All right. Let's go ahead and get started. Good afternoon. This is Steven Mah on the Tools and Diagnostics team, introducing our next company in the synthetic biology space. It's my pleasure to welcome Stephen Dilly, CEO of Codexis, to the conference. We'll keep it interactive. If there's any questions from the audience, please just raise your hand or just ask, or if you prefer, email me at stephen.mah@cowen.com and I'll ask the question. But you know, before we begin, Stephen, you know, could you give us a quick rundown on Codexis for you know, for those that might be new to the story? I know there's been a kind of a pivot with the business model and the value proposition and strategic focus. So maybe just kind of level set the audience would be great.
Sure. And thanks for the invite, Steve. Great, great to be here. So Codexis is in the business of biocatalysis. What we do is we build and evolve really well-characterized enzymes to stimulate chemical reactions. And our sort of heritage business has been in small molecule manufacturing, where we've made enzymes for difficult chemical steps. A couple of claims to fame in that would be making an enzyme to help Merck make Januvia, sitagliptin. Another one would be a real fire drill in 2020 when we scaled an enzyme for Pfizer, a few tons of enzyme for Pfizer, to enable them to make Paxlovid to meet the demand of the pandemic. And that business is mature. There are about 12 commercial products using our enzymes right now. 60% of our revenue from that business, though, comes from three enzymes and four products.
What we're aiming to do over the coming years is grow that to maybe six or seven enzymes and seven or eight customers out there. Just to set expectations on that, at the moment, that business is mid-30s in top-line revenue, about 60% margin. Over the coming years, we'd expect that to grow significantly, both in terms of revenue and bottom-line efficiency. So we've learned from that, and we've learned how to build enzymes, scale enzymes, evolve enzymes. And now our attention is super-focused on RNA because we see that as the coming wave of therapeutics. We did a deal at the end of last year with Aldevron, where we placed our HiCap RNA polymerase for manufacture of mRNA in their hands. That's great because they can scale it. They can get to GMP production. And we can leverage their channel.
We see that as exciting. But the place where Codexis is laser-focused right now is the fully enzymatic synthesis of siRNA. We call this our ECO Synthesis platform. In a nutshell, it's a suite of enzymes that, if you introduce the building blocks in the right sequence, can pretty much make any siRNA fully enzymatically. As a stepping stone between here and there, we have our RNA ligase program, which enables people to do short-mers by chemical or enzymatic synthesis and stitch them together. You know, the sort of punchline here is that when we are selling enzymes in the small molecule manufacturing space, the going rate is a few thousand dollars a kilogram for the enzymes. They're unrefined enzyme produced in bulk. When we're talking about the RNA space, it's thousands of dollars per gram.
So it's a completely different world, and the demand is similarly high. So that's where we're going.
Okay. All right. Well, great. That's so loud.
Sorry. All right. So maybe let's take a step back and, you know, tell us about the engine behind all this innovation. You know, I, I know you've been using machine learning and AI long before they became a buzzword. You know, maybe, maybe give us a little bit of background on the CodeEvolver® platform and how that's kind of driving, you know, kind of your, your, your enzyme business.
So absolutely. CodeEvolver is a directed evolution platform. What it enables is us to take a wild-type enzyme or an enzyme that we know what it does and refine its characteristics in a very defined way. And, you know, it's based on AI, machine learning, in silico modeling, but just as important, a very robust expression system and an analytical framework that allows us to do cycles of evolution to see, you know, what's working, what's not working, making individual and combinations of mutation steps to see what happens, and, you know, sometimes taking steps backwards to go forwards, that kind of stuff. And as you say, we've been doing machine learning and AI before it was even called that.
What we've learned over many, many years is that the quality of the output is directly related to the quality of the dataset that you're operating from. And so one of Codexis's most sort of treasured possessions is our vast library of enzymes that have been evolved at different stages where we can look at different reactions, different characteristics. So if we're starting something new, we're usually starting from ahead of anyone else. So the evolutionary steps are shorter.
Right. Right. All right. And then maybe let's go into, you know, your three business areas that you're, you know, you're going into now. So I think you've broken them down into one, sustainable pharma manufacturing; two, life science tools and enzymes; and then three, the ECO Synthesis for RNAi manufacturing. Maybe just give us a sense of the total addressable markets for each of those.
Right. So starting from, you know, the top, the principal determinant of the total addressable market to us is the quantity of enzyme we can produce and how much we can charge for it. And as I say, in standard biocatalysis of small molecules, when there is usually a chemical alternative, we're talking about a few thousand dollars per kilo. We can get up to ton quantities, but a program there is doing well if it hits high single-digit millions of annual revenue. So $7 million-10 million a year from a would be a very successful program in pharma manufacturing. When we looked at the sort of life science tools arena, they're more exquisitely engineered enzymes. They have to be purer, and so on and so forth. The price point goes up, but the quantities go well down.
