CRISPR: Policy, Platform, Trials (#11)

Hi everyone, welcome to 16 Minutes. I'm Sonal, and I'm here today with the 11th episode of our podcast where we quickly cover recent headlines the Asics and Z way. Why they're in the news, why they matter from our vantage point in tech. Sometimes we cover multiple items, sometimes we go deep on just one or two topics. So this week we're covering all the latest news, not research, in CRISPR. Really quickly, CRISPR stands for Clustered Regularly Interspaced Short Palindromic Repeats. Phew, what a mouthful. Basically, it's a DNA editing technology that's not a single tool or protein, but is really a platform for genetic editing and can be applied to all kinds of bio uses, including diagnostics and therapeutics in medicine. By the way, if you want to more deeply understand the technology challenges and startup opportunities in the world of gene therapies, including CRISPR, please visit a6nz.com/gendeck for a video primer with downloadable slides. I also always include relevant background links in the show notes. You can find those at a6nz.com/16minutes. Okay, so now onto the news summary before I introduce our experts. First, in the state of California, the governor recently signed a bill that will become law in January, so just a few months away, that requires warnings on gene editing kits. Second, the Alliance for Regenerative Medicine, or ARM, which is an international community of small and large companies, nonprofit research institutions, et cetera, very recently made a statement that they do not support germline editing. And that's especially relevant on the tails of the news, really scandal, from earlier this year, first revealed by MIT Technology Review, that a Chinese research scientist, He Jiankui, announced the birth of twin girls, plus a third child that was supposed to be born last month, all with edited genomes. And finally, beyond these policy aspects of CRISPR, the other latest news is of the first CRISPR patient to ever be publicly identified, a 34-year-old woman in Mississippi who volunteered for a study to treat sickle cell disease. And then there was the news of another clinical trial where CRISPR will be used to treat people with LCA, which causes inherited blindness. So that's the context for this episode. And joining me are A16Z biogeneral partner Jorge Conde and deal partner Andy Tran. I can't believe we're trying to squeeze this huge topic plus a bit of relevant history too into 16 or technically 20-ish minutes. But first, really quickly on the China CRISPR baby scandal, we still don't know if that scientist actually succeeded in modifying those genes. We do know he was later fired from his university for violating safety and other norms, and that there's an ongoing criminal investigation. But beyond the broader debates around all that, what did he actually do, Jorge? And what were the specific debates related to the ethics of pacing such CRISPR experiments?

So in that case, what that experiment did is it edited a gene called CCR5, which would reduce the susceptibility of contracting HIV, developing AIDS. It's a very laudable thing to protect people against infection of HIV. Is editing the CCR5 gene permanently the right way to go about that? And so I think there were a couple of debates that emerged there. I think the two most notable ones is, number one, you weren't curing a disease, you were essentially providing almost like a genetic vaccine. And by the way, it has some unintended consequences in that having a mutated CCR5 might make you less susceptible for contracting HIV, but it could make you more susceptible to contracting something like West Nile virus.

Which is just a great example of how complex biology is.

That's right.

Complex dynamics. Dynamic interactive system. You can't linearly take one thing and not have it, you know, have effect on other aspects necessarily.

That's right. So that was really the first issue that came up. The second one is, of course, the more meta question, which should we be doing germline editing in the first place? And the reason why there's the risk to the embryo or to the eventual child or individual, and there's also the question that those are permanent changes, that when you make edits to an embryo, those will be carried down to future generations.

Right. I mean, that's where the phrase of designer babies comes in. It's not exactly the same thing because that was a case where it was like a preemptive vaccine, essentially, but take it a step further and then you're actually designing custom, like, features you potentially want through that type of germline editing. So let's talk about this alliance news. I mean, it's kind of a big deal that 13 major companies in the space are declaring proactively that they're not going to do germline editing. They do, however, endorse therapeutic applications of somatic cell gene editing. What specifically is somatic cell gene editing?

So think of it this way. In your body, you kind of have two genomes. You have the genome that essentially runs your body, and then you have the genome that you're going to pass on to future generations.

And so the latter one is a germline.

Yes. So if you make an edit to the cells that make up the parts of your body that aren't the reproductive cells, so, you know, eye cells, liver cells, lung cells, that's somatic editing. That, by definition, will change the DNA in you as an individual, but will not be passed down to the next generations. Now, in contrast, germline editing is making edits to either the cells that exist in the egg or sperm, or more specifically, from a practical application here, making edits in an embryo. And so I think when the Alliance says we're not going to do germline editing, I think to put a more precise point on it, they're not prepared to do germline editing yet.

