Great. Good afternoon, everyone, and welcome to TD Cowen's I&I Summit. My name is Tyler Van Buren, Senior Biotech Analyst at TD Cowen. For this next session, we are very excited to have a fireside chat with Maze Therapeutics. From Maze, we have Jason Coloma, the Chief Executive Officer. Jason, thank you very much for joining me. It's a pleasure to have you. Yeah, thanks, Tyler. Thanks for having me. For those of you in the audience, if you have questions, you can submit them through the web portal, and we'll do our best to get them asked. We'll go ahead and get right into it since we have 20-plus minutes. Jason, I thought it would be good with ASN just passing, maybe discussing some brief highlights or takeaways from the conference or observations that you guys noted. Yeah, absolutely. For those not familiar with our story, we develop small molecule precision medicines, primarily focused on kidney disease. Last week was the American Society of Nephrology, which is the largest conference dedicated to kidney therapies, as well as for overall kidney research. We had seven abstracts, Tyler, that are accepted, which was great, a great showing for our team and our scientists and had presentations as well. Probably three things to highlight in terms of what we were able to kind of highlight, especially for APOL1-mediated kidney disease. The first thing is that we had published on a genotyping study. One of the key takeaways there was it really did reinforce what we were seeing in terms of the literature on the percentage of people that had the variants. What we reported was the previous literature had described in a general population, around 13% of the individuals that had been genotyped had the pathogenic variants that cause APOL1 kidney disease. In our study, because it was enriched for kidney disease patients, we had about 18%, so a little bit higher than the literature, but mainly explained by the fact that we had enriched for kidney disease. I think suffice to say 13%-18% in that range, having the variants that end up causing disease, just reinforcing the literature. The other thing that was important from that study is that sort of the breakdown. We've always talked about a million individuals in the U.S. have the variants, at least 250,000 individuals could benefit from a therapy, and a split between those that have diabetes and do not have diabetes. We reinforced the literature in the sense that about 40% of the individuals do have diabetes. It is about non-diabetes, 60%, diabetics, 40%, also reinforcing the literature. The second thing that we did is there has been some discussion, I think, in the sense that should we, in a broad AMKD patient population, whether or not diabetes, should there be an effect in diabetes? There has been some previous literature talking about the, let's say, decreased incidence of disease with those with diabetes. We and other sponsors as well have been starting to educate the field and investors as it is not about incidence, it is about progression to disease. What we published on was the fact that those that have progressive disease, it is aggressive and progresses at about the same rate with or without diabetes. Conceptually, if a therapy should have an impact, whether or not an individual has diabetes or not. The other thing that we had previously published on related to that was we had found a protective variant, N264K. Just to remind folks, that protective variant protected individuals, whether they had diabetes or not. Sort of illustrating if we truly are us, as well as some of the other sponsors mimicking the protective variant, which is the hypothesis, we should see an effect in either population. What's nice to see at ASN last week was another group, Natera, reinforced that finding in terms of N264K, which really demonstrated, again, this concept that it should work in both patient populations. The last thing that's probably important is, I think, prevailing sort of mechanistic understanding of the hypothesis of how APOL1 causes disease. Just to first off say, the genetics were known for a very long time. No one really understood function. Us and other sponsors worked really hard to better understand that the pathogenic variants that cause disease cause an aberrant, if you will, function of this pore channel in the podocyte, ultimately leading to that nephrotoxicity. We want to be able to, if you will, block that pore, if not, if you will, disrupt the assembly of that pore ever to be formed to begin with. We published some additional data to demonstrate and reinforce that the mechanism by which APOL1 is causing disease is truly that it is acting as a pore channel, and so that an inhibitor of that function should continue to see benefit in terms of the clinic. That is important because there were dozens of different hypotheses on how APOL1 could end up deciding disease. Now we have a better understanding of that. It is just really nice to see everyone that is working in APOL1, not just us, sort of the common understanding about at least 250,000 individuals could benefit. You should expect to see a signal whether you have diabetes or not. The prevailing hypothesis that it is acting as an aberrant pore channel seems to be sort of coalescing around that overall scientific hypothesis. It is just really great to see. Obviously, we are really happy to see all of those sort of get reinforced and accepted and being able to talk about the larger community at ASN. That's great. Very interesting. Thank you very much for recapping that for us. Maybe we'll get straight into the Maze 829 program, your APOL1 inhibitor in development for AMKD. Can you just give a high-level overview of that compound and how it compares to others, obviously, in particular, inaxaplin in terms of its properties? Yeah. We are developing a small molecule. We have a dual mechanism APOL1 inhibitor. We are advancing it for the potential treatment for what is called APOL1-mediated kidney disease. The idea is not only that you might be able to inhibit the pore itself, but also what we call a dual mechanism is that we also have, if