Sana Biotechnology, Inc. (SANA) Earnings Call Transcript & Summary

All right. Can you hear me? Very good. I think you can, right? So we're on webcast, so I'm going to go ahead and get this going. First of all, thank you for coming over on a beautiful New York spring morning to spend a little bit of time with us. And thank you to everybody who's joined us on the webcast as well. And our goal today is relatively straightforward. We're really optimistic that we have the potential to build a meaningful and durable company that has impact on tens or thousands of patients per year and creates value really for all of our stakeholders. And so what we're going to try to accomplish today is 3 things. One is to tell you a little bit about why we're so excited about where we're going and what some of the areas are that we're going into. The second is how we think we're going to learn in the very near term, in many cases, whether or not we're on the right track and really help you to understand how we see risk changing and where we see changes in how we'll think about allocating capital and investing and moving forward. And then the third is to introduce you to some of the people who will be really instrumental in ensuring that we can both leverage the opportunities that sit in front of us and also navigate the inevitable challenges that will come with it. So it's pretty straightforward. Hopefully, we'll get that done. To do that, of course, we'll have to make some forward-looking statements. We spend a lot of time writing our risk factors. Take a look at our 10-Q, which is on file with the SEC. There's a lot of information in there. So when we started the company, our goal was really to turn engineered cells, the power of cellular therapy and gene therapy into a reality. And it's a reality that hopefully is accessible for tens of thousands of patients or hundreds of thousands of people per year. And in doing that, we've really built 4 platforms over the last several years. One, that's what we call hyperimmune modified cells and ability, hopefully, the hide cells from immune recognition and allogeneic and autoimmune rejection. On second is to be able to scale this, we have to be able to make things repeatedly. And we've built capabilities around stem cells that we think can be very, very valuable. The third is we don't just want to replace cells. We actually want to repair them in the body when at the right time. And to do that, I think most of you recognize you can do almost anything you want to a gene inside of a petri dish in a lab. And the real challenge has been in vivo delivery. And so we'll tell you a little bit. And so we'll tell you a little bit about our fusogen capabilities, which are geared towards in vivo delivery. And once we get to the cell, we don't just want to randomly do things. We want to be able to deliver really any payload and to be able to modify a specific cell in the -- sorry, specific gene in a specific cell. We call that genome modification, and we'll show you some data on the progress we've made around that. So we're making progress across all 4 of these areas, and it will be fun to share that with you. The most near term though, is our hypoimmune platform, where we are optimistic we'll have proof of concept from 2 different studies as we move through the rest of this year, hopefully showing that this platform really does work as we hope it does. First is a drug called SC291. This is a gene-modified allogeneic CAR T cell targeting CD19 that we're moving forward in B-cell malignancies. We'll tell you a little bit about that and how we see success shortly. And the other is a gene-modified islet cell once you go into type 1 diabetes. And if we're successful with that, I think -- our view is that all the component parts have come together where you can now comfortably say that a cure for type 1 diabetes is inevitable. Right now, it's possible, but it will really change it to an evitable. So we won't be done. If those things do play out as we hope, not only we go after CD19 in type 1 diabetes, but our allogeneic CAR T cell field. We've got SC262, hopefully with an IND later this year, and SC255. Those target CD22 and BCMA, respectively. We'll tell you a little bit about why we're so excited about the opportunity in autoimmune disorders as part of that. And we have both the scale to do this and actually the drug products in our shelf to move quickly, and we think it can be a really important part of the company. And I talked about SC451 already, which is type 1 diabetes. And finally, we're going to talk a little bit some things we haven't talked about before. One is SC379, which is a glial progenitor cell drug. We think they can be really quite disruptive in the treatment of CNS diseases and some of the progress that we made and moving forward with our fusogen platform, both with SC299, which is a CD19 CAR, and also with delivering specific gene-editing payloads to specific cells inside the body. So that's a little bit about who we are and where we are. And it's a lot of stuff coming in the very near term. So just to start, transplant rejection has been a huge problem ever since the advent of transplant medicine. It's true, whether you're looking at solid organ tumors or bone marrow transplants. And the field has really struggled with how to deal with this. immunosuppression has been helpful, but it's 1 very toxic crimini patients. They end up with infections or cancers. And two, it's not tolerated that well. and that's really limited the field. Others have tried going down the path of autologous cells that has worked somewhat. It's difficult to scale. And of course, it's only usable in a very small number of cell types, cells that exist in suspension, bone marrow cells, T cells, things like that. And so our goal has been to really design a platform that will overcome the immune rejection of foreign cells. And we think if we can do that, you have a additional vast array of opportunities ahead of us. And the way we went about it was first and foremost, was not to approach this randomly, but to hire a team of really great transplant immunologists who understood the challenges ahead of -- in the field. And the real key has been that you have to overcome both adaptive immunity and innate immunity. And adaptive immunity, which is our B and T cells, is relatively easy to deal with. The field figured this out a while ago. And the challenge has really been trying to figure out how do you deal with the innate immune system, in particular, NK cells. So our team has been at this for a while. And I have to say they pioneered a lot in the field. We have foundational electro property that goes back almost a decade. And this is just some of the -- these are just some of the high-quality papers that have been published from this group, including 3 over the last 6 weeks or so in Science and Nature journals where they really, I think, begun to untangle the web of what is this challenge around immune rejection. And we'll share some of that data with you. And we'll also share some new data with you. This is just an example of some things we haven't shown before. And this is a nonhuman primate model of diabetes where we've shown that we can transplant cells and the animal will be euglycemic with no immunosuppression despite the transplant of allogeneic cells. And we're very excited about where we can go in diabetes. And we've now gone and you'll see more of that data later. We've shown that we can overcome allogeneic rejection in mice, in humanized mice in nonhuman primates. We've done this across multiple cell types. We've done this in multiple geographies. In fact, we've done over 40 nonhuman primate studies just to understand this. And I would postulate to you that if you were a mouse or a mouse with a human immune system or nonhuman primate, the problem of allogeneic rejection would be solved because none of us are, right? And so the real challenge is can we do that in people. And if we have just in near-term opportunities where others in the field have already validated the biology, have already shown you that you can transplant cells and have potentially curative effect. We get the opportunity to go after a host of different diseases. In blood cancers, over 100,000 patients per year are still dying of diseases that are readily addressable by CAR T cells. Autoimmune disorders. You could argue -- you have over 5 million patients that suffer from diseases where there is clear proof of concept that B-cell depletion, and not just B-cell depletion, but better B-cell depletion, will lead to increased effect. And type 1 diabetes, which already has shown some -- the effect of transplanting cells with immune suppression. And those are big, big opportunities that sit in front of us, and all of them have very near-term proof of concept. So to understand the proof of concept and where it's coming from, I'll just take a step back and make sure you understand our medicines. So the first place we took the hyperimmune platform was in allogeneic CAR T cells. And what we do is take a donor -- a healthy donor's white blood cells, take out their T cells, gene modify it in ways that we'll describe in more detail later, and insert a CAR, which will allow it to recognize and then kill the cancer. And we can do this across multiple different targets. In the first drug, it's CD19, right? So when you do that, you actually end up with a bit of a mixture of fully edited cells and not fully edited cells. And in fact, only about 40% to 50% of cells have all of the edits that we need to hide them from the immune system. That sounds like a bug. But in translational sciences, we can turn it into a feature. And what -- when a patient gets CAR T cells, the first thing they do is they get left for depletion, right? It's an autologous CAR T cell that will knock down the immune system, but it's pretty gentle and it comes back relatively quickly within a few weeks. And what we see consistently with most strategies with allogeneic CAR T cells is they're gone within 2 to 4 weeks as the immune system recovers, right? And so people have tried to just knock immune system down more robustly. But here, what you'll see is we'll lock the immune system down. It will come back within 2 to 4 weeks. And as the NK cells recover first, if they come first and then the T cells, you will eliminate all of the cells that aren't 100% edited. So at the end, if things work like we think they will, you have basically 100% of the remaining CAR T cells will be fully edited. So you'll know probably at 1 month. if all you see is CAR T cell, definitely at 2 months, definitively at 3 months. If it's 100% CAR T cells that we have actually overcome the challenge of immune rejection. And that all of that data that you've seen in mice -- humanized mice primates will translate into humans. And that opens up a huge array of opportunities for the company. It doesn't mean we have a drug, right? What it will mean is that we have solved this biology problem. And now we need to turn it into medicines. So the way you turn into medicines was keep peeling back the evidence. So we'll have [ Immunoevasion ] data, we can figure that out very quickly within a few patients. we'll do that. We have an ongoing enrolling study called ARDENT, which we'll talk about for the CD19 CAR. And if you have [ immunoevasion ] and you see cancer cells killed and you know they persist, that should lead to durable complete responses. We will know early evidence of just the complete responses and persistence. Again, as we move through this year and into next. And the durable complete response is 6 months and now, functional cures. Hopefully, we can begin to see that as we move into next year. And our goal is very simple. We want allogeneic cells that behave at least as well as autologous cells. We're not trying to make them kind of good. We're not trying -- it won't be success. We want them to work at least as well as autologous cells, and we have reasons to believe that they could be better. And Terry is going to walk you through because sometimes I hear from some investors, they're a little confused about whether or not cellular persistence really matters, which is -- it's a hard argument to make that area under the curve of your active pharmaceutical ingredient isn't important. But we'll show you a lot of data from the CAR T cell field that backs that up. And if it works in CD19, as I mentioned, we won't be done. We will be moving into CD22. We'll be moving into BCMA. In both cases, we have CAR T cells with what looks like best-in-class data in the autologous CAR T cell setting. So as you move that into our platform, it's a very low biology risk to making 3 different medicines. And then we have a host of targets to go forward. And to remind you of what this really means, you have data already and approvals already in multiple lines of therapy in lymphoma, myeloma, leukemia. And what we've seen is that you still -- patients still cannot get drug. If you went over to Memorial Sloan Kettering and asked them what percentage of their patients can get a BCMA CAR T cell today for myeloma? They would say it's 1 or 2 out of 10, right? It's just nowhere near where -- that's before you move into those light blue circles of expanded indications. So on the right, you see 250,000 people in the U.S. and Europe get these diseases every year where you have clear definitive proof of concept. Over 100,000 people die every year. Last year, just over 10,000 people got a CAR T cell. And of those, only 30% to 40% have really benefited. So our goal is to take that little blue circle and expand it so that, that big blue gray circle becomes a whole lot smaller at the end of the day, and it's right in front of us. But we won't be done. And over -- for a while for the last 25 years, the field has been trying to figure out where do you -- how do we deplete B cells. And when you do where can you have benefit for patients. And not only has it been important in cancers, but it's also been shown across a host of different diseases to be highly impactful in autoimmune disorders. I'm sure many of you have seen the data from an autologous CAR T cell as an example from Germany over the last year or so, and we'll go into that more depth later. But one of the things that's been shown is that the depth of B-cell depletion correlates directly with both the depth of response and the durability of that response. And with that autologous CAR-T data, what we've seen is the potential to -- for people to have durable long-term remissions off all therapies or maybe even curative. And you could see that CAR T cells could do for autoimmune diseases, what they have done in blood cancers. But to do that, you can't manufacture drugs at scale that won't even -- at a level that won't even meet demand for third-line cancers. You have to fundamentally change the way you approach the CAR T manufacturing. We already have the scale to move forward here. We make what is likely the autoimmune dose, almost 1,000 batches for dosing per manufacturing run. That means if you have 100 runs per year, you can treat 100,000 patients. It's totally transformative. And we have the drug product from our oncology study sitting on our shelves, right, ready to go. So this is something where you can see us, you're going to see us move very quickly. Doug Williams will outline this for you. And where you can see proof of concept across multiple studies with something like a little settings, I should say, from something like a basket study in the near term. So stay tuned. I think this is going to be a really exciting area for the field. And if we truly have what we think we do, which is an allogeneic cell that performs at least as well as an autologous cell, we'll be in a great position to really go after that. And of course, we won't be done. Stem cell therapies are ultimately -- we want to move away from these donor-derived therapies into stem cell-based therapies. We have a single cell that becomes your entire drug product. It's very scalable and reproducible. So type 1 diabetes is simply put, a disease of the immune system and the beta cell, right? Your immune system kills your beta cells. The field has shown that if you replace beta cells, and immunosuppress a patient, you can have a patient who is functionally cured. Our goal is really simple, which is to make [ promise ] stem cell a product where a patient will be glycemic normal glucoses with no immunosuppression and no insulin. And they have that last for years and years and years and years. And we'll have an IND from that product in the near term. We're also looking to take stem cells in the glial cells. And this is an area many of you probably don't know as much about. So we'll spend more time on basic biology, but they're implicated in the majority of CNS diseases, and glial progenitor cells are by potent in your brain. They can become either astrocytes or oligodendrocytes. And they target diseases of hundreds of thousands of patients. And we have an opportunity to move forward with an IND next year and really go after both rare genetic disorders where you have the opportunity to get the clear proof of concept early as well as very large and prevalent diseases that affect people that we all know like progressive multiple sclerosis or Huntington's disease, and Steve Goldman will tell you a little bit about this later. And we have a chance in the near term in diabetes for another proof of concept. So the idea here patients have been getting primary cadaveric islet transplants in hundreds per year for well over a decade, and they work to keep patients off insulin for long periods of time, but they have to be heavily immunosuppressed. So the demand for them hasn't been that great because there aren't that many people for whom lifelong immunosuppression is better than lifelong insulin. Our goal did an investigator-sponsored trial to take these cadaveric islet cells, introduce our gene modifications to make them hyperimmune and transplant into a type 1 diabetes patient without immunosuppression. You will know, if you can transport cells, we've got a week or 2 and regular allogeneic cell will be rejected. If these cells are live at 2, 3, 4 weeks, let alone at 1 and 2 and 3 months. Again, we'll have proven that we can overcome the barriers of immune rejection in the allogeneic setting. And we've also have shown at least in 1 autoimmune setting that we can do this. And this is an opportunity for us to get rapid learnings around how our platform is really performing, and we'll get into kind of how we define success in key measures outcome. If you were really excited -- if you start with a lower dose. So we're looking for cell survival and C-peptide, which is a measure of insulin production. If we happen to have glycemic control, we'll be doing cartwheels particularly for that patient. So that all sits in front of us, and that's all stuff we're doing, where we're replacing cells. That's not all we want to do. We want to be able to fix cells in vivo as well. And we have this program. We call them fusogens, which allow cell-specific delivery, and we put them into something called the [indiscernible], right, which is like a -- it's a think of it like a virus-like particle where you've got both targeting and you can put packaging in it. So what we're going to show you is that we can do cell-specific delivery. And hopefully, we'll have an opportunity to prove that in humans with an IND slated for later this year with SG99. And that we can deliver diverse payloads. And those include things like typical nucleases used for gene editing. They include base editing. This is just an example we'll go into. And it's a whole lot more that we can do with these. They're quite powerful tools that are able to really think about this change, any cell gene in your body in a specific cell. That's the goal. That allows you to get more potency and efficiency. It allows you to have much more safety, and most of your drug product that's nonspecific, will go into nontarget cells. And it actually allows it to be scalable because if 90% of your drug is going to liver and 9% of it is going to some other cell and 1% is getting to your target, you have to manufacture a lot of drug to have the impact. This is better efficacy, better safety and better manufacturability if we get it right. So that's basically what we're going to talk to you about today. These are INDs this year and next. This is not the full platform. You can see we've got a lot of activity inside the company, we have a lot of -- a number of opportunities for clinical data over the very near term, let's call it a couple -- next couple of quarters. And as we move through next year. And we're quite excited to tell you about some of what that is. So the people you're going to meet today -- just as a way the -- we've kind of built the company by trying to put together 2 types of people. They're really creative innovative scientists that have been thriving in academia with the really drug hunters, and people who know how to turn innovation into drugs that sit inside of the industry. And you'll meet a little bit of both of them. Sonja Schrepfer runs our hyperimmune platform. She was formerly a transplant immunology professor at UCSF. Terry Fry runs our T cell as a therapeutic platform. Terry, has formally ran the T cell program, cell therapy programs at the National Cancer Institute and it's been at the University of Colorado as well. Doug Williams, those of you who don't know him, has been really someone who has been a leader in research for over 2 decades in our field. This played a large part in building companies like Immunex, Seattle Genetics, Biogen idec, and has been developing immunology drugs in particular, through that whole time, including multiple blockbusters. Steve Goldman is someone who's been leading the field in [ GlioBiology ], and is a real thought leader as well in how things work around with stem cells and turn these stem cells in the controllable medicines. And Ed Rebar , who you'll meet is someone who's been in the gene editing field since its inception in the late 1990s and understands not just how to genetic but how to turn genetic reagents into real medicines. And I think that's something that hasn't always been straightforward. I almost forgot Richard Mulligan. And if you don't know, Richard, he's someone who ran the Harvard Gene Therapy Institute made many of the fundamental discoveries that allowed viruses to be used to deliver genes, and has been part of the core of our scientific team since the outset of the company. The last person is not here but will be available for Q&A, as Gary Meininger, our Chief Medical Officer, who recently joined us from Vertex, where he ran their cell and gene therapy clinical development. So those are the people you're going to meet today. Hopefully, you'll be -- you'll have a chance to learn a little bit more about us and to share some of our excitement about what we can do. So with that, I'm going to turn it over to [ Sonja ] .

