Accelerating Gene Editing Therapies for Rare, Neurological Conditions

In June, the National Institutes of Health’s National Institute of Neurological Disorders and Stroke made a five-year, $22.8 million grant to a group led by The Jackson Laboratory to develop gene-editing therapies for four rare, neurological conditions. The use of a platform approach to develop therapies for multiple indications follows other efforts on going at the National Institutes of Health in the area of gene therapies. We spoke to Steve Murray, associate professor at The Jackson Laboratory, about the promise of gene-editing, the work being done under the grant, and why the work could have broad implications for treating rare genetic neurological conditions.

Daniel Levine: Steve, thanks for joining us.

Steve Murray: Appreciate it. Appreciate the opportunity to talk about our project that just kicked off.

Daniel Levine: We’re going to talk about rare neurological diseases, gene editing, and a recent five year, $22.8 million grant from the National Institute of Neurological Disorders and Stroke to fund work that the Jackson Lab is leading to develop gene editing therapies for four neurologic conditions. Perhaps we can start with gene editing. What makes this a promising approach for treating genetic neurologic conditions?

Steve Murray: Yeah, so there’s numerous things that I think of why the community and why society in general is very excited about the potential of gene editing as a therapeutic approach. I think, at the most fundamental level, what makes it promising is the fact that it targets the underlying genetic cause of these rare diseases. So, a lot of our therapies currently are treating symptoms. Some of the more advanced gene therapies we have today can replace the missing gene. All of them, however, don’t address the underlying fundamental reason why those mutations cause disease. So, I think where gene editing comes in is it actually has the potential to change the genome in a way that fixes that underlying cause. The other advantages are, in some cases, and particularly the approach that I’ll tell you about more that we’re taking, is it has the promise to be more precise than a lot of the other approaches that have been used. Gene therapy as a strategy has been incredibly successful. One of the advantages of gene editing is that we can make the exact change that we desire in the genome. When you overexpress or you replace a gene product using gene therapy, there are some potential downsides of overexpressing that gene in a way that’s not necessarily identical to the physiological condition, normal physiological conditions. And then, I think, perhaps the biggest potential benefit is the nature of—that a change the genome is permanent. It’s a single shot and it’s durable. So, it has the promise to deliver all of these things that, in theory, a one-time treatment could be a permanent cure for a patient. And I think that’s the important promise of the approach.

Daniel Levine: This is a technology that’s rapidly developing. How good a therapeutic tool is it today?

Steve Murray: Right. So, it’s brand new as a therapeutic tool. I don’t believe there are any approved therapies at this point. You probably saw the recent news that CRISPR Therapeutics is moving its lead sickle cell treatment into the clinic, and that would be an ex vivo approach. Just to back up a little bit, the ex vivo approaches seem to be moving a little bit more rapidly. This is that we take cells from a person, modify those cells, and deliver them back to the person. And this is similar to approaches that have been used for, for example, CAR T therapies for cancer. So, this is obviously a place where there’s been a lot of interest because we have that technology in place and that platform in place. However, in vivo, direct somatic gene therapy or gene editing therapies, those are newer. There are a number of really promising clinical trials ongoing right now. So, for example, Intellia Therapeutics has made some really amazing progress in collaboration with Regeneron on transthyretin amyloidosis. They also have some really new promising clinical data on hereditary angioedema. So, the things are moving fast and a lot of these first generation therapeutics are based using the basic nuclease activity of gene editing platforms that basically knock out a gene, for instance, or down-regulate a target. But I think that’s just the tip of the iceberg in terms of the technology and how it could be used for therapeutics.

Daniel Levine: Well, what are the biggest challenges today to its therapeutic application and what obstacles need to be overcome, particularly if you’re targeting cells in the brain?

