Episode 69: Dr. Martin Burke on Democratizing Molecular Innovation

Introduction to Let's Talk Chemistry

0:00Hi, you're listening to Let's Talk Chemistry, a podcast by Chemtalk. On today's episode, we interview Dr. Martin Burke, professor of chemistry at the University of Illinois Urbana-Champaign and founding director of the Molecule Maker Lab. Dr. Burke is known for pioneering block chemistry, a modular approach to synthesis that opens up molecule building to non-specialists, robotics, and AI. This has paved the way to creating

0:32new treatments and digital platforms, where making molecules become easily programmable and accessible to anyone. We hope you enjoy!

Interview Introduction

0:50Hello, I'm Porvi from Let's Talk Chemistry, a podcast by Chemtalk, and I'm joined by Nina, and this is your primer for our interview with Dr. Martin Burke, a professor of chemistry at the University of Illinois at Urbana-Champaign. Dr. Burke works at the cutting edge of automated molecular synthesis and artificial intelligence. His advocacy is for the democratization of molecular innovation, essentially focusing on taking synthesis out of specialized labs and allowing anyone to create molecules.

1:21Right, and this is something that Dr. Burke is super passionate about. But before getting into the robotics and AI, let's talk a little bit about the drug discovery and synthesis process. Our audience has probably heard of organic synthesis and is somewhat familiar with how molecules are constructed. But it's easy to overlook the intricate steps and specialization that go into traditional chemistry. Yes, and the traditional field of synthesis is so specialist dependent. It can be so frustrating

1:52to treat every single molecule as a unique problem that requires thousands of different reactions under customized conditions. And as we'll hear later in our interview with Dr. Burke, his lab decided to completely flip the script. Instead of trying to build complex machines that can mimic artisanal chemistry, they decided to create a simplified chemistry that machines can easily execute. And adding on to the conversation of automation, in many of his works, Dr. Burke specifically mentions the concept of molecular prosthetics. What does his research mention about treating human diseases

2:28specifically? So, while most traditional drugs work by binding to a hyperactive protein and turning it off, many human diseases are actually caused by a deficiency or a loss of a protein's function. And Dr. Burke's concept of molecular prosthetics involves finding small molecules that can actively replace or replicate those missing cellular functions. And through this approach, his lab has actually been able to successfully transition these therapies straight into clinical trials. The intersection between automated technology and clinical medicine is truly a complex space that affects

3:04every one of us. So stay tuned to hear from Dr. Burke and his direct perspective on these issues as a leading physician scientist and innovator.

Dr. Martin Burke Introduction

3:19Hi, and welcome back to Let's Talk Chemistry. Today we're joined by a very special guest, Dr. Martin Burke. Thank you so much for joining us. And can you start off by introducing yourself and telling us a little bit about what you do? Sure. Hi, Nina. Thank you for having me on. And sure, my name is Marty Burke. I'm a professor of chemistry at the University of Illinois at Urbana-Champaign. My lab has developed block chemistry, B-L-O-C-C. The double C is a wink to the carbon-carbon bonds that we love to make.

3:53We understand that carbon-carbon bond-based chemical matter is really, really powerful. You can do all kinds of important, exciting, and impactful things with this class of chemical matter. And we are very excited. We figured out a way to make such molecules that's very friendly to robots. It's also very friendly to artificial intelligence. And perhaps most excitingly, it's friendly to anyone who wants to do it. So our mission is to democratize molecular innovation. And we're very excited that

4:24we're on our way to achieve that and looking forward to partnering with anyone who wants to join forces to create molecules that matter. Awesome. Great. And we'll definitely get into that more

Dr. Burke's Background

4:32as we go into the interview. I just wanted to learn a little bit more about your background and how you initially got interested into chemistry. Sure. So I grew up and I thought I wanted to be a doctor. So I was very much fascinated by medicine. I had an amazing physician when I was a kid, Dr. Ford, who I was very much admired and kind of looked up to and was inspired by. And I also grew up in a very small town in northwestern Maryland, about an hour northwest of Baltimore. And only in retrospect, kind of realized that we actually didn't have any

5:05scientists in our town. So we had a lot of amazing people and it was a wonderful place to grow up. But we actually had more cows than people. And so it's funny how sometimes you grow up in an environment, you get exposed to certain things and you dream about certain things and you don't dream about others based on your experience. I think that's a truism. And so I always wanted to be either a doctor or a baseball player, but the Orioles never called. So I actually ended up sticking with doctors. So I went to undergrad pre-med thinking I wanted to be a physician. But then I was very fortunate to go to Johns Hopkins and met all these amazing scientists and discovered chemistry and as

5:40a way to potentially think about having a really huge impact. So I got really excited about that. I think it was the impact factor that was a huge draw for me. And so ultimately went to graduate school as like an MD-PhD. But a large part of that was like the non-decision. So it kept options open. And so, but once I got into a lab and started learning how to make molecules at the graduate level, I was just absolutely blown away by the kind of power of organic synthesis and worked with Stu Schreiber, who was doing just phenomenal work in terms of thinking about making molecules that could have a huge impact on the world. So I think in many ways, I was just

6:15absolutely drawn to the power of molecule making and decided to not become a doctor, but become a research scientist and then started my lab here at the University of Illinois.

