Circio Holding ASA (CRNA) Earnings Call Transcript & Summary

Welcome to Circio and our R&D webinar. My name is Erik Digman Wiklund. I am the CEO at Circio, and we have an exciting program for you today. We will start with a short introduction by me to the company as well as circular RNA in general. Following my introduction, there will be a talk by Alex Wesselhoeft. Alex is a pioneer in the field of circular RNA and did many of the important early studies on this in the space, showing you can express protein from circular RNA in mouse models. And also is one of the founders and the scientific founder in particular Orna Therapeutics, one of the important companies in the circular RNA field. After Alex's talk, our own Thomas Hansen, will give a presentation on the technical development of our circVec platform; followed by CSO, Victor Levitsky, who will talk about where we're planning to deploy this technology in our R&D strategy. And at the end, we will host a Q&A session with all the presenters. [Operator Instructions] So with that, I move into the first part of the talk. Circular RNA is emerging as a novel format of RNA that has the potential and is expected to disrupt both vaccines and gene therapy going forward. And why is that? Well, the core reason is that by being circular, circular RNA is more durable than linear mRNA. Linear mRNA gets degraded by exonucleases. They cut up the RNA from the side. You can see it on this little cartoon here, by the Pac-Mans that we've drawn. They chop up from the end the mRNA. circRNA doesn't have a free end, so it's automatically resistant to these enzymes that degrade the RNA. This means circular RNA is more stable. So the circular RNA will have more durability. You can also engineer the circular RNA to express more protein. And you achieve then two goals, you get more durable expression and you get higher expression. And this leads to a massive therapeutic advantage, and this is why everyone is excited about circular RNA. In addition, by virtue of being more stable, you can build in other functionalities, and Thomas is going to give an example later today of how we deploy this and sort of have multifunctional circRNAs that can do more than one thing in the same product. Now it's not just us at Circio who are interested in talking about the advantages of circular RNA. In the last couple of years, there have been several companies forming. Especially in the U.S., you've seen some very substantial financing and partnering deals. Companies like Merck and Pfizer have moved into this space. One of the first companies to launch was Orna Therapeutics based in Cambridge outside of Boston in the U.S. And that's based on the work of Alex, who you will be hearing from soon. So today, you'll actually be fortunate enough to hear two of the pioneers in the circRNA space, both from Thomas Hansen, our VP of Research. He's one of the early people who worked on the biogenesis and functionality of circular RNA. And then Alex, he was able to take those discoveries and translate it into protein expressing circular RNA. As I mentioned, Thomas here is one of the pioneering scientists that published also some of the early work. You can see, particularly this Nature paper from 2013, which elucidated the first function of a known -- of a microRNA and this now has more than 7,000 citations, I believe, which is very substantial. So at Circio, we've developed a platform technology that we call circVec. And this is differentiated from all these other companies. So the other companies I've showed you more or less do something very similar. They make synthetic circular RNA or they make a circular RNA in the factory, they package it and the product is this synthetic circRNA. What we are doing in Circio is make circular RNA from DNA-based vectors, either from a DNA or it can be from a virus. And that DNA or virus carries the instructions in the form of DNA for the cell to make the circular RNA itself. So we provide the instruction for circular RNA generation inside of the body or inside of the body cells. And this is an approach that, so far, it's really only us that are taking, and I think we are the leaders in this space. Now you're going to hear a fair bit of technical material later and also it may be confusing to some of the more lay audience here, what is the difference between DNA and RNA, et cetera. So I'm just going to spend a couple of slides here to tee up the discussion and explain the background. What is called the central dogma of molecular biology is that DNA makes RNA and RNA makes protein. To take a step back again, DNA, this is the genetic code, sort of the recipe. It sits inside of the nucleus of the cell, and it carries all the instructions needed to make a human or an animal. It's the complete recipe. Now that recipe is instructions for how to make proteins. The way this works is that the DNA then gets what we call transcribed into RNA or messenger RNA. And this messenger RNA is then what actually gets translated into protein. So there's -- it's an intermediary molecule that sends the message from the DNA to actually make the protein. So when we talk about mRNA vaccines or synthetic circular RNA, this is -- these are products made in the RNA format. We're going one step earlier. We're providing the DNA and then the cell will, from that DNA, make the RNA that then makes the protein. So this is the big difference between what we are doing and what other companies are doing and sort of how this biologically functions. Now you can, with technology that's been developed and due to some of the pioneering work of Alex and others, you can now make circular RNAs that are mRNAs. So you can make circular mRNA, if you want to call it, if you want to call it that. Naturally, circular RNA do not act as mRNA or express protein. Naturally, circular RNA has structural and regulatory roles. But now we can engineer circular mRNAs. And on this cartoon, I highlight some of the difference between the two. So a linear mRNA, if you look at the cartoon on the left, you can see there is what we call a 5-prime cap. This little ball, the 5-prime cap is where the translation initiates and starts. So we call the translation of protein production starting point cap-dependent. mRNA is linear. It has the cap. It has this poly-A tail at the end. This gives some protection, but really mRNA has a short half-life. It gets broken down in hours or maybe it can last for a couple of days at the maximum. The reason why mRNA is unstable is that the cell wants to be able to up and down regulate genes. If the mRNA was too durable, the cell would have problem switching a gene off if it wanted to. So therefore, it's actually important biologically that mRNA has this sort of short half-life. That means you can switch it on and switch it off. In terms of the therapeutic, this is a major disadvantage. And that's why we want to switch to circular RNA. Now a circular RNA, because it doesn't have an end, it can't have a cap. So normally, you have this 5-prime cap to start the protein production. With the circRNA, you need to start it from something else. And then we use, what you can see on the cartoon here, what's called an IRES element or an internal ribosome entry site. This is a type of sequence that's actually derived from viruses that is highly efficient at starting protein production or protein expression. So this is what we need to put into a circular RNA in order to make protein. And a lot of the optimization and development that goes on is around selecting a powerful and good IRES element. And you will hear both Alex and Thomas later talk about IRES elements, so I just wanted to make it clear to everyone what IRES elements are and what the 5-prime cap is on an mRNA and the way the protein production start is different. The advantage of circRNA is that, one, the IRES element can be more powerful. They can make often more protein if you have a good IRES element than an mRNA can. And then, of course, the extended half-life. It can last days or up to weeks as opposed to hours for the mRNA. So this is why circular RNA can both last longer and make more protein. So with that, I wrap up the introductory part, and I'm pleased to hand over to Dr. Alex Wesselhoeft. As I mentioned before, Alex was the scientific founder of Orna Therapeutics. Alex has recently moved back to academia and is now setting up an RNA research lab at Mass Brigham in Cambridge in Boston. So thank you for joining us, Alex, and I give the word to you.

