Choosing the Right Viral Vector for a Gene Therapy

Daniel Levine: Guangping and Phil, thanks for joining us. We’re going to talk about your recent review article in Nature that looks at viral vector platforms for gene therapy, the current state of vectors and their relative strengths and weaknesses. Perhaps we can start with some definitions and basics for listeners. You talk about four basic gene therapy approaches. What are they?

Phil Tai: They’re kind of loosely termed, but the four basic gene therapy approaches. One is, of course, gene addition where if you have a mutation in a critical gene that yields a disease, gene therapy aims to then give back function to it by introducing an exogenous gene. Then there’s gene replacement therapy, which is if you have a mutation and that gene is critical, but you can’t supplement it with another gene, the therapy acts to replace that mutated gene. The third one is a gene knockdown approach. So if there is a mutation in a gene that causes toxicity, for example, Huntington’s disease, which is characterized by a repeat sequence that causes a normal Huntington protein to then become toxic to the cells, the strategy is to knock down that gene’s expression to cure the disease. And then finally we have gene editing. Gene editing is something that I think holds a lot of promise for these types of treatments, because, the way that the most recent platforms for gene editing and I’m speaking about CRISPR CAS platforms, they can essentially be engineered in such a way that they can target any position within the human genome to correct any mutations. And so, you know, there’s certainly a lot of hope in terms of treating these rare diseases that may impact people’s health.

Daniel Levine: What is a vector and in the context of gene therapy, what’s its role?

Guangping Gao: Please, Danny and Phil, may I add an additional clarifications–gene replacement and gene addition. So for gene replacement it’s really adding a functional gene back when your cells have a mutated gene that is not functional. So you want to use a functional gene to replace the gene function of a mutated gene. And during gene addition, you know, it doesn’t matter, if your cells have any mutation or not. And you just, by adding a gene, either endogenous or exogenous, such as an antibody, or anti-cancer block, then by adding this gene you fight the disease, and it may not be a genetic disease. So I just want to make that clarification. Thank you.

Daniel Levine: Anyone who’s heard about gene therapies has heard the term vector. In the context of gene therapy, what are vectors and what function do they provide?

Phil Tai: Gene therapy vectors is at the heart of what we do here in the Horai Gene Therapy Center at UMass Medical School. So vectors, you can think of them as essentially the vehicles that deliver a genetic payload into cells. The gene therapy vectors that we are really intensely working on are based on the adeno-associated virus, a class of viruses. And these are very small, non-pathogenic viruses that do not replicate on their own. Based on this sort of safety profile, essentially what you can do is you can gut the viral genome of any endogenous or genes that are native to the virus so that they don’t replicate on its own. And then you can replace it with a genetic payload. Oftentimes it’s a therapeutic, a transgene. Then you can use this as a means of delivering a therapeutic gene product.

Daniel Levine: What makes for a good vector?

Phil Tai: It’s a very interesting question. Obviously a good gene therapy vector is one that can target an afflicted cell type that defines the disease, and does so efficiently. So there are many types of human diseases. And so, each one of these diseases may require unique treatments. And so when we really think about gene therapy, we like to tailor the vector to the disease. So there are certain diseases that may afflict highly replicative cell types or cell types that are non replicative, so you can choose viruses and then vectorize them to target these individual cell types. And then there’s also differences in tropism profiles that certain classes of vectors are very good at targeting. So you can tailor the gene therapy just based on optimal vector profiles.

Guangping Gao: May I add a few points here? First of all, when you ask a question, what is the vector? A vector is a vehicle to deliver a gene drug, and you can use either a non viral vehicle that we usually call a formulation to deliver a drug, just make a delivery–small molecule drugs. You could also use a biological vehicle. That is what we call the viral vector. So, so that’s the definition. They basically deliver your gene block into a human efficiently and safely. The second, it’s related to your other question: What do we think are good vectors? Personally, I believe there are four factors I would consider as good, gene therapy vectors. Number one, it’s efficient–be efficient to deliver genes into human body, to different cells, and that is a very important and most critical step. Secondly, it’s stable. So once you have the delivery, in many cases, you want to gene be there forever, and one shot works forever. That’s what we want. And so that’s the second feature. The third feature is we most likely want the gene to be stable in cells, but not integrated because that could potentially generate some integration, what we call genome toxicity. And so that’s a third. Of course another important feature of a good vector will be immunogenicity and immunotoxicity. So those are the features we, as gene therapy researchers, we prefer for a good viral vector. Thank you.

Daniel Levine: You had talked about the kind of disease specific selection process. One thing I don’t think you talked about was the size of the payload itself. How limiting can a vector be in terms of the size of the genetic material you’re using it to carry?

