there ain't no arsenic in them thar hills

January 24, 2012 8:11 PM

A strange bacterium found in California’s Mono Lake cannot replace the phosphorus in its DNA with arsenic, according to researchers who have been trying to reproduce the results of a controversial report published in Science in 2010. (Via Bad Astronomy.) Previously.
posted by IvoShandor (31 comments total)

We've done our part. This is a clean demonstration, and I see no point in spending any more time on this.

I knew it was too good to be true.
posted by Renoroc at 8:15 PM on January 24 [3 favorites]

Bah... and that was a fun one too - though almost immediatly after it was unveiled people were poking holes in it, so this is not a huge suprise.

In other news, some Russian whack-job think there's life on venus.
posted by Artw at 8:17 PM on January 24 [2 favorites]

Leads to interesting questions though--apparently the little buggers have an extremely high tolerance for arsenic, so how would increasing the arsenic over a period of 1000 generations while decreasing the phosphorus change the cell/DNA? Could they be persuaded to convert arsenic into something less toxic to other bacterium? Could they reproduce with Joakim and thus justify a FPP for him?
posted by BlueHorse at 8:27 PM on January 24 [1 favorite]

It's been a long time since I took plant physiology, but I recall that doing experiments with low phosphorus-tolerant plants was hellishly difficult, since it was so easy to contaminate the labware with 100X the normal phosphorus level that the organism needed to survive. Trace amounts of detergent on glassware or dust from laundered clothing were enough to keep some species ticking. I wouldn't be surprised if the original arsenic paper was done without an expert in low-phosphorus protocols chiming in to ensure quantitation of P was accurate.
posted by benzenedream at 8:35 PM on January 24 [6 favorites]

This is hardly surprising, if you followed the controversy around the initial paper. That said, I really admire Rosie Redfield for following this up so carefully, when it would have been easy to just assert that the initial paper couldn't possibly be true, and leave it at that. I also really like her blog and the fact that she's documenting her work as she goes. I hope that this doesn't cause problems getting it published and I would love to see other scientists with similar blogs.
posted by pombe at 8:37 PM on January 24 [2 favorites]

I like how they just slip in this small paragraph about the fight for open science:

Redfield and her collaborators hope to submit their work to Science by the end of the month. She says that if Science refuses to publish the work because it has been discussed on blogs, it will become an important test case for open science.
posted by spitefulcrow at 9:05 PM on January 24 [9 favorites]

I saw a seminar by Felisa Wolfe-Simon and was really rather impressed, I think she has some good evidence for arsenic playing a role in these bacteria. But probably not in the dna. After all, the dna of extremophiles may code for specialized proteins, but it's still made of the same four bases as all other dna, and as such, will have the same affinity for arsenic.
posted by 445supermag at 9:14 PM on January 24

If you can't deal with the next experiment telling you that you're full of shit, science is going to be a tough game for you to play. The sad part is the number of people who will equate "not being right" with "being deliberately fraudulent."
posted by Kid Charlemagne at 9:51 PM on January 24 [7 favorites]

This is ridiculous; science flavored pompous grandstanding.

Redfield et al haven't reproduced Wolfe-Simon's results, but they also haven't reproduced her experiments either. From the looks of what is available of their methodology they would be very unlikely to find arsenic in DNA even if it was there. We have known that nucleosides with arsenates on them are incredibly fucking unstable for a long time, which is why everyone doubted these results to begin with. They left the DNA in CsCl gradients long enough to spin them down but it looks like they assume that in the event of a break the arsenate would stay attached to one strand and have not done anything to demonstrate that it would not simply result in the logical two broken strands plus undetected inorganic arsenate.

Really though, the only data that they can logically say contributes to disproving Wolfe-Simon's hypothesis is the fact that they were unable to grow the critter at the levels of phosphate Wolfe_Simon says she was able to, which is admittedly significant but who can say it isn't some failing of theirs. Everything else that Redfield et al have done with their elevated levels of phosphate that are not relevant to Wolfe-Simon's results is just sour grapes.
posted by Blasdelb at 10:04 PM on January 24 [4 favorites]

After all, the dna of extremophiles may code for specialized proteins, but it's still made of the same four bases as all other dna, and as such, will have the same affinity for arsenic.