All right. Can you see these slides?

Yes. Thank you. It works from our side.

Fantastic. All right. Yes. Thank you for the kind introduction. And I'm just going to talk a little bit about circular RNA at a high level today, distinguishing between RNA format approaches and DNA format approaches, why you would want to circularize RNA, why you would -- the main challenges that continue to face the circRNA field, and its potential as a technology format for expressing proteins of interest in tissues of interest. So I'm Alex Wesselhoeft. As mentioned, I'm one of the scientific founders of Orna Therapeutics. I led the molecular biology group there for about 4 years and recently moved back to academia, where I'm starting a research lab focused on solving the next-generation problems of circular RNA and lipid nanoparticle delivery vehicles. All right. So I'll just jump right in here. So why circularize RNA? I think this was fairly well detailed. Essentially, it makes it more stable is the answer to that. But there is more nuance to that answer. With linear RNAs, of course, you have multiple degradation pathways, including decapping at the 5-prime end of mRNAs and then exonuclease mediated decay at the 5-prime and 3-prime ends, in addition to endonuclease mediated decay. And all of this together in mammalian cells results in a fairly short half-life of even modified state-of-the-art mRNAs. And I have a picture of the human exosome complex, which is this very complicated protein structure involved in the regular turnover of RNA within cells. So there's these -- basically, there's these two main RNA turnover pathways, exonucleases and endonucleases. And the thinking here is that if we can get rid of one of those pathways, then we can improve the stability of RNA-based therapeutics, which has been a consistent limiting factor in their use and application. When your RNA only lasts 6 to 12 hours in a cell, it can be hard to reach the therapeutic index that you need to see efficacy from a drug. So the goal is to circularize the RNA to prevent rapid decay. And then you get more protein per RNA, which means you can dose less material or you can dose the same amount of material, they get more protein out of it. And that ultimately would be predicted to lead to a better therapeutic index. And this fundamental concept and these advantages are applicable to both DNA format circRNA and RNA format circRNA. I spent most of my time -- pretty much all of my time working on the panel on the right, which is the RNA format circRNA, developing new tools to circularize RNA efficiently, figuring out how to increase protein translation from them, applying it in different systems. But the principles of circularization, enhanced stability and better therapeutic index would equally apply to DNA format circRNA, one of the focuses of Circio. And if you can generate circRNA efficiently from a DNA vector of IRES or plasmid, for example, then each molecule that is transcribed from that DNA vector is more potent. And therefore, you would need less material as part of your dosing strategy, you have greater potency and, of course, as many of you probably know, with the safety concerns of viral vectors, having greater potency per transfected DNA sequence may ultimately present as a safety benefit. So circRNA fields is not necessarily new, although it has recently in the last 5 years or so, become quite popularized. In the early days, this was in 1991, there were natural circRNAs that were identified, that resulted from human genes and other eukaryotic genes through a back slicing process, which was typically seen as an error of transcription. There are also circRNA genomes that certain viruses like the hepatitis delta virus have, plant viroids like to have circRNA genomes for some reason. But when it comes to eukaryotes, it's this back splicing process that tends to generate these circRNAs. And then, Thomas Hansen here, really opened up, I would say, a lot of the circRNA fields by showing that one of these circRNAs actually has an appreciable function. So this was work showing -- when I asked, I believe, has a microRNA sponge, that was basically a circRNA that had evolved to manipulate the microRNA environment of the cell, which was a very exciting step forward for a few in 2013. And then more recently, I and others have been working on synthetic circRNAs in RNA format and trying to figure out how to stabilize mRNA, which at the time was a promising but not validated technology for therapeutics and vaccines in order to increase their potency and increase their -- or widen their applications. So since then, a lot of things have happened in the circRNA space. This is a semi comprehensive but not a fully comprehensive map of what's going on in the circRNA space today, and some of the lineage of the work that has led to the existing entities. And in purple, these are basically -- these are publications or concepts. And then in green, these are companies. And so a lot of times, you can see actually a lot of these concepts are being translated directly to industry, which is a pretty interesting observation, the rate at which these ideas are being brought into an industrial setting. On the top, you have your DNA format technologies. On the bottom, you have your RNA format technologies. And there are multiple companies that are active in both of these spaces and are continuing to develop both RNA and DNA format circRNA technologies. A huge amount of investment has gotten to these companies. I think we can recall the slide with all those hundreds of millions of dollars of investment. I think in total, it's probably about a couple of billion of investment has gone into DNA and RNA format circularization technologies, yet we're still really at the beginning of the path for developing circRNA technology, and it took mRNA -- well, if you kind of go back in the early '90s, it took mRNA like 20-odd years to get their first approved drug in the pandemic. And so I think we're near 5 of circRNA and so it's going to be pretty exciting to see where these companies go with this new format. So these are 2 different platform technologies. You have the DNA format and RNA format, but they share a lot of the similar challenges. Of course, how you make it is fundamentally different. DNA format, you're using viral vectors, maybe you're using plasmids, but generally delivering via virus is more common, I would say. And on the RNA format side, you have to think more about your purification process of the RNA itself and the circularization process of the RNA in a manufacturing setting instead of in a cellular setting. But everything else is basically shared. You have your IRES, you have your circularization elements that are sequenced elements that you can build in to the vectors themselves, the precursors, codon optimization, off-target functions and figuring out some form of nonviral delivery can apply to both of these areas. So I've spent most of my time in the RNA format side. And the rest of the presentation is really going to focus on some of the work that I've led in the past on -- around the IRES and the circularization elements as these common features of these two platforms. So first and most obvious challenge is how do you circularize something because you can't just make a circular RNA. You have to make a linear RNA first. And that's true whether you are transcribing up a DNA template in a cell or whether you're transcribing off of the DNA template in a test tube. First, you generate a precursor RNA and then either through enzymatic ligation or ribozymatic as processes, you will then circularize that molecule. And the earliest, probably most efficient -- or earliest, let's say, semi efficient circularization method was really pioneered in the '90s by Puttaraju and Been where they permuted a group I intron, which is a type of intron that doesn't require any proteins, to assist in splicing. So they permuted it, meaning they cut it in half, took the second half, put it in front of the first half and then inserted whatever they wanted in the middle of the two halves. And they found that using this approach, you could circularize relatively small things, in this case, you have 124 nucleotides circular RNA on the left, 338 nucleotides circular RNA on the right. It's not terribly efficient, maybe 50% circularization efficiency for these very small RNAs, and you can't really encode a protein in that. So limited utility. And I think that's sort of what held back the development of this technology originally was the inability to really efficiently circularize large things that encode proteins. But this was a major step forward for the synthetic circRNA field and was the foundation for the work that I got into when I was at MIT, which -- this is essentially a synopsis. You can take that permuted group I intron strategy. You can engineer in a bunch of accessory elements, such as duplex regions and spacers. You can pick your intron carefully and at the end of the day, you can get very efficient circularization, shown in this figure on the right here, where we are moving from -- in the previous slide here, we're moving from 50% efficiency of a very short RNA to about 95% circularization efficiency of a protein-coding circRNA. So this is about 1.5 kilobases. So this was very exciting. And once we could make these circRNAs, we could start to study them, study their transition properties and so on and so forth. So since then, additional work has happened, and this is work that Orna Therapeutics has shared, and they've shown that you can generate very big circRNAs. So this is an example of a full-length dystrophin, circRNA. So it's about 12 kilobases. It's actually a little bit shorter than 12 kilobases. But the chromatogram, you get peak merging at this point because these are really big RNAs, but you can sort of see the two peaks here and peak on the right is the circRNA peak and peak on the left is the nicked/precursor uncircularized RNA peak. And so this was actually fairly efficient, maybe about 50% circularization efficiency of this very big, probably biggest RNA you would want to realistically encode or realistically circularize. These could form really nice nanoparticles. And indeed, when you put them in a soft retranslation assay, you do get full-length dystrophin protein being generated from circRNA. So this is probably the biggest thing anyone has ever generated protein from for circRNA. And then, of course, if you transfect them into primary human myotubes, you also get translation in a cellular system of these very large circRNAs. There are -- these are smaller variants of micro dystrophin, Becker dystrophin variants. When I say small, I'm talking about like 7 kb instead of 12 kb. So they're still pretty big. And then this one is the full-length dystrophin on the right, yes. So the second challenge is figuring out how to get protein out of circRNA. And so of course, if you have a circular structure, you don't have a cap. And so you need