Precision BioSciences, Inc. (DTIL) Earnings Call Transcript & Summary

We did a survey a couple of years ago with about 2,000 people living with chronic hepatitis B. In the U.S., the #1 thing that they wanted, their #1 priority was reduction of liver cancer risk. Outside the U.S., especially in Africa and Asia, areas of the world where people are discriminated, their #1 priority was losing surface antigen, because if they test negative for surface antigen they can get jobs, they can go to school and they can immigrate. For someone living with chronic hepatitis B infection, I think achievement of functional cure would literally change their lives outside of the fact that they would live a longer life, that they've reduce their -- their risk of liver cancer would be significantly reduced and their risk of cirrhosis would be significantly reduced. We would also see great improvement in quality of life. People would be less scared that they were going to die young. They would not face discrimination anymore. They would not face stigma and isolation from the community. I think it would be a huge burden off the shoulders of people living with chronic hepatitis B infection.

The goals of current treatment with chronic hepatitis B is to suppress replication, which is quite effectively done with nucleoside analogues. Interferon alpha is still used in some parts of the world, but the effects are somewhat unpredictable and only a minority will respond to a finite course of interferon. So the goal of treatment is to reduce replication, and that reduction in replication is accompanied by a lessening of the severe outcomes. Probably by about 70% nucleoside analogues target a late stage of Hep B replication. So they don't eliminate the virus. They don't provide a cure -- they certainly don't provide a sterilizing cure. But they will reduce some of the morbidity from the disease as long as maintenance viral suppression is maintained. We know that nucleoside analogues do not eradicate cccDNA, neither do they actually reduce transcription from the existing pool of cccDNA. So if we use more sensitive markers like pre-genomic RNA and serum or co-related antigen and serum, we can still indirectly detect the pool of cccDNA existing in those patients on nucleoside analogues and ongoing transcription. All our efforts are looking at new parameters, new lateral directions of treatment because we're not achieving functional cure with nucleoside analogues and we're struggling to achieve functional cure with some of the newer capsid inhibitors, mollifying RNAs, entry inhibitors in development at present. New gene editing therapies offer great promise. They would target cccDNA and hopefully also target integrants of chronic Hepatitis B to truly affect a knockdown of HBsAg because of a choke on hep B replication, but also the expression of HBsAg from integrated viral genomes, which clearly has a profound effect on immunologic tolerance. If we had gene editing therapies that would actually scissor out the important expression of HBsAg and possibly HBx from cccDNA and knockdown expression of HBsAg, everything would be turned on its head. The current place of gene editing therapies would be to enhance the possibility of functional cures.

If we want to achieve functional cure, we need to lead to a sustained deactivation or eradication of cccDNA. We can't do that at this time. None of our current medicines, either the antiviral or the immunology approach are able to achieve that. I think the only way we'll ever achieve functional cure in a large proportion of our patients is to completely inactivate or eliminate cccDNA. To achieve that we need to target cccDNA directly, and for that we need a gene editing approach. I think a gene editing approach such as ARCUS would have a huge benefit if we were able to successfully eradicate or deactivate cccDNA after a single or a short duration of treatments and so that the person would no longer need to be on long-term treatment.

Good morning, and welcome to the Precision Biosciences 2023 In Vivo Gene Editing Day. During our presentation today, we may make forward-looking statements about our product candidates, market opportunities and our liquidity. Actual results could differ materially, and we're subject to all the risks inherent in conducting preclinical work, and our SEC filings including our most recent Form 10-Q, state important risks and other factors that could cause our actual results to differ materially.

Good morning. Welcome to Precision BioSciences' 2023 R&D Day or In Vivo Gene Editing Day. My name is Michael Amoroso. I'm the President and CEO of Precision. And I want to thank our investor community today for taking the better part of their morning to spend time with us and see our progress. I'm joined today by a top tier of talent. First on my team, our Co-Founder, Dr. Jeff Smith; Dr. Cassie Gorsuch, who leads our Discovery; Dr. Wendy Shoop, who leads our Mitochondria programs. We're joined by an amazing cast of some of our key investigators and thought leaders from around the country, Dr. Mark Sulkowski, who leads liver and hepatitis diseases; Carlos Moraes and Dr. Michio Hirano, who are experts in the Mitochondria disease area. We've got a wonderful agenda for you planned today. Today, we will talk a bit about Precision 2.0, our new identity, In Vivo Gene Editing singularly focused platform. We'll talk about ARCUS, our proprietary gene editing platform and its differentiation in vivo through data. Cassie is excited to show you and Wendy a bunch of new data today. And last, we'll unveil some of the progress against our clinical development plan. Hopefully, we'll have you understand why we're focusing ARCUS with specific types of edits in certain diseases versus the others in the gene editing landscape today. Precision 2.0, the transformation to a singularly focused in vivo gene editing company is complete. The last 18 months after assembling a really talented management team, we have had a strategic focus down a path of continuing to hone our core competency and exploit that for shareholder and stakeholder value. You may remember back to the end of 2021, early 2022, when we divested our agriculture business. Recently, you've seen that we have sold our lead asset in cancer and our CAR-T infrastructure to a partner, enabling Precision for the go-forward future as a single-focused, one-platform, right-sized in vivo gene editing company, what we believe is our core capability. That will be propelled through over 25 years of in-house experience in gene editing platforms beyond just our platform today, a proprietary gene editing platform wholly owned by Precision, ARCUS, and a wonderful strong IP foundation for that platform. You will see now on the back of our last accretive event through CAR-T, we now have a cash runway that enables us into the clinic with our own organic programs. So let's talk a little bit more about the gene editing landscape today. The gene editing landscape is early, it's nascent, it's limited. You see from the visual in front of you, it's focused heavily on the liver and it's focused on disabling or knocking out genes, gene deletion. We believe ARCUS can do more. You see below the line in front of you that there's other tissues of the body, and there's 2 reasons that we believe ARCUS can do more. One is an ability to deliver to those other tissues of the body. But an even more -- and even more important advantage that we believe is ARCUS is derived by nature to do a more tailored and sophisticated type of editing beyond just gene deletion and knockout. ARCUS was created to restore or enable function. Let's talk a bit more about that. You see our gene editing tree. ARCUS will focus on sophisticated gene edits. At the bottom of our tree, we see gene deletion or knockout. We at Precision refer to this as a commoditized gene edit, a more simplistic. Most editing technologies -- all editing technologies can do this today. Most editing technologies will leave a scar and disable function. But as you go up the branches of the tree, you will see what we call more sophisticated edits, where ARCUS will focus, where we believe we have a leg up on the competition. You'll also see the difficulty of these editing intended types goes up. Some of those areas, elimination, gene elimination, you'll see an example in our HBV program, where we cut out an entire viral genome, or gene excision at the muscle through our DMD program, where you'll see an excision -- and we'll talk to you about how important the repair mechanism of perfect re-ligation is to enable that more sophisticated edit in the ARCUS Advantage -- or adding function through insertion, adding a DNA template, or replacing, removing a portion of a defective gene, and inserting a DNA template. Let's quantify the more that we believe ARCUS can focus on, the top half of the tree. First, you see on your left-hand side the genetic medicines market, growing up to 2030 to over $25 billion, up to $35 billion. Please focus on the yellow points. That would be the gene editing growth within the genetic medicine market. And we know what fuels that, the prominence -- the promise of permanence. Today, conventional gene therapies deliver to the nucleus, but we have waxing and waning issues when it comes to the desired therapeutic effect. Gene editing offers us potential permanence. And you see the growth of the gene editing market. Being able to gene edit beyond just simple knock out, we believe ARCUS can own a great proportion of this market. But it's not about dollars and cents. It's about the human beings that it represents. On your right-hand side, you'll see gene editing is no longer for ultra-rare diseases and just a niche. Up to 500,000 patients just represent the faces of the different patient types we'll be looking at through Precision's clinical development plan, either organically on our own or with and through partners. ARCUS can do more. Now, before I tell you about the key 3 critical advantages of ARCUS and open it up for my esteemed colleagues today, let's go back one moment to gene editing 101, the left side of your screen. On-target gene editing, we talk about this all the time. You read about it in the regulatory environment, biodistribution, germline editing, making sure that when you permanently alter the genome, patient safety is first and foremost. We agree. We've talked to you for several years about the unique signature that ARCUS creates when they cut so that we can clean up the genome when we pick a clinical candidate to go into the clinic, making sure we have inconsequential off-target editing and we're solely focused at our on-target site. Now, on-target specificity is a non-negotiable when we talk about clinical candidates. Today, we're going to spend our time on gene editing 102, the right side of your screen. What happens and what is the differentiation between gene editors in the landscape prominent today in development once you've bound to your intended target site and you make a cut. And the short answer is how you repair and the state of readiness for repair makes a difference, in what we'll talk a lot about today, the therapeutic or desired outcome. When you simply knock out the bottom of the gene editing tree, repair mechanism doesn't really matter as much. But when you're focused on a sophisticated edit, excision, insertion, replace, how you repair is everything. Staying consistent with the framework, and I promise this is the last slide that the non-scientist will talk to you about scientific definitions. But it's very important for Jeff, Cassie and Wendy's presentation today to keep us in the right mind frame of exactly where we're identifying the ARCUS advantages. Now we're talking about on-target editing. Efficiency, a term we use all the time, simply stated the percent of cells that are edited. But to our investor community, we ask you -- we ask for your help. This is not enough. Let's drill down. Ask the next level of question, "What happens once that intended on-target edit occurs? Once that cut is made?" Do you have, on your left-hand side, a defined outcome, predictive, the intended or desired edit with the right repair mechanism, therefore therapeutic effect? Or do you have, on the right hand of the slide, the random outcome? And this occurs when you do have on-target editing, but you do not have the right repair mechanism, and more often than not, a random event. Maybe the insertion that was desired did not occur, therefore therapeutic index is compromised. Now you say, "Wouldn't everybody pick the left side of the slide?" And the short answer is yes, but you have to have the right tool for the right job. And that's where the pathway of how we repair becomes the central focus of what we will talk about today. Defined outcomes very simply use homology-directed repair. This is when the cut occurs, the DNA sequence matches with the DNA template, matching sequenced order for a defined outcome. Now what if you don't have a DNA template for insertion? Well, very simply the cut is made, matching sequences, complementary ends come back together for perfect re-ligation, critical for gene excision, like in our DMD program and a random outcome. An on-target edit occurs without the proper repair mechanism, which more often than not leads to a random outcome. Now I know I speak a bit in absolutes, but the key here -- and this is biology -- all editors have a compilation of defined and random outcomes. But ARCUS, where we will focus it, has a higher propensity for defined outcomes, and therefore, therapeutic index. Now, the 3 key ARCUS advantages that allow us to focus on sophisticated edits. The cut. We've talked about the repair mechanism. Why -- why does ARCUS lend to an HDR more often? Simply stated, ARCUS leaves a 3 prime overhang, the correct DNA replication sequence. It repairs through HDR, and in the presence of a DNA template, insertion will occur. Or it perfectly re-ligates for an excision in the absence of a DNA template. Size matters. Size is a practical advantage, sure, to be able to have non-viral or viral delivery, LNP or AAV, LNP to the liver, AAV to the other tissues of the body. But the size of ARCUS, about 1,500 base pairs, enables us to do really cool things in science -- and you'll hear that from Cassie and Jeff today -- of packaging multiple ARCUS in one AAV in our muscle programs, or packaging an ARCUS plus a DNA template into one AAV for gene insertion. Finally, let's keep it simple. The only gene editing platform today taking advantage of simplicity, critical for our HBV and our mitochondria programs. ARCUS is a single component. It recognizes the binding site and cuts with one enzyme. This allows simplicity of delivery. And Jeff and Cassie will talk to you more about this. More defined outcomes is what the ARCUS advantage translates into. Now let's see how it translates into our development plan. First and foremost, our lead program, HBV. We've spoken to you a lot about HBV. We're very excited, and Cassie's going to share some new data. This is an elimination program taking advantage of simplicity. Our newly announced program, primary mitochondrial myopathy. This is an area where simplicity of being able to get to mitochondrial DNA, mutant versus wild type, is a huge advantage for ARCUS, where guide RNAs cannot do that. You'll hear more about that from Wendy. You see our partnered programs, insertion hemoglobinopathies with Novartis, our DMD lead program with Prevail or Lilly and other undisclosed programs with Lilly focused on insertion CNS tissue. And ultimately, iECURE, our partner spinout of UPenn, who frankly may have the first in vivo ARCUS program CTA approved this year. This is very exciting in an area of OTC deficiency, where newborns unfortunately have toxic ammonia levels that lead to their mortality. Finally, before I hand it off to Dr. Smith, the path forward for Precision has never been more clear: 3 platforms to 1, singular focused on in vivo gene editing, singular focused on sophisticated edits. The others can compete at the liver for knockout. We're going to apply ARCUS where its attributes and advantage enable the best therapeutic index. And through organic development with our HBV programs and now primary mitochondrial myopathy and partnered programs, we will continue to take advantage of sophisticated edits and drive stakeholder value, most importantly for our patients. Now, I'm excited to ask Dr. Smith, our Co-Founder, to join us and go a little deeper on the ARCUS advantages.

