What is CRISPR with Amy KOSKI
In this, the very first podcast, I chat with Amy Koski at the Oregon Health and Science University about CRISPR/Cas9 gene editing. Amy is the lab manager at the Center for Embryonic Cell and Gene Therapy at OHSU. The lab made headlines in August 2017 when they successfully repaired heart disease genes in human embryonic cells using CRISPR/Cas9.
Dustin Driver: Hello and welcome to Let’s Get Mental. I’m your host Dustin Driver. This is a podcast about science, technology, and society. We’ll discuss some of the biggest questions facing humanity today and ponder our shared future. In this, the very first podcast, I chat with Amy Koski at the Oregon Health and Science University about CRISPR/Cas9 gene editing. Amy is the lab manager at the Center for Embryonic Cell and Gene Therapy at OHSU. The lab made headlines in August 2017 when they successfully repaired heart disease genes in human embryonic cells using CRISPR/Cas9.
Dustin Driver: What is CRISPR/Cas9? CRISPR is an acronym that stands for clustered regularly interspaced short palindromic repeats. Basically, they’re short, repeated bits of DNA in a sequence, and they were first discovered in the early ‘80s by a scientist studying archaea, which are single-celled organisms that have been around for at least 2.7 billion years.
Dustin Driver: Archaea are super simple organisms, and they don’t have an immune system like we do. Instead, they take DNA snapshots of the viruses that infect them, kind of like those ‘do not sell to this person’ polaroids posted in convenience stores. So, if a virus tries to stroll into a bacterial cell, archaea can identify and attack it right away based on those snapshots, and those snapshots are called CRISPRs.
Dustin Driver: To take the snapshots, archaea use an enzyme called Cas9 to precisely slice the viral DNA. Researchers saw this and realized right away that Cas9 could be used to target and remove specific genes in a DNA sequence, and since then they’ve figured out how to use it to do just that, and how to make replacements for the genes they cut out.
Dustin Driver: At OHSU, they figured out how to target a gene that causes hypertrophic cardiomyopathy, which is a serious heart disease that affects people in early middle age. Because it doesn’t strike until later in life, many patients don’t know they carry the gene until they’ve already passed it on to the next generation, so the mutation persists in the population. The lab, led by Dr. Shoukhrat Mitalipov, used Cas9 to snip the gene out of a developing embryo, then the cells in that embryo repaired themselves using a healthy copy of the gene from the other parent.
Dustin Driver: In this interview, you’ll hear us talking about embryos, but we’re really only talking about eight cells in total, so it’s very early on in development. After that, the experiment is over. They haven’t gone any further than that. They’ve only managed to repair the gene, like I said, in about eight cells, which is still a tremendous achievement, and we’ll find out why in the interview. Without further ado, here’s my chat with Amy Koski at OHSU.
Amy Koski: My name’s Amy Koski, and I am the center manager and the clinical research coordinator for the Center for Embryonic Cell and Gene Therapy, working with Dr. Mitalipov. I’ve been here for roughly about four years, and we’re deep into figuring out how to cure and end suffering of inherited diseases.
Dustin Driver: Specifically, I think the first trial that hit the news back in August, that was successful, was treating myocardiopathy --
Amy Koski: Hypertrophic cardiomyopathy.
Dustin Driver: Hypertrophic cardiomyopathy. Which is a weakening of the heart muscles, if I understand it correctly?
Amy Koski: Yeah.
Dustin Driver: And it’s an inheritable disease. This study was able to correct the genes in vitro, or while a fetus was developing.
Amy Koski: We call it ex vivo. There’s no fetus actually developing at that time, but what we do is we correct it at the earliest stage of life, so when you’re a one-celled zygote, prior to fertilization, we’re able to inject a CRISPR construct that was designed specifically for our donor and that specific mutation, the MYBPC3 mutation for hypertrophic cardiomyopathy. When we co-inject the CRISPR enzyme along with the sperm, in the process called ICSI (intracytoplasmic sperm injection), we’re able to correct the mutation in those developing embryos, and we can see that about three to four days after fertilization when we do molecular analysis.