The claim is that the DNA backbone contains arsenate groups in addition to phosphate groups; in other words, the arsenic atoms are covalently bound to the deoxyribose groups from which the base groups extend. In this scenario, the bases would be the same--there's a different element in the backbone, however.

A major objection to this idea (and what should have been the absolute roadblock objection if they had the right reviewers reading the original manuscript) is that arsenate bonds are highly unstable in aqueous conditions. There's no way that backbone should stay intact. The original claim was extraordinary from not only a biological point-of-view, but from the point-of-view of basic chemical thermodynamics. It required more rigorous proof; at least a set of NMR spectra demonstrating stable arsenate bonds in DNA.

They left the DNA in CsCl gradients long enough to spin them down but it looks like they assume that in the event of a break the arsenate would stay attached to one strand and have not done anything to demonstrate that it would not simply result in the logical two broken strands plus undetected inorganic arsenate.

What do you make of their data showing no degradation of the As+ DNA after 2 months storage?
posted by mr_roboto at 10:09 PM on January 24 [2 favorites]

Science wins again!
posted by darkstar at 10:32 PM on January 24 [1 favorite]

"What do you make of their data showing no degradation of the As+ DNA after 2 months storage?"

That experiment would only be able to detect arsenate related cleavage if the arsenate were added in specific locations to make bands, which is ridiculous, or if the arsenate incorporation were pervasive enough in the presence of the levels of phosphate they are adding, which is also ridiculous. Also the two months in Tris EDTA buffer are immaterial to the stability of an arsenonucleoside in CsCl anyway.

"Science wins again!"

Sigh...
posted by Blasdelb at 10:41 PM on January 24

Also the two months in Tris EDTA buffer are would be immaterial to the stability of an arsenonucleoside in CsCl anyway.
posted by Blasdelb at 10:52 PM on January 24

That experiment would only be able to detect arsenate related cleavage if the arsenate were added in specific locations to make bands, which is ridiculous, or if the arsenate incorporation were pervasive enough in the presence of the levels of phosphate they are adding, which is also ridiculous.

I don't think anyone expects sharp bands; but with random degradation, you'd expect band broadening or the appearance of a new, diffuse band. No?

Also the two months in Tris EDTA buffer are immaterial to the stability of an arsenonucleoside in CsCl anyway.

Why? The reactive species is water, which is present in both environments.
posted by mr_roboto at 11:03 PM on January 24

Yeah, looking again, Redfield is looking at a 30 kb band, and Wolfe-Simon claimed As insertions every 5 kb. So if there were rapid degradation of arsenate bonds, the 30 kb band should go away.
posted by mr_roboto at 11:19 PM on January 24

Could they be persuaded to convert arsenic into something less toxic to other bacterium? Could they reproduce with Joakim and thus justify a FPP for him?

Not unless they were using nuclear reactors, as arsenic is an element.
posted by delmoi at 11:28 PM on January 24 [1 favorite]

"I don't think anyone expects sharp bands; but with random degradation, you'd expect band broadening or the appearance of a new, diffuse band. No?"

I'm glad you're not expecting sharp bands, but to show any difference on a gel from random degradation you would need a significant amount arsenate incorporation to generate a significant amount of cleavage, which no one would expect at the levels of phosphate the organism had access to for this experiment.

"Why? The reactive species is water, which is present in both environments."

You don't think 6.6 M salt could possibly have any effects on the stability of delicately hydrolysable bonds?

"Yeah, looking again, Redfield is looking at a 30 kb band, and Wolfe-Simon claimed As insertions every 5 kb. So if there were rapid degradation of arsenate bonds, the 30 kb band should go away."

Wolfe-Simon claimed As insertions every 5 kb under the much more extensively phosphate starved conditions she believed her organism was growing in, and that Redfield was never able to replicate, which is again the only meaningful finding here.