Thank you, Steve. So I'm happy to show you today some details about the hyperimmune platform. So on this slide, what you see is Cristiaan barnard, and I met him when I was in school because my background is in solid organ transplantation. And Cristiaan barnard transported 55 years ago, the first heart into a human being. And I show you this slide because you see the patient Lewis Washkansky on those pictures right after the surgery, and he looks quite happy. And well, however, the immunology kicked in a few days after the transplantation. And Lewis Washkansky was immunosuppressed, meaning he had to take medication to dampen the immune response that he wouldn't reject the transplanted organ. However, immunosuppression is every day, struggle because if we give too much immunosuppression, the patient has the risk of infection, malignancies, toxicities on kidney, liver, other organs. But if we don't give enough immunosuppression, the patient has the risk of rejecting the transplant. So back then, the surgical team thought that he would need more immunosuppression. They increased the immuno suppression and unfortunately, Lewis Washkansky died on the 18th post-operative day. So that shows you really the struggle we have in the field for not only solid organ transplantation but cell and tissue transplantation. For cellular transplantation, of course, we do have the option to transplant autologous, meaning that the cells can come from the patient and then we transplant the [ varices ] cells back into the same patient that wouldn't lead to immune rejection. However, there's struggle with manufacturing. And of course, there's the access limitation that not every patient has the possibility to access those therapies. What you see on the right side of that slide is the allogeneic cellular therapies. And the idea is that we have a healthy donor and we could generate multiple doses for many different patients really off the shelf for anyone, anywhere, anytime. The problem is when we think about allogenic transplantation that the draft or the cells we are transplanting are subject of immune rejection and the patient would need immunosuppression. So I want to show you how we thought about how can we overcome this allogenic barrier, meaning the vision -- transplanting cells without the need of immunosuppression. And we learned from nature because doing pregnancy, the fetus is expressing 50% of the proteins from the father but the mother's immune system is not rejecting the fetus or so the immune system recognizes the fetuses allogenic. And what you see here on the right side is that Sana's hyperimmune approach leverage that knowledge, and we are using a healthy donor cell. And the first thing we try to overcome an adaptive immune response, meaning T cells, B cells response, antibodies is that we are wiping off this fingerprint on the cells and the fingerprint, those are the MHC or HLA molecules. And by doing that, the T cells are not seeing the transplant itself. So you can easily overcome that adaptive immune barrier. However, because we are creating that cell -- that missing cell, there is no fingerprint on the surface, innate immune cells, NK cells and macrophages. They are sensing the missing self, and they are killing. So we had to overcome this innate immune barrier as well. And we show you data today that we are overexpressing the CD47 molecule that's a "don't kill me" molecule, and we can overcome that killing by innate immune cells. When we started in the field, you see the histology picture, all those purple dots, those are immune cells. So that's how it looks like when a craft gets rejected. When we then started the fetal-maternal border where we know the fetus not rejected by their mother, we saw that there are certain molecules overexpressed or down regulated on the fetal-maternal border. And we started each of those molecules alone and how that molecule affects the adaptive and the innate immune system. And then we came up with a list you see on the right where we then started to think what is the minimum amount of molecules we need to combine to achieve an hyperimmune cell product? And as Steve already said, wiping off the MHC molecules, the field is recognizing that you can overcome the T Cell response with that. But the goal was really then how can we protect the cells from the innate immunity. And we studied certain molecules, and I want to show you a few of them, where you see in the gray boxes, the molecules we are studying. And those molecules when you over express that on your cell transplant, what they do need is the corresponding receptor on the NK cells. It's like a key and lock principle. If you over express the molecule, of course, it has to find its a receptor on the NK cells. And the issue with the NK cells is what you see on that slide in the flow blots that, for example, I used the HLA-E to show you that the receptor CD94 is only expressed on 28% of the NK cells, meaning you don't -- you're not able to inhibit the overall primary NK population. For the PD-L1, you see the example here, only 3% of NK cells are expressing the receptor, the PD-1. So when you look at those flows, you see that CD47, the receptor [indiscernible] alpha is expressed on all NK-cell subpopulation. And that leads to the protection of the cells that are not killed by NK cells. What the videos here will show you are in green the target cells. We have a missing [ self ] cell. And then we over express the molecules HLAC, [indiscernible] PD-L1, [indiscernible] knockout of CD47 and added NK cells. And what you see in those videos is that over time, because not all of the NK cells are expressing the receptor for those molecules, NK cells are killing the target cells. And you see that the CD47 is a total protection because the receptor is expressed on all NK cells. I always get the question when I go on meetings and share those data. Oh, could you combine those molecules? Is it maybe an additive effect, and you just combine 3 or 4 of them, and then you get protection? So I wanted to show you these videos where we played around combined certain molecules. And what you see is that even with combination, there is not such thing like an additive effect of those receptors on NK cells, meaning that the single over expression of CD47 is protective because the receptor is 100% expressed on the NK cells. So what we started is that we wanted to have an allogenic model and mouse model. So we created [indiscernible] stem cells, [indiscernible], and then we transplanted them, on the bottom row into black 6 mice. So there is no immune response. But when we transplant those cells into the allogenic biopsy in gray, you see that the gray graph is much higher than the blue. So this is T cell activation. The other graph shows you the IgM antibody binding. When we then induce the hyperimmune edits, so MHC class 1 class 2, knockout CD47 over expression, you see that the gray and the blue are at a similar height that shows you that we were able to overcome that adaptive immune barrier. What we show you here on the right side is that Y-type cells don't get cut by NK cells, so the line here is up. When we are disrupting the MHC Class 1 class 2 molecules, and we have the knockout, you see that NK cells are killing. When we over express CD47, now the line stays up again. So that shows you the protection of that CD47 molecules from the NK cell killing. We then moved on and we wanted to use a more translation model. So we choose the nonhuman primate model, the NHP model. I show you data here from 8 NHPs. In group 1, we first transplanted unmodified allogeneic iPSCs from 1 NHP into another. After 6 weeks, we then didn't cross over experiment where the very same NHPs received some hypoimmune transplant. And in Group 2, it's the vice versa experiment. We first transplanted hyperimmune iPSCs. 6 weeks later, we transplanted the Y type. And then we did immune studies and surviving studies. And what you see here is the transplanted cells we labeled, we could follow them in vivo. Group A is the transplant of first hyperimmune IPCs into the left lack muscle of the NHPs. And after 16 weeks, you see here the image, and you see in blue, the line, how it stays up. So those iPSCs cells were able to survive in the NHP without immunosuppression. After 6 weeks, we had transplanted the unmodified Y types into the other leg muscles. That's what you see in the second row, and in gray here. And you see that the unmodified [ iPS ] cells are rejected quite fast. And what those data show you is this is a concept of immune evasion. So also the immune system was checked, now the unmodified iPS cells, you see that the hyperimmune cells are surviving cost the immune system is not seeing them there, evading the immune recognition. In B, you see the transplant of first unmodified iPSCs as expected, they get rejected quite fast within 2 weeks. That's a normal electronic response. And then 6 weeks later, you see that we transported hyperimmune, and those hyperimmune cells are surviving. That shows you that hyperimmune cells can be transplanted into sensitized recipients because the immune system of those NHPs knows the antigen but is not seeing the hyperimmune cells after transplantation, and therefore, they are surviving. We then did a bunch of immune analysis, and I want to show you the upper row first, the unmodified NHPs. The gray bars are the injections of the Y type iPSCs. And what you see is in the T cell activation, high gray bars in the beginning, like 1 week after the transplant, then it declines over time after the cells are rejected. After 6 weeks, you see in blue, there is no [indiscernible] So there is no T cell activation against the hyperimmune. If you look at the IgM antibody production, there is no de novo antibody production. But after the unmodified transplantation, we see first IgM antibodies and then they switch to IgGs. I'd like to show the slides because that is normally what we expected patients, and it underlines how similar the NHP immune system as to the patient's immune system. What you see on the bottom is we first transplanted hyperimmune, there is no blue graph coming up in the T cell activation. There's no antibody production. Six weeks later, when we transplant unmodified iPSCs in gray, you see the bars coming up and you see T cell activation, antibody production, high gray bars. And that shows you that the hyperimmune transplants we did are not modifying the recipients in 1 system. So the NHPs are capable of mounting a normal allogenic immune response. So there is no immunosuppression after our cell transplant. We then moved on into a differentiated cell types. So I want to show you here our data also in the NHP model with primary islet cells where we use primary islets, created the hypoimmunes and its transfer those islets into the muscle location of an monkey without immunosuppression. And you see the picture here is the follow-up 10 months later. So those islets are capable of surviving up to 10 months in this NHP without any immune rejection. When you compare to the Y type results on the bottom, you see that the islets are rejected quite fast within 1 week. Another differentiated cell type we did was the T cells for our SC291, the hypoimmune CD19 CAR-T program. So we used here humanized mice. So the immune system of the mice got humanized, and is capable of mounting a human in 1 response. We then induce the CD19 positive [ AM6 ] tumor. And what you see after 1 month, I said the tumor here is everywhere in those mice. When we are dosing with unmodified CAR Ts, in the middle column, or with SC291, that's our hypoimmune CD19 CAR, you see that the tumor gets cleared. So most of the CAR T studies stop there. And what we learned at that time point is that the hypoimmune [indiscernible] don't seem to impact the function of the CAR T cells. But if we look a little bit later, what we see is because the mice, they do have a humanized immune system, they were checking the unmodified CD19 CARs. And you see here how the tumor reoccurs in those mice. On the right side, in the right column, you see hypoimmune, they don't get rejected, so the tumor is under control. So we performed that study for quite a long time. And after like 2.5 months, I just looked at the data, the tumor was still cleared in hypoimmune. So I picked up the phone and called Terry Fry, you will present after me. And Terry, how long do you want us to look at the mice? When will you be convinced that hypoimmune persists and is surviving in those mice? And Terry gave the great advice, why don't you just rechallenge the mice with tumor and see what happens. And that's what we did. At day 83, we challenge the mice with tumor. And within only 4 days, you see how the hypoimmune cells cleared that tour. So that shows you that the survive persisted and the functional after 3 months. This is one of my favorite slides. So what it shows you is we could find those CD19 positive cells in the bone marrow and in the spleen of the humanized mice. And when we look at day 95, so 3 months after the injection. What we found is when we gate around the CD19-positive population that the phenotype, all surviving cells had the hypoimmune phenotypes. So [indiscernible] CD47 over expression, showing that there is an in vivo enrichment and the hypoimmune phenotype. Those are the CD19 CAR T cells that survived and cleared the tumor after 3 months. Another model we did is the iPSC-derived islet cells. So here, the iPSCs were either unmodified or they had the hypoimmune edits. Those are human [indiscernible] islets and we transplanted them into humanized mice. As expected, under white type when you look at the column here, the white type cells, they do get rejected by the humanized immune system. Hyperimmune does survive the follow-up years 1 month. To get on functional readout we induce diabetes into those mice. So we used STC. And at the day of the transplant, what is labeled with day 0, you see that the glucose levels are high. In the Y type group, where the cells are killed the glucose stays high. However, when you look here on the right side, the hypoimmune islets, they are surviving and they are functional. And you see that the glucose levels over time are dropping, and the C-peptide was up of the 500 picomole liter target. After overcoming this allogenic barrier in the preclinical models, I showed you, we were really interested in the question, could we also overcome the auto-immune in type 1 diabetes. So we used a very common type 1 diabetic autoimmune model than not short mice. And what we did in that model is we used not islets and translated them into autoimmune, not mice. So this is the same strain. The such shouldn't get rejected. It's like an autologous transplantation in a patient. However, what you see on the left side is that the cells still get killed because of the autoimmune disease of the mice, of the recipient. And you see there is no survival, the glucose levels we induced diabetes here again, they stayed up. On the right side, you see when we introduced the hypoimmune edits into those islets. The cells are surviving when the glucose levels are dropping, showing that those hypoimmune edits can also overcome that autoimmune barrier in that mouse model. We really wanted to study that in vivo with human cells as well. And for that, we need an allogenic transplantation. So we need the transplantation of islets into a mouse with the same immune system. And the study design here was that we used type 1 diabetic patient PBMCs, and we generated mice, humanized mice. So the mice have now the immune system of the type 1 diabetic patient. At the same time, we use the PBMCs and reprogram them to induce pluripotent stem cells iPSCs. Then a bucket of them were left unmodified, Y type, and the other half, we created the hypoimmune edits, the [indiscernible] class I, class II [indiscernible] CD-47 over expression. Then the subs were differentiated into islet cells, transplanted into the very same mice. So now we transplant the islets into the mice with the same immune system. So there shouldn't be any rejection. But what you see is when we do that with the Y type that in the pictures, none of the cells are surviving because of the autoimmune disease in the type 1 diabetic patient we used. So the cells are all getting killed. The mice, we induced diabetes, again, you see there's no glucose control. The glucose levels stayed up very high. When we did the very same experiment, but we created the hypoimmune -- we transplanted the hypoimmune islets, you see that the islets here are all surviving in the images. You see also here the graph, the BLI graph. And the C-peptide was above the 600 picometer per liter, and the glucose dropped over time in that model, showing you that in that humanized autoimmune model, also we can overcome the autoimmune. And then this is a slide Steve already shared with you but I wanted to give you some details. This is the nonhuman primate study is still ongoing, where we induce diabetes. And what you see here that the glucose levels in the beginning get really high. For 2 months, that NHP was then controlled by giving insulin. We did the cell transplant. We tapered down the insulin, and he's now since 20 days without insulin that NHP was not immunosuppressed. And if you look at the C-peptide levels, before we gave the STC, so before we maintained diabetes, he was in the normal range, 50 days later and on the day of the cell transplantation itself, the C-peptide is done. And then 7 days after cell transplantation of hypoimmune. And 14 days later, you see the C-peptide levels. We don't have longer follow-ups yet, but we are still working on it. As Steve already mentioned, we transplanted different hypoimmune cells into different NHPs. I think it was a totality of more than 40 NHPs we did. And I showed you the primary islets the iPSCs, but we also did [ IP-CDYRPE ] so cardiomyocytes. And all the high-level summary was that all of those cells were able to survive and immune evade. So there was no immune recognition by the adaptive or innate immune system. And with that, I conclude that the hypoimmune platform does have the potential to overcome the transplant rejection barrier with no immunosuppression through the concept of immune evasion. And as I showed you, we did so many different preclinical models that I was really running out of ideas, what else to design to further study it. So it's really time for us to move into the humans, and I'm very excited for the 2 clinical studies coming up later this year to prove the high premium platform. And with that, I hand over to Terry Fry, who will show you how to translate the hypoimmune platform into medicines. Thank you.