MD and PhD Impact

6:25Awesome. Yeah. And did you find that having the MD and a PhD kind of shaped how you do research today? Yeah, definitely. I would say I loved the whole experience. It was just like a phenomenal learning experience. You go through certain periods of life where your learning curve is pretty much vertical. And so, yeah, I was, I would just learn so much. And I think for me, the part where it really made a huge impact was around problem selection. So, you know, as chemists, I think, you know, also one of the things I always tell my students is like, when you learn how to make molecules, you make yourself a very powerful person. I think that's one of the coolest things about chemistry and

6:58about kind of molecular synthesis in general. And so in order to try to, you know, maximally utilize that power, we all have to figure out which problems we want to work on. Right. So for me, the MD was kind of like just a treasure trove of experiences to start to think about areas of human health where hopefully small molecules could make a big impact. And so I'll just give you one example. When I was on my clinical rotations, once I met this amazing young woman, she was a college student in Boston, she had cystic fibrosis. And I was in her, in her kind of room, drawing up on the

7:33whiteboard, probably in excruciating detail, everything I had just learned in classes about her CFTR protein and how it was not working properly. And at some point in this conversation, you know, she stopped me and says, look, it sounds like you know exactly what's wrong with me. Why can't you fix it? And I remember thinking like, oh my gosh, like that is such a profoundly important question and actually gets to the heart of a really interesting area where, you know, we know small molecules are really good at binding the proteins and turning them off. But in her case, the problem wasn't excess of a protein function, but in fact, some kind of deficiency of protein function.

8:06So she's missing an ion channel in her airway that's responsible for transporting chloride and bicarbonate. And so that got us thinking about, okay, might there be a way? We know how powerful small molecules can be. Could we discover small molecules that could act like a missing CFTR channel protein, for example? And that turned out to be like a really important theme, a project that we worked on for many years. And I'm very excited to tell you that's actually now entered clinical trials, some of the work that we did using originally the block chemistry to get that started. And so that's just one example where the kind of experience as a medical student did have a

8:40really big impact on the types of problems that we picked to work on. Yeah, that's such a great achievement. And also like, so a huge congratulations for that. I feel like your experience through college or wherever your interests take you, they all like build up. And it's amazing to see that you get, you like your lab developed a drug that is in clinical trials right now. We'll definitely talk more about that later. And I wanted to get into

Block Chemistry Explanation

9:09block chemistry, which is something else that your lab heavily focuses on. Can you tell us a little bit about what exactly is block chemistry and what kind of inspired you to get involved into this? Sure. Awesome. Yeah. So one of the things amazing about small molecule synthesis, as we already talked about, is extremely powerful. You know, you can either make molecules that are already known to have a certain function, or you can make new molecules, right? And then try to find new function. And all of that together, given the tremendous amount of impact that small molecules

9:43can have on the world, whoever can make them has tremendous power. I think that's actually something that is really, you know, fascinating. And so it's very interesting to think about right now, there's like 8 billion of us on the planet, okay, which is phenomenal. When you try to wrap your head around, that's 8 billion imaginations, right? It's like our greatest natural resource on our entire planet. And yet, I've done the math, okay, those of us who can actually make molecules, we fit in one building. Wow. So 8 billion imaginations, and those of us who could actually wield this power are hanging

10:19out in one building. And you know, unfortunately, sometimes we all come from similar backgrounds or have similar kind of, you know, histories. And like, that's a huge missed opportunity. If you just think about the huge gap between this amazingly powerful thing, and then only one building full of people get to do it. And so we want to shatter that. That's where we started to get very, very excited. Like any field that figured out how to shatter those types of barriers to entry, and allow in a sense, like democratization of their ultimate discovery engine, everybody wins. There's so many cool examples

10:50of that in history. So here's a case where small molecule synthesis is like this ultimate superpower, and a very, very small fraction of the earth, you know, people on planet have access to it. We want to shatter that, okay? Because we think if we could democratize small molecule innovation, the world becomes a really exciting, better place. And so I think that's our mission. And when we started dreaming about that, you say, okay, well, you got to figure out how to transform the field from this current, very artisanal, one at a time, specialist dependent, manual process to something that could

11:27be very friendly to robots. Okay, that was our first takeoff point. And it's really interesting to think about, you know, there was a time where it wasn't clear if humans could go into a lab with their hands and make these complex molecules. That was an unanswered question. And so it was really important to do whatever it takes to try to make them. And then there's these like unbelievably important, you know, pioneers in the field, like R.B. Woodward, for example, who, you know, showed the world that this was possible. And that got all kinds of exciting momentum and imaginations about, you know, how might we do this better? And there's like 100 years of history then,

12:02at least, you know, down the roads as it goes all the way back. And I think that, you know, he was such a flamboyant, creative, artistic person that it led to almost like a culture in synthetic organic chemistry that, that almost emulates that kind of artistic expression of oneself every time you decide to make a molecule. The process of planning that was systematized with E.J. Corey's retrosynthetic analysis, which, you know, again, ultimately it's the, by design, each molecule becomes its own unique problem. And then you figure out in his, you know, this case by

12:33taking it apart in order to figure out how to make it in the forward direction. But ultimately, it all lands in the same place, which is artisanal customized synthesis for every individual molecule. If you start dreaming about a robot that can do this, it doesn't fit. Like there's a huge gap between that artisanal approach and the goal of automation. And so we realized that early on. And what we decided to think about was, so rather than try to build robots that could wrap around that

