Good afternoon. Welcome to this panel discussion, ADAR RNA editing. My name's Keay Nakae . I'm the Director of Research and one of the Senior Biotech Analysts here at Chardan. Joining me on the panel are executives from our participating companies, Dr. Paul Bolno, President and CEO of Wave Life , Dennis Hom, CFO of ProQR, and Dr. Sriram Sathy, CSO at AIRNA. Gentlemen, thank you all for joining us today. I think most of our audience is familiar with ADAR, adenosine deaminase acting on RNA. It's an endogenous enzyme that naturally edits one selected adenosine on double-stranded RNA, flips it to an enzyme, which is then read as a guanosine base during translation. This results in amino acid alterations and often entails changes in protein function, thus opening the door for the development of novel therapeutics. Two years ago, we hosted our first panel on ADAR.
I'm happy to get the band back together again two years post to get an update on what's happening in the field. Again, thanks all of you for joining us today. Paul, let me start with you. Your lead program, similar to ProQR and AIRNA, is targeting alpha-1 antitrypsin deficiency. It does represent an attractive initial target. It's a validated target caused by a single point mutation. There are easily measurable biomarkers. You're the first in the clinic. You're the first with data. Tell us what you've evaluated to date and what you're seeing.
Thank you everybody for sharing lunch with us, and thank you, Keay, for bringing the band back together on editing. I think, as you mentioned, two years ago, we were all here and you read about adenosine deaminases and flipping bases, and we were talking about kind of what I call the theoretical, you know, could you drive editing and base correction that could actually make a physiologic protein that could actually recapitulate the human wild type protein and do all of those wonderful functions? I think we had the confidence of this is working in a dish and it's starting to work in mice. Flash forward over the last year where we took the idea from then preclinical data a year ago at this conference, we're excited to bring this into the clinic and drive data to where we were delivering data a little over a month ago.
It's pretty astounding. What we've learned along the way and what we'll continue to learn is really stepping back and thinking about this target. It's so important when we do think about target biology to really think about what's the driver for treating the disease and what's the modality and the mechanism with the modality treats that disease. Coming into the treatment of alpha-1, as you pointed out, this community, patient community, physician community, investor community, had a lot of different sets of expectations based on IV protein replacement therapy. If we think about the gold standard of IV protein replacement therapy, it's really pouring water into a bucket that has holes in it.
This notion of how do you get from a ZZ phenotype, so these are patients who are going to be at risk of lung disease and liver disease who go on to progress, how do you transition them phenotypically to an MZ patient, so a patient who now has low risk of lung disease, low risk of liver disease? That's where when we think about protein replacement therapy and a number of companies in the space, people have in their head 11 µM, and then suddenly with inhibitors, people are like, do you need 15 µM or 20 µM?
This notion of what I call kind of rising this bar on protein replacement, that was always a notion of how do you put protein into a patient's body who can't make that protein so when they do have this acute phase response, they're going to consume that protein and it's going to go away. If you think about the 11 µM and why there's been debate of whether or not that's the right bar, it's because if you assume that a patient has 11 µM because you infused it and they have an acute phase response, they may actually have zero at the time of the event. There's nothing. We have to step back and when we look and think about our data in the context, which is super exciting, we will talk about the acute phase response and the ability to generate more protein than seen with inhibitors.
The notion in editing has to be a different mindset. The mindset in editing is you're actually correcting that mutation on the transcript, so you're fixing it within the patient so that now truly a ZZ patient becomes an MZ patient. That means they should have, you know, still to be MZ above 11 µM and in these data that we presented, we saw 12 µM of total protein, we're up to 13 µM as we continue the study, so we saw MZ-like protein total levels. Most importantly, we saw the exact correlation between what was supposed to happen with editing, meaning a shift from Z protein to M protein. We saw that where patients went from 0% M protein at the beginning of the study to 44% of the protein being M protein after the single dose.