All right. Thank you, [ Sonja ]. So I'm going to walk you through where we're at with the allogeneic CAR T cell platform. And I think it certainly is not necessary to tell this room about the success that has been seen with autologous CAR T cells. And I think the emerging real-world experience, it's pretty clear that CAR T cells have consistently outperformed current standards of care. And what we're seeing is that CAR T cells are being moved earlier up in therapy, and we're seeing an expansion of indications. And -- but there's no doubt that we've seen challenges among them, scalability, that is the ability to get these products into patients. And the other important component that Steve didn't mention is the second bullet, under challenges. I'm a pediatric oncologist, and I've been in the unfortunate situation of trying to manage patients during the 4 to 5 weeks between when they get a [ freeze ], and when a CAR T cell is delivered. And during that time, you have a patient with progressive cancer, you have to administer chemotherapy, other agents that cause toxicity and some of those patients actually don't actually get to the infusion. And the last issue that I've spent a lot of time studying is that in spite of the success, it's certainly less than half of patients that will ultimately enjoy durable remissions after CAR T cell infusion. So no doubt, the allogeneic CAR T cell offers the opportunity to address the issues related to supply chain and time to treatment since these products are available immediately for patients. But as Steve mentioned, I think 1 of the biggest challenges is that we're not seeing the durability of permissions that we see with autologous CAR T cells. And so I'm personally very excited to get the hypoimmune platform that [ Sonja ] just described into an allogeneic T cell and into patients to see whether we can address these challenges. So Steve walked you through the manufacturing and the team has spent a lot of time optimizing this -- the manufacturing platform to enable consistent generation of high-quality CAR T cells. And I'll just walk you through this briefly. We begin with PBMCs, apheresis products from healthy donors, which are then T cell selected, and then using a novel nuclease Cas12b. Three genes are disrupted, Class I, Class II and the T cell receptor to prevent the induction of graft versus host disease. And then we use plenty of our vectors to introduce both CD47 and the CAR gene. These cells are expanded and then [ aliquoted ] anviled for delivery to patients. And what the team has been able to achieve is a process that consistently results in high-efficiency gene editing with 85% of cells approximately with full knock out of Class I and Class II, and following a selection step, 99.5% or greater of the cells being deficient in the T cell receptor, which is important for the prevention of graft-versus-host disease. And when we look across numerous manufacturing runs and we sort of estimate the number of doses that we think we can generate per manufacturing run... [Technical Difficulty] On the webcast, can you guys put the slides back on? All right. So I think where I was at, was saying that with the manufacturing runs, we estimate that per batch that based on what we expect a therapeutically active dose to be from autologous cells. We're looking at about 450 to 500 doses in an oncology patient and based on the emerging data in autoimmune diseases where the biologically active dose seems to be a bit lower. We're looking at over 900 doses per manufacturing run. So certainly, scale where we, as Steve said, expect that we can dose many, many patients with just a few manufacturing runs. So since we're editing the genome 3 times, I mean, one of the really important elements of the program has been to look deeply at the genomic integrity of the products that undergo the manufacturing process. And we've done that by looking at a number of different types of analysis. The first is looking at editing specificity. And this was done using 2 orthogonal approaches. And the second is to look at chromosome structural analysis, using 3 orthogonal approaches. And then finally, set of functional studies to look for acquisition of oncogenic properties. And I'm not going to walk you through all the data that ultimately went into the IND. But this is just a piece of that data, just as an example, looking at the specificity of the editing, and this is just showing analysis of 284 candidate sites. And so out of analysis of a number of donors, only 6 low size showed up. Only 2 of those were recurrent. And when the team dug deeper at those, both of those turned out to be due to artifacts. So the bottom line is, I mean, this nuclease, this is a Cas12bs novel nuclease looks to be very specific in the context of T cell manufacturing. So the other important thing to figure out is whether or not the edits that we're introducing, in particular, CD47, which was a novel introduction into T cells, impact the functionality of the T cell product that was generated. And this is a figure that was taken directly from the recent publication that [ Sonja ] mentioned earlier, looking at a comparison of the activity of hypoimmune CAR T cells relative to, call it, standard CAR T cells, allogeneic CAR T cells in this experiment. And across numerous donors and in a couple of different tumor models, there is no evidence of impact on T cell function. And in fact, one of the things that we noted early on that in most of these experiments, if anything, the hypoimmune CAR T cells seem to outperform the "standard" CAR T cells. And last week, Crystal [indiscernible] presented some data at ASGCT that potentially points to the mechanism here. And what she showed is that when you knock out CD47 in a CAR T cell and put them into NSG animals, the activity of the CAR T cell is markedly diminished. And they have attributed this to eradication or depletion of CAR T cells by muring myeloid cells that are present in the NSG animals and are capable of cross-reacting with CD47 on the surface of the T cell. So it suggests that even beyond the allogeneic immune rejection component of the hypoimmune product, there actually may be a benefit in terms of potency with the overexpression of CD47. So as Steve mentioned, the trial called the ARDENT trial is open and enrolling. It's a fairly standard safety trial with the Phase Ia component of dose escalation. The starting dose is the dose where we would potentially expect to see some biologic activity based on experience with autologous CAR T cell products. The disease indications are showing -- are shown there. I'll note that we're including CLL in the Phase Ia portion. The other important point is we're using standard lymphodepletion. There's no enhanced lymphodepletion with the goal of really understanding the ability for this hypoimmune platform to induce cell persistence. And then once we complete the Phase Ia dose escalation, we'll then embark on a series of Phase Ib expansions in a number of different disease indications. So [ Sonja ] already walked you through this experiment. And as you pointed out, the fact that one of the things that emerged from this was the ability to demonstrate enrichment of the hypoimmune cells in the humanized animal model due to protection from allogeneic immune rejection. And so as Steve mentioned, we expect that we'll be able to look at those same features in recipients -- in human recipients of SC291. And the expectation, as Steve mentioned, would be that with recovery of the immune system, particularly NK cells, which come up quite early, I'll show you that in the next slide, and T cells that there'd be depletion of the non-hypoimmune fraction with persistence of the hypoimmune component. And I think we'll be able to see that certainly out at 2 to 3 months, but even potentially by 1 month, we would expect that we would see some degree of enrichment. And the reason I think we'll see that earlier is because we've known this for a long time in the transplant field. And this is also true for CAR T cells. The first cell that comes back after lymphodepletion are NK cells. And in fact, they come back quite quickly, and they often rebound to levels that are higher than what you see prior to lymphodepletion. And when you wipe out Class I, Class II, these cells become literally sitting ducks for NK cell immune rejection and soon you walk you through. And so the expectation would be that CAR T cells that aren't protected from innate immunity would be rejected quite rapidly in the first month. So with this immunovasive property then, our goal is to do an infuse CAR T cell that avoids early allogeneic immune injection resulting in cell persistence, at least comparable to what we see with autologous CAR T cells with the goal of sustaining remissions in patients that achieve them. And I'll just walk you back a little bit. Again, I said I'm a pediatric oncologist, the first FDA approval for CAR T cells was actually in pediatric oncology with Kymriah in 2017. And I was involved in some of those early trials in the early work at the NCI. And the CR -- the remission induction rate in kids with leukemia was pretty remarkable. It's about 90% of patients achieved remission. Unfortunately, if you look at the data from the ELIANA trial, this was the trial that led to the licensure of Kymriah, 50% of the patients achieving remission ultimately relapse within the first year. And the reason for the relapse really fell into 2 different buckets. One pattern is the leukemic resistance. That is loss of the CD19 antigen that's being targeted by the CAR. Obviously, not a problem with the CAR T cells, that's leukemia intrinsic effect. But the other pattern that we saw was that with early B-cell recovery within the first 3 months after CAR T cell infusion, that's evidence of lack of a functional CAR T cell persistence. The vast majority of those patients ultimately relapse of leukemia that retained expression of the CD19 protein on the surface, again, indicating the problem with lack of cell persistence. I think it took a little bit longer for the lymphoma field to identify that same pattern. Although if you look at this early data from Cam Turtle, from the Fred Hutchinson Cancer Research Center, this was a trial in which they looked at a couple of different lymphodepletion regimens, what they observed was that a subset of patients had durable persistence of CAR T cells, other patients, those not getting fludarabine, actually had rapid clearance of the CAR T cells. And if you look on the right-hand side of this slide, what they were able to demonstrate was that with persistence of CARs, the durability of the remission went up in the patients that were treated in this trial. As Steve already showed you this, this is data from the ZUMA-1 trial. I'll point out that this is actually a car with a CD28 costimulatory domain. Many people think that this car doesn't persist. But in fact, if you look carefully in the bone marrow, you actually can find CAR gene in patients for long periods of time. And in fact, if you look at this data and patients who achieve remission, you can find gene mark cells in 95% of them at early time points, and over 50% even at late time points. And so we really do believe, and I think the field has shown us that persistence matters, certainly in leukemia. I hope I've convinced you that it's also true in lymphoma. And it also is emerging as an important feature in myeloma. I'll talk about that in a little bit. And so we believe that with immunovation that if we can get CAR T cells to persist out to 2 to 3 months, it would certainly would be longer than what we're seeing with current allogeneic CAR T cell approaches. But we actually think that we have the potential to see persistence longer than that in these patients. So as Steve mentioned, I mean, the manufacturing platform in many ways, is modular, the gene editing, the cell modifications that we're doing can certainly be applied to other therapeutics and simply by swapping out the CAR itself, we can then generate CAR T cell products targeting multiple antigens. And the next 2 programs after the CD19 program will be a program targeting CD22 in CAR refractory patients and then a follow-on program with a BCMA CAR. And importantly, in both of those cases, we've in-licensed fully clinically validated CAR constructs. I certainly learned across a lot of CAR development work in my lab and taking a number of those in the clinic that it's good to have human data to confirm activity. But then we also look forward to expanding into other targets in the future. So one of the things that we're certainly recognizing is -- and I mentioned this briefly in my comments about leukemia, is that across multiple diseases, leukemia lymphoma, et cetera. It's really less than about half of the patients who receive CAR T cells that ultimately achieve durable remission. And so with an increase in the use of CD19 CAR T cells, we would expect that this car failure population will grow. Not a big surprise in the right-hand side of this slide are the patients who actually -- for home CAR T cells fail, the overall survival is quite poor. And this is certainly an unmet need. So when I was at the NCI, kind of before we even recognized this issue with relapses post CAR, in about 2015, we submitted an IND and embarked on a trial with a CAR targeting a different B-cell associated antigen, CD22. And as we opened the trial, we increasingly saw patients that were coming off CD19 CAR T cell therapy. We published the initial paper shown in the middle there, in 2018, in Nature Medicine. And what we demonstrated was that we were able to induce remissions in over half of the patients, the vast majority of those patients who had received prior CD19 CAR therapy. This CAR construct that was taken into clinical trials in lymphoma at Stanford, demonstrating comparable remission induction rates in a lymphoma population on the far left there and recent data published at the DOB Oncology Conference, and the updated data looking at all of the -- all CAR failures. They're looking at about a 50% 6-month CR in this patient population. So certainly, a CAR that is quite active in CD19 CAR failures. And so what we look forward to doing that is obviously expanding the allogeneic CAR T cell platform to CD22 with the program that we call 262. And so this is just some data, just demonstrating that we can incorporate the CD22 CAR into the hypoimmune platform and demonstrate activity. The model is a little different than the models that we've shown previously. It's now [ 6 ] tumor model. But instead of injecting entirely single population of tumor, we inject a mixed population with 1 component of that population lacking the CD19 antigen. And when we treat mice ingrafted with this particular mix population with CD19 CARs, as you expect, you get some early activity, you get clearance of the CD19-positive fraction and then you get outgrowth of the CD19-negative fraction. If instead you inject a CAR T cell targeting CD22 that is uniformly expressed across all of the now [ 6 ] tumor cells that are ingrafted in these animals, you get complete clearance of these tumors. So the second -- the third place will take the hypoimmune platform is into multiple myeloma. And certainly, as Steve mentioned, I mean, this is a huge unmet need with a large addressable market and a large number of patients that are unable to receive CAR T cells due to access problems, as I mentioned, the emerging data in myeloma is corroborating what we're seeing in lymphoma, which is that some degree of cellular persistence certainly matters in terms of durability of remission in these patients. So as you all are aware, there are 2 commercially available CARs. And one of the things that we're noting is that different than CD19, the quality of the binder seems to make a difference in the sense that we're seeing differences in outcome depending on the commercial CAR product that's being used to treat patients. And so as I mentioned earlier, we think it's very important to utilize a binder that has been validated in humans. And so the construct that we've in-licensed came from [indiscernible]. And CARs using this binder have been infused into over 100 patients with data presented at the last 2 ASH annual meetings. This is that data shown on the left-hand side of this slide, looking at overall response rates that are very comparable to the best-in-class BCMA CAR that is currently available. The other thing I'll point out is that when you look carefully using minimal residual disease, or MRD monitoring in the bone marrow, one of the things that's very important is looking at complete eradication of tumor cells in these patients after CAR infusions at early time points. And in addition to the response rate data shown in the graph, it's important to note that these patients actually achieved 95% MRD-negative remission induction. So again, we've taken this particular CAR construct, incorporated it into the hypoimmune platform, tested it across multiple tumor models. I'm only showing 1 here. Compared it to best-in-class commercially available control CAR construct. And across multiple donors have seen very comparable efficacy in multiple tumor models. So just in summary then, we think we're well positioned to deliver CAR T cell solutions for multiple populations of cancer patients. We've got a platform. manufacturing capabilities that can generate large numbers of high-quality products from single manufacturing runs. We're incorporating clinically validated CAR constructs, both CD19 and CD22 and then BCMA with an emerging pipeline of next-generation targets that we're currently working on with the goal of really generating off-the-shelf medicines at manufacturing scales that will address a huge unmet need. So in terms of upcoming milestones, as I mentioned, the IND for 291 is clear. The trial is open, and enrolling. It's received a fast track designation from the FDA. And really, at this point, we're just looking forward to early clinical data coming out this year, which we think will really give us a lot of information about the hypoimmune platform, for the first time in humans. And then looking forward to submitting an IND by the end of this year with 262, the CD22 CAR for CAR refractory patients and then the BCMA CAR expected in 2024. And so with that, I'm going to turn it over to Doug, who's going to walk you through the autoimmune program.

