The Promise of Gene-Based Therapies for Neurodegenerative Conditions

February 26, 2021

The ability to target the underlying cause of a disease and make a lasting correction makes gene therapy an attractive approach to treating neurodegenerative conditions. The advent of Zolgensma, a gene therapy for the treatment of the rare neurodegenerative condition spinal muscular atrophy, serves as a model for this approach. A recent review article in Nature Neuroscience looks at the advances in the development of gene therapies for neurodegenerative diseases and considers their challenges and promises. We spoke to article coauthor, Subhojit Roy, professor in the departments of pathology and neuroscience at the University of California, San Diego, about the pace of activity in this area, why he believes it’s so promising, and its potential to extend beyond monogenic diseases. This episode is part of our ongoing Platforms of Hope series that explores advances in gene therapy and gene editing.


The ability to target the underlying cause of a disease and make a lasting correction makes gene therapy an attractive approach to treating neurodegenerative conditions. The advent of Zolgensma, a gene therapy for the treatment of the rare neurodegenerative condition spinal muscular atrophy, serves as a model for this approach. A recent review article in Nature Neuroscience looks at the advances in the development of gene therapies for neurodegenerative diseases and considers their challenges and promises. We spoke to article coauthor, Subhojit Roy, professor in the departments of pathology and neuroscience at the University of California, San Diego, about the pace of activity in this area, why he believes it’s so promising, and its potential to extend beyond monogenic diseases. This episode is part of our ongoing Platforms of Hope series that explores advances in gene therapy and gene editing.

Daniel Levine: Subhojit, thanks for joining us. We’re going to talk about gene-based therapies for neurodegenerative diseases and the recent review article you co-authored in Nature Neuroscience that looked at the promise and challenges of such interventions. I should not use the term gene therapy broadly to include all forms of genome manipulation. One of the first gene therapies to come to market is Zolgensma, a gene therapy for the rare neurodegenerative disease spinal muscular atrophy. Is it surprising to you to see one of the first gene therapies to emerge was a treatment for a neurodegenerative condition?

Subhojit Roy: It was a surprise for me because the field has been doing gene therapies for a long time. So, in the field of neurodegeneration and Alzheimer’s, gene therapy has been tried for a while and there have been no successes so far. I don’t know what the reason is and it’s different for every disease. For example, in Alzheimer’s we have been doing gene therapy trials at UCSD by injecting these kinds of growth factors into the Alzheimer’s brain and that hasn’t worked for various reasons. Then in Parkinson’s disease, we’ve had many gene therapy trials and all of those have failed. So, yes it was a surprise. It sort of came out of nowhere that this gene therapy, in these children with spinal muscular atrophy, was so incredibly successful.

Daniel Levine: What makes neurodegenerative diseases inviting targets for gene therapies?

Subhojit Roy: That’s a great question. One of the biggest selling points of gene therapy is essentially that it can target the etiology of the disease, which means that it can actually target the cause of the disease. If you think about it, that’s not that common. Most of the time we are targeting symptoms of the disease and not the actual cause. If you go after the gene, the advantage that you have is that you can actually focus on targets that are responsible for the disease to develop in the first place. I think that’s one of the biggest advantages of gene therapy: you focus on the actual target rather than targeting peripheral things. The other advantage of gene therapy is the permanency and the long-term correction. These are not new things. In the sixties, people said in review articles that these are the two biggest draws of gene therapies. For neurodegenerative diseases it’s definitely a huge plus to have a gene-based therapy.

Daniel Levine: As you note in your article, many challenges still remain. I’d like to walk through a few of those starting with delivery. How much of a challenge is it getting a gene therapy across the blood-brain barrier and targeting desired tissues and cells?