This is still the lowest dose cohort to ultimately about 65% M protein at the multi-dose, still in the lowest first cohort. Our approach using novelty of chemistry gives a stable drug that gets in the cell, stays there subcutaneously because it's GalNAc conjugated and can be there to drive editing in a highly durable feature. If we think about what editing is supposed to do, it's supposed to correct Z protein to M protein. Most importantly, and this gets to the fundamental treatment of alpha-1 antitrypsin deficiency, is that when a patient has an acute phase event, meaning when they have an IL-6 mediated response that turns on transcript that is trying physiologically, normally these patients would then produce protein to protect their organs, these ZZ patients have nothing. If you get to MZ, you can start mounting a response. That's the exact response we saw.
These patients went from, you know, 0 M to now nearly 11 of M protein and over 20 µM of endogenous protein production, stayed there for about a month and then came back down to their baseline levels from the study. I think this notion of being able to think about the treatment of alpha-1 antitrypsin deficiency in the setting of editing is highly important in ultimately driving a therapeutic forward. We saw we were there in the clinical data, as I was saying, at the lowest dose cohort. That's the 200 mg cohort. We're excited about the 400 mg cohort, which will come in Q1, where again it's about driving how can we get not just more M protein, but really durability. We're probably now at monthly, bimonthly, we think quarterly.
By the time we get to the first quarter at 400 mg, we're going to be able to see how long we have editing. We were sustaining editing on the single dose out past two months.
Great, yeah, I think with a little bit of time, people are starting to appreciate what actually happened in that one patient with the exacerbation and saw that your drug is allowing it to do basically its day job. What will you report next for WVE-006?
Yeah, the next.
Q1, I think.
Yeah, so the next step for WVE-006 is again 200 mg, to your point, doing its job, responding to those, and again, highly important we talk about these acute phase events because in the absence we don't use lipid nanoparticles. By using GalNAc, there's no risk of the irritation being driven by the nanoparticle. It was nice to see that editing happens very, very quickly. I mean, two weeks after that first dose, that patient could mount a 20 µM response to an acute phase event and it was stable and durable past that. To your point, the 400 mg gives us a real opportunity again to continue driving editing efficiency. Where and how much more M protein do we make? What happens? We saw a 60% decrease in Z protein. How does that impact the decrease and continuing decrease in Z protein, which protects the liver?
Ultimately, it lets us test durability. Do we see this now going from bimonthly to even less frequently? We still have one more cohort above that. Highly encouraged given that again this is the first lowest dose and you know we could see activity at the single dose portion too of that 200 mg low dose, but 400 mg gives us another opportunity to look at more efficiency.
Paul, since you are furthest along, give us your view on where you think the bar is for an FDA approval.
Yeah, and I think if we think about the approval thresholds, everybody has in their head, you know, 11 µM. I think the field got a lot of consternation going with the inhibitor discussions of does it need to be more? I think we have to look at inhibitor being a little bit different in that it's a dimerized protein. This idea was the real debate wasn't of how much protein, it was as you created this larger protein, is it going to behave the same way? It didn't necessarily in our minds change the threshold for the protein as it's endogenously made. I think the key is really thinking about phenotypic registration. The whole premise on that 11 µM was all grounded on the human clinical genetics of an MZ phenotype.
I think the benefit we all have in the RNA editing space, and frankly editing space more broadly for AATD, is really one for thinking about what is the phenotype of an MZ patient where really they have better clinical outcomes, whether that's focusing on M protein, the totality of reduction Z protein, which potentially allows you to build liver into labels. I think this acute phase events, being able to demonstrate, because it always comes up where they say, great, you get this MZ-like phenotype, it's all in preparation for what? It's all in preparation for an acute phase event. You can see people asking the question of, how do we know how your drug's going to respond in the setting of an acute phase event?
I think the fact that we've demonstrated that we get high levels of protein, high levels of M protein in response to an acute phase event, durability of response during that event, and really protect, we think we'll continue to support that. You know, we plan to engage the agency along with our collaborators at GSK.
