All right, I think we could get started here. Hey everyone, I'm Chad Wiatrowski. I'm part of the Tools and Dx team here at TD Cowen. I'm excited to welcome Stephen Dilly, CEO of Codexis. You know, Codexis has gone through a lot of evolutions, and now today is kind of pioneering a novel approach to siRNA manufacturing. There is a lot to talk about. It's an interesting market. There are a lot of things that are, I think, underappreciated about siRNA manufacturing. It will be interesting to dive in and see what you guys are up to now. I guess for people that are newer to the story, since there has been some evolution, can you kind of walk through where you've been, where you're at today, and where you're going, and sort of the why behind some of the pivots you made?
Sure. Thanks for having us here. Codexis has a long history in enzyme engineering and optimizing biocatalysts. That is using enzymes to drive chemical reactions. That started in the field of small molecule API production. During the history of Codexis, it has branched off into a lot of different areas like biofuels and food and feed and other things. Really, after a good long look at the potential of everything in our pipeline, we said we are the best in the world at engineering enzymes and directed evolution. We need to be competing where that matters the most. What we saw was the emerging field of oligonucleotide therapeutics led by the coming wave of siRNA therapeutics.
That was a very happy coming together of circumstances because these are a very important modality that will address some very prevalent diseases, where there is a challenge with the conventional route of manufacture, which is the chemical manufacture using phosphoramidite chemistry, particularly when you get to scaling to supply the large markets. We saw an opportunity to build an enzymatic platform that has much more efficient scalability, can actually perform better, and by that I mean produce purer material faster and cheaper. The other beautiful thing was it was really hard, and we did not think others could get there nearly as fast as we can. That has really played out over the last couple of years.
With siRNA, siRNA as a new modality, it seems like the existing infrastructure with phosphoramidite chemistry is, you know, it's there and it's appropriate in amount. As they expand into these larger indications, as you're talking about, I guess what's the problem there? What's the bottleneck? How does ECO Synthesis address that scaling?
Yeah, there's a lot of fun numbers on this, some of which I will give you. You know, if you look at the worldwide supply or demand of siRNA right now, it's maybe 500-700 kg annually. There are drugs in the pipeline that could be more than that on their own. One of the challenges of the chemical approach to making siRNA is it's incredibly inefficient in terms of the use of raw materials. It takes about 7,000L of solvent to make 1kg of siRNA. That solvent is acetonitrile. Some of it's toluene. These are nice organic solvents. That means, you know, petrochemical derivatives that need to be burnt off. The environmental impact of that is huge. The economic impact of that is huge. It's not scalable.
There's not enough acetonitrile in the world to supply the market size that's likely to happen in five to ten years' time. There is an actual availability argument. There is also a cost and effectiveness argument. The beautiful thing about the enzymatic platform is it's aqueous-based and infinitely scalable.
I guess the ECO Synthesis platform, you've released a lot of data, like last year at TIDES USA and TIDES Europe. If we zoom out and kind of look at what did you guys demonstrate to prospective customers last year at those scientific conferences? Looking at this year, what are you planning on, you know, unveiling and how is that influencing those discussions?
Right. Absolutely. We need to address this market in a stepwise fashion. We need to grow into the opportunity. One of the ways that we're gonna do that is using something that we call the ligase technology. This is a single family of enzymes which are used to stitch together fragments of siRNA molecules to make the complete entity. What we've discovered is there is a significant market opportunity in doing that because it enables the chemical manufacturers to make shorter strands of RNA out of the gate, stitch them together to make their final drug. That can dramatically improve their efficiency and, to some extent, their scalability. That's been our first step in the market. We had our first commercial order of a double-stranded RNA ligase last year. We're fulfilling that right now. We have others in the pipeline.
