MeiraGTx Holdings plc (MGTX) Earnings Call Transcript & Summary

As a reminder, this conference is being recorded, and a replay will be made available on the MeiraGTx website, following the event. I would now like to turn the call over to your host, Zandy Forbes, President and Chief Executive Officer of MeiraGTx. Please go ahead, Zandy.

Thank you very much. Good morning. I'm Zandy Forbes, the President and CEO of MeiraGTx, and thank you for joining us today. We're going to present some of the unique technologies that we've built Meira and discuss how these tools can transform the field of genetic medicine. In particular, I'm going to talk about technologies that allow us for the first time to precisely and specifically control the activity of genetic medicines, using orally dosed pills. I'll show how we're applying these technologies to some of the intractable problems that small molecules, injectable biologic drugs and unregulated gene therapies have not solved. For example, these technologies can be used to check cancers in the brain, a site that today's biological oncology drugs cannot reach because they can't cross the blood-brain barrier. I'll show how this technology allows us access to the world of therapeutic peptides and hormones that are involved in metabolism response to exercise, satiety, fat and sugar homeostasis and aging. We can activate therapeutic drugs in the eye with eye drops rather than injections, as well as deliver to the eye appropriate doses of therapeutics that cannot be safely continually dosed. And we can potentially solve the problem of vaccine boosters, which is on everyone's mind today. Following a single IM injection of a regulated vaccine, we can boost either antibodies to COVID or antigenic spike protein messenger RNA using a pill. One vaccination than any number of oral boosters at any time needed with no more vaccine manufacturing, no more supply chain issues, no refrigeration. We can use combinations of the above, boosted with a pill to produce broad, durable, immune protection. These are just some of the applications we're developing right now. But even with all of that, we're just scratching the surface of this technology. The opportunities are boundless. One day, not so far away, we'll look back and wonder with complete bafflement how we ever put genes into people that we could not completely regulate. Before we begin, please note that we'll be making forward-looking statements as part of this presentation, which statements are subject to certain risks and uncertainties that may cause actual results, performance or achievements to materially differ from those forecasted. Certain of these risks are described on this slide, there it is, of today's presentation and in our most recent filings with the SEC. And now I'm going to move on to give a really brief overview of Meira and describes the strategy, which led us to where we are today. So when we set up Meira 6 years ago, we intended to build a pharmaceutical company that was going to innovate in the area of gene therapy in order to change how it could be used, so we wouldn't only be focused on rare diseases and gene replacement, but rather develop technology that would allow us to cost effectively treat a broad range of serious disorders. So to that end, our pipeline was selected to try and avoid some of the pitfalls we saw in the gene therapy technology at the time, such as large doses with potential side effects, difficulty in manufacturing and some safety issues, when you deliver systemically. So we chose therapies all of which have human proof-of-concept that gave local small doses in somewhat immune protected areas, which brought us to the eye inherited retinal disease, the salivary gland and Parkinson's disease, where we are focused on a GABA agonist just in the subthalamic nucleus. All of these areas of local delivery are also sites that we could use to regulate gene therapeutics in the eye to potentially regulate with an eye drop and in the salivary gland to potentially produce hormones and peptides, which is secreted into the serum rather than the saliva. But we have this pipeline, this derisked pipeline that we currently are taking through the clinic, 2 programs entering pivotal. We just had data on our salivary gland last week -- I'm sorry, my slides going there. Just one minute, salivary gland last week, and we have Phase II data in our Parkinson's program, which will enter a new IND very shortly. But what I'm really going to focus on today is our vectorology capabilities and our gene regulation platforms that has allowed us to produce a pipeline of novel drugs, which can be switched on and off with small molecules in a precise specific dose responsive fashion. Before I go on to that, just to mention, before we started Meira, one of the first things we did was focus on manufacturing. So we have in-house 2 GMP viral vector facilities, we have our own internal GMP plasma production, we, by the end of next year, will have internalized all our QC, our analytics for release and stability, we have quality systems that cover all our facilities that are approved and can support IND through commercialization. And over the past 5 years, we've developed a process that allows us to very rapidly fit any capsid, any genome to this platform process and transfer it into GMP. So this gives us the capabilities to initiate an IND with our own manufactured material and take that all the way through and with scalable within the same facility, using the same process, all the way through to commercialization, which vastly reduces your regulatory risk in clinical development and you see the flexibility of any of that material. So today, I'm going to focus on the viral vector design platform and spend a lot of time talking about our gene regulation platform. So before I do that, let's just look at what a gene therapy involves. So as we know, there is a capsid and the capsid drives tissue tropism. This isn't so much tissue specificity, but you see differential transduction efficiency with different capsids, which can drive potency if the capsid is particularly efficient in the tissue of interest. So we do have some programs that in capsid engineering, where we use directed evolution and screening to find capsids of particularly efficient at transduction of our tissues of interest. Now a very important aspect of the vector gene -- of the gene therapy, obviously, the vector genome, what does that entail? Between the ITRs, which allow packaging, you have your therapeutic transgene, but around that transgene are a number of elements, which can be optimized to allow you to switch that genome in the right cells at the right levels to have the maximum potency, efficacy and safety. So these regulatory elements, which include promoters, enhances, introns, codon optimization, Poly A and you can even have knockout siRNAs those are super important for determining the level of activity of your gene, which drives potency, the more potent your gene, the less of it you need, down comes the dose, increase safety. So all of those elements are required in any vector genome and must be optimized whether that vector genome is going to be regulated or not to have the most potent and effective therapy that you can. Now the therapeutic CDNA is transcribed to make messenger RNA, and it's at the level of the messenger RNA that riboswitch regulation works, which I'll talk about later today. So we have developed a completely new method of regulating genes, which involves RNA stability. When the small molecule is absent, RNA is completely degraded. When the small molecule is present, the RNA is processed. It's the same as RNA coming from a therapeutic transgene without the switch and you get therapeutic protein produced. So this is an overview of the vectorology toolkit addressing many of those elements of viral vector that I just mentioned. The promoter. The promoter, I'll talk about in the next section, but this is the basis of the control of cell specificity and cell levels. And we have 4 different platforms for mining the genomes from viruses to human to design bespoke promoters, large promoters, small promoters, strong promoters, cell-specific promoters that allow us to really precisely determine how and when a viral vector is expressed. Then there are all these optimization tools. We have this deep toolkit, some tissue specific, some not, that allows us to further tweak and optimize the viral vector. I mentioned we have capsid evolution. We currently have an ongoing screen in NHPs, which is looking to capsids not only to get to the back of the eye, but capsids that really efficient in transduce the front of the eye, where we can potentially put genes to be regulated by eye drops to secrete the therapeutic drug into the vitreous. We also focus on manufacturability and have a vectorology platform that ensures that as we make these optimized vectors, they are manufacturable and can go into our platform process to the GMP manufactured at Meira. And finally -- or not finally, but on top of all of this overlaying all of this optimization and promoter enhanced control, we have developed a gene regulation switch, which allows those gene therapies to be turned on and off using oral small molecules. And not just turn on and off, we've now developed technology that allows these optimized gene therapies to be regulated tightly in response to a dose of a particular small molecule, and that's what I'll show you as we proceed through the presentation. So first, promoter optimization and engineering. So promoters enhances regulatory elements exist in the genome of eukaryote viruses, they exist out there. And we have many platforms for designing, engineering, seeking and putting together promoter enhancer combinations. We -- some of our promoters, particularly in our eye program, we built by hand, and we did that in human organoids so that we know that the promoters we're using in our specific gene therapies work in human cells -- in the right human cells and express at the right level in human cells to have a therapeutic effect, and I'll show you one example of the importance of that. We also have a moderate scale screening system, where we can rationally design promoter enhancer combinations and screen different combinations that we seek from different databases or our own design to see if we can enhance the expression of particular promoters. We also generate promoters enhances by transcription factor binding site shuffling, and we do larger scale barcoded promoter screens, using fragments of genomic sequences to get promoters seeking synthetic promoters that may work in some tissues much more strongly than others. And additionally, we've recently bought in-house in silico screening, where we have -- we use