Bench philosophy: Gene Drives

Genes Taking the Fast Lane
by Steven Buckingham, Labtimes 04/2016



The idea of fooling Mendelian inheritance with gene drives is not new but has been hampered by inefficient gene drive constructs in the past. With CRISPR-Cas, the idea is acquiring new momentum, envisioning scary scenarios.

Changing the genome of a cell is routine in the laboratory today. Reading that ten years ago, would have made you sit up. But now we have moved up a grade. For the first time in human history, we now have the capability to alter not just the genome of a single organism but the genome of every single member of an entire species – out there in the big wide world, not just in the laboratory. This has been made possible by a technique called gene drive.

Gene drives are the logical next-step development of the CRISPR technique. ­CRISPR has been one of the kindest gifts that Nature has ever given the life scientist, making genome editing almost routine. But what the gene drive technology does is to take CRISPR and put it through a sort of chain reaction. It allows a CRISPR construct to hop across chromosomes and propagate itself down through the generations. And because of its ability to hop from one chromosome to another, it over-rides the stabilising, conservative effect of Mendelian inheritance. Almost inevitably, every member of the species will end up with the genomic change. Thinking through the possibilities is dizzying. Thinking through the dangers is nauseating. Whereas CRISPR makes DNA changes at the level of a single genome, gene drives make a change at the species. It is CRISPR gone nuclear.


Kevin Esvelt explaining the gene drive concept to students of the Academany program (bio.academany.org) that has been initiated by synthetic biology pioneer George Church. Photo: Rachel Soo Hoo Smith

A bold idea

Okay, I had better calm down and explain how gene drives work. The basic idea behind a gene drive was suggested by Austin Burt from the Imperial College, London, in the early 2000s. Burt envisaged introducing a sequence into a genome that encodes an endonuclease to cut a target site. If a chromosome containing the construct finds itself paired with a chromosome lacking it, the endonuclease would cut the sequence in that paired chromosome and copy itself into the cut site as the cell responds to the damage. In other words, it can copy itself into the matching chromosome. Natural inheritance would then, of course, pass the inserted sequence on to the next generation, so eventually the construct would make its way through the entire population.

Only almost-rans

But what sort of sequence could be used to accomplish this? At the time Burt made his original suggestion, there wasn’t a known construct that could do this efficiently. There were some almost-rans. They knew about elements called Homing Endonucleases – genes that wander around the genome until they finally end up at a specific preferred site. And of course there were transposons. But none of these was efficient enough to spread through an entire population.

But then along came CRISPR and that changed everything. When a technique starts appearing regularly in the popular newspapers, you know it is having a big impact. CRISPR is rightly achieving fame, even outside the laboratory, because of the enormous power it brings to alter the DNA of a target species quickly, cheaply and conveniently. CRISPR is adapted from a bacterial immune system, which recognises foreign DNA and cuts it up, so rendering it harmless. The CRISPR system as adapted in the laboratory consists of three elements: an RNA sequence (the CRISPR proper) with a guide-RNA (g-RNA), a nuclease (Cas9) and a primer to direct mismatch-based repair of the cut DNA. The CRISPR is an RNA-based targeting system that recognises a specific sequence in the DNA using the g-RNA. The Cas-9 nuclease then cuts at a site near the recognition sequence and then the cell’s own natural repair machinery kicks in. By adding primers that almost match the cut region, but with a chosen mutation, the experimenter exploits the imperfect matching performance of the homology-directed repair mechanism to introduce the desired change.

Easy and efficient CRISPR

There are several reasons why CRISPR is so important. It is very efficient – it is a direct genome-editing approach, so you don’t spend a long time selecting for transfected cells, for example. It is also very easy. The CRISPR part of the complex naturally takes care of the binding and activation of the nuclease, so all the experimenter has to do is to supply the appropriate guiding-RNA sequence. And by using a mixture of g-RNAs you can mutate several ­genes ­simultaneously.

Dream or nightmare?

So far, so good. But then in 2014, Kevin Esvelt, now at the MIT media lab, suggested that CRISPR could make Burt’s dream – or is it a nightmare? – come true. An element encoding a complete CRISPR construct is inserted into the genome. The element is expressed and it does its work on the target gene. But here is a crucial point: if the element also includes a homology-directed repair sequence, the element will also insert itself into the complementary chromosome. Effectively, we are using CRISPR to insert a CRISPR construct. CRISPR driving CRISPR. Chain reaction. And in one deft move, we have created a homozygotic mutant!