Steve Murray: Yeah, that, at least in my estimation, is the biggest challenge. I’ve been involved in a consortium for the last five years, known as the Somatic Cell Genome Editing Consortium. This is a Common Fund at NIH-supported program that seeks to advance gene editing as a potential therapeutic platform, but is really focused on providing support to build the fundamentals. That is the technology, advance the technology in a way that allows those who are going to translate it to do better. And really one of the major emphases in that consortium was delivery. So, getting the gene editing cargo to the cell of interest, doing so in a manner that is efficient and also at the same time is reasonably well targeted, is the major challenge thus far. It’s not a coincidence that some of the most promising therapeutic approaches thus far have been ones for which our delivery technology has matured. So, for example, ex vivo, we take the cells out, we know how to manipulate those cells, or in the case of Intellia’s lead products, they’re both targeting the liver. So, we’re very good at delivering lipid nanoparticles, for instance, to the liver, and that’s just the nature of the lipid nanoparticles. They migrate to the liver almost naturally. But moving from that to targeting other cells of interest is of great interest in the field, and I think is the biggest challenge right now. The other one is really understanding what we need to do in terms of ensuring both, and documenting and demonstrating safety and efficacy. I honestly believe gene editing is an incredibly safe approach, but it’s new. And so regulatory agencies are understandably quite cautious in this. And so generating the data that is necessary to prove to the FDA, for instance, and the rest of the world, that this is not science fiction. This is really a safe and effective approach, and getting some momentum in that area is something that we need to continue to work on.

Daniel Levine: And just to stay on the issue of delivery, is there any expectation whether this would be given through an infusion, or would you have to infuse it directly into the brain, or deliver it intrathecal?

Steve Murray: Yep. I think all of those are being explored. So, my general sense is that the technology for a standard sort of IV infusion to get editing cargo into the brain is still several steps away. Although there are some promising technologies, for example, using focused ultrasound as a way to open the blood-brain barrier.  That was one of our colleagues in the consortium that we worked with. There are some areas that people are working on in that space. If I were to speculate and project, I think intrathecal or direct intracranial injections, et cetera would be at least the first step in the process of delivering directly to the brain. That’s what we’re using in our project. And the other major challenge around delivery is exactly the mode of delivery. So, what we’ve proposed in our project, and maybe we’ll go into this in a little bit, but it is to use a viral vector delivery. And that’s because viral vectors have been proven safe and effective. There’s FDA approved viral vectors for, as I mentioned before, gene replacement therapies. However, it’s not the ideal mode for delivery of a gene editing cargo, although we do think it will be safe, because I mentioned that gene editing has this potential to be a one and done, a single shot therapy. We want, in a perfect world, the therapeutic delivery mechanism to be as transient as possible. So, viral vectors work so well for gene replacement in the brain because the AAV, the adeno-associated viral vector platform, actually continues to be expressed in the brain over the long term. So, ideally we could reduce the amount of time of exposure to the editor because we don’t need it. So, and, and we wouldn’t necessarily want to have it persist. However, we think the approach we’ve taken will be very safe in this respect, but this is part of the ongoing process of developing newer and better delivery mechanisms over time.

Daniel Levine: You’ve long used gene editing to create mass models of diseases. How different is gene editing as a research tool compared to being a therapeutic tool?

Steve Murray: Yeah, it’s interesting. The editing strategies, the approach we use to design editing changes in the genome, whether it’s for somatic editing, or for creating mouse models, fundamentally, at the very base level, they’re the same. We do a few things in mouse models that are somewhat different. Specifically, we use an approach called homology directed repair. So, we’ll use editing to generate a cut in the genome, and then we’ll use a donor template that is a piece of DNA that can elicit the specific change we want. In some cases, when we engineer mouse models, we can make significant changes. So, we can change it, swap out entire genes put an entire human gene into the mouse for example. And I think that technology, theoretically, can work for therapeutic editing. Many of the cells in the body aren’t constantly proliferating, which is what you need in order to do that kind of work. So, when we edit in a mouse embryo we can take advantage of certain DNA repair machinery processes that we don’t necessarily aren’t able to do in the cell. And then of course, there’s that delivery issue. We can deliver a lot of things to an early mouse embryo using a big microinjection needle. But to deliver the same sort of material to billions of cells in the body is a much greater challenge. The other aspect that I think is fundamentally different between therapeutic delivery and making mouse models is how we consider off target effects of genome editing. In a mouse, for example, we would engineer a mouse, and we’ll end up with a reasonably large number of animals that come out the other end that are edited in, probably edited in the way we wish. For those that are edited in a way we don’t want, we simply don’t proceed, and we don’t continue with that particular line. While in a human, every potential non-specific edit matters. So, our tolerance for off target or unexpected or undesirable outcomes from that editing for generating mouse models is much higher. So, we’re willing to accept that, and we just simply screen for what we care about in a human. We want to make sure that the edit is as precise as possible, and that’s not necessary in a mouse.

