 This is a shot of, for Mount Rainier, when I was snowshoeing this winter and just a beautiful winter day. So just, you heard that I've got a biotech venture going right now. I'm a senior vice president and an equity holder in sauna biotech. The only way things make it from the academic sector into the healthcare sector is by commercialization, of course. And so that's what we're trying to do is to actually get some of these research discoveries into treating patients. I wanna just stop, frame my lecture today by saying that I consider what we're doing to be like working on growing the third leg for medicine. So medicine, I hope will be a three-legged stool. And of course it started out with small molecules and these were things that were either isolated by chemists or synthesized by chemists in the laboratory. That's why I guess we still emphasize organic chemistry to our undergraduates. But really in the eighties and more so in the nineties we had the advent of biologics and large molecules too big to be synthesized by chemists. And so we use cells as our factories to produce our pharmaceutical agents. And my point is I think eventually cells and genes are going to become the third leg of very complex biologics, really living sorts of therapeutics that will potentially permanently change the cellular composition and permanently change the genome of patients. I will make a couple of predictions on this predictions being as you heard me say earlier I seem to be a very poor predictor as much as I'm surprised. I think this is going to be revolutionary. I think it's gonna change the way we practice medicine. I also think this is going to be a slow revolution. This, these are complicated things just orders of magnitude more complicated than molecules. And we have to be careful to promise enough that we continue to get the capital investments that we need to make this happen but not promise so much that people get disillusioned with us and we lose our credibility. And so that's always a fine line to try to walk. Now, this is a cold Newcastle slide to come to Texas Heart and say, heart disease is bad and you don't want to get it. But everybody knows here that the ischemic heart disease is number one cause of death in the world and that here in the States it's the number one cause of hospital admissions and Medicare expenses. And the pathophysiology is also super well known to you all in the sense that coronary artery disease leads to typically plaque rupture and thrombosis. A sector of the left ventricular myocardium becomes ischemic shown here as a cross section that's darkly shaded. Heart has a very high metabolic rate highest in the body perhaps. And so within 15 to 20 minutes of the onset of ischemia, irreversible cell injury starts to happen. A way front of cell death sweeps through in ischemic basket or bed from the endocardium towards the epicardium. This is over now between three and six hours is our best estimates in patients. And so things happen very rapidly and that is of course the whole rationale underlying, you know, the reperfusion therapy. For patients who have suffered a myocardial infarction, of course the heart is about the least regenerative organ in the human body, vying perhaps with the central nervous systems for that dubious distinction. And of course scars form as a result the only new tissue the heart is capable of generating. And patients are left with contractile deficit and this sets off a downward spiral too often of contractile dysfunction. And you all know the data that once you're diagnosed on average you have about five years to live diagnosis of heart failure that is which makes it more malignant than most forms of cancer. So our view is in a sense simplistic. The execution has not been simple but the view is this is a classic disease of cellular deficiency. And we know now that we can make these cells from stem cells. So we ought to be able to remuscularize the wall of the human heart. And so I had that idea 25 years ago and I am still working on it. So it tells you that reducing it to practice is way trickier than I imagined. So let me just summarize maybe the last 15 years or so for people who have not been obsessed with this as I have. The field really got started with adult cells because they were less complex, easier to work with more readily obtained. And so initially from bone marrow then adult cells from the heart itself and then branching out like to stromal vascular cells and things like that from adipose tissue. For a long time we thought that these cells had considerable plasticity that they could adopt to their new environment to become pericable cells of the new tissue. We now know that this doesn't happen. And in fact, I'm sorry, these cells don't really engraft long-term. And so to the extent that they have cellular effects these are through paracrine mechanisms. That is cell-to-cell signaling, things that are non-cell autonomous that are say secreted into the environment. And could be growth factors, it could be exosomes, it could be extracellular matrix molecules, things like that. And this has really been very mysterious. So there is something