All right, great. Sriram, let's switch over to AIRNA. Your company itself is relatively young, the people involved with it have been in this space for a while. Maybe highlight who they are and what they've done.
Yeah, I think the field goes back to 2012 when Thorsten for the first time showed that you could generate a DMNA fused to an RNA and edit RNA in an effective manner in a test tube. Of course, now we have clinical data saying that this could be actually proved in the clinic and it's working. It's interesting that it's the exact same year where CRISPR showed that it can be evaluated in preclinical models. We've advanced much more rapidly in terms of CRISPR generating modalities moving into the clinical proof of concept on approval with CASGEVY ex vivo editing approaches. I think the company is really founded based on the foundational work from Thorsten. As you've seen more recently, it's now shown that you could use endogenous enzyme and an oligo to do it without requiring the strida purity of the oligo.
We have also presented some of the data in most recent conferences early this year preclinically. We are excited to move this into the clinic with the CTA filing expected end of this year.
Yeah, you showed some data at ASGCT for your lead, again targeting AATD. Maybe share some of the key findings there.
Yeah, I think three key findings are consistent with the field is really understanding what kind of chemistry backbone modifications and what kind of optimization of engaging all of the different isoforms of ADAR, especially ADAR exists in this p110 as well as p150, which is what we talked a lot about the response to the infection. We can see that you could even in a dish add interferon alpha and induce the production of this cytosolic form of ADAR p150. We showed that with our oligo, we can engage both of these isoforms effectively. This provides, I think, from a clinical patient standpoint, that ability to respond to that infection in a mechanistic manner.
The data that we showed at ASGCT not only highlighted the platform in its ability to engage both of these isoforms of ADAR, but also demonstrated in vivo that you can reach much higher levels than what would be theoretically meaningful levels in terms of MAT reaching up to 39 µM, total AAT reaching up to about 63 µM in those mouse models. NSP data demonstrated that that durability could be expected with a half-life extending more than 30 days in monkeys, demonstrating that the molecules will potentially substantiate less frequent dosing in the clinic. We want to support dosing frequencies of more than a month.
As you finalized your drug candidate, you're preparing to enter the clinic. What do you think, number one, you've learned over the last, say, 12- 24 months in that process? How do you think that plays into your company's ability to differentiate itself based on your platform and design approach? Not just in this initial indication where there's multiple drugs entering it, but in the broader landscape of ADAR.
Yeah, I think in terms of the design space, as we sat around looking for molecule optimization, there are probably a billion different molecules you could design given just that one edit that you need to make, given the different options of the backbone, the sugar, the base modifications that you could make. One of the seminal works from Jin Billy's lab was to show that actually direct base sparing of the oligo is not optimal for maximizing editing efficiency. There have to be some more mismatches that are required for optimizing the editing efficiency. We really incorporated all of those principles, which could be, you know, translated from target A to target B to target C. We have now progressed on our pipeline as well to show in vivo proof of concept, which hopefully we'll be able to release sometime next year.
Okay, great. Dennis, let's move over to ProQR. One of the things we like most about the potential for ADAR is the ability to affect protein-protein interactions and to translate a protein variant and the applications for that. We're really happy to see that that's where you guys are initially starting. Talk about your lead, AX-0810. I know you just had some news about it, but talk about that program and your targeting of cholestatic diseases.
Sure. Thanks, Keay, for inviting myself. I'm excited to be here. I guess, like the rest of the panel, really excited about the broad potential of RNA editing. As you pointed out, our lead program does indeed introduce a variant to modulate protein function, not just correct protein mutations like what they're doing. AX-0810 is designed to edit NTCP and specifically block bile acid transport into the liver for cholestatic disease. As you mentioned, just yesterday, we announced clearance to initiate our phase one trial in healthy volunteers. It's a big step forward for ProQR, but I think also for the field because it's the second target to enter human studies and help validate ADAR editing as a field.