So, you know, that's a modest market. And we've determined that we call that non-core. And really what we're trying to do is leverage others. So for instance, we had an arrangement with seqWell where we engineered a transposase for them because they know all about that market. They're going to do the heavy commercial lifting. In the RNA space, we're seeing the happy sort of coming together of high price, high quantity. And we think that is many hundreds of millions of dollars of annual revenue, from a relatively modest slice of that market, which has led us to the, I think, really important conclusion that we have to evolve ourselves from being an enzyme producer to actually producing siRNA ourselves. And our first step on that road was the commissioning of our ECO Synthesis Innovation Lab.
Right. Right. Yeah. So the RNAi, you know, it seems to be a growing therapeutic modality now. You know, it's, it's been around for some time. But, you know, maybe, maybe compare and contrast what the existing solution for, for manufacturing siRNA versus your solution, just to kind of level set everyone.
Right. So the existing solution, the chemical way of producing siRNA, is phosphoramidite chemistry, which has been around for more than four decades. Well, actually, that's a lie. The original way of making oligos was enzymatic back in the 1950s and 1960s, but the enzymes were wild-type. There was no way of evolving them, so they couldn't get very long. So people went off into this detour of chemistry. Now, phosphoramidite chemistry works, but my goodness, it's hard. In order to produce a kilogram of siRNA through conventional phosphoramidite chemistry, it takes about 7,000 L of solvent. And these are things like acetonitrile or toluene. It's a very intensive multi-step process. There are lots of side-chain reactions. It's enormously capital-intensive.
So for instance, one of the things that's in the public domain is Agilent, the big player in the space, recently spent between $700 million-$800 million on a facility in Colorado, about 200,000 acres, which will produce about a metric ton of siRNA a year. We see the demand by the end of this decade approaching 20 or 30 metric tons. So, you know, that's a huge capital investment to get there conventionally, not to mention the sort of toxic byproducts, both economic and actual on the environment. The drastic contrast with our platform is that we're an aqueous-based solution where the enzymes do the heavy lifting. You know, you're moving from a situation where you need exposure containment and disposal facilities for your toxic solvents to a water-based system using enzymes.
We see both the efficiency when it's running, but also the dramatic change in the CapEx demands to be a real edge for us. We should be intrinsically scalable because we don't have some of the physical limiting factors of phosphoramidite chemistry, which keeps them to batch sizes of about five to 10 kilos. We should be able to get into the tens to hundreds of kilos in a batch with the enzymatic route. When we talk to the players in the space, including those already doing phosphoramidite chemistry, we see a general recognition that, you know, something has to change. Now, people are working on tweaking the chemical route and looking at solution phase, looking at other things. But those will be incremental changes because this is a mature technology that's been around a while.
And so that's where we see the disruptive enzymatic platform as having its, its potential.
Got it. And give us a sense for the scale, you know, to make, like, the batches you're talking about, like, do you need, like, a bioreactor or, like, you know, is it, like, existing infrastructure which is available, or has it to be some sort of new manufacturing? And I appreciate you guys have just got, you know, achieved gram-scale synthesis recently, so, but, you know, what's your sense on how a large-scale production facility might look like?
Well, I'll start with what the small scale looks like and then go to the large scale, right? The funny thing is when we do a show and tell at the moment, and we bring people in that are used to looking at phosphoramidite facilities, and we show them the lab where we're making siRNA, they go, "Is that it? Where is it?" You know, it's because it's a packed column with an immobilized enzyme in resin chucking out siRNA. Now, that's intrinsically scalable. You can put it in a very moderate footprint. Yes, it will be standard bioreactors that can do this. We think in the future with our ECO Synthesis innovation lab, this is a 1,000 sq ft footprint, essentially, in a corner of one of our buildings partitioned off.
That'll allow us to make GLP material at multiple gram scale. Now, then we say if we want to move into GMP manufacturing at the kilogram scale, that's a few thousand sq ft to do that. So we're talking vastly smaller and without this enormous infrastructure around it because, as I say, the enzymes are doing the heavy lifting here.
Okay. Got it. And so, yeah, maybe we spend a little bit more time on the, on the gram-scale synthesis. I know, I know you've achieved that. What's sort of the next step, on the, is it making the ECO Synthesis innovation lab and then going to multi-gram scale or kind of, you know, kind of outline, like, the, the pathway toward forward here?
