How do you prevent genetically modified organisms from jumping ship?

How do you prevent heavily reengineered bacteria from jumping the petri dish or fermentation tank to contaminate or overwhelm their unconfined natural relatives?

For two teams of researchers, the answer lies in significantly altering the genetic structure of those bacteria so that their biochemical machinery can't function without synthetic nutrients that only humans can supply.

Researchers are expanding their genetic engineering tool kit. They no longer are merely tweaking individual or small numbers of genes in an organism to introduce or emphasize desired traits. They are reengineering an organism's entire genetic structure. Those changes could give an organism a competitive advantage, explains George Church, a Harvard University geneticist who led one of the teams.

For instance, bacteria reengineered to withstand multiple viruses to avoid contaminating the compound the bacteria produces "could become an invasive species in the wild," he said – not necessarily bad for the environment, but "impactful."

"You want to get out ahead of these things rather than wait until you have a problem," he said during a briefing on the two sets of results, which are set to appear in Thursday's issue of the journal Nature.

The studies touch on a key to-do item set out last year in a report from the Woodrow Wilson International Center for Scholars' Synthetic Biology Project, which focused on research needs dealing with synthetic biology's environmental implications.

Synthetic biology is a burgeoning field in which scientists reengineer large portions of an organism's genetic instructions, encoded in its DNA, or design a complete set of instructions from scratch. With bacteria, the goal is to turn these single-cell organisms into diminutive chemical plants tailored to produce compounds for pharmaceuticals, produce biofuels, or help clean up oil or chemical spills, among other uses.

The Wilson Center report, based on workshops underwritten by the National Science Foundation, identified seven areas for research into synthetic biology's impact on the environment, including a focus on the type and extent of controls needed to keep modified organisms in check.

One line of defense against breakouts are the tanks and vats used to house heavily modified microbes. But these can leak, with microbes hitching rides off-site on workers' clothing, for instance.

Researchers have come up with other ways of controlling the spread of modified organisms, such as introducing biological kill switches. But these still offer organisms biological escape hatches, at least in principle. The organism can randomly mutate in ways that blunt control efforts. It can mooch a replacement for the synthetic compound it needs from natural analogues in the environment. Or it can incorporate DNA from other wild organisms into its own genome, allowing it to use nutrients found in the wild.

The two teams involved with the new studies applied complementary approaches to overhauling the genetic structure of a strain of E. coli bacteria that Dr. Church’s lab had altered to enhance virus resistance.

These were not merely genetically modified organisms; they had become genomically recoded organisms, or GROs. The changes to the bacteria's genome involved one out of every 64 segments along the entire length of the organism's DNA that corresponds to a specific amino acid needed to produce protein. DNA carries the coding for 20 types of amino acids.

In modifying the bacteria's DNA to thwart escape, the two teams altered the genetic code to include information for producing amino acids not found in nature but created in the lab.

One team focused on genes that coded for proteins crucial to cell functions, modifying them in ways that produced proteins that required the presence of the synthetic amino acid in the protein itself. The other team focused on 22 genes deemed essential to a bacterial cell's functions and tied the genes' expression to the presence of synthetic amino acids.

For the bacteria to survive, these synthetic amino acids had to be present in the medium on which the bacteria fed.

In both cases, the number of escapees was so small as to be undetectable.

At this point, it's unclear how quickly this could take off as a GRO containment strategy.

Church and Yale University colleague Farren Isaacs, who led the second study, note that they have received "nibbles" already from companies that use E. coli bacteria to produce chemicals. The two researchers have founded a company, enEvolv, to custom-engineer microbes for industry.

While in principle the approach could be applied to genetically altered crops, the researchers suggest that the initial applications probably would come in processes where the microbes are physically isolated as well. Plants have far more genes than microbes, making the task of introducing a synthetic leash more challenging.

Microbes already are designed to do their jobs and die off, notes Paul Winters, a spokesman for the Biotechnology Industry Organization, an industry trade group in Washington.

But "the method the papers describe is a more efficient, potentially more cost-effective method" for keeping genetically modified or genomically recoded organisms from running rampant.