there, but when this has been tried in the clinic the efficacy has really not been robust. And so we have shifted to many years and decades fairly to using polypone stem cells because of their ability to generate definitive cardiomyocytes. Of course, there are two types of polypone stem cells. The first type were the embryonic stem cells which were controversial when they were first derived back in the late 90s. And then the more recent reprogrammed cousins the induced polypone stem cells that where we learned how to turn the clock back on adult somatic cells and make them very primitive cells that could turn into anything in the human body. So these make definitive cardiomyocytes and we learned the rules for first generating the cells of the scales that we needed and then how to transplant the cells into animal models of schemic heart disease and begin the process of learning to remuscularize the heart wall. So we basically worked our way up the food chain starting in rodents, mice and rats moving our way up to guinea pigs. And now our work is mostly done in large animals like non-human primates and pigs. So how might these work in terms of mechanism of action? What we show is that this new myocardium can couple electrically and it can beat in sync with the host myocardium. And the assumption is that in so doing it we're actually bringing new force generating units into the heart that can improve the contractile squeeze. These cells being cells they also could have paracrine mechanisms of action and we haven't ruled that out in any way. I think probably both things are true. And I suspect you all are aware that clinical trials for human cardiomyocytes derived from pluripotent sources are really just beginning. It's very much in the early days. So we've learned how to grow these cells at scale and to manufacture them under pharmaceutical grade conditions in good manufacturing practices. So this is a shot from our cell manufacturing facilities in 2021 where we're working in clean room conditions our research staff basically are working in these Tyvek space suits and we're able to use stirred tank bioreactors to grow the cardiomyocytes at scale. So this is a three liter stirred tank bioreactor that you see here. What you can't see if we zoomed in that there's a single use disposable wet unit on the inside of this that has like a paddle wheel that turns and we grow the cardiomyocytes in suspension. Although the recipe is shown here we take a pluripotent stem cell doesn't really matter whether it's an embryonic or an IPS cell. Making cardiomyocytes is not that hard. Once you've spent five years learning how to do it we teach undergrads to do it in two weeks and they do quite well. You just turn on the wind pathway first this induces mesoderm then you shut off the wind pathway and let the cells roll on their own momentum after that they make cardiac vascular progenitor cells and then self-restrict to form cardiomyocytes. So this is what these little cardiac bodies look like that float around in the bioreactor like snow globe. And if I turn on my video you should see that these guys each of these little cardio balls start to beat. They have a couple of thousand cells in a spheroid like this and we'll zoom in and you can see each one of them is synchronized because they form gap junctions from cell to cell within the cardiac particle but each one of each ball is marching to its own drummer of course because they each one is essentially its own pace making unit non-electrical continuity with its neighbors. So we can make these at three liters we can generate eight to nine billion human cardiomyocytes at greater than 90% purity. And just for a frame of reference a typical infarct in a patient that would induce heart failure would result in the loss of about a billion cardiomyocytes. So we think it's possible to talk about scaling up human size doses at least for early first in human kinds of stuff with the technology we have today. I'll skip the small animal work that we did in development I wanna just cite some work that was published four years ago in Nature Biotechnology where we learned so much by moving into large animals that we did not learn by spending decades in small animals. This is from a non-human primate study using pigtail macaque monkeys as the host for this we induced infarction in these animals with an angioplasty balloon in the LAD and then we came back two weeks later after putting the animals on triple drug immunosuppression and we either injected a vehicle into the infarct or we injected 750 million human cardiomyocytes. On the left here is a vehicle injected heart everybody knows what an infarct looks like you can see the blue collagen here and you can see the red residual myocardium and it's very easy to see the cellular deficiency and why there's not a good squeeze in this ventricle. Looking over on the right-hand side you can see this is a human cardiomyocyte engrafted heart and the human cardiac muscle cells are stained within an antibiotic specific for human troponins in green and you can see we're getting decent islands of human cardiomyocytes that are forming upwards of