Daniel Levine: At the top of this discussion, I mentioned that grant from the National Institutes of Neurological Disorders and Stroke. The Jackson Lab is leading this effort to validate new gene editing-based therapeutic approaches for four neurologic conditions. How did this effort come about?

Steve Murray: Yeah it’s an interesting story. So, the lead principal investigator on this project is Cat Lutz. These four areas and neurological diseases in general have been an area of her research for a number of years. So, she’s led the development of a lot of these mouse models. She’s also led the repository here at Jackson that houses many of the mouse models, even if they weren’t built here at Jackson and they were built by other people, and then we distribute them. And so, she’s led a lot of the efforts to both characterize those models and use them in a preclinical setting. So, for example, our lead project, or our trailblazer project, I believe is what we call it, is for spinal muscular atrophy. And so, she was part of a lot of the preclinical testing for one of the approved therapeutics for spinal muscular atrophy right now. So, we had the models. We had the models; we had the expertise to characterize those models. What we needed to do is partner with some experts who understood the editing better than we did. So, we were incredibly fortunate to already have some ongoing collaborations with Dr. David Liu at the Broad Institute. And he happened to already be working on projects on all four of the disease areas that we chose to include in this grant. And we were fortunate enough to find a great partner in putting together this proposal. And then finally, we brought in an additional partner who is an expert in the delivery who Cat and myself have been working with for a long time. This is Dr. Steven Gray from UT Southwestern. So, his expertise and is in gene therapy and building those viral vectors that’ll allow us to deliver the gene editing cargo to the brain is his area of expertise. And also navigating some of the early approval processes for moving towards a clinical trial. He’s done that a number of times in ultra rare disease gene therapy applications. So, it’s really a great partnership with people we already knew, a place where we already had great models, a place where we already had editing proof of principle experiments underway, whether they’re both in vitro or in vivo, and so, sort of the planet’s aligned in a perfect way to come up with this proposal.

Daniel Levine: What are the conditions you’ll be looking at and other than being neurologic, are there shared features between them?

Steve Murray: Yep. So, the four areas are, as I mentioned, spinal muscular atrophy, and this is our trailblazer project. This is the one for which we’re most advanced, I think, in moving to the clinic. So, in this case, there was actually a really fabulous publication by David’s group, led by Mandana Arbab, using a base editor to treat spinal muscular atrophy. So, that’s our preliminary data, that’s the data that allows us to at least anticipate moving to an application for a new drug or for an investigational new drug in the five-year window. We’ve also added into this Huntington’s disease. So, this is a repeat expansion neurodegenerative disease; Friedreich’s ataxia, which is the most common genetic cause of ataxia; and finally, Rett syndrome, which is a neurodevelopmental disorder. So, we have these four. They’re at different stages in terms of the maturity of the proof of principle. All of them have at least in vitro proof of principle that we can modify the genome in a desirable way.

Daniel Levine: What makes a particular disease best treated through a gene editing approach rather than a different therapeutic modality and how does that question fit into the disease selection criteria for this project?