Well, good morning, everyone. Very nice to be here and very nice to be a part of Sana. I'm now in my second month in the company. And I have to say that as you're seeing this morning, the quality of the science that's being done at Sana is really pretty exceptional. And the people as well who really are moving that science along, including the folks back in the labs, many of whom I think are probably listening to the webcast, and I'm happy to be presenting their data as part of the presentation here today. So I'm going to talk about autoimmune and inflammatory diseases. I may lapse into the jargon of ANI. So that's what I mean, if I happen to say that. But I've been involved in this area for a very long time. Steve said 2 decades, it's actually 3, but I don't want to date myself. And I do look back to really what I think is the first renaissance in the treatment of these types of disorders, which was the launch of the cytokine inhibitors back in the late '90s, followed in really the same vintage as the B-cell depleters, rituxan being the first of a whole line of molecules that were very efficient at depleting B-cells, and have now moved into being mainstays of treating these variety of different autoimmune diseases. So just to reorient you again to the construct that we'll be taking forward here, SC291. This is a hip allogeneic CAR T cell. It has all of the edits and the knock-in of CD47 and the CD19 CAR. You heard Terry describe that in the context of the oncology studies we'll be carrying out. And the opportunity here is really enormous. This is a collection of more than 75 distinct diseases, some you've heard of, some you've probably never heard of. But many of them have an underlying B-cell pathology. I think that's been proven very clearly over the last 25 years. Lupus nephritis is one example. It's particularly pertinent in light of a paper that was published at the end of last year. It's a large opportunity with more than 100,000 patients in the U.S. who suffer from this disorder. And one of the great things about SC291 is that drug is available right now. We actually have material that's been manufactured, which could move very quickly, and allows for us to move very rapidly into this disease indication or disease indications, shall I say. So the evidence for B-cell depletion and autoimmune diseases, as I said, goes back 25 years to the launch of rituxan, really. I think that's where the story starts in a very clear way. And if you think about what's happened over the years, first, there was Rituxan, then there was ocrelizumab, and now there's obinutuzumab. Each of these drugs, each of these antibodies is a little bit better at depleting B-cells, and the efficacy has improved along with the depth of B-cell depletion. Well, there's really no better way to deplete B cells than a CAR T. If you want depth of depletion, a CAR T that targets CD19 is really the way to go to really maximize the depletion. And you can see on the right-hand side of the slide, some of the diseases where B-cell depletion has already shown itself to be an important treatment paradigm. A large number of diseases, and these are all life-long diseases. There's an accumulation of damage in these patients that occurs over a lifetime of not only the disease itself, but the treatments that we impose on them. So there's a huge opportunity here where B-cell targeting. We know that B-cell targeting is effective. So there's a road map sitting right in front of us in terms of where to go, what to do and the diseases to focus on. The -- I guess paper that got everyone buzzing in the field is really this paper from Nature Medicine that came out at the end of last year, autologous CD19 CAR T cell therapy in a population of very severe lupus nephritis patients. You can see the SLEDAI scores for the 5 patients who were enrolled in this study. All of them have very active disease on a collection of meds that really were unable to control their disease. Three months after the CAR T treatment, complete resolution of their SLEDAI scores. So remission is induced very quickly in this population of patients. It's quite remarkable the speed of onset of this type of major clinical effect in these patients. What was also quite striking is if you look at all the various biomarkers that have been used routinely, proteinuria, anti-double-stranded DNA antibodies, fatigue, which is really kind of a crushing, quality-of-life issue for these patients, all of them resolved almost completely within a few months. I think the encouraging side of the data as well is just how well tolerated the therapy was in this population of patients with only a little bit of CRS, fever in 3 of the 5 patients for a couple of days duration. So very benign from a safety perspective, no ICANS in these 5 individuals. And they're off all drug therapy now. So 3 months after the induction, if you will, with the infusion of the CAR T, these patients are completely off any other drugs and are in complete remission in terms of both the SLEDAI scores as well as all of the biomarkers, quite remarkable how these have all normalized. And we're now 18-plus months of drug-free remission in these patients with the last follow-up that was reported at one of the recent medical meetings. And what's fascinating here from an immunology perspective is that the recovery of B cells occurs about 100 days out pretty uniformly across all of these patients. But the recovery of these B cells -- these are naive B cells. They don't induce flares. They don't induce disease. It's almost that these patients that have a complete immune reset as a result of the CAR T therapy. So quite an exciting set of observations. A more recent publication that came out on an individual case report in anti-synthe syndrome. Again, with an autologous CD19 CAR T, you can see the decline in the diagnostic autoantibody that's associated with this disease within 6 months back below baseline. And then the inflammatory consequences of this disease. You can see on the MRI with complete resolution again in 3 months. So 2 diseases, 2 very quick remissions. And these are remissions in these patients. It's not clear yet how durable these are. But certainly, with the 18 months in counting in the lupus nephritis patients, it's quite encouraging that these will be long-term durable remissions in these patients. So our plan is obviously to utilize the existing SC291 drug supply. We think there's an opportunity, and we've seen others who have announced this to focus on sort of a basket study that is obviously to include lupus nephritis patients, but to really pick and choose amongst that laundry list of B-cell responsive autoimmune diseases the road map, as I described it, and to pick diseases where we think there's the greatest likelihood of success as part of that basket. The HIP technology, we hope the persistence and the ability to evade both adaptive and innate immune responses, we think may give us the opportunity to explore lower lymphodepletion. If that's true, it really opens the door to a much wider cross-section of patients if we can dial down that lymphodepletion dose as part of the preparative regimen for the infusion of the CAR T. And then, again, we think we're uniquely positioned from a manufacturing perspective to really be able to address what is a very large market and to be able to go out and produce materials to treat tens to hundreds of thousands of patients given the manufacturing technology investments that have been made already. So really an exciting opportunity. We're clearly not going from a standing start on this program, given that we've already filed the IND for SC291. So moving into autoimmune inflammatory diseases is obviously, something that we're quite excited about. And we'll have more to say about the specifics and the timing of the program as we get closer. But suffice to say, we see this as an enormous opportunity for the platform. Again, looking forward, if we focus only on the CAR T. portfolio, and I think Terry alluded to this already in his talk, there's a number of validated targets. To some degree, all we need to do is to swap out the binding domain in these CAR constructs, and we can iterate quite quickly through CD19, CD22, BCMA and others. That's the near-term opportunity in hematologic malignancies. Complemented, as I just said, by the opportunity in the autoimmune diseases, which is in its own right, an enormous stand-alone opportunity. And then as we look further out into the future, new targets, solid tumors, we think that this is sort of the gift that keeps on giving. And the investment we've made in the manufacturing technology will serve us well to be able to address all of these various markets as we go forward. So I'm going to switch gears now and talk about a completely different topic. And to some degree, a separate platform that's being developed at Sana. And this is the islet cell program, SC451 for the treatment of type 1 diabetes. I'm sure all of you are familiar with type 1 diabetes. It's a very large unmet medical need. It's a disease where we know the causation, and that's an autoimmune destruction of the beta cells in the pancreas, which results in an inability to control blood glucose levels. About 4 million people, just north of 4 million people in the U.S. and EU -- and there are no curative therapies. We can manage these patients to some degree. But certainly, cures have been elusive for this population of patients. That said, there's been an enormous amount of progress that's been made over the last, I guess, 100 years since the discovery of insulin. And that includes some of the tried and true things like dietary restrictions to manage the disease, but also innovation around diabetes -- or excuse me, insulin administration with pumps and real-time monitoring of glucose levels in patients. So a lot of progress, a lot of improvement in the standard of care. But this is a disease that's characterized by a number of long-term complications because patients can't perfectly manage their diabetes. And that results in stroke, heart attack, a variety of different consequences that develop later in life. And that's particularly true for folks that are diagnosed in the first 2 decades of life. There's a shortened lifespan in those individuals. The graph on the right is some recently published data, which really highlights the fact that the ability to achieve this target hemoglobin A1C level is really difficult to do. And we're looking at 2 different time frames here. What's odd to see is actually things have gotten worse, particularly in patients who are sort of in their adolescent years between 10 to 20, you see that blue curve showing the spike in hemoglobin A1C levels. And for reasons that are unclear, patients seem to be worse off in terms of managing their disease, even though all of these new innovations have come online. I think the path forward has really been identified here with the ability to transplant islet cells. That's really the ultimate solution to the problem. The current standard using allogeneic islet cells suffer from the fact that there's limited availability of the islets to begin with. So only a handful of patients really have the opportunity for this. But it's also a situation where these patients are subject to lifelong immunosuppression. And as Sonya mentioned already, that comes with its own set of consequences, but it's promising in that for some of the Phase III data that's been reported, roughly half the patients can achieve insulin independence. Obviously, the challenge for them is to stay on the immunosuppressive therapy if they come off, they obviously reject the graph and lose the ability to get the benefits of this transplantation. But this is the path forward, allogeneic islet cell transplantation ideally in the absence of immunosuppression. So we sort of see the problem, I think, in sort of 3 major categories. In order for allogeneic islet cell transplantation to become the norm, there's 3 barriers we've got to overcome. The first on the far left, I think [ Sonja ] very elegantly highlighted those. There's allogeneic transplant rejection. So the immediate rejection of the allogeneic cells. As well, there's the autoimmune rejection of the beta cells. The immuno cells and deletion of the beta cells in the pancreas. We have to be able to scale, right, in order for this to be a broadly applicable technology we need to be able to treat nearly 4 million patients or more than 4 million patients in the EU and U.S. And finally, there are improvements that could be made and probably should be made in the way in which the islets are transplanted. So the location of transplantation and trying to move away from this portal transplantation procedure, which has its own set of issues and limitations. So our solution, SC451, is an islet cell population that is allogeneic. It's derived from iPSCs. It has all the hypoimmune edits. And we think this solves many of the problems that are incumbent on the allogeneic islet cell transplantation field. You've already seen that as far as solving the immune problems of allogeneic transplantation that both allogeneic and autoimmune rejection, the preclinical data that [ Sonja ] shared with you earlier, shows that we may have solved that problem. And certainly, the preclinical data suggests that that's the case. So problem #1 appears to be under control with the hypoimmune edits. The second problem, which is scale, the ability to treat enough patients to really make an impact, we're focused on iPSC-derived islet cells. And the process here involves creating a master cell bank, sort of a tried and true technology in biologics drug manufacturing. In this case, what we'll do and have done is to actually edit those iPSCs to knockout HLA Class I and II and to knock in CD47 expression, the -- as [ Sonja ] described, the HIP modifications that we do routinely. Those cells can then be differentiated into islet cells and can form the basis of the transplant itself. So this is the method that we've developed to create these hypo immune glucose responsive islets that can then be transplanted. And the third leg of the stool here is implanting these in the intramuscular site that I'll talk about in some of the slides that are coming up. And we think this offers a lot of advantages for patients and really represents a step forward in terms of the implantation site and how simple that would be in comparison to portal administration. So I won't go through this data in any detail because you've seen it already. Again, our slide with the primate pops up again. It's an important slide here, and it is more evidence that we're able to overcome the allogeneic rejection. We've seen that in mouse models. We've seen it in nonhuman primates. We're now moving to the stage where we're hoping to see that in humans before too long. But suffice to say that there's an overwhelming body of data showing that we can overcome the allogeneic rejection in these preclinical models. The same is true when it comes to the autoimmune rejection. So the other form of rejection that's important to overcome in that first sort of barrier that I described in one of the previous slides. And again, I won't go through the specific data, except to say that there's very elegant mouse model that has been developed by [ Sonja ] and her team really highlights the fact that you can transplant these HIP-edited islet cells and see glucose control and see them circumvent the various forms of immune-mediated rejection, both autoimmune and allogeneic. So the second major barrier scalability. Again, the approach that we're taking is to create an off-the-shelf platform that allows us to make large quantities of islets that we can transplant into patients. We have an iPSC line. Actually, we have several lines. But these have all of the appropriate donor characteristics. They're compliant with 21 CFR. They have pluripotency in genome integrity. All of this is being very carefully monitored. We're creating GMP hypoimmune master cell banks. Again, these are the cell banks that allow us to basically carry out the necessary edits to create the hypoimmune modifications in those cells. We have a safe harbor site that's been very carefully selected to maintain the phenotype of these cells. And then finally, we are able to actually differentiate these iPSCs into islets and clusters that then form the basis after cryopreservation of the materials that will be transplanted into patients. So this is a very rigorous process. There's a lot of analysis that goes on behind the scenes here to look at both the genome stability of these lines as well as the phenotypic characteristics of the cells that actually represent the drug substance. And just a little bit of data from some of the laboratory studies, some of the early studies showing that when we differentiate these iPSCs, we can generate populations that have a high degree of beta cell purity, which is obviously what we're shooting for, a higher degree of endocrine purity. So a mix of the various different endocrine cell types that you typically find in the pancreas, they cluster into these islet-like clusters. And in vitro, they produce insulin in response to glucose stimulation. So phenotypically, we're able to derive cells that meet the criteria we're looking for transplant purposes. And again, when you do transplant them, they do work. And this is some data that [ Sonja ] showed a few moments ago in this in this model where we're able to actually transplant the cells see the C-peptide levels increase. And you can see in the far right-hand panel there, the glucose control labeled panel that we are able to control glucose levels in these animals post transplantation. So proof of principle, proof of concept that when transplanted, we can reverse the type 1 diabetic process in these animals. Safety. Always something that you think about with respect to deriving cells from iPSCs. We've been very careful to build in sort of multiple layers into this process. I think given the route with which we're going to use for transplanting these cells into the muscle in the forearm. Certainly, if there's any uncontrolled cell growth that's observed, the first step would be to simply remove the graft from the site of administration. There are other levels and other layers of protection built in as well, a safety switch that has been incorporated. That allows us to give an approved agent sort of dial down and kill off the proliferating cells over time as well as because we have high levels of CD47 on these cells, we're able to administer an anti-CD47 antibody. And then very quickly, you can see the removal of those cells following antibody administration. So really, 3 levels, if you will, of thought around making sure that we have adequate safety measures in place for moving this program forward. And last but not least, the third barrier we think to really making this available on a wide scale is the intramuscular site for islet cell transplantation. As you know, many of the prior studies have used portal vein infusion as the method of administering. This is an invasive procedure requires an interventional radiologist. So there's some limitations associated with access there. There are issues related to bleeding and portal vein thrombosis that are not insignificant certainly measurable consequences in this population of patients. And there's a phenomenon known as IBMIR, and you can read the definition there of what that is. But it's basically a process whereby complement activation causes sort of an immediate loss of up to half of the cells that are actually infused. So cell loss is a consequence of this process. And frankly, it's difficult to monitor the graft with this route of administration. So we think the ideal approach to this is the intramuscular islet cell transplantation approach. And we can deliver this more readily, certainly scalable for a larger population of patients, which we hope to access. We think that it will overcome this IBMIR phenomenon. And we can monitor this. We can monitor graft noninvasively. So they're easy to assess the progress of the graft, post implantation. And there is evidence clinically and certainly an abundance of evidence with the data in nonhuman primates that [ Sonja ] shared that this approach to engraftment actually works. So onto the investigator-sponsored study that we plan to initiate here soon. And the Phase I study is being carried out by very experienced investigators in the islet cell transplantation field. So a lot of skill in the art. And the process that we'll be following is actually pretty straightforward. I think Steve mentioned this at the outset, where we'll be harvesting donor cadaveric islets. They'll put them through their paces with the HIP modifications through the standard process and then transplant these into a type 1 diabetic patient without immunosuppression. Important to highlight that this is without immunosuppression. And back to the slide you've seen a couple of times -- in some ways, what we're doing with the IST is very analogous to what was done in this primate experiment that we've seen before. And what you can see here is the cell transplantation, dialing down the insulin levels. And these animals -- this animal is now insulin and independent with roughly 3 weeks of follow-up here, C-peptide levels normalize. This is very much what we're hoping to see in the patients that will receive these edited islets. So a little background work was done just to make sure that the editing process and the HIP modifications didn't have a negative impact on the composition of these islets as well as their ability to produce insulin. This was an experiment that's been done several times now. And if you look at the middle panel there, the colored bars, you can see that the distribution of alpha, beta and delta cells is equivalent across either edited or unedited cells that were just manipulated in the same way through the process. And in both cases, on the far right, you see very robust insulin secretion in both the wild type as well as the HIP islet cells. So quality, if you will, is maintained even through the processing and the hip modification. So we're on track to actually initiate this study later this year. And the goal of the study, very straightforward endpoints is to look at cell survival and immunoevasion to look at C-peptide levels in these patients and hopefully to see some enhanced glucose control. And again, this is against the backdrop of no immunosuppression. So a lot of progress in this program and more work to be done, but we've demonstrated now the ability to deal with both the allogeneic and autoimmune rejection phenomenon, demonstrated activity in relevant preclinical models, the manufacturing process is well underway in its development. And again, I mentioned already that the IST will start later this year, and we hope to have some preliminary data. So very exciting times for the diabetes program here, anticipating that SC451, the actual product candidate, will have an IND filed sometime in 2024. So I'm going to turn it over now to Steve Goldman. Steve, as was mentioned, is involved in glial progenitors. And in my past life at Biogen, we had a lot of interest in the MS franchise and trying to come up with approaches to be able to remyelinate. We were trying to do that through pharmacologic approaches, antibodies, small molecules, going after oligodendrocyte precursor cells. I think what you'll see from Steve is light years ahead of that kind of work and really an opportunity to reverse some of the damage that's been done in a number of not only autoimmune diseases that affect the brain, but many other diseases of the brain that involve myelin as well. So Steve, real pleasure.