Subhojit Roy: If you talk to people about gene therapies, the most common thing you will hear is that it’s all about delivery and delivery is difficult or doesn’t happen. I beg to differ. I think that I’m more optimistic when I actually look at the data and I look at what has been done rather than just listening to people. I feel that there’s really good evidence in various model systems, including non-human primates, that we have the ability to deliver genes diffusely into the central nervous system using the current technology that we have. I get a lot of pushback when I say this, but I can show you data that has shown that culture. It’s nuanced and the field has really moved fast in the last few years. For a long time, the thing was that it was difficult to cross the blood-brain barrier using viruses, which are the clear front- runner in this. Recently, people have figured out ways to overcome that. For example, the traditional way that viruses were put into the central nervous system, to get them diffusely into the brain, was to actually put it into the CSF. If you put it in the CSF, the cerebral spinal fluid, the viruses along with their payloads, which is the gene therapy, eventually gets into the brain. The way they try to do this is by injecting these viruses into the CSF in the usual way, which is by a lumbar puncture, the lower part of your back, and that’s just a routine neurologic technique for doing that. But by that technique, very little of the virus got into the brain. So, it’s been all over the place that the viruses cannot get into the brain. Yet, in the last few years, people have used the same procedure but are injecting the virus much higher up. It’s near the neck and still goes into the cerebral spinal fluid, but it is much physically closer to the brain, and it’s called an intracisternal injection. With these intracisternal injections, people have already shown that in large non-human primates, and forget about mice, mice we can easily inject things in their bloodstream and get the viruses into the brain, but in non-human primates you can get this diffuse delivery of genes into the brain. Actually, right now in neurodegenerative diseases, there are two clinical trials that are doing just that. One in Alzheimer’s where they are injecting the gene APOE e2, which is known to be beneficial for Alzheimer’s disease, into the cerebral spinal fluid using the AAV viruses and intracisternal, near the neck, delivery. They have already shown in non-human primates, that if you do the same procedure, you’ll get a diffused distribution of this gene into the entire brain. Can it be better? Of course, it can be better. Still the technology in larger animals is nowhere near what we can do in small animals. So, there’s a lot of room for improvement there. I’m less pessimistic than a lot of other people out there, maybe because I entered the field late. To me, it seems like we can do it with the tools we have in hand right now.

Daniel Levine: Another concern is potential off-target effects of these therapies. How big a challenge is that, and what’s known about our ability to mitigate such problems?

Subhojit Roy: By off-target do you mean off-target of the viral injection, or off-target of the specific therapy that you use?

Daniel Levine: Actually, both.

Subhojit Roy: With the viruses the big advantage that we have is that the blood-brain barrier, while it makes it difficult for things to get into the brain, also makes it difficult or almost impossible for things to get outside the brain. The advantage with the viral injections is that, if you do put it in the cerebral spinal fluid, like these clinical trials are doing, then you really don’t get much of the viruses going outside the brain. So, you don’t really have to worry about it so much. The second point, with the off-target effects of CRISPR, that’s something that the whole field has to deal with, and a lot of things have been written and said about off-target effects. The truth is that the off-target effect and what happens when you deliver this gene therapy is going to be very specific for each gene therapy. Every CRISPR therapy that you have will have its own set of off-target effects. So, you can’t really say, there was this paper published where they showed that you have these random deletions and insertions because they used CRISPR, that that will have anything to do with the specific effect of what I’m doing.

Daniel Levine: There’s also the concern about triggering an immune response, particularly with the use of a viral vector. What do we know about that?

Subhojit Roy: That’s actually a huge concern in the whole field, which suffered a lot because of some of the historical things that happened in the field, and that was because of the immune response. Since then, a tremendous amount of work has been done to mitigate the immune response. We have thousands of patients that have been treated without much symptoms. Every time you treat them, you have to give them other medications to suppress the immune response because there is going to be an immune response for sure. Recently there’s also been some mishaps with immune responses. So, clearly the field has to develop better sets of viruses, and there’s lots of people working on it who know much more about this than I do. So, the bottom line is, should we worry about it? One hundred percent! We should work towards fixing the problem, which we are. Is it a game stopper? The answer is no. We have many, many patients that have already been treated with these viruses that we have in hand right now and they’re fine.

Daniel Levine: There’s a growing set of tools to manipulate the genome. I wanted to walk through the main categories that you wrote about and have you explain what they are and offer any insight into the relative strengths and weaknesses of the approaches. Let’s start with gene expression, the delivery of a gene to restore lost function.