to figure out a creative way to get protein -- the [ protein categories ] and retranslate without the typical top structure cap structure of the known mRNAs. So the sort of most obvious approach is to take in IRES from picorna viruses because the picorna viruses, many of them have evolved these complex, large RNA structures that initiate translation in a cap independent manner. So some of you might be familiar with the EMCV IRES. EMCV IRES, shown on the left, it's this fairly large structured element, that is commonly used for research applications to start to generate bicistronic constructs and so on and so forth. But you can stick this into a circRNA. And then indeed, you can get protein translation out of a circRNA that's set, have these picorna viral IRESs. So that was an exciting first step in terms of demonstrating that these could actually be translated. When it comes to the diversity of translation initiation elements, there are -- I think this diagram basically shows pretty much all the caps that exist. There are some new caps out there as well. I think I read something about a benzylated something or another recently. But the cap diversity is really focused on the first three nucleotides, the methylation patterns or the modification patterns of those nucleotides focuses around the m7G and the triphosphate, the first nucleated here. And depending on your modification pattern, you can have a cap 1 or a cap 2, these can be co-transcriptionally integrated into the mRNA when you're transcribing it. And that's pretty much it. That's all you can do with caps. But with IRESs what we found early on was that there's a lot of functional diversity in the IRESs. So this is just a sampling of six IRESs, they're all picorna viral IRESs, showing that you have a wide variety of function -- or wide range of function in this particular cell type from these different IRESs, going from high to low. And sort of analogous to these different types of caps, there are different types of IRESs that engage different host factors. But notably, all -- these IRESs, just to use this example here, CBV3 and poliovirus, these are basically the same type of IRES and to actually come from the same type of virus as well. So these are very closely related IRESs. And you can see that these have pretty significant functional differences. One of them is pretty high, one of them is pretty low. And if you look at the sequences, there is a huge amount of sequence diversity even between these 2 IRESs. So once we sort of took a look at this and we're trying to decide, okay, what's the best way to optimize this IRES element, similar to how you would optimize the cap, we sort of realized that, well, there's a huge amount of sequence diversity. These sequences are like 700 nucleotides long. What can we do? What's the best way to figure out the best ones? And ultimately, we decided to take the empirical approach and just screen a lot of IRESs, which was quite challenging because these are large elements. You can't just put an oligo library into a plasma and call it a day. You have to individually synthesize every single one and test them in an arrayed screen. And so that's what we ended up doing. And what we found largely mirrored our original results with far more data points. This is a screen of actually about 3,000 IRES elements, and only the top 1,000 are shown, each individually synthesized and tested. And we tested them in primary -- this is data generated by Orna. We screened them in primary human myotubes, hepatocytes and T cells for maximum relevance. And what we found was that the original EMC vector IRES that people like to use is actually a very weak IRES at least in circRNA in certain cell types. And we found IRESs that were 40x or more potent than that cloning vector MCD. And so that was the first observation. Second observation is that there -- the best IRESs are quite rare. So you can see there's like four data points at the top here. There's a few here, there's a few here. But a lot of the IRESs were kind of average, and the really good ones took a lot of work to find. And when we did find them, they were far, far more potent than the ones that we were originally using. So that was -- those are very exciting moments for the team. So how is this ultimately relevant to therapeutics? So that's all very kind of discovery-oriented molecular biology. But I just want to tie this back into the ultimate goal of this work, which is to develop drugs. And again, this is data that's been presented by Orna. On the left is a diagram showing the potency of different IRES elements in different cell types as well as their structural identity. So in the center, this shows the structural similarity of these different IRES elements. You can see that there's a good amount of diversity in sequences. Then there's 3 rings of purple, which can be hard to see. That's the potency of the IRES elements in different cell types, so 1, 2, 3 different cell types. And then the outer ring is taxonomy. So of course, the structural similarity matches up pretty well with the taxonomy, that's expected. There are clusters of potent IRES sequences within the broader screen and the most potent sequences can be hard to predict just based on where the taxonomy and structural similarity shakes up. Now basically, we can use -- we can do all the screening work and then we can stick whatever we find into an actual relevant construct like something that encodes a chimeric antigen receptor, for example. And an in situ CAR T approach is one of Orna's lead programs, where the goal is essentially to turn CAR T-cell therapy into an injectable drug, so you don't have to go through a cell manufacturing step. And so the diagram on the right basically shows the product of this technology development, where -- this is a tumor model. These mice have been inoculated with cancer cells. It's a disseminated B-cell tumor, so [ non-sick ] cells. And then they're being injected IV at 4 doses over this time period, 5 to 15 days -- 5 to 12 days, and we're injecting lipid nanoparticles with circRNA inside, encoding our CAR T. And we can see that with our base IRES, there's not a lot of potency. But when we look at this condition, which is one of the newly discovered materials, these mice are cured. So after 4 doses, these mice have been cured of their cancer with no need for cell manufacturing. But I think the most exciting piece of this data is actually this piece right here, 0.1 mg/kg. That's an extremely low dose for a lipid nanoparticle RNA drug. And I would say, unprecedented dose. There are other people working on in situ CAR T approaches. But to my knowledge, 0.1 MPK observing tumor clearance has never been achieved before. And so this really just shows how this basic molecular biology work can translate can basically be the difference between a drug and not a drug. And so it'll be exciting to see where this goes in the future. Okay. So tying this back into applications and where circRNA might be going from both an RNA format perspective and a DNA format perspective, again, DNA on the top in blue, RNA in the bottom in purple. There's a lot of overlap between the applications of these two technologies. You could imagine using both of them for an infectious disease vaccine, for oncology applications. There are a lot of different ways that you can envision using RNA for oncology, whether it's a cancer vaccine or potentially, like I just shared, an in situ CAR T approach. And then there are things where each platform might be a little bit more specialized for a specific application, like DNA format would probably be better for genetic disease because if you use RNA, you just have to redose a lot, whereas DNA, you'll have that stable expression. RNA would probably be better for autoimmunity, where you might want something a little bit more transient. So far, RNA in general has only been proven essentially for infectious disease, maybe a little bit of cancer vaccine, but essentially just vaccines. And so it will be interesting to see what new doors circRNA is able to open in terms of application spaces. No matter what your technology platform is, having better potency out of your drug product is always a good thing. And that's ultimately the solution that circRNA offers. It's the ability to get more of your important thing out of the thing that you dose. And if you can get viral vectors that are 10x, 100x more potent than traditional AAV or adenovirus approaches, it's just ultimately going to be a superior platform. And the same is true of circRNA and mRNA. If your circRNA, if you're able to dose at 0.1 MPK, whereas you would need to dose at 3 MPK for an mRNA, then you would use the circRNA every single time. So that's ultimately the promise of circRNA is a new solution to the question of potency and solutions to safety and therapeutic index and applications all stem from the central question. All right. This is my last slide here. And so I touched on all -- most of these points at the presentation, just to tie it all together. The core advantage is increased potency, as I mentioned. Extensive optimization can lead to tangible results and can often be the difference between success and failure. And in my mind, I would say that there's probably another log fold improvement, at least in both RNA format and DNA format circRNA applications. It's still a new technology. There's still a lot of development that can be done. And because you have things like the IRES and circularization elements, you have more levers that you can pull that you can with mRNA, which is a relatively mature technology at this point. Circularization as a concept in terms of manufacturability of RNA format RNA or circularization from DNA vectors is largely solved. Things -- you'd probably boost it a little bit more, but it's largely a solved problem. And I just find this interesting because most of the newer circRNA companies seek to differentiate themselves based on their circularization process, but not necessarily the circRNA that they are creating through their process. So they're sort of solving problems that are already solved. In terms of things that need to still be done in the circRNA field, I think purification is still an open question and manufacturing in general scale up. These are things that haven't been fully achieved yet. CircRNA does not have any clinical experience. And then finally, not necessarily a circRNA problem or question, but targeted delivery and the ability to access different tissue types. That's going to be the next frontier of nucleic acid-based medicines. There are a lot of new companies out there that are starting to try to solve that question, but I don't think anyone has yet come up with a convincing solution yet. And this solving the problem of targeting different tissue types will open the door to a whole host of new technology applications that are not currently available. And so with that, I will leave the slides there and pass it back to the Circio team.

