NEB Podcast #78 -
Transcript
Interviewers: Lydia Morrison, Marketing Communications Manager & Podcast Host, New England Biolabs, Inc.
Interviewee: Craig Martin, Ph.D., Craig Martin, Ph.D., President and CSO, Waterfall Scientific, Emeritus Professor, Department of Chemistry, University of Massachusetts Amherst
Lydia Morrison:
Welcome to the Lessons from Lab & Life Podcast brought to you by New England Biolabs. I'm your host Lydia Morrison and I hope this episode offers you some new perspective. Today I'm joined by Dr. Craig Martin, president and CSO of Waterfall Scientific™, an Emeritus Professor in the Department of Chemistry at the University of Massachusetts Amherst. Dr. Martin has studied RNA polymerases for about 40 years and he joins us to explain DuoTether™ technology and its implications for RNA production efficiency.
Craig, thanks so much for joining us today. I know you have a background in academia. Could you tell us about your academic experience and how that research has led you to where you are today?
Dr. Craig Martin:
So my lab at the University of Massachusetts Amherst has studied the simple model system RNA polymerases called T7 RNA Polymerase. And we've studied that for almost 40 years, looking at mechanism, structure, function, how the polymerase works. We never had any goal of commercialization, but it turns out that that foundational knowledge that we and some others in across the country have developed in this system have been hugely foundational in terms of now using this enzyme to make RNA for therapeutics.
The need to scale and achieve GMP synthesis manufacturing has become huge and continues to grow. And understanding this is not an enzyme where a substrate comes in, it's converted to product, then you're done. It's a complicated process. And it's important, as we develop systems for manufacturing RNA, that we appreciate that it's a complex process and that we understand all the details.
Let me add one more thing to that.
Lydia Morrison:
Sure.
Dr. Craig Martin:
So about a little less than 10 years ago, I was in the position of using this enzyme to make RNA and we started running into problems because we were starting to try to make large amounts of RNA. And so at that point, my lab pivoted to using our foundational knowledge to developing better ways to make RNA. We were aware of the burgeoning RNA therapeutics and vaccine industry and we thought we could have an important impact on that. Then the need to scale and achieve GMP synthesis manufacturing has become huge and it's important to understand the system as you're developing that platform.
Lydia Morrison:
So that brings us to current day and you have co-founded a company, Waterfall Scientific. Could you tell us about that company?
Dr. Craig Martin:
Waterfall Scientific aims to commercialize technology that was developed at the University of Massachusetts, something we call DuoTether technology. This includes the use of DuoTether technology, both in match mode and in continuous flow, a scalable flow reactor.
Lydia Morrison:
Could you explain the DuoTether technology to our listeners?
Dr. Craig Martin:
So the first part of Tether is that we are tethering the RNA polymerase to the DNA at the site that we want it bind to initiate transcription. Now the polymerase does that binding pretty well on its own already, but by having a local high concentration of polymerase around that site, we drive promoter binding at the expense of polymerase doing other things, such as making double-stranded RNA. That's one part of DuoTether.
The other part of DuoTether is imagine that we now have the enzyme DNA complex that is effectively a catalyst for RNA generation. We can now tether that to a surface and now we have a surface immobilized catalyst. And so now we can reuse that catalyst many, many times, as in our magnetic beads batch application of the DuoTether technology. Or we can use it in a continuous flow reactor, as many chemists have done in non-enzymatic systems to generate RNA continuously with the same enzyme and DNA.
Lydia Morrison:
And obviously, the production of high quality RNA is really important for things like RNA therapeutics and RNA vaccines. I'm curious, what do you see is the potential of those therapeutics and vaccines?
Dr. Craig Martin:
Sure. The potential for RNA therapeutics broadly defined is huge. I'd like to see mRNA therapeutics replace all the current protein biologics. But more importantly, there are protein biologics that don't express well in vitro. Problems with folding, post-transitional modifications in the in vitro production of those proteins. Or proteins that don't store or transport well. It makes much more sense to deliver RNA to the cells and now have the patient's cells make the proteins because the cells know how to make the proteins, add all the modifications and stability is now native inside the cell.
As we saw with the pandemic response, mRNA is much more nimble than other sorts of therapeutics. And the real promise of RNA therapeutics is right now with a protein biologic, different protein biologics are very different. It might be a small protein, it might be a transmembrane protein. It might be a multi-subunit protein. And even for the same sized protein, the formulation, the storage, the expression is all going to be different and so there's a lot of work that goes into getting that done well.