Thanks folks. So I'm a neurologist and of course, I study brain disease. In my case, exclusively brain disease. But neurology by definition, study of neurons, right? And neurons, of course, are the primary conductor cell type of the brain, allowing impulse to go from 1 part of the brain to the next. And so that tells you something right there that neurology, the study of neurons and brain disease. And so the assumption was always historically that brain diseases were from loss of neurons or dysfunction of neurons. What we've come to appreciate really in the last 10, 20 years is that really the majority, perhaps of diseases of the nervous system are primarily diseases of the support cells in the Brain glial cells. And that's what the focus of the CNS group at Sana is the study of the glial cells and replacing glial cells as a means of treating brain disease. And so really, by way of education, if you will, for those of you who aren't familiar with the brain, basically neurons are the conductive cell type of the brain. But there are 2 primary support cells in the brain. They are the oligodendrocytes the astrocytes. The oligodendrocytes to the myelinating cell type that Doug just mentioned. Basically, they provide myelin, which is the insulating sheath around axons. Axon's being neuronal fibers that allow impulse to be conducted from 1 neuron to the next. And astrocytes are the support cells of neurons that provide direct metabolic support and also maintain the blood brain barrier, basically allow fluid transport in the brain. And so these 2 cell types are both derived from a common progenitor, that's the glial progenitor cell. The glial progenitor cells by potential. It persists in our brains, right? And together, these cells are actually more abundant than neurons. So in the adult human, depends on what part of the brand you're talking about, but the majority of the cells are glial cells. And -- but basically, we tend to run about 1/3 astrocytes, about 1/3 of oligodendrocytes, about 1/3 neurons across the brain overall. But there's a substantial proportion, about 5% of the calls in our brains are persistent glial progenitors. And so I'm going to really discuss is the use of the glial progenitor cell is a therapeutic vector. And the bottom line is at this point, we know that the glial cells that derive from the glial progenitor really formed the basis for most of the myelin diseases, many of the neurodegenerative diseases, we can -- we've learned how to make these cells to produce them under GMP conditions at scale and purity and have looked at them in a broad variety now of diseases where we see the potential to replace endogenous disease glial progenitors and their derivatives, which essentially fresh healthy cells that then allow the production of new astrocytes and oligodendrocytes in the brain. And so that allows a really broad platform -- and a broad set of diseases to be targeted with 1 common product. And that's the theme of everything I'll show you now. And so as I mentioned, the glial cells occupy most of the brain, the oligodendrocytes are primarily in the white matter of the brain, okay? And that the white matter underlies the cortex. The cortex is based -- is essentially comprised of neurons but also of astrocytes. Astrocytes sites are also though in the gray matter. And so astrocytes in the gray matter and the white matter, oligodendrocytes in the white matter. The astrocytes sites of the gray matter are supporting synapses or supporting the connections between neurons. And so what happens as a result is if there is a neuronal disease that also involves glia and we replace the glia that can rescue the synapse and therefore, rescue the neuron. And that's another theme that we'll -- what we'll focus on. So it's a broad set of diseases that are diseases of myelin. First and foremost, the pediatric leukodystrophies, the storage diseases and hereditary diseases, congenital where myelin doesn't form properly. These are relatively uncommon. But as a group, they're actually the most common set of neurologic diseases of children. And unfortunately, most of them are lethal with very few treatments available for any of these diseases on this list. In adults, of course, we think in terms of multiple sclerosis as a common neurologic disease and progressive multiple sclerosis, which essentially is a phase where the patients essentially develop a neurodegenerative disorder downstream of losing myelin. Because myelin and oligodendrocytes are required for neuronal viability as well. But then even the white matter diseases of the elderly diabetics of hypertensives, all of these represent a loss of myelin that results in significant neurologic dysfunction. And so it's a broad category of disease. But over and above the myelin diseases, we have the neurodegenerative diseases, things like ALS, Huntington's disease, frontotemporal dementia, which we've always thought it was neuronal but in fact, there's so much synaptic dysfunction because of the astrocytic disorder in these conditions that if we replace astrocytes, we replace actually much of the synaptic function neuronal viability. And so it's a category of disease that we look at favorably in terms of using, again, this 1 common cell type of glial progenitor to treat. And so this is our first use of the glial progenitor, if you will. We isolated these cells initially from fetal tissue, and long since I've moved on to stem cell derivatives, which we'll get into. But this is just a proof of principle where we've taken glial progenitors and transplant them into neonatal mice. So these are human cells, and it makes the point that these cells are extraordinarily migratory, right? It's a developmental phenotype where these cells are intended to be in a developing brain, and their job is to migrate through the brain, right, and set up shop. And that's what they do. If we put these cells, in this case of neonatal mouse, these are all human cells. Every red dot here is a human cell in a single 40-micrometer section. And you get the idea that these cells really take over. And in fact, they do. So in this case, we've gone into a hypomyelinated mouse. So hypemyelinate mutant called the shiverer mouse. That doesn't make any myelin. And so no myelin -- in these animals, we'll see die young, very quickly, typically by about 20 weeks. We can rescue these animals by transplanting these cells in neonatally and in this case, we end up with these mice where all the cells, the donor cells, have migrated through. They've taken over, they've literally kicked out the deficient population and then they go ahead and myelinate and this is all human myelin, which is interesting, makes for an interesting disease model because we can look at the effect of basically different human disease-derived cells in these models as well. That's a topic for another day. But the bottom line is we can completely re-myelinate these nervous systems and these animals are rescued. This is an example of that. Nick, if you could turn this first one. Thank you. And so here's what shiverer mice look like. This is at 18 weeks of age about 2 weeks before they die. And they die like clockwork typically. And what happens is as these animals are growing older, they get larger. And as they get larger, exons get longer. And since those actions aren't myelinated, there's more and more conduction failure. So first, they become a toxic, imbalanced lose truncal stability, then -- and they developed leg weakness versus the high limbs and the fore limbs and they start seizing and then they die of status epilepticus typically recurrent seizures. Next over here -- and this is after transplant, and this is 13 months, we looked at this guy, he made it up to 2 years, he was fine. And what's remarkable here is not just the rescue of the animals, but early on, these animals the transplanted animals, develop a neurologic dysfunction, takes a while for the grafts to take. But then they slowly but surely get better and then all better and then back to normal and rescued. And we have every reason to think that that's what we'll see as we go to patients as well. We've - tractly looked at these cells in rat models and squirrel monkeys, where we see the same type of effect. And so because of that, we realized that, okay, this -- these cells really have therapeutic potential. And the use of fetal tissue would never be really scalable in any reasonable fashion. So some years back, we developed the methods for making these cells from pluripotent stem cells, both from iPS cells and embryonic stem cells. And here, we're looking at what these cells look like in this 7-step preparation protocol, and we've developed that allows us to make really highly, highly in rich population of glial progenitors from embryonic stem cells or iPS cells. It's a process that takes a while. It's a several month process. When we first published it, it was out to about 160 days. Now we have that down to about 120, but it's still a long process -- that's one issue. But the flip side of that is, by virtue of the long differentiation, it turns out that it's an extremely efficient population at the end of the day, and that almost all the cells are glial progenitors, those that are not or fibrous astrocytes. There's nothing left in these preparations by way of undifferentiated cells. And so that gives us a lot of possibilities in terms of manipulating the cells at the iPS or EF stage, whether putting in hypoimmune edits, which are a little different in the brain than the systemic edits that have -- that Sonja has discussed. But it's the same idea. We can put in hypoimmune edits. We can engineer for competitive advantage and the safety switches that can be put in as well is needed. But basically, it's a cell population that we can edit upfront if needed, and it's not necessarily needed and our first go around, we'll be using really unedited cells. And yet -- and you'll see that that's more than sufficient in terms of both safety and efficacy. But of course, we are looking for the potential for really manipulating these cells further for -- in a disease-specific fashion down the road because there is such a long list of diseases that potentially this 1 product can treat. The cells really are remarkably safe. And it's because of that long differentiation process and the efficiency by which these cells become glial progenitors. And so these are single cell RNA plots, where we've looked at and differentiated ES cells in this case. And then the cells after they have been -- and this is a different cell population, of course, with a match cell population after they have generated glial progenitors in vitro. And you see the really complete lack of overlap of the gene expression patterns, right? In terms of the glial progenitive stage in vitro versus the cells from which they derived, right? And again, it's about a 4-month derivation process. But then we went further, transplanted them and then took the cells back out of the brains and then do the single cell RNA sequencing of the human cells that were removed back out of those brains, and there's essentially no overlap whatsoever of the integrated gene expression pattern in vitro versus in vivo because there's a lot of context-dependent differentiation further after these cells are transplanted. And certainly none in terms of overlap with the undifferentiated ES cells. And so accordingly, we have looked at hundreds and hundreds, probably well over 1,000 animals with -- that have received grafts from this cell line. I've never seen anything malignant, anything by way of undifferentiated tumor genesis. Yes. So this is where the cells look like when they come out. In terms of basically the composition, the cells removed back out of these brains, the human cells, again, basically extracted back out and then analyzed. And basically, all the cells are either astrocytes or oligodendrocytes or persistent glial progenitors cells. And those persistent glial progenitive cells provide advantage because they serve still as a reservoir for making new astrocytes and oligodendrocytes as needed. And so everything I've shown you up to now, though, basically were comprised of graphs of human cells into mouse, right? And the human cells are taking over the mouse. What happens if human goes in the human? And of course, we want to be able to mimic the clinical situation. And what we find is that -- and this is quite remarkable, we, in this case, set up a chimeric brain with human Huntington's disease mutant glia and then transplanted in 9 months later, 9 months later, healthy cells. And these healthy cells then outcompeted the Huntington's mutant -- Huntington's-expressing glial progenitors and literally replace them over time and essentially kicking them out. And that results in the human on human replacement, which tells us it's predictive that this should work every bit as much in people going human into human as, of course, in these experimental models. So it's well published set of studies in terms of the use of these cells for treating the myelin disorders, but also in the astrocytic disorders of the neurodegenerative diseases, Huntington's, say, even frontotemporal dementia schizophrenia in all these cases, we see the replacement of host cells by donor and that can -- basically effectively replace that disease cell population with new cells. And so the opportunity is to treat a broad set of diseases with 1 product, but we're focusing initially on just 3. And they include Pelizaeus-Merzbacher disease, which is a relatively uncommon disease of children, where it's the closest thing in clinical practice that we'll see to a shiverer mouse, where basically these poor kids don't make myelin. They make a very dysfunctional myelin, I should say. And we're looking to potentially replace the dysfunctional oligodendrocytes of these children with new oligos made from the glial progenitor. Secondary progressive MS as an adult target. And in this case, we're both trying to put in new oligodendrocytes and also getting some degree of repair of lesions that have already occurred. And then Huntington's disease as largely an astrocytic disorder that also involves dysmyelination, where we think we can replace these populations and effectively treat the neurodegenerative disorders as well. And just running through these really quickly, Pelizaeus-Merzbacher. We see this loss of myelin, okay? So here, we're looking at it another image of the brain, but it doesn't have the kind of white matter we saw earlier on, right? It's very, very thin white matter, and this is what needs to be replaced in these kids -- patients with chronic progressive multiple sclerosis. And you can see normal brain, but then after the initial stages of relapsing remitting MS, we see some loss of myelin. But what happens is in relapse and remitting MS patients progress to a later stage of progressive multiple sclerosis, secondary progressive, which essentially transitions to a neurodegenerative disease, right? Because neurons are lost because oligodendrocytes aren't there to support them. And this is the phase of secondary progression that we're trying to rescue. We're trying to prevent this ongoing loss of neurons due to the loss of oligodendrocytes at the same time potentially to reverse the deficits of these patients. And then Huntington's disease as a -- really, in this case, an astrocytic therapeutic target because by, begin, rescuing the synapse by virtue of putting in functional astrocytes, we see the ability to rescue the neurons that are at those synapses as well. And that's something -- again, we've gone from a number of different disease models at this point that are validated at that point. This is what a normal brain looks like at the same level as the Huntington's brain. And it's a remarkable degree of atrophy that these brains undergo. But first and foremost, there's loss of the Striatum, which is the putamen and caudate. These are the basal family of the brain, the motor control regions. And it's really by injecting cells early on that we expect to rescue these brains. And so basically, we're looking at a really encouraging set of studies at this point. Everything we've done. Animal models has played out well, and at the same time, the -- we have scaled manufacture of these cells, certainly enough to entertain a number of IND enabling -- of IND and Phase I studies at this point, but we're undergoing the IND-enabling studies at this point. And our goal is certainly an IND, but perhaps several INDs, but based on the same product in '24. And this really cuts to the chase. We are all undergoing age-related white matter loss. It's one of the major causes of adult dementias. Alzheimer's gets more credit than is due in that sense because about half of all adult dementias come from loss of white matter. And aging itself is a disease when you get down to it. What we find is that young cells will outcompete old cells, young glial progenitors put into an isogenic brain will outcompete older. And so we see the ability to potentially replace these cells in an ongoing fashion in a broad variety of diseases as well as ultimately in the setting of aging. And I'll close with that. And next is Richard Mulligan. And so when I was a grad student at Rockefeller, this guy was one of my heroes. And so that's already -- you really were. So that's a testament to -- and you are that much older than me. But that's a testament really to the science at Sana. And here we go. Thank you.

