For being with us this afternoon at the Citizens JMP Conference. Excited to introduce Korro Bio next. Korro is a company focused on novel RNA-based therapeutics, RNA editing technology, and presenting today to us is Ram Aiyar, the company's CEO. Ram, thank you.
Thank you so much, Jason, and the JMP team for having us. Some customary disclaimers. So my name is Ram Aiyar. I'm CEO of Korro Bio. I've been at the company for a little over 3.5 years, really focused on bringing this novel modality into humans. If there's three things you want to leave with today and know about Korro, the first one is we're working on a technology that can change a single base on RNA, converting an adenosine to an inosine. And we do that with super high specificity across every cell using a synthetic oligonucleotide. That's number one. Number two, we've built a platform called OPERA, and that platform uses a combination of four pillars to have that high level of specificity and reversibility. And then the third thing I want to leave you with is the proof is in the pudding.
So we've nominated our development candidate for our lead indication in alpha-1 antitrypsin deficiency. And we nominated that in December of last year, and we are on track to hit a regulatory filing in the second half of this year with clinical data anticipated in 2025. And we have runway now extending into the second half of 2026. When I joined Korro, the vision for Korro was to really expand and bring this gene editing technology into large patient populations. And so when you think about DNA editing or genomic changes, it makes a lot of sense to do that in rare Mendelian diseases where you know the root cause of the disease and you know exactly why something has gone wrong, and you can come in and fix it.
But when you think about large chronic diseases where there is a genetic influence, it's less clear as to how you can do that by changing DNA. So you want a transient and reversible way to make changes to the genome without actually touching the genome. So that's what we do. We transiently make a single base change on RNA by converting an adenosine to an inosine. In the pharmaceutical industry, there are many ways to knock down or abrogate a pathway. So when you think about LDL cholesterol, when you think about all the bad things that you find and identify in the human body, you want to find ways to shut it down or silence it. But there's no real way to activate other than in rare cases of small molecules or in rare cases of gene therapy.
There is no real way to increase or activate a certain pathway. Korro's entire focus and its pipeline is on activation of a pathway, whether it's a specific protein or a cascade of proteins that need to happen. The third component that we've done to enter into large patient populations is to have an intellectual property base that's large. All of those discoveries that we've made on our platform are internal to the organization. Again, the focus of the company, although the technology is really cool, although there are things that we can do that's pretty interesting, the focus is on entering and bringing drug candidates to the clinic. We want to take a very stepwise approach, remove layers of risk. So as you think about our pipeline, our initial focus is taking these oligonucleotides to areas where you can deliver them, especially in the liver.
You have multiple ways to get there to hepatocytes. You have multiple ways to get to the central nervous system. So that is where we're focusing initially on indications where you can show high likelihood of success into areas where you can deliver oligonucleotides, specifically in the liver and specifically in CNS. So why go this direction? I always get asked the question, why are you guys touching RNA when you can touch DNA? It almost feels like a stepchild. When you can touch DNA, why not go to why do you really want to spend time on RNA? It's sort of like asking which of your kids is better. Who do you like better? So they both serve different needs, and they both serve different areas. I think over the last two decades or so, that has come true.
You're starting to identify not just on the genome, but also on the transcriptome, causal mutations that are made either in genetic code or as they get transcribed and translated, somewhere along the way, changes are made. And so you find these genetic variants in large chronic indications, and you start to see causality being implicated in them. Before the advent of RNA editing, there was no real way to use that genetic information and translate that to something more meaningful. So that's where RNA editing steps in. So for folks in the room where I'm going to spend a little bit of time on mechanisms, so please bear with me. We won't go too much into the scientific details.
But at a very, very high level, much like siRNA or much like approved modalities like antisense oligonucleotides as well as siRNAs, we co-opt an enzyme that's present in every cell. This enzyme, its sole job is to make that single base change from an adenosine to an inosine on RNA. The enzyme is called ADAR. It's expressed in every cell. It's present at high levels in every cell. And at any given point in time, this process is ongoing in us. There are some specific genes that if this process doesn't happen, the protein is not produced, and it's deleterious to the body. In other cases, it happens due to environmental factors at 10%-20%. So there's an entire range that at any given point in time, 2%-3% of all cells in our body undergo this editing.