Subhojit Roy: So, that’s the simplest. For example, in the trial that you mentioned, the SMA trial with Zolgensma, that’s exactly what they did. It’s very easy to restore the function of a gene that has been lost by introducing that gene. Technically it’s very simple because all you need is a strand of DNA. In terms of biology, it makes perfect sense. If you have a loss of something, then you just put more in. That’s the simplest form of gene therapies. I do feel that those are the kinds of gene therapies that are going to be most successful. I keep coming back to these two clinical trials that are ongoing in Alzheimer’s and Parkinson’s; in Alzheimer’s with the APOE e2 that I discussed, and in Parkinson’s there is one with a gene called GBA. In both of these cases, what they are doing is introducing a gene that produces a certain protein that they think will improve the pathology.

Daniel Levine: How about DNA editing? Within the realm of neurology, is there a particular approach that’s most advantageous for those conditions?

Subhojit Roy: I’m a big fan of DNA editing but there are risks. The main thing about DNA editing is that there’s no going back. Once I edit your DNA, we can’t go back and fix it. So, that’s the biggest risk when doing a clinical trial with DNA editors, but, like everything else in life, that’s also the biggest advantage. What we’re really talking about is a single shot therapy, what’s called one and done for the rest of your life. Now, there are other therapies like RNA-based therapies that are very popular. The problem with RNA therapies is that if you really think about how it’s going to work, it doesn’t seem practical. This is a therapy that’s going to work for two to three months and every two to three months you have to repeat the procedure. Now, if you are really saying that you have to do intrathecal, which is into the CSF, injections in a 65-year-old man, he’s going to come to the clinic every three months for that procedure. It’s a procedure that’s routinely done but it’s not like a pinprick. I see the advantage of RNA therapies in that it is transient, if something goes wrong you can stop it. To me, DNA editing therapies are a big advantage. I see that as not so different from a surgical procedure, where also you cannot go back. If you take out your colon, you can’t put it back. So, it’s pretty much the same thing. In never case, we have to weigh the risks and benefits and ask ourselves, is editing the DNA in this person and giving him the potential to live the rest of his life with the cure, is that worth the risk of having some abnormality that we can never fix?

Daniel Levine: I think anyone reading your article might be surprised to see the size of the list of gene therapies in studies right now for just a handful of conditions like Huntington’s disease, ALS, and SMA. Is there anything that surprises you in what you’re seeing? Anything that suggests one approach may be more promising than another?

Subhojit Roy: I was pretty surprised too. I’m very interested in this field and when we started putting all the trials together, it was surprising to me that there are so many trials that are ongoing. Is there any that seem better to me? So, I’m a little biased towards DNA editing for neurodegenerative diseases, when I think about the practical aspects of having to treat older patients. I feel like that’s the way to go. I see the advantage RNA therapies, especially the popular antisense oligonucleotide therapies, is they actually diffuse into the brain. Delivery is not an issue as long as you can put it into the CSF. That’s a big advantage there, which the DNA field has to match up. Do I see one as better than the other? I think that if you talk to people working in the field, people are excited about ASOs and RNA therapies and I can see why they’re excited. But to me, I think the future is in DNA editing because of the permanency and the logistics of how the DNA therapies would work.

Daniel Levine: Our audience, who is focused on rare disease, I imagine, thinks of these as tools for monogenic diseases. As you’ve mentioned, you’re involved in work using CRISPR CAS9 to develop potential treatments for Alzheimer’s disease. What do you think the potential is for gene therapies to address neurodegenerative diseases that are more complex than a monogenic condition?

Subhojit Roy: The first thing to remember is that no matter how complex the disease is, the therapies that we have, regardless of whether it’s gene therapy or antibodies or whatever else you are giving, is that you always focus on a single target anyway. It’s either going to be amyloid or it’s going to be tau for Alzheimer’s, for example, and for Parkinson’s is going to be one of the genes in the pathways for dopamine metabolism. It’s not that gene therapy is very different. The mindset that people have is that gene therapies have always been relevant for monogenic disorders, which they are. I think that monogenic diseases are essentially going to be completely curable once we have delivery and safety figured out. If you think about it, monogenic diseases have to be curable with this technology but with more complex diseases it really comes down to what is the most important target for [treating] the disease. Whether you’re talking about an immunotherapy or gene therapy it is the same question. What is the most important target that you’re going to focus on?