With RNA therapeutics, you simply change the sequence. So a membrane bound protein or a soluble protein of the same length, the RNAs look almost identical and so the packaging is the same, the production's the same, everything's the same. And so that allows a much more nimble response and it allows a much less expensive development because we don't have to reinvent everything for every new therapeutic. Personalized therapeutics, the potential of them is huge. And that's really only addressable with mRNA technologies, again, because we can switch the sequence and come up with a new therapeutic very, very efficiently and safely.
Another potential in the mRNA therapeutics and vaccine space is something called self-amplifying RNA. So that holds the promise of very long times between dosing. I should note that with proper formulation, even non-self-amplifying RNAs have demonstrated long durability in the cell, but this has the ability to extend that even further. So those are the things I can see in the near-term horizon that are important for the potential for RNA therapeutics and vaccines.
Lydia Morrison:
Lots of promise that you outlined there. But I'm curious, what are some of the current manufacturing challenges? Obviously, creating these RNA therapeutics and vaccines really high purity RNA at a larger scale than traditionally has been produced. Could you talk a little bit about what kind of challenges there are in producing that high quality RNA?
Dr. Craig Martin:
Certainly. So for the mRNA manufacturing side of things, there are three main challenges. One of those is purity. I entered the space in particular because current manufacturing approaches create double-stranded RNA as a byproduct, an unintentional byproduct of the reaction. This limits mRNA expression in vivo and it's important to recognize that the RNA polymerase doesn't normally do this. It's only because we're trying to make the RNA, asking the RNA polymerase to make large amounts of RNA. So double-stranded RNA contaminant is one challenge to purity.
And then there's more routine purity problems. You have to get rid of the DNA that was used to make the RNA. You have to get rid of the RNA polymerase that was used to make the RNA. There's also a couple enzymes that are added either during or after the reaction, pyrophosphatase and DNAs. All of these things have to be removed from the pool of RNA after production of the RNA.
Another manufacturing challenge that's related to that of course is expense. All of these purification steps increases the number of what we call unit operations to achieve removal of the above contaminants. The purification also reduces yield and so that increases the expense of the product. All of these purification steps involve people, they all involve sometimes very fancy and expensive equipment and so those are expensive components. This all to remove things that we introduced into the system or that we unintentionally introduced into the system because the RNA polymerase made double-stranded RNA. If we can make the RNA without those contaminants, we can drop the expenses substantially.
And then the last challenge, or one of the last challenges in manufacturing of RNA therapeutics and vaccines is lengths of RNA. So when people try to make very long RNAs, such as is needed in self-amplifying RNAs, often they have trouble making RNA using our system. We've recently discovered that that's actually a result of the conditions as well. The enzyme can make very long RNA, but not when you try to increase yield and not think about the effects of what you're doing to increase yield. So in particular, we have some evidence in my lab at UMass that when you have more than one RNA polymerase on a piece of DNA, normally they'll tread along one behind the other and that's just fine, and they won't interfere with each other. But if the leading RNA polymerase flows down and the trailing RNA polymerase catches up, it can knock that polymerase off. The longer the RNA, the more likely this is to happen just statistically. And so we have evidence that that is limiting, contributing to the inability or the lower ability to make long RNAs, such as is needed in self-amplifying RNAs.
Our DuoTether system is obligatorily one to one. There is no chance of a second polymerase coming up behind one polymerase and knocking it off. And we've demonstrated that we can make very long RNAs with the DuoTether system.
Lydia Morrison:
Yeah, I could see how the system both decreases the need for cleanup because it makes it much easier to pull away the DNA and the RNA polymerase, as well as decreases the amount of double-stranded RNA that you're seeing. I don't know, is it even possible to say that it makes it impossible to create double-stranded RNA? Does double-stranded RNA still happen?
Dr. Craig Martin:
I would like to think that it completely eliminates double-strand RNA in production. I'll be happy if we can get it down substantially lower than what people have now. Again, different applications are going to require different threshold levels of double-stranded RNA. If we can make all of our RNA at well below the most stringent threshold, then that's enough.
Lydia Morrison:
So can you explain a little bit more about why it's important to get rid of the double-stranded RNA?
Dr. Craig Martin:
Sure. So first, it's important to understand what the double-stranded RNA is. The RNA polymerase is actually making it and what happens is that your correct RNA gets released from the protein and now the protein has two choices. It can rebind the DNA and make another piece of RNA, which is what you want it to do. Or it can rebind the RNA and now start using that RNA as the template. That is the source of double-stranded RNA.