The reason it can do this is because it recognizes double-stranded RNA. And so it forms as we think about RNA or messenger RNA in this case, you think about a single strand nicely laid out on a table. That's not the reality within a cell. It's a hairball. And so it forms these secondary structures. It folds on itself, and it attaches to other things. And so when it folds on itself, it creates these double-stranded structures that ADAR can recognize. And so in one such instance where there's an adenosine being exposed, ADAR recognizes that adenosine and converts it to an inosine. So we create those similar double-stranded structures so that we can co-opt this enzyme and really target it to that specific adenosine. We've been able to show that we can go after every cell and most cells in the body. We can go after any adenosine.
When you think about the four alphabets or five alphabets that are there in RNA and targeting any adenosine on that RNA transcript, that's close to 20%-25% of the transcriptome that you're going to actually go after. So it gives you a large opportunity space to really go after things. So how do we create this platform that we call OPERA? It's based off of four pillars. The first one is a very deep understanding of this enzyme called ADAR. Korro is the only company that has been solely founded based on RNA editing technology. One of our co-founders, Josh Rosenthal, has been studying this enzyme since 1995 when he was a grad student. That's when it was discovered. And so we have a very deep understanding of what this enzyme does and how we can co-opt it. The second pillar is around chemistry.
We've been fortunate that over the last two decades, companies like Alnylam, Ionis, Dicerna, Arrowhead have worked on chemistry to be able to chemically modify RNA but also deliver it. And so we leverage that expertise that's there on this chemistry. And so when you look at our team, there are folks that have worked on approved products coming from each of those companies. So we have a strong expertise in oligonucleotide chemistry and how we can design these compounds for its specificity. The third component, you can't talk about oligonucleotides without talking about delivery. And so here we've taken a very pragmatic approach, i.e., for the first few compounds that we have in development, we are leveraging existing delivery where people have shown they can deliver oligonucleotides to a specific tissue type at high potency and high safety margins.
And so for our lead program, we use a lipid nanoparticle. For our subsequent programs in the liver, we'll use GalNAc conjugates. For programs in the CNS, we'll decide whether we do it naked or with lipid delivery. So we want to leverage things that already work from a tissue type. And then the fourth component, again, we won't go into a lot of it today, is oligonucleotides actually are alphabets that are lined up. Even the chemical modifications that we put can be coded into a language. So we use a lot of algorithmic tools there. I use the word AI to design some of the compounds that we're working on. In fact, one of our first lead compounds here was heavily leaned on through this artificial intelligence approach. We have a patent portfolio that's pretty broad.
It covers both the design of the oligonucleotides plus chemistry plus delivery plus, in certain cases, the specific proteins that we can go after that are non-natural. So we have a very broad patent portfolio that helps us prosecute on our pipeline programs.
So when you sort of step back and say, OK, where can RNA editing really fit? When you think about central dogma going from DNA through the protein, we have an opportunity to intervene in the pre-mRNA phase and go after proteins, non-coding regions, and think about changing in gene expression. Or you can go further down towards translation. Once an mRNA is made, make a specific amino acid change, change the protein, and really change the structure function. So that's where we spend most of our time, which is changing the amino acid sequences and therefore the structure of the protein and therefore its function.
And again, always focused on upregulation of a certain pathway. That's what our pipeline really does. The first two programs in this case are going after and repairing the amino acid sequence, alpha-1 being the lead program, LRRK2 being a program in CNS going after Parkinson's disease with a specific mutation. And then every other program is really focused on how can RNA editing be different, i.e., creating a de novo protein with altered properties that will alleviate the pathology. For our lead program, as I said, we are on track to have a regulatory filing in the second half of this year, starting a clinical study, and then having a data readout in 2025. So as I said, we can target any adenosine. We've designed our first-generation compounds called CORDS. It comes from the platform. It's single-stranded.
And with this single-strand first-generation compound, we can go after any target, super high specificity, super high editing efficiency. And we, as I said, leverage existing chemistry as well as delivery. So this is a data set where I can show on the left-hand side, we look at hepatocytes, go after multiple targets, editing at greater than 80%. On the right-hand side, looking at CNS targets, also very unique in terms of how we go after it, we can achieve close to 50% editing and above. These are both in vitro cell lines that we look at. One of the things that we have shown that others have not shown is that when we deliver these oligonucleotides, we don't take the enzyme ADAR away from its day job. So on the left-hand side here, you see multiple oligos going after an endogenous site called COG and so forth.