Daniel Levine: What’s the approach you’re taking in Alzheimer’s?

Subhojit Roy: In terms of what’s the appropriate target?

Daniel Levine: Yes.

Subhojit Roy: To me, you can’t get away from the amyloid pathway. I want to make it very clear, by the amyloid pathway I do not mean the amyloid beta aggregates. I’m agnostic about whether the aggregates themselves are the cause of the disease or not. Instead, I’m talking about the process by which these amyloid beta aggregates are formed, which is the metabolism of this protein called amyloid precursor protein and the cleavage that happens to that protein to create the amyloid beta. I think that you cannot get away from APP(Amyloid precursor protein) processing as a key event in Alzheimer’s disease for a variety of reasons. I’ll give you a couple of them. As you probably know, every familial form of Alzheimer’s disease has a defect in this amyloid pathway. If you put that aside and say, that the familiar forms make up only 5-10% of the cases. What about the other 90% of cases, the sporadic disease? There was this population in Iceland where people didn’t get Alzheimer’s disease and they randomly screened whole genomes in the population. The gene that came out that is a protective mutation for Alzheimer’s disease is in this APP, which is the first molecule that is cleaved to make the amyloid beta. That’s a mutation where others have shown exactly how it is protective. Then, in Down syndrome, for example, you have triplication of chromosome 21. You may know that pretty much all Down syndrome patients get Alzheimer’s disease if they live long enough. So the region that is triplicated, chromosome 21, has the APP gene in it. I and others who think about APP all the time just assumed that there was triplication of the APP. There’s been a lot of pushback in the field asking, could it be different because there’s a lot of things being triplicated? What happened was there was; there were three or four patients with Down syndrome that did not get Alzheimer’s disease. When people sequenced those patients, they found that the region containing the APP gene was not triplicated. It just happens that there’s slight variations in the regions that are triplicated in the Down syndrome patients. What is the chance that all of this is a coincidence? To cut a long story short, the APP processing and the process of forming beta is the most important thing in Alzheimer’s disease. People are focusing on immunotherapy and microglial therapies and I think that’s very important as well. Maybe that will be good for Alzheimer’s disease, but I can’t see that as the etiology of the disease, which is my main interest. I want to focus on the trigger of the disease and to me it is the APP gene. Remember, I didn’t say amyloid beta, it’s the APP gene that is the target.

Daniel Levine: What’s the path forward for that work?

Subhojit Roy: We are developing a gene therapy for the amyloid precursor protein. The one thing we have to realize is that APP is a highly conserved protein. It’s there in pretty much every animal. So, we can just get rid of it. That’s a very risky technique and there is actually an ongoing clinical trial that is just removing APP using ASOs. Here is the approach we are taking. My lab and other labs have discovered that the very end of the APP molecule has these five amino acids, which trigger the entire amyloid cascade. We’ve been working on the biology for a while and by sheer luck we discovered that if you use CRISPR CAS9 to cut out this small portion of APP, you can ameliorate the pathology of the disease. The advantage of this is that you keep about 90 percent of the molecule intact. So, the business end of APP, so to speak, is intact in our setting. We are only deleting the last few amino acids. You can think of it as having gangrene in your toe, and not killing the whole man, just cutting the toe off. That’s the concept.

Daniel Levine: Subhojit Roy, professor in the departments of pathology and neuroscience at the University of California, San Diego and coauthor of the recent review article of gene-based therapies for neurodegenerative disease in Nature Neuroscience. Subhojit, thanks so much for your time today.

Subhojit Roy: Thank you for having me. It was a pleasure.

Thanks to Pfizer, Inc., Bluebird, and Novartis Gene Therapies for their support of this podcast, part of our Platforms of Hope: Advances in Gene Therapy and Gene Editing series.

Stay Connected

Sign up for updates straight to your inbox.