Right. So this is all about enrolling partners, showing people we can actually do this because, you know, this time last year when I was talking about this, this was a sort of twinkle in our eye. This was an idea that we could probably do it. And a lot of people said, "Lovely idea. Show us. Come back when it works, right? Come back and show us." Gram scale was actually the first step on that where we showed that we could include the real sort of building blocks with the real modifications people are using in current therapeutics at significant length with what we were realistic process conditions. The next step from that is to make people's siRNA therapeutics and show them to them, right? So that's where we're going.
And by the time we come back to Boston in May, we'll be doing that, showing full-length siRNA out of the ECO platform. Now, when we're talking to potential partners that are adopters of this technology, they're also equally interested in the enabling parts of the platform. So there's the little ECO engine, which is the sequential synthesis of the siRNA by adding building blocks with the deblocking enzyme in there. There's also how you make the building blocks themselves, the nucleotide quadriphosphate, enzymatically. There's also how you make the starting material enzymatically. And then the sort of cherry on top is how you do the enzymatic conjugation to the targeting moiety. And we're making, you know, major progress on all of those such that, it's not just a matter of making something that's a sort of abstract concept. It's making a therapeutic, showing it to people.
That's really, we think, going to be extremely powerful because then they can test it, look at the purity, and really think about the pros and cons of adopting the enzymatic method.
Great. Okay. So the synthesis lab's going to support sort of, you know, potential partners kicking the tires and.
Yeah. And GLP as such that, you know, if it's a small company early on, we can make their siRNA in the ECO lab to get them through preclinical testing, right? And then we can talk about how do we scale and where do we go to do the GMP. And, you know, at some stage, that will be in-house at Codexis, but in the meantime, it'll be through CDMOs, through some of the, you know, relationships we've cultivated over many years in the pharma manufacturing space.
Okay. Got it. Okay. So the idea is the technology is you're not necessarily going to do all the manufacturing yourself. You have the potential to, you know, outlicense and do a tech transfer to CDMO or to potential partners.
Yeah. So the technology is tweaked to match the customer. So, you know, think about it. If you are, you know, you've just raised Series A or whatever in a company with a, a really exciting siRNA early-phase molecule, what you want is someone to take the headache away from you, make your siRNA for you. That's what we'll do with the innovation lab, and then we'll scale it with them. If you are a CDMO, what you want is a kit where you can service multiple customers at scale. At that point, it'll be reagents, enzymes, columns, that kind of stuff we'll be supplying them with, and, you know, a user manual. And also what we're finding in our, in our early interaction with, with collaborators is there's a to-and-fro between the engineering the enzyme, playing with the process, optimizing around any given molecule.
And then, you know, there are going to be some sort of innovator companies, big pharmaceutical companies that are going to want to bring this in-house so they can make it themselves, and they'll pay us for the, for the reagents. So bespoke. Now, all the time, with the siRNA, we shouldn't lose sight of the other part of this platform, which is the double-stranded RNA ligase. Because right now, in an effort to make phosphoramidite chemistry more efficient, people are manufacturing shortmers, so six or seven long, the final molecule somewhere around 21 long with overlapping strands, and stitching them together using wild-type RNA ligases.
We have an offering that, you know, is ready for prime time now where we've evolved a series of ligases optimized around the specific reactions people would want to drive, with their shortmers, which is a very rapid entry because it's a semi-enzymatic synthesis. And, you know, people are doing this right now on FDA-approved moieties and under GMP. The agency has an opinion about which bits are starting materials and which bits are under GMP. That gives us a rapid market entry into getting people used to the concept of using enzymes.
Okay. So that double-stranded RNA ligase, that's, that's going to be potentially outlicensed for other applications aside from the ECO Synthesis, right? Because it's part of the ECO Synthesis enzyme suite, right?
It's part of the ECO Synthesis enzyme suite. And there are even situations where you would see using a ligation step within an enzymatic synthesis, particularly for very long constructs, right? And so we see it as a, a really important entry point into that. And, you know, we initially talked about it as the first generation that we'd evolved, and then there'd be a second generation enzyme that would do everything. And what we're learning is just like pharma manufacturing, there is an opportunity to optimize and, and perfect the enzyme for the specific reaction it's doing. Because, you know, if you want a platform when you're in discovery phase that can make anything, that's a different problem set from you want a platform that makes three tons of this specific molecule, right?
And so we have the option to optimize around the specific molecule when it's scaling very rapidly, particularly if people have done their initial small scale with a ligation step.