AI to further evolve promoters, using machine learning methods, which has been really successful, actually. So first of all, bespoke promoters. This is an example of 2 programs that are currently ongoing in clinical development in the eye, partnered with Janssen in the area of achromatopsia. Achromatopsia involves a complete lack of cone function because a channel, which is made up of CNGA3 and CNGB3 is missing. You lose either A3 or B3 and your cones don't work and you're blind from birth. But CNGA3 and CNGB3 don't bind one to one to make a channel. Rather, they join together in 1 unit of B3 to 3 units of A3 create a channel. That means that different levels of B3 and A3 are required in order to make -- in order to rescue the respective mutations. So we first made a CNGB3 construct using the cone arrestin promoter, a promoter that activates CNGB3 in all cones. It worked really well. CNGB3 is only 1 out of 4 of the units in the channel. And you can imagine, in the absence of B3, you just need to add a little B3 to associated with 3 A3s you get working channels. However, when we put the cone arrestin promoter in our A3 construct, it did not work. And so we had to design through promoter engineering in human retinal organoids, the strongest pan cone-specific promoter, opsin are created, taking pieces of red and green and other opsin promoters and joining this together iteratively to find a promoter that was expressed in every single one of the cone types of the human retinal organoid. This is a human retinal organoid, by the way, and these are the cones coming out of it. When we did that, we were finally able to rescue the CNGA phenotype because we had sufficient CNGA3 expression to join with the B3 in a 3:1 ratio, any less than that, and that ratio is hard to get, and you don't get rescued. And by the way, when we put this super strong promoter that we developed into B3, it blocked activity because too much B3, you don't get the 1:3 ratio and your channel doesn't function. So this is just an example of how critically important your promoter is for the efficacy as well as the potency of each of your products. So quickly, our medium-sized screening, we rational design screening, we take tissue-specific promoters or constitutive promoters and mix and match them with enhancers, which can come from that database is like ENCODE and FANTOM. We have a marker here driven by a commercialized SV40 and then a second marker here, and this is where we put our promoter enhancer combination. We've screened over 300 of these sorts of promoters. We do it in live cells. We sought the cells, and we look for a ratio of RFP to GFP to see relative promoter potency. We can also do higher throughput screening with barcoded fragments of human genome. We can put those barcoded fragments either in vitro or in AAV into mice, and again, look for expression at high levels in particular tissues or particular conditions. We can similarly do our transfection factor binding site shuffling and also barcode that and these are the sorts of places we get the libraries for our barcoded screens. Now I'll move quickly to our AI platform. Really exciting, so we train a model to predict the promoter activity from DNA sequences, then scan 0.75 million enhancer sequence and try and predict the best location of those enhancer sequence relative to certain promoters. And when we test, we're looking for the highest predicted strength here. And when we test for the highest predicted strength, this is our baseline. We see improvements in our in vitro expression driven by these predicted promoters. We can also use genetic algorithms to evolve these promoters. So this is a number of generations, and this is the predicted expression, right? So we evolve the promoters. And here, we have supposedly better expression. When we take those promoters in vitro, what I'm showing you here is here, if the baseline promoter is down here, here are the promoter improvements from the other methodologies I've mentioned. And here, a further improvements using this in silico evolution. We can also separately mutate individual bases in silico and see improvements with individual mutations in our promoters. And now we're combining these single mutations to see if we can further improve our promoters. So these various platforms give us a large library of promoter enhancer regulatory elements that allow us to quite specifically control the transcription of our genes to allow us to really address where they're expressed and to make sure when they're expressed that at potent levels that are efficacious and safe. So first example is our search for smaller and stronger promoters than CAG. So CAG is a commonly used ubiquitous promoter, (chicken beta actin CMV enhancer. It's used in Zolgensma and Luxturna, but it's large and it's ubiquitous, so there's no cell specificity. We using the technology that I've mentioned earlier, we've got over 40 constituent promoters that are tenfold or more stronger than CAG. And some of these are self-selective, so they are more highly expressed in some cell types than others. And that's indicated here. These are the different cell types, and this is a heat map showing the different promoters. This marks the sizes of the promoters, we're trying to get them smaller. You can see strong promoters in particular tissues over others. And here, these are CAG variants. And what I'm showing you is in comparison to CAG in 293 cells, CAG is here, CBh is very similar to CAG, we're able to really improve the expression from these smaller changed CAG promoters to be several fold higher than CAG and stronger even than CMV in 293s. Then we have tissue selectivity. So we have a whole library of CNS promoters, 10 promoters that are stronger than CAG in human and mouse neuronal cell lines. I'll bring your attention to CAG in neurons, CMV, CAG, CMV and these neuronal promoters that are used in the clinic in some cases. And here, you can see stronger than CAG, we have promoters that are significantly stronger than even CMV. Likewise, in the liver, these are liver-specific promoters currently used in some liver-specific programs in the clinic. That's CAG. And here are our liver promoters significantly stronger than the CAG that has been used extensively. And just to show you the selectivity of these promoters, this is 203, so kidney cell, we'll call it expression versus liver cell expression. And you can see significantly shifted towards liver expression. And now muscle. So we've developed a library of small and selective strong muscle promoters shown here. 9 muscle promoters that are stronger than tMK and equivalent to or stronger than CAG in muscle. tMCK is currently used in some muscle gene therapy programs in the clinic. So here, we've got a muscle-specific promoter that's 17 fold higher than tMCK in mouse muscle. This is over time in vivo, in a living mouse, you can watch as the gene therapy switches on. And here, we have the muscle-specific promoter, tMCK and here is our novel, small engineered promoter. And you can see even here at 35 days that hasn't fully switched on, you've got a very big differential from the regular, I'll call it, muscle-specific promoter. And here, this is just to show you, we have some very small strong ubiquitous promoters that work very well in muscle here in comparison to CAG and CMV, you've got these unusually strong promoters that activate expression in muscle. This is a leg muscle of a mouse that's been sectioned. This is expressing CAG. And you can see here that driven by one of our new strong muscle expressing promoters, you've increased expression. So what I've shown you is the technology that has allowed us to build libraries of synthetic, novel promoters, smaller size, greater strength cell selectivity comparable to CAG, CMV, et cetera. Many of these, we've got libraries of many of these for different cell types for the eye, for other tissues, 40 constitutive promoters that are up to tenfold stronger than CAG. 10 neuronal promoters, 12x stronger than CAG, 13 liver-specific promoters up to 4x stronger than CAG, 9 muscle-specific promoters, some of which are very small, which is stronger than tMCK and synthetic muscle-specific promoters that are durable and strong. And we have in silico screening that allows us to further evolve and optimize any of these promoters. So we have a very deep toolkit for controlling our gene therapy vectors in a way that allows us when they're on to be expressed at high, potent and safe levels. Which brings me on now to our gene regulation platform. So gene regulation has been studied for many years. In order to be able to switch on your gene therapies with a pill, an orally delivered pill. So you would deliver a gene therapy, install into the body, a little factory for making whatever therapeutic it is, and that is only on when your oral small molecule is delivered. And there are various different ways that gene regulation systems have been built, one of which is to try and regulate via small molecule responsive transcription factors. Now I've just described to you the importance of promoter and enhancer and the transcriptional regulation that allows you to have gene therapy vectors that express in a potent and safe fashion. One of the big problems with trying to regulate with a small molecule that transcription is that you lose all of that promoter specificity that we have engineered. So what tends to happen when you regulate transcription with a small molecule, you lose that cell specificity and you lose that strength and rather, you get low level expression at low dynamic range everywhere in response to a small molecule, which is obviously not feasible. Likewise, there are methods that use transgenic proteins to regulate transcription factors, which are transgenic, for example, but those have the problem of being immunogenic. So we decided to rather than use transcription and protein-based methods for regulating our genes, we would use RNA shape. And we have developed an RNA shape-based regulation system that is overlaying on top of the better optimization and promoter enhancer regulatory element control that I've just mentioned. This is not just an on/off switch. It's an entire system for regulating gene therapies in a dose-responsive fashion to oral drugs. Very importantly, we have unprecedented high dynamic range. What does that mean? Previous switches tend to have low dynamic range. The dynamic range is the difference between on and off. And previous switches that you'll see published, whether transcriptional or with riboswitch tend to have dynamic ranges around 2- or threefold, maybe tenfold, sometimes a little bit more. A twofold dynamic range means that when the gene is switched on with the small molecule, it's at a low level. But when it's