Donald G Comb Lecture - Carolyn Bertozzi

Sal Russello:
Morning everyone, and welcome. How's everybody doing?

Audience:
Great.

Sal Russello:
Good. I want to welcome everyone to the 2025 Don Comb Memorial Lecture. I'd like to recognize a few individuals. First and foremost, our speaker, Dr. Carolyn Bertozzi, who Tom will introduce momentarily. Welcome. I'd also like to welcome members of our board, members of the Comb family and a few invited guests. We're joined here by Michael Comb, board member and Don Comb's son, founder and CEO of Cell Signaling Technology. We're joined by Dave Comb, former NEB and CST employee, member of the board. And we're joined by Ben Comb, Don's grandson, Michael's son. We're also joined virtually, I know, by Jim Ellard who's ill today. So just want to say hi to Jim, hope you feel better. Jim is a former CEO, as you all know, and chairman of our board.

Today's an opportunity to reflect and celebrate those things that our founder believed in and worked tirelessly to support, a commitment to innovation and the ability to allow scientists the freedom and flexibility to drive new ideas forward. In today's uncertain funding climate, it's companies like NEB that have to remain steadfast in support of our values and those things that have and will continue to differentiate us. And as you all know from our town hall, you've all seen me present this slide on dozens of occasions, in times of uncertainty, it's really our purpose that remains constant. And that purpose for me encompasses Don's ideas about what New England Biolabs could be, and ideals for what New England Biolabs should be. And that is to build a sustainable business focused on enabling the scientific community, fostering curiosity and giving back to those closest to us in the world around us.

Okay, we're going to get started. A few quick acknowledgments to those individuals that have helped to make today possible. First and foremost, our research leadership. I want to thank Rich Roberts for his vision, and of course, Tom, Executive Director of Research, for turning that vision into real practical, scientific projects. So, thank you both. And I want to thank our scientific team who ensure we remain at the forefront of science. I'd also like to acknowledge two individuals that helped put on the meeting today from our scientific staff, James and Leah, who you'll both hear from a little bit later this morning. I'd like to thank Erin Varney for... Where is she? I'd like to thank Erin Varney for her work in coordinating the event as well as the team that produced the event, Kari, Deana Martin, Dave, Dave, and Mileidy. Without further ado, we're fortunate to be joined by an amazing speaker today whose work many of us have followed for years. I'd like to ask Tom to come up and introduce Dr. Bertozzi.

Tom Evans:
... to so many people here today. These are always fun presentations. And whenever I give these introductions, I always think, because it's in honor of Don, "What would Don think of the incoming speaker?"

And I know there's a few reasons he would particularly like listening to Carolyn's talk today. First, she took some basic research in chemistry, and then she turned it into applied research that led to more basic discoveries, and she'll talk more about that in a little bit. The other thing is she's passionate about science. When we were walking around today, it was just curiosity was clear, and that's, I think, what makes a great scientist. And Don was always asking questions. And I think the third thing that really struck me, and I didn't really think about it until this morning, when Carolyn walked in, one of the first things she said was, "Can I go in the winter garden?"

And I'm pretty sure that would've appealed to Don. So I'm going to introduce Carolyn, even though I don't think she needs any introduction. She started her really educational career at Harvard. She got her BA in Harvard. She then went to UC Berkeley where she got a master's and a PhD in Mark Bednarski's lab. And it's there where she learned about carbohydrates a little bit, where she was involved with viruses that interact with carbohydrates.

She then started her own lab. She's now at Stanford and she's won a lot of awards. I had them tattooed on this arm, it goes up this arm down the other arm. But I will want to point out that in 2022, she shared the Nobel Prize with... Oh, I love this as well, because I always like to try to name things and marketing won't let me do that. And Carolyn coined bioorthogonal chemistry. I am so jealous. But her co-noble laureate, Barry Sharpless coined click chemistry, which I also think is an amazing name. So three amazing scientists shared the prize and two amazing names came down with that. So with that, I want to turn it over to Carolyn. Thank you.

Carolyn Bertozzi:
What an honor to be here. Thank you so much for the kind introduction and just to be here on this stage with such an amazing group of people who've had such an impact in the world, celebrating the life and the impact of Don Comb is really an incredible honor for me. I think we would all agree that the modern life sciences enterprise, both in academia and an industry, was built on the shoulders of New England Biolabs. And, this is a company I have been familiar with since I started graduate school in 1988. And even then, the NEB catalog was the Bible that you had on your shelf. Everybody had one. You had to have one, because when you wanted to kind of look up something about molecular biology or how restriction enzymes work, that book was like a primer. It was a textbook. And unfortunately, we don't get them printed anymore, at least I haven't seen one at Stanford for a while. So my students have to enjoy all the learnings of NEB on the web. But I wish I had kept that textbook, actually. I think it would be a collector's item now.

