those performed some decades ago with the normal DNA!...on the other hand... Just a few thoughts (not directly related to Elio´s proposal)... I wonder if there´s any evidence of As-containing sub-cellular fractions from a bacterium (e.g. E coli) growing under sub-letal concentrations of As...and It was strange for me not to see UFC/ml as a measurement of bacterial GROWTH (doesn´t acridine orange also interacts with DNA form death/metabolically inert cells?)...and we know that OD is not reliable for monitoring growth, especially when changes in cell shape ocurrs.

[Mark O. Martin]

I think that Dr. Badger is on the right track. I mean, I appreciate the commentary from Professor Silver and Dr. Sanders. Rosemary Redfield has much to say on this, of course. Data do not care about the "coolness" of a result; only journals and researchers and editors and reviewers do that. That data need to speak for themselves.

As Carl Sagan supposedly said, "Extraordinary claims require extraordinary evidence." I'm just a middlin' to low functioning microbial geneticist. I think that the physical biochemists or analytical chemists would have the best guide to answering this question. A revamped Meselson-Stahl experiment would be one way to begin?

I would think that Dianne Newman would have some thoughts on this, as well, since she has done much work with the biochemistry of arsenic.

Finally, back to the "I wish, I wish" aspect of this kind of research. The concept of a "Shadow Biosphere" that is not "life as we know it" is of course fascinating. I couldn't tell you where, but I seem to remember reading something by Joshua Lederberg a loooong time ago suggesting that a search for non-DNA based life should be carried out in the presence of blazing "hot" 32P.

Even as an undergraduate at the time, I wondered about the generalized damage of ionizing radiation to any kind of polymer. And that was before I learned about the overall doughty-ness of Deinococcus radiodurans!

So: I think that this controversy can serve two great functions. First, as a reminder of the peer review process (and its strengths and possible weaknesses). Second, an opportunity for those who doubt these results to present "this would convince me" kinds of suggestions to researchers.

[David Sanders]

Elio,

It's been 16 years since we met at Tufts. There is no need to look for arsenate in the DNA. The data in the paper indicate that the bacteria are growing on phosphate and sequestering the arsenate. Interesting resistance to arsenate--bad paper. Authors and reviewers should be sanctioned.

[Simon Silver, University of Illinois, Chicago]

Elio:
You invite comments.
Firstly, it is not our job or possible to come up with ways to disprove arsenate in DNA (and other bio-molecules). That is not how labs work and publish. One cannot prove that the world is not flat. The various blog reports on this already offer many such suggestions for experiments, but it is the job of the authors and not the readers to provide data supporting their conclusion.
The Science paper’s data seem reasonable and sound, if often lacking wanted controls and conditions.
The conclusion that arsenate “substitutes for” or “was incorporated into” (their repeated phrases), phosphate in the bacterial DNA and other biomolecules is undoubtedly wrong and there are no data in the report to support this conclusion.
Understanding that the basic conclusion of this report is wrong, several lessons are to be learned:
1) Each author is fully responsible and there is no way to "substitute for" such responsibility. The first author (FWS) and the last and senior author (RSO) have both repeatedly stated in interviews, press conferences, and Q&A sessions in Science (e.g. 24 December issue, vol. 330, 1734-1735) that there are no major problems with their results and conclusions, which they continue to believe.
2) Ron Oremland and John Stolz, senior microbiologists for whom this is the fourth report on arsenic microbiology in Science in a decade, should state widely and openly that “we screwed up.” There is no credible alternative for them; and to try to convert this into an argument pro and con, or a normal scientific process where new experiments test a hypothesis just doesn’t work, when they have already participated in blogs and news conferences and interviews.
3) The journal Science dropped the ball and either did not competently review this manuscript or went ahead and published after critical reviews. Science has a responsibility to its readers and continuing need for credibility to own up to its error in its job. Otherwise, it can not argue convincingly with creationists about the adequacy of data on evolution. One more tolerant view circulating on how Science screwed up is that the reviewers were geoscientists with little or no understanding of cellular microbiology and standard methods. Each of the ten scientists thanked in the acknowledgments for discussion and criticisms has a problem of dissociating from this sad and ugly situation.
4) NASA, which funded the effort, has a problem, as does the US Geological Survey, where the major flawed work took place. Maybe unfriendly congressional committee questioning is where this will be handled.

[Jonathan Badger]

Can't we do a version of the Meselson–Stahl experiment, only using As76 as the "heavy" label?