Thanks, everyone, for being here. My name is Yanan Zhu. I'm one of the Biotech Analysts here at Wells Fargo. It is my great pleasure to introduce our next fireside chat. We are privileged to have the Beam Therapeutics team with us. Here on stage are John Evans, CEO, and Pino Ciaramella, President of the company. Thank you, Pino, and thank you, John, for being here. Thanks for having us. Great. I was wondering if you could start off by providing an overview of the company and any upcoming catalysts that investors should keep in mind. Sure. Beam Therapeutics is a next-generation gene editing company. We use an important form of CRISPR called base editing, in which we use the same precision targeting ability of CRISPR to edit any spot in the genome. We have modified the system so that it doesn't make double-stranded breaks where you lose control of the sequence. Instead, we're making precise single-base changes at the target site, so much more precise and much more predictability in terms of the therapeutic outcome that you're going to have, all for a durable one-time therapy for very serious diseases. We're applying this very broadly, but probably our core franchises are in two areas. One is hematology. This is where we take the blood cells, and we edit them directly to try to cure disease. Lead programs in sickle cell disease. BEAM-101 is our version of correction edit for sickle cell and an ex vivo transplant that is mid-clinic. We've shown dramatic proof of concept there. We believe that's a best-in-class product moving rapidly through the clinic. That is then followed by two more waves of technology, which will come to try to treat sickle. They have applications beyond that, which would try to bring that same functional cure to many more patients, first by getting rid of chemotherapy in the transplant process. We call that the escape technology, moving to the clinic in the medium term, followed by in vivo delivery directly to the bone marrow. A series of technologies that I think could be transformative for these patients. The second big franchise at Beam is our liver franchise, where we're going to be using simple lipid nanoparticles to do in vivo infusions, just a couple of hours in the chair, to then edit the liver directly to correct a variety of severe genetic diseases. The lead program there is for alpha-1 antitrypsin deficiency. BEAM-302 is our program there. BEAM-302 is a single-base correction of the point mutation, the single-letter misspelling that causes alpha-1, a very severe condition. As with sickle, we have shown dramatic proof of concept for BEAM-302 in patients in early phase I data. That was back in March. We believe that is a best-in-class profile and now, again, starting to accelerate in the clinic for a very high unmet need, very large patient population. Behind that, we have BEAM-301 for glycogen storage disease and a variety of other liver programs beginning to line up behind that. Great. Let's jump into the AATD program, which seems to be the primary focus for investors now. Can you give us a quick review of your phase I/II study, including the data that you reported for the first nine patients in the 15, 30, and 60-week group? You have since dosed more patients. Can you talk about initial observations from the additional eight patients you dosed? Sure. I'll do a flyover here of the data we presented in March. We're doing a dose escalation, where we give increasing amounts of the drug over time. We did 15 mg, 30 mg, and 60 mg, three dose cohorts, three patients each. That was a nine-patient data set. We saw very clear dose-responsive effects. Specifically, at the 60 mg dose, that third dose cohort, we saw therapeutic effects, beyond all the thresholds we would think would be needed to cure the disease. Specifically, what that showed was patients were going from a baseline, where these are patients who are only making the mutant form of the alpha-1 protein. We call that the Z protein. These are patients who have a ZZ genotype. They're maybe in the mid-single digits for a level of alpha-1 protein in their body, and it's all Z. Z is less functional. It is less effective. It also causes toxicity both for the liver and systemically. There may be 4 μM-5 μM o f Z. On therapy, after the 60 mg treatment, we turned that into 12.5 μM of total AAT. Obviously, a dramatic upregulation there. The mixture of that AAT is almost all what we call M, which is the normal form of the protein. We had about 11 μM of M and just over 1 μM of Z remains. Not only turning on a lot more AAT in the body, but it's all that normal functional M protein and much less of the toxic Z. We think all of those components are important, both higher total levels, higher functional AAT in the body, production of M to a high degree, and elimination of Z as much as possible. The reason we think all those are important is because when you look at people who are carriers of the disease, they have one copy of the Z mutation. Their other copy is not Z. It's some other allele. They don't have the disease. They are not progressive. Their lungs are protected and function normally. Their livers are not getting toxicity. Whereas in patients, the Z protein is both causing liver toxicity and failing to protect the lungs from progressive toxicity over time. Our 60 mg dose cohort looks like a medicine to us. It gets patients well out of that disease phenotype and into the zone where we would predict that they would not have any further progression. From here, as you noted, we're going to do some additional dose exploration. This is our one chance to do it in the latter part of phase I. We're going to do a few different things. First, we're adding more 60 mg patients because, again, we think this is a potential drug already. We want to push for more exposure to see, can we pick up more editing? Can we drive levels higher without sacrificing safety? The safety profile here has been very clean so far. We're basically following two axes on that