Okay. We're going to now turn to something very, very different but equally interesting. As Steve -- we have from the get go had both the HIP technology platform and importantly, the in vivo -- platform. And what I'm going to do is just introduce you to the -- platform really giving you the "why?" Why is this interesting? And how do we do this? And then the final stuff comes after me where Terry and Ed will tell you that this really works and really works very powerfully. The why is very straightforward, for someone like myself who's been doing this for a longer period of time, we don't need that anymore. Longer than before some of you guys were born, the evolution of gene therapy has always been to kind of -- the aspiration to treat more patients, more effectively, more diseases. And what we've seen is a natural progression of the 2 approaches to gene therapy the ex vivo gene therapy and in vivo gene therapy. And the ex vivo approach basically exemplified by removal of -- from the patient transduction of them in the -- and Re-transplantation of them. It's been very, very powerful. Terry has always -- already gone through the whole value of the technology for CAR T cells. You've also heard the value of the technology for treatment of diseases like data thalassemia, sickle cell anemia. Terry has also showed you the next evolution of the ex-vivo therapy to go away from autologous cells to allo cells, increasing the accessibility, the number of patients we can treat. But nevertheless, ex vivo technology has also -- always been looked upon as limited in its scope of application. Not only is there the need to manufacture the drug product in a way outside the body, but there's also just a logistical problem getting enough cells to treat enough patients. In vivo technology has been very, very powerful as well, but we're now beginning to see some of the limitations. And that's really -- it's those limitations that have caused us to be very, very interested in the technology I'll talk about. In vivo gene therapy has been very successful, particularly using vectors like AAV and LNPs. AAV, very useful for treating a number of rare diseases for local delivery of -- factors and so forth. LNP is very, very valuable for local delivery of vaccines, as you know, and also for the treatment of diseases of the liver because LNPs largely go the liver, but as the in vivo therapy feel it has evolved, you see the limitations. And the limitations are essentially that, you really more than ever are getting the sense that we really need to direct in vivo gene therapy to the desired target cells. In the case of AAV, for instance, to try to treat a disease that's -- diseases like muscular dystrophy, we found that you really have to push the system. You have to push very large amounts of the gene delivery agent. And what you see as a confluence is you're hitting cells other than the cells you wish to hit, in particular, liver cells. So if one can get a better AAV system, we'll have a better therapy, okay? In the case of LNPs, they're not exactly amenable to, at this point, to delivery of genes or payloads outside of the liver. People are working in both AAV and the LNPs to give this technology, specificity. But we're not there yet. So there's a safety issue of not getting into the proper desired itself. From a regulatory point of view, I think people are becoming more and more concerned about the off-target effects, the problems with gene editing that are surfaces about lack of specificity and so forth. So it's very, very important to all of us think to be able to limit transduction, limit the payload delivery to the cells of interest. A second major limitation of existing in vivo approaches is when you get into non-desired cells, you can have immunological consequences. And now a clear understanding that if you introduce genes, for instance, immunological and to preventing cells, you have the opportunity to mount immune response against the Trasgene and the vector itself. And so those are a couple of things. Lastly and probably most important for a gene therapy guy like myself is the potency is very much affected by the inability to, in fact, transduce only the target cell of interest. That is to say, the body provides a sync for the delivery. And if you want to get the virus in a certain location, you don't want to go into another location, like in the case of [indiscernible], you don't want most of the virus to go to the liver and not them use the muscle, the cells of interest. So the system that we've developed takes off on this principle, and it's absolutely designed to address these issues, to get absolute cell-specific gene transfer and this higher cells of interest. And so now the why -- or the how. The how is basically to take a page our of the playbook of envelope viruses. So envelope virus, as you may be aware of, because they're the progenitor of the lentiviral vector systems that are used for many ex vivo active patients. And what's showing here is that the way in which these lentivirus vectors and envelope viruses are generated and infect cells. And what we're going to show you is how we're able to use this general approach -- this general mechanism of gene transfer to gain absolute cell-type specificity. Envelope viruses are produced from infected cells in such a way that they uniquely take off pieces of the cellular membrane with them. And on the cellular membrane are presented viral proteins, envelope proteins or fusogen proteins as we call them, that predominantly almost exclusively dictate the kinds of cells that can be infected. Shown on the right side, the virus particle generator, the envelope viruses, or as we call them, the fusosomes, are then able to infect cells solely on the ability of that envelope presented on the surface to interact with a cellular receptor on the cell to be infected. Now notably, most envelope viruses have envelopes or fusogens that allow this viruses to infect a wide variety of cells. However, interestingly, in the development of lentivirus vectors and retrovirus vectors, it's been possible to demonstrate that you can shuffle these envelope proteins or fusogens. That is if you express alternative envelopes on the surface of infected cell, the virus is called viral serotypes that come out cell then we'll have the specificity of the new -- And so what we've been able to do, is to reengineer viral envelope proteins, so they will have specificity. And I'll just very briefly show you how that occurs. People have attempted this for a number of years. In fact, I hate to admit, but we tried this maybe 4 years ago. In my early days, when I knew Steve I guess. And it never worked. We weren't using the right envelope protein. The innovation here was to use a 2-part envelope system, a fusogen system where 1 component dictates the attachment of the viral leg particle to this. And the second component allows the fusion to occur that's necessary for the virus to enter the cells. And the reason this is important is many have found through the years if you use a single envelope or fusogen protein that has to do both of those tasks, recognize the recipient cell and induce fusion. It's hard to kind of change the specificity without marking up the efficiency of the cell fusion. So we've used an exotic paramyxovirus fusogen that can -- is consists of a G protein or attachment protein and the -- protein of fusogen protein. And what we've done to confer specificity upon this fusogen is to first mutate the receptor binding domain of the G protein so that G protein no longer able to recognize as normal receptor. So it's blind and it can't see anything, so it can't allow infection to occur. And then what we do is we have panel 2 the mutated G protein, a protein sequence that confers the ability to recognize the desired antigen on the recipient cell. Now that antigen is the antigen that specifies the type of cell that we want to infect. So in the cases, you'll hear about fun-- We're talking about very well-characterized antigens specific for desirable cell type, the CD4 cell, a CD8 cell, a CD3 cell. You'll hear later from Ed a more complicated system where there isn't 1 antigen that you can choose. But the point of the story is that basically, although -- this is designed to give you a complete specificity, you're only as good as the antigen that you pick. And so the next slide just points out that remarkably, I think remarkably anyways, it works extraordinarily well. And well, it's not a plug and play system. It's near that with good protein engineers, which we have -- we're able to take a number of different scaffolds that bind desirable proteins and append on to the mutated G protein in such a way that a new protein will probably recognize the desired antigen on the recipient cell and still has the fancy footwear necessary to interact with F protein to elicit entry cell fusion. Again, I didn't probably make that clear, but normally the way this system works is after the G protein recognizes its receptor, it tickles the F protein a very sophisticated way that allows the entry the curve. So in summary, we have a nearly plug and play system. It's very versatile in terms of the types of antigens we can argue, and I've shown here a list of some of them that are of interest to us. It just shows that we can do this. Takes a little bit of work. We can almost always get things that work very well. It worked very well. It's all about potency. I'd like to say, it's a titer, stupid. It's very, very important to have very potent patterns. Some of you go back, the analysts in the audience and go read papers for the last 20 years, you might say, hasn't this been done before? That kind of looks like I see these papers. They're all in human gene therapy if you want to go look at them. And the problem is they just -- they don't make the grade. They're insufficient to out in vivo gene therapy to occur. The system we have is so damn potent, it really approaches the kinds of potency that traditional lentivirus vectors, like Terry has talked about, that allow very high titers and a very efficient entry in to the cells. So summary system works, works extremely well, is specific, applicable to all kinds of things. The final thing I'll point out is really an introduction to Ed's talk, in fact, is that the fusosomes platform is very versatile in terms of the kind of payloads that delivery vehicle can introduce. Some of the issues with AAV as an in vivo delivery vehicle and also LNPs is they're not really amenable to more sophisticated gene-editing payloads. They're not necessarily amenable to putting in DNA, transferring DNA. They're not necessarily amenable to putting in RNA, just RNA. And because of the nature of the virus-like particle, very much like a lentivirus particle, and the sophistication about how you generate these package -- well virus-like particles, we are able and we have demonstrated these delivery of not only DNA which, again, is what Terry is next going to talk about, standard lentivirus, like which that allows a proviral integrations of cells, but also -- editing protein, editing agents and a mixture of all 3. And so there you have it. I'm going to turn over -- the mic to Terry, who's going to talk about the traditional DNA fusosome, and then we'll hear from Ed.