Thank you, Michael. Good morning. I'm Jeff Smith. I'm the Chief Research Officer and Co-Founder of Precision BioSciences. I'm also the Co-Inventor of the ARCUS platform. I've been in the field of gene editing for over 27 years, so I know a few things. But for me, the -- my personal connection to gene editing really started in the late 1980s when my grandfather was diagnosed with an early form of genetically-linked macular degeneration. Watching him and his sister, my great aunt, lose their sight, and now watching it impact my mom's vision has always given me a very deep belief in gene editing and the significant impact it can have on people's lives. I ended up doing my thesis at Johns Hopkins, developing the first customizable gene editors called zinc-finger nucleases. I also demonstrated the first in vivo gene editing using zinc-finger nucleases in collaboration with Dana Carroll. Now -- then after I finished at Hopkins, I moved on to Duke University, where I recruited my friend, Co-Inventor, Co-Founder, Dr. Derek Jantz from Johns Hopkins. And together, we set out to build a better gene editor. Now the Holy Grail in gene editing was to stimulate editing using a process called homology-directed repair, or HDR, which Michael just spoke to. Why HDR you might ask? Well, let's think of it this way. Certainly everybody here at some point has had to put something together. Let's use an IKEA dresser as an example. I'm sure during the process of putting it together, you realize that there's probably lots of different ways the pieces can be put together, especially if you're not following the directions. It turns out if you wanted that dresser to function and look like the picture on the box, that you better follow the directions. Biology is the same way. If you want a repair that's functional at the end, then you better follow the instructions. Homology-directed repair is the only natural process that provides a set of instructions in the way of a DNA repair template and actually diligently follows those instructions to get a defined outcome. In contrast, non-homologous end joining can be viewed as putting the dresser together without the instructions. You end up sometimes getting it to function, but you get a lot more random outcomes. Now, the other thing that we figured out that makes homology-directed repair very nice is we can provide our own set of instructions in a DNA template and drive the repair to be the therapeutic outcome that we desire. Sounds pretty great doesn't it? The only problem is that most gene editors on the market today don't drive a very high efficiency of homology-directed repair. This is because most gene editors on the market today were evolved to delete DNA and not insert DNA. We ended up realizing that we need a new foundation for a gene editor. We needed something that was simple, efficient, safe and reliably provided defined outcomes, preferably by homology-directed repair. We looked for a couple years, looking at a variety of different enzymes. And the result was, of all places, we found the answer in pond scum. Green algae called Chlamydomonas reinhardtii of all places. There, we found a homing endonuclease called I-CreI. Now, homing endonuclease are unique in their evolution and that they evolve to safely insert DNA into a genome using homology-directed repair. Now, we at Precision learned how to re-engineer I-CreI while maintaining that natural function to create ARCUS. And we've spent the last 17 years refining it to be able to do sophisticated edits such as gene insertion, elimination and gene excision. Now, dogma in the gene editing field still maintains that most editors generate a very low efficiency of homology-directed repair and that there's no homology-directed repair in non-dividing cells, which is what you would find if you were treating adults. But I'm here to tell you they're wrong or at least they're using the wrong gene editor, because Cassie and I will show you today data showing that ARCUS is able to drive a very high rate of homology-directed repair, and even to our surprise, able to drive homology-directed repair in non-dividing cells. The practical implication of this is that we will be able to with ARCUS treat more people that need those sophisticated edits, like my own family, and be able to enter markets that our competitors cannot. So Michael spoke to you as far as 3 key advantages that ARCUS has that lead to those defined outcomes: the cut, the size, the simplicity. We're going to go a little deeper into the background on each of those, starting with the cut. So Michael ran you through definitions of defined outcome versus random outcomes. And I would like to take the opportunity to show you a diagram to try to illustrate this fact. So on the left, if you had homology-directed repair driving defined outcomes where the majority of your outcomes all look the same and are all therapeutic, you would still, because it's biology, get some outcomes like that bottom deletion that are a little random. In contrast, on the right-hand side, you'll see random outcomes generated largely by non-homologous end joining. In this case, you can still get a few therapeutic outcomes, but you notice that they're fewer and a little more random, plus you have more small, big deletions and random insertions. All of these uncharacterized events bring with it a bigger safety risk in random outcomes. I think it's pretty easy to understand that the therapeutic outcomes and defined -- on the side of defined outcomes has a greater therapeutic effect than the fewer results in the random outcomes. And I think it's easy to understand why we think that patients and regulators would prefer defined outcomes. So now how do you generate more defined outcomes? You start with a nuclease that drives more defined outcomes. So let's look specifically at ARCUS. And what makes ARCUS very special is its cut. So what I'm showing here is, first of all, that ARCUS generates a staggered cut. And what I mean is it cuts the DNA at 2 different spaces to generate 2 single stranded sequences of DNA, shown here in the pink and the blue. Now, first of all, for insertion what's very important is the fact that those overhangs, those single stranded overhangs in pink and blue run in the same direction as DNA replication. Now, it's important for people to understand DNA only replicates in one direction. And this is important because see those pink and those blue single-stranded DNA, they are what find the beginning on the DNA template, on those instructions. And the fact that they run in the same direction as DNA replication allows biology to now read the instructions, to read through that DNA template and read in the insertion. Now, in the case that there is not a DNA repair template, then biology also looked out in the case of ARCUS that -- notice how those pieces fit together in only a very specific way, like 2 puzzle pieces coming together. ARCUS evolved very specifically to do this, because in the absence of a DNA template, it does not generate a scar or a missing piece when it comes back together. Now, Precision is able to take advantage of this by -- if -- when we make 2 cuts in a genome and excise the intervening piece of DNA, as long as we've cut those 2 sites to generate compatible ends so that those 2 puzzle pieces match, they come back together in what we call a perfect re-ligation. And this is very important for giving us more defined outcomes from excision. Now that we understand how ARCUS's cut is special, let's look at how it compares to all the other editors to understand how it's unique. So on the left-hand side, as a reminder, ARCUS's cut. And notice again that staggered cut that leaves that 3 prime, 4 base pair overhang. The DNA replication runs in the same direction as those overhangs. Now, in contrast, let's look at the top far right at Cas9, which is the most common nuclease for CRISPR. In that case, it generates a blunt cut, which is not good either for perfect re-ligation or really for driving homology-directed repair. Instead, those blunt cuts mainly drive that non-homologous end joining, putting the dresser together without the instructions. If you look at the zinc-finger and TALENs as well as the Cas12 right below that, they do make a staggered cut. But if you notice, it's flipped from what ARCUS does. It's a 5 prime overhang. It doesn't read in the same direction as the DNA. So even if those overhangs could find a sequence in a DNA template, they're in the wrong direction and DNA replication can't actually read the instructions. It can't read into the DNA template. So it's not -- it won't stimulate homology-directed repair. And at the bottom, base and prime editors depending on their version can cause 1 or 2 cuts. But in neither case are they actually set up to be able to drive homology-directed repair or perfect re-ligation for large insertions or excisions. Now, speaking of base and prime, it is very interesting that recently there have been some concerns raised as far as generating a double-strand break in DNA. This is very interesting because base and prime, their most efficient forms generate DNA breaks. But DNA breaks really shouldn't be that big of a concern because we get them every day in our normal lives from things as simple as going out into the sun. In fact, there's some estimates that say up towards of 50,000 DNA breaks in all of us at any given moment. So why aren't DNA breaks a bigger concern in our day-to-day lives? Well, it's because Mother Nature has lots of different ways to repair those DNA breaks. So in this diagram at step 1, if you had a DNA break, at step 2 it turns out that there are lots of different types of proteins that can find those breaks and process them. And depending on which exact type finds it, that determines which path the DNA goes through to repair the break. You can almost view those orange Pac-Men in step 2 as the randomizers and the random outcomes that we've been talking about because there are multiple different options and they're all pushing them in different repair paths. So how does ARCUS provide more defined outcomes? It's simple. It's by skipping the randomizing step at step 2. You'll notice that ARCUS generates cuts that are much more similar to what I already showed for ARCUS. Notice the overhangs. This allows the DNA to -- once it finds its instructions in the purple strand, the DNA template, to be able to read in the direction of DNA replication represented by the green arrow to read in the new repair, a defined repair. So let's see how this plays out in the real world. So this is a great experiment in that this is a true head-to-head comparison of ARCUS versus CRISPR. In this case, everything else is the same. The site, the dose, the repair template are all kept identical for exact comparison because only the ARCUS and the CRISPR enzymes are different. And you'll notice that we always see a large difference between ARCUS and CRISPR. In this particular case, in these primary cells, I'm showing you a 17-fold difference of greater insertion for ARCUS over CRISPR. And this experiment really emphasizes the importance of that 3 prime overhang cut that is unique to ARCUS. In this case, on the left, we have a cut that has the ARCUS overhang and showing a good insertion rate. If we take an enzyme called TREX1 and we chew off that 3 prime overhang to make it look blunt, to make it look more like a CRISPR, you see the precipitous drop in our efficiency of insertion, proving that 3 prime overhang is very important. So to summarize, ARCUS makes this unique 3 prime overhang cut, and that drives homology-directed repair as well as perfect re-ligations, both of which lead to more defined outcomes. And as I showed you on the first slide, more defined outcomes leads to a better therapeutic effect and potentially less safety risk. So now let's look at size. So most of our competitors deliver using a lipid nanoparticle, or an LNP, and that's very good for delivery to the liver. But if you want to be able to do gene editing beyond the liver, then you're going to need to use a viral vector. And the most common viral vector is adeno-associated virus, or AAV, which does allow you to go to, example, eye, muscle, central nervous system, as well as other tissues. Ideally, you would want a gene editor that can use both, and ARCUS is one of the few editors that can use both. And let's see why. So what is shown on this next slide is the relative sizes of the different editors compared to ARCUS. The size of each circle represents their size in an expression cassette. And the key thing -- if you see that white dotted line that just flashed up, that's the AAV packaging capacity. Now, if you want to be delivered beyond the liver, then ARCUS needs to -- or sorry, a gene editor needs to fit within that white dotted line. You'll notice that ARCUS fits very comfortably, whereas most of the other gene editors are too large and would not be able to be packaged. The one gene editor that can fit just barely fits. So the result of this is that ARCUS is able to be delivered to a variety of tissues using both editors. And this is a great table showing all the different animal models and deliveries that we have successfully done gene editing with ARCUS, of course the liver with LNP and AAV as well as many others. In particular I want to call out to Precision's knowledge. The central nervous system editing in the non-human primates is the first example of that, that we can find. So we've talked about where you can deliver based off size, but even cooler is what you can deliver with ARCUS. And in this example I'm showing that ARCUS is so small that you can package 2 ARCUSES in one AAV. Now, why is that useful? It allows you to do sophisticated edits that allow you to make cuts at 2 different sites in the genome and excise a piece of DNA and then bring back, as I was already talking about, that perfect re-ligation. Cassie's going to be talking to you today how we're doing exactly this to restore function to the dystrophin gene by pulling out a very large segment of it to treat Duchenne's Muscular Dystrophy. Another very cool aspect is instead of a second ARCUS, you can supply that DNA template, those instructions for what you want to happen to allow you to do gene insertion beyond the liver. And Cassie will also be talking about some of our insertion programs. So finally, simplicity. What am I talking about with Simplicity? I'm talking about the fact of how many components need to be code delivered to the same cell in order to get an edit. In this case, ARCUS is again very unique in that it is the only single component editor, which allows us to have better efficiency, and as I'll show you in a moment, be able to go places some other editors are not. In comparison, let's look at some of the other editors. You'll notice, mainly on this slide, that all of them are multi-component editors, multiple components need to be co-delivered. So what does this look like in an example? So on the left, we already showed you that ARCUS can comfortably fit with the DNA template all in one AAV. And this allows us a very efficient delivery to each cell that gets that, gets everything it needs to do a targeted insertion. In contrast, on the right-hand side, you'll see CRISPR. If you had to deliver the CAS, the DNA template, and the guide RNA, this is going to take at least 2 AAVs, if not 3. And so now you have to try to get multiple AAVs all to the same cells, which you can understand would lead to lower efficiency. You can counter this by raising the concentration of all of your components, but you're also raising the concentration of your delivery vector, in this case AAV, which brings with it its own safety risks. The other very interesting aspect is that ARCUS is able to go into places that other editors cannot. ARCUS in this case is able to get into the Mitochondria. And the Mitochondria generate most of the energy for your body. They have DNA, which is pretty cool. And Wendy will be talking to you later as far as how we're using this to fix problems in mitochondria to treat diseases. Now, a guide RNA cannot actually get into the mitochondria. So any CRISPR-based technology, such as CRISPR base or Prime Editors cannot actually compete in this market. So to sum it up, the cut, ARCUS generates that unique 3 prime overhang, which drives homology-directed repair in the presence of a DNA template and Perfect Re-ligation in the absence to give you large insertions and large excisions with defined outcomes. In contrast, CRISPR runs -- makes a blunt end cut and causes non-homologous end joining in both cases. Now Base and Prime, neither one are actually set up to do large insertions or excisions. In terms of size, again, ARCUS is the smallest gene editor, which gives us the advantage that we can fit 2 nucleases or a nuclease and repair into the DNA template, allowing us to do sophisticated edits beyond the liver. And as you saw, the other competitors cannot. Finally, Simplicity. We are the only single component editor which lowers our dose and keeps us at a high efficiency compared to the complexity of some of the other competitors. All of these things add up together to provide more defined outcomes. And with that, I'm going to hand it over to Dr. Cassie Gorsuch, our VP of Gene Therapy Discovery, who will run you through our internal and external programs to show you how The Cut, The Size, The Simplicity are being taken advantage of in our programs.

