Scientists and security agencies are thinking hard about the safety of this sort of gene editingby Philip Ball / December 9, 2015 / Leave a comment
When engineer Theodore Maiman announced the first laser in 1960, reporters were quick to disclose the perils. “LA man discovers science fiction death ray”, announced the Los Angeles Herald. Four years later, Arno Goldfinger was threatening to emasculate James Bond with the new “weapon”.
Don’t be surprised if some Bond villain is soon holding the world to ransom with a “gene drive” that will spread a lethal poison across the globe in the bite of a mosquito. This is the latest nightmare invoked by biotechnology, thanks to the new ability to edit DNA with pinpoint precision. Of course, just because a danger features in a Bond movie doesn’t make it fantasy— plenty of movie villains have threatened nuclear apocalypse. Both scientists and security agencies are thinking hard about the safety of this sort of gene editing, and it’s right that they should. But right now, the perils of gene drives come more from inadvertent than malicious release into the wild.
Gene drives are one of the possibilities raised by the technique known as CRISPR-Cas9, a method for highly selective and efficient gene editing devised in 2012. This exploits a natural DNA-snipping enzyme called Cas9 to target and edit particular genes. The target sequence of the DNA is recognised by a matching sequence on an RNA molecule carried with Cas9. This enables a modified form of the gene to be pasted into a genome in place of the existing gene. In principle, CRISPR-Cas9 offers a very powerful way to cure diseases caused by mutations of one or a few specific genes, such as cystic fibrosis. Clinical trials of genetic therapeutics are not far off.
Proposals to use CRISPR-Cas9 to address human disease at the embryonic stage—to “correct” mutations in embryos—are controversial, however, because they would for the first time entail making inheritable changes to the genome, with consequences for future generations that are not yet known. Preliminary experiments on human embryos (not destined for development into fetuses) show that there may be complications (see my Prospect blog on this subject), for example owing to the tendency of the editing machinery to introduce changes elsewhere in the genome.
A gene drive deploys CRISPR-Cas9 to spread a particular gene variant through a population, by ensuring almost total inheritance of that variation in a population of sexually reproducing organisms. Because sexual reproduction combines chromosomes from both parents in the offspring, an organism modified by gene editing that mates with a wild, unmodified organism would produce offspring with just one altered chromosome, diluting the effect and slowing the modified gene’s spread in a population. But suppose the modified organism is equipped with a CRISPR-Cas9 system that acts on the unaltered chromosome in offspring, giving that the modified gene too. Then inheritance of this gene would be almost 100 per cent, and it would spread rapidly.
This method has been proposed for combating malaria. Here, mosquitoes would be equipped in the laboratory with the gene-editing machinery needed to insert genes that confer resistance to the malaria parasite (Plasmodium falciparum), and would then be released to spread resistance in the wild. The principle was demonstrated earlier this year by researchers at the University of California at San Diego, who created such a gene drive in fruit flies. Now they have teamed up with another California-based group to apply the same idea to mosquitoes, using a gene drive to spread two malaria-resistance genes in a contained population. Some researchers think that practical methods to attack malaria and other insect-borne diseases such as Lyme disease might be available within just a few years.
If the idea of introducing a turbo-boosted method of genetic modification into the wild sounds alarming, it should. In 2014, before it was even clear whether gene drives would work in insects, a group of US researchers recommended some safety guidelines and called for regulation and extreme caution before unleashing such a powerful technique in a natural ecosystem. The subsequent publication of a gene-drive system in flies led the same researchers (including those who did that work) to recommend lab containment procedures.
That all sounds worrying enough. But what if a bioterrorist were to use this technique to develop lethal mosquitoes? They could, for example, use CRISPR-Cas9 to stitch into the insects’ genome the ability to make a toxin fatal to humans, which would be delivered with every bite. Then a gene drive could be added to the insects’ genome to send the mutation racing through the wild population.