You guys can hear me? Okay. So I'm actually -- it's pretty exciting to stand up here and tell you about the progress that we've made in this program. We've been talking about this program for a while. And I think what you'll see is that we're really in a position now to test this approach, this novel gene delivery platform in humans with a real potential to see activity. And I think not only certainly, would this be important and critical in terms of validating this platform that Richard just described. But I think if we're successful in the CAR T cell space, this would certainly be transformative. I mean this is a novel platform. This could be hard. It's going to take some time, but this would really, really transform the field because of a couple of things. One, you overcome all the complexities to the manufacturing with this -- the ability to deliver a CAR gene in vivo with a single intravenous infusion. The other component is that we're going to do this without lymphodepletion because we don't want to give the lymphodepletion to damage the T cells inside the patient. And then the third element of this is that because we're not removing T cells from the body, exposing them to high concentrations of cytokine activation, et cetera, we would expect that the quality of these T cells would be superior to what you see with ex vivo manufacturing, and potentially would be a reason why we might see activity even in the absence of lymphodepletion. And so what we're talking about here is really with a single intravenous injection of this fusosome particle that we deliver a CAR gene to T cells in vivo resulting in expression and then all the activities that we associate with CAR-positive T cells with ex vivo manufacturing, that is to say, activation, expansion, eradication of tumor. And so with the initial programs, the gene will be delivering. It's a CD19 CAR, a well-validated CAR, and that's deliberate because really what we'll want to do is to isolate the risk of this program on the fusosome itself. That is the ability to specifically and efficiently deliver genes to T cells in vivo. So as Richard mentioned, we've been very successful at identifying fusogens that is, say, the protein that actually allows the fusosomes to target cell types -- and this is an example of sort of a campaign to identify binders. It is the case. And certainly, there's a variability in potency. That is to say here, in this case, the ability to deliver GFP to cells. And so this allows us to identify optimal binders. We can also select binders based on cross-reactivity, for example, in nonhuman primates that allows toxicology studies that I'll talk about in a few minutes. We can then take these binders, put them onto fusosomes and then package a CAR gene and then test these on nonclinical studies. And so this is sort of the -- an example of one of these models, where we're delivering a CAR gene in vivo. The difference in this model is that the human cells that are ingrafted in these mice, the human T cells that are ingrafted in these mice are not gene modified. And so these mice then receive a single injection of the fusosome and then we're able to measure CAR T cells and a CAR on the surface to these T cells. And then by gating on the different T cell fractions, C4, CD8+ T cells, we can demonstrate a couple of things. One, specificity since we don't see evidence of CAR gene in the CD4+ compartment. And the second is that we deliver genes to CD8+ T cells in a dose-dependent manner. We can then look at the ability for these cells to eradicate tumors. And what we find is that across multiple donors. We've done lots of these types of experiments with multiple CARs. You can see that the CAR T cells are highly functional, and able to control tumors the same way as ex vivo CARs. I will point out that as one might expect, the kinetics of this is a little bit slower. It takes longer to see control of these tumors, which is expected because we've got to generate the CAR T cells inside the mouse. The binder that we've selected for the lead program, 299, cross reacts in nonhuman primate CD8. And still again, it allows us to do nonhuman primate studies. We've done a number of these studies. This is an example of one of those studies with an earlier manufacturing process in which we were able to demonstrate depletion of B cells. Importantly, the reason this is a CD20 CARs because the CAR the CD19 CAR that we're utilizing doesn't cross-react to nonhuman primate B cells. So this surrogate allows us to measure pharmacodynamic effects, that is to say B-cell depletion in these animals. And what we see is B-cell depletion in a subset of the animals, 4 out of 6 of the animals. And importantly, when we looked at later time points, this is immunocytochemistry for CD20+ B-cells in secondary lymphoid tissues. We see sustained depletion lasting out this necropsy in this animal shown on the right, was done at almost 60 days after fusosome infusion. The other thing that we're able to do is to use insight you approaches to actually directly measure CAR gene in secondary lymphoid tissues. And what I'm showing in this slide is lymph node biopsy is day 9 and then spleen at day 35. CARs detected using 2 different imaging modalities, standard IHC, the brown and then immunofluorescence, where you see the red CAR+ cells in these tissues. And what you see is evidence of CAR+ cells or gene transfer in secondary lymphoid tissues, day 9, day 35. Importantly, we have not seen, using this particular modality, evidence for CAR gene and liver, an organ that we would expect to see large amounts of the systemically administered fusosome. The other thing we can do is use co-staining to confirm that the cells are CAR positive or actually the target cell type that we're going after. And this is an example of one of those staining protocols where we can demonstrate the red on the far right there, which is the CAR gene, localized to the CD8+ cells within the secondary lymphoid tissues, again, demonstrating specificity. So as Richard mentioned, it turns out that it's difficult to make this material at very high concentrations or potency. And so one of the areas that the manufacturing folks within Sana spent a lot of time working on is improving the manufacturing process. And they've made really outstanding progress in this and have -- this has resulted in improvements in really 3 different areas: potency, yield and purity. A lot of the impurities in the product comes from the producer cell line, both whole cell DNA and whole cell protein. And importantly, this is now been consistently observed and observed at scale -- 200-liter bioreactors, we're seeing 50-plus x improvement in titers. And I want to take a moment here before I go to the next slide and just remind everybody that Richard talked a lot about potency as being the most important element here. For the drug product, the potency is defined as the ability to transfer genes into cell lines, right? That's an analytic that needs to be done to actually measure functional particles. But the important component that you want to then look at is how does that functional potency defined as transfer into cell lines to translate into transfer of genes into your primary cell of interest, in this case, T cells. And what we've seen with the manufacturing improvements is at a given dose in this case, defined as a concentration of particles per T cell, we actually see improved gene transfer with the new manufacturing process relative to the old manufacturing process. And that's shown on the left, across a range of IUs per PBMCs, and then the right across 5 donors at a single MOI that or a IU per cell. So again, with both the improvement in titer, the potency of the drug product itself and the improvement in gene transfer efficiency, we're looking at a couple of things. One, because of the 50x improvement in titer, it means that we can deliver the same amount of functional particles in proportionately less volume, right? Which is critical in terms of dosing patients. And then the second thing is that we see improved gene transfer efficiency. Collectively, right now, we're estimating that we can potentially generate equivalent numbers of CAR T cells to what you see delivered with ex vivo CAR T cell manufacturing. So we're on track to file an IND for this program in 2023. The manufacturing runs are underway. The GLP toxicology study in nonhuman primates is on track for completion. The Phase I study in which we'll go after acute lymphoblastic leukemia is our first indication -- has been -- is in the planning stages. And again, I already mentioned that last bullet point that we're anticipating that we're going to be able to deliver comparable CAR T cell dose to ex vivo manufacturing. So just very quickly, I'm going to walk you through some areas that we're looking at beyond this program, certainly, interested in targeting other T cell receptors or other T cell types, delivering other payloads or different CAR genes. And certainly, thinking about the potential here, in autoimmune disease, where one might expect to see biologic activity. We already know that the CAR T cell dose with ex vivo CAR T cells seems to be lower. And so there's a potential that we actually might see that autoimmune disease might be an optimal place to take this after we'd certainly generate safety and data in humans. And so just quickly, this is just data demonstrating that we can deliver CAR genes to other T cell types in the same mouse model that I described earlier. In this case, using a CD4-targeted fusosome, demonstrating that CD4+ CAR T cells are able to eradicate tumors in this non-human -- in this mouse model with NUM6 and then second thing that we've looked at and not a surprise that we can swap out the genetic payload deliver other CAR genes. In this case, a CD22 CAR that I've already described earlier in the allogeneic program, and demonstrate the ability to control tumors in these animals. So in terms of upcoming milestones, we've demonstrated the ability to deliver genes specifically into CD8+ cells with minimal evidence of off-target effects. We've shown tumor clearance in the preclinical models across multiple models and multiple donors. We haven't seen safety concerns in the nonhuman primate studies we've conducted. We've seen market improvement in potency and the ability to target additional T cell subsets and to deliver different CARs. And we're on track now to file an IND this year with the expectation that we'll be able to generate early clinical data next year. So with that, I'm going to turn this over to Ed Rebar, who's going to close this out.