switched off, it's only 50% off. So those sorts of dynamic ranges are not useful therapeutically. And we thought, only when you have a very high dynamic range, can you then get a really granular dose response to your small molecule. And that's one of the very important features of our switching technology, our gene regulation technology that gives us that incredibly granular control of gene expression of the therapeutic. So this is not just one switch. This is a platform, which allows us to control the expression of any gene in any vector, it can be applied to cell therapy, and I'll show you how we've used it to control gene editing. So we could optimize many cassettes for every gene. And in addition, I'll show you that we have libraries of small molecules such that we can design synthetic aptamers to new small molecules, and we can then regulate each gene with its own small molecules. We can then choose the PK and drug characteristics of a molecule that we want for each particular indication. For example, within the blood-brain barrier, if we want a gene to be activated there, we have a blood-brain barrier penetrants for molecule. So what I'm going to talk about is not just a switch. It's an entire technology that transforms genetic medicine and the delivery of biologic by allowing us to totally control the activity of a gene therapy using the dose of an oral small molecule. Here are some of the potential applications that we could apply this technology to. First of all, obviously, we can vectorize biologics, we can vectorize antibodies, we can obviously replace genes. And we can do -- we can regulate any of those. In particular, we have regulated PCSK9 antibodies, anti-IL-17 antibodies, and this allows you to potentially dose gene therapies or antibodies precisely with a small molecule and for gene therapies. I'll show you that it may improve the safety and consistency of dosing between patients. If you are able to regulate those gene therapies with a small molecule, not just by the dose of the AAV. There's potential to use this technology for passive and active vaccines. So for an active vaccine, for example, we could take our messenger RNA for the antigen, the sequence for that, and we could potentially have multiple multivalent, different messages for different spike proteins in COVID, for example. And we can inject that and boost the antigen time after time after a first IM delivery as needed with an oral pill. Similarly, for passive vaccines, we could put into our AAV, the sequences for multiple neutralizing antibodies, not just one. And then activate those systemically whenever needed with an oral pill. We can do the -- we could use -- do this sort of passive or active vaccine for COVID-19 or flu or even universal HIV. One of the problems that an intractable problem in biology today is how do you get biologics across the blood-brain barrier. It's one of the biggest challenges in oncology and neurodegenerative disease. We can install into the brain with one injection, the gene for a particular therapeutic, for example, Herceptin and then activate that gene with a blood-brain barrier penetrant small molecule. Likewise, for an antibody to amyloid, if you so wished, we could put that antibody within the blood-brain barrier and then activate it with a small molecule that crosses the blood-brain barrier. I'm going to tell you today about how we can address metabolic disease using regulated therapeutic hormones and peptides. These are naturally short-lived, and they're difficult to develop into long-acting drugs and dose effectively and appropriately. In addition to dosing, now with an oral pill, we can combine multiple natural peptides and regulate them together. So for the first time, we can deliver, for example, combinations of GLP-1, PYY, GIP in one vector with an oral pill. And of course, we can regulate insulin so that rather than injecting insulin, you can have a dose response to a small -- to incident to a small molecule delivered orally. In the eye, which is where we have a lot of expertise in vectorology and in clinical development, you can tightly control therapeutic protein expression in the eye, and we're doing that with eye drops. We're currently putting our small molecule into eye drop formulation so that next year, we will actually be able to regulate a VEGF inhibitor or a complement inhibitor in the eye with a daily eye drop. I'll show you how we regulate gene editing and just to point out that this is not only for AAV, we can regulate in multiple viruses, we can use this in cell therapy, we can have kill switches, on switches, off switches, and finally, down here, this changes the paradigm of pricing in gene therapy. Now you can deliver the gene therapy and then you pay for the pills that activated. So what is our gene regulation system? As I mentioned earlier, promoters, transcriptional regulatory elements are critical for potency, efficacy and safety of your gene therapy. So we use RNA shape to create the switch, which allows us to regulate the expression of the protein encoded by the therapeutic transgene. So what we did -- this is our therapeutic transgene, unregulated, and you'll see lots of controls in our experiments. And when it's unregulated, that's what it is. What we do is we put a sequence of RNA into the middle of -- into a position in the transgenic sequence DNA sequence. And in that cassette, 6, an RNA binding sequence called an aptamer that binds to a specific small molecule. When that small molecule is absent, this RNA folds into a shape, it's from this transcript, folds into a shape, which results in the entire transcript degrading and no protein is expressed, nothing. There's no fragments of protein. No, it's just completely degraded. However, when a small molecule binds to that site in our cassette, this entire cassette is spliced out. This message is processed normally with splicing events, which will move that cassette and the messenger RNA is formed that is stable and this message is identical to the message from the unregulated transparent gene. And you have protein expressed at high levels. The use of this aptamer binding the small molecule as the RNA sequence involves what we call a riboswitch. So what is a riboswitch? Riboswitches are used throughout the bacterial world. They're frequently used, and this is a riboswitch in bacteria, responding to a metabolite SAM. So they're used extensively and what they are is their pieces of RNA or sequences of RNA that in one configuration allow a function, but when the small molecule is bound, as here, it changes the configuration of the RNA and changes the function occludes or activates a particular RNA function. So what actually goes into a riboswitch? Riboswitch involves an aptamer, which is the region that binds a small molecule and something we call an expression platform. The expression platform is the region that determines a function. And the function -- I'm going to focus on is splicing today. Now, many people have spent many years, the last 30 years trying to take these riboswitches from bacteria and move them into mammalian systems to regulate mammalian genes. However, there are no mammalian riboswitches. And when people have tried and groups have tried to regulate eukaryotic or mammalian genes in mammalian cells, these riboswitches haven't worked very well. They tend to have low dynamic range, as I mentioned earlier for transcriptional switches. So low expression, but particularly, they don't turn off completely. So you tend to have 2, threefold dynamic range, which means that when genes regulated by these riboswitches are off, they're on 30%, 20%, 50%. So rather than take bacterial riboswitches and try and make them regulate mammalian genes, what we did at Meira was think of what a riboswitch is and designed base by base, a mammalian riboswitch in an expression system, which is a splicing cassette that allows us to have a really sharp binary responses for molecule, leading to a clean and high dynamic range, on/off, and that's what I'll show you here. So that cassette, I mentioned earlier, and showed you in green is this region. It is an intron/exon/intron cassette between the 2 exons of your transgene CDNA. So when we were thinking about what did we want to make a riboswitch, we decided that something really simple would be a hairpin that's shown here. What we wanted to do is have a hairpin that is stabilized, as shown here, when you have a small molecule, but without the small molecule is open. So when the small molecule is bound, the hairpin forms. How do we turn that into a riboswitch? We put a functional region of message RNA here, such that when the hairpin forms shown in right here, it's occluded by that hairpin. And what we decided to use is the functional sequence is a splice site. So the 5x splice site of this alternative exon/intron boundary. So that when there's no small molecule and hairpin is open, this is the splicing series of events that occurs. And this alternative exon is included in the transcript. When the small molecule is bound and splicing is blocked, this splicing event occurs, the entire cassette is spliced out, and you get a perfect message, which is the same as a message that comes from the transgene without a regulatory cassette in it and protein is expressed. How do we do that? In this alternative exon, which is spliced in the OC state when no small molecule is present, we have a stop code on. And what that means is that this message is immediately degraded by nonsense mediated decay. It never gets to a state where it is translated messenger RNA. So alternative exon included, entire message degraded, alternative exon and the entire cassette, spliced out, perfect message and protein is expressed. So how did we build this splicing riboswitch? If you consider a hairpin, the length of the hairpin determines how easy it is to stabilize. If the hairpin is long, it becomes stable even without a small molecule. And that means that the gene is always going to be on. It's going to behave as if the small molecule is there because the happen is stable. However, you can delete base by base, until you come to a situation where the hairpin is so short, but is always open, and the gene is never on. So we did deletion and sequencing of hairpins, and this is what we found. So this is the gene, cone one, so it has no splicing cassette in it. This is luciferase cell margin, this is luciferase. We put the splicing cassette with different hairpin length in it. When they happen is long, you have the gene on even without the small molecule shown in red. As you reduce the hairpin size, you eventually come to a state where you only really get the hairpin stabilizing when you have the small molecule. So here, you have -- this is a fairly loose