Tony Perry - What is CRISPR?

Kat - Let's take a closer look at exactly what's involved in CRISPR - or to give it its full name, CRISPR/Cas9. To find out more about this molecular toolkit and what it can do, I went along to a meeting organised by the Progress Educational Trust, discussing new genetic techniques and the impact they could have on human health and human genomes in the future. I caught up with one of the speakers, Tony Perry from the University of Bath, who told me more about the story of CRISPR, and some of its potential applications.

Tony - The prototype comes from bacteria that give you a sore throat, actually. It's sore throat bacteria Streptococcus pyogenes. It seems to be a kind of defence system and it turns out that even bugs get bugs. And so, these Streptococcus pyogenes, these sore throat bacteria themselves get invaded by viruses. The CRISPR/Cas9 was evolved as a way of cutting the viral DNA and thereby defending themselves, the Strep pyogenes, against these viral invaders.

Kat - So, what are the components of this self-defence system? How does it work and how have researchers then hacked this to do more precise editing?

Tony - The system really is like a pair of molecular scissors and a molecular satnav. The satnav is a small RNA molecule and it guides the molecular scissors to precisely where you want it to go in the genome. The reason it does that is because like a satnav, it's programmable. So you get to tell satnav what the coordinates are in the genome and it will then take the scissors to that point in the genome and the scissors will cut there. When they do so, when the scissors do so, they make a double-stranded break called the DSB and then the cell has got two types of tool kit that it can use to repair that double-stranded break. One type of repair that the cell can use or one tool kit is a simple kind of paste mechanism...

Kat - Just get the ends and glue them back together.

Tony - Absolutely, just get the ends and glue them back together again quickly. It does have to be a rapid process because these breaks are anathema to cells. If they're not repaired quickly, the cell can die and even worse, it can give rise to cancer and death of the organism in multicellular organisms. Well, that's very useful for research because you can use it to knockout genes if the cut is in a gene. It's probably not going to be so useful for clinical applications anyway. The second toolkit that the cell has or mammalian cell has to repair a double-stranded break is called homology-directed repair. We can actually add in our own designer piece of DNA which is used for repair. And so, we can design that, produce it, fabricate it in vitro in the laboratory and make sure by sequencing that is exactly what we want. And then we can introduce it into the cell of the same time as we've introduced the molecular satnav and the molecular scissors, the CRISPR/Cas9. And so, the cell in the right circumstances will use this extra piece of DNA that we've added as a repair template. What you end up with is in effect this sequence that you've added precisely inserted in the targeted position of the genome so that it's been used to repair the genome and you've introduced change that  corresponds to the DNA that you've added in.

Kat - So, we've got the CRISPR which is the RNA, the guiding system, we've got the Cas9, the scissors. And then this DNA template, and that could be anything I can imagine.

Tony - Can't quite be anything you can imagine as yet. Perhaps one day. One constraint at the moment is the size if you want to introduce a piece of DNA. So, you make the cut where you want the cut to be. So, you can introduce a large piece of DNA where you want it to be. But there probably is an upper limit at the moment today on how much novel DNA you could introduce. We're probably in the many thousands of base pairs or many thousands of characters in the genome and we don't know whether the sky is the limit or not. But what we do know is that as you increase the length of information that you want to introduce, the efficiency drops. And so probably, if you wanted to introduce a really large piece of DNA for whatever reason, you might have to do that in multiple steps.

Kat - So, we've got this technology which seems like an incredibly powerful tool and I know that scientists all over the world are using it in all kinds of ways in research in the lab, to understand how genes work and how genes get turned on and off, and all this kind of thing. But what we're talking about at this meeting here is editing the human genome. Tell me about the two different approaches that people are trying to think about - not doing necessarily at the moment - but the two sort of different ways that people are thinking about doing this at the moment.

