I remember back in around 2007, the head of the lab I was currently working in jetted in (literally; he had been on the other side of the world) in a flurry of excitement about a new technique he wanted to roll out comprehensively to engineer mutations in the lab’s model organism of choice, the zebrafish. The technique involved the use of zinc finger nucleases, and he thought it would revolutionise the whole tricky business of trying to introduce mutations into the genome (i.e. so they could be inherited by any offspring) of model organisms. He was half-right.
I also remember one of the postdocs in his lab being very annoyed because she had been doing a big screen using ENU mutagenesis – a hugely laborious process of mutagenising, screening and sequencing mutations, hoping you’d get something interesting – for a number of years, and now it looked like this was going to be abandoned. As for specifically mutating a gene you were interested in, this mostly involved something called homologous recombination: indeed, I myself used a type called BAC-mediated transgenesis simply to make a modified version of the gene of interest with a fluorescent tag attached to it. It was also a pain in the proverbial, if not on quite the same scale.
This is a something of a hazard in current molecular biological research: sometimes the field moves so fast that you can invest a lot of time, effort and money into something that may well be superceded by something better a few years down the line – and most projects take a few years to yield results.
Zinc finger nucleases (ZFNs) are part of a group of nucleases, often referred to as “molecular scissors” which are the hottest new techniques on the block as far as genome editing goes. They target unusually long sequences of DNA at the desired place in the genome and create specific double-strand breaks (DSBs), using this to insert, remove or replace DNA. The cell’s own natural repair mechanisms then fix the break (this type of break actually does happen naturally). There are currently four families of engineered nucleases being used: Zinc finger nucleases (ZFNs), Transcription Activator-Like Effector Nucleases (TALENs), the CRISPR/Cas system, and engineered meganuclease re-engineered homing endonucleases (catchy). I’m not going to go into technicalities here (see Wikipedia for an overview); they all work in similar ways, but I’d like to consider the potential implications of the one of the newer techniques, that of CRISPR.
CRISPR is coming to be seen by many as something of a game-changer as far as genome editing goes, but why? Simply put: because it is very easy, very fast and very cheap compared to almost any other method out there. ZFNs looked promising, as my old boss thought, but they’re expensive and actually quite difficult to engineer. Usually the sticking point with these things is getting it to effectively target the DNA you’re interested in (and nothing else). CRISPR works a little differently to ZFNs: it relies on an enzyme called Cas9 that uses a guide RNA molecule to home in on its target DNA, then performs the DNA editing. Most of the system is composed of generic components that can be bought off the shelf; often, you’ll only need to order the RNA guide, which isn’t going to break the bank. As of 2013, in fact, the number of papers citing CRISPR exceeded those citing TALENs and ZFNs, and took off almost exponentially. Similar leaps in both funding and patent applications occurred around the same time (data from here)
I already discussed a success for human therapy in this blog: engineering immune cells to target cancerous cells holds a lot of promise. This was done with TALENs, but CRISPR is catching up. Genome editing also holds the possibility of altering human embryos – with all the ethical implications that holds. I doubt many people would particularly complain if this were used to correct disease-causing genes – but the whole “designer baby” issue starts coming into play, since, after all, if you can alter defective genes, you can alter any other gene too. The ease and power of this form of GM modification could have big implications for agriculture and pathogens too. This is an obvious drawback to the attraction of CRISPR: it is, in a sense, almost too easy. Almost any fool can use it (well, I exaggerate, but not by much). Which usually means that somewhere, sometime, some fool will. And do something foolish. Researchers are charging in to use it and bioethics committees are running to catch up with the potential misuses of this technology, or, as is most often the case When Things Go Wrong, unforeseen consequences and careless mistakes. The other major concern, particularly when editing pathogens or human genomes, is that, however precise the method, you may well be cutting the genome in places other than your target site.
What is also not so widely discussed outside the scientific community is how much easier this is going to make the process of research itself. One thing that immediately struck me about CRISPR was that it could really be used to get over a problem that plagues researchers in my field, that of evolutionary developmental biology. This uses non-model organisms for comparative research into the evolutionary origins of embryological features. For example, you could study how the vertebrate brain become more complex by looking at fish, frogs, chickens and mice brains as they develop. These are all “model” organisms; ones that have been used for decades, and which generally have fully sequenced genomes and (with the exception of the chicken) a range of existing mutants and genetic tools available. The chicken itself is an obvious target for the new ease of genome editing. But other ones are the non-model organisms favoured by evolutionary developmental biologists: spiders and beetles compared to the fruit fly, lampreys as a very primitive form of vertebrate, sea urchins as a very simple animal, etc. Making any type of genetic alteration in these animals is difficult and time-consuming; CRISPR may change this almost overnight, with the potential to revolutionise an already vigorous if somewhat niche field. Indeed, the journal Development has a publication this week on the use of CRISPR in lamprey. I am really looking forward to the results that are going to come out of this: I predict a flood of work characterising genes in species that could not be practicably engineered before, and a whole raft of other non-model organisms being adopted.
Reference: Nature 522, 20-24