In terms of mechanism, I mentioned that the edit is designed to block bile acid transport, but just as important, the edit is designed to allow the normal transport of other molecules. This is where the modulation of protein function comes in. Interestingly, this is also a potential for AX-0810 to be disease modifying because it is designed to reduce the toxic accumulation of bile acids, which is a common pathology amongst cholestatic disease. Before I go into some of the preclinical work that we've done, there's quite a bit of genetics and pharmaceutical targeting around NTCP. It really helps validate this approach for us.
In terms of genetics, there are naturally occurring variants of NTCP where individuals show an impaired transport of bile acid function and naturally have quite high, like you can see, up to 40 x elevated levels of serum bile acids with no negative consequence, which is very interesting. In particular, no pruritus. Additionally, there is a peptide inhibitor of NTCP that's been approved for hepatitis. In that phase three trial, those patients showed elevated levels of bile acids, showing that indeed they were able to block NTCP transport of bile acids into the liver. That resulted in improvements in liver health, even without virologic response. It's important to consider that without the removal of the virus, you're still seeing liver health improvements with an NTCP blocker.
Preclinically, we ourselves have demonstrated that about a 15% level of editing in both mice and in non-human primates has translated into roughly a twofold increase in serum bile acids, which we know from the literature and from the clinical studies done in this other blocker can translate into functional effect. Just to wrap up, from a safety perspective, our GLP tox studies have not shown anything concerning. It's essentially a profile consistent with other GalNAc conjugated oligos.
Okay, not to front run your event coming up to talk about this, but you know, there previously was some guidance about when you would show first data. Maybe update or tell us about that.
Yeah, great question. As you mentioned, I can't front run too much, but we have an event scheduled for November 3rd, which will walk the street through the specific trial design, biomarkers that we expect to read out, that sort of thing. At a high level, I can tell you, as all phase one studies, the primary endpoint is safety and tolerability. This particular mechanism lends itself to measure of PD markers of target engagement in a healthy population, which will be treating healthy volunteers. In that case, we'll be able to look at whether blocking or editing this target will indeed block bile acid transport into the liver by looking at increases in serum bile acid, which I talked about already. We will be looking at potentially a threshold of 2x the level of serum bile acids and increase. That should tell us, do we have target engagement?
Great. Another area of focus for your pipeline is the CNS. You guys have presented preclinical data on some of your Axiomer candidates, which show good editing efficiency, good distribution. Why more specifically do you guys believe the CNS is an attractive place for ADAR?
That's a great question. As you know, RNA editing is not limited to the liver naturally, right? The brain has robust endogenous ADAR expressions, so in some ways, it makes it an ideal target organ. RNA editing can target transcripts precisely and reversibly. I think everyone knows that, and therefore, it's somewhat unique amongst genetic medicine approaches. In terms of what work we've done, as you mentioned, we've shown that our editing oligonucleotides are distributed broadly across the CNS, both cortical and subcortical regions, and continue to redistribute over time. That speaks to durability. From an efficiency standpoint, we've actually seen up to 60% editing in vivo. These data come from, I would note, intrathecal delivery of a tool oligonucleotide that we use into non-human primates. We've gained a lot of this knowledge through our partnership with Lilly, which is focused primarily on CNS, but also liver diseases.
It covers 10 targets in both, as I mentioned, liver and CNS. In addition to that, we have a collaboration with the Rett Syndrome Research Trust. Together, these two collaborations have sort of enabled us to do quite a bit of work in CNS. We haven't publicly been able to talk too much about it because those targets are still confidential to Lilly.
I think I know for Rett's, your lead candidate there, again, I think interesting in that it's trying to translate a wild type-like protein. Again, a very interesting application for ADAR. Paul, let me switch to your pipeline again. I don't want to front run your R&D day, but you've also talked about PNPLA3 as an attractive target. Tell us why.