13:05highly artisanal world of synthetic organic chemistry, where there's like thousands of different reactions and each one run under a thousand different conditions. And there's like million different starting materials where people have tried, but I would argue, haven't actually succeeded in a way that's scalable to automate chemistry in its artisanal form. So rather than try to build machines that can do chemistry, we decided to create chemistry that machines can do. That was the key flip. And so then say, all right, what if you just tried to dream up a new way to make molecules that's very friendly to robots? What would it look like? Okay, well,

13:40in the limit, you'd say, well, how about we just use one reaction? So rather than use a thousand, what if we could only use one? What if we could prefabricate the block so that everything you need is dialed in up front, all the functional group oxidation states, stereochemistry pre-installed? And what if we could do this in a way where it's very precise? So you avoid random oligomers and you want to control the, you know, blocks as they get assembled together. Well, there's actually a pretty good inspirational role model with peptide synthesis or DNA or RNA, where this has actually

14:11been achieved and it's been automated. And so then we kind of said, okay, how do we do something similar for carbon bonds, which are, of course, the key backbone of small molecules? So long story short, that's what we invented, was a way to control the reactivity of a boronic acid so that you could take palladium-mediated coupling between halides and boronic acids and do it over and over and over again in a controlled manner and then build up your small molecule just using one reaction iteratively, okay? And this is where my first grad student Eric Gillis discovered that MITA and methylaminodiacetic acid can wrap around a boronic acid, creates a transition from SP2 to SP3 for the

14:47boron center, removes the p-orbital, and now it doesn't react, okay? So it's like an off switch. And so you can take bifunctional blocks, kind of having like a halide on one side and a boronic acid on the other, turn the boron off, and now like an amino acid that's maybe protected with an FMOC or something, you can iteratively stitch the blocks together through carbon-carbon bond formation, okay? So we call that block chemistry. And again, we put double C as a wink to the carbon-carbon bond. And it's actually very exciting because it's very, very friendly to automation, okay? That part was

15:19by design, the automation part we were aiming for. I'll just fast forward and say we kind of got really lucky with the AI part. So we didn't see the AI thing coming, I'll be honest, 20 years ago. It was not even on our radar screen. But the same chemistry that's friendly to the robot because it's very modular turns out also to be very friendly to AI because AI loves systematic modular data. And so because we're building molecules in a modular fashion, we're also creating data that's very friendly to building artificial intelligence models, which is something we're really leaning into. And if you put all of

15:51that together, if it's good for automation, it's good for AI, it's good for anyone. And that's where I think we've gotten very excited. Yeah, that's super interesting hearing you talk about like actually making chemistry that machines can do, like making it easier for machines to synthesize everything. And yeah, it was just super interesting to hear about and how you're really working to like change the field of, I guess, synthesis. And we talk a lot on the podcast about like interdisciplinary collaboration and how you have to like have good communication to, in order to like transfer

16:26information to physicists, engineers and everything. And it seems like the thing that you're doing is like trying to make synthesis not as specialized as like, you don't need a PhD to do this. Like anyone can just bring like two building blocks together and create a molecule. So that's super, super interesting. And yeah, it was great to hear from you talking about that. I was also curious about,

Challenges in Block Chemistry

16:51are there like any major hurdles that you had to go through to get to where you are today in block chemistry and like, what was that like? Yeah, absolutely. There's been a lot of hurdles and I would say we still have hurdles ahead. We're kind of at this very exciting moment where we've seen that it can work. And now we're also seeing like all these opportunities to make it even better, faster, stronger, which I think is an exciting time. So I think the first hurdle, the biggest challenge that we had was just finding a good way to turn a boronic acid off and on that could be very

17:23controlled because that was the key. And this is where Eric Gillis, who was my first graduate student, we looked at lots of different trivalent heterotomic ligands. So the concept that we had drawn up on paper is that it was known that in order for a boronic acid to transmetallate to palladium, you needed that boron to have an MTP orbital. And if you remember, boron is usually SP2 hybridized. And so there's like this MTP orbital. That's where all the chemistry seems to happen for a boron. And then it gets activated and then transfers the carbon. So the idea was simple. We said, what if

17:55we just remove the P orbital? Like if we took a boronic acid and we rehybridized the boron center from SP2 to SP3, in theory, we got rid of the P orbital and it shouldn't react. So then we said, okay, could we find like trivalent ligands that had three different binding sites that could wrap around a boronic acid and like turn SP2 boron into SP3 boron? And could that then lead to reversible attenuation of the reactivity of the carbon to which it's attached? Okay. That was the basic idea.

18:26But when you have these ideas, you kind of draw them up on paper and you have the concept in mind, but now you need an actual real physical thing that does that. So yeah, so we looked at lots of different trivalent heteratomic ligands and we failed and failed and failed. And then we discovered that MIDA actually works really, really well. So that was a very exciting time. The problem with MIDA though, that we ultimately found is that it's stable. The boronates are super stable under lots of conditions, but others that we really want to be able to use, for example,

18:57to do SP3 cross couplings, the MIDA boronate is not stable enough. So then a next really important chapter in the whole story, one of my postdocs, Daniel Blair, who's now at St. Jude, discovered a next generation version of this, which we call TIDA, he discovered it's about a thousand times more stable than MIDA. And TIDA is actually just four methyl groups on the backbone of the MIDA ligand. And for reasons that are kind of surprising and really interesting from a physical organic chemistry perspective, the nitrogen boron bond flips from being dative to now covalent.