exploration. One is to add more single-dose potency. We're going to go to a 75 mg c ohort. The second is to stay at the same dose but add more frequency on the schedule and do a two-dose regimen of 60 mg each separated by eight weeks. This just tests different sort of pharmacodynamic ranges for the drug, what happens if you add more drug, what happens if you add it a second time, and see what we get in terms of AAT levels, transformation, creation of new M, and then obviously safety. Our goal is to bring all of that back by early 2026 and get a read on where we think the optimal dose and schedule is for a registration push. Right. Thanks for that overview. I do want to touch on safety because that's something that's notable from your data report. Safety has been very good with grade 1 observations only. Could you talk about any specific design feature of your LNP that might account for this profile? How about your expectation for tolerability profile with a second dose? Yeah. Thanks for the invite, by the way. The LNP has obviously four components to it. One of the components is an ionizable lipid, which is an ionizable lipid that we have licensed from a third party, which is biodegradable. The nature of biodegradability certainly helps to that tolerability component. It's important also how you assemble the LNP together. At the company, we have extensive experience, actually, of doing LNP from various prior lives as well as what we've been doing here. I think we paid a lot of attention, actually, making an LNP that we had introduced all the features and the manufacturing process that we felt was important to be able to do that. Thankfully, the data is backing that up and gives us the opportunity to do that. The second dose, actually, we're now expecting to see an additional tolerability component. In part, in order to maximize that fact, we are going to give the second dose eight weeks apart. What that means and the reason why we've chosen the eight weeks is because by that time, by the time the second dose comes, the lipid from the first dose will be completely washed out of the body. Therefore, the second dose will be a second first dose. There will be no accumulation of the LNP. Our expectation is that it will behave just like the first 60 mg dose. There was the opportunity, like John said, to explore a situation in which you increase the AUC, the area under the curve, without pushing basically the max higher. Obviously, with 75 mg, it would be pushing the Cmax as part of that. I think we've already disclosed that with 75 mg, we're seeing some initial tolerability data, which is as good as the 60 mg, so grade 1 or below. I think it's really behaving as one of the best LNP franchisees that I've seen. Got it. Thanks for all that color. Let's talk about the 12.5 μM because that's something you highlighted. Obviously, you could get even higher with further studies, which remains to be seen. Just wondering, for this level, you said it's a therapeutic level already, right? In your interaction with physicians or patient community, what's the sense and their feeling towards this level as an outcome? I think the patient and physician response has been overwhelming. I think there's a lot of enthusiasm for what we've shown. This got picked up in The New York Times. We got patient emails from around the world. I can tell you that the physicians have never seen something like this. The field to date has really only had augmentation therapy. In augmentation therapy, you're basically adding some extra M protein into the system. That's the normal protein, but it is going to gradually wash out of the system. It isn't regulated normally. It doesn't turn on when you're sick, which is what your gene will do. Our numbers are, in some ways, a floor. They will go up when the body needs more. With augmentation, that doesn't happen, and you aren't affecting Z. The body is still making a lot of Z, so the liver is going to be sick. It's going to interfere with lung function, et c. It's a good option for patients, but it's not perfect. It's not a cure. This BEAM-302 is the first time that the field has seen a potentially curative, fundamentally addressing the root cause of the disease medicine. I think you can see in our enrollment rate, which has been very swift, that there's a lot of people signing up. We have wait lists. We have a lot of investigators who have been eager to get involved. When we do informal and early market research, we do test these sorts of profiles. I think it seems clear to us that the target product profile that we're showing of a one-time correction that gets you into these carrier ranges, that is raising your M and eliminating Z, and you only take once and you're then freed of chronic therapy is very popular. Got it. Could you be a little more specific in terms of in your market research, when you put forth the profile into the carrier range, did you give a specific micromolar number? Also, from that research, do you get a sense of the proportion of patients who will be pursuing a therapy? Yeah. I think it is a significant proportion. I'm not going to give specific ranges or numbers. It is not only a large proportion, but we think it's the winning proposal. It's the winning TPP when you kind of compare it across different options. In terms of levels, I think there's obviously a lot of debate about that. Physicians in the community, they are used to numbers in the mid-teens, let's say, for augmentation therapy. The augmentation gets you to about 16 µM. It's really apples and oranges, if you think about it. That is a number that will not go up when you're sick, whereas our numbers will go up to that and higher when people are infected. They operate differently. The 16 µM is really about 11 M on top of a remaining 5 Z. The Z is the baseline. That's already there. It's still being produced by the body. The augmentation just goes on top of that. You haven't corrected the toxic protein. In our case, we're really substituting