Archaea use a different fundamental protein

[syzygy42]

Ara Pacis,

I've sometimes thought that in the distant future, the best evidence for the conditions of early life would be found in the ejecta from the Late Heavy Bombardment either on the Moon or Mars.

[Jens]

The second problem with the concept of a singular LUCA is the difference between the compositions of the membranes: Bacteria have cytoplasmic membranes composed of phospholipids in which a fatty acid with a straight unbranched lipid (or fatty) portion is linked to glycerol through a ester linkage. In contrast, Archaeal phospholipids differ in several important ways.

It's interesting that you mention this, because I happened to go to a lecture a couple of weeks ago where the speaker was talking about that issue. I'm not a specialist by any means, so I may wrong on the details, but I think he was saying that the archaea have lipids with a different chirality than those of both bacteria and eukaryotes, and yet it is believed evolutionarily that both the bacteria and eukaryotes branched off from archaea. So it almost sounds like the chirality changed happened twice. I think he was suggesting that maybe originally it was mixed, and that later they became unified and just happened to fall two one way and one the other.

[Nereid]

Awesome thread, awesome discovery!

These questions do raise the issue of "weird" life living a parallel existence here on Earth. I think Paul Davies was the first to put the idea forward in the literature. The argument rests upon the fact that most of our knowledge of the makeup of the microbial world is dependent upon PCR to amplify DNA sequences or upon shotgun cloning of isolated DNA.

Is this the "shadow biosphere"^WP, or closely related?

The code is made up of 64 triplets called codons that specify 20 amino acids and three stop codons.

I thought I read somewhere that there are 22 amino acids in use across the three domains (though one is found only in the archaea?), and that the codons are not universal, in the sense that a given triplet always codes for the same amino acid (or start or stop), beyond one pulling double duty. If so, how does the biochemical machinery "know" what to do, when a codon may mean more than one amino acid (or a start or a stop)?

[ShinAce]

A lecture I attended a few months ago was about 'traffic flow' in ribosomes making proteins. Even though there are multiple codons for a single amino acid, some are faster than others. If a protein requires a slow codon at the end of the chain, you'll often find a slow codon at the beginning near the start codon. This bottleneck at the beginning helps keep the traffic flowing smoothly down the rest of the chain and helps extra ribosomes from latching onto the start codon where they will sit idly for some time.

It was a shame that it was given to physics students who missed the point entirely. I did get lots of time to speak to the presenter as a result. Despite his work, he wanted to focus on the role of protein folding.

The lecture was titled "effeciency in protein synthesis" so everyone thought it would be about energy. And it was. Too many ribosomes stuck on a strand of rna is not efficient use of resources.

Edit: sorry nereid, I misread what you wrote. Make note that ribosomes are not 100% universal, thankfully. That's why we have antibiotics.

[syzygy42]

I'm not a specialist by any means, so I may wrong on the details, but I think he was saying that the archaea have lipids with a different chirality than those of both bacteria and eukaryotes, and yet it is believed evolutionarily that both the bacteria and eukaryotes branched off from archaea. So it almost sounds like the chirality changed happened twice. I think he was suggesting that maybe originally it was mixed, and that later they became unified and just happened to fall two one way and one the other.

You heard correctly: one of the main differences between the lipids is the glycerol phosphate used to attach the fatty acids (Bacteria) or fatty alcohols (Archaea). Bacteria (and Eukaryotes) use glycerol-3-phosphate and Archaea use glycerol-1-phosphate. These are stereoisomers with a chiral center making them mirror images of each other. There are two arguments against his mixing suggestion. The first is based upon the observation that fatty acid and membrane synthesis in modern Bacteria (e.g., E. coli have around 20 genes involved in this process). I would imagine that Archaea have a similar number of genes. Thus, it seems unlikely that a common ancestor to both domains carried duplicate membrane biosynthetic systems. The second, and this is just recall from a talk, is that the two different kinds of phospholipids cannot form a functional membrane. Take the second reason under advisement: I have yet to find a citation in which someone has actually done the experiments.

You raise a very good point. While our probes into the evolutionary history of genes have made stupendous advances in the past decade with a deluge of data, our knowledge of the development of cell structure and function is sparse. The most prevalent model for the development of the Eukaryotic cell is that it descended from a symbiotic merger of Bacterial and Archaeal cells, with the Bacteria donating most of the metabolic genes and the Archaea donating most of the informational genes (i.e., replication, transcription, translation). The devil is in the details: there are about a half dozen different models with each emphasizing a different aspect.

Originally Posted by Nereid

Is this the "shadow biosphere"WP, or closely related?

Yep. The idea has been tarnished by the arsenic fiasco. When I heard about the NASA news conference, I was excited and hoped an announcement of some such creature. I had been thinking about the shadow biosphere for a while, so when I watched the news conference and read the paper, I was very disappointed .

Originally Posted by Nereid

If so, how does the biochemical machinery "know" what to do, when a codon may mean more than one amino acid (or a start or a stop)?

In biology, there are always exceptions, and as you point out, the universal code is not quite so universal. For the most part, there are a few oddball organisms out there who have slightly rearranged their code for their genomic genes. Mitochondria and, I assume, chloroplasts have codes that have substantial differences from the "universal" one. This is not surprising since their translation system only needs to be good enough, thereby tolerating some changes. This is not to say that there is no selective pressure on the mitochondrial genes.

There is one exception that is found throughout the three domains: a stop codon is used to code for selenocysteine in a few genes. The mechanism of how this is accomplished is very different from the normal translation process. First, a serine-tRNA^Ser is modified to make selenocyteine-tRNA^Ser, replacing the oxygen with selenium. This Sel-tRNA^Ser is delivered to the ribosome using a specific protein that substitutes for the factor that delivers aa-tRNA. Additionally, a sequence in the mRNA is required to signal, "Insert the Sel-tRNA^Ser at this stop codon." About half a dozen different proteins participate in this process. It's so rare that this is generally ignored unless one is examining a gene in which this is known to happen.

Originally Posted by ShinAce

It was a shame that it was given to physics students who missed the point entirely. I did get lots of time to speak to the presenter as a result. Despite his work, he wanted to focus on the role of protein folding.

Twas a shame, but I can understand why he may have chosen to emphasize that aspect of his research in his talk. He might of thought that kinetic modelling of translation would be more interesting for the students -- at least more accessible if they had ever taken a biology course. Protein folding is a difficult subject with a lot of history, structural detail, thermodynamics and modelling. Nevertheless, control of translation elongation rate is important in protein folding. If you take a protein and denature it by heating or chemical treatment (i.e., turn a protein from folded to unfolded) and then try to get it to refold back into its original functional form, many never do. They fold into a non-functional form which is stuck at an energy minimum. For some of these proteins, people have demonstrated that the ribosome translates some specific codon, or codon pairs, slowly. This allows time for the nascent protein that is just emerging from the ribosome to make a critical fold. Translate those codons too quickly and the following protein segment can interact with the unfolded segment, preventing proper folding.

I am happy to see that there is renewed interest in translation. Once the basic mechanisms were worked out, many people left the field to get into the mechanisms of transcription control, which became easier and easier with the new DNA technologies. Several labs did continue to develop the system, and now it is apparent that there is a whole bunch of interesting biology going on.

A similar thing goes on in transcription in which RNA polymerase will initiate transcription, transcribe a short distance and pause from seconds to hours, preventing new initiation from taking place. For those genes where RNA polymerase pauses for an extended period, they can be turned on by a trans-acting factor binding to the paused complex, releasing the pause, and allowing transcription of the gene.