Thank you, Alex. Very impressive work. I believe this is -- the work you presented have been instrumental in bringing circRNA forward to actually being a clinical relevant platform technology. I will try to share slides. Hopefully, you can all see my screen. Yes. So just to briefly introduce myself, I am Thomas Hansen. I'm the Head of Research at Circio. And I have, together with a very talented team of researchers located here in Stockholm, been developing our circVec technology over the past a little more than 1.5 years now. So I'd like to present where we are, what we've been doing and why we also believe that, similar to what Alex has been describing, why we believe that circRNA will be the future technology, both in a DNA and an RNA format basically. Right. So I'd just like to emphasize what's already been mentioned. circRNA is a natural molecule. And I think for two reasons, that's important. First of all, expressing a circRNA in cell happens all the time. We have circRNAs found in all living animals, all plants. So this is a natural molecule. So expressing that will be quite natural to the cell. And in addition to that, we can learn a lot from nature. So we can actually harness what nature has already developed for us. So in that respect, we can take a quick step forward by just looking at how nature has evolved over time. And particularly in human, circRNA are very, very prevalent and in high abundance. So this is also something that typically correlates with organism complexity. The more complex the organism, the more circRNA do we have. And of course, from an academic perspective, that is in itself very interesting, but this is not our focus. I'll be a little -- maybe a little technical in the first slide here. So I'm just going to show you basically a growth decay model. It's very mathematical. It's not so important actually, but it's basically just to show that all expression level, all molecules within the cell, the expression level, especially governed by two things. It's the production rate, so how many molecules are you producing per unit time and of course, the decay rate, so how quickly is that molecule then degraded afterwards, again, right? So there's two approaches to achieve high-level expression. And the obvious one is, of course, increasing the production rate. But sometimes what people forget is, you can do the other set. You can do some of the back way approach. You can reduce the decay rate and that would actually achieve the same thing. And that's where circRNA come in and we already touch upon. Then naturally, they are more stable then orders of magnitude, almost more stable than mRNA. And then intrinsically, we will achieve at least half of the goal here to get long and durable expression. So naturally, we reduced the decay rate by just having the circRNA format. In addition, we are then working, and this have been sort of the core of our technology development, to enhance production. So we get increased production rate or increased expression rate by enhanced production and reduced decay. So that is basically why we use circRNA and this is sort of bioinformatically sort of simulating how we believe the expression profile would look over time from when we apply the vector into a cell and then following expression. And the numbers here are empirically based. So we measured the half-life of our circRNA molecule. It's roughly 130 hours, whereas you have less than 10 hours for mRNA. So that's a 15x improvement in half-life. And how would that look over time? So here we sort of conservatively set circRNA production rate quite low. So therefore, you actually get, at a very early time point, better mRNA expression, but quickly circRNA will surpass mRNA levels, and it will accumulate, and the area under the curve will be way higher for circRNA-based expression. So this is sort of the theoretical expectation of the expression profile over time that we have been working towards optimizing and getting that red line as high as possible. So going back to nature. So how does this work? Erik and Alex already mentioned, back splicing as this is the mechanism of which natural circRNAs are being produced. This is the root of biogenesis. And very generally, what you can see here is how gene expression basically works to the top. You have linear splicing happens all the time. This is how mRNA are produced in cells. You have splicing of these blue exons, but then a sort of a competing alternative splicing event, you back splicing. That happens at sort of specific loci within the human genome. And in that sense, you can basically look at those specific loci, you can cut them out, take basically a naturally occurring DNA sequence, put it into sort of a small expression vector, and then that vector will actually produce your -- the circRNA of choice here. So the first step we basically did was try to identify the loci that are naturally occurring in the human genome that seems to give rise most effectively to circRNA. So we selected a handful of different loci that we're showing high level of circRNA expression. So this is basically the x-axis where you can quantify circRNA expression. You have thousands of different loci to choose from and you look at how much circRNA do we get per transcriptional event basically. Selecting the best ones, testing how well do they perform then when expressing it in a small cassette. And using that approach, we identify at least one locus that seem to be superior in all our tests, the IR1 here that seems to be more than 5x better than other loci we've been testing. So that's sort of was the basis of our circVec design, based that on a natural sequence. So this is all what has been given to us by evolution basically. But of course, being scientists, we are very interested in learning what are we looking at. Is it actually circRNA, we are measuring? And there's different ways you can approach this. So one approach is that you can take these flanking sequences that are required and necessary for the back splicing. You can -- as you can see to the left, you can remove one of them and see how would that impact your circRNA signal. And of course, taking that away, if nothing happens, then likely you're not looking at a circRNA because you removed a critical element. But what we observed is if you take away what is required for the circRNA biogenesis, the circRNA goes away. So that's sort of validation that what we're actually looking at is indeed a circRNA. Maybe a little more straightforward approach is to actually take the RNA that's being produced, subject that to some enzymatic treatment, in this case, treating it with an exonuclease called RNase R that will remove all linear RNA from your -- in your test tube basically, and you can then see what's left. And what we observed is that the mRNA disappears, but the circRNA that we are expressing sort of persists and is resistant towards this treatment. So once again suggesting that we are actually producing circRNA from our vector system. So we're pretty confident that, that is the case. As Alex very elegantly showed you that various elements are critical for effective translation from circRNA due to the fact that it's cap independent. So of course, similar to what Alex and Orna has been working on for a number of years, we are also interested in identifying effective IRES element for high-efficient protein production. And just to briefly show you that this is critical, it's black and white, basically, if you don't have an IRES element, you don't get any protein. If you put in an IRES element, you can get protein from circRNAs that are then expressed here from our circVec platform. So that is clear and expected. But then the -- what was a bit more unexpected to us is actually now we have the circRNA sequence that we want to end up with, but we need to engineer that into a linear DNA expression cassette. So how do you do that most effectively? And it was a little surprising to us that, that was actually quite critical how you go about engineering that or designing that linear cassette. So we tested a few different things. Consistently, we saw specific designs that were superior. We're able to file IP on those design rules. And I think that's a critical piece of information that we never have at Circio as to how do you effectively from a DNA vector actually express circRNA that are then able to encode