Okay. Great. Thanks. Coming through. Okay. So I hope you all enjoyed Terry's talk and are as excited as we are about our ability to generate CAR T cells in vivo. We've been very pleased with the performance of our fusosome platform for doing this high-efficiency gene transfer. But our aspiration is to go beyond that as both Richard and Steve indicated in their talks. What Sana aspires to is the ability to do genome editing, therapeutically in vivo. And what that means is basically the ability to go to a patient, go to a chosen cell type, point to a spot in the genome and say, we want to make this precise change right there, and that's what the slide indicates. And if we can do that, right, we could open up a wide variety of new types of indications that we can go after. And also, we could take current indications. So it's a small number that other entities are addressing that are being performed ex vivo and bring them into the in vivo setting where you could dramatically improve patient access. So this ambition is not without its challenges. I'd say the 3 key ones are indicated here. Those are achieving cell-specific delivery. This is critical, not only for safety but also potency and manufacturability. And I think Steve and Richard did a great job of describing that earlier. Also delivering the appropriate payload type, right? So gene editing can require delivery of protein, RNA, DNA or some combination of those entities. And also delivering sufficient editing reagent, right? Ideally, what you'd like is for 1 delivery event, 1 unit, a particle or a virus to be able to reliably edit that targeted cell. And what's really exciting about fusosomes is that they seem to have the capacity to address all of these considerations, right? So in terms of achieving cell-specific delivery, Richard described very, very nicely how we're able to retarget fusosomes to go to cells of our choosing. And I'll provide some data in a few slides showing how we can make that be a very highly specific delivery event. In terms of delivering the appropriate payloads type, fusosomes have the inherent capacity to package both RNA and protein. And by virtue of this natural process of reverse transcription of the packaged RNA, DNA as well. And so they just inherently allow you to deliver any type of molecule you might need for -- do editing. And finally, in terms of delivering sufficient editing reagent, fusosomes have a size and a capacity that's larger than other options, which allow you to deliver more editing reagents as well as larger gene templates that might be available, say, via AAV. So I'm just going to spend a couple of slides on just making maybe a little more clearly on how gene-editing fusosomes differ from integrating fusosomes. So basically, at top is the integrating fusosome. This is kind of the archetypal parental fusosome. This is what Terry described in his talk. These particles are very good at integrating transgenes randomly in the genome. On the bottom is gene editing fusosomes and what those are, are specialized for packaging and delivering gene-editing reagents. And so what they offer is the capacity to make specific edits at chosen sites in the genome. And it's this capability that allows them to address potentially a much wider range of application types. Now I don't want to disparage the parental fusosome. It's actually very good at what it does. And if you want to do something like make CAR T cells, I think it's the best platform around. But if you want to do something more site specifically like knock out a specific gene or correct a chosen mutation in the genome or integrate a transgene at a known location for some purpose, what you'll want is a gene editing fusosome, and you'll have the capacity to do more useful therapeutic things. Okay. So with this rationale, Sana has begun adapting fusosomes for gene editing. And we focused initially on nucleases and base editors. I think those are the most advanced editing reagents for therapeutic application. Here's an early study where we just tested our ability to package these reagents and fusosomes. The experiment was -- we ran a production run. Took some of the sort of crude, unpurified supernatant from those producer cells containing the fusosomes and then titrate at these reagents onto tester cells just to see how well we could either edit in the case of nucleases or convert a base in the case of base editors, those target [indiscernible] And these are the results in titration curves, and this indicated that we can make very highly-potent particles. It's not shown on the slide, but if we ran a similar study using our integrated fusosomes and just asked for gene transfer, we can generate highly similar curves. So we do well relative to that benchmark. We've also taken these fusosomes and have started performing in vivo studies with them. And this is one early such study where we looked at knockout, our ability to knock out the TTR gene in human cells in a mouse model bearing a humanized liver. And so the model is sketched out on the left. The study overview is provided on the right, and the really critical aspects of the study overview are that we tested 2 different types of fusosomes, one carrying a nuclease targeted to TTR; and one carrying in Irrelevant nuclease, that was the control. And we tested them in these humanized liver mice providing them on day 1 of the experiment. And then at day 17, looking at the gene modification in the liver cells as well as the secreted levels of this TTR gene product in the blood. And the results are shown on this slide. So on the left, we're monitoring gene editing percentage via sequencing, and we're able to show a very high level of editing, 55% in the treated mice versus mice treated with our negative control fusosome. On the right, we monitored the level of transthyretin. This is the TTR gene product in the bloodstream. And you can see at day 0, the pre-bleed mice, there was no difference in levels. Whereas at day 17, we see in the range of 60% to 70% repression or a reduction in that level of TTR. And that's consistent with our gene editing efficiency, and it was a really nice result for this highly early study. Okay. Now I should mention at this point that this was a good model system to perform our initial in vivo studies in, very convenient to test these reagents. But TTR is not a therapeutic target for us, and so we then transitioned our focus for our in vivo studies to what we are endeavoring to make therapies in, and that's HSCs. So HSCs or hematopoietic stem cells are a very important stem cell type in your blood compartment. They give rise to diverse progeny that comprise the blood and immune systems as well as some other cells. And given their critical role, methods for efficient editing of these cells have the capability of potentially addressing many different types of diseases. And this is just a partial list on the right-hand side. I should mention that many of these diseases are in fact being addressed at this time or there are many entities endeavoring to address these diseases using ex vivo methods. But what we want to do, of course, is to do this in an in vivo fashion. And the reason for this is -- is highlighted on this slide. I think, actually, this has been pointed out several times by prior speakers. Basically, an ex vivo process is much more cumbersome than an in vivo process. And moreover, there's substantial safety considerations, given the conditioning that's involved for ex-vivo therapies. So we're very excited to pursue this vision, and I just want to point out that Sana has really these diverse necessary capabilities that you will need in order to be successful here. So we have Fusosome expertise, a team that's been working together for 5 years to design and retarget fusosomes to cells of our choosing. We have Genome editing expertise. This involves both the development of therapeutic editing reagents as well as their qualification for IND filings. We have HSC expertise, both deep knowledge of the receptor biology that you'll need to know in order to target HSCs. Richard alluded to this. It's an area that you really need to know what you're doing in order to focus your modification on HSCs or cell compartments that contain HSCs, and also the uses and limitations of HSC preclinical models. And finally, we have licensed foundational IP. So with that, I'll close out by just providing 3 data slides that I find tremendously exciting and that I think give a sense of our capabilities at this point in where we're heading. So first of all, we've shown that we can make editing fusosomes and efficiently modify CD34 cells, and this is just an example of that. So a slide very similar to the one I showed 3 slides ago. Fusosomes were made in producer cells, applied to tester cells. We got very high-efficiency editing of our target gene B2M. But in this case, the tester cells were CD34 cells. We then took these fusosomes, and we've tested them for the ability to modify bone marrow cells in vivo. An overview of the study model is indicated on the left. This is a mouse model that allows ready [ engraftment ] of human CD34 cells. This mouse model is treated with the CD34 cells. After 7 days in [ engraftment ] it's then treated with fusosomes. And after another 5 days, we look at those -- we look at either all the bone marrow cells in those mice or just a subset that's thought to be enriched for HFCs, and we see how well we're doing for the modification. And those results are indicated on the bar charts on the right. And what we're able to show is that we're able to modify in vivo anywhere from 20% to 40% or even close to 50% of the targeted cells, indicating a very high degree of potency in this in vivo setting. Now I should point out that we can't formally say we've targeted the HSCs. You have to do a longer-term study in order to approve that definitively, but we're very excited by our ability to get to the cell type that we expect will include that HSC compartment. Okay. So finally, this theme of safety and specificity has been highlighted multiple times throughout this series of presentations, and it's in this context that I find this last side to be the most exciting one, at least for my section. This is -- so in these prior studies that I just talked about, we achieved important proof of concept targeting CD34 cells and targeting bone marrow, but that was with fusosomes that were actually somewhat broad in their tropism. They hadn't been refined to be pinpoint-specific for just a minor type of cell. In this study, we took fusosomes that actually had been engineered to target a very precisely chosen compartment of cells. This is the cells represented by that blue sliver in that right-hand pie, right? So if you took the whole set of marrow cells, you could point to a particular type identified by expression of a chosen receptor, and that would be that blue sliver. And we said we wanted to target just those cells. And the fusosomes team generated fusogens that were specific for that receptor, and we tested that fusosomes type in these mice. And the results are shown in the bar chart at bottom. We were able to show that of that 0.4% sliver, we were able to get 7% of those cells successfully targeted; whereas for that remaining 99.6%, that bulk of the cells, the remainder cells in bone marrow, we were able to largely avoid them having gene transfer to them of only 0.4%. So basically, we were able to show a 184 (sic) [ 178x ] preference with this targeted fusosomes system. And I just want to underscore that. What does that mean, right? Well, that translates directly. If you're worried about, say, safety, right, the ability to suppress gene delivery to all the other cells you might get into by more than 100-fold, that directly translates to enhanced safety in terms of potency or manufacturability, right, that means that you have 100x more potency for the cells you want. It means you have to manufacture that much less. So this is actually a very important result. Finally, just as a control, I should point out that we tested our non-targeted fusosome and saw the expected no cell preference. Okay. So just to summarize, a lot of exciting progress. I wanted to highlight where we're heading. So the next is to develop gene editing fusosomes for a therapeutically-relevant gene. We're also headed towards Nonhuman Primate Proof-of-Concept, as well as filing an IND for an in vivo therapy likely in the beta [indiscernible] And with that, we can go to Q&A.

All right. Thank you, Ed. Thanks, everybody, for the attention. We kind of went through a lot today. Hopefully, you were able to see a little bit about why we're so excited about the technologies we've put together, the data we've generated and our ability to translate those hopefully into important medicines for patients. And we're on the precipice of really understanding how well the preclinical data that you've seen translates into people. So with that, we're going to open it up for questions, and I'll just be joined on stage by our scientific team.

Any questions? And if you just wait a second here, Nicky will come with a microphone to make sure that, a, I can hear you because I have bad hearing; and more importantly, people on the webcast can hear you.

Salveen Richter, Goldman Sachs. Maybe to start here on CD19. Is there a possibility of de novo generation of mechanisms that could down-regulate CD47 expression in lymphoma patients and lead, therefore, to decrease persistence or an allogeneic reaction? And then I have a follow-up to that.

Terry, do you want to take that?

Yes, sure. I want to make sure that I understand exactly what you're talking about. So you're talking about de novo down regulation of 47 on the lymphoma cells.

Yes.

I mean I'm not aware of any data to indicate that that's happening. I mean but, certainly, something we want to pay attention to, but I'm not aware of that happening.

And to the extent that it did, I can't imagine it would be important for what we do. CD47 has nothing to do at this point with how we target the cancer cell.

Right.

Right? It's nothing to do with the interaction at this point, although Terry showed you [indiscernible] between the CAR T cell and the tumor cell. What it has to do is with how we prevent transplant rejection of the patient's own immune cells from killing the CAR T cell. So variable expression on the target cancer cells should have 0 impact. We don't have SIRP- [ alpha ], the cognate receptor on the other cells that shouldn't impact it.

And can you just speak to the translatability of the HIP modifications to humans? Just your thoughts on how that will translate from the animal studies.

Is that for Terry, you think? So yes. I'm not sure I got the question. Did you get the question?

You're talking about in vivo delivery?

First on CD19. Sure. On CD19, whether you can talk about the translatability of HIP modifications into humans.

Yes. I mean the models we showed you, those are used in the field in the CAR-T studies. I think what's different about our studies is that we not only use NSG mice, which don't have an immune system, we really use also humanized mice. So showing you that allogenic variant how to overcome that. I haven't seen that in the field when I look at preclinical data from others. Also, they might have allogenic strategies. They're still studied in immunodeficient mice. So I feel we at least have this component of an allogenic barrier in. And then also the follow-up, if you look at our data, we have a quite long follow-up of 3 months. And the 2 more study I showed you with a rechallenge. But of course, it is a preclinical mouse model, but I cannot think about any more closer to the clinic than what we showed you. So I think we have to see.

Ren Benjamin from JMP Securities. I guess just one starting off, can you talk a little bit about the donor cell characterization and how you're thinking these cells? Is there the criteria that you're using so that you're getting this type of homogeneous product? And along with that, on the patient side, is there a variability of expression of CD47 or sort of alpha expression, right, that might make you think about the ideal sort of patients to select in running these clinical studies?

Terry, do you want to take the first one? And maybe Sonja, you take the second? Does that make sense?

Yes, sure. I guess I want to make sure I got the first question. So if you don't mind, just kind of repeating that for me just so I understand.

So you're taking healthy donor cells.

Yes. So donor characteristics. Got it. So I mean so we haven't -- so what I would say is that we definitely recognize that when you make CAR T cells from cancer patients, right, that there's variability in quality. A lot of that driven by the therapies that those patients have received. But we also recognize that there's a lot of donor-donor variability, even with healthy donors when you run sort of the preclinical models. And so we're looking very carefully at product quality relative to donor characteristics. I wouldn't say that we or I think anybody else in the field are at a point where you can select the super donor, so to speak, that everyone is talking about, but it is something that we're spending a lot of time looking at. There are other selection criteria that we utilize that are relevant to the hypoimmune platform itself that we incorporate into our donor selection.