Ah, because that's my question. When you read it, especially in the headlines, it seems like a declarative, hey, Hey, this is not a good thing. So you're saying it's really a timing thing, not necessarily a no thing.

Yeah. If you read through the language in terms of the principles that they've stated as part of their call here for action, a lot of the language is around it being currently inappropriate.

Right. They say, we as therapeutic developers utilizing gene editing technologies are not modifying human germline cells for use in human clinical studies. They have not matured to the point where human trials of edited germline cells are appropriate. Many important safety, ethical, legal, and societal issues involved with this type of gene editing remain unresolved. How does this affect future development? Because it's essentially moratorium, that's saying no more.

Frankly, this moratorium does nothing to really affect the state of therapeutics. All the CRISPR clinical trials are actually somatic cell gene editing. The majority of the work actually is how do we use this technology to really edit for certain diseases, things like thalassemia, the eye, congenital diseases, and cancer immunotherapy. All these are for actual therapeutic treatments.

Oh, good. So not so bad. So why don't we take a quick step back briefly into some of the history that's relevant here? I also think it's interesting because people think of CRISPR as an overnight success, but gene therapies themselves have been decades of work in the making, and some of it may have parallels to the recent policy news. But first, Andy, do you want to share some of the history?

Yeah, sure. So in the '70s, that is when we had the discovery of the recombinant DNA, when we had the technologies to cut and paste genes. So in the '80s, what a lot of people used these recombinant DNA to— is for actual different genetic applications that were outside of the human body. Now, if you actually compare that gene therapy 1.0 versus all these new age technologies of CRISPR, the early days of gene therapy was very rudimentary. You would basically cut a healthy copy of the gene and shove it into a virus and stick it into the human body, hoping that it would somehow randomly and stochastically integrate into the cell. Fast forward into the late '90s, we had a very unfortunate circumstances of this patient known as Jesse Gelsinger, and people realized that there are a lot of immunogenicity and various toxicity concerns to this type of technique. You can't just randomly stuff genes. You actually want to remove the source of mutation, actually delete and edit it out. And that spawned the whole generation of genome editing. And so in the early days, you have these systems known as zinc fingers or talons, and basically these are large gobbler proteins that can essentially go to the site where you want it to go and take it out, take out the mutated gene. But the problem is to engineer these zinc fingers, you know, take PhD students on order of months and tens of thousands of dollars.

It's not cost-effective to actually do this at scale.

Even at the experimental level, it's way too expensive.

Right. And now what was the big significant insight that went into CRISPR and Cas9 proteins?

Well, it's interesting. You mentioned a lot of this being built on the foundation of work over decades. CRISPR is this kind of unique thing in that there quite literally was a bit of a eureka moment.

Oh, I love these kinds of stories.

Yeah, because what ends up happening is that scientists discovered that some of these bacterial systems use the mechanism of CRISPR as almost as a way to create an immune memory. So things that they had seen before that had attacked them but not killed them, they would essentially edit their DNA to store portions of that DNA as sort of a new memory. And so when that was discovered, that led to the eureka moment like, well, wow, this is a very efficient mechanism for editing DNA. Could this be used for other applications? And I'm obviously oversimplifying it, but it's a big part of the reason why there's so much back and forth over patents.

Since then, there have been tons of advancements, including different types of proteins like CasX. There's like, you know, people are mining proteins in all kinds of places, searching for other proteins.

The New York subway system.

Are you kidding?

The New York subway system?

Oh my God. So basically gross places where things grow. People are mining proteins.

I don't know about San Francisco streets, but you know.

Let's not go there.

Let's keep this PG.

So what we're really talking about here now is the next phase of CRISPR, CRISPR as a platform.

You're precisely right. The first generation of CRISPR tools utilize, you know, Cas9, Cas12, and there's new proteins that are coming along. Every single day. And so the interesting thing, the CRISPR-Cas proteins are not only used as a cutting tool anymore. You can actually reprogram it as a platform and just using it as a DNA search engine or DNA targeting machinery.

That helps with drug discovery.