Dustin Driver: So, essentially what you’re doing with CRISPR is you’re able to target the defective gene, snip it out, and then there’s sort of a natural process in that the DNA repairs itself using a copy from the other parent that has a good copy of that gene, and it swaps it in there.
Amy Koski: Exactly. The coolest part about what we found, and it was very unexpected, actually, is that when we put the CRISPR in, the Cas9 acts as molecular scissors. The CRISPR you could think of like a magnet. It locates what we want, it sticks onto it, the Cas9 comes in and literally induces a break in the DNA by cutting, and that break is then recognized by the machinery of the embryo itself and comes in and goes through HDR, directed repair. And it uses the maternal copy of the DNA, which is free of the mutation, and reassembles.
Dustin Driver: Cool. So, in this case, the repair happens naturally. It sees a break in the DNA, and there’s a natural mechanism. From what I understand from other methods of using Cas9, you can actually program RNA that carries a completely different gene to go in and insert into what has been taken out.
Amy Koski: Exactly. In theory, when CRISPR/Cas9 was developed, you design your CRISPR/Cas structure, and that includes a single guided RNA, and you send that in, and that is the template that, after you make the cut, it’s supposed to recognize that template to correct itself. To mirror image. In our situation, it absolutely doesn’t happen. We tried to put in single-guide RNA. We did inject it. We know it’s there, but the embryo machinery chose to use the full copy of the maternal DNA over our synthetic guide RNA.
Dustin Driver: Wow, okay. That’s really interesting. Before we get too deep in the weeds, let’s take a step back, and let me explain how I view CRISPR working, and you can tell me whether or not I’m totally crazy, or if it’s correct or not.
Dustin Driver: From what I understand, CRISPR is actually a bacterial immune defense system. What a bacteria will do is it will take a DNA snapshot of an invading virus and use that snapshot, the next time that virus invades, to basically defend itself against being infected and killed by the virus.
Amy Koski: Correct.
Dustin Driver: Part of how it does that is basically analyzing. It’s snipping a piece of the DNA out very, very precisely. When you’re using CRISPR/Cas9, you’re able to reprogram that Cas9 pair of scissors to snip out the piece of DNA that you’re looking for, and then the cell can either repair itself or, as we were just discussing, you can engineer a new gene or grab a different gene to insert back into the sequence.
Amy Koski: Correct. Somatic cell therapy, which we’re reading about in the news everywhere -- in mice, or even in some human trials -- is trying to do exactly what you talked about. They’re locating something that has gone wrong, and then they’re trying to fix it with a new template and saying, this is how you should be acting. This virus that you have or this cancer that you have is inadequate, inappropriate -- do this instead. Whereas in the embryo, when we try to use this template, it does not want to do that. It preferentially selects the other, healthy copy of the DNA, and that’s actually great, because then we’re not altering the human genome, we are just providing the mechanism to correct itself.
Dustin Driver: That’s really interesting. Obviously, when the study was going forward, did you anticipate that it would self-correct, or were you thinking that you would have to have the correct gene to put in there?
Amy Koski: When we first began this study, it was very, very early in the whole CRISPR/Cas9 developing, and we definitely thought we needed the single guide RNA. It was part of our design, and we worked extensively on building that template. We didn’t want to change what was already existing in the human population. We just wanted to correct back to what it is. So, you make multiple versions of that template with different base pairs, and what will be accepted, what won’t.
Amy Koski: There was some literature, early, where they had seen this self-correction. One of our newest postdocs that we have here actually observed this phenomenon in mice, quite early, using CRISPR in the embryo, and she published that paper at her home institution.
Amy Koski: So, there were whispers of this happening. We didn’t think it would happen. Honestly, we weren’t exactly sure what would happen. We anticipated results, but we definitely were guided by our first few studies, what we were seeing. Mosaicism, getting an embryo that has some cells that are corrected and some cells that aren’t corrected, was also something that we were looking at, and it turned out that we were able to use Mosaic embryos to tease apart this question. We not only figured out how to stop the Mosaicism but also use the Mosaicism to figure out the self-correction and the lack of the template being used.
Dustin Driver: When you’re saying Mosaicism -- a little background. So, when you have a developing fetus, how many dozens of cells in the very early stages, and you try to apply your Cas9 treatment, it may not take through all of the cells.
