DNA is nature’s Houdini, an escape artist par excellence. It moves not only from parent to child but also from species to species. Zebra fish and cows share a genetic sequence that likely hitchhiked its way to each on a tick; aphids captured the fungus gene that gives them their color; and even the human genome is littered with bacterial DNA. This remarkable slipperiness is among the many reasons—in this case, a legitimate one—that people worry about genetically modified organisms. The list of lab creations that have escaped into the wild, “Day of the Triffids”-style, is long. A 2010 study by scientists at the University of Arkansas, for instance, identified engineered herbicide-resistance genes in more than eighty per cent of the canola weeds growing along roads in North Dakota. Meanwhile, in Oregon, the genes from a trial plot of engineered golf-course turf somehow ended up inside a different species of native grass, thirteen miles down the road. In both cases, the usual physical containment methods—hedges as pollen barriers, rows of conventional crops as buffer strips—had proved an insufficient obstacle to roving DNA.
Today’s issue of the journal Nature suggests that biologists may have designed an effective way of confining engineered genes. The new technique, which is described in a pair of related papers by George Church, a molecular geneticist at Harvard, and Farren Isaacs, one of his former postdocs, who now runs his own laboratory at Yale, relies on nothing so straightforward as a wall. Rather, it forces an organism that was created in the lab to depend on the uninterrupted supply of an artificial nutrient. Take away its food, in other words, and the G.M.O. dies.
Every known organism on Earth, including humans, depends on a standard set of twenty naturally occurring amino acids. These are the component parts of proteins, which control or catalyze virtually all of a cell’s processes. But there are also thousands of “unnatural” amino acids, variations on the base organic compounds, some of which occur naturally but are not used in protein building, others that can be created only in the lab. In their papers, Church and Isaacs describe a process for genetically engineering a strain of the E. coli bacterium to live off of one or two of these synthetic amino acids rather than on the canonical twenty that are found in nature. As Church described it, “We’ve locked that dependency into a number of different essential points in the metabolism of the cell”—forty-two separate sites on the E. coli genome, to be precise—in order to insure that there is no way for the organism to work around its need for the artificial nutrient.
Once Church and Isaacs had designed their metabolic prison, they tested the ability of the bacterium and its genes to escape. Church’s lab removed a trillion of the engineered cells from their home environment, which was brimming with synthetic amino acids, and transferred them to petri dishes, which were coated with the standard nutrients used to culture bacteria. If any E. coli managed to thrive in the petri dishes—either by mutating so that they no longer required the synthetic compound to survive, or by finding a substitute for it in the standard nutrient blend—they were considered to have slipped the net, potentially able to survive outside the lab. None did. Church’s group then tried a variation on the same theme, testing whether the lab-created bacteria could absorb enough DNA from their wild cousins to regain the ability to live off of natural amino acids. Here again they failed. Either the genetically engineered E. coli couldn’t take in enough wild DNA to become viable, or they took in so much that their entire genome was effectively overwritten. If they had been carrying an engineered gene for herbicide resistance, for example, it would have been obliterated before it was passed on. Church was quick to point out that he and Isaacs will have to scale up their tests before they can be sure; as population size grows, so, too, do the odds that the bacteria will find a way to adapt.
Despite their technique’s effectiveness, both Church and Isaacs see any first industrial applications taking place firmly within the confines of a stainless-steel vat. The chemical company DuPont already uses a strain of E. coli that has been genetically engineered to turn corn syrup into trimethylene glycol, a solvent that is used in the production of plastics and antifreeze. The new method could prove a useful additional safeguard, reducing the chances of the DuPont E. coli slipping into the wild and teaching other organisms (or, no less significantly, the company’s competitors) their trick. It could also have applications in the dairy industry, which uses genetically engineered Lactobacillus acidophilus extensively in cheese starter cultures and yogurt. “These factories are not perfectly sealed containers,” Church pointed out. “I don’t want to be an alarmist or anything, but I think the point is that these organisms do spread.” The geneticists’ hope, though, is to eventually apply their findings under less controlled conditions—in designer probiotics that change how our guts absorb fats and sugars, in specialized microorganisms that are capable of cleaning up oil spills or landfills, and even, with a little more work, in a new generation of G.M.O. crops.
The same technique that allowed Church and Isaacs to edit the E. coli genome so heavily can also be used to confer any number of abilities on any number of organisms, from virus resistance to antimicrobial properties. Released into the wild, these organisms—Church calls them G.R.O.s, genomically recoded organisms, to distinguish their extensive reëngineering from the single-gene tweaking found in many G.M.O.s—would have unknown effects on an unprepared environment. Church emphasized his belief that it is “important to couple a safety mechanism with the productivity mechanism”—to code a potentially crippling dependency into a genome that also confers significant competitive advantages on new forms of life. If biotechnology is to deliver on its potential, Isaacs added—disease-resistant crops, personalized drug treatments, green biofuels—then the safety features built into these organisms must be equal to the task.