We have multiple customers on the hook where, you know, we will be building them a ligation solution for their particular problem. You know, the beautiful thing about these enzymes is their value add is orders of magnitude higher than the old biocatalysis enzymes. That is a real economic opportunity. We could build a business around that. That buys us the time to move further in our toolkit. I say toolkit because we have moved from an idea that we would make siRNA end-to-end sequentially with enzymes to saying, you know, sometimes that might be a good idea, but sometimes it is better to make it in fragments. There are some pieces that right now are better made chemically, some that are better made enzymatically.
We can tailor the solution to the exact molecule that we're making over time and evolve it over time to optimize as we scale. What we're hearing from the customers that we're engaged with is they really don't have that much of a problem right now for the smaller molecules that they're making. When we get to 100kg demand for a specific molecule, they absolutely need ligation. When they get to 300 or 400, they need a full enzymatic solution. We are growing with the customers into that solution.
Just speaking to that, like you've recently announced there's seven plus, you know, ongoing discussions, including a lot of large therapeutic developers, but there's sort of a multi-prong approach to kind of the go-to-market. How is that commercial strategy evolving and just how are you addressing kind of specific needs from one customer to another? What does this look like commercially as a product in the next few years?
There are as many solutions as there are customers right now, pretty much, particularly with the big innovative players because they have different capabilities in-house. There are some that have started with ligation and say, look, we just want to improve right now our chemical footprint. They start with ligation. They use one of our ligases. They like what they see. They start talking to us about when can we eventually introduce enzymatic fragments and the full enzymatic process over time. That starts off as a, you know, as a deal around supply of ligase and then move to supply of enzymes and reagents and scaling for the full ECO process. Others come to us and say, we think we have ligation taped in-house. We're gonna give you fragments to make enzymatically. Let's scale around that. Let's look at reproducibility.
Let's look at the path to market. There are others that are coming to us, particularly smaller early phase companies wanting a complete solution, wanting support for their preclinical work, for their scale-up, for their GLP, then GMP production. We really have, this is why I keep calling it a toolbox that we can tailor to not just the specific molecule people have, but the capabilities of the company that we're working with.
Outside of like the technical expertise and the really, you know, innovative things that you're doing, there are also a lot of times in these new modalities, there's a lot of value in the CMC and regulatory components. Could you just give some granularity as to, you know, what capabilities you're building out and how that would be of use to potential customers?
Yeah, absolutely. And all of this takes time and has to be done in a logical stepwise fashion. We see much of the value add in what we do in building the complete analytical package to support the initial introduction and then scaling of the molecule and passage through the FDA. We also see the ability to write the CMC documents on behalf of some of our partners being crucial or to be able to support the bigger partners in their writing. You know, our vision for the future is that, you know, small companies, maybe small VC-backed companies that are getting serious about developing a molecule would come to us as their development partner, and we would actually build their molecule for them. We'd build their analytics for them.
We'd write their regulatory dossiers for them, and we'd stay with them all the way through development. The reason we wanna do that is we wanna get paid on the way, but we also want a piece of the value of the asset itself. I mean a share in the value in terms of sales. That's crucial because, you know, there are some of the 500-odd siRNA molecules that are gonna become very big in terms of quantity, but there are equally important, a whole bunch of them that are pointed at what I would call boutique indications in a few thousand patients where the price point is gonna be compelling. The revenue from the drug is gonna be enormous, but they're only gonna need a few hundred grams to supply the market.
The only way we can capture value there is by being partners in the value of the asset. Now, why are people gonna sign up for us? They're gonna sign up with us when they know that we de-risk their way forward. We de-risk their way forward by making it cheaper, by making it more affordable in terms of the capital investment and all that kind of stuff. Also, they're less likely to make mistakes. The kicker on this is because we are working right now with the big guys, what happens in biotech is the big guys tend to acquire the small guys. One of the important steps in that is the due diligence. CMC is a huge part of that.
Being able to say Codexis does our manufacturing, Codexis does our CMC support will be a major positive factor for the small company wanting to get acquired.
Is there anything like being included in like having a master file or being included or specced into these processes? Is that something that Codexis would be a part of?