a centimeter or so in maximal dimension we can see these by sort of negative imaging if you will by late gadolinium enhancement because the infarct goes from being gadolinium positive to having these dark islands of new myocardium that exclude gadolinium from their intracellular space. So you can imagine a heart like that might have a better squeeze to it and indeed that's what we found here we're looking at left ventricular ejection fraction by magnetic resonance imaging and before infarction macaque monkeys have ejection fractions like healthy human patients sort of in the 65% range or so after the infarction this is two weeks post infarction pretreatment where you can see these are pretty big three hour LAD occlusion infarcts and so that we lose about 25 points on the ejection fraction down an average of around 40% or so but the groups are you can see are quite comparable prior to treatment and then the animals that receive human cardiomyocytes are in blue and the animals that receive vehicle are in red rather sorry and the vehicle treated animals show no significant improvement over four weeks and if we extend them out to 12 weeks we see again no significant improvement and this is natural history of the disease but the animals that we see getting human cardiomyocytes show perhaps a 10 point improvement in ejection fraction by four weeks and in a couple of animals that we've let go out to three months we get another 10 to 12 points of ejection fraction improvement basically getting them back into their pre morbid levels. So this is like the go signal that gets me out of bed in the morning if human cardiomyocytes can do this for a non-human primate they should be able to do this for a human primate. We have some problems still to solve but it should be possible to remuscularize a human heart I think and change the trajectory of a patient's illness. This is the thing that keeps me up at night and I'm gonna spend a lot of time talking about electrophysiology with you all and there is a toxicity that occurs with these potent stem cell derivatives and this is a rhythm disturbance that we've come to call engraftment arrhythmia. A representative rhythm strip is shown up on the top you can see this is from a non-human primate going along happily in sinus rhythm and suddenly it starts with like a PVC and then we get this run of nasty looking ventricular tachycardia at a very high rate and these rates can go up into the mid 200s for example tolerated reasonably well in a non-human primate not so well tolerated in a larger heart like in the pig. We discovered this years ago when we put our human cardiomyocytes into non-human primates this is the time course of that the vertical line here represents the two weeks post-infarction at the time at which we transplanted and you can see these animals go into very substantial periods of engraftment arrhythmia. Here this is the number of hours per day on the vertical axis so you can see like in this one animal over 20 hours a day of V-tach. This was also true when you put monkey cardiomyocytes from stem cells into monkey hearts so it's not a species specific thing same sort of time course and it is also true when you put human cardiomyocytes into the pig this is work from Michael Flam's laboratory. The two vertical arrows here indicate two pigs that died from cardiac events and so these engraftment arrhythmias we learned can be, excuse me, can be lethal as well. So this is the good news is animals seem to come out of the woods. At the end, this is like a healing in process. I view this as a cellular maturation process and so once they come out of the woods they're stable but it's a nail biter along the way and we didn't want to do this to a patient. I will say we were of course bummed out when we found this because we were hoping to jump the patient soon and we realized no we needed to fix this but we didn't see it in mice we didn't see it in rats we didn't see it in guinea pigs probably because their hearts are all too fast and so I think we had to go into a large enough animal that the physiology was more akin to the human and so that was a good lesson to learn as well. So we thought, well in the end, good catch we didn't do this to a patient. So we wanted to know what the pathogenesis of this was so we started out by doing electrical mapping and so what you're seeing here is a single electrode with a biosense Webster system a single electrode map of activation in the left a left ventricle of a macaque monkey that has that is in the middle of a rapid VTAC. I think you guys know how these things work but these three balls represent cusps of the aortic valve the base of the heart is up here the black zone is unmappable so we don't have any data points the apex of the heart is down here at the lower right and you can see of course everyone knows that activation should break out at the apex and the heart should squeeze towards the base that's not what is happening here and if I turn on the video and you see the propagation map that it breaks out here in the anterior septum and it spreads centrifugally I guess you'd say out from there and this was a consistent