Arsenic life?

Scientists claim to have discovered a strain of bacteria that seems to be able to grow on arsenic, a well-known poison.

SCIENTISTS have always believed that phosphorus is essential to life. But a recent finding, which was reported in Science Express, the early online version of the reputed journal Science, has claimed evidence that would demolish this long-held view. A team of U.S. scientists, led by Felisa Wolfe-Simon of the NASA Astrobiology Institute (and which includes Paul Davies, well-known for his popular science books), has isolated a strain of bacteria called GFAJ-1 from Mono Lake located in eastern California that seems to be able to grow on arsenic a well-known poison by incorporating it in its DNA (deoxyribonucleic acid) and other important biomolecules instead of phosphorus. However, the evidence so far is not absolutely conclusive and some scientists have expressed reservations as well.

All living things on the earth depend on the following six elements: carbon (C), hydrogen (H), nitrogen (N), oxygen (O), sulphur (S) and phosphorus (P). There are trace quantities of other elements, which are necessary for some critical cellular functions, like non-protein components of an enzyme needed for its catalytic activity.

These trace elements sometimes are known to substitute for one another whenever there is a chemical similarity between the swapping elements. For example, as the authors point out, tungsten is known to substitute for molybdenum, and cadmium for zinc, in some enzymes, and copper as the oxygen-carrier instead of iron in some arthropods (insects) and molluscs (marine invertebrates such as squids, snails and slugs).

However, the six basic elements have never been found to be swapped even though arsenic (As), for example, is chemically similar to P. Arsenic, in fact, occurs just below phosphorus in the periodic table and is a chemical analogue' of phosphorus, which means that its structure and function are similar to that of phosphorus.

Arsenic and phosphorus have similar atomic radii and near-identical electronegativity, a chemical property that defines the tendency of atoms to attract electrons and become negatively charged ions. However, while P-based molecules are stable, As-based molecules are unstable. While P-compounds take a very long time to break down, As-based compounds undergo hydrolysis (breakdown of a molecule in the presence of water) within minutes or seconds.

The most common form in which P is present in biological systems is the phosphate group (PO4) containing one atom of P and four atoms of O. Phosphate forms the backbone of key biomolecules DNA, RNA (ribonucleic acid), and the energy carrying adenosine triphosphate (ATP), which is contained in all cells and phospholipids that form all cell membranes. To recall, for example, the spiral ladder of a DNA molecule has its rungs made of pairs of nucleotides which carry the genetic code of life. The sides of the ladder, DNA's backbone, are long chains of alternating sugar and phosphate molecules (Figure 1).

Now, biochemically, arsenate (AsO4) is known to behave exactly like phosphate. But it is precisely this physical and biochemical similarity of the two that is the cause of arsenic toxicity. It arises when biological systems use As instead of P in metabolic pathways that fail to distinguish between phosphate and arsenate. While arsenic may indeed get incorporated in some early stages of such biochemical pathways, processes downstream of these generally cannot tolerate As-based molecules because of their easy breakdown by hydrolysis.

As a result, downstream metabolic processes fail, causing toxic effects. But are there biological systems that have inherent mechanisms to counter this intrinsic instability of As-based compounds? The bacterium GFAJ-1 appears to be such a system that is able to thrive and reproduce in an arsenic medium.

Alternative biochemistry?

The obvious question is: has an alternative biochemistry been discovered? The claim of the researchers has been questioned by several scientists who remain sceptical. For instance, in her blog, Rosie Redfield, a Canadian microbiologist from the University of British Columbia, says: Basically, [the paper] doesn't present any convincing evidence that arsenic has been incorporated into [the] DNA (or any other biological molecule). She is, in fact, preparing to write a strong letter to the editor of Science.

Similarly, according to Steven Benner, an astrobiologist at the Foundation for Applied Molecular Evolution in Florida, that GFAJ-1 uses arsenic as a replacement for phosphorus is, in my opinion, not established by this work.

Mono Lake from where the arsenic-tolerant organism has been isolated is a very salty and alkaline waterbody with high dissolved arsenic concentrations. According to the data of the Los Angeles Department of Water and Power (LADWP), the water of Mono Lake contains 88 mg of phosphate and 17 mg of arsenic per litre. So, bacteria living in it would have already adapted to tolerate arsenic, but there is enough phosphorus in the lake for the usual phosphorous-based life. But Felisa Wolfe-Simon speculated further that some microbes might have adapted to using arsenic itself for their metabolism in such environments.

Despite criticism from other scientists, who argued that no cell would be able to cope with arsenate's instability in the presence of water, Felisa Wolfe-Simon and her team went ahead looking for such bacteria in Mono Lake. What the researchers did was to take sediment from the lake and put it into a similarly hypersaline and alkaline medium that also contained glucose (a carbon source), ammonium sulphate (nitrogen and sulphur source) and a mixture of vitamins and trace minerals necessary for bacterial growth. But this basic medium had no added phosphate or arsenate.