Okay, you know, you mentioned, you know, some of these full-length siRNAs are, you know, 21+ nucleotides. I mean, give us a sense of the error rate of making RNA via phosphoramidite versus DNA, you know, because, you know, I cover a number of DNA synthesis companies where, you know, it can go upwards of 150 nucleotides without, you know, having, you know, too many errors or an acceptable error rate. What's the difference between DNA synthesis chemically versus RNA synthesis chemically?
A lot of it is around the side chain reactions, and it's the impurity profile rather than the error rate. I mean, there are errors, through any sort of chemical system, but really it's those, it's those harsh impurities.
Okay. You just, you just get more impurities on the side chain the longer.
And the reaction slows down. And really sort of practically, the handicap on phosphoramidite chemistry is going too long, right? That it's very good for sort of brute-forcing it on relatively short strands. And so you can see a sweet spot here where people are making certain things that are sort of six or seven mer and then stitching them together. That's what people are already doing because of the efficiencies of scale. The reaction does break down. You do get more degradation as you go longer, right? And, you know, looking at the enzymatic DNA space, you know, that's always been said to be the advantage is sequential synthesis of very long strands very accurately. And, you know, like our friends at MAI doing that, for instance.
Right. Right. Okay. All right. So, what's your, I mean, a lot of those companies have actually not done so well, you know. Can you maybe compare and contrast why this will be potentially different in the RNA space versus the enzymatic DNA space?
Yeah, because the demand is just dramatically different. And the alternative has this big challenge of scalability that, you know, there's real live demands for at the moment, several hundred kilos worldwide, which we see going up, as I say, into the metric tons as new approvals come through. And the pipeline's already full of these new modalities. So the question for those who can't ask is how people producing the siRNA now for the hundreds of kilos. It's largely through phosphoramidite chemistry. Many of them are doing phosphoramidite chemistry with enzymatic ligation. So no one's doing fully enzymatic synthesis yet, only us. So all of the modalities you see out there, like, you know, inclisiran is probably the biggest drug right now.
That's being manufactured, you know, in a big chemical plant using tons and tons, well, 7,000 L per kg, you know, of, of solvent, massive carbon footprint. But the beautiful thing about siRNA therapeutics is that the annual dose per patient is very low. So, for instance, inclisiran is about 600 mg a year split between two doses. The, the fact that the cost of goods is, is high in a per kilogram basis, it goes out in the wash because of the sheer efficiency of the molecule. And one of the things that gets us excited about it is these are better medicines, potentially, things where people are being dosed twice a year and having exquisitely good control of their lipids or their hypertension. And Steve, you asked the question, you know, a little bit what's happened about this modality that's been around 25 years.
Why is it suddenly hot news? A lot of that's about tissue targeting and working out how to put it in the CNS, how to put it in the lung, how to actually deliver it to treat more prevalent diseases. And so that's why we see this, this dramatic increase in demand. And what's happening now is the big companies are tapping out the infrastructure and the supply chain. And when we talk to small innovator companies, you know, the big companies might be paying $250,000 or $500,000 a kilo for siRNA. You're a small company. You're paying $3 million because you're at the bottom of the line and you're trying to find a slot. And so we see ourselves servicing a very broad part of that ecosystem.
Right. So how's that affected your partnership discussion? So it sounds like that's actually a pretty good value proposition if you're, you know, a small biotech, you know, developing a, you know, siRNA platform. I mean, can you tell us how maybe your partnership and BD discussions have evolved since the gram scale achievement?
Yeah. Well, it's been really busy. It's been a number of different constituencies. There've been the sort of big kahunas in there that are sort of watching us and waiting and sort of will adopt, but clearly interested in what we're doing. There are the players in the CDMO space that want to get into the business but don't particularly want to build all that infrastructure that we see as a very, very fertile ground for discussion. And then there are the small innovator companies saying, "Can you make this for us?" And, you know, the ECO Synthesis lab, we've, you know, started work on that. That should be up and running in the second half of this year.
As I say, we should be able to service, you know, about 10 customers per year at the, you know, tens of grams scale out of that facility, within our existing footprint in Redwood City, California. This is our way of, you know, catching them young, early, and then taking it forward.
How do you envision monetizing these partnerships? Would these, you know, just be like a fee for service, or would you actually, you know, take some downstream economics, or how should we think about that? I appreciate they haven't been signed yet, but.
So with big solvent companies, what we're talking about is a very healthy cost per, you know, whatever of enzyme plus a license fee for using the technology and so on, that kind of structure. We're also, you know, over time, we'll be willing to look at, you know, downstream economics as well that scale with the value of the asset. And also, you know, to tailor that to the customer we're talking to. So, and again, it's going to be different when we're providing people with the kit, essentially the enzyme, the reagents, and the columns and the user manual versus if we're actually making you your siRNA when we're going to charge you like a CDMO would for that, right?