Amy Koski: Yeah. What we have in the lab is a one-celled human egg, right? And the sperm. When those come together and you fertilize, you quickly get two cells and then four, and they go up. We stop at roughly eight-cell stage. Six to eight cells is all that we have, so there’s no fetal development. It’s truly still in the cellular development stage.
Amy Koski: In those individual cells, it’s possible that maybe one of the cells would show the correct DNA but another one would not, and that has always been the case with CRISPR/Cas9, so it’s why when you’re doing it in somatic cells, outside of the embryo development, they can’t get 100% correction. They’re looking for efficiencies and trying to increase that efficiency, and in our case, you’re never going to use the technique in an embryo and potential progeny and human child, if it is Mosaic. There’s no purpose to engage in such a therapy if the outcome is no different than just having a baby through natural means. So, it was imperative, actually, that we stopped this Mosaic embryo from occurring. We were able to do that when we injected at the time of fertilization. Once we saw that in the lab, that was the first step: how do we stop this from happening, because if we can’t, there’s no reason to go forward.
Amy Koski: We stopped that, actually, very easily. We were injecting after fertilization, we would get Mosaic embryos, then we injected at the same time as fertilization, and we had no Mosaicism. It was great. It was this glowing moment. We were all super excited. But then we couldn’t figure out why we weren’t seeing our template, because in the template we add in a marker that allows us to find it when we’re doing sequencing. We could never find it. And it was like, what is happening here? We know that we have more wild-type embryos, something is working, but what’s working?
Amy Koski: We actually went back and created those Mosaic embryos and analyzed every individual cell within that embryo, and that’s when we determined, oh, here is that mutant sperm. We can see the mutation in one of the six blastomeres from the embryo, but in the other five it’s not there. And so, we knew that we were getting this maternal copy being used to correct the embryo.
Dustin Driver: Okay, perfect. So, in a lot of ways that makes it a little bit more simple. At least in this specific case, you can just target the gene itself without having to worry about engineering the replacement RNA.
Amy Koski: Exactly. It actually does make it more simple, because you’re using the body’s own natural machinery to correct itself. We’re not, as humans, inducing a change to become something different. We’re inducing the body to correct itself.
Dustin Driver: That’s very cool. And the mechanism for that repair that happens is still a mystery. We’re not sure how the developing cells are grabbing that direct copy of the gene, or is it just something that happens naturally, biochemically?
Amy Koski: The why has not been teased out. We could speculate that over the course of human history it makes sense that if -- we know if you have a break in your DNA, there’s machinery that will come in and it will correct that. Errors happen, that’s why we age, right? Those types of things come into play. It’s not far-fetched to assume that this would happen. If you were a DNA machinery, would you rather select a twenty base pair thing to fix yourself or a complete copy?
Amy Koski: So, the answer is no, we don’t know why, but it makes sense that it is occuring.
Dustin Driver: It’s a natural part of gene repair.
Amy Koski: Correct.
Dustin Driver: It’s built into the system. That’s very good. I want to jump back a little bit. You were talking about Mosaic cells, and the reason why you wouldn’t want that is because the mutated gene, or the defective gene, would be continued to be passed on?
Amy Koski: Correct.
Dustin Driver: Unless you get all of it, you’re saying there’s not a good reason to go through with such therapy.
Amy Koski: Exactly.
Dustin Driver: What would happen if, say, that therapy was done and there was a Mosaic result? Would you get half of the syndrome? We’re not really sure.
Amy Koski: We’re not sure what would occur. There’s a technique used in fertility clinics right now called PGD. It allows you to develop the embryo, take one of those cells that I was talking about, the blastomeres, and actually analyze what’s going on in that specific embryo. You would be able to tell, does this embryo contain the mutation or not? That’s why we went through extensive rounds of our science and our data generation, to prove that if you actually inject the CRISPR construct with the sperm, we never saw Mosaicism. It never happened, and we did multiple, multiple cycles of this data generation. It was a very important point to drive home, that we really understand that if you do it this way you’re not going to have Mosaic embryos. In our hands it never occurred.
Dustin Driver: That’s amazing to have 100% success rate.
Amy Koski: It was really incredible, actually.
Dustin Driver: That could be something that’s super successful or very useful when you’re doing IVF, say.
