Genetic Kill-Switches Promise to Keep Bioengineered Bacteria in Check

Custom microbes have all kinds of potential, but we need a way to turn them off.

The future-world of widespread customizable, programmable bacteria is a bit closer this week with the release of an MIT-led study describing two new synthetic gene circuits capable of keeping such genetically-modified utility-bacteria from escaping and going rogue. The biological kill switches even have cool names: "deadman" and "passcode."

Genetically modifying bacteria is hardly a new idea—these were the first organisms to have been modified in the first place and have enjoyed a reasonably long history in medicine as tiny protein factories. But, as bioengineers get better at manipulating bacteria and genes, their microbial utility will only grow. More recently, bacteria have found use in applications ranging from mining to pollution cleanup to water quality monitoring.

"With the advent of synthetic biology, genetically modified microorganisms are being increasingly used for biomedical, industrial, and environmental applications," the study, published Tuesday in the journal Nature Chemical Biology, notes. "Deployment of these engineered microbes in large scales and open environments calls for the development of safe and secure means to restrain their proliferation."

As James Collins, the study's lead author, told Motherboard in an interview, the initial motivation was in limiting bioengineered bacteria employed in a medical diagnostics setting, where the bug is introduced into a patient's body in the hope of sniffing something bad out, e.g. checking on the state of an infection, sniffing cancer, etc.

"You'd like to be able to eliminate it from the patient should the patient have an adverse reaction to the microbes," he said. "Similarly, you'd like to have bacteria that can deprogram themselves once they leave the body, so they don't contaminate the sewage system or the broader environment."

"The need is real, I think, and is present," Collins continued. "There are a number of research groups and companies introducing microbes for a variety of real-world applications."

As bioengineered bacteria isn't a new idea, neither is containing it. Traditionally, this has been accomplished by producing bacteria that only grow and thrive in the presence of externally supplied metabolites or amino acids. But these aren't foolproof schemes and don't account for the possibility of the metabolites already being in the environment naturally, or being in the environment because you happen to already be feeding some other bacteria strain (resulting in "cross-feeding"). How do you keep your GM bacteria straight?

On top of that, schemes based on some externally supplied metabolite tend to require large-scale genetic editing—which may not always be practical—and a degree of programmatic specificity that may make the engineered bacteria impractical to use in a wide range of applications. There are other options, as it turns out.

One promising and more recent approach uses gene circuits to modify the bacteria's genetic expression and to control its production of toxins. The circuits then function as on/off switches or, better, self-destruct mechanisms.

"Upon loss of the biocontainment signal, the circuit blocks essential gene expression or induces toxin gene expression to kill the cell," the paper explains. "These circuits offer the promise of complex environmental signal integration but are hindered by a relative lack of programmable environment sensors to enable their use under non-laboratory conditions."

On its face, the Deadman killswitch is similar to historical methods of limiting bioengineered bacteria, but rather than the target bacteria requiring a certain metabolite to persist (needing to be fed, in a sense), Deadman is based on chemical signaling, a single on/off toggle. So long as a specific signal is provided to the bacteria, it can go about its business, but once removed, a "rapid and robust target cell killing" ensues. The circuit switches to its "dead state," in the words of the MIT researchers. (Note that this is a lot like the classic dead man's switch, in which the sudden absence of something makes another thing happen, like a train engine stopping or a nuclear weapon launching.)

Crucially, the signaling molecule employed by the Deadman mechanism, known as ATc, is not found in nature. So, should an engineered bacterium get loose, it could only self-destruct.

The Passcode mechanism is a bit different. Rather than a simple signal, the cell's survival is based on several combinations that can be programmed into a genetic circuit. The microbe is always checking for the presence of a particular two environmental conditions and, should they not be present, the cell will self-destruct. The neat thing is that these two inputs can be a wide variety of things, depending on the specific circumstances.

In addition to protecting ecosystem and agriculture from escaped bioengineered bacteria, the kill-switches may have a somewhat ominous bonus utility: IP protection. "In addition to its use as a biocontainment system, the Passcode circuit may find particular utility as a tool for intellectual property protection," the study notes, "where unauthorized growth of strains without the appropriate passcode molecules would induce cell death."

The promised land of bioengineered bacteria is a ways away, and, as Collins noted, Deadman and Passcode are mostly pre-emptive. Seems a reasonable precaution.