19:32And when it does that, it's like a thousand times more stable. And that opened up the door to SP3 cross coupling, which now that gets us into very interesting, like natural product, like chemical matter being made on robots and all these types of things, which was actually a really exciting advance. And then the last one I'll mention in terms of challenge that's in the rear view mirror was in order to automate, you don't just need to be able to stitch the blocks together. You need to be able to purify after each coupling. And it's a really interesting problem. So with DNA, RNA and peptide synthesis, the way that problem

20:07was solved was putting the blocks on a resin and then washing away the reagents right after each step. We didn't really want to do that because it kind of limits a lot of what you can have as your first block because it has to have an attachment point to the resin. And unlike DNA, RNA protein, some molecules, you don't have like a common attachment point. Okay. So here's where serendipity can, again, be really powerful. So one day, Eric noticed, Eric Gillis, that any compound that had the mitaboronate attached, when you try to run a TLC, it behaves really weird. So instead of like

20:42normally, right, different compounds move at different, you know, degrees on the TLC plate, we call this the RF, right? With mitaboronates, it was all or nothing. So in certain solvents, they would just stick to the baseline and they never move. You could flush the TLC, you know, solvent three or four times or just sit there. But then you switch to other solvents and they all move up to the top, like with the solvent front. So it was this really weird, like binary illusion on silica gel. And so we were like, whoa, that's like really cool because now you can imagine like

21:13catch and release type purification, which could be very friendly to a robot because you can use the same simple, like, you know, take your product, everything you want still has a titaboronate attached or middle borne, push it through a little plug of silica gel, the thing you want gets stuck, everything else flushes off, then just switch the solvent and the thing you want now gets released. It's kind of like streptavid and biotin in biology or something like that, right? And so that was a very cool serendipitous discovery that led it, made it possible now to automate purification in a

21:43general way. And so now that was kind of the key breakthrough that allowed us to put this all into robots and to make the whole process automated. I think in terms of challenges moving forward, where we're still looking for really great solutions, the sp3 coupling problem is like super, super interesting. There's some things you can do. Actually, Kathy Cruden was the first one to show that you could couple a chiro non-ricemic pinacobronic ester with retention of stereochemistry. She has amazing Jack's paper from like 2009, where she showed that first time. And then the field has exploded. And now there's so many cool examples of sp3 coupling, but we're probably like 10% of where

22:18we could be in terms of what's possible in that space. And I think as sp3 couplings, particularly stereospecific sp3 coupling, gets more and more rich in substrate scope, we're going to be able to get more and more natural product-like molecules kind of flying off on the robot. We also want to learn how to make the very complex polycyclic stuff. So this technique is really good at making linear things, but turns out, of course, you can make linear things that then fold up. And so we're trying to learn the rules, kind of like, almost like an alpha fold for small molecules or something, where you could make linear things. And based on the linear structure, which ultimately

22:52comes back just to the blocks, you could predict how it's going to fold up into a three-dimensional complex shape. That would open up really interesting opportunities for like complex three-dimensional architectures, etc. I think we need more general conditions. I think we need, you know, the capacity to make longer oligomers, which right now we usually make three or four, but we'd like to be able to make longer things, etc. So these are just some of the exciting questions, I think, that lie ahead. But what's currently possible already allows us to access a lot of interesting function. And so we want to put that power in the hands of anyone who wants to wield it. Yeah, that was so interesting

23:23hearing you talk about like discovering Meta and TIDA and about all of like kind of the ingenious solutions that you guys went about to go through these, get through these hurdles. And it kind of reminds me of what you were talking about earlier when you said that there's 8 million or billion imaginations and yeah, just humans working together. That's really, really inspiring. And it makes me think, are there like, I just want to hear your perspective. Are there any discoveries that like humans have to

Human vs. AI Discoveries

23:56find on their own versus like what discoveries can AI systems find that humans might not find on their own? Yeah, so I love this question. And where I've landed in all of this, I think there's something super interesting about human machine interface. So in a sense, intentionally creating AI tools that complement and amplify human creativity and innovation and problem solving and willingness to form teams and join forces. And, you know, if you ask, I think it's a very interesting question,

24:26like what are uniquely human superpowers that could be combined intentionally with what AI is also uniquely good at? And that what makes that interface so interesting? Anyway, this is what the type of stuff that your question prompts for me. So this is what we spend a lot of time thinking about. I would say one in my mind, like really cool, like opportunity is, I don't know about you, but I can't think in 17 dimensions at the same time. I wish I could, but it's not something that I've been able to figure out how to do. Whereas this is a real strength for AI multidimensional kind of thinking and handling data that's that complex and actually be able to find trend lines

25:00in such a complex space is super interesting. So we would love to figure out how to be able to take like human imagination and creativity and teamwork and problem solving and passion. And, you know, there's like these really interesting things that humans are good at. And then combine that with this extraordinary capacity for AI to deal with highly complex multidimensional problem solving. Because if you can actually, in a sense, amplify your own human capability into multidimensional