switch. It's easy to turn on, and it's sometimes on when you don't have the small molecule, only it has background, 24 dynamic range. Take another base away, 60 fold, another base 800 fold, Here, G17 baseline over 2,000 fold dynamic range, really, really low background, high expression. When you take away more and you've got this situation, you can't form a hairpin and you can't express the gene. So this is how we developed our riboswitch. And in understanding this, you can understand that we have switches with a range of dynamic ranges. And here we could optimize further. So this is the regulation cassette, the hairpin, the intermittent effects on the intron sequences and this particular cassette had a 1,500 fold dynamic range in full length. And we can mutate and delete sequences within these interims, as an example and change the dynamic range, double it, we can even get to almost 5,000 fold dynamic range in this particular case. So we can do optimization of these sequences to increase dynamic range, decrease background and keep expression high. And here's an example of another deletion series. And what I want you to look at here is the types of sequences, the types of constructs we're interested in. So here, our sequence is that of constructs that express at around 100%, just under 100% of our unregulated construct. And when you go over here, so 70 to 90, when you go over here, the dynamic ranges are in the several thousand fold. Here are really easy to switch on switches. So these are somewhat looser switches, okay? They have high expression, over 100% of the nonregulated construct. But look, the dynamic range is low, okay? They're not off completely. Now these may be useful in some cases, if you want very high expression of an antibody maybe and low level background is not important. And then there are these sorts of switches, where they're only 30% on, but look how high their dynamic ranges, 5,000 fold here. And I'll show you ways that we can further double this. And these may be important when you need extremely tight expression. And remember, this is 32% of the endogenous expression. So if we are able to have really strong promoters and potency, which allows us to have therapeutic activity with 30%, we're able to regulate therapeutic levels to very, very high dynamic ranges. And here's some examples of -- in cells, the types of things that we're looking at. So high dynamic range means that when gene is off, expression is low, when it's on, expression is high. And I've said, we've got different switches with different dynamic ranges. So on the top, these are no inducer on the bottom inducer. This is the unregulated gene. Next is G17, remember that hairpin that regulated 2,000 fold dynamic range. And here, no background, but it's only about 60% of the unregulated expression. Here, G15, higher dynamic range -- sorry, lower dynamic range because there's actually background here, but you've got high expression. Now if you put them together, you maintain the high expression of G17, you move the background and you go back to high dynamic range, another way of controlling your dynamic range. Here is just an example of switching on in old cells, how quickly it occurs. So here in a cell sorter, we give us some molecule, no GFP, 6 hours later, you've got GFP expression in almost all cells. Here again, a time frame of induction, remove the inducer and your gene therapy is switched off. Now everything I've said so far is about our switching base cassette. And that's what I'm going to focus on when we talk about the in vivo experiments that we've done. But we have also developed switches with multiple RNA functions. So we've got switches of aptamer modulated RNase P cleavage that the 3-prime end, twister ribozyme we've used to regulate. Aptamer-modulated U1 interference of the 3-prime end and aptamer-modulated polyadenylation. So we've made on switches, we've made off switches. Here are 2 examples of an on switch, RNase P, here causes degradation when you have it at the 3 prime end in the absence of a small molecule. You change the configuration by adding the molecule that finds the aptamer, you don't have the degradation, and you have protein stabilized. Here's an example of twister ribozyme that's used as an off switch because ribozyme makes these cuts. So we have an off switch. Here, it's in the on configuration. We add a small molecule, we get this break and you have no protein produced. So we do have these switches. The dynamic range is in the tens to hundreds. And so it's not as individual switches as strong as our splicing cassette. But like you can join spicing cassettes together, we can take different aptamer-based configurations and riboswitches and join them together. And this is an example of the 3 prime RNase P and what you can see, this is G15 or loose switch. If we add the RNase P to it, we double its dynamic range. We can now go to G17, which is already starting at the high dynamic range. We add the 3 prime regulation, you double dynamic range. So we have a large toolkit, where we can regulate with different switches, combinations of switches, we can actually have switches that respond to different small molecules. And so we've got a large toolkit that allows us to really precisely control gene expression of our gene therapies. Now what I've shown you is that we've got a rationally designed gene regulation cassette that, for the first time, created a mammalian riboswitch. We did this by creating a riboswitch that drives Hep information and therefore, correct transcript splicing. When there's no inducer, RNA is degraded, when the inducer is present, stable RNA is produced and proteins expressed, and we have an unprecedented dynamic range of 5,000 fold. All of this was invented at Meira, and we developed iteratively over the last 6 years to have this platform technology today, which is broadly covered by quite an extensive IP portfolio. Now I've described the cassette and the riboswitch, but one and the importance of dynamic range for dosing. But that dynamic range also was important for another reason. It has allowed us, for the first time, to screen for small molecules that bind aptamers in the context of mammalian cell. So how do we do this? So first of all, what we initially show, this is with our G17, is that the aptamer part of our riboswitch within the cassette is interchangeable. So we can use these natural aptamers warning and these are a number of different natural guanine aptamers and you can regulate the switch, right? This is the dynamic range of the switch switching on. Adenine aptamers, here and this is the unregulated. You can put adenine aptamers into that aptamer region of the riboswitch and you can get regulation with adenine aptamers. TPP aptamers. Basically, what this is showing is that you can change the aptamer within your cassette, our cassette, and it will respond to different small molecules. So what that means is that we can take that up to a region, and we can mutate it, and we can randomize it, and we can make any aptamer we like, and we can then screen libraries of aptamers within our riboswitch, within our cassette and screen for small molecules that activate gene expression of our cassette within a mammalian cell. So there's functional specific binding of the small molecule to the aptamer in the context of the mammalian cell. Why this is important is, up until now, up until this high dynamic range cassette that we're able to use for screening, the way that you get small molecules to bind RNA is a methodology called SILEX. And what SILEX is, is you put small molecules onto a column and run randomized RNA down that column and then you take out the RNA that sticks to a small molecule. The great problem with that is that when you take those RNAs that have bound to the small molecule of the column and you put them in the context of a mammalian cell and want them to have a specific function there, they tend not to work because RNA shape on a column is very different from RNA shape in a transcript in a mammalian cell. So here's what we do for screening. There's our cassette, exon one exon 2 of a marker gene. Here we go. So we can -- here's the aptamer region that binds its small molecule. We can take many different aptamers and put them just into that region. We can mutate around that region. And we have created large libraries of many, many different aptamers, different aptamer classes with different aptamer mutations. So randomized sequences, natural sequences and site directed mutogenesis. Then we can also screen particular aptamers or libraries of aptamers with libraries of small molecules. So we've screened non-compound libraries, novel fragment based libraries. And when we have active compounds that bind to certain aptamers, we can then modify those compounds to create new compounds with specific desired properties like crossing the blood-brain barrier, for example, and then re-screen those to find new compounds that specifically bind to certain aptamers. Importantly here is only when the small molecule binds to the new aptamer and drives Hep information and drives this splicing regulation, do we get the gene switched on. So this screening method allows us to select functional small molecule aptamer binding. Here is some of the data to show you how we get these small molecule aptamer pairs, and we involve them to be highly potent and highly specific, both small molecule and riboswitch. So example here of -- we have riboswitch here, a particular aptamer, and we screened a compound library. And what we find, these are different small molecules that activated the particular aptamer that we're screening. Here you go. And we get a series of small molecules that come out of this screen, and all of these show strong activation of the cassette by binding to the aptamer and Hep information, okay, one library. Another library, and this is actually a library that I'm going to show you in vivo data from. Again, here we go, we find switch should be may screen many thousands of molecules. And out of that, we -- how many we got some, which drive aptamer binding to the small molecule and gene expression. We can then -- with small molecules that work, we can then change the riboswitch library. So within the aptamer region, we make point mutations or randomized in the small molecule binding site, and we do that iteratively. So this is an example of a first-generation aptamer. This concentration, quite low dynamic range. It would probably require much more to get to that high expression. We then mutate and we get a slightly higher dynamic range in expression, we mutate again, and we get a much higher dynamic range. Importantly, it's not just a dynamic range because that's inherent in the switch, but it is the dose at which you see