We absolutely need to do that. We need to, we are currently going through the conversations with the agency around how we precedent our process, how we bring it in on stream with existing modalities. Having a DMF that can be cross-referred to, having precedent for what people want to do de-risks it. Now, one of the nice things is, one of our first double-stranded RNA ligase relationships was a drug moving from phase two to phase three. Having that in the clinic and being able to point to that is quite important in saying, you know, this is a real thing that actually gets to people. Having the regulatory framework, having precedent we can point to, it is all super important.
Obviously your prospective customers are therapeutic companies, so you're largely insulated from a lot of the headwinds just impacting the broader tool sector. I'd be remiss to ask, like, what are you doing to ensure your raw material supply, just given tariffs and all the noise?
That's a really interesting situation in that we've been looking at, you know, how do we bring on shore our raw material supply? Typically, the raw materials for making siRNA enzymatically are things called nucleoside quadraphosphates, or NQPs, which are derivatives of nucleoside triphosphates, which people often use for making mRNA constructs. Now, those have been synthesized chemically, often from China or India. We have been looking at, you know, how to bring that on shore. We have an enzymatic method now of how you can manufacture those. We're really looking for the appropriate partner, but there is a sort of tangential benefit of tariffs and all the rest of it, which is some of those companies are coming to us right now trying to maintain their business by giving us very sweet deals in the short term.
We're actually getting a bit of a tailwind there in terms of our raw material supply. But over the long term, we absolutely need to bring that on shore and have a full enzymatic platform for the raw materials as well as the final product.
Just staying on the theme of kind of like more high-level secular themes, how is AI impacting your business today? How do you view the future of ECO Synthesis in light of just AI continuing to permeate kind of all aspects of everything?
We're kind of the OGs when it comes to that in that we've been doing AI only, it wasn't called that since about 2007. What we've learned is AI is only as good as the way you apply it. You know, I had a fantastic conversation with one of my lead scientists last week where I said, you know, are we gonna get scooped by, you know, some of these larger learning models and AI in terms of protein design? Will it just all be done in silico? He took me through some of the math around, you know, the average enzyme maybe is a 400 amino acid long protein. Each of those has 20 possible amino acids, even if you don't include some of the non-native ones.
You do 300 or 400 to the power of 20, that's more molecules than there are in the entire universe. That's an incredibly complex system. AI just isn't there yet in terms of being able to model it from ground zero. The way that AI really works is if you can reduce the problem to a bite-sized chunk that it can cope with, with experimental data. Our advantage is that we have a really good machine learning AI platform applied to vast databases generated over several decades of existing proteins and knowing what happens when you make modifications in specific places. The problem with iterative AI on its own is it hallucinates. If AI is drawing conclusions based on previous AI conclusions, it goes off at a tangent. It can often be completely wrong.
We actually find real utility in having these large databases and the interaction of artificial intelligence and machine learning with real experimental data.
It's not a perfect analogy, but there's been a number of companies in the DNA synthesis space pursuing it, pursuing an enzymatic approach. It has sort of similar advantages, like being a clean solution, things like that. What's different about siRNA doing it through an enzymatic approach? How does, like, pricing and lab space versus traditional steel infrastructure play along in making the value proposition worth it in siRNA manufacturing?
Right. So we built an enzyme that is exquisitely good for making DNA oligos, which we licensed to MAI that has now been acquired by our friends at Maravai. And, you know, we looked at that space long and hard, and we really did not see a compelling market opportunity because the quantities are so small. You know, you can think of, you have to ask the question, can I, when does it matter that I can make you an exquisitely accurate long DNA oligonucleotide? You know, and it might matter in full gene sort of synthesis and all the rest of it, but, you know, then you are talking about solutions that require vanishingly small quantities. Otherwise, it is pretty much a commoditized market and it is a, you know, a race to the bottom competing against, you know, bigger technologies, in terms of DNA.