25:31thinking, but still keep your own passion and energy and creativity and problem solving, there's something there that I think is like really, really interesting. And we should try to do that on purpose. I think that's one. The other one is the ability to predict stuff. So I think one of the things with education we go through, of course, like undergrad and graduate school, we're like, we're trying to learn how to have, in a sense, visceral instincts based on hopefully physics and chemistry and biology, fundamental science, that give us the ability to look forward and predict stuff, right? Like when you generate a hypothesis, it's kind of like a

26:05little, it's like a prediction, right? I'm hypothesizing that this will be true. And then, of course, we do experiments and test that. That ability to look forward and imagine data that doesn't exist yet. That's like really interesting. And I think there's a huge opportunity there to intentionally leverage AI to enhance our capacity to do that. That said, I think there's a huge gap in the way people are currently doing that, that we're trying to think about how to fill. So obviously, like ChatGPT and Claude and a lot of these large language models, they're getting

26:37really, really good at doing certain things. But I would argue that without the capacity to generate new data on demand, there's going to be limits. Like OpenAI and Anthropic, they don't have labs yet. Okay. And I think that that opportunity for, let's say, an AI agent to understand its own limitations, to have a sense of its own uncertainties, and then intentionally request new experimental data to fill in the gaps. And then in a sense, self-learn through that iterative

27:09process is very, very interesting. So trying to create, you know, systems where uncertainty minimization can actually be the target objective function. So you could, in a sense, build new kinds of foundation models that are primarily built on new data generation intentionally per request from an uncertainty minimizing AI agent. That becomes very, very interesting. And I think that's the kind of stuff you can dream about once carbon-carbon bond formation becomes roboticized.

27:35Yeah, awesome. And that seems like a great, is that something that your lab is heading towards? Like making, having robots make more data and then using AI to like look at that data and everything like that? Yes. And we've also realized that these types of things require tremendous resources. And to be able to generate the quantity and quality of data to do things like that, there are certain things that fit in a typical academic lab model, right? And we can apply for grants and we can try to do that as an academic lab. But something like this, where you say you try to build a new type of foundation

28:10model, primarily with new data generated on demand through robots, it just takes a different level of resource to pull something like that off. So we've spun out a new company to try to do that. And in concert, in partnership with the academic community. And this company is called Excelsior Sciences. We just launched recently. And its mission is to build such foundation models that could be super impactful for lots and lots of type of research associated with small molecule science. It's not just medicines. You can

28:41think about the same kind of approach working for things like materials or even like cosmeceuticals or even like quantum sensors or quantum computer components. And like small molecules do so many amazing things. So with the company, the idea is to try to get those scaled resources that could actually build really powerful new kinds of foundation models that could be very, very helpful for problem solving in small molecule science. So we're very passionate about that and excited to have the chance to do that. Yeah, that's so interesting to hear about. And as like a chemist background,

29:12because you come from a chemistry background, now you're working with all like these robots and like coding from AI and everything. How do you kind of like balance this? And how do you make sure that like you know enough to understand other specialists in the field, but like also, yeah, just like balance everything that you have right now. Yes, really interesting question. I think, I mean, at the end of the day, I'm a very ambitious person. I want to try to make an impact. That's my driver. And if the problem that I get excited about kind of pulls me out of my comfort zone, my instincts are to go try to

29:47learn everything I can about that different area so that I can try to figure out how I could leverage the other space. That said, I think it's really great to know what you know and what you don't know. And this is where I love to collaborate. I love finding like people who are really, really good at what they do and then join forces, right, to try to create synergy in terms of how I think about problems. So I'm a huge fan of forming teams and collaborating. And therefore, even if you do like read constantly and learn everything you can, you can also benefit tremendously for someone who's been

30:18thinking about this their whole life or their whole career, right? Because that's really special. And so that's where I think getting teams of people together who come at things from different angles, with different backgrounds, with different experiences. There's just something incredibly special about interdisciplinary research. And so I've become a huge fan of that. And so, yeah, I mean, having no fear to go outside your comfort zone, and yet at the same time, finding friends and collaborators who can help you journey off into the adventure and be able to know what really experienced, thoughtful, you know, career driven, visceral instincts can do,

30:55because that's also special. And I think that's kind of where it gets interesting when we team up. Right. And so inspiring also to learn from other people, and like pick their brains about problems also. Yeah, I wanted to transition a little bit to another part of your work on molecular prosthetics.

Molecular Prosthetics Explanation

31:12Can you kind of explain like what molecular prosthetics is to a wider audience? And how does this kind of differ from like traditional drugs? Yeah, awesome. Thanks. So yeah, this gets back to that story I was telling you about, about this amazing college student I met when I was doing my clinical rotations as a medical student. So the idea of molecular prosthetics was actually driven, primarily inspired by that conversation. But then we got more and more excited about the basic idea. So many human diseases are caused by some type of hyperactive protein.