Hi. I'm Cassie Gorsuch. I have the pleasure of heading up our In Vivo Gene Editing Research Group. And it's really a privilege to be here with you today to share some of the progress that our team has made. As Jeff mentioned, I'm going to walk through how we are applying these ARCUS advantages in the context of our therapeutic programs. I'm going to going to start with the cut and why that matters for insertion. When you think about gene insertion, it's important to understand that efficiencies and outcomes are context dependent and ARCUS is efficient at insertion across a variety of different contexts. I'm going to walk through these different contexts and explain how and why there are different challenges associated with each of them. So first, Dividing Cells in Culture. In culture, delivery is off the table. It's pretty easy to deliver the gene editing components and the gene insertion template, the DNA template in culture. Dividing cells more readily, rapidly, more easily take up DNA to insert them onto the chromosome. So this represents sort of the easiest context for a gene insertion event. Moving into a slightly more challenging context would be still in culture, so no concerns on delivery, but now non-dividing cells. As Jeff mentioned, and we'll talk a little bit more in the future, non-dividing cells are a challenge and the field would have you believe it's really impossible to insert in this context by homology-directed repair. Moving into the In Vivo context, now we have to get the components where we actually want to insert them. So delivery becomes a big challenge. But the In Vivo context follows the same paradigm as the In Vitro context where dividing cells are going to be slightly easier than non-dividing cells. And these 2 contexts are important when considering patient populations that you could impact. So dividing cells represents infants or children where their tissues are still growing and dividing versus non-dividing cells which would represent more of an adult patient population. Being able to insert into either of these contexts is really important when thinking about the types of patients that you could treat. I'm going to show you examples where ARCUS is not only efficient at insertion in each of these contexts, but provides the defined outcome that we're looking for across this entire spectrum of difficulty. So first, starting with that easy context of dividing cells in culture. In this experiment, we delivered a DNA template along with an ARCUS Nucleus, and we achieved 76% gene insertion, which is really exciting. But we've heard a lot about efficiency as the starting place. How did you get there? What was the mechanism by that by which you achieved that efficiency? When we drill a little deeper we see that 92% of cells that achieved an insertion did so by a homology-directed repair. This was a really exciting result because we know that driving that insertion through homology-directed repair can give us the higher frequency of a defined outcome and a therapeutic impact. Let's talk about that therapeutic impact a little bit more. One other really important consideration for gene insertion is being able to insert your desired gene into both copies of the target allele. What that means is every person has two copies of most genes within their chromosomes. And so if you're trying to insert a gene into that target gene, it would be best if you could insert into both copies. What we found in this particular experiment was that ARCUS does that at a really high rate. So 91% of cells that achieved an insertion did so at both copies. We call that Bi-allelic editing. This matters whenever you think about that therapeutic outcome. So shown on the right here is a Mono-allelic editing, meaning an insertion in one allele versus Bi-allelic editing, which is insertion in both alleles. You can see that Bi-allelic editing results in a much higher level of therapeutic gene induction, which really just means we turned on the gene that we were trying to turn on, our therapeutic outcome, our defined outcome. So together, these data really suggest that ARCUS inserts efficiently through the HDR mechanism in a Bi-allelic way and gives rise to this therapeutic outcome that was desired. Before we move in to the next context, our non-dividing cells, I want to level set based on some of the things you've heard around what the field thought was possible in this context. You can see across a variety of different sources from published literature to patent filings that the field is pretty well aligned that the ability to insert with high efficiency is low and that the ability to do that through homology-directed repair is basically impossible. Now much of this literature is based on the fact that most gene editors create blunt cuts. And so we hypothesized that perhaps ARCUS could get around these restrictions or limitations because of the unique 3 prime overhang cut that it uniquely creates. So to test that hypothesis, we conducted an experiment in primary human hepatocytes. So these cells don't divide in culture, so they represent that non-dividing context that I was talking about. In this particular experiment, we delivered a DNA template for the insertion and the ARCUS nucleus and we achieved almost 30% gene insertion. Again, super exciting, but that's just the first step. How did we get there? When we look at the mechanism by which those insertions occurred, we found that 57% of them, more than half, were by this homology-directed repair pathway. Just to remind you, this was thought to be impossible. So this is really exciting data that the team collected. And we think this is really attributable to that unique 3 prime overhang, bypassing the limitations, that randomizer step that Jeff talked about. Moving into the in vivo context, we wanted to conduct an experiment to demonstrate efficiency of insertion in dividing cells in vivo, but also demonstrate long-term durability of those insertions. So this experiment was conducted in infant non-human primates in partnership with our partners at iECURE for the OTC Deficiency Program. In this particular study, non-human primates were dosed with 2 AAVs, one carrying an ARCUS nucleus and one carrying the DNA template. Those AAVs were co-administered and then we measured gene insertion efficiency at 3 months after dosing and out to 1 year after dosing. And excitedly, we found that that gene insertion efficiency was sustained or potentially even slightly increased over those 2 time points out to that 1 year mark. What I think is really exciting on this slide is looking at the right hand. Each of those blue dots represents a different cell in the liver of these animals. The cells that are also stained in red represent cells that achieved that gene insertion and are now expressing that OTC gene. And what's neat is you can see these clusters of cells, these pockets of red cells. And what that's really telling us is that gene insertion happens in a cell and as that animal was continuing to grow and the cells continuing to divide, we see that cluster start to expand. So this data really demonstrates that the insertion efficiency is a very high and importantly, it's maintained over the long term. So we think that this really highlights the durability of a gene editing approach for gene insertion versus a standard conventional gene therapy approach. Now, finally, this is the most challenging context I'm going to talk through. So this is insertion in non-human primates in non-dividing cells. We were extremely excited when we got this data and saw that we achieved up to 45% gene insertion efficiency. So this particular experiment utilized an ARCUS LNP and a DNA template delivered by AAV in non-human primates. And just to level set around this level of gene insertion, when we look at others taking a similar approach, but using a CRISPR based editing system, we find that these levels of gene insertion are actually 3 times higher than what's been achieved with a CRISPR-based approach. And again, we think that this is really attributable to those 3 prime overhangs that stimulate homology-directed repair and lead to high therapeutic outcomes. So in summary, across an entire spectrum of difficulty where context really matters, we've demonstrated that ARCUS is an ideal gene editor for therapeutic gene insertion and this is achieved by high efficiency gene insertion rates, high homology directed repair rates, and the ability to edit in a Bi-allelic way. And all of these attributes contribute to a defined outcome leading to a therapeutic impact. Here's how we're leveraging this attribute of ARCUS in terms of our therapeutic programs. You can see we have a number of insertion programs in partnership with Novartis for hemoglobinopathies for insertion in hematopoietic stem cells, with Prevail for an insertion program in the liver, and then of course with iECURE for our OTC deficiency program, which we expect that they could file CTA as early as by the end of this year. We're really excited about all of these advantages that we have found in terms of ARCUS's ability to insert with high efficiency through HDR in dividing and non-dividing cells. We've demonstrated that the durability of those edits can be persistent over the long term in non-human primates and that the efficiency is greater than 3 times what's been reported with a CRISPR-based approach. And finally, this insertion ability, the ability to insert a gene long-term really provides a lot of value for patients and the ability to provide a one-time treatment that's durable over the long-term. So now, I'm going to pivot into our next ARCUS advantage of size and Jeff mentioned we're really leveraging the -- this attribute of ARCUS in the context of gene excision. So this is just a reminder when we say size matters, what does that mean? Size matters, when you're thinking about delivery of a gene editor outside of the liver and today the delivery vehicles are really limited to LNP in the liver, but if you want to go anywhere else in the body, you really need to consider using an adeno-associated virus. The challenge with AAVs is that their packaging capacity is pretty limited. And so the size of your editor matters if you want to put it into an AAV. Jeff already told you this, but just as a reminder, an ARCUS nucleus is one of the smallest gene editors, so small that we're actually able to package 2 nucleuses and 1 AAV. We're using this advantage of ARCUS in the context of our program, partnered with Prevail for Duchenne Muscular Dystrophy. The therapeutic approach here is to deliver 2 ARCUS nucleases in 1 AAV vector. Each of those ARCUS nucleases recognizes a unique target site in the introns surrounding Exons 45 to 55. This is a really critical region in Duchenne Muscular Dystrophy patients because mutations often occur in this region, so up to 50% of patients have disease-causing mutations in this hot spot region of the gene. So if you could remove that region of the gene, you could really impact a huge number of DMD patients. So the goal is to deliver these 2 nucleases to create the 3 prime overhang cut at each site, excise that Exon 45 to 55 region, and allow those 2 overhang cuts to come back together perfectly, resulting in a perfect re-ligation that will restore the dystrophin reading frame and result in an ARCUS edited dystrophin protein. So looking at the dystrophin protein. It's important to realize this is a really big protein. It has a lot of important protein functional domains. And all of these functional domains are really essential for proper function of this protein within the muscle. So at the top, each of these different circles or ovals represents a different protein domain in the dystrophin protein. There was a recently approved drug from Sarepta using a micro-dystrophin approach that showed really exciting promising results in the clinic for these DMD patients. What you'll notice is that in order to fit that micro-dystrophin into an AAV vector, the majority of the natural protein had to be removed. When we look at the resulting ARCUS-edited dystrophin protein, you can see especially compared to the Sarepta product, a number of functional domains are actually maintained in our approach comparatively. We think that this could provide a really exciting advantage using this ARCUS-edited dystrophin approach because we preserve the majority of the natural protein domains, thereby hopefully preserving even more function of the resulting protein that is expressed. So that's our therapeutic strategy, our approach for DMD. Let's get into the data now. So to test the ability of that ARCUS edited protein to restore function in the context of DMD, we completed a functional mouse study. In this particular study, we delivered a single AAV with our 2 ARCUS nucleases, and then measured a number of different endpoints to determine if this editing approach could restore function in this mouse model of DMD. We compared treated mice to untreated mice, untreated diseased mice, as well as untreated control healthy mice. Looking first at the ability of those ARCUS nucleases to create the excision event at the genomic level, we measured the ability -- we measured excision within the quad of these mice. You can see in this particular example, we achieved about 15% gene excision, and this is an over 500,000 base pair region of the genome. So it still kind of blows my mind that this is even possible up to this for this frequency. When we look a little deeper at how those excision events came back together, so we talked about how those 3 prime overhangs can find each other perfectly, what we found is that in -- 72% of genomes where the excision occurred, those 2 overhangs came back together seamlessly, perfect re-ligation. And this is important because it means that the variety of outcomes is smaller when most of them come back together perfectly. So the number of outcomes that you need to characterize is a much shorter list when 72% of them are achieved by a perfect re-ligation. When we've looked in the field and seen others taking a similar approach for an Exon 45 to 55 deletion using a CRISPR based approach, most of the data that's been reported shows up to about a 30% perfect re-ligation. And so this really suggests that those 3 prime overhangs are driving that perfect re-ligation at much more substantial frequency of the time using an ARCUS-based approach. We next looked at the ability of those edited genomes to actually express the dystrophin protein. And here I'm showing a variety of different tissues where we measured that ARCUS-edited dystrophin protein. You can see in every case, whether it's skeletal muscle, heart or diaphragm, we see that the editing actually resulted in the restoration of this dystrophin protein. It's really important to note that we were able to achieve high levels of dystrophin protein expression in both the heart and diaphragm. These are really key tissues when thinking about this patient population because this is how patients die, through cardiac failure or respiratory failure. So the ability to reach these really hard to reach tissues is essential for providing a quality of life for these patients. We next wanted to test whether this dystrophin protein was going to be functional in these animals. It's not the native dystrophin protein, so is this particular version of dystrophin going to provide a functional