Steve Murray: Yeah, I think in many cases, the major feature that really supports a gene editing approach is that the disease is caused by a discrete number of specific mutations or would be alleviated by a relatively discrete number of types of edits. So, for example, there are ultra rare diseases or other rare diseases that are caused by a single gene mutation in a single gene, but there aren’t necessarily common mutations in that gene. There’s many mutations in that gene, and they’re all different. And that’s a little bit harder to attack for using a gene editing approach, at least the gene editing approach that we’re using, because each mutation would need a separate strategy associated with it. For these four, it’s the same mutation in all cases, or in most cases, in Rett. In Rett syndrome, we’ve selected a small number of more common mutations, but for SMA, Friedreich’s ataxia, and Huntington’s disease, there’s a single target that we can go after. And I think that’s one of the key criteria. The other, these are all they’re rare, but they’re not ultra rare. So, they’re common. So, the potential impact in terms of the number of patients is higher. And so, we’ve selected that. And then I think one of the most important considerations that we spoke to, given the goals of this program is to move at least our trailblazer project to an investigational new drug application by the end of the five year period, is the animal models had to exist and they needed to be well validated. And in all four cases, that’s the case. So, we have good models of disease for all of these, and in fact, really great targets to test our therapeutic approaches. So, while in other work that we’re doing at Jackson we’re building and improving models of rare disease that don’t currently exist, it would be difficult to do that while also developing the editing strategy while anticipating investigational new drugs. So, we needed to start with models that were well validated.

Daniel Levine: This is a five year project. What will the team be tasked with during that time?

Steve Murray: Yeah, as I mentioned, really taking proof of principle experiments that have already been done for all of these and developing a more extensive set of preclinical tests that would evaluate, you know, time of administration, routes of administration, different dosing schemes, et cetera. All of this would be done in vivo with the animal models to identify the best arrangement that could be then put forth for the investigational new drug application. And each of these will require a little bit different approach. We’d also be testing and confirming the potential for off target editing and what would be basically our safety. And then there’s toxicology, et cetera to really understand the pharmacodynamics, et cetera, and also the potential risk of off-target effects. And we would be assessing those carefully along the way.

Daniel Levine: You’ll be working with your Jackson lab colleague, Cat Lutz, you mentioned earlier, to develop, validate, and optimize in vivo mouse models for each disease. I know there’s some existing mouse models in these conditions. Is a mouse model a model for a disease, or are there things you’d specifically need to do to create a mouse model to test gene editing therapies?

Steve Murray: Yeah, it’s an interesting question. I think all in these cases, we really do have a good model. All models can be improved. There’s always limitations to the models that we have. One of the things that we’ve been doing more of, and we’re actually doing this for Rett syndrome in particular, is testing to confirm that or developing new versions of the mouse model where it’s the human sequence that’s been targeted to the mouse and not a similar version or the version in the mouse. So, the reason for this is that the gene editing approaches that we’re putting together will have very specific sequences that while the mouse is very similar to the human, may not be an exact match. So, the material, the editing material that we generate, we want it to be equivalent to the exact material we would use for a person. So, in that case, we have to change the sequence of the mouse at the very least, to match the human sequence, at least around the area that we’re editing. And for Rett, we have to develop that. For Huntington’s, for Friedreich’s ataxia, for SMA, we have that already. We have the model that is actually the human sequence in the mouse. So, we actually have already the ideal version in terms of the target sequence. In theory, all of these models could be improved in terms of the exact disease, the replication of the disease condition, although they all are, in our view, suitable for these purposes right now. But like anything, that can always be improved.

Daniel Levine: We use the term gene editing as if it’s a monolithic approach, but there are different technologies that can be used to edit genes. How broadly are you looking at technology in this effort?