you will, differentiated biology relative to other compounds in the literature and the fact that we not only block the pore, but we also disrupt the assembly of that pore to begin with. That ultimately, in terms of differentiation, leads us to, at least in in vivo proof of concept models and kidney models, we show that we had a significant potency advantage relative to any molecule that had been described in the literature before and then further reinforced by other work that we had done. The idea, of course, is that previous compounds have shown clinical proof of concept in FSGS patients previously. That's great because we know that the mechanism can demonstrate clinical benefit. Our goal with the Q1 26 readout, Tyler, is number one, what we want to be able to do is be one of the first companies, or if not the first company, to demonstrate clinical proof of concept in broad AMKD with or without diabetes. We set that minimum threshold in terms of seeing a signal, at least 30% reduction of proteinuria relative to baseline. Happy to describe how we landed on that number. What we've been communicating to folks at ASN and this week as well is if we see, if you will, a signal and show clinical proof of concept in that patient population in Q1 2026, we could be in a position to potentially start planning for a phase IIb/III study, much similar to, obviously, what Vertex had done with their data a couple of years ago. That's great. Just to be clear with the Q1 update, we're going to get data from both the diabetes and the non-diabetes AMKD cohorts. Yeah. When we set off on Horizon, our goal was because it had shown clinical proof of concept in a rare population, we'd like to show for the first time and hopefully broad AMKD with or without diabetes. That was the intent. If we see that signal at 30% reduction, to us, that signifies that we have demonstrated clinical proof of concept, as we've talked about that with nephrologists. I think based off of the strength of that data, should we get there, the idea is then we can start initiating planning for what a, if you will, 2Be3 design could look like. Got it. Have you said approximately how many patients we should expect per cohort? Yeah. I think for us, in terms of the readout, people should expect an order of magnitude of what Vertex described in their first study. That is why we describe it as an initial set of data that should allow us to, if we see that signal and that percentage of patients, we would hope to, again, think about that 30% number, look at the totality of the data. If we see that signal, then we can ideally push as fast as possible to something that could look more like a registrational study. Got it. Is it possible that we see FSGS patients in Q1, or do you think we need to wait for a pure FSGS update later? Yeah. I think our goal for Horizon on the first readout is to show clinical POC in that broad AMKD patient population. We are enrolling patients that have FSGS, of course, if we have that. Just to set expectations, our goal is to focus really the investors on we could be the first company to show POC in that broad patient population. If we do, that gets us from a rare population of FSGS to a broader set of patients who can benefit. Got it. When you were recapping some of the ASN highlights, you talked about the variant being protective both in diabetes and in the broader AKD population. As we think about proteinuria reductions, is that variant equally protective in diabetes versus non-diabetic patients, or is there still somewhat of a difference that might manifest in the proteinuria reductions? Yeah. I think what's interesting from the genetics, it progresses at the same rate and equally as aggressive. That's interesting that whether you have diabetes or not. Conceptually, again, if you could intervene with a medicine, you should see an effect in both. The N264K, specifically the protective variant that you just mentioned, yes, it does at the same rate protect those with or without diabetes. We know that's nice that we have the natural, if you will, genetic experiment showing that. That's why for all of our programs, Tyler, we do look at not just pathogenic variants, we look for protective variants as it might allow us to better understand function and mechanism. As we design our small molecules, as we did with this program, as well as for SLC6A19, if we can mimic that protective effect, that should allow us to reap that benefit you see clinically. That is the goal here. I think what's nice is we'll hopefully try to demonstrate that in Q1 2026. Okay. Can you just elaborate on how you picked that 30% threshold, especially with the 48% with inaxaplin out there in that very small FSGS data set and your confidence in potentially being able to achieve significant proteinuria reductions? Yeah. The previous literature described a rare patient population. This is broad AMKD, so more heterogeneous in disease, both with or without diabetes. The 30%, probably four things to highlight for investors. 30% is a number that if you look at some of the analyses, it correlates very well to eGFR slope, which is most likely just like in IGAN, the approvable pathway for particular proteinuric diseases. The second is in the guidelines, so KDIGO, 30% as being clinical significance. UACR specifically is highlighted as what you should be measuring for this more moderate disease. The third is if you just look historically, things like other classes like SGLT2, they had 30% as their bar, eventually short of translating to, if you will, outcome data and eGFR benefit. The last is, which I'm not sure if you saw this, Tyler, but Parasol, which is the group at University of Michigan, which was instrumental in working with the agency to help define for FSGS that rare patient population, that proteinuria alone should be the approvable endpoint. The Parasol group has taken on two kidney indications, one of them PMN and the other APOL1-mediated kidney disease. In that press release, they did describe a couple of papers that even referenced UACRs, the proteinuric measurement that people should be