31:45Okay, so one of your proteins is doing more than it should or doing something different than it should. And that that excess of function is causing a problem. Okay, a classic example is the BCR-ABLE fusion protein that drives chronic myelogenous leukemia. This was the discovery with the Philadelphia chromosome. We get this very interesting kind of unnatural protein that forms in a human where it's like a kinase that gets turned on constantly. And so it's hyperactive and it's

32:15sending a signal to cells, divide, divide, divide, and that leads to cancer, this chronic myelogenous leukemia. So in that case, the idea is you find a small molecule that turns the protein off. And actually most drugs, that's how they work. They bind to a protein and they turn it off. So if your disease is caused by a hyperactive protein, okay, that strategy can work great. And that's why obviously a lot of drug discovery focuses on that. But there's many diseases that are caused not by too much protein function, but deficient protein function. You're missing

32:49something. And in those cases, this classic pharmacological strategy falls apart, right? There's nothing to inhibit. You actually want to turn something on or even just create, you know, replacement function, et cetera. And, you know, there are strategies for that. So people obviously, and there's tremendous momentum and excitement using things like gene therapy or an mRNA and, you know, other types of biological approaches. And I think those are going to continue to advance, which I think is really exciting. But there's still a huge gap. There's a lot of

33:19challenges where we can't currently address with those approaches. And from a global impact, scalable solution type of perspective, small molecules are amazing. Like you find a small molecule that helps make people healthy. Once it goes off patent, it's like the whole world can use it. This is one of the things Jim Flynn always talks about is that Deerfield is an investment firm in New York. Look, for a very short period of time, you could have very expensive small molecule because it's on patent. But as soon as that patent expires, once you discover a small molecule, you can't undiscover it. So it's like the world has it forever as a really nice way to try to treat

33:54human disease. So small molecules are very scalable, practical. Once you know which one is going to do it, you can make metric tons of them and, you know, they can get all over the world. They can be shipped on train cars. I mean, it's like small molecules are really powerful. So the thing we started to think about, and this was again, going back to when I was a medical student, knowing how powerful small molecules can be as medicines, by the way, they also like orally available, don't get recognized by the immune system, penetrate cells, all these unique things that small molecule chemical matter has as its properties. I mean, there's a reason why nature went there. If nature

34:28could solve all of her problems with DNA, RNA, and proteins, she would have stopped. That's kind of how we think about it. So there's a reason why secondary metabolism happened. And so given that there's all these diseases caused by loss of protein function, and we know small molecules are like super powerful as medicines, we started to dream about, you know, could we find small molecules that kind of act like proteins? So in a sense, like, could you replace the missing protein with a small molecule surrogate, almost like a prosthesis on the molecular scale, right? So just like if you lose a,

35:01you know, lose your left hand, a simple prosthetic device can actually give you back a lot of function. And it doesn't have to be perfect. That's actually the cool thing about that analogy is like your right hand can work harder and be creative. And so even a pretty simple replacement for a missing hand can actually provide a lot of recovered function. So then we said, what if that's true at the molecular scale, right? So you don't necessarily need to perfectly replace a protein with a small molecule. You just need to get back some of that function. If that's true, well, then this could

35:32actually be very effective. And so we started dreaming about this idea of molecular prosthetics. That was actually before I even started my academic career. That was my, when I applied for academic positions, that was kind of the primary thing that I was proposing. And we started with cystic fibrosis. It's back to that story I was telling at the beginning. And there was this amazing natural product called amphotericin B, which was known to form ion channels in both yeast and human cells. And long story short, we were able to use the block chemistry to understand how amphotericin works, separate this channel forming ability from its notorious toxicity, which allowed us

36:07to start to think about replacing the CFTR protein and cystic fibrosis with amphotericin or a derivative. And that's the program I was excited to take, has now made it into clinical trials. And so the block chemistry was actually kind of invented for that problem, trying to solve that problem. It just then took on a whole life of its own and turned out to be super exciting as a parallel track. Since we've been actually really leaning into the molecular prosthetics idea, we have another program that Tony Grillo in my lab launched, where he proposed we try to get iron to

36:39move across the bilayers. There's actually very interesting diseases caused by loss of iron transporting proteins. And so he had proposed a whole program in the group to try to find a small molecule that could move iron across bilayers. Long story short, he discovered this amazing natural product, hinakitiol can do it. It's from the bark of the hanoki tree in Taiwan. And so we published that paper in 2017. And that actually, that program also has advanced tremendously and is headed towards the clinic as well, which we're very excited about. And so, and now we think there's like a whole host

37:11of opportunity. Like you start to think about all the different things that proteins do in the body. How many of them could be replicated with a small molecule? And I think that's like a very interesting question. There's clearly going to be limits there, but it's actually probably a lot broader of a space than we all might think, especially if imperfection is enough, which actually opens up the door to kind of, you know, pretty simple, imperfect mimics of proteins, perhaps getting back to physiology in a meaningful way. So we're very excited about the concept and we think there's a lot of, you know,

37:42headspace to do some interesting things there. Awesome. And can you take us kind of through the process of how do you like design these molecular prosthetics and how do you like ensure that they actually have these functions that you're looking for? Yeah. In the two cases I told you about, we were able to discover a natural product that kind of has some of the targeted function. So we didn't design them. We discovered them. And the way we were able to do that was by developing an assay that would respond in a, you know, quantifiable way if the missing protein was, you know,

38:19replaced successfully by the small molecule. So we took, for example, a yeast and knocked out the ion channel that was required for growth. This was like a potassium channel, TURK1, TURK2 and yeast. And then we just started looking at different small molecules to see if these yeast, which couldn't grow because they couldn't transport potassium in that case, could we discover a small molecule that could cause the yeast to grow again? So it's kind of like almost like a functional complementation experiment, but using small molecules instead of proteins to replace proteins. And that's how we

38:51discovered that amphotericin could work. And so then Tony Grill in my lab had actually used a different yeast strain missing in an iron transporter and did the similar type of experiment. And that's how he discovered that henokitio could do it. So that's one way to do it actually is if you set up an assay where a protein is missing in a yeast or some other model organism, and then there's a quantifiable output because of that deficiency of protein function, let's say the cells don't grow. Ah, okay. So you could screen molecules and discover one that all of a sudden the yeast start growing again. Then you've discovered a potential molecular prosthetic, and then you

39:25can go study the mechanism and see if that's really what's happening, et cetera, et cetera.