outcome. We did this in the calf, and I want to point that out on this slide because if you'll notice, the calf is actually the lowest of all of our tissues in terms of dystrophin restoration. So this represents the highest hurdle to demonstrate a force or therapeutic outcome. Here I'm showing you the results of that force measurement. In blue, the blue line represents the treated animals, and you can see a significant increase in the ability of these mice to generate a force output in the calf compared to the untreated diseased animals shown in pink. What was really exciting about this data was that we found that the ARCUS treated animals achieved 86% maximum force output levels compared to the healthy animals shown in gray. So this was a really exciting result that demonstrated that we could achieve really broad delivery using a single AAV to deliver 2 nucleases. We could excise a huge region of the dystrophin gene, resulting in a nice protein output that was functional. Most importantly, in my mind, is this slide. When you think about durability of a gene editing approach for DMD patients, you need to think about the ability to edit muscle satellite stem cells. These cells are the cells within the tissue that give rise to all new muscle. So after you exercise or if you have a muscle injury, muscle cells will die and they need to be repopulated and that's done so by muscle satellite stem cells. In this particular image muscle satellite stem cells are indicated by that purple marker for PAX7. It's a rare cell type, so you won't see very many cells on this slide that actually show that purple stain. The blue stain indicates the editing event after ARCUS nucleases have cut out that region. So what you can see in this highlighted cell, the circled cell, is this is a muscle satellite stem cell that has been edited with our 2 ARCUS nucleases and is now expressing that edited version of the dystrophin transcript. So this is evidence that we can edit in muscle satellite stem cells and really provides a lot of excitement and enthusiasm that the editing event and the functional outcome that we've achieved here could be durable over the long term. So we're really excited by this data that we've collected in partnership with Prevail for this program. It's a really exceptional demonstration of our ability to leverage the size of ARCUS to put 2 nucleases in 1 AAV, deliver that to a variety of different tissues within the body, and exhibit this really meaningful therapeutic outcome. As I mentioned, my favorite part of this data set is really that satellite stem cell editing that has been tough for the field to really demonstrate and really provides hope that this approach could be durable for patients. Okay. So the last section that I'm going to talk through in terms of our ARCUS advantages and how we're leveraging them in the context of our therapeutic programs is Simplicity. When we talk about Simplicity, just as a reminder of what Jeff and Michael already told you, we mean that ARCUS is the only gene editor that is a single component. It has the ability to recognize the DNA target site that we've engineered it to, and cut that target site all within one component. No other editing platform can do that. What that means for us is that this will streamline delivery. There's no need to co-package multiple components, so streamlines delivery could lead to a higher efficiency overall. I want to talk first about how we're leveraging this unique attribute of ARCUS in the context of our hepatitis B program. Hepatitis B is a leading cause of morbidity across the entire world. It's a really challenging global disease burden. There are more than 1 million patients in the United States and more than 300 million globally. The real problem here is that chronic hepatitis B leads to long-term liver complications, serious liver complications, including cirrhosis, liver failure and liver cancer. And unfortunately, a lot of these liver complications often result in death for chronic hepatitis B patients. The sad truth is that the current standard of care today almost never leads to a functional cure. Patients are required to maintain daily nucleoside, nucleotide analogs, and that's good at suppressing the viral antigens, but it almost never leads to a functional cure. So what do we mean by functional cure? So functional cure is a field-aligned goal for what we would like to achieve by 2030 in the HBV field. And it's our goal too, at Precision. So a functional cure is defined as sustained, undetectable, circulating HBsAg and HBV DNA off treatment. And that's the really important thing that's been really hard to achieve in the field is allowing patients to come off their therapy while maintaining, eliminating the virus. We're really excited to offer a gene editing approach as a really unique differentiated approach towards an HBV elimination strategy. We think that this provides the highest chance towards that functional cure. And many of our KOLs agree with us including Dr. Sulkowski who we'll hear from in a little bit. So before we get into the data, let's talk a little bit about the HBV viral life cycle. So hepatitis B virus infects hepatocytes in your liver. Once the liver cell has been infected, the virus sets up shop in the nucleus and forms two different viral reservoirs. The first one is cccDNA or covalently closed circular DNA. This mini chromosome really persists within the nucleus for a long time within these chronic hepatitis B patients and exists outside of the host chromosome. What's challenging with HBV is it can also randomly integrate into the host chromosome. So that's the second viral reservoir that can form this integrated HBV DNA. Both of these 2 viral reservoirs, cccDNA and integrated HBV DNA, give rise to the expression of HBV RNAs that can be translated and secreted in the form of HBV DNA and S-antigen. And those are those two important biomarkers that I mentioned earlier that we think about when we think about a functional cure. So our approach using an ARCUS editor towards a functional cure is to deliver an ARCUS nucleus by mRNA that targets the HBV sequence. Once inside the cytoplasm, the mRNA is translated into protein, and that protein can bind cccDNA and integrated DNA. And after the cut occurs in cccDNA and integrated HBV DNA, this will result in the loss of HBV RNAs, HBV DNA and secreted S-antigen. This approach is really differentiated from a lot of the antiviral approach that are currently being pursued clinically today. You can see in this schematic, there's a number of different drugs that have been developed that target different parts of that viral life cycle, but really it's only ARCUS and gene editing approaches that go after the heart of the disease at the cccDNA and the integrated HBV DNA. Even amongst other gene editing approaches being explored today for HBV, ARCUS is unique and that it is the only one that will result in elimination of cccDNA after cutting. When we think about selecting that ARCUS target site, we had to make a couple really important considerations always with the patient in mind. What you're seeing on this slide is a schematic of the cccDNA genome. We selected an ARCUS target site that falls within the cccDNA in a region that is highly conserved across different types of HBV genomes. So like many viruses, HBV can often mutate. There's a number of different genotypes, subgenotypes. They're geographically localized, and we wanted to ensure that our approach could be accessible to the highest number of patients possible. So we selected this target site with that in mind. It needed to be conserved in the vast majority of patients. The other important consideration when selecting a target site for our ARCUS Nucleus was integrated HBV DNA. We know that integrated HBV DNA is the primary source of S-antigen expression in many patients. And so to turn off or turn down that S-antigen expression, you need to be able to edit with an integrated HBV DNA. The challenge here is that when HBV integrates onto the host chromosome, it doesn't do so in the same way almost ever. And so what you're seeing here is a number of different sequences that have been sequenced from patient biopsies. And you'll notice the regions of the HBV sequence that have actually been contained within those integrations is different all over the place. We selected an ARCUS target site that was conserved in the vast majority of integrations and would likely turn off S-antigen expression. So this was our 2 important considerations, must target cccDNA and integrated HBV DNA in the vast majority of patients. So what actually happens after you cut cccDNA or integrated HBV DNA? I want to walk through quickly the molecular mechanisms that occur within the cell and how those can lead to therapeutic outcomes. So looking first at cccDNA, when you cut cccDNA, there are two potential editing outcomes that lead to a therapeutic effect. The first one would be elimination, complete loss of the cccDNA. The other potential editing outcome would be what we call an Indel or a small insertion or deletion that happens within the cccDNA. We found that both of these editing events actually leads to loss of HBV DNA and S-antigen. In the context of integrated HBV DNA, which only secretes S-antigen, not HBV DNA, after a double-stranded break occurs, this is repaired through this Indel formation, small insertions and deletions and we found that this also leads to S-antigen loss. Okay. So it's a lot. HBV is a complicated indication. We're really excited about our therapeutic approach and how it's different. Now, let me show you what it actually looks like in practice in the data. We first tested our ARCUS editing approach in the gold standard model within the field, which is HBV-infected primary human hepatocytes. We found that after we administered this ARCUS mRNA in these HBV infected PHHs, 85% of the cccDNA was completely eliminated. Of the 15% remaining cccDNA, we found that 30% of it had now been mutated to contain small insertions and deletions. And this represents about a 90% overall editing efficiency in PHHs. So editing efficiency was very good, very high. But what does that do towards that therapeutic outcome? What does it look like for defined outcomes and therapeutic effect? We find that, that 90% editing efficiency led to an 80% sustained loss of HBV DNA and about a 77% loss of secreted S-antigen. Importantly, both of these endpoints -- the reduction in these endpoints was maintained long-term long after all of the editing had occurred and the ARCUS protein was no longer present. We developed a model that had integrated HBV DNA that secreted S-antigen so that we could test specifically the ability to impact integrated HBV DNA. In this model, we found that we were able to achieve 87% editing efficiency, and that led to a sustained 80% reduction in S-antigen. So together, these results really drive home this two-pronged approach for our ARCUS therapeutic strategy of eliminating cccDNA and then activating integrated HBV DNA, leading to sustained reductions in these viral antigens. One of the big challenges in the HBV field that is a challenge for everyone here is that there are a lack of good in vivo animal models. And this is really due to the fact that HBV that infects humans, the virus that infects humans doesn't really infect a lot of other species, including mice, in most research non-human primates. So to overcome this challenge, we got creative. What we did is we utilized a virus that we know very well, adeno-associated virus, to help mimic an HBV infection. We know that AAV can transduce across species. We know that AAV sets up shop very similarly in that episomal DNA way within the same cells that HBV infects. And we know that we can replace a part of the AAV genome with the HBV genome to actually give us a target site to monitor and that S-antigen biomarker. So what we did is we encoded part of the HBV genome into an AAV vector and then administered that like an HBV infection into these animal models. We first did this in mice. And after we treated these AAV infected mice with an ARCUS LNP, we found a 40% reduction in viral copy numbers. So this represents that cccDNA elimination editing outcome. Of the 60% remaining viral DNA, we found that 86% of it contained mutations. Both of these editing outcomes combined led to a 96% reduction in secreted S-antigen. So what you're looking at in this graph is a time course of the entire mouse study. The LNP ARCUS was dosed at week 3 where that arrow is indicating. You can see that really rapid steep decline in S-antigen by week 4. And all of the editing actually happens between week 3 and 4. And by week 4, there's no more ARCUS protein present. What's really exciting to see is that, that S-antigen suppression 96% was maintained for 4 weeks after all of the editing was complete, out to week 7, which was when we took the animals down. With the success that we saw in the mouse model, we decided to test the same episomal editing strategy in the context of a non-human primate, and using a non-human primate allowed us to really determine the translatability of our ARCUS editing approach with that LNP. So here we found after administration of an LNP containing the ARCUS nuclease, we saw a 66% reduction in viral DNA. So that's the elimination side of the editing. In the remaining 34% of viral DNA, 44% of it had contained indels. And together, this represents greater than 80% overall editing efficiency in a non-human primate model. So 96% S-antigen reduction and 80% overall editing in a primate are pretty good. But we know the bar is going to be really high for HBV. Our goal is elimination and functional cure. So with that in mind, the team has really put in a ton of effort around mRNA optimization to ensure if we can achieve the highest level of potency possible through our LNP approach. So what you're looking at here in the graph is the left bar represents the mRNA that we've been previously using in all of the studies that I've shown you so far. Through a whole bunch of work from a lot of different people on the team, we've been able to achieve an 8x increase in overall ARCUS protein expression from the same LNP dose. So this means that we now have much bigger bang for our buck in terms of the amount of LNP that we deliver and we think that, that could be a really beneficial optimization strategy for improving efficiency. So we've implemented this strategy into our HBV program and generated a lot of exciting data that we'll be sharing later this year at a scientific conference. So I might be biased. I worked on this program since I started at Precision. I really believe in our approach for HBV. I think we're really differentiated because we are targeting both cccDNA for elimination and an activating HBV DNA. And we've demonstrated that, that two-pronged approach can lead to really drastic decreases, sustained decreases in both HBV DNA and S-antigen. The team has put forward a lot of effort to improve these -- our mRNA strategy to ensure that we're delivering the best patient -- the best product for patients. And a number of people are working literally day and night to continue to push this program forward with an expected CTA and an IND in 2024. With that, I'd like to hand it off to my colleague, Dr. Alan List, our Chief Medical Officer, for a conversation with Mark Sulkowski on our hepatitis B program.

