The genetic code common to all life is not set in stone. We can change it at its most fundamental level for our own purposes. Genetic engineers have invented a new way to quickly, precisely and thoroughly rewrite the genome of living bacteria.

The technique could make drug-producing bacteria immune to viruses, prevent laboratory engineered organisms from genetically contaminating wildlife and enable scientists to construct proteins that do not exist in nature.

Farren Isaacs of Yale University led the team that this week proves it is possible to make numerous and very precise changes in the genome of living cells. "In the process we are recoding organisms that could have completely new functionality."

"I think it's a tour de force, one of the top 10 papers of the year," saysFrederick Blattner of the University of Wisconsin-Madison, who was not involved in the study. "Even though the genes are essential, they can be altered."

Ultimately, the study paves the way for re-coding the universal genetic code.

Stop sign swap

Isaacs and his colleagues systematically replaced one three-letter sequence in the genome of Escherichia coli with another.

Three-letter genetic sequences are known as codons, and they can either code for an amino acid – the building blocks of proteins – or act as stop signals. A cell's internal machinery reads copies of the genome, and as it goes along it links the corresponding amino acids together into protein strings until it reaches a stop codon.

The researchers sifted through the E. coli genome and identified all 314 TAG stop codons. They then designed fragments of single-stranded DNA that, with the assistance of viral enzymes, would replace the TAG stop codons with TAA, another stop codon.

Isaacs submerged a billion E. coli cells in pools of water brimming with the bits of DNA and viral enzymes, and zapped the mixture with electricity, opening pores in the bacteria's cell membranes for the DNA to pass through. The process is known as multiplex automated genome engineering, or MAGE.

Pizza party

Isaacs then isolated 32 strains from the mixture, each of which had around 10 TAA codons instead of TAG codons at different points in the genome. The next challenge was to combine these partially rewritten genomes into a single genome with 314 TAA stop codons and no TAG codons. How did they do it? Think promiscuous pizzas.

Imagine the E. coli genome as a pizza cut into 32 slices that together contain all 314 TAG codons, but not on one slice. Isaacs encouraged the bacteria to swap DNA by manipulating the bacterial equivalent of sex - a process called conjugation, in which two bacteria exchange fragments of DNA but do not produce any offspring.

By systematically pairing different bacteria, the team gradually moved ever closer to producing one strain with nothing but TAA stop codons. Isaacs calls this process conjugative assembly genome engineering, or CAGE.

Isaacs and his colleagues have so far winnowed the 32 strains down to four, in which nearly all TAG codons are replaced by TAA codons. The emergence of the final strain is imminent. Theoretically, CAGE should take only about five weeks - far faster than any other technique to produce an artificial genome.

The required equipment is much cheaper than for other methods, too. Whereas Craig Venter's synthetic genome cost $40 million, an "evolution machine" to perform MAGE might cost only $90,000.

Rewriting the genome, wholesale

James Collins of Boston University says the paper represents an important advance for synthetic biology. "The team shows it is possible to introduce in a rational way large-scale changes in the genome of an organism. They have very cleverly swapped one punctuation mark with another punctuation mark, which opens up the possibility of rewriting the genome wholesale."

Even more remarkably, it opens the possibility of rewriting the genetic code.

Theoretically, once a particular codon – say, TAG - has been removed from a genome, the cell's protein-making machinery could be reprogrammed to assign TAG to an amino acid, instead of being a stop signal.

And if the amino acid were not one of the dominant 20 found in nature or a completely new, synthetic amino acid, then the cell could produce entirely novel proteins. "That's the vision," says Blattner, "completely refactoring the genome to where it is quite substantially different from any other life forms."

There are other advantages. Genetically engineered organisms whose genomes are written in a brand new genetic code cannot mix promiscuously with other organisms if they escape into the wild. A new genetic code would also confer immunity on bacterial cells against viruses, which attack by incorporating themselves into the host cell's own DNA. That could help with applications like drug production: the bacteria we engineer to produce drugs like insulin are routinely attacked by viruses.

"The code is totally arbitrary," says Blattner. "We can manipulate the fundamental aspects of life."