39:30We're starting to think about the question you're asking, which is a really important one. What if there is no natural product or other molecule already known that does it? How do you design molecules, small molecules that act like proteins? Okay, that's a very interesting question. And that's where creativity can become really, really powerful and start to think about everything that's been learned. I mean, obviously, there's a whole, you know, fields of chemistry, organocatalysis and host-guest chemistry and all. There's lots and lots of interesting things out there where people try to build small molecules that do interesting things. And we think that's like a treasure trove.

40:03It's just now different. Instead of trying to design molecules that do chemistry, you're trying to design molecules that act like missing proteins, which is a very interesting design problem. So we think it's a huge opportunity for creative people to take a run at. You're a nice answer. And talked to before about how you move some drug candidates into actually entering clinical trials. What does it feel like to see like an idea move from the lab to actually actively being used to help patients?

40:35So I think for our lab, we're really passionate about what you just said, is that at the end of the day, we're hopeful that maybe some of the things that we discover have real world translational impact. I think I'm kind of hardwired that way to try to do that. Like when I was in medical school, as I was telling you, I think I had my quarter life crisis when I was like, you know, maybe 22.

41:01You know, we're all trying to dream about what we can do and try to make an impact. And for me, it was, I think, you know, going through like anatomy and pathology class and kind of seeing very real physical thing that the human body is and how this, you know, transient time that we all have, et cetera, et cetera. I mean, it was very kind of existential for me. And like, I just got very, very passionate about trying to figure out how we could do something to try to make a difference in that space. And, and so for us, yes, if a molecule, you know, something we helped contribute to or start something that contributes to leads to a compound that ends up in the clinic and hopefully

41:35makes a difference, that's obviously like just tremendous, like hopeful thing that we aim for. And it's, it's really, really hard to do. It's obviously a big challenge. We approach that whole thing with a lot of humility because to get a molecule all the way across to actually be an FDA approved drug, as you probably know, it's like a huge challenge. But we're very excited that, okay, so five companies have spun out of the lab so far and together they have seven drugs that have made it into the clinic. One of them that we're very, very hopeful about is called Draxon Racib,

42:06which is the treatment for pancreatic cancer that Revolution Medicines, which was the first company that I was able to found is now just showing a big readout from phase three, showing almost a doubling of lifespan in people who had refractory pancreatic cancer. So this is really starting to give us hope that some of these things might make it across the finish line. And I think for the students to see that, you know, work that started in academic lab and really trying to do basic research may have in some cases the potential to translate ultimately to things that hopefully can

42:41make such a difference. It's very, very inspiring. And so, you know, we love in our lab to think about both. Obviously, we love the published papers. And of course, the most important thing, the students build a portfolio of accomplishments and go off and do all these amazing things themselves and launch their careers. But I don't think it has to be one or the other. Like academic labs can do research that hopefully some of it might have translational potential. And for us, this is a real passion. And we're very kind of grateful and excited that some of the work has led in those directions. And we're fingers crossed that some of them will make it across the finish line. Yeah, I hope to see it too. As we wrap up, you spent your career working to make molecular

Translating Lab Work to Clinical Trials

43:18innovation more accessible. What are some projects initiatives you're most excited for right now? And what are you doing to further democratize molecule discovery? Okay, awesome. Yeah. So this is our mission and will stay. So I think for a long time is to actually bring that hope to a reality. So anyone in the world can make molecules and innovate in this space and hopefully find molecules that matter. I'll tell you about some of my favorite things we're doing currently that I think help capture both the ambition and hopefully excitement.

43:53One of the things we've done is create what we call a digital molecule maker. It's kind of like scratch for molecules. I don't know if you know the history at MIT of development of the scratch program to be able to empower kids to be able to code. I learned about this actually because my kindergartner at the time found out I didn't know how to code and he started laughing me at dinner one night. And actually then I think he was kind of embarrassed. It was like, I didn't know how to read. So he's like, dad, please don't tell my friend you don't know how to code. It's like really embarrassing because he had learned it in kindergarten using scratch. And this was Mitch Resnick's

44:26creation at MIT. And he took me over to the computer actually and showed me scratch. And I was like, oh my gosh, this is like incredibly powerful and brilliant. Like coding used to be this very specialist dependent thing, right? And only highly trained computer scientists could do it. All these ones and zeros looks very complicated. And then they just made it like child's play, right? To drag these little blocks of code together and this little figure dances around or jumps over a river. And as a kindergartner, you know, you just made that happen. That's like incredibly awesome,

