The Defense Advanced Research Projects Agency (DARPA) has released more detailed specs regarding its call for proposers to contribute to the Safe Genes project, whose goal is to develop tools to reverse the malicious or accidental negative effects of gene editing in the general population.

This would involve the development of systems which can apply a kind of ‘Digital Rights Management’ to a genome, rendering it resistant to a variety of genome editors, and also able to reverse the effects of gene editing on an already-affected genome, based on the distinctive signature of various editing tools.

The first phase of the project will last two years, and contributors will need to develop countermeasures to the effects of genome editors on mice or insect embryos. The proposers are being asked to consider various de-contaminant possibilities, and to offer their own suggestions, but the inhibitors developed should be ‘capable of inhibiting multiple classes of gene editors in multiple species’.

Possible delivery mechanisms suggested include ‘small molecules, antibodies, interfering RNAs’.

Additionally, the solutions must be specific to an individual subject, and not be capable of continuing to operate if transmitted to another subject, or at a species-level.

The second phase of inhibitor testing would involve live subjects and, eventually, testing at the level of a gene drive.

DARPA invites new approaches to the potential way that the inhibitor system would be delivered, but suggests that external application, oral delivery or delivery via nucleic acids are possibilities.

One interesting aspect is that the inhibitor system must be non-toxic, not interfere with the normal process of cell reproduction, and produce ‘minimal off-target effects’.

The project’s scope appears to be both remedial and preventative as if it might be possible to develop a type of immunisation against external genetic interference – at least from editing tools that leave identifiable signs behind.

What is unclear is whether it would be necessary for an inhibitor to run through all the lines of genetic code in a genome using a wild-card search for known editor signatures, or whether a heuristic-style search would seek out the kind of behaviour modifications common to all gene editing activity.

There also seems to be the possibility to compare a genetic checksum registered when the genome was unmodified to an altered checksum after editing has taken place, and use this as a less invasive basis for establishing that genes have been artificially modified.

Though not addressed in the latest proposal request, any remedial agent would also need to not interfere with any valid gene editing that has taken place as medical science begins to investigate the possibility for therapy via direct manipulation of RNA.

The agency’s broad research remit makes it prone to initiate speculative work that’s very exciting on paper, but sometimes science-fictional in scope, and the latest release certainly follows the trend.