this high expression, and we can evolve. Here's another example, and these are a much higher dynamic range switch, okay? This again, it's from this library of small molecule. This is unregulated. This is the baseline. First generation -- sorry, first generation, second generation and further optimized switches. So with this screening technology, we're able to create libraries of very specific potent small molecules that bind exclusively to their own aptamer sequences that we can transfer into our regulation cassettes. And this is an example of this is a compound that I'm going to be talking to you today about. And we have made derivatives of this compound and screen that library, we're looking for desired tissue distribution, blood-brain penetrance, PK, stability, feasibility, those are the sorts of things we're looking for, and we currently have a quite large number of molecules is all different compounds that each regulate in a very dose responsive and high dynamic range fashion. Out of the library that we created from compound A of 30, just to give you some numbers of 73, 57, had 50 fold dynamic range 51, 338, over 1,015 had over 2,000 fold dynamic range. The aptamers are then interestedly evolved to make them more specific and potent. And each compound is then goes through full ADME, PK, tops metabolic profiling and PK in -- sorry, pharmacokinetic profile in mice, rats, dogs and NHPs, our first compound is currently completing long-term NHP dosing and will be in the clinic in a Phase I study next year with this entire pipeline of molecules coming behind it. So having described the switch and the -- having described the switch, the aptamer and the small molecule development methodology, now I'm going to show you data in a number of different targets and in vivo. So this is just to show you to start off with how dose responsive our switch control is. So I say we have precise control -- sorry, precise control of transgene expression with single doses of orally delivered small molecules. What I want you to see here on this set of panels here is 3 things. First of all, from left to right, each of these is an increasing dose of AAV, 5E10, 1E11, 2E11. So here, within each panel, there are 3 types of constructs. The blue is the unregulated optimized transgene that viral vector. The pink is a tight switch. You can see very, very low background here. The dark blue is a looser switch. There is some level of background. This is liver delivered AAV by tail vein injection. So what you see is constitutive expression from the unregulated optimized construct, and then you give a single dose of the small molecule, baam, you switch on the gene and then it switches off. You give a higher dose, baam, you switch on more, 100, baam, higher, 300, baam, higher and the switching happens in low dynamic range and high dynamic range to the same extent. You increase the viral dose, you get the same dose response, and you can quite readily get to levels of the constitutive -- the unregulated construct using viral dose and small molecule dose. This is just with a single dose of this gene, you see get switched on. Now here is everyone's favorite example of in vivo expression. And what I'm showing you here is the regulation of EPO using an oral small molecule in a dose responsive fashion in vivo en masse. So in this case, the AAV is going down in its dosing. So 1E11 here 5E10, 1E10, 5E9 and these are -- this is just PBS. So what you can see is that every viral dose you have, a clear, high dynamic range, dose response to your small molecule, 300, 100, 30, 0, and that is recapitulated whatever your viral dose is. Now what this -- what I want to show you here is how tight this regulation by the small molecule is, okay? So what I'm showing you in the next slide is a blow up of this peak. Yes. So what you can see is in that cohort of mice that had tail vein injections, all of the exact amount of AAV, there's some spread in the amount of our gene product that they're producing Luciferase in this case. But this is the unregulated -- this is the regulated construct, very low levels here. You give a single oral dose, baam all the mice express the same amount at the same time. Using this rider switch overlay of small molecule regulation, we're able to restrict variability in Expression and Re and give very consistent dosing, single dose. This is a luciferase dying off over time. Again, slightly higher dose of AAV. And look what happens. Very consistent dose response to this small molecule. Now this is an even better illustration of how tightly our regulated constructs respond to the dose and PK of a small molecule. I just want to remind -- I just want to remind you what the liver delivered constructs look like really sharp peak and baam, it comes down again. This is exactly the same dose of virus but injected into the muscle. And look, it doesn't go to baam up and down. It spreads out, right? The loose construct, the high construct. And then we look at the PK. And this is the blue line is the distribution of the small molecule in liver goes in. This is oral. It's taken orally, and then it goes down rapidly. But look what happens with muscle, it goes in, but it accumulates, and then it goes down. And so when you look at the gene activation in muscle, it activates and it goes down slowly. So it seems that our gene regulation cassette is allowing us to really dose the gene therapies in a very, very granular fashion. I mentioned earlier all the different types of applications for these gene regulated gene therapies. And what I'm going to show you now, these are the constructs that we have currently built in-house and that we have in our fridge that are vectorized, optimized and regulated with cassettes that we've optimized for particular expression. So vectorized, optimize and regulated construct. We have a PCSK9 antibody that we can regulate and switch on with an oral small molecule, a VEGFR2 antibody, which we will be regulating in the eye with eye drops, anti-amyloid, antibody IL-17 antibody, PD-1 antibody and HER2 antibody. All of those we have made, and I'll show you examples. Therapeutic Hormones and Cytokines, all of these we have made and regulated many of them showing in vivo activity. EPO, human growth hormone, PTH, insulin, we can regulate tightly with a small molecule, GLP-1. And combinations of gut peptides, GLP-1, GIP and even including PYY. And then I will show you that we can regulate new cases, very tight knockdown of RNA using regulated CasRx. So first, I'll go to antibodies. These are each antibodies that have been vectorized and optimized for high expression and the cassette has been inserted into that antibody in the optimal position for the highest dynamic range expression. We see very clear dose response activity, anti-PCSK9, anti-PD-1, anti-VEGFR2, anti-HER2, anti-amyloid, anti-IL17. I show you this graph really not that we're developing every one of these, but to show you that we have the tools to dose responsibly activate any antibody and vectorize and optimize it. Here is the PCSK9 in vivo. It seems and throughout what we do, the in vitro dose response is very closely correlated with what we see in vivo. So anti PCSK9 antibody here is the clear dose response, very tight regulation. And here is a mouse experiment. The antibody gene is injected into the mouse, IM, not liver. These are all IM. And then we add the inducer. So this is now daily delivery of the inducer, so not just one dose of the inducer. And you can see at each day, there is a dose response. 30, 100, 300. Day 8, as you have more antibody accumulating gets higher. Day 15, it gets higher. Day 21, it's starting to tail off to a level which is probably ongoing with daily dosing, another antibody, very, very high expressing antibody. Dose response in vitro, and this is only going up to 9 days. But here, very high expression in a completely dose responsive fashion. This is our PD-1 antibody. Next, I'm going to just mention because it's incredibly tight and really amazing are regulation of Nucleases. Nucleases, you don't necessarily want them all the time. We can regulate Cas9, it's quite big, but we have and can regulate it, and the data I'm going to show you here is the regulation of CasRx. So this destroys knocks out RNA. So we have the CasRx cDNA sequence with our cassette in it. And here, we have a promoter driving a guide RNA that targets an important metabolic gene, okay? The unregulated CasRx here with the guide, has almost 100% knockdown of the RNA. Here, in green, pink and blue, these are different tightnesses of switches regulating the CasRx. So 3 different dynamic range switches, the tightest in green, middle and a little bit looser here. So what you can really clearly see that there is a very clear dose response to the small molecule with increasing knockdown of your RNA, as you add small molecule, increased dose a small molecule until you get to the dose where you have 100% knockdown. So using CasRx, we can precisely regulate the amount of RNA that is degraded using the guide RNA. I'm not going to go through that. Okay. Now, I'm going to move on to therapeutic hormones and peptides. And this is really -- this is really interesting because the world of therapeutic peptides and hormones is really involved in body homeostasis in energy homeostasis in things in metabolism and exercise its entirety and aging. And many -- while we do have many hormones and peptides that have been turned into drugs, one of the issues with these small proteins is that they're ephemeral. They disappear quickly. And what our gene regulation technology allows us to do is to actually dose those quick, disappearing peptides and hormones in a physiologically effective fashion because we no longer have to inject a PYY that's going to disappear in 2 minutes. We can activate with a small molecule, PYY, when we want it and to the levels we want it, for example. So here are examples of some of the peptides and hormones that we've regulated, and I'll show you in vivo data. EPO, human growth hormone, PTH, PYY, GLP-1 and insulin. Insulin, I just want to say that again. We can potentially activate insulin with a pill to exactly the right dose when you need it. So we have really tight regulation of these hormones and peptides, and that's demonstrated here. So this is the EPO optimized with the switch in the optimal position to give the high dynamic range. And here, you see good dose response, high dynamic range. Again, this is in different cells. So in HEK, C2C12 cells. So these switches work in multiple different tissues and multiple different cell types. Back to the EPO slide. Here, you have recapitulating in vitro, this is with the multiple doses of AAV at every dose, you get that dynamic range dose response coming up, in this