I thought what I would do is just give you the story of bioorthogonal chemistry, which was the work recognized with the Nobel Prize. It was also the work that really launched my lab when I was a brand new professor back in the mid 1990s. And yes, we coined the term bioorthogonal. It didn't stick because you can't hardly say it. And then a few years later when Barry coined click chemistry, that really stuck. That was a brilliant bit of branding. So I wish that we had chosen a more pronounceable word.

But here's a factoid that I learned in Stockholm a couple years ago, which is that Barry actually did not coin the term, click chemistry. It was his wife. His wife was a writer. So she is gifted with the written word. And when Barry was describing the qualities of this type of chemistry that he wished there was more of, and he was like, "You take two things and they just come together and it's like something where they get connected together, and then those two things are now one thing and I don't know how to..."

And she was like, "Click." That's what happened. Anyways, there you go.

And this also has given me the opportunity to read up on Don as a human being and his back story. And I find this so inspiring. I had heard, certainly, the name Don Comb, but I actually didn't know much about the person until this opportunity. And I pulled out this particular quote down here from one of the news bits about him. I don't know if you can see this from where you are, but what it says is, "His vision for how a company could be run was unconventional but highly effective. People and passion over process and profit." And boy, right about now, do I wish there were more people that shared this sensibility. So, I'm looking forward to my return to Stanford tomorrow, so that next week in our group meeting, I can share this story with my own lab.

Okay, so let's talk about the back story and what led my young lab in 1996 to start working on this brand of chemistry that we call bioorthogonal. And I'm an organic chemist by training. And so what that means is I like to make molecules. And when you learn organic chemistry, there's another book on your shelf, the Compendium of Organic Transformations. That was a classic textbook that we all used to own. And there's a thousand reactions in that book. There's lots of different ways to take atoms and make new bonds and build complexity from simple starting materials.

But as much diversity of chemical reactions as existed in the 1980s when I went to graduate school, one thing all those chemistries shared in common was the assumption that you would perform them in this type of a vessel, which I know this is a bunch of biologists. Are there any people here who are chemists? Oh, I'm surprised and pleasantly surprised. There's nothing to be ashamed of for those of you who didn't raise your hands. Are the rest of you molecular biologists or biochemists? How many of those are there? Are there any glycoscientists? Oh, proud, represent. Okay, great. Good to know.

So for the chemists, you all are familiar with this. And this is a gadget that was designed to give the chemist complete control over all of the parameters of that reaction, which means the chemist can choose the solvent from dozens of possibilities. The chemist can choose the concentrations of the reagents and what the identities of those reagents are. The chemist can control the temperature and the pH, even the atmosphere. People who do chemical synthesis just assume it's possible to purge air from their reaction vessel and replace it with nitrogen or argon in case they have reagents that are sensitive to oxygen or to moisture. And it's that level of rigid control that allows the chemist to have thousands of possibilities to choose from in making bonds between two starting materials because a lot of the chemistries in that compendium are subject to interference. And those chemistries won't work if you can't get rid of moisture or oxygen, or if you can't exclude everything else and make the temperature just right and so on and so forth.

So that's great if you're doing conventional chemistry. But the more I learned about biology, which started when I was an undergrad. I was actually a bio major for two years thinking I might go to med school like every other person I knew. And then, I stumbled into organic chemistry and changed my mind about that and veered into chemistry. But I had taken a number of biology courses, and it seemed that if you could do chemistry in less controlled environments like biological environments, there might be interesting ways to probe and manipulate biological systems that would teach you about how molecules work in their native environments. And that interest became even more intense the more I learned about what chemistry could bring to biology during my years as a grad student and then later as a postdoc.

And those were the years where, as you suspected, I became interested in glycoscience. I got exposed to carbohydrate chemistry and glycobiology. And of all the different biomolecules, I think the glycans still remain today one of the most difficult to study because they're not easy to visualize, they're not easy to sequence. All the things that we now take for granted because of the brilliant work of Don Comb so that you can sequence DNA, you can sequence RNA, you can build and splice and dice DNAs and RNAs, that makes them much easier to study. But glycans, they don't follow the same paradigm so you don't have the same tools.

But chemical tools I thought maybe could fill some of those holes. And so this got me interested in, could you do chemistry specifically on sugars in living systems to study them? And if you could do that on living cells, which are the environment in which glycans operate, could you do it inside even living organisms, model organisms in the laboratory, or maybe even humans someday for purposes of disease detection or disease treatment? And these ideas were kind of swirling around in my head during my years as a postdoc and then as a new professor.

So, we came up with this idea. We coined the term bioorthogonal chemistry. People asked me, "Well, what does that mean? What's the definition?" So here's the definition. Bioorthogonal literally means not interacting with biology, which means not interfering with biology.