Amy Koski: Correct. The whole purpose of this scientific generation is to actually provide, in the future, a means for families who want to have genetically-related children but carry these genetic mutations in their family line, and they want to stop it. They want to have healthy children, and they deserve that. We’re providing another choice, another option, we hope, in the future, for these families, no matter what disease they carry, that if they want to have a healthy, genetically-related child, they can come in, and this is another one of their choices for their healthcare. That’s the end goal.
Dustin Driver: That’s great. Let’s talk a little bit about this hypertrophic cardiomyopathy and how it presents. It’s a defect in the heart that leads to, later in life, heart troubles. The gene can be passed on from generation to generation, because the symptoms don’t present until you’re older, so you’ve already had children and passed the gene along.
Amy Koski: Exactly. Typically, what we see in families is that somebody will come in, and they’ll have a history of, like, my mom died of a heart attack at kind of a young age for heart failure. Oh yeah, Uncle Bob might’ve had some heart failure too. And what they find is in their thirties or forties they are showing signs of heart failure, or they themselves will have a very quick and unexpected heart attack. Then they end up being rescued, which is amazing, because we have this great healthcare system, and usually some implants are put in to help regulate their heart rate, basically implanted defibrillators to restart the heart. But that’s usually in your thirties or forties, and most American families, most families in general outside of America, in the international world, are having their families in their twenties and thirties. So, you’ve already established your children. You might have one, ten, twenty children, and you don’t know that you carry this mutation yet, so you’ve already passed on this cardiomyopathy to your future children, and then they might show the phenotype, they might not. But they will then be able to pass it on to their children. You have literally generations in a family who are suffering from the effects of this mutation, this inherited mutation.
Amy Koski: What we hope to do is try to limit the number of future generations that have this mutation. You can imagine -- we don’t need to imagine, it’s happening right now if you look at places like India where they actually have a very, very high prevalence of this specific hypertrophic cardiomyopathy, the frequency of the mutation is increasing in the population. What you’re having are more and more children who are then having their own children, and you’re seeing this disease phenotype perpetuate through a population at a pretty rapid rate.
Amy Koski: Right now we’re looking at heart disease being the number one disease in America. If we can start limiting sections of heart failure and heart disease by monogenic, inherited diseases, and limiting the number of future generations that carry this mutation, we can start to bend the curve back, similar to when Polio was out, right? Everybody wanted to come up with a way to eradicate Polio, so we came up with a vaccine. Now, we’re not proposing a vaccine here, but it’s the same idea, that if you can start stopping the number of people who will get Polio, then you will limit the number of future people who will have it. So, this gene therapy in developing human embryos can have the same effect, we hope, as vaccinations did.
Dustin Driver: It’s an important distinction that you made there, that just because someone carries the gene doesn’t mean that it will be expressed.
Amy Koski: Exactly. Different inherited, monogenic diseases have different phenotypic expressions. Some of them, if you’re a heterozygous carrier, meaning you carry one mutant copy, not two, you might not ever express the disease. However, when you have children, you could potentially have a partner who also carries a heterozygous copy, and then you can express a homozygous child who will definitely have the disease phenotype.
Amy Koski: Other inherited diseases you do express when you carry one copy, and hypertrophic cardiomyopathy is one of those. If you carry one mutant copy, you will eventually express the phenotype and have these issues, and that’s why we’re seeing it later in life after people have already established families.
Dustin Driver: This is a monogenic disease that we’re talking about here. There are tens of thousands of these that we know of so far that possibly could be treated using techniques similar to this.
Amy Koski: Exactly. In the news we hear a lot about BRCA1 and BRCA2, the breast cancer genes, thanks to Angelina Jolie. She put it out there for all of us to learn about, which is great. Women and men who carry this BRCA1 and BRCA2 will get cancer, primarily ovarian, testicular, and breast cancers. It’s a matter of when that will occur. If you’re a BRCA1 and BRCA2 carrier, female, the likelihood of you having breast cancer in your thirties is 80%, and it just keeps continuing and going up.