your protein of interest. So that seems to work irrespective of what IRES element we're using. Actually also irrespective of what protein we are going to express. So it's a very general design rule that we have identified. And I think that was a critical step in the circVec development that we've been working on the past 1.5 years roughly. So I think this was sort of some of the early work basically that went into establishing what we now call the circVec 1.0 cassette. And that's basically shown on the top. So we have this cassette, we've shown that, that effectively expresses the circRNA, and we've shown that, that circRNA then effectively encodes the protein. So that's sort of -- that's been established. But now, of course, what our main focus, at least at sort of the in vitro development or platform development is to optimize the biogenesis that is how do we improve the level of circRNA that we get out of our vector system. And then secondly, once we get the circRNA expressed, how do we improve or enhance the number of proteins that are then encoded from that circRNA. So these are sort of the 2 obvious optimization steps very similar to what Orna has been working on. You want to get the best possible biogenesis. You want to get the best possible translation. We are quite different in terms of the -- how we do the biogenesis. So this is where we differentiate ourselves at least compared to some of the synthetic circRNA companies because this is a completely different endeavor for us. But just to show you the optimization that's been ongoing. This is basically all happening in the flanking regions. That is where the recipe for back splicing is positioned. So you can take these flanking sequences, you can manipulate them, you can rationalize or try to rationally optimize the sequence structure, and we've done that for a period of time, and we've identified now a few different sequence variants that give us around 5x improvement compared to this natural design that we started off with. And you can see that on measuring circRNA levels to the left. And you can see that to the right where we basically measure the protein that are then expressed from the circRNA that seems to correlate quite nicely as expected. And we now have this improved cassette for biogenesis established. Secondly, we also have been working on IRES optimization to testing, not as elaborate yet as for Orna, but we tested quite a few different elements being encoded from our circVec DNA cassette, I'm showing here roughly 9 different IRES elements that came after the circVec 1.0. And as you can see, we sort of incremented -- incrementally increased protein yield from our vector platform, and we now ended up with our most effective IRES element. This is now what we have coined the circVec 2.0 because we now have roughly one order of magnitude increase in protein yield per circRNA copy number. So that was another milestone in the R&D development. So that is an important step going forward for us. So now we have this improved IRES element. And I think sort of what Alex also hinted to and what we believe is the case. We have basically now a circVec platform that is likely, at least over time, due to the durability and the high-yield protein expression to be the preferred platform, either that is in a synthetic context on a DNA-based context that may apply to a different indication. But in any case, it seems the circRNA design will be superior when you develop these optimized sequences. And what you can see on the right is basically where we are at the moment, and I think this is critical data, where we bring in not only the enhanced biogenesis and enhanced translation. We bring that into the same vector that's now circVec 2.1. And what happens then over time when you also sort of leverage the fact that it's a more durable, more stable RNA that you're expressing. So what you can see actually even at the early time point, our circVec 1.0 vector as you can see starts off quite low actually, but it builds up over time, and it surpasses the mRNA after 5 days roughly. But as we were expecting in our modeling, we were expecting lower expression at the early time point due to less -- lower biogenesis or slower biogenesis. But now with the 2.x generation of circVec vectors, we are on par or even higher than mRNA at the early time point, and then it builds up over time. So at here, we are looking at 8 days after we introduced our vector system into cells, and we are almost 10x higher than mRNA expression. So that is -- I think that's very important data for us going forward. And of course, we're also very excited about now characterizing the circVec 2.1 in an in vivo model. So how we approach this is basically, we have developed a somewhat elegant system, I would say, where we basically in one mouse inject a DNA cassette that expresses an mRNA encoding FireFly Luciferase and then the right hand like we encode our circVec 2.1 expression cassette that also encodes FireFly. So you can basically using FireFly, you can monitor luminescence in real time in these mice. So this is the same mouse you basically see as a representative image here at the bottom that one day after this was injected intramuscular and up to 3 weeks after this was injected intramuscular. It seems the temporal profile is a little slower in vivo than we are used to see in vitro. We do see a little slower buildup of circRNA in this pilot experiment, so very low expression from the circRNA time point, but it builds up, and now it's slightly higher than mRNA here at day 21. And this is data from this week. So these mice are still running around happy here across the street. So we will get read out on a weekly basis. following the expression profile for the mRNA and the circRNA vector. So I think this is very exciting to see what happens and just to have now an established in vivo setup where we can actually benchmark our circVec designs against mRNA-based designs. Good. So with that, I think we are -- in terms of protein expression, I think we've achieved quite some milestones, at least in the in vitro development. So that's sort of is relevant in the scenarios where you may have a loss of function indication. So you have a protein that's missing, for instance. That will be good to be able to replace that protein in a long and a durable manner, right? But you also have the opposite case where you may have a gain of function scenario where you have an unwanted protein, a protein that's toxic, possibly that is cancerous. There's many different scenarios for that. And so we have also been developing what we call the remove and replace design. So basically, what we have been working on is in addition to having a protein coding aspect in our circVec design, we also have a remover element. So we can basically just position this at the end of the cassette as shown. So we can, from the same cassette, express a protein via the circRNA, and we can express this remover entity that can actually in a target-dependent manner remove the [indiscernible] then encode this toxic protein that you want to get rid of. And we already have proof-of-concept data that shows that positioning this remover is not affecting circRNA production, so we still get circRNA, we still get the protein expressed from the cassette, but we can target our gene of interest. So here, we get a significant reduction in the target gene that we were here aiming to reduce. So with that, I think we have established not only that we can encode circRNA immediated protein, but we can also -- for replacement, but we can also remove toxic proteins. And I think our CSO will take over in a minute and talk about where we believe that could be deployed in a meaningful manner. So I think that brings me to my summary slide. So I think we've been able now to optimize circRNA biogenesis and protein expression substantially. We now have this remove and replace concept is also established, so we can co-express circRNA with this remover element. We have ongoing in vivo validation. I showed you some early data from this pilot study we are currently doing in mice. And in this next section, our CSO, Victor Levitsky will take you through some of the more therapeutic relevant settings that we believe this could be highly relevant to deploy the circVec technology in. And with that, I thank you for listening, and I'll give the word to Victor.