Now, think about what happens when the altered animal mates with a wild-type animal. When the modified chromosome pairs with a wild-type chromosome, the wild-type chromosome will also be transformed, because the CRISPR element in the transformed chromosome will “CRISPR itself” into the wild-type one. Even though one of the parents was homozygous wild-type, all the progeny will be homozygous mutant. The checks and balances offered by Mendelian assortment are completely bypassed.

Driven to extinction

Impressive but scary. It means that for the first time, we have found a way to drive an element through an entire population, raising the possibility of a host of interventions. Take malaria, for example. Two ways have been thought up for using gene drives to solve this major world health problem. One idea was to drive an element that introduces a mutation that makes females sterile. Unless the wild-type gene escapes in some way, the species will be driven to extinction within a few years. Another idea that has been put forward is to introduce an immune response in mosquitoes, making them unable to carry the plasmodium that causes malaria. As well as modify organisms that vector diseases, gene drives could also be used to eliminate organisms that directly cause a disease, such as schistosomes, or perhaps just alter them so that they can no longer infect people.

But having the power to change the genome of an entire species sets alarm bells ringing. Even something that looks good, like eliminating the malarial-bearing mosquito – are we really sure we want to do this? Let’s face it, we have form when it comes to eliminating species, but doing it deliberately? That’s another matter. For one thing, there could be unintended consequences. What if the “malarial” mosquito brings some benefit, of which we are unaware? Perhaps there is some flower that relies on it for pollination. Or what if eliminating a bad species leaves an ecological niche empty for something even worse to take its place?

On the other hand, three billion people live in areas where malaria is rampant and 600,000 will die from malaria this year. I suspect they wouldn’t be quite as timorous as us about the loss off malaria-bearing mosquitoes.

Ecological engineering

The potential applications of gene drives go beyond human health – there is the possibility of an emerging discipline of ecological engineering. For instance, the idea of eliminating exotic species by directing a gene drive against some unique sequence in their genomes has been mooted. If the exotic species is isolated geographically, there is no real risk of the drive hitting the organisms that are living where they are supposed to be.

For writers of bio-apocalypses, species-level genome engineering with gene drives is a gold-mine of ideas. Happily, a lot of these cannot ever happen in the first place. For one thing, the rate at which a gene can be driven through the population is dependent on the generation time. In the case of short-lived, rapidly reproducing species like mosquitoes, a gene can indeed work its way through the entire population in a few years. But this won’t happen in slowly-reproducing species like, well, ourselves.

Nature's resistance strategies

Could a terror organisation drive the spread of a lethal gene through our food supply? Producers of agricultural seed regularly screen for rogue genes and would be able to stop a gene drive spreading. Indeed, Nature has some of her own resistance strategies up her sleeve. For example, the gene targeted by a gene drive could randomly mutate at the site recognised by the ­CRISPR, which means it will escape the drive. Of course, we could respond by introducing another gene drive but at least Nature has something of a chance. We can also purposefully design safety checks into our gene drive. For one thing, we can balance its power by introducing a second, immunising gene drive. This immunising drive would pre-emptively alter the target sequence, rendering it immune to the original drive.

Another approach would be to use a “split gene drive”, where only some of the components of the CRISPR element are encoded in the genome, with the rest being supplied by the experimenter. And if anything does go wrong, and some overworked postdoc releases something from the lab they shouldn’t have done, as a last resort, we can always release a rescue drive which re-inserts the native sequence.

Urge for regulation

But of course, the simple fact is, we really have no idea at all what would happen should any gene drive stock enter the wild: the potential power conferred by gene drives would give nightmares to anyone who knows anything of the history of our interaction with the environment. So great is the concern that, for once, regulation is being urged before the technology is tried, instead of after our first major ecological disaster.

The USA’s National Academies of Sciences, Engineering and Medicine commissioned a committee to look into the potentials and dangers of gene drives and this year, their Committee on Gene Drive Research in Non-Human Organisms published a 214-page report entitled “Gene Drives on the Horizon: Advancing Science, Navigating Uncertainty, and Aligning Research with Public Values”. While noting that “if the current pace of change in general genetics is thrilling, the pace of change in gene drive research is breath-taking” and acknowledging the unprecedented power of gene drive, the committee recommended that existing regulations are sufficient and that the responsibility falls finally on the shoulders of the lab worker and their adherence to Good Laboratory Practice.

Somehow, I don’t feel so reassured.





Last Changed: 30.08.2016




Information 4


Information 5


Information 6


Information 7


Information 8