We have two other sites shown here. On a regular basis in our bodies, in hepatocytes, this site gets edited. So COG gets edited about 30%, COPA gets edited about 20%, and Ajuba gets edited about 75% or so. And so what we're showing here is that at super high doses of our oligonucleotide, we do not alter the endogenous editing ability of this ADAR, i.e., we don't exhaust this protein, and it can be utilized over and over again. I spoke to you a little bit about the computational approaches. For this talk, I'm going to skip over this just because, you know, as we think about these mechanisms and models that exist to look at designs of oligonucleotides and leveraging them, we've come a long way over the last 3 years. Same thing on the computational chemistry.
We have a deep understanding of how these structures bind to the RNA. We understand where they fit. We can actually think of existing chemistries that we can apply in those cases in a novel approach to improve potency. On the right-hand side here, we show in vivo, just by making a single change, we're able to increase editing efficiency almost threefold. The last piece I'm going to touch on today on a platform standpoint is really focused on delivery. This is a question that we get asked all the time. Our lead program is in we're using a lipid nanoparticle. Is that what you'll use in terms of moving with all of our programs? The answer is no because really we're focused on the indication and the target and the patient population that we're looking at.
So here is an example where on the left-hand side, we can show we achieve greater than 50% editing with a GalNAc conjugate that is delivered sub-Q in mice. On the right-hand side is a surrogate target that we've taken called RAB7A, delivered via an LNP, also achieving greater than 60% editing efficiency, just to highlight that with the same target or with the similar targets, you can get editing efficiency by layering on delivery. So for the next 10 minutes, I'm going to focus on alpha-1. This is where most of the interest has been in terms of the company. As a lead program, it also drives the most value. So hopefully, I can take you through why we believe that we are delivering a potential best-in-class candidate for this patient population. So a brief note on the indication. Alpha-1 antitrypsin deficiency is a rare Mendelian disease.
It is caused by a single missense mutation that is a G to A mutation in a gene called SERPINA1. That gene is primarily expressed in the liver in hepatocytes. Its main function is to act as an acute phase reactant. So what do I mean by that? When you are normal and you have this M genotype, your hepatocytes secrete and make normal levels of M protein. I'll go into the levels in a second. The primary goal of this protein is when you have an exacerbation, whether it's a wound, whether it's something in the lung, whether it's something on the skin, whether it's something in the gut, you get this influx of white blood cells, in this case specifically neutrophils, that come on board and want to destroy the pathogen that's around the site of injury.
Alpha-1 is a way to stop that signal because if you let the neutrophils run amok, they're going to put neutrophil elastase, and then you start destroying your tissue. So alpha-1 is the stop signal to prevent further degradation. When you have the Z phenotype, or specifically in this case, the homozygous Z variant, you end up with a protein structure that starts misfolding. And because it starts misfolding, it starts to polymerize and aggregate and starts to get deposited within the hepatocytes. Progressively, over the course of 40-50 years, that deposit leads to cirrhosis and a loss of lung function. The other impact of that is now you have all of this protein stuck in the hepatocytes and not really out in circulation.
And so because you don't have enough out in circulation, when you get an injury, and in this case specifically to the lung, you end up getting this huge influx of neutrophils, and it doesn't have a stop signal. So the neutrophils actually physically eat lung tissue over the course of time progressively. And so that's why when you have this ZZ phenotype in these individuals, you have a risk profile of high likelihood of liver injury as well as a high likelihood of lung injury. And in some instances, patients have had both a liver as well as a pulmonary transplant, which is not a great place to be. People talk about levels of alpha-1 for circulation, specifically in the context of regulatory approval. So on the left-hand side is the phenotype. On the right-hand side is a ZZ homozygous individual.
You can see that progressively, you see a decrease in the amount of alpha-1 protein in circulation. You also see from an odds ratio standpoint that you have a pretty eightfold increase in either cirrhosis or in COPD for the ZZ phenotype. Our goal is to get to 50% editing or above, i.e., convert at least one of those Z alleles to an M allele. Hopefully, we can get above that 50%. We want to be in that 50%-100% range. Hopefully, I can share this preclinical data that will show that we can actually achieve that. The other reason to get above 50% is that 10% of all individuals that are on augmentation therapy, which is a standard of care in the United States, 10% of those individuals have a single Z allele. So KRRO-110 is our product.