Exactly. CRISPR can easily target an entire array of genes, and you can rapidly screen for certain variants there. And if you replace the cutting enzymes with different tools, like if you replace it with activator systems, you can actually activate genes or repress them. You can even replace it with certain deaminases that allow you to to edit bases at the single-level resolution. And there's even wacky applications where you can replace these cutting tools altogether and replace it with imaging modalities using CRISPR to very precisely image parts of your chromosome, for instance, right? And so there's a wild array of applications that are spawning about just using CRISPR as this software API layer, if you will, to build new applications about.

That's fantastic.

I think one thing that's really interesting about the emergence of CRISPR as a therapeutic modality—

and by the way, in layman's terms, that's basically the emergence of CRISPR as a way to treat disease.

Correct. The fact that CRISPR in relatively in short order is now being used to treat some pretty, pretty devastating diseases is a remarkable thing in and of itself. What's also remarkable is that the emergence of CRISPR as a way to treat disease is just one of many new ways that we now have to treat disease. So we have gene therapy, we have CRISPR now emerging in clinical trials, we have engineered cells that are being used to treat cancer. So we're in this moment where the armaments that we have to go after disease are getting far, far more powerful and far broader. And I think this is just an incredible moment in time.

But I have to ask, what's hype and what's real? We are in this great moment, but when we hear the coverage around CRISPR, what are the things that people are exaggerating the claims about? What are some of the frequent misconceptions that you see come up?

Well, I think one of the misconceptions is that we're there, that it's done. I mean, this is early days. The second one is when we describe CRISPR, people, including ourselves, tend to oversimplify it and try to use, you know, analogies like find and replace, cut and paste.

Always the editor's favorite one.

Exactly. Obviously, biology is much more complex than that, so it has its own challenges, you know, off-target effects. You know, the genome is 3 billion base pairs long. If you make edits across that, you could have unintended consequences.

So that's where the news on California comes in. Speaking of unintended consequences, they're not banning gene editing kits, but they are making sure that there are warning labels which say the kits are not for self-administration. And of course, gene editing is obviously very different than permissionless innovation with the internet. And we know technology tends to come first and regulation follows. But I'm actually curious on whether this also stifles innovation in this area or not. What's your quick take on the California news?

Well, this is an example where I think the legislation is probably a bit ahead of where the technology is. As far as I know, you can't buy a CRISPR kit.

So it's preemptive.

It's a bit preemptive. And so just like it was with recombinant DNA protections in municipalities back in the 1970s, people are trying to be thoughtful about what could go wrong and how can we be protective here. I believe that the city council of Cambridge, Massachusetts, where Harvard and MIT are, were worried about that stuff you know, getting flushed down the toilet, and then you'd have mutants in the sewers, right? So, like, there was an actual moratorium for a while on figuring out how to use recombinant DNA technology, because these are powerful tools, and used improperly, any tool can be a weapon.

Right. It can go both ways.

It's true for gene editing. It's true for software. That's true for baseball bats, right? Like, lots of things can go wrong if used improperly.

You did a really great conversation with George Church, who is a pioneer in gene editing, and he made a very interesting argument that I want to quote back to you. He said that there will be something that looks like a pause, but it will probably be an acceleration. And it's so counterintuitive to me that he said that because it's almost like putting a moratorium on things actually speeds things up. Explain that.

He was referring to is the moratorium that was placed on recombinant DNA. And as he experiences firsthand, his research actually accelerated. It forced people to think through how to work with this technology. In a very directed, in a very deliberate way. And so his argument there is that the investments were made in the equipment, the investments were made in the expertise, and that actually helped accelerate technology. I think similarly here, there's so much that needs to be perfected with this technology that to the extent that we get better at how to target specific cell types or specific organs, to the extent that we get better at how to ensure that there's error-free editing, To the extent that we get better at ensuring that we can do things like redosing CRISPR and not having some sort of an immune response, all of those things are going to advance this CRISPR as a modality for treating disease in a very meaningful way. If and when we get to the point where those things are dramatically improved, you will reach a point where the other issues around ethics and societal concerns will also catch up. And so what looked like a pause in many ways may in fact be an acceleration. What's interesting to me is if you sort of go back when genome sequencing became sort of an available thing, the federal government went ahead and passed legislation called the GINA Act, the Genetic Information Nondiscrimination Act, sort of as a way to preemptively say, well, look, if we get to a point where everyone's DNA is available and you could basically search for it online, you know, I can't discriminate against you based on something that's in your DNA. At least for the purposes of healthcare. And so I think we're seeing a version of that now, which is, look, it's not widespread today. You can't do it in a garage today, but would you be able to do it in a garage 10 years from now? Perhaps you will be.