finding. So in low risk to be sure but by sort of clinical standard mapping at that time this is several years ago it looks like we didn't see anything that looked like a rotor of re-entry. So we went on to do some other studies again using clinical tools basically so by mapping it looked focal we tried pacing and we were able to capture we could take an animal that was in VTAC at say 200 and overdrive it to 250 or 300 but this never suppressed and so again this evidence favors automaticity. We tried DC cardioversion so you could shock the animal and after a very brief pause of a second or two it resumed the TACA cardio at the pre-cardioversion rate. This is evidence in favor of automaticity. Finally, since this was an evanescent phenomenon we could bring the animal into the laboratory when it was in sinus rhythm and we could do programmable electrical stimulation and we never could induce this TACA cardio. So typically re-entry TACA cardio can be induced with program stimulation. So again this is evidence favoring automaticity. I was a little uncomfortable more than a little uncomfortable going through it was like a fork in the road in terms of how we were gonna approach this problem if it's re-entry versus impulse generation. And we basically had to make a call and we said, okay based on this evidence we're gonna pursue automaticity and pacemaking mechanisms and not re-entry type mechanisms in terms of trying to think our way through the problem. It wasn't like we had genetic evidence it was like clinical grade decision-making I would call it. So we started to really focus on arrhythmias here and I wanna just give a shout out to two junior faculty members in our group Kenanakumara is an interventional cardiologist who has really helped us with all of our large animal catheter-based work and that sort of thing. And then Lauren Nidig is a veterinarian and a large animal surgeon and so forth. And she has really helped with the management of these animals and all surgical models for cell delivery and so forth. Our work on arrhythmias we discovered them in the monkey but the monkey is super expensive, very low throughput and it's small they're only 10 kilos or so and so we're always having trouble trying to fit our catheters and things like that in the circulatory system of the monkey. So pivoting to the pig has proven to be much higher throughput. It's not cheap everybody knows that but it is less expensive. And as I mentioned it's compatible with human delivery catheters and so forth. We're still doing primate work. Really we've been using the primate a lot for immunology because you could do allogeneic transplant from monkey stem cells into unrelated monkey hearts. And so we're doing a lot of immunology work that I won't have time to talk about today. So the first thing we wanted to do was to see was there anything in the medicine cabinet in the pharmacopoeia that might help us find a drug approach to suppressing and graphing arrhythmia. So we tried antiarrhythmic after antiarrhythmic and I'll save all the screening that we did but we found two things that helped. The first was amiodarone. And so here you can see three examples of animals responding to these again are pigs with 150 million human cardiomyocytes transplanted in them, they're immunosuppressed and what you're seeing is examples of breaking of the rapid ventricular tachycardia back into sinus rhythm. Not perfect, but this was the first signal of any kind that something might be able to help. Amio is the only drug we have seen that can actually convert from the VT into sinus rhythm. The other drug that we found that's been helpful is evabridine. Excuse me, people know the mechanism action of evabridine, it binds to HCN4, which is the gene that encodes the funny current basically, that it's important in the pacemaking mechanisms. We'll talk more about HCN4 later. And what this does, it helps us with rate control. So you can see it doesn't convert it from ventricular tachycardia to sinus rhythm, but it converts it to what my cardiology buddies call AIVR accelerated anti-ventricular rhythm. I think of this as just slow VTach. The thing is the animals don't necessarily die because they are in tachycardia per se, you guys get this. They die because the heart doesn't fill, the heart doesn't empty and the heart doesn't perfuse when it's going this fast. And so rate control, they can bangle on pretty well in AIVR, they don't like rapid VTach like this. So how does this play out at the group level? We found, I'll show you some Kaplan-Meier curves where we saw changes in survival. So here are the control animals. You can see this was in our hands with the dose of cardiomyocytes that we gave. There were quite a few lethal events. And so this was, we had our group sizes were nine and eight. And so almost two thirds of the animals died from cardiac events. None of our animals died of cardiac events. These two deaths were from opportunistic infections, the pneumocystis pneumonia and cytomegalovirus. And so that's like intention to treat. And then if we look, how many animals died from cardiovascular disease, it was zero. We had 100% cardiovascular