Tony - As I see it, this whole topic can be sort perhaps rather artificially but sorted into whether the changes that you make to the genome are heritable - in other words, they can be passed on to successive generations - or are not heritable. So, the two main types of non-heritable genome editing changes that you might make work together with existing technologies that are variously well-established.

Kat - So this is effectively just normal cells of the body that are broken. You're repairing them in some way.

Tony - You can repair cells in situ using gene therapy and this would be coupled together very nicely with CRISPR/Cas9 that you here are doing the editing as it were of the patient's cells in situ. And there are one or two clinical trials that are already in progress to use this kind of in situ approach of somatic cell gene therapy.

Kat - What sort of diseases are people starting to look at?

Tony - Well, the ones that I am aware of are for example retinopathies. So, you can go in and you can edit genes that predispose to certain types of retinopathy.

Kat - That's problems with eyesight - the back of the eye.

Tony - Exactly, problems with eyesight and this is quite a good place to start because the eye is relatively accessible and a great deal is known about the eye. And so, this is something which lends itself to this application of CRISPR/Cas9. So, this is an example, perhaps the first of many, but it's an example of somatic cell gene therapy combined with CRISPR/Cas9. So, somatic cell gene therapy at the moment, there are probably over 600 clinical trials on-going in the US and it's a technology that's been around for some 25 years or more since the first successful somatic cell gene therapy. So, we might see this with the CRISPR/Cas9 which is very new, being combined with very well-established clinical technologies. 

The second non-heritable type of application is a cell therapy where you take cells from a patient who's affected with a particular genetic condition that predisposes to a disease. And you then edit those cells in vitro in the laboratory and then you can confirm that you've made the edit that you want to make. And you can put those cells in effect, you can put them then back into the patient because you've now fixed as it were the genetic change that predisposed to their condition. And that can happen either by taking cells and directly editing them and putting them back in - this is largely speculative although there are one or two on-going pieces of work. The applications for that are for example, where you can recover hematopoietic cells from the bone marrow of patients. And then you can potentially edit genes in these blood stem cells and then transplant this bone marrow back into the patient so that now the patient's bone marrow cells have got their repaired version of the gene, and the problem is fixed. 

Another way, perhaps a little bit more futuristic but is conceivable, is combining CRISPR/Cas9 with the whole technology of induced pluripotent stem cells. So induced pluripotent stem cells was something that was discovered by a method that was reported by Shinya Yamanaka. He won the Nobel Prize in 2012 for this. Basically, what you can do is you can take somatic cells like skin cells and if you do the right treatment, if you subject them to the right factors, you can cause them to start behaving like embryo-like cells. From those embryo-like cells, you can ask them to then respecialise into whatever cell you ask them to specialise into. So now, we have a system if we combine this iPS cell technology where we could, if we have a patient - perhaps a patient with a neurological disease that has a genetic predisposition - we could take skin cells from that patient because although skin is not the brain clearly, those cells in many cases would also carry the same mutation. It's just that there's no manifestation of the mutation there because they're not neurons. But we can take skin cells, generate from them iPS cells, embryo-like cells, we can go in with CRISPR/Cas9 - so combining the CRISPR/Cas9 here - fix the problem and then ask these embryo-like cells - these iPS cells - to differentiate to become neurons which we can do in many cases in many different cell types. And then we can transplant the neurons which are derived in effect - it's a bit around about but they are derived from the patient - back into the patient. But now, those neurons have got the fixed genome. So, that's another example of non-heritable because the neuron is not the germ line. So, that's another example of how CRISPR/Cas9 might be used in non-heritable therapy. Not CRISPR/Cas9 but this kind of iPS approach is being used, has been used in clinical trials for example in Japan.

Kat - Tony Perry from the University of Bath. And we'll be hearing more from him in next month's Naked Genetics podcast, when we take a look at whether CRISPR could be used to genetically engineer humans, and - getting to the heart of the burning questions - I ask whether my friend could hack her genome to grow a tail.