Yeah, I mean, we could talk about PNPLA3 in terms of size and correction. I mean, I think the benefit of several research days, so we don't have to front run the ones we've done, really do demonstrate the breadth of editing. It's interesting how we say, taking clinical data now and applying what do we learn in the clinic and how does that de-risk what we're doing next is really the fundamental feature, I think, of thinking about translation. I think the real opportunity, I've fixated on that picture where I look like I had a lot less gray hair back then. The translation on being able to take preclinical data and translate it is really important and fundamental.
I think we learned a lot if we take AATD to a mouse that has 12 copy numbers that, you know, why we wanted to wait for the mouse to actually get well below 11 to treat it taught us a lot that actually shows us modeling really well, the mouse model to the human model. I think sometimes we have to be really careful because the data we use for a lot of these extrapolations are always preclinical systems that we're trying to predict clinical benefit from. The more time we have to take learnings from our preclinical data, generate human clinical data, and bring that human clinical data back creates the efficiencies that ultimately release a flywheel. I think our now translation from clinic from target to clinic, I mean, for our obesity candidate in siRNA is 18 months from generating mouse data to human clinical trials.
To your point on PNPLA3, I think stepping back, we've generated editing data in lung as it related to CFTR and fixing mutations with substantial editing there without conjugates. We've demonstrated a potent distribution in CNS around editing and building out a portfolio. To not front end research day, I think one of the highlights for October 29 is really going to be what did we learn on alpha-1 antitrypsin? How is that de-risked the ability to think about our chemistry, which is not just stereochemistry? I know that's back in that time. We have stereochemistry is important to rationally design single drugs. I think the real evolution wave back in 2020 to 2021 is the advancement of phosphoroguanidine that's independent of stereo control.
We control it, so we make a single drug, but really the intellectual property around how do you design medicines that can stabilize oligonucleotides, give them the ability to distribute, have good retention and potency, and ultimately the N3-uridine, which is our modification for how do you get site-specific base editing. If we translate that to this research day, it's going to be a continuation of the clinical data. How do we build beyond alpha-1 antitrypsin with PNPLA3? How do we look at extrahepatic and the growth of the pipeline there? As it relates to PNPLA3, this is really an exciting opportunity to think about correction, not in the environment where it's about what I call the dynamics of an acute phase response and trying to see how do you fix something so that it can respond to these acute events.
Here's how do you fix an enzyme that actually is responsible for a good portion, about a nine-fold increase in hepatic diseases, liver diseases across all causes, alcoholic cirrhosis, MASH. It's really a unique genetic population that happens to be substantial in size. It's about 9 million patients who are homozygous for this particular mutation. If you can fix that mutation, you decrease their risk of liver disease and the ability to demonstrate how we can drive that forward in a way that's going to be important. We'll reiterate this at research days. This isn't front running. Look at this as kind of sensitizing. Is this notion that, you know, siRNA, so there are these targets where, and this is really, I think, the advantage we have is we think about an armamentarium of different modalities and how do you pick the right modality for the right indication.
This is one where if you knock it out, actually you could reabsorb fat in the liver. You actually need this protein to be physiologically functional to protect the liver. How do you do that? You edit and correct it. By correcting the protein, you actually restore its physiologic function back and therefore can think about treating patients that are genetically stratified based on their status and ultimately be able to think about treating them for a variety of liver diseases, MASH being one of the ones that's interesting in a way that's distinct from how we think about medicines that treat obesity and actually decrease fat to treat this. This is actually an underlying genetic driver for these patients.
We're excited about the opportunity on research data to highlight, obviously, PNPLA3 beyond alpha-1 antitrypsin and really go on the journey of how do we continue to de-risk the modalities both in the liver and very importantly outside the liver for both editing and siRNA.