survival. And so we actually have decent P values on this. So we at least have a drug cocktail that we think makes a meaningful difference in survival. How does this work? Part of it is we're suppressing the heart rate that goes along with this engraftment arrhythmia. And so some of that is the amiodarone. So if you look here, so the infarct happens here at the first dotted line, the second dotted line is the cells being transplanted just to look at the natural history. The animals typically go into ventricular tachycardia over the first week. And then it gradually slows down until it gets to the point where the sinus rhythm is able to overtake. And you can see we get nice rate control with this. And I think that's part of the reason that none of the animals died. And here's animals that got a sham transplant. So this is not a, you know, there's some arrhythmia associated with the infarct model, but not a lot. It's very easy to tell what's engraftment arrhythmia. And if we look at the fraction of the day that the animal spends in arrhythmia, you can see we suppressed this. This is probably the amiodarone effect. So this, you know, on average, when it gets bad, it's 75% of the day or so, and that's suppressed to maybe 50% of the day. It's not, I'm not saying this is great. I'm saying it's better and it's giving us some insights into mechanism as well. Amio, as you know, is a pretty promiscuous drug. And so it binds to so many ion channels that I don't think it really tells us what the culprit channel is that carries the current. But it's a very good for what ails you kind of drug. So that was our first foray. We could use pharmacology. We could make some impact. The next thing we wanted to do was to really get after the, you know, try to find what the currents were that were carrying this, what channels were carrying the arrhythmic currents. And I wanna contrast an adult ventricular action potential here with the action potential of a stem cell derived cardiomyocyte. Everybody knows the adult action potential waveform, a rapid upstroke as the cell depolarizes a little notch, plateau phase, and then gradual repolarization, and then a flat baseline here in the phase four of the action potential. It really, you know, ventricular cardiomyocytes have no automaticity. They're waiting, of course, the signal from the sine wave trial node. In contrast, the stem cell derived cardiomyocytes at their immature fetal-like state have a very prominent upslope in depolarization because they are behaving like pacemakers. Every cell in the fetal heart early on is a pacemaker cell. And then it has the phase zero upstroke. It doesn't have the notch. It doesn't have the long plateau phase, and then it repolarizes. So very, and this is, of course, because they have a very different ion channel repertoire. So here's the notion. We made the bet that this was automaticity. An automaticity could result from one of two reasons. First is that there are currents that are normally absent in mature cells, but they're present in our graphs. So there's a culprit current that we wish was gone. Or there are currents that are normally present in the mature ventricular cells that are absent from our cells. So there is something that we need to goose up in these cells in order to fix this problem. So you could imagine this is setting the experiment for either loss of function or gain of function designs. And if you follow our logic, these are currents, and then these currents are conducted by channels. These channels are proteins, and therefore they're encoded by genes. And that means that we should be able to use gene editing of the graph because we have complete control of the graph, either by knocking out things we don't want or overexpressing things we do. And we should be able to hack the action potential to the point that we make engraftment arrhythmia go away. So the way it starts, and I'll run through this part quickly, we wanted to know how our cells matured. And so we grafted human stem cell derived cardiomyocytes into rats, and we let them mature in situ for up to three months after transplantation. We did laser capture micro dissection like this. So you can see we've zapped out the graft in this region. We made RNA, we did RNA sequencing, and we looked at their expression profiles, and we watched these human cardiomyocytes mature in situ. And we separated out all the rat signals from the human signals bioinformatically. So we're sure that we've got pure human cardiomyocyte maturation data. And we got this time course of maturation, which is basically this vertical axis in a principal component thing. It shows these cells as they mature going forward. And we'll talk a little bit about some of the genes. So this lab launched a project that we call, Project Medusa, which is one of these acronyms that you make up a little after the fact to have some fun with it. So Medusa stands for modifying electrophysiological DNA to understand and suppress arrhythmias. And here are the three people in the lab who were most involved in leading this, a really talented postdoc named Silvia Marquiano, Hans Radica, a senior molecular biologist