Significantly, however, the medium had a background contamination of phosphate and arsenate (at 0.3 mg and 0.04 mg/l levels respectively) and, as we will see, it is this phosphate contamination that is the source of controversy on the claim.

The organism was grown in this medium with increasing concentrations of arsenate (from about 14 mg/l to about 0.7 g/l). These arsenate-enriched media were first inoculated with the organism-containing sediment and then subjected to serial dilutions so that the concentration of any original phosphorus was reduced to the point where, in order to make DNA, the microbe would have to use arsenic to survive. It is in the inoculated medium with the highest arsenate concentration (about 0.7 g/l) that the interesting results were observed.

The culture was subjected to seven tenfold dilutions over many months. At the sixth stage of dilution (or a millionfold dilution), the researchers observed a gradual increase in turbidity, signalling the growth of bacteria. After one more tenfold dilution when the culture was introduced into a Petri dish containing the same medium, one colony of microbes was found to be growing.

This colony was then inoculated into the same medium with the highest arsenate concentration (of 0.7 g/l). The arsenate concentration was then gradually increased to about 5 g/l (much higher than the arsenate concentration in the lake, which is about 28 mg/l).

It was found that starved of phosphorus in an environment flooded with arsenic, the colony finally grew at the rate of about one doubling every two days with the arsenate concentration at 5 g/l. If the arsenate was replaced by phosphate (even at a concentration of about 0.15 g/l) it was found to grow faster. The growth in the culture in which neither phosphate nor arsenate was provided was, however, relatively insignificant and the researchers used this as the control for comparison.

By a technique called 16S ribosomal RNA sequencing, the researchers identified the growing bacterium as belonging to the genus Halomonas, of the gammaproteobacteria group, and has been christened GFAJ-1. Halophile means salt-loving and, according to a NASA information sheet, many of these bacteria are known to tolerate high levels of arsenic. But GFAJ-1 seems to be more than that it appears to be a distinct member of the species which, it seems, can thrive in an arsenic environment probably by incorporating As into biomolecules that are key to life.

Scanning electron pictures show the bacteria to be rod-shaped and when grown with arsenate and no added phosphate they are about one micrometre across and two micrometres long but bloated in comparison to bacteria grown with phosphate and no added arsenate. When deprived of P, bacteria grew slowly and they looked bloated they had 1.5-fold greater intracellular volume because the bacterial structure was found to have many vacuoles, or membrane-bound sac-like structures that serve to contain cellular waste products, which become numerous when the cells are under stress (as in this case).

Different techniques

Felisa Wolfe-Simon and her colleagues used different experimental techniques to establish conclusively whether GFAJ-1 was really incorporating As into its biomolecules or just cleverly managing to recycle the trace levels of P in the medium (arising from contamination). According to the researchers, data obtained by very sensitive mass-spectrometry methods known as ICP-MS and nanoSIMS which showed the distribution of various chemical elements within the organism's cells established that there was a clear difference between cells grown with As and those grown with P. However, as pointed out in her blog by Athena Andreadis, a professor of cell biology at the University of Massachusetts Medical School, the data reveal an interesting aspect of the apparent substitution of P by As in this extremophile.

Even in such an As-tolerant microorganism, there is extreme reluctance to incorporate As instead of P. According to the paper, under normal growth conditions the intracellular As:P ratio was about 1:500. When P was specifically excluded by increasing the As concentration to much higher levels than what is obtained in Mono Lake, the As: P ratio was only 7:1. Notwithstanding this, the authors have argued that the level of intracellular P in high As cells was consistently less than the quantity needed to support growth. It is this assertion that has been challenged by critics of the work.

By introducing radioactive As into the growth medium, the researchers have obtained more specific information about the intracellular distribution of As. They have found intracellular As in proteins, metabolites, lipids and nucleic acids. The fraction of As in the aqueous phase, which contains the nucleic acids, was only one-tenth while the remaining was incorporated in the organic phase containing proteins, lipids, and so on. This, Athena Andreadis points out, means that under extreme pressure the bacteria will harbour intracellular As but they will do their utmost to exclude it from the vital chains of the genetic material.

However, the paper argues that this fraction is consistent with the fraction of phosphorus that the organism incorporates in the aqueous phase (containing DNA/RNA) when cells are grown in P-rich medium.

To confirm that As was getting incorporated into the DNA, the team used gel-purified extraction, a well-known technique in molecular biology, to isolate and concentrate DNA from GFAJ-1 cells. This technique ensures that the extract is not contaminated by any other cellular material. Using high-resolution secondary ion mass spectrometry, or NanoSIMS, on this concentrated DNA extract showed that As was indeed present in their DNA. Our NanoSIMS analyses, combined with the evidence for intracellular As by ICP-MS and our radio-labelled AsO4 experiment, demonstrated that intracellular AsO4 was incorporated into key biomolecules, the authors state in their paper.