Okay. Good morning, everybody. Thank you for joining us today. Joining me is Dr. Mark Sulkowski, who is the Chief of Division of Infectious Disease at John Hopkins University School of Medicine and also an internationally recognized expert in HBV. So Mark, thank you for joining us today. I'd like to begin by addressing current HBV therapeutics. Clearly, there has been tremendous development over the last decade. Could you begin by briefly describing the current status of treatment and why a functional cure is so elusive?

Sure. Thanks, Alan. First of all, it's great to be here to talk about a truly innovative approach to hepatitis B cure. I've been treating patients with hepatitis B at John Hopkins for the last 20 years and using the same therapy. These are polymerase inhibitors that target a late stream. And I tell patients, "Maybe you'll get a functional cure, but it's unlikely and we can't control that." I could probably count on 2 hands the number of patients I've treated that have achieved that outcome in the time I've treated them. So of course, that's led to a lot of excitement in terms of developing new therapies. There's an unmet need. We're curing very few of the nearly 300 million people. But most of the approaches that we've taken thus far have been downstream. I think it's very important to go back to that lifecycle that Cassie presented. We've got 2 issues. One is this DNA template, the cccDNA, that when actively transcribed produces intact virions and a tremendous amount of surface antigen. The second issue is the integration which produces surface antigen, and that's thought to play a role both as an immune decoy by flooding the immune system with the target S-antigen, but also potentially in the pathogenesis of parietal cell carcinoma, which is the big threat to patients. So as we look for new therapies, we've targeted the viral lifecycle. I mentioned polymerase. They work downstream in terms of the reverse transcription of RNA to DNA. We've looked at capsid inhibitors. They're also downstream in the replication cycle. And none of them actually interfere with the integrated component. And even silencing. While we can silence that, that's temporary with siRNAs. So really to have a meaningful approach that could render patients cured and not at risk for clinical outcomes would be important. And I think we need to target the DNA, which is what this virus is.

Perfect. That's a good segue into our next question. Could you elaborate on the differences in the biologic relevance of the HBV covalently closed circular DNA and the genomically integrated DNA, and importantly, the need to eliminate that cccDNA?

Sure. I think it's really the heart of the problem. So this covalently closed circular DNA is really a very stable -- I think Cassie referred to it, and many of us, to a mini chromosome. This is the template for viral replication with transcription and then translation of the pre-genomic RNA. As I mentioned, this is a very protected component of the virus. Nothing we're using both in the clinic today or really in clinical trials is getting at that cccDNA, but yet that is the problem. The second element -- so if we could target that and eliminate transcription at the source would be a major advantage. We would not see intact virions. We would not see proteins in the blood like the surface antigen. But even if we just went after cccDNA, we still have this integrated. Now that's not replication competent virus, but this is pieces of hepatitis B DNA that produce surface antigen. Now, perhaps we could leave that alone, but there's several problems. One is that integration events are thought to be part of the cancer hepatocellular carcinoma pathogenesis. So by targeting that, we can reduce the likelihood, hopefully, of having further disease. So it's important to target both of those. It's important to go after the source of the problem, which is really the cccDNA.

Okay. Cassie, anything you'd like to add?

No, I think that -- as you mentioned, I think there was a really important finding in the field that those integrants really are disease contributing and your therapy really does need to address that. You can't leave that off the table. And so that was an important consideration for us. As I mentioned, in selecting our target site, we knew we needed to eliminate cccDNA. But we also learned, as the field learned, that you have to inactivate that integrated DNA as well.

Perfect. Let's move on and talk about the ARCUS strategy. So how is the strategy by ARCUS to eliminate the DNA, the cccDNA, how does that compare to the other gene editing platforms that are out there to achieve a functional cure?

Well, I think there's several components to it. You've heard a lot about the elements of the ARCUS approach that make it important. I think one of them is that it's important to recognize that the liver at different stages of the virus, a natural history for a patient, many hepatocytes can be infected. So in the early stages of infection where patients have very high levels of virus in the blood -- we're talking about billions of international units -- when we look in liver biopsy, we see that 95% of liver cells are infected. At later stages, we have more integration events. So when you look at a patient who may have lived decades with the virus, they now have perhaps fewer hepatocytes, but more integration. So we need to target both of that. And that efficiency in delivery to a -- really getting into the source of all those infected hepatocytes is important. And then it's also important when one thinks about targeting integrated DNA, which is integrated into human DNA. We need to be sure that when we cut, that we've got that HDR that we talk about, that repair that's really important for those patients. So by targeting all the stages of the lifecycle, both in early and late and targeting both forms of DNA, I think is critically important. ARCUS, as we've heard this morning, really lends itself to addressing both of those areas.

Is there any other editing -- gene editing platform that's out there that's trying to address both?

So certainly, that's one of the unique advantages. As I mentioned earlier, you could make the argument to go after just cccDNA. But that's a problem, because when one looks at the virus lifecycle, these integration events are part of it and it's part of the disease pathogenesis. So clearly targeting both is important.

Okay. My final question, and that is what would eliminating HBV, the virus completely and delivering a functional cure mean for the patient?

So I mentioned my 20 years or more in practice for treating people with hepatitis B, and I can tell you every one of those people wants to achieve a functional cure. Those who are taking oral daily therapies for life, they do that in hopes of preventing liver cancer, liver cirrhosis outcomes. But they live with a burden and a stigma that comes with hepatitis B and they all want to be cured. Every appointment I have we talk about what's in the pipeline. To deliver a functional cure to people with hepatitis B delivers not only a medical outcome, that is they have less risk of liver cancer, which hangs over their heads as they think forward. Many people have said that "I think I may get cancer someday. My father did, my brother did." And then it removes that stigma that occurs of living with that and takes away the burden -- both psychologically and for societal the stigma of the infection. So a functional cure is an extraordinary meaningful outcome for people with this infection and it's one, as I mentioned earlier, that our current strategies deliver infrequently. And I tell patients, "I can't promise you success with our current polymerase inhibitors."

Right. We have a little bit of time left, so maybe you could just comment. I think everyone has wondered can you eliminate all virus. Or probably the better question is, do you have to eliminate it? What can you expect of the immune response when you have deep suppression of the virus?

It's really a great point. And I think it is worth recognizing that patients with hepatitis B spend their entire lives with their immune system fighting the virus. And there's an interesting phenomenon. When you look at people infected as adults, let's say a 25 year old who acquires hep B, their immune system largely can render that silent and deliver a functional cure. Yet many of the 300 million people you heard about were infected at birth through mother to child transmission and their immune systems don't recognize that pathogen well. So the strategy for functional cure really includes knocking down as much of the infection as you can and then understanding that the immune system does have the ability to clean up the rest of the infection and can be very effective. So I think that in a sense if we can control most of the infected hepatocytes, reducing the S-antigen, reducing other viral proteins, we would anticipate that the immune system would have a greater advantage. We've now turned the tides in favor of the immune system.

Perfect. Well, that completes our time. But I want to thank you again, Mark, for joining us today for this discussion. And it's my pleasure to introduce next Dr. Wendy Shoop, who is going to be joining us by video in discussing our mitochondrial disease platform.

