Published online 14 July 2011 | Nature | doi:10.1038/news.2011.419


Genomes edited to free up codons

Redundant sequences could be used to encode artificial amino acids.

DNAResearchers have removed all of one type of stop codon from a bacterial genome, and plan to reuse it for artificial amino acids.S. Kaulitzki/

Researchers have removed nearly all the snippets of a particular sequence from a living bacterial genome. This could enable the sequence to be reintroduced and used to encode artificial amino acids.

"For the first time, we're showing that you can make genome-wide codon changes," says Farren Isaacs, a bioengineer at Yale University in New Haven, Connecticut. "We're going to be able to start introducing completely new functionality into organisms."

The technique, published today in Science1, exploits redundancy in the genetic code. Amino acids, which make up proteins, are coded by three-letter combinations of DNA called codons, and multiple codons sometimes encode the same amino acid.

Cutting of the TAGs

Isaacs and his colleagues chose a 'stop' codon, TAG, which, along with TAA and TGA, signals the end of an amino-acid chain and the release of a newly made protein. Because all three codons do the same job, the researchers decided to erase all the TAGs from an Escherichia coli genome and replace them with TAAs. That would leave TAG free to encode a new amino acid.

Isaacs and his colleagues began by engineering 314 short artificial DNA strands to replace the instances of TAG in the E. coli genome. Each strand was identical to E. coli code, except that all the TAGs were replaced with TAAs. The researchers then used electric pulses to open up E. coli cells and overwrite the original genes with the new DNA. Several iterations of this technique2 resulted in 31 strains of the bacterium with 10 of the modified genes, and one with 4.

Then the team designed a hierarchical scheme to begin funnelling all 314 mutations into one cell. The researchers paired up the initial 32 strains, and one partner donated its mutated genes to the other. The resulting 16 strains were paired up to make 8 strains and then again to make 4, condensing the mutations along the way. The team dubbed the process conjugative assembly genome engineering (CAGE).

After two more rounds of CAGE, Isaacs says, a single strain of the bacterium will contain all 314 mutations and will be TAG-free. Then the researchers will snip out the code for the molecule that 'reads' the codon naturally and TAG will be free to code for new artificial amino acids. Some labs have already created such amino acids, as well as the coding machinery needed to incorporate them into proteins. But thus far those unnatural amino acids have had to compete for codons with natural amino acids.

"The big advance here is that we will have a host that will allow the incorporation of unnatural amino acids at much higher efficiencies," says Isaacs.

Genetic isolation

Such engineered organisms would be genetically isolated from other organisms and so immune to viruses that rely on traditional protein translation — important for keeping industrially useful strains healthy. And altered genetic information wouldn't be able to contaminate natural organisms, because outside the lab, the code would be gibberish, because natural amino acids in place of the artificial ones would create non-functioning proteins.

The team is eyeing 12 more redundant codons that it can co-opt in a similar way, says co-author George Church, a geneticist at Harvard Medical School in Boston, Massachusetts. "Now we can move quickly towards changing genomes," he says, with new genomic edits perhaps possible in weeks.


The approach is cheaper than trying to design whole genomes from scratch, says Church. By modifying existing organisms, most of the work is already done.

Scientists at the J. Craig Venter Institute (JCVI), which has labs in Rockville, Maryland, and San Diego, California, who last year 'created' the first bacterial cell controlled by a synthetic genome3, say the method is a positive addition to the field, but is only practical if the desired genome is similar to an existing organism. "Ultimately, we at JCVI would like to design cells from scratch, and only a whole-genome de novo synthesis approach would make this possible," said a spokeswoman for the institute in an e-mail.

The two approaches will probably one day be used in concert, says Isaacs. "It wouldn't surprise me if these technologies merged and we started to see hybrid approaches that are even more powerful than we're seeing today," he says. 


If you find something abusive or inappropriate or which does not otherwise comply with our Terms or Community Guidelines, please select the relevant 'Report this comment' link.

Comments on this thread are vetted after posting.

  • #62345

    I think the original human genome project sequenced many individuals in parallel, so the first sequence published was based on more than the sequence of just one person.

Commenting is now closed.