Right. Right. No, I'm just trying to, trying to think of a, you know, you know, retaining the IP versus, you know, one, once something's kind of kitted, I mean, is there are risks that it would be reverse engineered or.
We've got a lot of IP around this. And, you know, we don't worry. We've been all over that. And it's not just the enzymatic IP. It's also the process. There's a number of sort of non-intuitive things. And as I say, there's the kit itself is no use without the reagents. The reagents can be really prohibitively expensive if you don't make them enzymatically. You know, there's all kinds of guardrails on it. So we're trying to stay ahead of the game, but it's a completely valid point.
Okay. All right. No, I appreciate that. All right. So maybe just, you know, just sticking on the ECO Synthesis platform, you know, you mentioned, you know, some of the timelines and goals, but, you know, kind of maybe can this outline kind of the roadmap between here and having a, you know, fully GMP qualified facility?
Right. And I think the roadmap that has to run in parallel is the ability to pay for it, right? Is the ability to be in command of our own financial destiny. And so, we made some big changes in Codexis last year where we really focused the business. We monetized a lot of non-core assets. We built ourselves the runway to the end of 2026, by which time we think we should be healthily cash flow positive. Now, we then did a debt financing with Innovatus to both give us more cushion on our cash runway. Most of that debt is going to sit on the balance sheet and giving us more wiggle room, but also allocating $10 million to scaling up the ECO Synthesis innovation lab.
So the way we see the cadence is very soon from now. We expect to ink our first commercial customer on the ligase platform. We come to the middle of the year, the TIDES Conference here in Boston. We're showing full-length siRNA. We expect around that time to sign our first validating partnership in terms of a technical collaborator, one of those, you know, CDMOs I was talking about. Then we start signing up in the second half of the year, small companies working, wanting to work at a GLP level. And then all the time we're having conversations with the sort of the bigger players around their adoption of the ECO platform. We need one or two, what we call early access customers this year, growing to three or four or five next year. And then full commercial launch in 2026.
That's when, you know, we should be providing the full ECO platform with the enzymatic build NQP building blocks, with the starting material, with the targeting moiety. So it really is ready to go. That's a hell of a lot of work, Steve, but we think we can do it.
Okay. Got it. And then, yeah, you know, obviously, you know, the recent funding, $30 million+ an option for another $10 million, you know, should we expect any additional financing things to get you through that roadmap, or are you there?
So we're in the lovely position where we don't need to. That doesn't mean we won't, but it means we won't need to. If we finance, it will be because we have a very clear use of proceeds to accelerate our value trajectory, right? We can get where we need to get with what we have now, which is, you know, an excellent position to be in.
Okay. Well, great. I don't know. There's one, one question here.
Help me understand what you mean when you say full commercial launch. Because it sounds like it's more easily, commercialized using CDMO partners, right? Using their kits. So why is there even a necessity for you to be the CDMO manufacturer?
So the question is why don't we just stick with leveraging CDMOs and providing them with kits? Why do we have aspirations to become a GMP manufacturer ourselves? It's really capturing parts of the market that we wouldn't have access to otherwise. So there's the small-scale innovator that needs a small amount to go forward. You've then got the step of moving from GLP to GMP. If we can do that in-house, the tech transfer then for the massive scale to the CDMO is so much easier. Many of the drugs in this space are boutique drugs, which we could foresee supplying ourselves through their life cycle because they only ever need a few kilos. It just makes us that much more of a player.
You know, having had a lot of experience of tech transfer of biologics and build and setting up new facilities, it's always better if you've run it at half scale and you're just trying to double it rather than trying to do the first, you know, iteration out of house. Also, you know, think about the idea of the ECO Synthesis lab at large. That allows us to tinker with the process. So we can, if you bring me an siRNA, I put it under my generic platform, I look at how it works. I might say it's great, but step 16's a bit slow. I'm going to play with the process conditions. Having the ability to do that in-house so we have optimized process for tech transfer we see as absolutely compelling.
The other part of it is, you know, the return on investment of being able to make GMP siRNA on the ECO platform is very, very favorable indeed. We're talking about, you know, investing tens of millions, not hundreds of millions to, to get a GMP facility that could do multiple-kilogram scale. Not yet. We're not quite far enough along the validation cycle, but pretty soon we'll be in a position to, to entertain that.
All right. Appreciate it. Is there any other questions from the audience? Otherwise, I want to thank Stephen for being here today and congrats on all the progress.
Thanks, Steve.
Looking forward to TIDES in May.
Excellent. We do, too.