who's been with me forever, and Alessandro Bertero, a brilliant postdoc from Italy who's now got his own lab back in his home country. So the workflow goes like this. We start out with the universe of ion channels. We focus on ones that we know, we of course focus on things that we know because EP is pretty mature. And then we look at the RNA expression after engraftment and we get idea for, huh, this is one that looks like it's missing. And here's another one that we wish weren't there. And so we mash all this together and we come up with a list of things that we want to manipulate by gene editing. We prioritize this and we test it in vitro to see what the cells behave like. But the rub is that we really don't have any in vitro models for this engraftment arrhythmia. Neither do we have a small animal model of engraftment arrhythmia. What we have right now is only large animal models for it. So we set out with the lowest throughput experiment ever where we did an in vivo screening model in the pig where we tried to eliminate everything that we could. For example, you don't have to infarct the animal to get engraftment arrhythmia. So we did normal hearts without engraftment arrhythmia. We used a relatively low dose of cells, 150 million because we're gonna be generating many different cell types and learning how to differentiate them in bulk is very time consuming. And so we did these immunosuppressed pigs and we transplanted cells into their hearts and we tracked them by telemetry to see if engraftment arrhythmia came or went. And so I'll skip this. This is a teaching thing that I've been teaching. I've been teaching electrophysiology to people in finance and we can talk about that later. We transplant our cells either surgically or with catheters. There's a Myostar catheter that we were using for a long time that's been discontinued. So we've had to switch to surgical delivery but you're seeing here, there's the LAD in a thoracoscopy. We've done a sternotomy to expose the pigs heart. The LAD is here. You can see down in the LAD bed, there's five per string sutures for we've transplanted cardiomyocytes. And also there's a Myostar catheter doing an injection into the anoreceptal wall of a pig. We get nice looking graphs of human cardiac muscle. This is about a month after transplantation. This is stained for a human specific cardiac troponin and you can see sarcomeres and things like that. So it doesn't seem to matter. The model is about the same, whether it's catheter or surgically based, which was lucky because the catheter went away. So this was our gene editing strategy. We started out with wild type cells and whenever you see today a pig with swirly eyes, that means you get engraftment of rhythm. And we made a list of all these different genes that we were interested in. And then we evaluated many of them in vivo. And then we started doing things in combinations, in duplicates, triplicates and ultimately quadruplicates. And the only reason you would do that is because it's not working when you're doing single gene knockout. So I will, this was a four year study that I'll take you through in about seven minutes or something like that because I wanna keep an eye on the clock here. So the first channel that we were interested in the one would be probably high on everyone's list was HCN4. This is, some people call this the pacemaking channel. It's responsible for that funny current that contributes to the phase four depolarization. And it's also the target of avabradine, which as we told you worked pretty well in pharmacologically. So we knocked this out in the stem cells, made cardiomyocytes and we transplanted them in. And let me just again run you through the first, the wild type animals cause you'll see the same wild type curve, typically within a week after transplanting the cells, they go into a very rapid heart rate. You see one black dot here, that means we lost one animal, although it died from cardiac complications or had to be euthanized. And then it comes back down as the cells start to mature and graft. And then here's the arrhythmia burden. Again, it shoots up to maybe 75% of the day. And then it gradually comes back down as the cells mature. And I think it's pretty easy to say, to see that these are small numbers and of two here, but we didn't have any meaningful impact on the engraftment arrhythmia as the, by knocking out HCN-4. That was like, okay, well, maybe we need to do some other things. And I'll skip some of the other single gene things. All of our single gene interventions looked just like this. They didn't do anything. And so the next thought was, well, these cells, their resting membrane potential is quite positive compared to, they'd say minus 50 to minus 60 millivolts compared to an adult cell that's like minus 80. And there's a channel called KCN-J2 that encodes a depolarization current that sets the resting membrane potential. So we thought, okay, let's overexpress KCN-J2 in combination with this HCN-4 knockout. And then we'll reduce the depolarization current and we'll increase the hyperpolarizing current. Okay, so we tried that. We knocked this into a safe