mouse with IM delivery. Here, I've got a lower dose of EPO, and this is just showing you the -- not only the serum levels of EPO, but it's a dynamic range within the level of -- within the dose that we're using. So we're getting like 900 fold from the baseline. Now you look at efficacy. On the left-hand side, these are all mice were 2.5E10, okay? So they're all this dose. And these are anemic mice. So this is the normal hematocrit of the normal mouse cohort. And these are anemic mice, these have PBS. These have no -- sorry, these have -- are regulated on a regulated construct with no drug, and these have drug, but no construct. These 2 down here. This is the anemic mice. When you add a low dose of our small molecule, you get hematocrit restored partially over time. You increase the dose, you get a dose responsive increase in hematocrit, as you see a dose responsive increase in EPO. In this particular case, we didn't treat with 300 which would have gone potentially to bear. But we did another experiment here in another experiment to E10 back to genome, so the same as this. But this time, here are the baseline anemia. But this time, we treated with 300, look what happened, Hematocrit goes way over the normal rescue. If you use a lower viral dose with 300, you get quite rapid rescue. All this to say is that we can really precisely through viral dose and small molecule dose, determine the amount of EPO that is produced and the exact Hematocrit results. Another example, PTH, PTH, optimized construct, cassette, tight dynamic range in vitro. Likewise, in vivo. Remember, it's really hard to detect PTH. But here, you have low dose, 100, 300, you've got that good dynamic range like in vitro. And then look at serum calcium. This is just 3 oral doses, okay? 3 days. You already at 100, you see increased calcium, 300 increased calcium. So we're getting physiological changes in response to oral doses of small molecules, and these are dose responsive physiological changes. Insulin. Insulin is really hard to vectorize. So we've spent a lot of time vectorizing, and we had a really strongly expressing insulin. Okay? We can now regulate that incident. We've optimized the cassette. And here again is the dose response of insulin expression in vitro. This is the constitutive. So when we switch it on with the small molecule, it's a little bit higher than the baseline expression of insulin. Now I'm moving on staying in the diabetes area to gut peptides. And I mentioned how, in general, this regulation system is useful for peptides and hormones, but combinations of peptide is really, really important and really interesting. So when, as you know, GLP-1 is one of the largest classes of drugs used for diabetes, their outcome study is showing that it increases survival, and it's recently been approved for obesity. And the gut peptides are very powerful regulators of metabolism of sugar and fat and satiety and a multitude of other important homeostatic mechanisms in the body. However, they work in combination with one another, and the timing of those combinations is important. So for many years, many groups have tried to make GLP-1, GIP coagonist, GLP-1, glucagon coagonist. And I remember in the early 2000s being at a diabetes meeting where Amylin was presenting data on exenatide before it was approved. And in that meeting, they also presented data on a combination of gut peptides, which included, I think, GLP-1, amylin and PYY. And it was the most extraordinary regulation of obesity that anyone has seen. And the question was, well, why don't you develop that? And the problem was, you needed to infuse constantly because these combinations of gut peptides are so ephemeral that you can't get therapeutic levels at the right time at the right place, so you need constant infusion. Meanwhile, lap banding is known to change the profile of these therapeutic peptides. And we know the types of combinations that have the most beneficial effects on different metabolic diseases. What are gene regulation platform allows us to now do is not just regulate an individual construct and deliver it with a small molecule. It allows us to, at the same time, in one construct at these combinations. So we can regulate blood sugar levels, using a construct that switches on GLP-1 and GIP, natural forms of them. We have constructs that include PYY, one of the most potent metallic peptides. And this is really interesting because we're looking at putting these genes into Salivary Gland to secrete into the serum where they're able to do their job in regulating all those metabolic functions. So here is GLP-1, again, very hard to vectorize. These are small. And so we have spent a lot of time actually building constructs that have very high GLP-1 expression. Here are some of them. And this really good one here. And like our other peptides, really tight dose response when you put in our regulation cassette. This shows you that when we have GLP-1 plus GIP, dose response, dose response, type regulation. And this is actually showing you the activity of GLP-1 versus -- against its receptors here, the activity of GIP, when there's only GLP-1 present as in here, one copy very little, 3 copies, a bit more activity, GLP-1, GIP. You've got higher expression, GIP, you've got no expression. Likewise, you look at GIP activity, when there's only GLP, no expression. When there's GLP-1, GIP, you have GIP when there's GIP alone versus GIP. And likewise, this just shows you, you can regulate PYY to levels higher than constitutive. And here is in vivo example. So this is a GLP-1 GIP, IM delivery, 2E11 in mice, and these are mice that have been given a bonus of glucose. In red, you can see mice that are normal, they have no -- sorry, they're not normal. They've got this construct in them. So these are mice that contain the construct, but there's no inducer, right? However, when you give the inducer that was given daily for 3 days prior to this day when they were given glucose, look what happened, give the glucose and you immediately drop the glucose level, much faster than the rather slow tail reduction of the mouse with no inducer. So finally, so I've shown you regulation of a broad range of different therapeutic areas. Where is our expertise at Meira? And where do we really focus in this regulatory space. So obviously, the eye we have a huge amount of expertise in the eye with respect to vectorology with respect to our clinical development. So we have a deep vectorology toolkit with organoid technology, promoter technology, capsid engineering is currently ongoing in NHPs to create capsids that effectively transduce cells at the front of the eye, which can secrete therapeutic molecules into the vitreous. We will be putting our small molecules into ophthalmological formulations next year so that we can actually activate these genes that are delivered closer to the front of the eye, not needing subretinal injection using topical delivered eye drops. And what sort of indications are we addressing? We have a regulated construct for VEGFR2 now for wet AMD. We are working on anti-complement, our strategies for dry AMD. In uveitis, we have a program where we regulate cytokine inhibitors. And in glaucoma, we're taking 2 approaches to regulate water flow, actually using aquaporin knockdown and fibrosis. So we envisage using our gene regulation technology to deliver therapeutics to the eye using our small molecules formulated in the eye drop that allow us to treat these large eye indications with eye drops rather than with multiple injections. And also to be able to use these sorts of targets, which you can't just put in with a gene therapy and not regulate because of the safety issues of cytokine inhibition, complement inhibition and changes in water flow. Now it won't have escaped your notice that we have a program in the clinic, in the salivary gland for xerostomia. So we have a lot of experience, not only in the clinical development of salivary gland gene therapy, but also in the vectorology that allows you to deliver durable, strong expressing genes to the salivary gland. The parotid is one of the largest secretory organs in the body, and it's been shown that with specific signal peptides, you get secretion of peptides for the parotid into the serum rather than saliva. This has been shown in pigs and mice for EPO PTH and human growth hormone. So we have experience in vectorology and clinical administration. Its small doses locally delivered, durable expression and worse comes to worse, if something goes wrong, you can always remove a parotid. Now we've -- I've shown you targets for salivary gland regulation, the gut peptides. You can also treat genetic endocrine deficiency disorders. PTH hyperparathyroidism is an unmet need because of the short life of natural PTH 1 to 34, and the treatments give you hypocalciuria and impaired renal function and renal failure. You could potentially use this same PTH for osteoporosis, growth hormone deficiency, EPO and, of course, insulin. Currently, we are taking our first small molecule that looks really safe through IND-enabling studies to have an IND next year. And by the way, this small molecule, when we do, by distribution, has good exposure in the salivary gland. So we see developing the salivary gland as a secretory organ from which to secrete regulated gene therapy delivered therapeutics as an entire platform to address some of these disorders. So I've shown you that for the first time, we've created an entire system for regulating genes using small molecules, not just on and off, but specific granular regulation based on the dose of orally given small molecules. We regulate to really high dynamic range, which allows this precise, granular dose responsive activation. We can regulate any gene. It's not vector dependent. We can deliver in lenti, in AAV, naked DNA, we can apply this to cell therapy or gene editing. I've shown you 15 targets which we have built and vectorized. We have the toolkits to rapidly vectorize and optimize and regulate any antibody, any biologic. We have libraries of small molecules with specific aptamers that they bind to potently. So we can potentially regulate different genes with their own small molecule based on the specific aptamer that we put into our cassette. Both regulated viral vectors as well as small molecules will be in IND-enabling studies in the next few months, in 2022. I've shown you that our expertise at Meira is focused on the eye and the salary gland and why both of those tissues are excellent targets for gene regulation. And I think I presented technology, which is truly transformational. We hear that word a lot, but has endless opportunity and really changes the potential of genetic medicine in general. So I just want to open the call up to questions.