and replacing toxic protein with normal protein, really turning it into 11 M. It will go up on therapy. I think that that's really where we get the confidence to say we're already in what we believe is a functional curative range. It's not to say that we can't go higher. I think we probably can. We'll be exploring that and pushing on those in the ways that Pino mentioned. Back to the clinical genetics, we know that a person with the profile that we have just produced using this drug would not have any progression and wouldn't have the disease. Got it. Got it. Speaking to potentially also getting even higher levels, you did dose three patients at the 75 mg cohort. Of course, the data will come early 2026, along with the data from the double dose 60 mg cohort, which you haven't dosed a patient yet, I assume. Just on the patient you dosed on 75 mg, I think you did mention something about dose response. You haven't given any detail, but you did mention dose response. Can you give us a little more color on what that means and how to think about that? Yeah. In addition to the safety that Pino mentioned, we also do see signs that as a fold change, what percentage does the alpha-1 level go up for a given patient after therapy? We do see that 75 mg may be giving us more than the 60 mg. Than 60 mg. That's right. I think it's early days, and there's a lot of follow-up still to do. Got it. Very helpful. Definitely looking forward to the early 2026 data. Your competitor reported data yesterday. This is the RNA editing approach. This obviously had a lot of investor discussions. I was curious if you could comment on your thoughts of that data. Yeah, please. Yeah. I think the data sort of indicates and supports that the profile that we've already disclosed with our program is, frankly, superior to what has been shown yesterday. I think what we were talking before is very important, that quantity is only one aspect of what we need to look at. Quality, and by quality, I mean how much M you actually have, is really the critical data point here. Unfortunately, for Wave, what was disclosed is that the total level is roughly in the same ballpark as we are, but actually, the amount of M is, or actually, I would say the amount of Z that is retained is significantly higher than what we are seeing. Depending on whether it's the 200 mg or 400 mg, I think it's between 40% or 60% of Z that is residual. A significant component of those 12 µM is now actually Z, and M is only a fraction of that. I think it's between 4 μM and 7 μM, if I remember correctly, the data. That is the issue because Z, in my view, is the offending protein. Ultimately, what you want to eliminate is the Z because Z is not only a clear function of an accumulation in the liver, because obviously, there's going to be more Z in the liver than what you see in the blood, because the Z is not secreted that well. Also, Z in the blood in itself actually contributes to inflammation of the lung. It is actually a chemoattractant for neutrophils. For example, it forms polymers. The elimination of Z has a direct impact on both liver and lung disease. You really want to eliminate that from the system. The other aspects of the data, at least from my interpretation, seem that already at 200 mg of antisense oligos, the system that they're operating seems to have hit a certain saturation level. It remains to be seen whether additional dosing may unlock that, but at least that's from the initial data that they've disclosed. Yeah. Maybe I'd add only one thing to that, which is we talked about total levels. There's another key readout, which would be functional AAT. In functional AAT, you directly measure the concentration you're getting of protein that is effective at inhibiting neutrophil elastase and binding to neutrophil elastase. That's the role of AAT to protect you from infection. We've quantified that. In our case, with the 12.5 μM, you're getting 11 μM of M. Our functional is right around 11 µM or more. It's nearly all functional at that point. The issue with having a lot of Z left over is Z is much less functional. You get a little bit of activity from Z, but not much. It wasn't quantified in the data set, but we would predict the functional AAT readouts would be lower as a result. Maybe one additional point to add is that to some extent, if you look at the preclinical data, it's probably what we are seeing could have been predicted. They seem to have reached a 50% editing sort of ceiling that they couldn't break through. It may be a consequence of the mechanism of action. In that case, the production of Z mRNA is left untouched, so you have a continuous production of Z mRNA. Now you're asking the system to be able to take care of all of that amount in a meaningful way so that you actually see the overall reduction of Z. It seems that that may be difficult to achieve with that approach. Got it. Got it. Thanks for the cover. Maybe let's switch to the regulatory perspective. Congrats on the recent RMAT designation for BEAM-302 in AATD. Does this allow you to interact with the FDA on a frequent basis? Is the dialogue starting, or has it started? When could we hear some updates from there? Yeah. It does. We've had a lot of interactions with the FDA. We often get questions about, is the FDA still open for business? Is everything on track there? We think it is. In fact, the RMAT designation for alpha-1 came in early before it was even due. We have many of the same reviewers all showing up to all the calls. I think it's all on track. We won't obviously give much specifics in terms of when we're interacting with the FDA or exactly what the dialogue is like. That is certainly something that is ongoing in parallel. One of the reasons I characterized the kind of late-stage dose exploration that we're doing in phase I as this sort of bake-off between these different possible doses and schedules is we want to finalize that because we want to move forward to a registration path. In parallel to