harbor and we put a constitutive promoter on and this didn't work at all because the stem cells hated this gene. We could never get a stem cell. So we presumed it's toxic to stem cells. So we had to find some way that it only came on as the cells differentiated into cardiomyocytes. And so what we did was basically a two for one where we knocked this potassium channel into the HCN-4 locus. And so the idea was to get sort of the natural regulatory environment for KCN-J2 that would mimic HCN-4. And I'll go through this quickly, but basically here in a wild type cell, you can see HCN-4 comes up after differentiation of cardiomyocytes from stem cells. And KCN-J2 is, here's the wild type. It's typically not expressed until much, much later in development. But in these ones where we swapped it out, KCN-J2 has sort of a natural regulatory profile of HCN-4. So we were high-fiving. We thought this was a very clever way to knock one out and knock another one in, et cetera. This had absolutely no impact. If anything, it was worse. We lost both the animals that we put this into. It looked like maybe it delayed the arrhythmia, but when it came on, it was really bad and really unstable. So we created a very chaotic situation. And so both these animals died from cardiac events. And so you could argue is the biology different, but the outcome is the same. And so this was not what we were looking for. We were looking for something that would make really a meaningful difference in rate, rhythm, and survival. Okay, back to the drawing board. How about if we edit things in triplicate? Again, so here we, so these things will be in the background of overexpressing KCN-J2 by knocking it into the HCN-4 logo. So the no HCN-4 and overexpressing of KCN-J2. And then there's a T-type calcium channel that's expressed in these cells early in, gets silenced later as adult cells called CACNA1H. I'll just call it the T-type calcium channel because it gets to be so long. So we did a triple edit where we knocked this one out in addition. And the first animal that we did died rapidly from ventricular arrhythmias. And so we were discouraged about going on with that. That didn't seem really good. So a different gene that's very important in cardiac biology is the sodium calcium exchanger. It's encoded by a gene called SLC-8A1. And so this NCX knockout, we thought also might help. And when we put that in, the animals died as well. So let me just summarize what I've showed you so far. You're getting a graph on arrhythmia with wild-type cells. You get it, we knocked out this mechanosensing gene called PASA1, by the way, that didn't help either. We knocked out HCN4 alone, that didn't help. We added the overexpression of KCN-J2, that didn't help. We added knockout of, where is he? I'm getting lost in my own. We added the knockout of the T-type calcium channel and that didn't help. And we added the knockout of the sodium calcium exchanger and that didn't help. So sort of as a last hurrah, this is about three years in and Sylvia is now like a very depressed postdoc. We can't have sharp objects around the lab any longer because of self-harm concerns. I'm never gonna get a job, et cetera. And we thought, what if we did this whole suite? Maybe we've gotta do all of these in concert. And that's what we did. So we knocked out HCN4, we knocked out CACNA1H, the T-type calcium channel. We knocked out the sodium calcium exchanger and we overexpressed KCN-J2. And lo and behold, some interesting things started to happen. So we've come to call these the Medusa cells just because everything else is such a mouthful that we'll just call these guys Medusa from now on. What we got was a cell that was quiescent at baseline, no automaticity, but something that's electrically excitable, kind of like an adult ventricular cardiomyocytes. So here you can see that we've plated cells on multi-electrode arrays. We get normal beating in the wild type cells and there is no spontaneous activity when you go for days and days in these Medusa cells. Here is a single cell patch clamp recording when it's being pasted at one hertz. So you can see a wild type cell that generates a very nice action potential. This is what it looks like if you stretch one of those guys out and I'll just point out that you see this normal phase four depolarization in this one and that's HCN4. Here we can, when we patch onto the Medusa cells, we can paste them and you get a pretty normal looking action potential except there's a couple of differences. One of which is that this is flat after it returns to a very flat baseline, it doesn't start depolarizing on its own. So here, if you look at the capacitance which is a measure of cell size, no difference between the two, action potential duration, no difference between the two but the Medusa cells are more hyperpolarized. They have a more negative membrane potential and that's because of the overexpression of that potassium channel as we mentioned and this is just two different ways of looking at it. So they're clearly more hyperpolarized. So this looked like a really interesting cell to transplant into the pig. And we went on to do this and we