[Operator Instructions] So the first question will come from Joshua Schimmer at Evercore.

Just curious whether you can regulate half-life of the protein expression? And if so, how might you do that? Does something like that could be important, say, for expressing insulin around mealtime versus basal?

Yes. So we can -- that's a very important question. So in the libraries of small molecules that we're creating, the duration of expression seems to be directly determined by the exposure of the small molecule. So for example, we could have a small molecule, which was long-acting or durable. And then we would get expression for that period of time. So we're building libraries with different characteristics. Really short acting, high metabolizing compounds, compounds that have a longer tail, greater exposure, flatter PK, compounds across the blood-brain barrier. So you can control the timing and amount of your gene therapy by the PK profile and the exposure profile of your small molecule.

The next question comes from Geulah Livshits at Chardan Capital Markets.

So in terms of the buying distribution and the cat exposure that you get with salivary -- to the salivary program. Can you give us maybe a little bit of color on how about what we can expect in the preclinical setting with respect to delivery with vector in salivary and the kind of exposure that you go to small modeling eventually is towards the end. And I just wanted to understand a little bit more about the translatability of the preclinical data across different systems like you might have seen.

Okay. So we do -- I'll address the small molecule part first and then go on to peptides in the salivary gland. So we've done rodent, dog, NHP, PK and exposure. And so far -- and we do that for all of our small molecules. So far for the most advanced small molecule, it appears safe. And our exposure is really good. As we go up, we obviously are doing IND-enabling experiments to take this into man next year. So we'll only then know the actual exposure in man. But so far, the exposure I showed you in liver and in muscle, and it looks at least as good, if not better, in salivary gland. That's the small molecule that's IND next year. Number two is the translatability into man. So there's a lot of proof-of-concept for signal pack tides that result in secretion of therapeutic levels of hormonal peptides and salivary gland in rodents and in pig models and in the durability of those effects. So PTH growth hormone, EPO and exenatide all when delivered with AAV 2 or AAV 5 into the salivary gland with regular promoters is all academic studies express at levels that are therapeutic for each of those different peptides, even in pig, which is similar to man. Now our constructs unregulated are optimized and expressed at much higher levels than those, I'll call them standard PTH, growth hormone, EPO and Exenatide constructs that have been tested preclinically. In addition, we have the combinations and the combinations are even higher expressing than the baseline expressing GLP-1, for example. So we know that in animal models and particularly pig, which will be a proof-of-concept for us, that we, at levels much lower than ours in unregulated gene therapies, we get really good therapeutic effects. So we think that the therapeutic range of our constructs and the regulatability of that within the range gives us a very, very high likelihood of having extremely efficacious expression levels when we put this into animal models and ultimately people.

The next question will come from Chris Raymond at Piper Sandler.

So just going to -- I know this has come up before, but it's a question on the clinical path. Maybe help explain how do you demonstrate the safety and the right dose for the small molecule and the gene therapy, do you need to do this separately? Or is there a pathway to sort of do this maybe sort of step-wise in a single study? Again, I know this is kind of a common question, but maybe give us a sense of your up-to-date sort of sense of the regulatory environment for that. And would there be differences depending on the target, et cetera, for how you go about tackling this?

So we have live versus small molecules and obviously, libraries of targets. And so what we're doing is for each small molecule. We take that through the tops PK, et cetera, the IND-enabling into a Phase I safety study in man alone. And once we've done that, we can then take it into the clinic into a Phase I/II in combination with the target of choice, which we've done the IND-enabling combination on.

The next questions will come from David Hoang at SMBC Nikko Securities.

Well, thanks so much for this really informative overview. Really, really interesting stuff you guys have here. So I had a couple of questions. One, I think you touched on a little bit about the programs are of interest to Meira, but just given the broad applicability of the platform. I was wondering if there are, I guess, any efforts to maybe develop programs for other partners that may be interested or any active BD efforts to license for potential other applications?

Yes. Well, thank you for that. And you're right. When I said the opportunities are boundless, that really is the case. And we do and are having discussions in particular areas. I would say that in metabolic disease, we're not yet in the size that we could do many, many large diabetes or obesity studies. And we do have discussions ongoing with potential partners. In the area of antibodies or oncology, we also have discussions with partners about delivering antibodies that cross the blood-brain -- or the antibodies are within the blood-brain barrier, like set of HER2 targeting or even those PD-1, PDL1s that I showed you. So yes, we do have interest in that, and we have had a lot of incoming interest actually about regulating everyone's favorite neurodegenerative disease target. Amyloid antibodies, if you so wish other antibodies, if you so wish, putting them into the brain and then activating them with blood-brain barrier penetrant small molecules. So yes, we are always talking to large biotech, farmer, about our technology. And this is really applicable to a lot of different areas. I think in the area of vaccines, that's obviously something that we would partner with and requires partnering in -- for a regulated cover vaccine that was a multivalent, had everyone's messaging it. So it was multivalent or everyone's mutualizing antibody. That requires a lot of partnership, and it's not something that while we are regulating those antibodies and constructs in-house, it's not something that we would necessarily obviously develop a loan, but we have had discussions about those sorts of things and the applicability to other types of vaccine.

I see. I appreciate all the color there, really informative. I had maybe just one more question to sneak in. I was struck by your comment on -- obviously, it's early days, but the potential to maybe change the pricing paradigm for gene therapy. The small molecule presents opportunities for, I think, long-term chronic dosing. So I was just curious as to have you had given any thoughts about kind of how a pricing paradigm could work for this type of Riboswitch technology? Is there a component for sort of the onetime Riboswitch administration and then it looks more like sort of your typical biologic or long-term drug that's given, I guess, chronically, perhaps for the life of the patient.

Yes. Well, I think you answered your question. So yes, being able to regulate your gene therapy, whether it's for an inherited disease or diabetes, you can -- you in-store into the body, the gene therapy. And what I described today is local delivery. So these are quite small doses and locally delivered. So not huge cost of goods. We -- I mean we can manufacture millions and millions and millions of these doses in-house at Meira even now. So cost of goods is low. You can imagine a charge for the procedure to deliver, and then you will charge for pill, which is a lot in injecting an antibody or GLP-1 or whatever it may be.

The next question will come from Gena Wang of Barclays.

Great. Thank you, Zandy, for a great R&D Day. This reminds me of my Rockefeller time Friday Lecture, which is a very exciting scientific discovery. So I have tons of questions on the science part, but I would just limit myself to 3 for now. So the first question is regarding all the promoter data you share, right, the fold increase. So what kind of test do you use to measure those protein levels? So the not in vivo one, I know you used reported gene GFP. But for the in vitro part, what kind of test was?

So it's relative -- so when we -- so these aren't in specific genes, I didn't show you specific gene data. So we have these libraries that are relative expression of reported genes. What I actually took out in the presentation was those -- the performance of those promoters, enhancers and regulatory elements when we use them in optimization. So we have those libraries, which tell us the relative strength of promoters or give us synthetic promoters and small promoters. And then when we're actually optimizing a vector, for example, our Wilson vector or whatever vector it may be, we then compare head-to-head for efficacy with all the different regulatory elements. What I showed you were platforms for mining the world regulatory elements to get ones that were useful tools.