I don't know which is worse the fact that she thinks this performance art is worth re-publishing/self-plagiarizing as a Science paper or the fact that the editors will probably feel strong armed into giving her a shot at reviewers who, per standard policy, will be mostly from other fields and have roughly the same level of knowledge of inorganic archaeal microbiology as most of the researchers blogging about it.
posted by Blasdelb at 11:42 PM on January 24

You don't think 6.6 M salt could possibly have any effects on the stability of delicately hydrolysable bonds?

Not a strong effect, no. Why would it? What's the mechanism? I'd expect a pH effect, of course.

Would you be satisfied with a result showing no band broadening after an hour in CsCl?

Wolfe-Simon claimed As insertions every 5 kb under the much more extensively phosphate starved conditions she believed her organism was growing in...

She believed this, but failed to demonstrate it. Redfield showed the organism couldn't grow in these conditions, and demonstrated growth similar to that seen by Wolfe-Simon by adding phosphate at levels that could reasonably be expected to arise from contamination. We're getting into Occam's Razor territory here...
posted by mr_roboto at 11:54 PM on January 24 [1 favorite]

In fact, if anything, hydrolysis should be slower at a lower water concentration.
posted by mr_roboto at 12:00 AM on January 25

"Would you be satisfied with a result showing no band broadening after an hour in CsCl?"

I certainly would, so long as she used a control that was actually appropriate to the question at hand, like a naked alternatively produced 5kb strand incorporating Arsenate.

"She believed this, but failed to demonstrate it. Redfield showed the organism couldn't grow in these conditions, and demonstrated growth similar to that seen by Wolfe-Simon by adding phosphate at levels that could reasonably be expected to arise from contamination. We're getting into Occam's Razor territory here..."

Redfield failed to grow the organism in her best approximation of those conditions, which in these kinds of bizarre corners of microbiology is only slightly more conclusive as it sounds. What was interesting about the growth that Wolfe-Simon saw was that it looked an awful lot like a dose response to arsenate in extreme phosphate limiting conditions, the fact that Redfield could show a dose response to phosphate is both obvious and interesting only as performance art. I wouldn't claim that we can solidly say GFAJ-1 can incorporate arsenate into its DNA, though based on what I know of the much crazier shit that archaeal extremophiles can do with their DNA I wouldn't be surprised. However, I will tell you that nothing Redfield et al have done has brought us any closer to answering that question aside from failing to grow the organism under their best approximation of the communicated conditions.

When Ronald Oremland, mentioned in the link, gave a talk at my department he mentioned that there were other people quietly doing work on arsenate incorporation, I really glad that bullshit like this hasn't yet scared away their funding.
posted by Blasdelb at 12:36 AM on January 25

Next: Debunking life based on old lace.
posted by dhartung at 12:36 AM on January 25 [1 favorite]

Why? The reactive species is water, which is present in both environments.

Cesium salts are pretty interesting, because the very large cesium ions tend to be solvated quite well, which reduces ion-pairing and increases the nucleophilicity of the anion. While a bit of chloride isn't the sort of thing you'd normally expect to go busting up bonds, it's not inconceivable its presence could facilitate the breaking of already-weak bonds.

All that said, the burden of proof lies with the person who made an extraordinary claim, and a failure to replicate results is pretty damning no matter what experimental nitpicking you can do. What will be really interesting is when the third researcher takes this up.
posted by Dr.Enormous at 2:32 AM on January 25

This must be what reading math threads is like for non-math people...

(Hey, I got a 5 in AP Bio, that counts for a tiny bit of something, right?)
posted by kmz at 5:38 AM on January 25 [1 favorite]

Could they be persuaded to convert arsenic into something less toxic to other bacterium? Could they reproduce with Joakim and thus justify a FPP for him?
Not unless they were using nuclear reactors, as arsenic is an element.

Not true. Chelation involves locking up metal ions in inert compounds where they can no longer cause harm to organic molecules/processes. I don’t know that anyone’s observed organisms equipped with chelation enzymes for arsenic but that doesn’t mean it couldn’t happen.
posted by spitefulcrow at 5:41 AM on January 25

I love it when you guys talk dirty.
posted by Floydd at 6:22 AM on January 25 [1 favorite]

I was reading Felisa Wolfe-Simon's wiki entry, and it said that "Wolfe-Simon left USGS in May 2011 to pursue her research elsewhere. Wolfe-Simon claims she did not leave voluntarily, but was "effectively evicted" from the USGS group."