Yes. Thank you, Thomas. With all this extensive background on the technology that you have just received from Alex and Thomas. What I would like to do is to discuss with you shortly potential clinical applications of this. And clearly, from what you just have learned, the technology creates possibly the best fit with clinical situations where increased protein expression. But maybe most importantly, extended longevity of protein expression should result in much better efficacy. And at Circio, we see 3 areas where this seems to be the case, represented by genetic rare diseases, vaccines and cancer gene therapy. And at the moment, we prioritize the first 2 areas of development because this allows us to define really clearly shaped and focused clinical development programs in the future. So with a rare disease, we see this as a major long-term potential for the company. And this is also the space where we can fully explore this concept of remove and replace technology that Thomas has described to you, and I will come back to this in a minute. With the vaccines extensive longevity of protein expression should also lead to a much better immune response, and importantly, a potential for very simplified administration of vaccines. So basically one shot of administration, one dose injection of vaccines. And as we generate more data, validating our technology in different context, we believe that cancer gene therapy also develop into an attractive area, especially for partnering with bigger pharma companies. When it comes to genetic disease, as you may know, that there are thousands of genetic diseases described. And we had to go through a very extensive exercise of analyzing many of them trying to identify the ones where our circVec platform would fit best. And this was based on several prioritization criteria including, of course, physiopathology, genetics of the disease, the size of the gene, which should have fit into our platform. So in simple words, the gene couldn't have been too big for our cassette allowing its expression. But also the frequency of this disease and its manifestation should allow a reasonable clinical development program in the future. So based on all of these criteria, we identified now currently 8 diseases and alpha-1 antitrypsin deficiency represents the current #1 priority for us because this is a disease which is quite well known, but at the same time, previous attempts of genetic therapy of this disease were not successful. And we believe that this is the indication where we can effectively utilize our replace and remove technology, and I will come back to this on the next slide. The second priority group is represented by different genetic diseases, but then falling into the same class of so-called urea cycle defect. So the genes which lead to the development of this disease are different, but they are all located in the metabolic cycle of urea, which primarily takes place in the liver. And this means that there is no demand for a very sophisticated delivery technology, which can be achieved with relatively well-known gene delivery techniques. And then proof-of-concept data, which would be generated with one of these diseases should be easily extrapolated to the other forms of urea cycle defects caused by mutations and other genes. Other diseases in the third priority group involve not only the liver, but also other organs and tissues and effective development of gene therapy in this space would require maybe more advanced gene delivery technologies that we are now analyzing in ongoing collaborations with our gene delivery companies. And then depending on which of the platforms will look more promising. We can then consider a clinical development program in one of those indications as well. So coming back to alpha-1 antitrypsin deficiency. It is important to know that this disease has actually 2 major manifestations. One manifestation in the lung which results in different inflammatory processes and eventually obstructive pulmonary disease in the lung due to the lack of the wild-type protein. So if the gene is mutated, and the wild-type protein is not generated then this pathology in the lung develops over time. But certain mutations in the gene, which actually is more effective and where the production of the protein takes place in the liver may lead to generation of a mutated variant of the gene, which accumulates and become toxic in this organ. So ideally, the most optimal treatment of this disease should do 2 things: substitute replace the missing wild-type protein and then prevent accumulation of the mutated toxic product in the liver. And we think that we can achieve this with our replace and remove technology, which Thomas has described already. Here is some additional data demonstrating proof-of-concept results already using our expression cassettes. So you can see here that when we compare -- when we look at replace aspect of this approach, then expression of the wild-type protein from the circular RNA construct over time becomes much more effective the expression, which is driven by mRNA on the top graph of this picture on the right. And then we can also express suppressive factors, which down regulate expression of the mRNA. This can be developed for any sequence. But in this particular case, we are specifically down regulating, suppressing the expression of abnormal mutated toxic protein, and this does not affect as shown on the slide, expression of the wild-type protein or a different less toxic variant or muted variant of this gene. So then this allows us then to address both aspects of pathology, which are intrinsic to alpha-1 antitrypsin deficiency. Most of the current gene therapy approaches are based on utilizing a then associated virus as gene delivery tool and then associated virus as a DNA virus, which can support expression of our circVec cassette. And we have just recently generated data demonstrating that, in fact, our cassette allows generation of circular RNAs from the [indiscernible] viral genome. But there are several complications related to the clinical use of this technology and all of them are related to the fact that the dose of AAVs has to be very high for sufficient efficacy. And this results in liver toxicity, very high immunogenicity and also very high manufacturing costs. So all these aspects can be addressed with our circVec approach because it has been already discussed many times today due to increased stability of circRNA and increase efficiency of translation, we can produce much more protein from these vectors. And then our modeling, which now has been introduced by Thomas, but now based on sort of more actual experimental data points that we can put into the model predicts that the area under the curve achieved the circular RNA will be significantly higher as compared to a similar vector including the protein using conventional mRNA. And yes, so the expected increase is approximately 25-fold higher. Similar considerations are applicable also to vaccination, where we believe it makes sense to utilize adenoviral vectors. And you may know that historically, the company has significant expertise in the development of oncolytic virus therapies and vaccination in cancer. So we are planning to combine all these approaches together. And then express circular RNA from adenoviral genome. Again, we have recently learned how to do that without affecting significantly the fitness of the adenoviral variants and the relevant IP has been already filed. The adenoviral genome is pretty large, which allows expression of several different payloads. This then can include either several different proteins, genes of the same pathogen. So you can vaccinate against several proteins from the same virus or the same bacteria. And you can also combine it with expression of different immunomodulators that can either boost immunostimulatory pathways or through the natural capacity of circular RNA to [indiscernible] inhibitor microRNAs that could also result in increased immune stimulation. And you can see on the slide that we are expecting some major milestones to be achieved at the end and early next year -- at the end of this and early next year. Just one additional illustration that already with our not very advanced circVec 0.1 platform, we were able to express higher levels of COVID spike protein, which we are going to use as a model in our vaccination experiment to boost immune response against the COVID virus. And you can see that even not advanced platform allows us to achieve higher levels of spike protein expression in vitro. And as we speak, we are progressing with experiments, which are utilizing circVec 2.1 vectors expressed from the adenoviral genome. And this will be used for vaccination and analysis of immune responses in mice. So this brings me to the end of our presentation. And just to summarize everything that we have said to you today is that we believe that we are at the moment the only current active player in the field of DNA dependent circular RNA generation and enhanced durability of protein expression, which can be achieved using this platform will allow us to reduce dosing of a variety of DNA formats, virus-based or DNA-based, which we did not discuss today significantly in detail, but that's another big area of gene therapy that can be covered also by this technology. And then using this, we can significantly decrease challenges, which are faced by the field of gene therapy. And generally speaking then, this approach may become the preferred major approach which should be used by any DNA-based therapeutics in the future. And with this, I conclude our presentation for today, and we'll open the chart and e-mail for questions and Q&A. Yes. Thank you very much.