that, we need to align with the FDA on what that is. Part of the early 2026 update that we're planning would be not just data, but we would expect to be able to give some color commentary on what are we hearing from the FDA, what is the path to market, and what is the way to get this filed. I think our expectation for a drug that is working on the fundamental mechanism of the disease in this way and has the dramatic effects like this is having is that we may have access to some sort of accelerated pathway based on levels: the total AAT, functional AAT, reduction of M, reduction of Z. Those are all independently predictive of clinical benefit in their own way. Of course, altogether, we've shown really correcting the pathophysiology of the disease. We'll see. We have to work through that with the FDA. I think that remains our sort of base case assumption. The alternative would be we have to go do a randomized trial. Those are also doable in this disease. That would take a little longer. I think a drug that is this dramatically active early in development, in my experience, tends to be a candidate for finding a faster path to patients and then following up, of course, with longer-term studies to build out the data set. Right. Thanks. That's super helpful. I certainly think one of the investor focuses, as you were saying, is whether FDA would be OK with using serum AAT as a surrogate marker for accelerated approval. You did mention the clinical benefit, and that's going to be the key rationale for FDA to agree to use this pathway. I was wondering, could you elaborate a little bit more, making a case, I guess, if FDA is weighing this information, you would need to present the case why this is the appropriate biomarker. Could you give us a sense of how you will make this case? Maybe I can take that. First of all, the important thing that I want to stress is that the data that we are able to generate is not just serum AAT level, which clearly is an important thing. It's the demonstration of M going up, Z being reduced, the functional activity are all consistent to basically the gene function that is what we are restoring. The reason why that is important is because that has never been possible to do before. Enzyme replacement therapy does not do that. It does not take care of Z. By the way, upregulation is also what we expect to see because the promoter region is exactly the same. It's really that totality of data package that demonstrates the restoration of the gene, unquestionably, frankly. I think even the data that we've already generated demonstrates that. There is a very strong understanding of the relationship between the gene function or the Z protein and why the disease is progressive. The science basically is absolutely well described in the literature and for many years that if you then now are restoring the functionality of the gene to the extent that we do, particularly when you are recreating essentially an MZ profile, as we are doing with 20.5 µM, but importantly, 90% of that being M and potentially even greater than that, I think it's going to be difficult from a scientific point of view to suggest that that's not evidence of restoration of the gene. Will they accept that that may be supported additionally by some clinical evidence? It's possible. Our most optimistic would be that gene function restoration should be sufficient or could be sufficient for a full approval. Should that not be the case, we would expect that at least an accelerated approval on the back of that should be certainly scientifically valid and supportable. At that point, you may support some confirmatory trial data, potentially with some clinical endpoint. In that case, we think that actually seated densitometry is the way to do it. It's not yet completely endorsed by the FDA. It's certainly endorsed by the EMA. There is definitely sort of a hint that they are considering that very strongly. The body of evidence, a seated densitometry, is the way to characterize the alpha-1. It's becoming so overwhelming that actually, I think they would find it difficult not to accept that. In that case, seated densitometry, there are publications already available with natural history data that looks at untreated individuals. You should be able to see with 100 or so individuals over a period of two to three years already see the benefit, basically, that those would do, particularly with the strong data that we've already generated. The effect size is fundamentally huge in this. I think that's our position. That's really the argument that we're going to be making. Great. Thank you for laying that out for us. It is a very convincing, compelling series of arguments. I appreciate that insight. Obviously, the company, we touched upon this multiple times. You will have an update early 2026. I was wondering two things. One, could it be as early as the January conference, for example? I was also, can you also give us a sense of the data, what it would look like at that time? A curious question, given that you talk about the importance of the M level can go up at the period of acute phase reactions. Any potential that you can capture an event like that by the next update? Yeah. I would say I wouldn't look to the health care conference. That's the very beginning of 2026. The guidance is really early, so you can think of that as the first quarter, thereabouts. I think in terms of what we would share, we'd like to see, obviously, all the same things we've been looking at here. You'd want to see total alpha-1 levels, functional levels, how much M is being produced, and how much Z is being reduced. We would have quite significant durability at that point across all of our dose levels, of course. The very earliest patients would be out to a year or so at that point, and even some of the 60 mg patients would be maybe half a year or more. Of course, less with the more recent cohorts. That would give a good sense of durability, which, again, we expect to be rock solid