Guangping Gao: Yeah, so I will start, and Phil can add, this is a very challenging question, or issues we’re facing as gene therapy researchers. Basically what happened is that different viruses have a different payload, or transgene capacity, and ours probably can handle up to six kb, which is a 6,000 base pair transgene cassette. And then the virus probably can handle up to nine kb, but it depends on how you design it, where seven kb is most likely going to be optimal. But adeno-associated virus, while it’s very attractive in other features such as general toxicity, such as no immunogenicity and a low genotoxicity, it only can handle up to a 4.5 kb transgene cassette. So that’s certainly an indication for many genetic diseases.

Daniel Levine: Is there an example? I mean, I don’t think most people know what a four kb or six kb is like. If you think in terms of gene therapies that are either in development or diseases that are targets, is there an example of a gene that’s…

Guangping Gao: Sure, I can give you two examples. One example is Duchenne muscular dystrophy. That gene itself is huge, probably about eight or nine kb or so. So the DNA itself cannot fit. So now scientists trying to revise it where they modify and engineer it into either a micro dystrophin gene or many dystrophin genes. So we can keep that, uh, in the AAV capacity. And a second example is cystic fibrosis. That gene itself is close to 4.5 or five, so basically with very limited space for regulatory elements, such as promoters and other signals required to promote expression. And so those are the limitations of our gene therapy for those diseases. Phil, please, could you speak more on this one.

Phil Tai: Yeah. I think Duchenne muscular dystrophy is actually one of the best examples for folks trying to package oversized genes for gene therapy. So as Guangping said that the dystrophin gene is fairly large, right? I think it’s 14 kb, the cDNA, but actually what is interesting about those strategies is that very early on, most biologists had identified a specific mutation that causes Becker’s muscular dystrophy, where you get a truncation in a spectrum, like repeats of the gene. Now this gene is still functional, right? So, it still allows for muscles to function, for muscles of the diaphragm to help the patient breathe. Those individuals were slightly compromised in terms of quality of life, but it still allowed for these patients to live a pretty long life. And so because of the discovery of this type of truncated dystrophin gene, scientists were able to really take advantage of this aspect because this can actually be packaged into the small AAV packaging size. And so a lot of the Duchenne muscular dystrophy gene therapy strategies employ these type of strategies to generate these micro genes that you can package into AAV. And they work fairly well, at least in mouse models. And a lot of these platforms are now being tested in clinical trials with some success.

Daniel Levine: You touched on immunogenicity, our immune systems are designed to fight pathogens. What are the consequences of using viral vectors in gene therapies?

Phil Tai: Yeah, I mean, that’s an excellent question. So we really believe that the host immune response is one of the major barriers and challenges for gene therapy, as you said. Our immune system is developed to take on viruses, and some of the gene therapy capsids that are in clinical trials were either isolated from human tissues or related nonhuman primate tissues. And so the chances that any one of us may carry some response to these capsids is fairly high. So if you look at any population we may be 50 to 80% seropositive for AAV. There are several serotypes that are currently in use. And so we really envision a future where some part of this gene therapy may involve personalized medicine. So if we find out that a certain gene therapy platform is incompatible because of the immune response risk, then we can tailor the gene therapy based on different serotype usages for the vectors that are used. Right. It’s a major problem. There are many, many strategies to try to overcome both the adaptive and innate immune response. This is some research that we’ve been really focused hard on as have others in the field as well.

Daniel Levine: How much ability do we have to re-engineer a virus to make it, a more desirable vector?

Phil Tai: Again, that’s a very good question. AAV was discovered in the 1960s. But even now we’re learning new things about these viruses and the prevailing thought really is that in order to build a better vector, we have to really understand the viruses themselves. And not only that, we have to understand the host and virus interaction. Because again, the host immune system is one of the major barriers. So the more that we learn about these things through basic science pursuits, the more that we can better engineer these viruses as vectors.

Daniel Levine: Manufacturing of viral vectors is complex. It’s expensive. How should the question of manufacturing weigh in vector selection?

Guangping Gao: That is a critical, very critical question. Thank you. Manufacturing and the correlation between manufacturing and the disease target is critical. First of all, it depends on how you’re going to give the viral gene therapy to the human patient. If you do it systemically you use IV injection to deliver genes to your target tissues, such as the liver, muscle, and the CNS, as some vectors can cross the blood-brain barrier and transduce the CNS. That usually requires a high dose of vectors in order to generate this effect of a trans vascular gene delivery. However, for some local injections such as ocular gene therapy, when you deliver genes to eyes, or you deliver to ears for some hearing defect, then you do not need a high dose of viral vectors and the volume and the dose are limited. So in that case, the manufacturing burden will be dramatically reduced. And if you do systemic delivery, manufacturing burden is a major hurdle for getting a gene therapy commercialized.

Daniel Levine: You looked at three major viral vector platforms, adenoviruses, adeno-associated viruses, and lentiviruses. I thought you could walk us through each of those and talk about those from their advantages and disadvantages. What are adeno viruses and how do they rate as gene therapy vectors?

