Precambrian: the oldest cyanobacteria-like fossils known are nearly 3.5 billion years old

New research casts doubt on the theory that genes contain 'signatures' that can provide an insight into how the genetic code itself evolved.

According to the coevolution theory, first mooted over 25 years ago, the DNA that makes up the genetic code in any organism contains a signature of its own development. In other words, genes hold a 'historical' record of the origins of life.

Stephen Freeland and co-workers at Princeton University have revisited the theory and say that this signature is illegible. There is more than a 1 in 2 chance that any record of how the genetic code developed has been scrambled beyond meaningful interpretation, the team reports in the Proceedings of the National Academy of Sciences USA1.

The genetic code translates one molecular language into another. Proteins, the workhorse molecules of the cell, are chains of interlinked small molecules -- 'amino acids' -- of which 20 types appear in nature. The sequence of amino acids along the protein chain is recorded in DNA by a string of different molecules, 'nucleotides', along the double helix.

There are only four kinds of nucleotide, denoted A, C, G and T. They are read in groups of three by the molecular machinery that makes proteins. This gives enough different permutations to encode all 20 amino acids.

Each group of three nucleotides -- AAC, say, or GCT -- is called a 'codon'. In almost every organism on Earth, a particular codon encodes the same amino acid, indicating that this universal genetic code sprung from a single evolutionary source.

Ccoevolution theory suggests that the genetic code was once simpler, because there would have been fewer types of amino acids on the early Earth. The idea is that as the earliest organisms got more complex, they made new amino acids from the ones that already existed.

Then, the theory says, a new amino acid would have usurped one of the codons that previously represented the amino acid from which the new one was made.

So by inspecting the present-day genetic code, we should be able to figure out the sequence of steps by which the earliest code expanded to encompass more amino acids. This amounts to a kind of evolutionary history of genes, extending the 'tree of life' back even before the earliest recognizable single-celled organisms.

But the theory would fall down if the relationships between codons and amino acids could have occurred by chance. For example, the amino-acid leucine (for which one of several codons is CTT) is said to have been made from valine (codon GTT). The single nucleotide difference between the two codons points to this evolutionary relationship, the theory claims.

But this depends on whether it is reasonable to suppose that valine was made from leucine rather than from some other amino acid. And Freeland's team argues that some of the precursor/product pairs assumed by coevolution theory don't make biochemical sense. The theory also fails to take into account the shortcomings of the protein-making machinery that translates the code, the researchers say.

After correcting these misconceptions, they find that there is a 62 percent probability that the codon relationships between precursors and products claimed as support for the coevolution theory could arise by pure chance.