given it's gene editing. Yes, we will be looking at inducibility. That one, you have to get a little lucky. It was nice to see in the data set from Wave yesterday that there was inducibility shown with a fortunate or unfortunate, I guess, medical event, but they got a good reading on it. We're definitely looking to see that. That's exactly what would be expected from these editing modalities. When the gene turns on, you're going to create a lot more transcripts. In the case of RNA editing, you edit a similar proportion of those transcripts, and that was what was seen by Wave. In our case, all the transcripts would be corrected, so you would just be generating a lot more M. I think that would roughly give you a pretty—oh, I guess the last thing that would be part of that data set, we will be dosing patients in what we call part B of the trial. Recall, most of the patients we've been talking about to date are in this part A, where it's the majority of patients where they are primarily lung disease with maybe some liver. We have been excluding the patients who really have sick livers. It's probably about 10% - 15% of the population where liver is actually the leading symptom more than lung, so we're going to bring those into this second set of cohorts. We talked about initiating enrollment already in that. We're going to start at 30 mg. The drug has been safe enough. We don't have to go back to 15 mg the way we did with part A. I expect to have some number of patients dosed as well and give a read on the safety in those heavy liver involvement patients at that time as well. Got it. What about the three to six patients for the double 60 mg dose? Do you expect most or all of the patients with data there? Yeah. We expect to include those patients. They obviously will depend somewhat on recruitment, but we think that that should be part of that data package. You would have 60 mg, six people; 75 mg, seven people; six people; and then some of the part B. We intend to share all of the data that we would have available at the time. Importantly, also maybe greater clarity on the regulatory cleanup as part of that, which I think is obviously of interest and important as part of that. As far as inducibility is concerned, we obviously are measuring CRP as well, just like you've seen from M. I don't know. We're hoping on the flu season coming up. We'll see how it goes. Interesting. Thanks. Lastly, I was wondering if you are able to comment on Prime Medicine's move to develop a prime editing treatment for AATD. Yeah. In this competitive landscape, you've obviously got RNA editing. I think there are multiple players out there. The preclinical packages look comparable. We've seen obviously one of those from Wave. Let's see more. It's a big disease. There's room for a lot of different kinds of modalities, although again, we feel quite good about where BEAM-302 and base editing stack up. You have next-gen augmentation therapy coming, which is interesting. There again, you're not going to be regulated normally. You're still leaving Z in the system, but you might have longer acting, less frequent dosing for the M that you're adding. That's interesting as well. There will be certainly other gene editing approaches. There's a few of those. Prime is one. We actually are partners with Prime. We have exclusive rights to prime editing for making transition mutations. That's a place where there's actually an ongoing legal dialogue with them over that program in terms of their rights to advance it relative to our development in alpha-1. Obviously, we'll let that play out. There will be other gene editing players come forward. It's not going to be an added one. I think the real question is really for BEAM-302, is there anything we have left unfixed? Is there really an unmet need? I think if we can edit the entire liver, if we're creating M, if we've really eliminated Z, at that point, I'm not sure how much room there is to improve with additional gene editing players. Again, patients benefit from having more options. Ultimately, that's a good place for the field to go. Great. Great. In the remainder of the time, let's discuss sickle cell. You completed enrollment for the pivotal study. You dosed 30 patients as of July and plan to provide an update at year end 2025. What will be the focus for that data? At what point do you plan to discuss the regulatory path with the FDA? Do you need, like, how much data do you need to start the filing? Yeah. We think that the regulatory path has been described already through the CASGEVY approach, and our approach is very similar. That's where the 30 patients data package is really very similar to what they've done. We think that there is already a path to that. Initially, we've also announced that we started to dose adolescents, and we plan to incorporate some of that as part of that. In terms of when we look to interact with the FDA, we don't provide the dates in particular, but we have basically the data package already to start some meaningful interactions with them. Very soon as we have greater clarity, we'll obviously disclose that. We're very confident about what needs to be as part of that BLA package in order to support the licensure, and because there's a precedent, basically. Yep. You had mentioned ASH. I think we'll have an updated ASH that really starts to show expanded durability. You have a number of patients out past a year. Obviously, we'd hope to see the same signs of strong differentiation we've seen previously, both in the process, a few cycles of mobilization, faster time to engraftment than what has been seen previously, as well as the hematology parameters, higher levels of F, lower levels of S in terms of eliminating sickle protein and resolution of anemia. Got it. Great. I think that's all the time we had today. I wanted to thank the team for a very wholesome discussion and all the insights that you provided for our audience. Thank you for being here. Thank you very much. Thank s everyone.