Guangping Gao: I will quickly walk through the advantages and disadvantages of those three vectors and Phil can supplement. So the adenovirus. That’s probably one of the most efficient viral vectors for gene delivery. It can transduce virtually any cell type, dividing cell, non-dividing cell, brain cell, liver cell, muscle. You can do a lot of things. That’s a major advantage. And the disadvantage there is this virus is probably the most immunogenic vector at both innate and adaptive immune response levels. So for this reason, and this is probably the most immunotoxic vector that can cause adverse effects for human gene delivery. The second vector is lentivirus, and its advantage is when you treat the cells with this virus, the virus intends to be integrated into host the genome and accomplish long-term gene transfer. This virus is primarily used for ex vivo or cell culture infection, for genetic modification in cell culture or ex vivo. And then you put modified cells back into the human to function as a living drug in the human cell, in human tissue, and the disadvantage of this virus is because it will require integration into the host genome. So even though scientists have developed all those non interbreeding or self inactivating virus systems, but if it gets integrated into an oncogene or proto-oncogene, it may cause some mutagenesis and so that is a concern. The other virus is AAV. The major limitation is really a packaging size. This is the smallest vector genome compared to the other two, but the major advantage for this virus are two: first it’s the least immunogenic compared to the other two viruses; and second, it can form episome stability, meaning that this virus does not need to integrate into the wholesale genome, but generates a circular episome format to stay in the cells for long term gene expression.

Daniel Levine: Phil, did you want to add anything?

Phil Tai: Yeah. We kind of touched upon some of this stuff when we were talking about the packaging limitations of the viruses. When choosing a gene therapy platform that also plays a major part in the design of the therapy and different aspects that Guangping had already touched upon, for example, adenoviruses. Those vectors are used in the Johnson and Johnson and AstraZeneca vaccines. So those you can sort of consider almost like gene therapy vaccines, right? And the advantage again is because they’re highly immunogenic, they can trigger a strong immune response, and you can get sort of a bystander effect for eliciting strong immune response to the COVID-19 SARS spike protein payload. So there’s certainly advantages again for the lentivirus platforms. So lentiviruses are RNA viruses and they need to integrate into the genome in order for them to remain stable within the transduced cells and because of this they’re very good at ex-vivo strategies. One of the best examples of these are the recent development of these CAR T cells that are used in cancers. And so what people do is they take these T cells out of the patients, and then they treat them with lentiviral vectors that engineer the cells so that they now have properties that recognize and kill cancer cells. They re-deliver these back, or re-introduce these back into patients to do their job. So all of these different platforms certainly have advantages and they need to be developed with the disease in mind.

Daniel Levine: You don’t cover non-viral vectors in their view, I’m wondering in general how you view them and what you see as their long-term potential. I think

Phil Tai: I think there’s a lot of potential. I think one thing that’s attractive about non-viral vectors, and some of these non-viral vectors are these nanoparticles, for example. They are also considered vectors. The advantage is that they’re not based on viruses. And so if you think about recognition by the immune system, they’re probably a little bit safer and less toxic to the cells, if they get to a point where they can be developed in means that are as effective as viruses. So one drawback of these lipid nanoparticles or non-viral vector platforms is that they don’t have all the advantages viruses have. The reason why viruses are very good is because they’ve evolved for millions of years alongside humans that make them very efficient at transducing or infecting cells. And so I think there’s still a lot to be developed in terms of non-viral vector platforms. But, just to say that one of the first platforms for gene therapy, they were non-viral in nature. So this began very, very early in the early 1970s where people were just trying to inject naked DNA into patients to see whether they would get an effect, but the problem is, of course, they just don’t transduce as well. They can’t enter into the cell as efficiently as viruses. I mean, you have a whole mechanism there that is available for viruses to get into the cells, to get into the nucleus, and to do their job.

Guangping Gao: I want to add two more points for the advantages of the non-viral system. The first one is that unlike viral vectors, where when you it give the first time you generate a neutralizing antibody that will prevent a second administration of the same virus, re-administration is one of the advantages for using a non-viral vector because you do not have this kind of response from a host. The second advantage is that in some applications, such as CRISPR-mediated genome editing, if you use a viral vector, such as AAV, you could establish CRISPR expression in the targeted cells for a long time. So you will keep having this gene editing process happening and repairing consistently. So in the long term, you may generate some off target effect. But if you use the non-viral vector, that’s a one time deal. Once they accomplish the genome editing process, the non viral vector will be gradually degraded and then the gene editing process ends there.

Daniel Levine: Guangping Gao is director of the Li Weibo Institute for Rare Diseases Research and director of the Horai Gene Therapy Center and Viral Vector Core, as well as a professor at the University of Massachusetts Medical School, and Phillip Tai is an assistant professor at the University of Massachusetts Medical School. Both are coauthors on the recent journal article on viral vector platforms within the gene therapy space. Guangping and Phil thanks so much for your time today.

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.