Hi. My name is Wendy Shoop and I am the Research Lead for our mitochondrial programs here at Precision. Today, I'm going to be talking to you about the third key attribute of ARCUS, which is simplicity. The project that I'm going to tell you about today was actually my PhD project that I just finished a couple of months ago. I can't tell you how many people have said to me after they finished their PhDs that their projects went and sat on a shelf somewhere and nothing ever came of them. And it is just incredibly humbling to get to take this project that I've worked on for the last several years and now be talking about treating patients. So actual real living people who are out there right now suffering, and we have an opportunity to greatly improve their quality of life. So I hope that you will share my excitement for this program as we go through the data. So as Jeff has already mentioned, ARCUS is the only gene editing technology which is a single component editor. And what that means is that both the DNA recognition and double-stranded break generation components are included in that single protein. This is unique from other technologies like CRISPR or base editors or prime editors which have multiple components. And we can leverage the simplicity of the technology in pretty unique ways. Specifically, we're able to go places in the cell that other editors like CRISPR cannot access. So we're utilizing the simplicity of the technology to target a group of diseases called mitochondrial diseases. As a group, these affect about 1 in 4,300 individuals, so are actually incredibly common. In fact, they are the most common hereditary metabolic disorder. We are specifically targeting a subset of these diseases, which include primary mitochondrial myopathies, or PMM. PMM is characterized by defects in energy production in the skeletal muscle, which is a really high energy demand tissue. And so, this imbalance between energy supply and energy demand ultimately leads to a fairly severe quality of life impact for these patients. So they experienced daily fatigue, muscle weakness and exercise intolerance that greatly impacts their quality of life. Unfortunately, there are no cures currently available for these diseases. However, we have the opportunity to utilize ARCUS as a transformative treatment option for these patients. As you may have guessed from the name, mitochondrial diseases are due to defects in these organelles called mitochondria. And if you remember back to eighth grade biology, the mitochondria is what's known as the powerhouse of the cell. And what that means is that mitochondria are what generate the majority of the cell's energy. And that energy is what we need to do everything. So for me to stand here right now and walk and talk and breathe, that all requires energy. Mitochondria are present abundantly throughout the cell and each mitochondria has many copies of its own genome, which we call the mitochondrial DNA. So this DNA that's included in mitochondria is completely separate from the DNA that's in the nucleus. And the genes that are included in mitochondrial DNA are crucial for the production of energy. And so, if there are mutations in this DNA, it can greatly impact the cell's ability to generate enough energy, which results in disease. Here's a schematic of the human mitochondrial DNA. Like nuclear DNA it is a double-stranded molecule. However, unlike nuclear DNA, which is arranged in these linear chromosomes, the mitochondrial DNA is actually circular. There are lots of different mutations throughout the mitochondrial DNA that are implicated in mitochondrial disease. Most of these mutations are point mutations. So there's only that single nucleotide that differs between the mutant and the wild type sequence. Our first indication that we're going after is caused by the 3243 A to G point mutation. We selected 3243 as our first target due to the overall prevalence of this mutation. So the estimated population prevalence is about 1 in 500 individuals. And about 1/3 of all mitochondrial diseases, which includes PMM, are attributed to 3243. As I mentioned, 3243 is a point mutation. So there's only the single nucleotide that differs between the sequence that we want to cut, which is the mutant, and the sequence that we definitely don't want to cut, which is the wild type. And so you can see on the slide is the 22 base pair ARCUS recognition sequence around 3243 and that 3243 position is highlighted in either red or blue, with the red being the mutant and the blue being the wild type. So I hope you can appreciate that in order to specifically target the mutant sequence, we need a nuclease or an enzyme that is really, really specific. And so we're going to talk a lot today about specificity as this is a really important component of this project. So the multi copy nature of the genome means that these mutations often exist in the state known as heteroplasmy. And what that means is that both the mutant and the wild type molecules are present in the same cell. The presence of the wild type molecules can buffer and offset the impact of the mutant ones until a certain disease threshold is reached. And so on the left-hand side of the screen, you can see 3 different mitochondria that all have different levels of heteroplasmy. However, in this example, the heteroplasmy here all falls below that disease threshold, and thus, none of these mitochondria exhibit any impairments in energy production, and therefore, are clinically asymptomatic. However, once the percentage of the mutation exceeds whatever that threshold is, that's when you start to see defects in energy production, and thus, clinical symptoms. There are a couple of really important differences between editing nuclear DNA like we've been talking about up until this point and editing mitochondrial DNA. And importantly is what happens within each organelle following a double-strand break within the genome. And so like Jeff has been talking about, nuclear DNA has lots of different double-stranded break repair mechanisms that can act to repair that break. And so we can leverage the double-stranded break repair to generate intended edits within the nuclear genome, such as an indel or an insertion or an excision. However, in mitochondria, there's actually no efficient double-stranded break repair mechanisms. And so there's really only one outcome that can occur following a double-stranded break within the genome and that is the elimination of the molecules that were cut. Mitochondria, though, don't like to suddenly lose copies of mitochondrial DNA and they actually work really hard to maintain a fairly steady state mitochondrial DNA copy number. So what this means is that if there's a sudden depletion in mitochondrial DNA, the cell will replicate whatever molecules are remaining in order to bring the copy number back up to that steady state level. So our therapeutic approach here hinges on those 2 key aspects of mitochondrial biology: there's no efficient double-stranded break repair and mitochondrial DNA copy number is maintained. And so our strategy here is to design an ARCUS nuclease that is specific for the 3243 gene mutation. We can deliver that nuclease to affected cells using an AAV vector and then allow the translation of that protein. At this point, though, ARCUS is present in the cytoplasm and it needs to get in to where the mitochondrial DNA is located, which is in the matrix of the mitochondria or that innermost compartment. So in order to get ARCUS into the mitochondrial matrix, we put this mitochondrial targeting sequence, or MTS, at the beginning of the ARCUS protein. This MTS is just a short peptide that tells the cell, "Hey, I belong in the matrix." Once ARCUS is within the matrix, it can then go find its DNA sequence and generate that double-stranded break. And in this case, the target sequence is present exclusively within the mutant mitochondrial genomes. After the double-stranded break is generated, the molecules that are cut will be degraded and any of the remaining molecules, which in this case are the wild type molecules, will be replicated in order to repopulate the cell. And this phenomenon is what we refer to as shifting heteroplasmy. It's the shift in heteroplasmy that is ultimately our therapeutic objective as we are seeking to reduce the percentage of the mutation below that disease threshold, and thus, alleviate the symptoms. While we are currently utilizing this strategy specifically for 3243, this general approach could be easily adapted for other mitochondrial DNA mutations as well as deletions. So in order for this to work, of course, we first have to get ARCUS into mitochondria. As I said, we can do this by fusing that mitochondrial targeting sequence to ARCUS. So what we're looking at here are images of cells that were treated with what we call this mitoARCUS or ARCUS that has that mitochondrial targeting sequence. And then we're visualizing either the ARCUS protein, which is in green, mitochondria, which are in red, or the nuclei, which are in blue. The merged image on the bottom right shows you the overlay of all 3 of those images. And what you can see is that the ARCUS or the green overlays with the red really nicely, indicating very efficient mitochondrial localization. Importantly, there's no green and blue detected in the same spot, indicating that there's no ARCUS protein present in the nucleus. So we've shown that we can get ARCUS into mitochondria. That's great. The next really important piece of this is generating a nuclease that can specifically cut the mutant 3243G mitochondrial DNA. And so in order to test the efficacy of our nuclease, we're utilizing cells that have a very high degree of heteroplasmy at position 3243. So these cells are actually 95% mutant. And that's depicted in this bar on the left by the different colors within the bar. So there's about 95% of that bar that's red, which corresponds to mutant mitochondrial DNA, and only 5% blue, which corresponds to wild type mitochondrial DNA. So we can then treat these cells with our developmental candidate nuclease, which we call PBGENE-PMM. And what we see is that after treatment, the cells have transitioned from being 95% mutant to essentially 0% mutant. So there's almost no mutant mitochondrial DNA present in these cells and it's all wild type. This indicates not only a very high degree of efficacy, the nuclease was able to cut its intended target sequence, but it also suggests a really high degree of specificity, because if the nuclease was cutting the wild type sequence, we wouldn't have gotten that repopulation. And so this suggests that the nuclease was able to make that very accurate discrimination. However, we wanted to test the specificity in a bit more of a stringent of a way. And so in order to do this, we utilized wild type cells. So these cells are wild type. They don't contain any mutant mitochondrial DNA. All of the mitochondrial DNA is that off target or wild type sequence. And what we're looking at here is we want to see if PBGENE-PMM cuts the wild type sequence. We don't want it to, but we want to look to see if it does. If it does cut the wild type sequence, that would translate as a loss of mitochondrial DNA copy number. And so what we're hoping to see is no depletion of mitochondrial DNA in these cells. And when we utilize a 3x higher dose of PBGENE-PMM and evaluate these cells along a time course, we don't see any meaningful changes in wild type mitochondrial DNA copy number, which indicates that while PBGENE-PMM can cut its on target sequence or its intended sequence, which is the mutant, it is not cutting the wild type sequence, which speaks to just the very, very robust protein engineering process that we have here at Precision. So in addition to specificity in the context of the mitochondrial genome, we also wanted to look at specificity in the context of the nuclear genome. So from those images that I showed you, we didn't see any protein present in the nucleus. However, there's obviously a lot of DNA there that we want to make sure we're not cutting. And so in order to test this, we identified 14 potential off target sequences within the nuclear genome and then evaluated those sequences for editing at a 20x higher dose of PBGENE-PMM. And across all 14 of the off target sites, we did not detect any nuclear off target editing, which again speaks to the specificity of the nuclease. So we've shown that we can get ARCUS to mitochondria. PBGENE-PMM is specific for its intended target sequence, so it cuts the mutant, but it does not cut the wild type sequence or any nuclear off target sequences. The last piece of this is seeing what happens phenotypically following our edit. And so the cells that we are utilizing that have a really high degree of heteroplasmy showed defects in energy production. And so what we're really looking to see is if our edit and our shift in heteroplasmy can improve energy production in the cells. And that is in fact what we see. So the cells that were treated with PBGENE-PMM show a greater than two-fold increase in energy production, which suggests that our intended edit or the shift in heteroplasmy is capable of greatly improving mitochondrial function. The last piece of data I would like to share with you is looking at this in an in vivo context. And so we wanted to see whether we could systemically deliver a mitoARCUS nuclease to an animal utilizing an AAV vector and generate shifts in heteroplasmy. And what we see is that across multiple tissues, including skeletal muscle, we do see very significant shifts in heteroplasmy following this one time AAV treatment. And importantly, the time point that we're looking at here is 6 months post AAV administration, which speaks to the durability of the treatment. So to summarize overall our approach for PMM, we have developed a nuclease specific for the 3243G mutation, which we call PBGENE-PMM and we plan to utilize a single vector one time administration to deliver this nuclease. The construct will include a muscle specific promoter, which will drive expression of the nuclease in our target tissues, which for PMM is skeletal muscle. The protein is localized exclusively to mitochondria through the use of the mitochondrial targeting sequence. And PBGENE-PMM has been shown to be incredibly specific for its intended target sequence. We've shown that it cuts the mutant sequence, but does not cut the wild type sequence or any of the identified nuclear off target sites. Importantly, we are able to access and edit mitochondrial DNA due to the overall simplicity of the technology. Mitochondria are currently inaccessible to CRISPR due to a lack of RNA imported into mitochondria, and therefore, ARCUS can go where CRISPR cannot. Overall, we believe that ARCUS is uniquely suited for applications such as mitochondrial gene editing due to the simplicity of the technology. There's no guide RNA required for ARCUS like there is CRISPR, and that allows us to access and edit mitochondrial DNA. Additionally, this treatment has the potential to be transformative for patients with 3243 associated PMM. And we can utilize this approach to treat other mitochondrial DNA mutations as well as deletions. I hope I have been able to share my enthusiasm and my excitement for this project with you and I hope you see that the unique aspects of ARCUS allow us to treat mitochondrial diseases whereas other technologies cannot. It has been a great, great pleasure working at Precision and getting to pursue my PhD at the same time. And so, I'm so appreciative of the team here and the opportunity and I really hope that we can utilize this to improve the quality of life for patients out there.

Well, welcome back. We're very fortunate today to have 2 experts in mitochondrial disease to discuss the mitoARCUS approach in primary mitocardial myopathy-- primary mitochondrial myopathy. So on my left, we have Dr. Carlos Moraes, who is the Esther Lichtenstein Professor in Neurology at the University of Miami Miller School of Medicine. And to his left is Dr. Michio Hirano, who is Chief of the Division of Neuromuscular Disorders at New York-Presbyterian/Columbia University Medical Center. So Carlos, we'll begin with you for the first question. I know you've focused your career on studying ways to target the mutant to mitochondrial DNA through gene editing. So well, can you tell us why ARCUS and its single component enzyme approach is uniquely positioned to accomplish this, particularly in the 3243 PMM?

Sure. Nice to be here. Thank you for inviting me. And it's nice to see Wendy present. She did an amazing job. And she actually just [ defended ] her PhD in collaboration with my lab, just had a baby. And her papers, her work presented here was just accepted in Nature Metabolism. So it should be out in about a month or so. So yes, I think more than 20 years ago I started working -- maybe 30 now, working on mitochondrial diseases. And in the beginning we thought the mutations in mitochondrial DNA could not happen, because it's such an important thing, you need it for energy. And -- but then we realized that the mutations could happen because they were heteroplasmic. So there was the mutant, but also the wild type that would compensate. And only if you had a lot of mutation, not enough of the wild type, you get the disease. So early on, everybody in the field realized, "Wow. If we can get rid of this mutant, we have a magic bullet, that will be fantastic." Right? But there was not such a thing. That was in the 90s. So -- but since then, I've been looking for something that could do that. So we started working with restriction endonucleases that recognized very small pieces of DNA. So they're not very useful. But we show that it could work as a proof of principle. And then the gene editing technology came about, in about 2010. And we start using the TALENs, right, because there were proteins that could get into mitochondria with a targeting sequence. But TALENs, they're very big and they're dimeric, they need 2 proteins. So my dream was very similar to what Jeff Smith said his dream. It have to cut, it has to -- the size has to be small and it has to be simple. And so I have this dream until I realized that something like ARCUS was existed. And I got together with Precision. We start to collaborate and we start to do this work, as you mentioned. And several people mentioned CRISPR cannot work because you cannot import RNA into mitochondria. There's no such a system for that. And so ARCUS was the perfect molecule to do that. And I think we just proved in a couple of publications now that it works beautifully.

And tell me about the mitochondrial targeting sequence. How does that help for efficiency?