Amy Koski: Breast cancer we’ve gotten really good at detecting, really good at stopping, and the life expectancy out of that is really great. However, ovarian cancer is kind of a silent killer. By the time you locate it, it’s usually too late. These women who carry BRCA1 and BRCA2 are definitely having children before their thirties and forties. Sometimes not, sometimes later, but if you’re in your forties and you’re a carrier, you’ve probably already expressed some type of cancer.
Amy Koski: So, this could be used to stop the transmission of this mutation to future generations, and you can literally think about -- everybody wants to cure cancer. This could be a cure for cancer for future generations. You are stopping it before it happens. You’re stopping it in one cell versus trying to cure cancer in trillions of cells in an adult who already has the disease.
Dustin Driver: There are tens of thousands of these diseases or genes that have been identified so far, and more continue to be identified, probably on a monthly basis now, with how cheap gene sequencing and how quick it is getting. I guess a lot of people have a fear of where does it stop, and how do you draw the line between a disease and an enhancement?
Amy Koski: This is a tricky question, right? And tons of bioethicists and families and scientists are trying to answer this question. In my perspective and what I’m doing, it’s a fairly simple answer for me. I want to provide choice to these families, and our center would really like to provide a choice that ends suffering. We’re not looking at changing the color of your eyes or the color of your hair. Those kinds of enhancements are very different from what we’re talking about here. We really are finding disease, disease phenotypes that are known. We know that when you get Huntington’s disease, which is another one of these inherited diseases, that you will have neurological problems. It’s not a question of if but when. That’s a lot of internal suffering for that family. It’s a pretty clear line for us. You provide a choice for healthcare options. Right? This is another healthcare choice, and we’re not saying everybody has to do this. We’re saying, you know you carry an inherited mutation, and here’s an option for you to help you have a genetically-related child without passing that on.
Amy Koski: There are diseases that get us closer to that enhancement vs. therapeutic. You start looking at muscular dystrophies. This is the one that I’ve been reading about. A lot of people have been talking about this. Muscular dystrophy, definitely a disease, definitely something that we will love to add into that list of ten thousand. However, if you design a CRISPR construct that can snip out that mutation, the way that you would help these people is by generating more muscle mass. Then question then is, if somebody got ahold of that, if athletes got ahold of that, could they use it for enhancement? That’s a tricky question, and it is something that we need to continue to talk about, but when we think about making superhumans or hair color, eye color, those are almost impossible to do. We’re looking at one monogenic disease, one small mutation. We’re not looking at multiple spots on multiple alleles, so when you make one cut, it happens with high efficiency. We see this correction. You start making multiple cuts in multiple places, you have no idea what’s going to happen, and it’s not going to be as efficient. That would also assume that we know all of the sequences to make those corrections. We don’t. I just don’t see that being the future of humanity. I think we’re better than that.
Dustin Driver: Yeah, I see that. And just from a technical standpoint it seems almost impossible, when you look at the number of variables that go into making a person or even a cat, to be able to have this Island of Doctor Moreau, almost lego-block-like view of building creatures seems completely unthinkable. It doesn’t seem imaginable.
Amy Koski: As a scientist doing this with some of the best people in the world, I cannot even imagine it being possible. It is highly tricky, and we don’t even know how to manipulate sequences that would reached desired effects. And nobody’s going to spend money or fund that type of research.
Dustin Driver: Unfortunately. For science fiction fans, it may be a hard pill to swallow, but unfortunately, it’s the way things are looking. On the other hand, though, curing disease is a sci-fi dream. Just fewer sick people is incredible. Who doesn’t want that?
Amy Koski: The changes that we’re talking about won’t just impact one person. They’ll continue to impact our population dynamics around the world. If we can start removing the number of people who carry the mutations, by correcting that mutation prior to them being born with this disease, then they no longer carry it to pass on to their family, and it stops.
Dustin Driver: Yeah, that’s fantastic. I immediately think of the complexity involved, and when you’re dealing with one gene and correcting one disease, I wonder -- obviously this is very very super early in the research phase, but can we know, ultimately, if changing one gene to fix this one inherited disease is not going to have other consequences? Given the complexity of the animal, the nature of what we’re working with here, it’s a big question that I have.