Okay. Very good. We have a few questions that have come in both for Victor and Alex and Thomas since we were just listening to Victor tell us about gene therapy and the plans in gene therapy. There is one question here on gene therapy in Ag. So gene therapy has previously not been successful with AAV viruses in AAV. Why do you expect that circVec can succeed where others have previously failed?

Yes. I think the simple answer to this question has been already provided during our presentation that we simply are going to produce more therapeutic product over a much more extended period of time, and this should result in better efficacy. And I think that was the major challenge. And this remains to be a major challenge for different gene therapy approaches. And in addition, again, as this has been discussed, if you consider this in the context of utilizing the adenoviral -- the adeno-associated viral vectors for gene delivery, then we can also significantly decrease toxicity and increase the therapeutic window, and this should result in better treatments.

Could you comment on when you would expect to be able to enter the clinic for an AAT circVec gene therapy?

Yes, our pretty ambitious aim is to be -- to arrive to this point in approximately 3 years from now. But of course, it depends on multiple factors, but that's the goal that we are defining for ourselves at the company.

Okay. Then we can move on to a question here for Alex. So, Alex, you mentioned that it had taken 20-plus years for mRNA to get to where they are today with a approved COVID vaccines and circRNAs only at year 5. Where do you think circRNA will be once you hit year 20?

Very hard to predict out to 15 years in the future, I'd say. But I feel that there's still quite a bit more development that can be done to the circular RNA vectors themselves in terms of increasing expression and a variety of other things. I would say that circular RNA is about as good if not better than mRNA pretty much everywhere as the technology stands today. And of course, everyone is using different versions of the technology. And most of those versions of the technology are not kind of the most developed or recent [indiscernible]. But -- so with another 15 years of development, I would see circular RNA essentially displacing linear RNA format technologies for pretty much every application. I don't really see a niche existing for mRNA if circular RNA can basically do everything that mRNA can do and more. But it will take significant development effort to get there for circular RNA as it has for mRNA.

So following up on that, what do you think is the most promising usage of circular RNA or what's frequently referred to as the killer app? And maybe you can comment on both the synthetic RNA as well as the DNA format, circRNA from that perspective?

So the killer apps, I think, will continue to evolve and that is dependent on the development of delivery vehicles, particularly in nonviral-delivery for synthetic circular RNA. The killer apps right now are basically defined by what your delivery vehicles can get to. And so I think access to local tissue environments, such as intramuscular injections access to the liver, access to immune cells, those are going to be the first places where you're going to probably see a circular RNA technology being applied. As delivery technologies become more mature, I would say, the killer app will also evolve. And it's -- the nice thing about the nucleic acid technologies is that they're pretty modular. So you basically develop the platform once. And then as all these things come through the line from companies developing delivery solutions and things like that, you just basically plug and play more or less. And so the application landscape is currently the narrow as it will ever be for nucleic acid therapies. And that's for DNA therapies, that's for RNA therapies. We are at a point in time where we have not yet undergone a delivery revolution. I think that delivery revolution is starting, and it will happen. And after it happens, companies like Circio and companies developing these really strong underlying format technologies will have their pick, basically, of the application space. Sorry, that was probably a little bit more abstract and complicated answer. Today, vaccines originate diseases.

And then as you say as delivery technologies evolve, so also the prospects of where you can deploy both the synthetic circRNA and DNA format circRNAs. And they will go hand-in-hand the development of these technologies, the DNA and RNA on one side and the delivery to bring those DNAs and RNAs where they need to go on the other side.

Yes. Absolutely.

So we have another question here from the audience to Alex. How competitive is circVec compared to Orna and others?

CircVec in reference to the Circio DNA-based platform, how competitive is it? I would see them as 2 different platforms with different application spaces. Of course, there is some overlap like they can both be used as vaccines. [indiscernible] originate diseases, I'd probably rather use circVec's tech because with a synthetic circular RNA, you'll be redosing pretty frequently. So you'd have to have a good justification for that. But I don't really see them as necessarily competing technologies. They are fundamentally different platforms and one's RNA format, one's DNA format.

And if we follow up on that competitive question, your vaccine comment, maybe, Victor, you could comment on how you see circVec could have an advantage or a differentiating factor when it comes to vaccine versus now mRNA, which is becoming the gold standard vaccine format?