Right. So most of the proteins that are in the mitochondria are actually not encoded by the mitochondrial DNA. Mitochondrial DNA encodes very important catalytic subunits of the energy production system. But most proteins are encoded in nuclear DNA made in the cytosol and they have to go to mitochondria. So during this billion years of evolution, mitochondria developed a system to import proteins very efficiently. And we have recognized targeting sequence that take proteins inside the mitochondria. And we optimize it to use one that's very strict so that all the protein would go to mitochondria and not to the nucleus or stay in the cytosol.

Perfect. All right. Michio, this question is for you. Could you comment on the clinical manifestations of the 3243 primary mitochondrial myopathy as well as the current treatments and their effectiveness?

So the clinical manifestations of the 3243 mutation are heterogeneous depending on the heteroplasmy level that Wendy described and the distribution of the mutation. However, the majority of patients, as Wendy described, have symptoms of myopathy with weakness, fatigue and exercise intolerance. And the exercise intolerance and fatigue are particularly debilitating. Many of these patients are unable to do activities of daily living, such as doing laundry. That will often leave them exhausted for the rest of the day and they have to spend the day in the couch recovering. So that's a very debilitating component in the vast majority of these patients. So that's a great concern to the patients. And as a clinician, I would love to be able to treat that. And this mitoARCUS has given me great hope and excitement about meaningful treatment for these people.

What are the treatments that you're using now for these individuals?

So currently, the treatments are mainly geared at symptomatic treatment. So if they have hearing loss, we give them hearing aids. We give nutritional supplements, which have a modest effect. Subjectively the patients feel a bit better, but objectively we don't see anything that we can measure on examination. So our treatments are very limited at this point. And we're very excited about mitoARCUS as a clinician because it's going after the root cause of the disease, the mitochondrial DNA mutation, in this case, the 3243 mutation.

Perfect. As a follow up to that, Michio, what would it mean for patients to shift heteroplasmy towards the wild type?

So as Wendy described, the mutations are recessive, meaning you have to have a high level of mutation, 80%, 90% of the mutation in a cell to cause a biochemical defect. So if you could shift that heteroplasmy down from 80% to 50%, you can change yourself from biochemically deficient and unable to make energy, the ATP energy to a normal appearing cell. So I think we can -- if we can have a similar shift in patients, we can convert them from patients with severe debilitating myopathy to being energetically normal and being able to recover ability to do activities of daily living and beyond. So we're very excited about this possibility.

That's exciting. All right. Finally Carlos, this is for you. Can you expand on the design of the mitoARCUS approach? And how it's able to safely target the mitochondrial DNA but not the nuclear DNA?

Right. I guess a concern in the gene editing field are off target effects, right? You might try to cure something and make some damage somewhere else. So this is a very important question. In the case of the mitoARCUS, I mean, there are some advantages. First of all, it's not going to the nucleus to start with, right? As I just talked about, we use very effective targeting sequence that take the protein to mitochondria. So that already eliminates most of the danger of nuclear damage. Now on top of that -- and that's something that Wendy did also. She couldn't go into much detail. But you can also attach to these proteins a nuclear export sequence. And that's an extra layer of protection. And that makes proteins that -- if they leak or by accident end up in the nucleus, they're going to be exported out. So I think it's much safer than the general gene therapy approach. And another thing that I want to mention about the mitochondrial approach, mitochondrial DNA treatment is that in contrast to nuclear DNA defects that you need to express these genes for the life of the patient. In the case of the mitochondrial DNA, you just need a pulse that you can reduce this mutant load, as Michio said, from -- maybe from 80% to 40% and you pretty much can cure the disease, at least the symptoms. So that, I think, is another great advantage of the mitoARCUS approach relative to the more traditional gene therapy approach, right?

Sure. Yes, there's no doubt that safety is the biggest issue in gene editing for all the regulatory authorities. And it sounds like what you described having the ability to exclude it from the nucleus and direct it to the mitochondria makes it probably the safest approach we could see. Any -- we have some time left. Is there any other comments that either of you would like to make about the mitoARCUS program?

I mean the experiments we did in the lab show that's very effective, and the limitation here is delivery. So if you can reach an affected tissue, you can eliminate the mutant mitochondrial DNA. And again, we go to the size. The size issue makes AAV a very valuable delivery method, and that's what has been used for gene therapy so far. But it also opens the possibility for other non-viral deliveries that in the future might come up. If we can get it there, we can treat it. That's what we learned from all these years working with it.

Michio?

Yes. As a clinician, I just want to comment that I think mitochondrial disorders have an advantage in therapeutic approaches, because we're talking about the energy production of the cells. So structurally, the cells are more or less impacted, especially in the early stages. So by recovering the ability to make the ATP energy, we can basically revive a cell that's functional. It's like having a car without being able to burn the gasoline. So we're putting the spark plugs into the engine and allowing the -- in the car case, the car is able to burn the fuel and make energy and drive the motor to drive the car. In the case of humans, we're taking the ability to digest and use carbohydrates and fats and make ATP energy. So we're restoring the spark plug if we can restore the mitochondrial functions.

A very good analogy.

I mean just to follow up on that, we have some mouse work where we show exactly that, that the cells they can really live for a long time being defective and not dying. So there is this window that you can treat and restore the function or the fuel or the battery or whatever analogy you want to make.

Very good. Well, thank you both for joining us today for the discussion. Now I'm going to turn it over to Michael Amoroso and he can give us some concluding results.

Thank you, Alan, and thank you to our panel. Wonderful presentations, helping qualify and help us understand the disease and therapeutic approach for ARCUS in these major diseases. So to our investor community, we thank you for your time today. And as we sum up and go into our question-and-answer section, ARCUS is differentiated for the sophisticated edit, the cut, the 3 prime overhang inducing HDR, or in the absence of a DNA template, perfect re-ligation; the size, being able to go to many tissues around the body as we saw through multiple experiments today pre-clinically, as well as what we can package in that size for delivery. The greatest gene editors in the world can't be great unless they can get to the target tissue. And the simplicity. And some great examples we just heard for simplicity for our elimination program for HBV and for primary mitochondrial myopathy. We thank the panels. Altogether, the ARCUS advantages lead to a higher probability of defined outcomes. So if we've done our job well today, there's only one question left, when? And I think we've been getting a lot of that question from our investor community. You see here the time tables for these programs. On your left-hand side, first the HBV program, lots of positive momentum this year, final clinical candidate, Cassie and team. We had an FDA interact meeting. This is a pre-IND meeting -- a pre pre-IND meeting, if you will. And this is really focused on characterization and on target versus off-target editing. We met with the FDA. That was a very, very productive meeting. And the preparations to get ready for GLP Tox, which will begin activities in '24. You've heard Cassie talk about the target. If the science goes well, as intended, would be to file a CTA and/or IND for HBV in '24. And when we say CTA and/or IND, we just haven't flagged exactly what market we'll start in yet. For our newest program -- and thank you to our panel -- PMM, final clinical candidate this year, proof-of-concept work. You've seen a wonderful job by Wendy and team. And now we're well underway in dose finding studies. Also in the '24 time frame, you could expect GLP Tox. And from there, the target time frame for an IND and/or CTA -- we also have to decide which market we'll begin -- would be in the '25 time frame. And we'll continue to give you updates along the way. Again, when I say target date, if the science continues to go as planned. Our partnered programs. You heard from Cassie today and the really exciting insertion, 1-year sustained insertion in the iECURE program for OTC deficiency, dire need for those children. They will be submitting a CTA this year, actually the first in vivo ARCUS CTA accepted if it is submitted and accepted. And then for our other programs -- and I know we get a lot of questions all the time and I'll preempt some of them. Wonderful partners at Novartis and Prevail Lilly. The teams continue to progress and do wonderful work together. Those time lines on time to clinic will come from the respective partners, Novartis and Lilly, Prevail. And I know they're as anxious as we are to get to patients. So with that, I thank you so much for your time and attention today. And I'm going to invite my colleagues to come back to the stage and we're going to answer your questions. So Cassie and Jeff, please join me. Mark, Carlos and Michio are in screen with us also. Thank you.

[Operator Instructions] And we do have our first question over the phone from the line of Maurice Raycroft from Jefferies.

I was going to ask you a question related to DMD for editing in the muscle satellite stem cells as it relates to that program. Can you project or estimate how efficient that process is? And I believe those data you showed are from the mouse pole. Would you expect that to translate to humans or what sort of evidence do you have there?

Maurice, thanks for the question. It's Michael. Good to hear your voice. I'm going to send that all to no better person for that than Cassie here, so Cassie, please. I know that's your favorite slide in the DMD section. So go ahead.

Sure. Yes. Thanks, Maurice, for the question. Yes. We think it's really exciting data. It is early days in terms of quantification of the results. And you're right. The data that we showed today is evidence of muscle satellite stem cell editing from that mouse model. We are working diligently. The team is working really hard to continue analyzing those results, determining what level of editing within that stem cell population would actually be needed to really drive that therapeutic outcome over the long term. But this has been a real challenge in the DMD field is even demonstrating the ability to transduce these cells much less edit these cells. And so we think that this is a really exciting result. Even if at this point, it's just evidence of we don't quite have the quantitation yet. In terms of translatability, we've -- a lot of the AAV work, whether it's standard gene therapy or gene editing has started in mouse. And we've seen pretty good translatability as a lot of those programs have moved into the clinic. So we're really encouraged and enthusiastic about that data.

Thank you, Cassie.

That's helpful and makes sense. And maybe one other question for me. For HBV, I think in the publication, you only use one ARCUS enzyme. Is this the go-forward plan? Or do you plan on using a few separate ARCUS enzymes poles PCC and integrated DNA? Is there anything else additional that you're saying about the initial clinical study design and ex-patients who all at this point?

Yes. So Cassie, maybe the first half and I'll comment on the second.

Sure. Yes. So the paper that we published in 2022 utilized in ARCUS nucleus that targets within the polymerase gene. We've stuck with that target site. That target site, as I mentioned today is conserved in cccDNA and integrated DNA. So it's one nucleus that targets both. It won't be too separate nucleases. We have made some really important iterations, on the nuclease since what was reported in the publication. And that was really aimed at 2 approaches here, 2 aims, 2 goals: one, to improve the specificity as was reported in the publication. We did observe some off-target editing with that particular nucleus. We've made some substantial improvements in the specificity of the enzyme that we're moving into our GLP Tox studies. The other side of it is efficiency. We've actually seen a pretty nice increase. While we were optimizing for specificity, we actually saw an increase in efficacy as well. And so our clinical candidate improves on both of those parameters compared to what was reported in the publication.

Thank you, Cassie. And Maurice, as far as readiness, of course, in '24, when we filed our CTA and/or IND, this is a question that also came in off the phone lines. We will be looking to have our sites ready to go as quickly as possible. The IRBs, of course, you know are different. So sometimes you have to have that full confirmation depending on the site. Sometimes you can start your contracting before. We'll be working diligently in parallel. And then the last piece of this question, Mark, I'm going to ask you to comment in a moment, maybe not protocol specific, but just generally. When we think of the approach of what we're doing here eliminating cccDNA, creating indels at the integrated disease. What would be some of the endpoints we should be looking for clinically? What would be some of the earliest time frames you would see such endpoints?

Sure. So unfortunately, in hepatitis B, we have some really good blood markers of on-target effect. So we would expect to see in patients who are, let's say, they're treated with suppressed DNA. We could measure viral proteins and there's, several we can measure in the blood quite easily, surface antigen correlated antigen and e-antigen depending on the population and we should see declines in those. Now, if we were approaching a patient who's not suppressed on nucleoside nucleotide analog, and their DNA is high, given the targeting of that source of intact virions, the cccDNA, we'd expect to see a decline as was seen in the publication of DNA as well. So we have some very good peripheral markers that we can measure of on-target outcomes.

Thank you, Mark. Appreciate it. Next question?

Your next question comes from the line of Justin Zelin from BTIG.

Thanks for a very informative session here. I really appreciate you bring this on together. I was just curious if you can talk a little bit more about PMM. If you could highlight what the regulatory and clinical strategy passed for the program moving forward and if you could highlight if there's any other competitive agents for this market?