Okay, well, let's talk about where we actually are right now then in practice with CRISPR. I mean, to me, the most exciting news is that clinical trials are coming of age. And how are people pacing relative to all these policy concerns that we've been talking about so far?

First thing that's worth noting, the clinical trials are in diseases where the driver of the disease is very well known. Take sickle cell anemia, for example. This is where you have a mutation in the hemoglobin gene that causes red blood cells to fold or sickle. That's where the name comes from.

And it disproportionately affects African populations, African Americans, and Ashkenazi—

Yes, Ashkenazi Jewish populations, I think, also have a high prevalence. And the strategy for treating this, you have a hemoglobin gene, and you actually have something called fetal hemoglobin gene. So the fetal hemoglobin gene is what's present in utero. When you're in the womb, and that's what helps sort of, you know, oxygenate the baby. And when you're born, fetal hemoglobin gene essentially shuts off and the hemoglobin gene takes over. So the thesis has long been that the way to treat sickle cell anemia— what this trial is trying to do is, can you edit the fetal hemoglobin gene to essentially reactivate it in adults? Basically turn it back on.

Right.

Right. And what folks have been, I think, very thoughtful about is figuring out what is the right application, what is the right disease, and where to intervene with CRISPR. The first thing is you do it in diseases where you really understand what the genetic driver of the disease is. So in that regard, you're not taking any sort of true biological risk.

And just for context, that's because when you think of biology, it's such a complex system that generally there aren't a lot of diseases where it's single-factorial.

Yeah, these would be called monogenic conditions generally, and that's exactly what that is. And what's really interesting about the two examples you highlighted, LCA and sickle cell anemia, in both cases they have that phenotype.

So you mentioned a few terms. We have in utero or in utero, which is in the womb. We have in vivo. We have ex vivo. Ex vivo, which is outside the human body. I've heard laboratory terms like in vitro. There's all kinds of different things. Can you kind of break that, how this plays out with CRISPR?

Yeah, I think most importantly, without getting into medical terminology lecture, is that for the case of CRISPR clinical trials, what's really important is this differentiation between the ex vivo and in vivo trials. What Jorge mentioned with sickle cell anemia and other trials like beta thalassemia, for instance, and then cancer immunotherapy, all of these are ex vivo trials. So ex vivo are just these gene therapies that take place outside of the human body.

Why does that matter?

It's super important because as CRISPR is a novel and still clinically developing tool, we want to make sure we hit indications where the biology is well known, but also in instances where we can well control it. When it's outside of the body, we can utilize CRISPR however we want and then utilize other analysis techniques like deep sequencing, for instance, to determine if we really did the type of genetic edit that we want without any mutations, right? So it's a really controlled environment where we can really analyze and understand if a cell and the drug really work perfectly before we put it back in.

By doing things ex vivo, you can essentially QC or quality control your work. If you want to abuse the editing analogy, this is almost like doing it ex vivo allows you to proofread your work before publishing. Exactly.

Well, I would also say the other thing I heard there besides the quality and containing and control is it also lets you do a lot more because you can actually do all the things you want to do that you have in your toolkit outside the body.

That's precisely right.

Because when you actually do it inside the body, you were very worried about the delivery mechanisms and various other things that can have other unintended consequences on the route to the gene. So that was the other opportunity I heard in there.

That's an important thing to point out. You have two components to this, right? One is, in many ways, like the cargo, like what are you going to deliver to the cells? And then the other one is the vehicle. How are you going to get it there? And so the reason why ex vivo helps is it minimizes in a real way the vehicle problem, the delivery problem. I mean, it's still a challenge. You still have to get the edits into the cells, but it does lower the burden there.

And then the delivery dimension is very important. And that really takes us to the in vivo world.

In life, because vivo is from the Latin root, vive, or like vitality or vivid, something having to do with life.

That's precisely correct. And this is actually what made the news recently is for the first time CRISPR is actually used in vivo in the human body. And, you know, the particular clinical trial was Adytas and Allergan, and they used it for particularly LCA10, Leber Congenital Amaurosis 10, the congenital blindness disease. And they picked it precisely because it's in the eye, because the eye is First, for one, it's more immune privileged. And then most importantly, for the delivery, it's in the contained cavity. You can actually inject it into the eye. So it's more like a hybrid approach between fully systemic, as if you were delivering to the lung, and then ex vivo. And so it's a means of how we bridge that gap.