Sriram, let's talk about how you think about your pipeline. Obviously, again, walk before you run. AATD, point mutation, validate a target, large enough population, you can enroll the trials in a reasonable amount of time, and there's biomarkers. What does the next thing look like for you that kind of meets those same kind of criteria?
Yeah, I think really AATD has really allowed us to build a platform as we, you know, play with different chemistries and different delivery optimization to really, we anchored ourselves on GalNAc because we wanted this to be, you know, AIRNA has taught us through multiple clinical development phases that that is actually the better approach. It's really good to see that, you know, three different companies are really focusing on GalNAc to deliver to the liver. Since we are still a private company, we've really remained focused on unlocking the liver. If you think about the white space in liver, you know, people say, hey, for a knockdown approach, liver is saturated. For an editing approach, I don't think so. We've identified more than a dozen different targets that are either a gain of function mutation that could be introduced into the population or a specific protein-protein interaction.
One example that we have announced outside non-confidentially is an E3 ubiquitin ligase binding pocket of an important protein where you can just block the E3 ubiquitin ligase binding itself and not affect the unintended consequence of knocking down the E3 ubiquitin ligase. Those types of precision edits are now possible. For the investor dollar concerns, really going to remain focused on the white space in the liver.
Great. Let me take advantage of the collective brainpower up here and have each of you ask somebody else a question. Dennis, why don't you start?
I guess I would be curious to see what the panel thought about how long it will take for us to get to the stage of siRNAs where it's now seen broadly as applicable across targets and not just narrow mutation correction, for example.
It's an interesting way of stratifying because I think you could say, and I like it of like, you know, how do we think about siRNA? Because we're working on si, so we'll have data next quarter on our...
You have a unique perspective.
I think a lot, we have to spend a lot of time, and I'm looking at our team thinking about balancing, you know, the ADAR portfolio and editing and correction, and where is that tool useful? Then where is the tool useful as it relates to si? I think to your point, the similarities on those are once you understand, because I think when we're really saying how do you get to the stage that si is in, it means that you understand how the enzyme behaves between a mouse and non-human primate and a human. Once you can...
Yeah, that's right.
Yeah, but once you understand the pharmacology of how that enzyme is going to translate, right? Like we know if we see X amount of time in silencing of a protein in a mouse for siRNA in Ago2, that it should reasonably translate to X amount of time in an NHP. We're seeing actually really reproducible translation of what's supposed to be happening with Ago2. I think the same is true now for ADAR. What does that catalytic efficiency of that enzyme look like? I think what we're going to realize, and we see this, and I think we need to be careful also creating generalizations.
Case in point for activin A in obesity is the program we have now where it's been intractable by those who are really viewed as, I think, would be given, I think, being viewed as experts in the field in terms of knockdown of the target protein and duration. We're the only ones who've demonstrated potent durable silencing of activin A, which is this biomarker, this protein that's produced for obesity in the liver, and shown in a mouse that that reduction translates to fat loss, so no loss of muscle, but just fat loss in a mouse and ultimately generating that data in a human. Others who've targeted Ago2 haven't been able to see that level of reduction of the protein.
I think the caveat in thinking about ADAR is realizing that there's a, when we talk about the generalization, we got to think about it in the terms of the protein. Then how do you and each company, you know, I think that's the benefit of each company, how do we each learn systematically about how our chemistry interacts with a given protein and how that interaction with a given protein translates to patients? Once we can establish that longitudinally, I think then we unlock what we're kind of saying is the power of si, which is how do you quickly go from a target to human clinical trials in a reproducibly fast way? That's about understanding the platform capability as well as then the biology and target biology. I think, I mean, we'll talk about this at our R&D day soon. I think we're getting pretty close.
I mean, I think our understanding of the catalytic efficiency of the enzyme, both between mouse to non-human primate and now with human data, how well that's happening. I always say it takes two to feel like it's not one target translating in a unique way, but how does that happen across multiple programs?
Exactly.
I think we're right at that precipice.