Okay. And those are the western block, I assume.

No, no. So we -- in a cell sorting.

I see. Okay. So it's a cell sorting. Okay. And then my second question is regarding the small molecule specificity. So I know you show a lot of data with the specific regulation, but most of the -- so far, we've seen mostly the reported gene expression. Just wondering if you've done anything, the cell wide, the, say, whole -- all the protein profiles comparison? And then also, or if protein, I know it's very complicated to test the whole -- like a protein like proteinosis the comparison is very complicated. But the other alternative would be mRNA level. Have you do the comparison to make sure those small molecules only going after your specific sequence the helping, where you mentioned that you do have pretty long sequence there and then -- sorry, not very long sequence. Whatever the specific area is relatively short. And we do understand secondary structure is very important, but I'm pretty sure the whole genome will have many other -- the RNA sequence may mimic that particular aptamer that you -- with the Riboswitch?

Right. So these aptamers are very, very specific. And it is not the case that there are load to sequence bind to these molecules. So we're not selecting molecules that bind to multiple aptamers. When we do those screens, right, and we mutate our aptamers, we don't get thousands of aptamers that bind to our molecule. We get 1 or 2. And then we can mutate them again. You change one amino acid. So we do get very high specificity. You change a base and you lose the binding because it's not just binding its finding and functional activity, okay? Then we also do with those small molecules, everything that you just said, not only regular talks and whether it hits our channels and receptors and kinases, but we do RNA seq. So we look at the whole transcriptome. And we see, so far, we are small molecules that we're bringing forward, there's no obvious changes that we've been able to detect.

Okay. So you did do RNA seq to do the whole genome analysis then do the comparison.

You do the whole transcriptome. Yes.

Okay. That's very good. Lastly, I think that, that was asked a little bit earlier also regarding the small molecule, I assume each will have a certain safety window. Just wondering how much -- like you do identify group of the small molecules, how many of these are well understood, the small molecule already there? And for the lead asset, what kind of work you need to do before you go to the clinic to understand better the safety before even do the clinical study to identify the therapeutic window?

So this is an interesting question, Gena, because we often get asked, why don't we use molecules where there's safety data, known molecules. And I was speaking to an adviser from the FDA in recent days, and that is not what FDA wants to see. We developed these molecules, their novel, synthetic molecules designed specifically to bind our particular aptamer sequences. They're not existing in libraries anywhere else. So we choose molecules with exposure of PK and safety parameters such that they are safe within the range that we dose. And remember, these molecules have never been developed to have some effect. So there's a therapeutic window, our -- where there's toxicity and activity. Our therapeutic window just requires activation, right? That's all it has to do. And so there will be side effects or there will be dose-limiting effects, but we do all of the toxicity, PK admit that you do for any new novel small molecule, it's as comprehensive as that.

Okay. So then for the clinical development, how would you -- that adding additional layers, you need to also identify the small molecule and the dose first, safety and the dose? And then with that dose, you move forward and then you need to identify the right dose to regulate the transgene. And then for the gene therapy part or the other component, you also need to identify the right dose, like how -- can you just walk us through what would be the thought process to identify the optimal dose for the combination of both?

So we do our Phase I study in man, which will give us safety and PK. So we know exposure, right? And we know for any vector with the cassette that the in vitro dose response is recapitulated in vivo based on exactly that exposure. So we know once we've done Phase I, the exposure in man, and we immediately can from our preclinical studies, having seen how that exposure results in expression in preclinical models will be able to translate into man. It's not that different from just doing a gene therapy.

The next question comes from Alec Stranahan at Bank of America.

Okay. Great. Zandy, [Technical Difficulty].

I can't really hear you, Alec. Sorry.

Is this better.

Kind of.

Can you hear me okay?

Try, yes.

Okay. Just one question out on [Technical Difficulty]. Can you build out process, formulations from a specific construct you patented [Technical Difficulty], just trying to get a sense of how you're building [indiscernible].

So everything I showed you was completely invented and built at Meira. So we have a broad patent portfolio that covers displacing switch, all the other switches, the screening technology for aptamer, small molecule evolution and selection. And for every indication, we have specific patents on those exact sequences as well as classes of different drugs. So we have a very broad patent portfolio as well as patents on each of the small molecules and classes of small molecules and their interaction with aptamers. Now what I will say is when we started Meira, our intention was to discover a way of controlling gene therapy with small molecules. We haven't yet done that. So we had some ideas. We put those ideas together, and we worked for a year with our patent attorneys to create the data for the strongest possible initial patent. When we did that in our first patent search, we got only As in that patent search, no X and Ys. What an X and Y mean? Is that there's something in the world similar and A means there's nothing even vaguely similar. So in our first fundamental patent, which is issued, we have very good claims and broad coverage. And then everything we've done since, we've patented it and continued to surround those patterns.

And the last question will come from Lisa Walter at RBC.

Perfect. Zandy, this is Lisa on for Luca. Congrats on all the progress today. I really appreciate the modularity of the riboswitch platform and really the immense possibilities that you outlined today. I'm just wondering from a regulatory standpoint, if it would make sense to identify and optimize an aptamer small molecule combo that works and use that in multiple programs. Considering that, that riboswitch has been derisked versus using a new aptamer small molecule combo for each program. Just wondering if you could share your thoughts there.

We can do both. So our first molecule, for example, that is in the clinic next year, the first small molecule that regulates all of the constructs that we've shown today. Would we take all of them into the clinic at the same time, using that, no. We'll prioritize, and we will optimize as we go. But we can use -- I don't think that changing the aptamer necessarily increases the risk of development. I think you do have to do Phase I for every small molecule. So you could synergize and do potentially multiple genes regulated by a small molecule. Ideally, eventually, every gene has its own small molecule, but to start with, would we do more than one with our first? Maybe. Yes. We can do that.

And then maybe just one more, if I may. Just -- I know you have a wet AMD program, and you mentioned this as an ideal candidate for your riboswitch platform. Just kind of what gives you confidence maybe that you're not going to experience inflammation issues seen with other gene therapy programs? And why is having the ability to turn on and off anti-VEGF expression, more ideal versus consistent expression as with other gene therapy programs?

So we have a lot of expertise in the eye and both in the clinic and in vectorology, and we've got a vectorology toolkit in the eye, which we are using to address large indications like wet AMD, dry MD, uveitis and glaucoma, which includes multiple intravitreal capsids, multiple suprachoroidal delivery technologies as well as the development of new capsids to transduce cells more towards the front of the eye than the retina. Now at the moment, gene therapy is delivering VEGFR2 target the retinal cells by subretinal or intravitreal. And yet, if you step back for a minute and you say, well, we want to deliver drugs to the eye. Let's not put them in rods and cones and retinal cells, why don't we put them in cells near the front of the eye, whose job is to secrete into the vitreous. And then we've got a platform for secreting drugs into the vitreous, but you don't necessarily want to have there all the time, and you can activate with a small molecule. Now are we 100% sure that you want VEGF expression for the rest of time or complement expression or anti-complement expression for the rest of time in an eye. I would say there's good data so far. But if you could switch it on with a pill and then -- sorry, with an eye drop and then withdraw that, should you have any issues, I think that has a big safety advantage. And that is still a safety concern, whether it's in the eye or elsewhere in the body about continual expression of gene therapies if they're delivering a therapeutic drug like a VEGF. So this is the first target, which we've regulated because it's fully validated and ready in the sort of gene regulated therapeutics from the front of the eye platform. I don't know that it's necessarily the biggest one because this technology can really be applied to all of those very big unmet needs in the eye because it's a platform technology, both the vectorology and the regulation and the formulation into eye drops.

All right. And congrats again on all the progress today. Great R&D Day.

Thank you.

Great. Thanks for the questions, Lisa. I think this concludes today's question-and-answer session. And I'd now like to turn it back to Zandy for concluding remarks.

Well, thank you very much, everyone, for attending. And I just want to acknowledge. I've spoken today, but the work you've seen has been done by some of the extraordinary scientists that we have at Meira, who are not only imaginative and hard-working and dedicated, but they've really produced an incredible platform. So thank you very much for attending, and I'm sure we'll speak soon.