That's why the eye is such an interesting initial target, because on the one hand, it's immune privileged, as Andy just mentioned. So there's less likelihood of an immune response. Remember, these are bacterial components that we're using.

Right. We don't want the body to attack itself, which it's supposed to do to protect itself from disease.

Yeah.

And the second one, the eye is believed to not be very leaky. So the likelihood of sort of the CRISPR leaking out beyond the eye is relatively—

What's in the eye stays in the eye, to use the Vegas line. Exactly. Pretty much.

There's no worry about off-target effects touching other organs and etc.

I thought you were gonna say CRISPR's in the eye of the beholder.

God, we're going to town with the cheesy analogies. What are other parts of the body that are sort of like the eye in that they're kind of self-contained that way?

Another one that a lot of people are focused on and target is the liver. And the reason why is that if you do deliver something systemically, it's going to pass through the liver first. And so the liver sort of becomes the first stop, the first port of call, if you will, of anything you inject systemically. And so a lot of the areas of research and focus for things like CRISPR and other new modalities tend to focus on the liver first for that.

That's so interesting. I had no idea.

In the case of LCA, you're injecting the CRISPR system directly into the eye. In the case of sickle cell anemia, you're taking the cells out of the body and editing them and then putting them back in. And so in both cases, the delivery of CRISPR is at some level contained.

What are some of the more practical considerations that are remaining just to help put the reality of this clinical trials news in context?

Yeah, I think one of the other areas that I think it's gonna require a lot of thought as these enter the clinic is how CRISPR is delivered as a medicine. And I don't mean sort of at the cellular level, I mean just in terms of treating patients. So CAR-T therapy for cancer patients, that requires an entirely new way of thinking about both not only delivering a medicine, but making the medicine in the first place. They take out immune cells, they send them to get them edited, they get edited, they get QC'd, and they get sent back and they get put back in the patient. So CAR-T is sort of innovating this area of what people call vein-to-vein delivery. Take it out, fix it, and then put it back in your vein. So that's a new way of thinking about delivering medicines. And I think as we think about CRISPR applications, essentially the system is going to have to find ways to make sure that patients can be treated in this way. So it might be as simple as an injection in the eye. It might be an infusion. And so as clinical trials start to come online, thinking about how we physically get these medicines to patients is going to be an important consideration.

Exactly. And then also, aside from delivery, actually making these drugs are really hard. Making these complex biological biologic vectors, right? These large complex cells, right? Having the manufacturing infrastructure and the software to really manage all this logistical supply chain.

It's engineering to your point.

How do we make these tools more precise? How do we actually deliver these systems more effectively? And even how do we have the infrastructure to manufacture and create such complex vectors and cells, right?

How do we do it in a high throughput way that you can do it on the scale of something that's fast and affordable, not only for very, very rich rich biopharma companies. There's so many different layers to this.

Yeah, it's super important, right? Currently, we don't really have all the tools to have the capacity to produce all these gene therapy drugs, right? And if you think about it, there's only 4 gene therapy drugs on the market. And so the next decade, we're projected to have, you know, 50 of these type of targeted gene therapies. So how do we really have the infrastructure to really develop for that? And so that will create a whole new ecosystem for, you know, scientists, engineers, and founders to really innovate So bottom line it for me, you guys.

How should we think about all these sort of emerging regulations and these first public patients coming out? How should we be thinking about the broader story of CRISPR beyond the news?

CRISPR really shifts the early days of old-school drug discovery into making biology this more precise engineering discipline. Everything now can be using CRISPR as its platform to have much more molecular precision. And so everything that we see in the whole drug development pipeline is really shifting that way. Not only is it the world's most efficient tool to knock in and out genes, But it really is a platform for something so much more.

It's important to remember that CRISPR is not just one thing. In many ways, really what we're talking about isn't CRISPR per se. We're talking about the fact that we now have technologies, among them CRISPR, that are going to enable us to very precisely edit DNA. And that's going to have a lot of implications for how we treat disease. It's also going to have a lot of implications for how we do other things in science, like diagnostics, basic research. And it's early days, so don't expect to see X-Men or superheroes walking around anytime soon.

Well, thank you for joining this segment.

Thank you so much.

Thanks for having us.