Yes. So I'll start, Justin, I'll start backwards on that a little bit. So I'm going to put it to my colleagues here in a minute to talk about what therapeutic options exist today. Cassie, I'll ask you to open up and then maybe go to Carlos here and Michio. As far as regulatory path, Justin, very early here, you saw that we -- we obviously have a clinical candidate this year. We're doing some of the dose finding. So those are some answers that we'll probably be giving you as we have more of those interactions. But Cassie, maybe we'll start with you and head over to our KOLs a little bit.

Yes. I think from a science perspective, it's a really unique and interesting program that I think Wendy really highlighted takes advantage of a number of the attributes that make ARGUS unique in the editing space. In terms of regulatory, there's, some unique considerations when thinking about this PMM program. Editing mitochondrial DNA is different than editing nuclear DNA. And I think Wendy did a really nice job highlighting what those differences are. But those also have implications from a regulatory perspective in terms of the types of studies that you need to prepare to support a regulatory filing. We're really encouraged by the specificity of the nucleus. I think it's immaculately clean. We see no off-target editing at the wild-type mitochondrial targeting site. And we've seen no off-target editing in the nucleus. And that's really controlled by the localization of the nuclease out of the nucleus that's hard to say. So I think it's a really unique, really exciting, really fun program from a science perspective that does present some unique challenges. But the team is working really hard to anticipate and to work with regulators to make sure that we're fully prepared to support those regulatory interactions. And maybe I'll hand it over to Carlos and Michio to really speak to, from their perspective, what the clinical landscape may look like.

Yes. So there are some other pharmaceutical companies that are working at the clinical trial level on primary Mito myopathies. I think there are 2 general approaches. One is to target some of the downstream effects of the primary mitochondrial defects. So, one process to use P-PAR agonists that increase the number of mitochondria and increased fatty-acid oxidation, but it's a downstream target. Another target is a small molecule that targets cardiolipin; which in turn stabilizes the monochondral oxidate phosphorylation super-complex. And again, it's a downstream target that doesn't go after the heart of the primary problem. There is a -- the second approach is a more targeted approach to primary mitochondrial myopathy, where you go after the primary cause and root cause like the 3243 mutation. There's a power program working on timed kinase deficiency, where they're using nucleoside therapy to work on a particular form of primary mitochondrial myopathy. So there are a number of companies and products that are being tested. I think that there are outcome measures that are being used. Some have been used previously in FDA approvals, such as the 6-minute walk test, which is a standard assessment of motor function. So I think that there are outcomes that can be borrowed from the neuromuscular field and applied to primary mitochondrial myopathies. So there's a path to follow there. And I think that this approach, I think, will be followed similarly here.

Just to add I think the question also was about some competitors' approach. And there are disorder enzymes like talents and their Mito-talents and the zinc finger. But they do have those problems that they are very large proteins, very large genes and they are primary. And then from what I've seen in the literature, I think Precision is far ahead also because they have this particular molecule against the 3243 that's very specific. And I have not seen anything like that from the other platforms.

Thank you, gentlemen. Thank you, Cassie. Next question from the line, please.

Your next question comes from the line of Andrea Tan from Goldman Sachs.

Michael, in the context of a seemingly conservative FDA and what's happening from a regulatory standpoint with other In Vivo gene editing companies, where we're seeing some clinical holds. Just curious if that's playing into your decision about which geographical territory to pursue initially? And then, for Cassie, I'm just wondering if you can speak about what gives you the confidence that you will have sufficient data to answer the agency's concern over off-target editing or potential impact to germline cells?

Yes. Andrea, very relevant questions and welcome back. So I'll start. The short answer is yes. We're actively monitoring the landscape. Starting from the very beginning, we've always talked about, first and foremost, that we can tag an ARCUS cut uniquely, which we think is very, very helpful for on versus off-target characterization. Second from there, as far as what markets will go into, the Interact meeting we just had, for example, is an example of we are taking the approach to get feedback early and often, starting even with the FDA. And that was very, very productive for us. Andrea, as you know, this is a bit of a crowded derive say, competitive landscape. So right now, we are leveraging U.S. and ex-U.S. regulatory guidance. But we haven't necessarily said what markets the world will think of starting with just yet. Cassie, I'd come over to you to see if there's anything you'd add to Andrea's question.

Yes. I would just echo what Michael said. We're really prioritizing regulatory feedback early and often, whether that's from the U.S. or ex-U.S. regulatory authorities. One of the really exciting results of our recent interact feedback on our HBV program was alignment around our specificity characterization for that particular nucleus. And really, that's a platform approach. And so the ability to align with FDA on specificity characterization for HBV will translate in a lot of ways to a lot of our other programs internally or partnered. And so we're really excited about that particular piece of feedback. And as we think about other challenges that we know the field is currently experiencing and people are spending a lot of time and effort on right now, like germline editing. It's another topic that we're staying really close with regulators on. We're prepared to do the work to demonstrate that we do not have germline editing. We're not going to generate a transmissible edit. And so we're prioritizing those regulatory interactions and looking for advice anywhere and whenever we can get it.

Jeff, is there anything you'd want to add about maybe some of the work we've done early with NIST even helping shape the guidelines and policy of off-target characterization?

Sure. That's a great comment. We've definitely -- so Precision has a very deep pipeline for specificity characterization. And as Michael alluded to, we've actually shown due to that increased insertion that we see. We are able to detect any off targeting 4x higher sensitivity than a lot of our competitors. And that's really allowed us the ability to engineer something that is very safe and that's been received well by the FDA. And certainly, in the conversations with NIST as far helping to set that bar high because I think specificity is going to be very important.

Yes. And Andrea, I think I would just conclude with an overall philosophy comment in this landscape. I think for us, there's -- watch the evolving landscape, see what someone else does and then maybe see the implications from a regulatory standpoint. I think prospectively, we've decided when we talk about permanence and altering the genome in the context of HBV or any of our programs, we're going to do the fullest panel of characterization work we can. So we're starting that work with regulatory feedback. But we're also starting that work prior to regulatory feedback. And we understand an area like Germline is super important. Thank you. Next question?

And your next question comes from the line of [ Mayberts ] from Precision Biosciences.

Michael, this is the question submitted by Robert Goff from U.S. Bank. How quickly after a CTA or IND filing would the company be able to dose for hepatitis B? And as a follow-up question, how soon after dosing would you expect to see data?

Yes. So I think it speaks a little bit to what I was getting at before. And I'll ask Mark to chime in, in a moment. But operationally, we will be working in parallel to onboard our sites when we're filing and waiting in our CTA filing our CTA and/or IND submissions and waiting on that feedback. There's a process. Every site is a bit different. Usually, you have to have the approved CTA and/or IND before the site can go through. But we'll be contracting in parallel. We want to start as fast. Patients are need as fast as possible once we get a CTA and/or IND approval. And then, Mark, maybe you could chat a little bit more. I know you talked about the endpoints, but serum levels, is it 3 months? Is it 4 months? Is it 8 months, 6 months? What is the time frame of when we can expect to see some of these serum levels and some first early signs of efficacy?

Sure. It's a really important question. As I mentioned, these markers are relatively easy to measure in blood with accepted widely used assays. So that's the first thing. The turnaround time is very quick on these assays. I think the second part of that is the translation from the data we have thus far for this platform in, say, for example, the mouse model. As you recall, we saw in that model very rapid declines in these markers of hepatitis B replication. Now if we anticipate that translating to humans of people with hepatitis B, we might see that within the first 4 to 12 weeks of dosing. Of course, we've got to do the studies and see. But given the reliability of the plasma and blood markers and the time lines observed in the mouse model, I'm optimistic that we'd have a readout relatively quickly from the first cohort to be treated with therapeutic doses.

Excellent, Mark. And I know our investor community is as anxious as we are to say, when are you starting in the clinic and how fast is there an endpoint to see if it's working and safe. So I think what you're hearing is, we think from the time we get in the clinic, it could be relatively fast. So there's a patient on the other end of this. We're always thinking about that. So thanks for the question. Next question to the line?

Your next question comes from the line of Kelsey Goodwin from Guggenheim.

This is Robert. One question from us on the platform. Can you characterize market with regards to persistent expression following AAV administration? And are you comfortable with the long-term safety profile following AAV?

Yes. I'm going to -- it's a great question. Thank you. I'm going to push this over to Jeff and Cassie. So please, I know we talked about this topic quite often.

Yes. The simple answer is I'm very comfortable. We have engineered ARCUS to be very specific to find its target site. And once that it is done, it has no more target site. It's largely not doing anything else. In addition to that, we have seen, and Cassie can touch on this even more that over time, because, as a reminder, we have, I think, the longest-standing safety data out there where we have delivered ARCUS with an AAV in nonhuman primates. And we're well over 6 years, worth of safety data. And when they have looked, they have seen that ARGUS any off targeting actually has, if anything decreased and not increased over time. Cassie, anything to add?

Yes. I think I would just add that, of course, it's always part of our development plan to characterize the pharmacokinetics of the ARCUS nucleus after an administration, whether that's by AAV or by LNP. We certainly know that the expression persistence will be longer with an AAV. And I think you're alluding to potential safety concerns related to that long-term expression. And I think the data that Jeff referenced of our long-term nonhuman primate studies really support the idea that expression, at least in liver, where we've characterized it best is actually quite transient. We see that the ARCUS can be detected about 1 month after the AAV administration. But by about 4 months, we actually don't detect any more ARCUS protein expression, which was a little bit surprising, but has been consistent across many, many nonhuman primate studies thus far, different nucleases, all delivered by AAV to the liver. I think at this point, we have a cumulative over 80 years of in-life observations and primates that had been treated with AAV. And as Jeff said, the specificity profile, if anything, over the long term, has actually improved. And so that gives us a lot of enthusiasm and encouragement that AAV will be safe and well tolerated as long as you characterize it well over the long term for a gene editing approach.

So Robert, it's a great question. And I'd say we brag about the team a little bit here. We think we've got the longest safety data. We know we've got the longest safety data in primates and that is through an AAV delivery, so great question. Next question from the lines?

Your next question comes from the line of [ Mayberts ] from Precision BioSciences.

Michael, our next question is also from Mr. Goff. In addition to the specificity characterization that Jeff already touched on, what lessons has the company learned from its, ex-vivo clinical trials that we'll use in the upcoming In Vivo trials.

Very good question, so the applicability from our ex vivo or the allogeneic CAR-T platform to In Vivo editing. Dr. Smith, you want to take that one?

Sure, my pleasure. So really, the largest and key thing that we learned from our ex-vivo programs was the very high rate of insertion for our CAR-T. And looking at that, the very high rate that that was done by homology-directed repair. And so really developing the CAR-T program really taught us how well ARCUS was performing for these applications that we're going to take forward in In Vivo.

Thank you, Jeff. Next question from the line, please?

Your next question will come from [ Mayberts ].

The last question we have is from an investor. How do we think ARCUS compares to epigenetic editing approaches?

Okay, great question. Jeff, I have a feeling you have an opinion on this.

Well, of course, it depends on the application. But if you ever want true permanence, then you want to have a change in the genome. So in the case of hepatitis B, there are some epigenetics out there looking at just turning off the virus. The risk of that is at any point, the virus could turn back on. And in our case, we're eliminating the virus. So I would think that that is a strong indication at least in that application as to which way you would probably choose.

Yes, it's a great question. I think we believe durability will be a big advantage of ARCUS genome. Next question?

And there are no further questions.

Terrific. Well, I just want to take a moment to thank the investor community sticking with us. We've had a pretty long day. But I hope it's been a rich day with science. I know many of you asked when I'm out on the street and meeting with you. So I hope you -- we're excited by some of the new data presented today. I hope you're as confident about ARCUS and Precision's path forward as we are. And I want to thank the panel of my team colleagues to our thought leaders who came in and got on airplanes because they're as excited as our approach. Thank you so much for your professionalism and your input and thank you to a team of wonderful people at Precision, our Precisioners.

And this does conclude Precision Bioscience R&D Day. Thank you for joining the webcast. Have a great day.