using, and the 30% number as a good minimal threshold to see clinical benefit for patients. Those are the four things we highlight as the rationale for 30%. I think that allows us to at least see where the bar should be. If we exceed that, of course, that would be great. Great. Preclinically, what have you seen with A-29 that gives you confidence that you could have significant reductions in proteinuria in patients? Yeah. So we've previously published, especially in particular the BAC transgenic mouse model, which has high levels of proteinuria. We've done that both in a chronic setting as well as preventive and shown that we can get to doses that get to an AC90 that effectively can get into that range of at least a 30% reduction. I think in terms of translation, we've seen that in in vitro models. We've seen that in multiple models. More importantly, we saw that in the BAC transgenic mouse data. We know why that might be translatable is because we were able to synthesize inaxaplin, bring that into that model based off exposure, where their clinical dose was, and we saw proteinuria reduction. We already know that they showed clinical proof of concept in patients. Very good confidence in the sense of this being a translatable model where we should see at least some level of proteinuria reduction as we move into the clinical trial. That's great. As we see this initial data and as we think about the development plan moving forward, how do you think about baseline proteinuria? In other words, does baseline proteinuria have a significant impact on the% reduction that you can achieve in patients? If people have higher baseline proteinuria, do you have a greater potential to have a more significant reduction, or is it independent of baseline proteinuria? Curious to get your latest thoughts there. Yeah. I think that, I mean, at least for APOL1-mediated kidney disease, we saw that in highly proteinuric patients, which was published in the New England paper, is that you had that significant drop, right? Those were highly proteinuric patients. We know that a group has not described, at least publicly, clinical proof of concept and being able to show that 30% could be significant. As we talk to investigators and we talk to nephrologists, setting the bar there across diabetes and non-diabetic patients, to them, they would see that as clinical proof of concept and allow us to think about justifying the progression into a more registrational study. Got it. Because the broader patient population will have obviously lower baseline proteinuria than FSGS, it's reasonable to assume that you'll have maybe lower proteinuria reductions in that population than FSGS. Correct. Yeah. Okay. All right. Okay. Yeah. I guess that's going to be somewhat challenging, right? Because if you look at the inaxaplin data, it's going to be in a different patient population than what you guys are showing, right? We won't be able to necessarily compare it head to head, but obviously, clinical proof of concept to have the first data in the broader patient population will be really special, right, to be the first. Yeah. I think it'll be incredibly impactful because then you get close to that, at least 250,000 instead of the, say, 10,000 patients that have FSGS with APOL1 variants, right? I think for us, I think that was always the goal because they've already shown clinical proof of concept, and the rare will form our sort of like data. They will also have data. They've also guided, Tyler, of course, that they should have data in a broad AMKD patient population by the end of 2026. I think that also will help inform that. I think as we think about the overall opportunity, what we always guide investors is that at least 250,000 individuals, this is at least the same size as the IGAN population, if not bigger. We have also described, of course, that in terms of how to think about this, different multiple agents would probably be necessary in order to benefit this patient population. We have seen that at least in other analogs that we describe to investors like ATTR-CM, where you saw diagnostic testing being set, multiple participants being able to ultimately make an impact for patients. We are even seeing that in kidney, of course. IGAN has multiple participants that allow us to kind of think about differentiating not just on a molecule level, but also the patient populations they can address. Great. I have a follow-up question on the market in a second. Just to be clear, the mechanistic difference from A-29 to inaxaplin, it is twofold. It is the increased potency and then also mechanistically the fact that you are inhibiting pore assembly. Is that in Vitro? In addition to blocking the pore. Yeah. Yep. The pore assembly inhibition, which is obviously a novel addition to your mechanism of action, is that you observe that in Vitro and you guys are confident that that's happening in vivo? Yeah. The idea there, of course, is that something that's probably important to point out, APOL1 in the podocyte rapidly turns over. It's less than an hour. I think some people reported like every 42 minutes that it's turning over. The fact that you want to be able to intervene, at least in vivo in the sense of not just inhibiting the pore because it's a little bit of whack-a-mole, you also want to just make sure, and ideally, that you're disrupting the assembly such that they never insert into the membrane of the podocytes, right? Conceptually, that makes sense to us. I think partially helps, at least to us and others, explain why we are more potent in particular in vivo systems. Great. 