With RNA, it's completely different in that we have a, an end market where people need large quantities. They need many hundreds of kilos of siRNA, which they can't make by other means. That's the real difference here. You know, Agilent, who are probably the biggest kahuna in the current siRNA production space, told the world they'd spent $728 million in building a plant in Colorado using conventional phosphoramidite chemistry. That will have an output potentially of around, you know, a metric ton annually. We see the demand at the end of this decade moving into the early 2030s being somewhere around 30 metric tons. That is an enormous capital investment, even if you get away from the run rate and the solvent use and all the stuff I talked about earlier.
That's why we say, you know, it could happen, but it's not likely, right? If there's a better solution where you can use conventional infrastructure, you can just have, you know, storage tanks and columns. You don't need all that stuff around, you know, handling explosive solvents and, you know, explosion-proof walls. You don't need any of that because our system is based in water. We have a fun thing, which is we call it the ECO Innovation Lab in our facility in Redwood City. We take people on tours of it. These are people that are used to looking at PAC chemistry. They're used to looking at siRNA production. We take them past it and I say, is that it? We say, yep, that's it. 'Cause it is so damn simple. The enzymes do all the work, right?
They don't have all those big stainless steel tanks.
Just diving deeper into that ECO Innovation Lab, like, what have you proven so far? What's the next catalyst? Is it pilot scale, GLP material? And when can we expect to see that? What's kind of the roadmap to scaling all the way up to like GMP?
Right. What we're saying this year, and we use the TIDES meetings as sort of our next sort of technical rollout. Those happen in the U.S. in May and then Europe in late fall, sort of October, November time. First of all, we're gonna be showing process parameters as good or better than phosphoramidite chemistry. We're saying, you know, even though the competitive technology's been out there in terms of, you know, purity and yield and all the rest of it for 40 years, we're gonna be better. We're gonna be better now after, you know, two years of trajectory versus 40 years. That's really, really important. What we've also been showing is increasingly tricky sequences that we can actually make using our platform.
Remember, our platform is a tethered enzyme that is the polymerase plus a blocking deblocking system where you rotate iteratively past it. It is the same enzyme that adds all the different nucleotide building blocks sequentially. It is the sequence you introduce them. That is why we call it the universal enzy me. What we will be showing later this year is inclusion of additional modified nucleotides. We really can show it is not just that we can address the current things that are in the market with this platform. We can also address the next wave of modalities. You know, this is something people can get excited about.
We've also got, believe it or not, tomorrow in Geneva at a thing called the RNA Leaders Conference, we're doing the next rollout of our ligase platform, which is coupled to machine learning, as you were asking about earlier, where we've built a capability where we can look at an RNA molecule and we can actually, using machine learning, artificial intelligence, and our libraries, work out where the optimal place in that molecule is to make what we call the nick, which is where you separate the components and to pull an existing ligase out of our library that will be optimized to doing that reaction. This changes the cycle time from someone coming to us and saying, can you build us a ligase for our reaction?
That's going off and scratching our heads and taking, you know, a few weeks or months to do it to say, you know, do you want it Tuesday? You know, here it is. We can, we can do this. And often, you know, that we can do the modeling in, within the same working day. It's, it's very cool stuff. And it's really, you know, addressing one of the customer requirements, which is responsiveness and cycle time.
Just from an operational standpoint, like you are positioned to reach, you know, cash flow breakeven positivity with cash on hand. Does that include any type of corporate development M&A strategy, or is this just all hands on deck on developing and commercializing ECO Synthesis?
This is all hands on deck on several things. We have said we get to cash flow positive within our current means at the end of 2026. That is next year. We do that by continuing to grow the pharma manufacturing heritage business, which is, you know, spinning off cash into the broader organization. Plus we have to convert some of our current discussions into contracts around the ligase, and that gets us to lift off. The longer term, the real sort of hockey stick in terms of our revenue potential happens around the full ECO platform when we move from stitching together people's fragments that they have made elsewhere to actually making the entire molecule.