44:57like inspiring thing. Okay. So then we said, what would that look like for molecules? Okay. So we created this digital molecule maker. We just published this in a journal of chemical education. So it's something we're really excited to start to get out there. And you can use it now in the classroom where students learn how to make molecules using a drag and drop. You just pull the blocks into a maker space and then you press a button and then you can 3D print your molecules from the, you know, digital molecule maker to a robot that can then, you know, put those blocks together and actually make the molecule that you designed on the digital molecule maker. You can

45:31then, let's say, put an AI agent into the digital molecule maker to be your friend and work with you and use, you know, data science to help you predict or these foundation models are dreaming about. So it's a very exciting platform that I think could get very interesting very quickly. And so we were presenting this actually at a conference several years ago. Uh, and, uh, we created to try to get this to be accessible. We created at Illinois, we call the molecule maker lab, uh, the MML, um, at the Beckman Institute. And the whole purpose here is to build a facility that anyone could access. Okay. So

46:04we're working very hard on that. So we had a conference to try to dream up what that facility might look like. Okay. And so we were able to invite people from lots of different backgrounds. And one of the people who was there, his name is Saif Williams. He was an eighth grader in Chicago and, uh, Ruby Mendenhall was able to invite lots of really awesome students from both local, but also from Chicago. And, uh, Saif actually saw the presentation on the digital molecule maker. And he's like, I can do that in Minecraft. So he knows how to like hack Minecraft. And so it was such a cool idea. And he

46:35proposed it and we were like, okay, this is awesome. So we brought him in as, as like a fellow for the, what became the molecule maker lab Institute. We got this grant from NSF. So he actually designed a Minecraft version of the digital molecule maker. And he took our blocks and put them into the Minecraft game. And then he was able to work with scientists at the national center for supercomputing applications here at Illinois. So there's a button you can push in the Minecraft game and print the molecule from Minecraft to the robot. And, uh, it was actually, he was able to put this together and make it work from his house in Chicago. Okay. And so then he was able to teach

47:09students in Pune, India, how to do it. And we had this big, this, another conference and ceremony and actually students from Pune, India, 8,000 miles away, were using Minecraft to then print, uh, molecules at the molecule maker lab in the Vecma Institute at Illinois. I think it's just one step, uh, in, in this very exciting direction, uh, of what we could start to dream about, you know, the future of organic synthesis can look very different than it has been in the past. And that's very, very exciting because once anyone can do it, I think the world is going to be a really exciting

47:42place. That's so cool. And you, you talking about scratches, like brought back so many memories, because I remember I did it like in my elementary school and yeah, that was so amazing to hear about and really goes to show that you can like learn, still learn from like anyone. Um, and to end this interview, what's one piece of advice you would give for younger chemists and scientists? I would say finding what you're both passionate about and good at is really important. And it's

48:18such a joy. If you find something that you really, really love to do, and it's something that you really feel like you can do, that overlap is amazing. Okay. And I think the joy of chasing a problem or trying to do something impactful is like almost even better than actually ever getting there. And so I just wish for everyone, they find something like that, because I think that's like a really, uh, you know, awesome way to kind of go through life is, you know, you really feel there's something

48:48ahead of you that is so exciting. You can't wait to try to figure out if you can do it. And so I would say for young people, try to find such a thing, right? Where you're, you're really, really excited about trying to make that happen, trying to discover that, trying to learn about, you know, what makes something tick and let that goal be a joy, right? I think the other thing I do, we do too much sometimes as young people is we fear failure and then it shuts us down in terms of what we actually try to do. So if I go back and give advice, I'd say spend a lot more time excited

49:19about possibly succeeding and as little time as possible fearing failure because the latter is worthless. It's like, it doesn't help you at all. And I do think there's something about human creativity and imagination and problem solving that joyful, excited, optimistic pursuit actually helps you be more creative and effective in problem solving. So don't be afraid of failing, be excited about the possibility of succeeding and you'll be amazed what you can do. Awesome. And talking to you, I can definitely tell you're so passionate about your, the work that you

49:51do and it's been a joy to learn from you and to hear about your work. So thank you for joining us. Thank you so much. Thanks for having me on. It's been great to hear from someone who is passionate about sharing the behind the scenes that goes into automation and developing new paradigms for medicine. I appreciate Dr. Burke recognizing the real and current value of the human machine interface where AI handles highly complex multi-dimensional data while human scientists provide the core passion, creativity,

50:22and teamwork. Yes, definitely. And molecular innovation is a perfect example of the intersection between chemistry, robotics, and clinical transition. I really appreciated the insight Dr. Burke shared about the skills chemists need to develop beyond chemistry to succeed, like having no fear of stepping outside your comfort zone to form interdisciplinary teams. Things like effective communication and collaboration are essential, yet they don't get as much practice as the analytical skills that we develop in lab. Absolutely. Conversations like this are a great reminder

50:57that behind every breakthrough in medicine is a huge, huge network of scientists, leaders, and innovators working for years to make it possible. To all the listeners of Let's Talk Chemistry, thank you for joining us today with Dr. Martin Burke, and I hope you learned something about the future of automated synthesis or developing the resilience necessary to lead the next generation of scientists. We'll see you next time. Thank you for listening to Let's Talk Chemistry, a podcast by ChemTalk. We hope you

51:27enjoyed it. For more information on today's episode and countless chemistry resources, please visit our website at www.chemistrytalk.org.