250,000 patients across the broader AMKD landscape with diabetics, non-diabetics, FSGS, I mean, that's clearly a very sizable population. I guess, technically almost exiting the orphan drug range. You mentioned the 13%-18% earlier. That's elevated, right? I think you previously said it was more like 11% incidence of AMKD. No, no. If you look at our previous, 13% of the general population in the Black community have the variants that could cause disease, right? That was the sort of overall number. Our genotyping study and the epidemiology around it was a little bit higher, 18%, but partially explained, I think, by the fact that we were looking specifically at those that had kidney disease. It was not a general population. I think in that range, it is in there. I think how do we get to 250,000? It is those that have proteinuric disease at a certain level that could be best, if you will, served with a potential therapy. Got it. All right. We're up on time here, but have to ask about 782. You guys had great data with 782, SLC6A19, as I like to call it. And just can you touch on both PKU? I mean, what you saw there clearly pretty striking relative to benchmarks. And then also CKD, which was a pretty interesting surprise with eGFR dip and just your conviction that mechanistically eGFR dip could lead to real efficacy in CKD or the broader almost SGLT2-like indications. Yeah. I think for the 782, which targets SLC6A19, in terms of PKU, this is a potential therapy that can work across the entire spectrum of disease, in particular the severe classical form, which is the majority of the patients. Over 60% of the patients have severe classical form of disease clinically defined as greater than 1,200 micromolars of Phe in their plasma. The idea is that we do not have a therapy that relies on, if you will, the older agents that have been approved, residual enzyme, or, if you will, cofactors to allow that enzyme to work. The concept, very straightforwardly, was, "Hey, why do not we just get rid of the toxic substrate itself?" If we can safely do that, we can potentially affect all of the patients, including those that are not served by, that are defined as severe classical. Because what we want to do, Tyler, clinically is that 1,200 number, you want to try to reduce them below 360 micromolars. Why is that important? That is when you can start to remove people off of their onerous medical diet. For some of the children, that means that you reduce the probability of seeing these neuropsychiatric or neurodevelopmental issues that happen when the plasma Phe gets into the brain, right? We had an agent. What we reported is that there was another investigator who had shown clinical proof of concept of inhibiting the target and having an impact of plasma Phe. They also showed that a non-invasive urinary biomarker, plasma Phe, if you will, excretion mapped very closely to that plasma Phe reduction. Why is plasma Phe important? That is the approvable endpoint for PKU patients and regulatory path, right? What we showed in the study was that we had clearly best-in-class properties and the ability to excrete that urinary Phe. We knew that the other investigator had shown at one dose a 10-fold increase, if you will, of that urinary Phe biomarker that translated in the patients on average 44% reduction of plasma Phe. We significantly cleared that over 42 weeks in terms of our urinary Phe excretion. That gives us the confidence, of course, to now run our patient study in PKU where we can exhibit best-in-class and the ability to reduce plasma Phe and ideally convert more patients below that 360 number. If we look at their data that they reported of the 19 patients that they reported, only six of them got below the 360 number. If we can do better than that, that would be excellent. On CKD, what we did, we generated the first genetic association to the target and being able to have an improvement on kidney health. We are equally excited about that concept in kidney disease. We not only demonstrated genetics, we validated some academic work in terms of target validation. We also showed for the first time in vivo proof of concept in the kidney model that allowed us to see and contextualize it to SGLT2s, which are an important anchor in kidney health. We showed that we can not only work with it in an additive sense, but we also can do a little bit better in the sense of lowering proteinuria in that model. We had genetics, we had target validation, we had in vivo POC. What we did based off of that, nephrologist gave us some feedback before we started the phase one, which was, "What you should do is collect serum creatinine so that you can calculate eGFR as an exploratory biomarker." Because what we know, and this is important for the investors to understand, is that all approved CKD agents or therapies show what's called this eGFR dip as it's indicative of renal protection in the long term. We would like to actually see that even in your healthy volunteer study because it reinforces all of your great genetic preclinical data and gives us even more confidence to do the POC study. Suffice to say, we did show that. We showed eGFR dip. We showed that there was a dose relationship. We showed it was true. Secondarily, we showed it was true to the agent because when we pulled them off the agent, the eGFR bounced back up. We also showed that it was in line where SGLT2s were in terms of seeing that eGFR dip as well. Now with that clinical proof of mechanism data plus the work that I described earlier, I think there's a lot of excitement in the nephrology community. Can we try to show clinical proof of concept in patients? The way we position that from a target product profile is there's a lot of people who aren't responding to current standard of care, including SGLT2s. We know up to 25% of the individuals are not responding adequately to SGLT2s. We know that the one-year discontinuation rates on SGLT2s is very high. Tyler, it's like 30-40%, partly due to the complications like hypoglycemia and urinary tract infections. Minimally, as a TPP, if we can be just as efficacious as SGLT2 but not have the liabilities, that could be very interesting. Moreover, if we can actually just exceed that in terms of efficacy, you're talking about something that potentially can change the way that kidney disease patients are treated today. Yep. Could be a mega blockbuster. That's my speculation. In other words. Wonderful. We are way over time, so we better wrap up. Jason, thank you so much for the fantastic discussion. Thanks to everyone for logging in. Yeah. Thank you, Tyler. Thank you, everyone.