However, critics have problems with the methods used for these measurements. They point out that important control and purification steps seem to be missing and contamination of As from the gel matrix could have resulted in the data observed. Might they not have presented assays using properly purified (washed) DNA because these turned out to not have any arsenic? asks Rosie Redfield in her blog.

The crucial questions are: exactly where is As located, how much substitution has occurred in the DNA, if any, and how does it affect the layers of DNA function (folding-unfolding, replication, transcription and translation)? Definitive answers will require at minimum direct sequencing and/or crystallographic data, says Athena Andreadis in her blog.

Synchrotron X-ray studies

The authors have provided further evidence by characterising the intracellular chemical environment in which As is located using synchrotron X-ray studies. The technique, known as micro-extended X-ray absorption fine-structure spectroscopy (mu-XAFS), provides information about the structure of molecules by probing how its internal chemical bonds respond when stimulated by a beam of light. Within the DNA extracted from GFAJ-1 cells, arsenic was found to be bonded to oxygen and carbon in the same way as phosphorus binds to oxygen and carbon in the normal DNA. But cellular arsenic being in DNA can't be the explanation, writes Rosie Redfield, because their DNA analysis indicated that very little of the cellular arsenic purifies with the DNA. The cells contained 0.19 per cent arsenic, but the DNA only contained 27 ppb arsenic. Felisa Wolfe-Simon, however, says: I really have no idea what another explanation would be.

Barry Rosen, a biochemist at Florida International University at Miami, says that he has this lingering feeling that the arsenic is simply concentrated in the bacterial cell's extensive vacuoles and not incorporated into its biochemistry. He would like to see the authors demonstrate a functioning As-containing enzyme, for example. Benner, on the other hand, suggests that the trace contaminants in the growth medium used in the cultures are sufficient to supply the phosphorus needed for the cells' DNA.

According to him, it is more likely that the arsenic is being used elsewhere in the cells, in lipids, for example. Arsenates in lipids would be stable, he says. He feels that what appears in the gel-purified extraction to be arsenate DNA may actually be DNA containing the standard phosphate backbone but with arsenate associated with it in some unidentified way.

One of the key issues that the paper does not address squarely or answer convincingly is the instability of the As molecules as discussed earlier even as the authors give some plausibility arguments. If you replace all the phosphates by arsenates [in the backbone of the DNA], says Benner, every bond in that chain is going to hydrolyse with a half-life of the order of minutes, say 10 minutes.

So if there is an arsenate-equivalent of DNA in that bug, it has to be seriously stabilised [by some as-yet-unknown mechanism].

One of the arguments for stability that the authors give is: Although arsenate esters are predicted to be orders of magnitude less stable than phosphate esters, at least for simple molecules, GFAJ-1 can cope with this instability. They have also proposed that intracellular regions (like vacuoles) or mechanisms that exclude water may also promote stability. But these do not answer the question of arsenate DNA instability.

One way to cope with the instability could be to increase the turnover rate of the unstable molecule, says S. Shivaji of the Centre for Cellular and Molecular Biology (CCMB), Hyderabad, an expert in organisms in extreme environments who isolated three new species of arsenic-resistant bacteria in 2004 and 2005 from aquifers in Chakadah in West Bengal. The stability perhaps also depends on the existing mechanisms already known to be present in arsenic-resistant bacteria, he adds. Normally they achieve this by converting it into less toxic arsenate and at the molecular level this has been attributed to genes which are involved in arsenic resistance.

[But] I think the most important message from the finding is that life forms could exist, survive and divide in the absence of phosphate by substituting with arsenic. The evidence provided in the paper is convincing and also opens up umpteen questions relating to cellular processes such as cell division, cell growth, cell signalling, and so on, which are governed by the phosphorylation and dephosphorylation of proteins. We know the specific enzymes that catalyse phosphorylation and dephosphorylation. In the presence of arsenic, what enzymes would regulate arsenic transfer or removal from proteins, which are very crucial for the functioning of a cell, says Shivaji.

He adds: Since arsenic is known to be present in greater concentrations at the depths of the earth or near thermal vents in the sea, it probably hints at the possibility that early primitive life originated in such environs. Later living forms adapted as they evolved to forms preferring phosphate, which is more abundant.

The obvious question that the discovery of GFAJ-1 raises is whether there are other similar organisms. [Wolfe-Simon] would be very unlikely to have just found the only arsenic life form on the first try. So it's got to be the tip of a very large iceberg, says Davies, the popular science writer and one of Wolfe-Simon's co-authors.

And indeed Felisa Wolfe-Simon is apparently already growing 14 or so other isolates from Mono Lake in As-rich media. She has also looking at several of the previously identified Halomonas organisms, close relatives of GFAJ-1 on the genetic tree, all known to be As-tolerant to see if they too can survive in P-free environment. This is fertile ground for thirty years of future work, says Wolfe-Simon.

But before scientists embark on this odyssey to understand how extensive a role arsenic plays in terrestrial (evolutionary) biology in general, the lingering doubts about the present work's conclusions await clarifications through more detailed work on GFAJ-1.