Yes, I think that it is very well known, very well established in the field of standard basic textbook immunology that long-term exposure to the protein to the antigen is very important for effective maintenance of immune response. And that is one of the things that classical adjuvants do. They create a deeper of the antigen and then they allow this prolonged stimulation like aluminum salts and oil solutions, they allow to release slowly the antigen to the system and thereby stimulating immune response too effectively. I think the danger with our platform is that it will become too good in doing that. And if it stimulates immune response too strongly for a too prolonged period of time, we may even run into an issue of tolerance. But let's take these problems in the order they arrive. And I think at the moment, it is really important that we can achieve this prolonged antigen exposure. And as we already brought up today during the presentation, this may also lead to more simplified regimens of vaccination where maybe one shot will be sufficient to induce a very effective immune response. So that's how I see this.

So you can enable effectively a single dose format potentially?

Yes. And where I also see a strong advantages, again, as we discussed today, that DNA delivery systems allow combining this approach with immunogenic viruses like adenovirus, and then equipping these delivery systems further with additional immunostimulatory payloads, which is pretty difficult to do with the conventional mRNA due to size limitations. But DNA platforms, they may allow doing all of that. So there might be a completely new wave of vaccines sort of coming next as the field progresses. But I think that at the moment, there is a significant potential advantage in building this simple concept of higher extended protein exposure when it comes to vaccination.

So we have one question here. I would like to format it a little bit and post it to both Thomas and Alex. During your time in developing circular RNA, what has been the biggest surprise or the most interesting piece of data you have seen? And second, what do you see as the biggest challenge that remains? So maybe Thomas can comment from the DNA format perspective and Alex from RNA. So why don't we start with you, Thomas.

Yes. I think from -- the biggest surprises, I think, from sort of coming from academia and going 10 years back, just working on understanding the biology of circRNA, I think it's quite surprising that we are already here. We didn't have an eye back then at all for therapeutic relevance. We were just trying to understand basic biology. And now we are drug-developing circRNA platform. So I think that's a big surprise that, that has happened in such a short time frame. But for the development at Circio, I think what I also tried to explain and that was sort of the basis of our circVec IP portfolio is how you actually compose that circRNA [ excellent ] for effective expression, which was critical achieve a functional and relevant protein level. And that was a little surprising to us, and that's also why we're able to IP protect that discovery, namely due to the surprising nature of that finding. So obviously, that was a big surprise because we actually didn't expect to see that. And it was very consistent across IRES element payloads. So it's always nice in science when something is a general thing that seems to apply to generally across IRES elements and protein payloads. So we've also got to touch about the challenges set before I leave the word to Alex. I think the main challenge we had at Circio is like when I -- I didn't touch upon that actually during the presentation, but the circVec 1.0 was actually not particularly effective in mouse, it turned out, at least not in murine cell lines when we tested that in vitro. It didn't express as highly as it did in human cells. So that was, of course, challenging because we actually like that expression in human cells. It was actually on par at least at the early time point compared to mRNA. But then in mice, it was subpar, it was below. So for 2 reasons going into this circVec 2.0 development was basically not only to, of course, optimize the platform, which is relevant, but just to enable a relevant in vivo setup because putting the 1.0 into a mouse would likely not yield an encouraging data point going forward. So I think that was a little challenging knowing that we need to develop something that also here applies not only to humans but also to rodents, basically. And that doesn't necessarily -- that's not necessarily the case when you optimize your platform, but...

Different cell types are different [indiscernible] organisms, it becomes complicated.

Yes. But luckily, the 2.0 -- the 2.1, this generation of circVecs seem to be pretty effective also in murine cells. So that's more encouraging going forward with, for sure, yes. Alex?

Yes. I mean I would -- well, the biggest surprise, let's say, when that was sort of in the early days of Orna, I'd say, it was COVID and the sudden interest in developing vaccines because we started the company shortly before COVID, a few months before COVID. And there wasn't a lot of investor interest in developing vaccines at all. And so that changed very quickly after it became evident that we had this global pandemic on our hands. But from a technology perspective, I would echo a lot of Thomas' comments, it's the cell type, tissue type species, variability, maybe variability is not -- it's maybe a pessimistic word, but specificity, I guess. And I would say that there is an extra layer of specificity of regulation when you're dealing with circular RNA as compared with mRNA. Because you're dealing with this complex IRES element that circularization element, especially if you're circularizing from a DNA vector, you may have additional -- another layer beyond that. And I think that is one of the challenges of circular RNA. People do IRES screens and HEK-293 cells, it's totally worthless. [indiscernible] as well not even around the experiment. And so picking your system is extremely important. Because you'll get the answer you look for, especially with circular RNA or there's this things may not translate from one system to another.

Thanks. Back to Thomas, what are the major milestones you expect going forward?

Yes. I think we are looking forward to a lot of milestones. So as I showed you some of the proof of concept in vivo data that we're working on basically expressing a reported genes such as FireFly Luciferase, just monitoring the expression over time, comparing that -- benchmarking that to mRNA expression and be able to understand our vector format better in a mouse model. I think that's a critical milestone, and hopefully, we'll be able to show early next year data points. I mean beyond what I've showed you today, how does this develop in a longer time frame and how do we see expression profile building up over time. And I think that will be something we'll be reading out then. In addition, we are testing different formats of expression, so different DNA formats, different viral formats. And we have that in the pipeline already. Some data we already have read out on in vitro, and we are moving into in vivo systems as well for those formats. So that will also read out early next year for some of them. And then, as Victor explained, we'll have -- we'll go in more therapeutic relevant directions, testing circRNA in a vaccine context, testing circRNA in a rare disease such as alpha-1 antitrypsin, we will also get data on how that performs compared to a classical mRNA-based gene therapy or vaccine approach. So these 4 pillars, I would say, would be the main milestones that we are looking forward to in Q1, maybe Q2 next year.

And finally, there's a question on IP, is your technology IP protected, and maybe I can comment on that. Yes, it is. Both -- there are several aspects of IP protection here. What Thomas referred to as the design and designing was surprised about that's kind of the core IP, the features of the circVec cassette, then we have filed IP around utilization in specific vector types. And we have filed IP relating to how we can further boost and optimize the translation. And we're planning on building out a portfolio of IP around our technology. And given this an early space, there is not so much IP, at least publicly available or published on vector express circular RNA at the moment. So that's an important part of what we do is generating relevant data for building a strong IP portfolio. So with that, I think we have come to 5:30 in Europe. That concludes our 90-minute slot, and we've handled all the questions. Thank you for attending. Thanks a lot, Alex, for participating all the way from Boston. It was very interesting to hear your perspectives. And thank you, Thomas and Victor.

Thank you.

Thanks.

Thank you very much.