That has many different potential revenue streams from reagents and enzymes and raw material supplies to the big players that wanna do it themselves, all the way to supplying directly siRNA made at Codexis to the customer.
Just on the past quarter, you had, you know, the remainder of a retainer with Pfizer that was supposed to be delivered, and you did receive that, but it was actually converted in kind of an upfront of a new deal. Could you just give kind of the details to that deal? Is that type of deal something we can expect to see going forward, or is this a one-off that was opportunistic?
The specific around Pfizer was one-off and opportunistic. And, you know, we were super pleased with that because it converted something that was a, you know, an accounting sort of glitch into, into actually a future source of revenue. And this was the ability to, you know, deepen the relationship there. There will be, you know, milestones and downstream payments associated with that. It was, we saw that as a significant win 'cause it was something that was gonna finish that now has become a revenue stream in the future. We are looking at, you know, being as creative as we can about deriving value from our, from our platform. So while I've talked a lot about our focus in oligonucleotide manufacture and evolving enzymes for that, we're perfectly willing to build enzymes for other people to enhance their business.
I mean, I'd point towards the transposases we made for seqWell. You know, we are not in next generation sequencing and we are not into building kits, but they could tell, they could express to us what they needed. It looked like a tractable enzyme engineering challenge. So we did. And, you know, now we will be deriving downstream economics from that, even though it's not part of our focus business. But what we have to be super careful of is we get paid along the way.
Obviously, ECO Synthesis is kind of, you know, the main focus here and what's really exciting, but you just referenced like an AGBT update. There's some other stuff going on. Just in the next 12 months, what are the catalysts to really focus on? Are there things like that that present some upside?
Really the things that I'm looking for is the continued growth in the overall revenue. I start at a very macro level in terms of how I think about the company, that the revenues are going up significantly faster than the expenses are, and, you know, that we have a line of sight into new markets. Now, some of the things that we are trying to do this year is converting, even in the pharma manufacturing, some of the mid-sized pharmaceutical companies into long-term contracts. There's people that don't tend to have their own in-house biocatalysis capability. Converting those seven, you know, customers that we've got hovering into real contracts around the ECO platform, being able to announce development deals there, the stuff that I talked about at the different scientific meetings. We've got quite a lot on this year.
For someone looking to, you know, invest in Codexis and gain conviction around siRNA, just really expanding into these larger indications, what are some of those specific indications that you look at, maybe even specific programs or updates to kind of look ahead the next 12, 24 months and gain that conviction around the space?
Right. Let's go right back to what does siRNA actually do? Once it gets delivered to the target cell, it will knock down a specific protein in that cell for six months with a single dose. That's what makes it a beautiful modality if you can get it to the target cell. The building of the siRNA itself, that's one problem, just as big a problem as the targeting. There is a lot of noise around that right now because getting siRNA into the central nervous system, skeletal muscle, cardiac muscle, lung, and other places opens up a whole vista of opportunity. Alnylam was talking about this at their R&D Day last week. Super excited. They intend to be in every major tissue by the end of the decade. Novo just signed a $350 million deal looking at targeting moieties. This is the coming wave.
One of the beautiful things for us is we have an ability to attach those targeting moieties in a very elegant way to siRNA molecules. This opens up the world of, you know, Alzheimer's, cardiovascular disease, diabetes, all these places where you can actually move away from the liver into systemic diseases. We actually see a lot of validation and reinforcement that, you know, siRNA is gonna be the modality of choice for managing a lot of chronic diseases 'cause you only have to give it twice a year. That is really compelling as, you know, I'm speaking now as a physician when you're trying to manage patient compliance and you can bring 'em into the office, dose 'em twice a year, forget about it. They don't have to worry about their lipids or their hypertension or whatever. That's groundbreaking.
That's why this is gonna be a super important class of medicines if the supply is available. That's the thing that we are working on.
That's awesome. Yeah, we're at time. So thank you for the time, Stephen. It's an exciting space and looking forward to all these updates the next coming year.
Thanks, Chad. Thanks very much.