That led me to this very interesting article that I think is a worthwhile read. She's a fellow through 2013, but is looking for a new home for her research. Needless to say, she's had a very rough time professionally since Dec 2010.
posted by anniecat at 7:38 AM on January 25

Why would they store the As+ DNA in such high salt conditions, I thought high salt made the compound less likely to degrade? Why not store it in lower salt conditions (or no salt!), where one might expect As- DNA to be reasonably stable but As+ DNA to break down? Wouldn't that be a better demonstration of the stability or non-stability of the As+ DNA?

If I'm reading this correctly, the problem with Redfield's methods is they work on the assumption that there is a very high level of As+ incorporation in the DNA. Problem is she can't necessarily make this assumption because of the conditions under which she grew the bacteria in the first place. Blasdelb, that's your assertion, right?
posted by schroedinger at 9:25 AM on January 25

This brings to mind a story about one of the main people who proved the geological theory of plate tectonics. When asked what his next achievement would be, he is reported to have said, "I guess I'll prove it (plate tectonics) to be wrong."
posted by Danf at 9:40 AM on January 25 [1 favorite]

Why would they store the As+ DNA in such high salt conditions, I thought high salt made the compound less likely to degrade?

They're storing it in relatively low salt conditions (physiological-like Tris buffer). The high salt comes in the cesium gradient columns they use to distinguish between As-containing and As-free species.
posted by mr_roboto at 10:24 AM on January 25

"If I'm reading this correctly, the problem with Redfield's methods is they work on the assumption that there is a very high level of As+ incorporation in the DNA. Problem is she can't necessarily make this assumption because of the conditions under which she grew the bacteria in the first place. Blasdelb, that's your assertion, right?"

Largely, yes.

The argument has never been that this organism will happily add arsenate into DNA. Wolfe-Simon's original results showed that, under conditions that are so phosphate limiting nothing should grow, her organism would replace phosphate with arsenate. Redfeild grew it under conditions that were phosphate limiting, but not so phosphate limiting that the organism would be forced into using arsenate in order to grow. The most logical explanation for why she wasn't detecting arsenate is that it wasn't there, or at least wasn't there in significant enough amounts. That finding does not however address the original claim that GFAJ-1 will incorporate arsenate when phosphate is absolutely not available.

With labs dealing with these kinds of super-fastidious conditions, that is assuming its real, it is not really conclusive that one lab was able to grow an organism and one wasn't. Once you get to this level of microbial persnicketiness, there are so many tiny factors that can affect growth or not growth. Perhaps GFAJ-1 can only incorperate arsenate in the presence of the tiniest amount of copper and the media in Wolfe-Simon's lab was run through a copper faucet and some point, or something similar with manganese, or platinum. Perhaps Redfield's Lab using a copper faucet or something similar at some point could also selectively poison the putative arsenate incorporation pathway. To say that the whole thing is discredited based on one lab's failure to grow the organism at the relevant conditions is, if anything, much more irresponsible than Wolfe-Simeon publishing her not quite conclusive data with as much fanfare as she did. Particularly when there are other labs working on the putative system.
posted by Blasdelb at 10:07 AM on January 27

Science News:

Wolfe-Simon, who says she can’t comment in detail until Redfield’s results appear in a peer-reviewed journal, wrote in an email 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.

Debunking by Jonathan Eisen:

The whole thing was about their claim that the arsenic was ending up in the DNA.

In their abstract, for example:

Our data show evidence for arsenate in macromolecules that normally contain phosphate, most notably nucleic acids and proteins. Exchange of one of the major bio-elements may have profound evolutionary and geochemical importance.

In their conclusions:

We report the discovery of an unusual microbe, strain GFAJ-1, that exceptionally can vary the elemental composition of its basic biomolecules by substituting As for P. How As insinuates itself into the structure of biomolecules is unclear, and the mechanisms by which such molecules operate are unknown.

I personally am hoping beyond hope that Wolfe-Simon was misquoted in the new story, but I am guessing that that is unlikely.

posted by grouse at 8:36 AM on February 2