Steve Murray: Yeah, we’re fairly targeted in what we’ve proposed to do. We’re using, and I think that’s a really important point to note that gene editing is a fairly broad space and [I] probably should have brought this at the beginning. But our focus in this project is to use both what we call base editing, and this is a technology developed by the Liu lab that allows specific modification of either an A base or a C base and changing it to either A to G or C to T. So, these specific changes don’t involve, and this is important, don’t involve a double stranded break in the DNA. And that modification, the advantages here, one, is that it’s very precise. We modify one single base. And secondly, the double stranded break approach has some potential caveats with it that we think the base editing approach helps to overcome. That being said, a standard CRISPR nuclease approach, as I mentioned before, there are multiple projects heading through the clinic right now, and they seem to be incredibly efficacious and safe. So, while we see base editing as providing greater precision, I think all of the approaches are incredibly promising in total. So, the other approach that we’re adding to this is a newer technology also from the Liu lab called prime editing. So, prime editing is exciting because it allows you to do more than just change that single base, or it allows you to make base changes that aren’t just C to T. For example, you could change a C to G or you could change a T to a G, for instance. So transversions are also possible with prime editing, although Liu Lab’s also developing base editors that can do some of this as well. But we think having this portfolio of potential editing approaches that don’t require double stranded break is a really great advantage. One of the things that I think is very clear is that the technology is advancing. It’s a five year grant. I anticipate that new discoveries will emerge during those five years that might be incredibly suitable for our project, and we could try to incorporate those, perhaps not the SMA because we intend to get this to an IND in the five year period. So, not a lot of time to pivot, but for the other projects, I think we have that opportunity to sort of reassess. The other thing I wanted to mention about this is the fact that what we’re trying to achieve can be different and is different. So, for example, with a nuclease approach, you can, for the example of the transthyretin amyloidosis treatment is just to remove that gene. And actually editing is incredibly efficient at that rather than changing the gene. And we’ve taken two, I think, pretty innovative approaches for Friedreich’s ataxia and Huntington’s that is probably unique. So, both of those diseases are due to an expansion of a repeat that’s in the genome. And so, patients who inherit that repeat, they have a larger repeat than people who don’t end up with the disease. And what causes the neurodegeneration is expansion of those repeats, both inherited but also somatically as well. So, a key therapeutic approach for those is to stop the expansions. And the way that we propose to do this is to actually install specific base changes into the repeats that prevent that repeat from continuing to expand. Other approaches might remove that entire repeat that require a double stranded break. There are potential complex rearrangements and outcomes that come with a double stranded break, and we’re able to avoid that using our base editing approach.

Daniel Levine: There are some other efforts at NIH, PaVe-GT and the Bespoke Gene Therapy consortium, which are looking across groups of diseases to leverage gene therapy to accelerate the development and reduce the cost of these potentially curative therapies. What’s the opportunity here to take the work you’re doing and apply it across other genetic neurologic diseases or rare genetic diseases more broadly?

Steve Murray: Well, I think that’s definitely the case. You know, this is funded as part of a broader consortium that includes other disease areas. And I think the expectation is by working within a consortium and sharing—although we haven’t had a kickoff meeting yet—I think we’ll soon get to know everybody who’s part of this program and that knowledge can be shared and that not only our group or other groups could take the experiences and the results that we achieve and consider how to apply them to other disease areas. I mentioned earlier that we’ve been collaborating with Dr. Liu’s lab in a number of other disease contexts already. So, this is something that we’ve already been considering for ultra rare diseases. they tend to be, at least the ones we’ve been working on, neurological in case, but there’s no reason why they have to be. For example, I’m definitely aware the Cystic Fibrosis Foundation is looking to help develop a treatment and a cure for the 10 percent remaining patients for which we don’t have an existing therapy. Most of those patients have mutations that are not addressable by a drug. And so, they’re eagerly looking to the community for research proposals and funding research proposals to apply base editing and prime editing and other gene editing strategies in order to target the remainder of those mutations. So, that’s just one space, but there’s many others. I mentioned sickle cell disease, for example, was one that CRISPR Therapeutics is working on now. There’s many other groups working on those on sickle cell and on other hemophilias, et cetera. So, there’s a lot of interest in a lot of disease areas.

Daniel Levine: Steve Murray, associate professor at the Jackson Laboratory. Steve, thanks so much for your time today.

Steve Murray: You’re welcome. Thank you. Thank you for having me.

This transcript has been edited for clarity and readability.