The reason double-stranded RNA is a problem, or the fundamental reason why it's a problem in therapeutics, is that most of the viruses that infect us, cold viruses, that sort of thing, are what are called double-stranded RNA viruses, so their genome is a long stretch of double-stranded RNA. Because most of the viruses that infect us are double-stranded RNA viruses, we have evolved in our immune system in our innate immune response a series of receptors that recognize long stretches of double-stranded RNA, 40 base pairs or longer. And they tell the cell, "If you see big stretches of double-stranded RNA 40 bases or longer, turn on the immune system because you're being attacked." And so that leads to inflammation, which if you're being attacked by a virus, that's a good thing. But of course, you're not being attacked by a virus, the cell just thinks it's being attacked by a virus.
The other thing the cell does to protect itself when it thinks it's being attacked by a double-stranded RNA virus is it starts turning off the production of protein from RNA translation. And of course, we're putting RNA in as a therapeutic in order to produce a protein and if that RNA is now turning off protein production, we now have a therapeutic efficacy or viability problem.
So for that reason then, we need to get double-stranded RNA down to low levels. Low varies a lot on the application. If your therapeutic requires high dosing of RNA, then you need that double-stranded contaminant very low. Or if you're going into cells that are particularly immune responsive, you need very low double-stranded contaminant levels.
Lydia Morrison:
I think that makes sense. Obviously, you don't want to introduce an innate immune response in someone when you're trying to serve them a therapeutic, so really important to keep those double-stranded RNA quantities low. Can can you detect the double-stranded RNA within your production?
Dr. Craig Martin:
Yeah, so that's both easy and hard. It depends how high a sensitivity you want. So to taking large amounts or in particular long lengths of double-stranded RNA, 4 or 500 bases, that's relatively easy and the field uses things like immune dot lots or what are called an ELISA assays or there's some new assays coming on the market. But what really matters is getting that RNA level down low enough that it doesn't trigger the innate immune response in the cells that you're targeting at the amounts you're delivering.
And so the ideal metric should be zero double-stranded RNA. So we'd really like an assay that is both sensitive and easy, and can detect very low amounts of double-stranded RNA. I don't think we have that yet and I hope that others will be creative and come up with those. But it's a fundamentally hard problem. So if you think about often, when we draw cartoons of double-stranded RNA contaminants on RNA, we'll show a 200 base RNA with a 50 base double-stranded RNA extension on it. If the problem were that, detection would be easy. They're easy to distinguish, those two things. But more likely, it's something like we have a 5000 base RNA and we now have a 50 base extension. That's a 1% compositional different and so detecting that is very, very hard. So current technologies can maybe detect really long stretches, by I don't know, hundreds or something. But detecting a 1% compositional difference and something like RNA is just a fundamentally hard analytical problem.
Lydia Morrison:
So ideally, you'd create a method that avoids the creation of double-stranded RNA all together?
Dr. Craig Martin:
Correct, right. So the goal is, our goal is to make detection of double-stranded RNA not important because we know that we're routinely making very, very low levels and it's not a contaminant that people even worry about anymore. That's our ideal.
Lydia Morrison:
Yeah. And I'm curious, how do you imagine the DuoTether method scaling for use in therapeutic manufacturing?
Dr. Craig Martin:
So, the scaling of RNA production has been important to us from day one. People often hear that I'm working on a fluidics or a micro-fluidic reactor for RNA production and they think, "Okay, that'll be fine for personalized therapeutics, but you'll never be able to make large amounts of RNA." That's actually not true. I can walk you through a little bit of the logic, why we think we'll be able to scale really well.
So right now, we're working with a prototype flow reactor that will probably work well for R&D production for RNA and the reactor's about 100-microliters in volume. But that should perform like a much larger batch reaction. So instead of a two-hour reaction in a one-mill volume in a batch production, for example, imagine a 100-microliter flow reactor that can process 20 reactor volumes in two hours. Now a 100-microliter reactor is performing like a two-mill batch reaction.
But it gets better than that. If you imagine a reaction progress curve for typical batch production, the rate of synthesis of RNA starts off fast and then slows down, and so over the last hour it's not producing as it was during the first hour. And in fact, it's leveling off to zero at some point. And that's because you're running out of NTPs and that's what you want to happen. You want to use up NTPs. But now if you imagine a flow reactor where we continuously supply, resupply the substrate, nucleoside triphosphatase to the reaction, the reaction never stops and more importantly, it continues at maximum production the entire time. So now over that two-hour reaction, we're making more RNA because it's making RNA at the full rate the entire time.
But it gets better than that because in a batch reaction, that reaction slows down and stops after, for example, a couple hours. Our reactor can run for much, much longer run times. So in the DuoTether batch systems that we published results for recently, we can do 20 repeat reactions over five days and we're getting constant production the entire time. You can now imagine our 100-microliter reactor that's performing like a much larger reactor and now it's running for 12 hours, 24 hours, five days continuously so we're producing much more RNA out of that small volume.