Still more futuristic and scary is the notion of a gene drive that releases some fatal genetic mutation directly into human populations. And while it is easy to underestimate the expertise needed to make methods like this work, CRISPR-Cas9 is becoming an increasingly standard tool for genetic engineering. “You don’t need a very expensive piece of equipment and people don’t need to get many years of training to do this”, one researcher told Nature. In November the US National Academies of Science, Engineering and Medicine held a webinar meeting at which scientists and representatives of the FBI discussed the implications of gene drives for biosecurity.
Actually, the mosquito scenario is pretty insane at face value. Why make a bioweapon that could kill anyone indiscriminately wherever mosquitoes thrive? And what are the chances that you could make a weapon like that without becoming the first victim? Bond villains are not noted for their sanity, but in the real world it is unwise to confuse psychopathy with irrationality.
It’s also important not to overstate the novelty of the risks. It is not as though gene drives are the first biotechnological weapons. “We already have the Biological Weapons Conventions, which is imperfect but does ban bioweapons”, points out Drew Endy, an expert on synthetic biology (the systematic engineering of organism) and its hazards at Stanford University in California. There’s other US legislation to place restrictions on such activities below the state level, he says. Nor are gene drives the first system that can insert itself into and modify the human genome—in fact, our genomes are full of sequences put there by viruses.
Besides, Endy says, achieving such things in humans is still a long way off. “By the time somebody could, in theory, wreak havoc with the human population’s germ line—that’s several generations away—we should be readily capable of printing entire human genomes from scratch”, he says. In other words, those risks (at least) will probably be overtaken by other developments.
We’re perhaps too inclined to fear new imagined dangers while forgetting the devastation that can be wreaked with old methods. The alluring frisson of Bond-style high-tech super-weapons can distract from the ugly effectiveness of very crude means of causing death and destruction, as we are currently all too aware. “Low-tech is probably faster, scarier and more effective”, agrees George Church of Harvard Medical School, a leading authority on biotechnological risks posed by engineering the cell, and one of the authors of the 2014 paper on gene-drive regulation.
What’s more, Church adds, it’s not clear why anyone with access to such sophisticated biotechnology and the intention to do indiscriminate harm would not simply make and release a deadly virus, rather than relying on organisms that reproduce sexually. “The gene drives spread much more slowly than conventional pathogens “, Church explains, “because the generation time for viruses is much faster than for insects”—they replicate faster.
Some researchers have called for a moratorium on gene drives—or at least on state funding for such research—while the risks are assessed. But it’s not clear how meaningful such a gesture would be, given that bioterrorists are unlikely to apply for federal funding. Perhaps more sensible would be a moratorium on research that builds such potential into species that could escape into the wild. “A moratorium on certain species, like fruit flies which have wild populations near labs, would help scientists demonstrate responsible biocontainment”, says Church. The team that put gene drives into mosquitoes took the precaution of experimenting on a species native only to India; if a few rogue insects escaped in California, they’d have no one to mate with.
All the same, it seems wise to have plans for clearing up the mess if such inadvertent release occurred. In November Church and his coworkers described two variations on gene-drive technology that could minimise that risk. First, experimenting on yeast, they showed how to make a “split drive” that couldn’t be inherited in wild populations. It splits the genetic machinery used for CRISPR-Cas9 gene editing into two components: part is inserted into the host cell genome as usual, but the DNA encoding the Cas9 enzyme is carried in a strand that resides outside the genome, yet which can still be read to make the enzyme. So offspring of the mutant organisms mating with wild ones would only get half the machinery, which is useless.
Second, the Harvard team showed how to use CRISPR-Cas9 to make a “molecular eraser” that can undo a gene drive in more than 99 per cent of organisms that carry it. “We advocate testing ‘reversal drives’, in advance of need, in case anyone develops a gene drive with inadequate approval or foresight”, says Church. Such work indicates that simply ceasing research on gene drives would be to bury our heads in the sand. By far the best way to deal with potential dangers of new biotech, whatever their source, is to develop safeguards and antidotes.