Sriram, let me modify that question a little bit and then have you answer it. It goes to what Paul said, you know, how do we engage this endogenous enzyme? For an siRNA, when you're going after a validated target, we assume you're going to knock it down sufficiently. Because it's a validated target, knocking it down sufficiently, we kind of know what the result should be. It's just a matter of, okay, whatever your tissue target is, can you get enough of it there? We're not so much worried about safety anymore either. What do we need to know better as a field, or you guys need to know better as a field, to get to that kind of same point where an siRNA is?
Yeah, I think it really goes back to picking the target, right? We don't want to make an edit to, you know, kill an enzyme activity where you could just take an siRNA and knock down the protein if that is viable, right? You have to really identify that white space. Second, understanding the haploid insufficiency of the transcript, understanding that maybe 50% edit is sufficient. That's where the human genetics for alpha-1 is very, very strong, right? We understand the phenotype of the SZ. We understand the phenotype of the MZ. We know the problem associated with the ZZ. We know exactly what the median levels of the circulating biomarkers are.
We can establish a relationship between the drug's potency in the mouse model, the PK duration in NHPs, and the now emerging human clinical data to start to draw that association between these molecules, how they are behaving in mouse to monkey to human. That translation really will help us accelerate this process. The front end needs to be very clear. There has to be a target where, you know, the level of editing of that target needs to be clearly associated with what is the level of therapeutic benefit that's needed. That should be a circulating biomarker like what we're talking about. It's bile acids or it is wild type AAT levels. That should be associated with clear clinical benefit. Therefore, you know, we don't need to wait for a randomized placebo-controlled study to understand whether this is actually trackable or not.
I think sometimes the models, and it's an interesting point, we have to be highly selective about how we use the models to translate. I think our learning out of the gate is, if I had to look back, the SERPINA1 model itself, as we think about the extrapolation, probably created more consternation about thinking about that initial translation. The benefit in hindsight of clinical data and how do we all learn from clinical is like this notion of, you have 12 copy numbers in a mouse, and depending on when you treat and how you treat and the response rate of the mouse, you could actually, in some ways, think you're going to see more and really not getting people indexed on what are you trying to do in terms of this concept of editing efficiency. How efficiently are you converting it from Z to M?
A little bit of that model got everybody kind of, I mean, I think about conversations we were having with you and your peers coming into alpha-1 antitrypsin data where the field's going like, we should be seeing 20 µM or 30 µM protein. I remember and remind everybody, anytime you go into any therapy, it's really about thinking about what's biologically relevant. What do you actually need to actually treat the human disease? What's the drivers for that? I think there's ample opportunities to do that. A lot of this translation, that's why I get to some of it, is what do you learn about the enzyme and how the enzyme behaves and how does that translate? The other side is how do we think about targets that are moving forward that best underline and validate the capability?
Ultimately, how quickly do you learn from clinical data to imply back to preclinical data?
Just one addition to that, I think all of the work that Jin Billy has shown so far is that humans have evolved to have an active ADAR function. We all benefit from that natural ability because unlike mice, humans and primates have a lot of endogenous retroviral transcripts that are expressed at high levels. When you have mutations in those ADAR enzymes, you see that there is an autoimmune consequence. That is, the inability of those ADAR to edit and repair these double-stranded RNAs that are naturally produced because of the pseudo-retroviral elements that are present in our body is an advantage. We are seeing that translation from mouse to human happen because of this natural biology of this enzyme. Also, the biggest advantage is the precision because we are using the endogenous enzyme.
The adenosine, the target adenosine, has to flip out of the double helix for it to be deaminated. That provides a very precise chemistry that we can apply around that edit site to make this exquisitely precise compared to a bacterial protein engineered to have that precision, which is a big advantage for us to think about other targets where we can introduce temporarily, transiently a gain of function mutation and withdraw the drug. The patient goes back to a natural state without having any of those deleterious consequences of editing of the DNA.