basically eliminated the engraftment arrhythmia. This is really quite interesting. So here, looking at heart rate, this is the same curves in blue that you've seen before but really rock steady in the animals where we transplanted the Medusa cells. And then similarly, the burden of fraction of the day spent in arrhythmia, very minimal arrhythmic events. There's a few PVCs and things like that but it's really nothing like these sustained polymorphous tachycardias that we were saying before. So that looked really interesting. Now, a trivial explanation would be what if we edited the cells so badly that they won't engraft? And then obviously that would be an explanation for no engraftment arrhythmia. So we looked histologically and we see nice looking graphs of these human cardiomyocytes in the pig heart and so they seem to engraft at least as well as their wild type cousins. So what I'd leave you with here is that the absence of engraftment arrhythmia is not due to the absence of engraftment. So what have we shown you here? I've shown you that you can do a fairly complex quadruple gene edit that looks like it pretty well eliminates this engraftment arrhythmia. And this is the cocktail of edits that we did. I'll just remind you, this is the last time I say it, I'll say it, I'll promise that we knocked out the funny current HCN form. We knocked out the T type calcium channel which is not expressed in the adult heart. Neither of these are CACNA1H. We knocked out the sodium calcium exchanger. This is expressed in the adult heart. We'll come back to that. And we overexpressed this hyperpolarizing channel that sets the resting membrane potential. The fact that we were unsuccessful with the edits in singlicate, in duplicate and in triplicate says to me that there is a lot of redundancy in channels that carry this arrhythmic current. In a way, since we're essentially hacking the action potential and the basis for the human heartbeat, you might imagine that nature has designed us with some failsafe mechanisms that you would have to go through in order to be able to eliminate this from a cell that you want to use for ventricular repair. To me, this provides pretty good data that automaticity really is the mechanism of this engraftment arrhythmia. We still don't know the why. You're like, why do they go crazy when we engraft them? I'm really interested in that, but we don't know the answer to that. So we got a lot of insights into mechanism. A question that we're asking ourselves, I'd be really curious what you all think, is could this be a therapeutically useful strategy? This is some of our ruminations so far. HCN4 and CACN A1H, the T-type calcium channel are really not expressed to any significant level in the mature cardiomyocytes. And then we're overexpressing KCNJ2, but that's an alocus that will naturally silence because the HCN4 gene is naturally repressed during maturation. So those three, I feel pretty good about because the cell will come to reflect the normal profile of gene expression. The thing we think hardest about is the sodium calcium exchange or encoded by SLC8A1. This is a really high abundance gene in mature cardiomyocytes. And in fact, for the adult myocyte, it's the main way that calcium leaves the cell. It of course enters during depolarization. And then during repolarization, that is the major route of calcium extrusion. We know from mouse genetics, Ken Phillipson's work at UCLA, that much to my surprise, you can knock this gene out in the mouse heart and the mice are surprisingly healthy. There's a little toxicity. They're not, the moms don't do as well with pregnancy. There's a little bit of reduction in ejection fraction, but it's not like a fluoride cardiomyopathy. And so the cell clearly has adapted and they change their action potential to actually let less calcium in with each depolarization so that they don't have as much that they need to extrude it. So this is a maybe right now. And the last thing I'd say to young people is like, this was a tough one. This is maybe the hardest thing I've had to do since I've gotten into science. And one lesson is, sometimes guys, it pays to stick with stuff until it gives because we had to really dig a lot of layers until we finally got this thing to open up. With that, I'd like to say thanks to the team who really made all this possible. Along the ways, if I've said I, I should have said we. If I said we, I should have said they, that sort of thing. I'm just fortunate to get to speak about these people's work. And this was when the team was pre-COVID at the university and then the folks whom I've circled here have moved to SANA to help start the cardiac regeneration program there. And I would like also just to say thanks to some partners in crime. Rob McClellan and Nate Stadecki. Rob's our head of cardiology and Nate's an engineer who's very helpful with our university heart regeneration program. And many thanks to our funding sources, the NIH of course, but other sources as well, who have helped us move this work forward as we, as we're getting closer to the clinic. And with that, I can stop sharing. And thank you all for your attention.