Amy Koski: I think the best answer is that when we induce the change, via the Cas9 making the break in that DNA, the machinery that it chose to fix itself was the DNA that’s already existing in the human population. We’re not actually putting anything new into that DNA strand. When they met over two years ago -- the International Committee of Medicine met to discuss about this -- they talked about, if this was going to go forward, the change that you made has to be currently present in the human population. That DNA code that you put in should not be different than what we already are seeing in our very varied human population. So, by inducing that break, and by that break being corrected with the maternal copy of the gene, we actually aren’t changing anything that could be skewed to cause off-target effects, and that’s kind of what you’re going at. We make one change, what else did we do by making that change? Because you’re using that DNA copy already existing in the cell, you really are limiting the potential negative consequences.
Dustin Driver: Yeah. And that could happen naturally anyway, right?
Amy Koski: It does happen.
Dustin Driver: Right. As you were saying, 8% of the population that you’ve studied would have this particular gene. Well, that’s because it’s not always expressed or passed down, so there’s already a little bit of a self-correction going on?
Amy Koski: No. So, if you carry this mutation and you have a child, depending on how many copies of that mutation you have, whether it’s one or two, heterozygous or homozygous, you will pass it on. There is no correction happening in the body. The DNA breakage machinery that we have naturally is actually to protect us when we’re exposed to sun. We have all sorts of DNA breakage events that happen in our somatic cells on a daily basis, and our body comes in and says, oh my gosh, I know how to do this, I’m going to fix you, I’m going to use this copy. That mechanism is well known in adults and people in your somatic cells. Trillions of cells are doing this every day. And when that goes astray is when you get things like melanoma. These are very tangible effects of our DNA machinery repair mechanism breaking down. What is new here is that we’re actually able to see that machinery happening in one cell, and that is very cool. That is a really novel finding that came out of the science that we’re very excited about.
Amy Koski: There will be some instances where you’re going to have a homozygous embryo, right? You will have two parents who carry mutation, and there won’t be a copy of this DNA sitting there, and that’s a big question. We really want to answer that, and that’s when we will be using a template design. How do we do that appropriately and make sure that that embryo wants that template, and that that template is representative of the healthy population already.
Dustin Driver: Wow, okay. That would be the next step, because if you don’t have a healthy copy of the gene to replace the defective copy, what happens? How does the machinery work? Will it actually refer to that RNA sequence that you’ve engineered and injected in and say, oh look, that’s a good one. I’ll grab that one. Or will it ignore it completely, and there’s nothing you can do?
Amy Koski: Yeah, and we don’t know yet.
Dustin Driver: We can wrap up. I just wanted to know, what are the next steps now? Where do we go from here?
Amy Koski: The next step is, we’ve done one study, and one study will never get us to a clinical trial. We need to improve all of our numbers. We really want to improve the efficiency of what’s happening. We were able to correct roughly 75%. Some of those were already wild-type non-carriers of the disease, of mutation, so we want to really increase that number and see if we can get the efficiency higher. We want to look at homozygous embryos. We would love to start looking at other mutations too.
Amy Koski: So, what we’ve done right now is say, for hypertrophic cardiomyopathy, who carries MYBPC3 mutation, very specifically, we think we know a way forward. But we need to make that broader. We need to know the way forward. We need to have it safe and efficient, and we also need to start looking at other mutations to make sure what we saw in this mutation isn’t just specific to that mutation site. We want to make sure that this can be used on any allele, any chromosome, and for any of those ten thousand mutations.
Dustin Driver: After we stopped recording, we chatted a bit about PGD: preimplantation genetic diagnosis, which is essentially screening embryos for genetic disorders before IVF implantation. The question was why not just screen embryos instead of trying to fix them. Well, the goal, says Koski, is to eliminate the gene from the population entirely, so that nobody has to worry about it in the future. You’re preventing untold amounts of suffering and also expense by removing the mutated gene from the family line.
Dustin Driver: Of course, Koski and Mitalipov aren’t proposing that this sort of gene editing should be mandatory. They just want to give couples the option or the choice to do this kind of repair if they want, and like we said during the interview, there are more than ten thousand identified single-gene genetic disorders, including cystic fibrosis, sickle cell disease, Fragile X Syndrome, muscular dystrophy, and Huntington’s disease. So the team at OHSU has their work cut out for them.