Yeah, I think simply, I mean, and I completely agree. I mean, I think ADAR, I have every confidence now that ADAR in humans translates as a therapeutic enzyme. I think unequivocally, like it's endogenous, it's catalytic, it's functioning as predicted. I do think to that, you know, to bring it back to your question, I actually, you know, I think we're there in the context of what we would think about Ago2. Nobody, Keay, to your point, having this debate of is Ago2 a therapeutic enzyme in a cell that's going to do what's predicted? I have a high degree of confidence that ADAR as an enzyme will be able to do what it needs to do. To that point, it comes down to how you harness it, how you think about targets, and how you advance it.
The enzyme, I think compared to like two years ago where there was more of at least where I was sitting theoretical saying this should work, like this is all of the premises of what a therapeutic enzyme should do. I think two years later it can say the enzyme is doing exactly what it should do. It's remarkable when you have human data and you step back and you go, the enzyme did exactly what it's supposed to do, which is pretty incredible.
For the investment community, when you think about the timelines of getting ADAR to where siRNA was, I think it's going to happen much faster, right? It's not going to take what was it?
15 years.
Yeah, 14, 15 years from the time, yeah, the time the first siRNA was approved. It shouldn't take us 15 years, right?
Yeah, that's largely because the chemical toolbox is quite well validated now clinically. We understand all of the modifications that people are using, and we know how to not trigger the immune response by doing that.
Paul, you want to squeeze in a question?
Sure. I think it's not dissimilar, but we've got a lot of colleagues that are out there and Keay, you can participate too. It's really, frankly, a question as much for the audience as I do think that all of us spend a lot of time with this community. I think as things evolve at various stages, there are some misconceptions about the enzyme. You're pointing it out on the time horizons and the translatability. I'm curious, as you hear, where do you think or hearing where you think the misconceptions still exist for RNA editing and ADAR?
When I have investor discussions, they center not around the ADAR editing anymore.
Okay.
They center around the indication. Have you picked the right indication to demonstrate proof of concept here for the technology?
That is specific to you folks because you're not going after initially a point mutation.
Right, right. I'm assuming these guys only get questions about alpha-1, not about how does this work.
Yeah, I think for us, the biggest questions that we are being asked is, is there an upper limit to how much editing can you see with RNA editing compared to DNA editing? I think the conception is DNA editing is reaching 90% in humans, which is not, I think, mathematically true from the models that we see because if they are getting 90% editing, then the levels of AAT should be 27 µM or 30 µM and not 12 µM. I think that comparison is inevitable given the amount of clinical studies that are happening in parallel in alpha-1. I think really convincing the investment community that the RNA editing is here to provide therapeutically meaningful benefits to these patients and it has a broad breadth to go beyond multiple targets is to me the next key step.
I think what's interesting is that I think investors sometimes overemphasize where you're at today. You've gotten there very quickly and, to the point you made, Dennis, this stuff usually doesn't happen that fast. Maybe take a step back and let them flesh it all out before we set the bar too high. All that said, Paul, to all of your points, theoretically, a lot of this stuff should work. We're building on the shoulders of other RNA medicines, but you do have to prove it.
I think that piece, and you brought up in a great way, I mean, I think the field now, which is great for an indication, means we're talking about targets as opposed to an enzyme. One is a lot more important. It's what we hopefully all do, which is make medicine. I think the ability to shift the framework from enzymatic biology to actually therapeutic indications and how do you unlock them is, frankly, I think a lot more exciting for all of us and a lot more beneficial to patients. It's a good place to be.
I would comment just like how siRNA evolved, the first generation and the second generation, our technology, RNA editing technologies evolved quite rapidly and we're learning quite a lot in a short period of time. It could be that the first drugs that go into the clinic won't be the best. There'll be subsequent ones. I think I should be stronger than maybe, right?
For sure.
Yeah.
Yeah.
All righty. We've reached the end of this session. Thank you all for the great discussion.
Thank you.