It's a CORDS or a single-stranded oligonucleotide that is encapsulated in a lipid nanoparticle. This lipid nanoparticle was specifically licensed from Genevant. It has clinical precedent. We are delivering this IV so that it targets the human liver, which is where predominantly alpha-1 is made. We are converting a certain proportion of the transcript in every cell into either the M version or we leave it as a Z version. The critical piece there is that we've been able to show and demonstrate that we can get a lipid nanoparticle to every hepatocyte through multi-dosing. We've also been able to show that we get a pretty high level of editing efficiency from a transcript perspective. So I'm going to skip over this slide. I wanted to highlight how specific 110 is. What is being depicted here is human hepatocytes from individuals that have a heterozygous phenotype.
Turns out that 1 in 23 Caucasians have this single Z allele. What we did is we delivered 110 to these hepatocytes in vitro. And we looked at every adenosine, 100 base pairs to the left or sorry, all 100 base pairs to the left of the target site, which is the E342K site, and 100 base pairs to the right. We tried to identify every adenosine that could be edited using this approach. And we show that even having almost 7 adenosines in the vicinity of the target, we do not touch any one of those adenosines. And this is the precision here we achieve is because of the chemistry that we put on the oligonucleotide. This is not the case for some of the competitors that are in this space.
I think when you start to think about from a regulatory standpoint, this becomes important as you think about how close you need the protein to be to the endogenous protein. I then want to highlight recent data that we've shared. This is dosing these mice once every two weeks using tail vein and then looking at data a week post-dosing. Here, that's the study design on the left-hand side. The mouse model that we've used here we've done this in multiple mouse models. The mouse model that we depict here is the NSG-PiZ mice. The reason we did that is to compare across others in the field. On the right-hand side, we take the liver, mash it up, look at what % of transcripts are SERPINA1 that have the A versus transcripts that have the I.
We've been able to show that even just a single dose, we've been able to achieve greater than 50% editing a week post that dose. We've then looked at the protein in circulation and so at those time points, so 7 days post the first dose, 7 days post the last dose. So on the left-hand side, the shaded bars indicate the Z protein, which is what you see in the control. Just after a single dose, we get about 50 micromolar of protein in circulation, out of which 35 micromolar of protein is the M variant. I don't think anybody has achieved such high levels even after a single dose within a week in terms of the M protein. What you also see is that over the multi-dosing, you see a progressive increase in the amount of M.
That's because of the accumulation of the oligonucleotide in the cells. On the right-hand side, we depict, OK, so we create the protein. Is it really functional? And you get a yes/no answer in terms of over time. You do see an increase in function. And you do see that the M protein that's created is functional. This was done at 2 mg/kg. And what I shared on the prior slide is that even at 5 mg/kg in these mice, we've not seen any changes in LFTs. We've not seen any other overt tox signal that's adverse, which puts us in a very good perspective because if you think about Onpattro or patisiran that's approved, they had no difference from their efficacious dose to their tox dose. One of the big questions for gene editing has always been, will this translate to humans?
For oligonucleotides, that mechanism has always existed. But we want to make sure we can showcase that using a surrogate compound. So what we did is we targeted the non-human primate SERPINA1 gene, edited an adenosine that's at a different location that is present even in the PiZ mice, and then looked at, in monkeys, can we actually edit it? And so the left-hand side is the mouse data where we show about 25% editing at day 4. And on the right-hand side is the monkey data where we've taken a core from a non-human primate with a single dose. We get north of 40% editing. The second thing that was notable is that oligonucleotides, ASOs, as you increase in species, also increase in stability. So you see at day 14, we have about 2%-4% editing in the mice, whereas we have closer to 18% editing in monkeys.
It just gives you a sense that the likelihood of success in the clinic is actually very high as you think about this modality. I'd like to end there. From an efficacy standpoint, we've shown we have the best-in-class data. From a safety standpoint, we have showcased that we can provide a therapeutic index that's better than approved products. And then showing translation in monkeys is something that nobody has really done. And we've been able to show that pretty rapidly. With that, I'd like to end. And happy to take any questions, if any.
Great. Thank you. Any questions from the audience? Go ahead.
I know that you told me that mRNA and pre-mRNAs here. But have you guys considered the utility of editing microRNAs or the regulatory RNA?
We have. So microRNAs end up having poly-A tails. And we can go and edit some of the adenosines. The biology gets a little complicated in terms of what else from a regulatory standpoint it can go after. So it is on our radar. I want to see how this looks in the clinic first. And then we can start expanding where to go into. But that's a great point. So we've looked at it. We haven't done anything in that space.
Great, Ram. Really, thank you. Appreciate you being here.
Thank you so much.