Another benefit of this of course is all of that production is done with the same enzyme and DNA. We're not throwing away the enzyme and DNA at the end of two hours. So there's a cost savings there as well. Even in our bead-based system, which is our initial solution to production of RNA, like I said, we can do repeat rounds of transcription over many days, all at the same emboli, and enzyme and DNA.
And finally, with current batch reactions or in current batch reactions, let's say a five-mill reaction, when you want to scale that for clinical trials, scale production for clinical trials, typically what you do is you just make the reaction bigger. You go from five-mills to 50-mills, or larger and this is called scaling up a reaction. This brings tuning challenges. As you dramatically increase the size of a reaction, now the reaction doesn't perform the same as it did at the smaller volume. So a lot of effort has to go in at that stage to making the system perform well and hopefully it performs the same way and the same purity levels that it did at the discovery stage. I emphasize hopefully. In contrast, with the flow reaction, we can scale out the reaction. So we don't make out reactor bigger. When we want to make more of a product, we simply make 50 or 100 parallel reactors of the same shape and size that we had at the discovery stage, we just parallelize that reaction.
And chemical engineers love this because those reactors now perform exactly the same as they did at the early discovery stage. So you get much better reproducibility of the product as you scale production with very little tuning required, so less expense. So in general, chemical engineers would rather scale out of flow reaction than scale up a batch reaction.
Lydia Morrison:
It sounds like your system has actually addressed a lot of the current manufacturing challenges in the RNA space, which is super promising. I'm curious where you imagine the field of RNA therapeutics, and vaccines and personalized medicine in the next five to ten years?
Dr. Craig Martin:
In five to ten years, I imagine a world where personalized therapeutics are widespread and relatively low cost. I also imagine that the breadth of conditions for traditional mRNA therapeutics expands widely. And that will happen because of two things, higher efficacy and lower development and deployment costs. So if we can get mRNAs that perform well more reproducibly, and if we can lower the cost of both the development stage and in the production stage, that will allow therapeutic manufacturers to tackle a wide-range of RNAs. And again, also the fact that you can just change the sequence of RNA should also enable a much more rapid development cycle.
I imagine in five to ten years, pandemic responses will be rapid and efficient. We already saw that recently, but I hope that to become much more widespread and common. I expect therapeutics, because of the reduced expenses in particular in manufacture and deployment and local production, I expect therapeutics to be more democratically spread across the globe. I also expect long non-cloning RNAs to be a part of the therapeutic landscape in five to ten years.
We've been talking today about mRNA-based therapeutics, but there's a lot of research going on right now in academia and in some companies looking at diseases that involve long non-cloning RNAs or could be treatable with long non-cloning RNAs. And my enzyme doesn't care whether the RNA is an mRNA or a long non-cloning RNAs, it can make them all. So I expect that in five to ten years, that will start to become a part of the therapeutic landscape.
Moving a little bit away from my expertise, the manufacturing, I expect that targeted delivery and expression of RNAs will increase dramatically. Targeted delivery expression of therapeutics in general has been the Holy Grail for a long, long time and I think that because of the ability to program intelligence into RNA, that's a whole other subject, I think that we'll be able to control expression of these therapeutics better than we can with current therapeutics. And so I hope that, particularly for things like cancer therapeutics, but for therapeutics broadly speaking, we want those to have their action only where they're needed and not other places and that's a huge goal that I think we'll become more achievable in the next five to ten years.
And then finally, I hope that mRNA therapeutics in the next five to ten years will become much more transportable and we'll have storage at or near room temperature. People think that RNA is an unstable molecule, that's not really true. It's unstable under certain conditions and in conflating the RNA, stabilizing it, I think we can develop therapeutics and there are already people developing things like this, we'll be able to develop therapeutics that can be stored and transported as well or better than current protein biologics, and hopefully as well as small molecule therapeutics.
Lydia Morrison:
Lots of potential for the future. Craig, thank you so much for being here with us today to share your new methodology for RNA manufacturing and your perspective on its potential impact for the future of RNA therapeutics and vaccines, as well as your thoughts around the future of RNA research and therapeutics and medical applications of RNA. We really appreciate your time, and thanks for being here with us today.
Dr. Craig Martin:
I'm happy to be here.
Lydia Morrison:
Thank you for joining us for this episode of the Lessons from Lab & Life Podcast. We invite you to check out the episode's transcript on neb.com for helpful links from today's conversation. And we hope you'll join us next episode, when I'm speaking with the winner of the 2025 Genes in Space competition, Nitya Johar of Skyline High School in Sammamish, Washington. Nitya and her mentor Marissa Morales will explain the experiment she is sending to space.