And another way to look at this is if you think of the entire universe of all possible chemical reactivities, it's a pretty vast universe. And within that universe as a much smaller galaxy that we consider the biological reactivity space. So that's basically all the chemicals in a biological system and all the chemical reactions they could undergo. And you could just think of your body. All the chemicals in your body constitute biological reactivity space. But somewhere in the universe of all possible chemistries, we thought maybe there's a different galaxy which are chemical reactions that have no counterparts in biology that can take place in a way that's totally agnostic to that biological space. And what could you do with those types of chemicals?

And what I thought you might be able to do with those types of chemicals would be at the outset to image glycans in living systems. That was sort of like the use case for bioorthogonal chemistry in my 30-year-old brain. And why did we want to visualize or image glycans? Well, the reason is that structures of glycans, we thought could have some diagnostic applications specifically in the context of cancer. Because even before I was in grad school, so it was back in the early 1970s and beyond that, people who study cancer tissues and do molecular profiling, even with the very crude tools available to them 50 years ago, noticed that there were some pretty dramatic changes in the patterns of the glycan structures on the surface of cancer cells compared to surrounding healthy tissues. And one of the more prominent glycophenotypes in cancer that had been reported at the time was an overproduction of glycans that terminate with a particular sugar building block that's called sialic acid.

So this cartoon gives me an opportunity to do a little primer on the glyco lexicon. So I realize very few people here are glycoscientists. So you probably don't know some of the fundamentals about this space. So let me be the one to tell you that there are building blocks, there are simple monosaccharides. They get linked together to make polymers that can be either linear or branched. And those are the glycans. The glycans are the polymers. The glycan is like DNA, and the monosaccharides are like nucleotides. And this audience obviously knows how many nucleotides, how many letters are there in your DNA. Would anyone like to answer the question?

Audience:
Four.

Carolyn Bertozzi:
There's four, right? Your middle school children are learning this right now, probably. There's four. And since you rely on the central dogma to make the enzymes that you sell, how many letters are in your proteins? How many amino acids? 20. The really modern hip people are like 22. But the textbooks would say that there's 20 plus or minus. Here we go. Ready? How many letters? How many monosaccharides constitute your glycans as a human? Crickets. Glyco people who raised your hands, you're all sitting on those hands right now. Anyone want to take a shot in the dark?

Audience:
About 60.

Carolyn Bertozzi:
We're in a double digit... No, okay? 9. What? Yeah, there's 9. So it's not as complex as you might think, right? One of the caveats is that they differ in different species, unlike the DNA and the proteins which are pretty conserved, it's pretty much the same building blocks in all domains of life as far as we know. The glycans vary a lot. The building blocks are different even. So there's been a much more rapid evolution of the glycome compared to proteomes and genomes, just FYI. But in humans there's 9.

And the way that we draw these structures is each of those building blocks gets a shape and a color that is the symbolic lexicon. And sialic acids are always these pink diamonds. So when you see a pink diamond, that's a sialic acid. And if you're a chemist, that's the chemical structure. And it's often at the terminus of a tree or a big glycan structure. And all of your cells have glycans that have a terminal sialic acid. We call them sialoglycans. And I like to think of healthy cells as having a pattern of sialoglycans that's like a well-manicured garden, the NEB campus, that's a healthy sialome.
But for reasons that we've been working on in my lab, and it's not the subject of this talk, but if you ever invite me back, I'll give you that talk, cancer cells were identified as having an overproduction of sialoglycans. So that well-manicured garden was left unattended in some way, and it became a tropical jungle of sialoglycans. And I read about this when I was in my training years, and I was thinking, "If there was some way you could image those sialic acids in living systems in a human body, let's say, non-invasively, then maybe you could actually detect a tumor or monitor the progression of a tumor because that's a molecular feature that distinguishes the cancer cell from the healthy cell. Let's make a technology that allows us to image those sialic acids."

And that was one of the very first projects I was trying to conceive of when I was applying for my first faculty jobs. But I didn't really have an idea of how to do it until I went to a conference during my second year as a postdoc, and that's where I met Werner Reuter, who was a German biochemist who worked in the world of carbohydrate biosynthesis. And I think this is a really good story for me to tell also, because the reason I got to go to this conference is my boss had agreed to give a talk there. It was in Southampton, England. And it was a bit of a weird boutique conference that didn't seem particularly interesting or relevant even to the work we were doing in my lab. And as the date approached, my boss, whose name was Steve Rosen, he really didn't want to go. And he said, "Okay, if anyone here is willing to go to England next week on my behalf, you can give this to get a free trip to the UK."

I was like, "Yeah, pick me."

So he sent me on his behalf to give his talk in England, and I just happened to meet Werner at this conference. And Werner was talking about some work in his lab on the biosynthesis of sialic acid. And so the way our cells make sialic acid is it starts with this precursor called N-acetylmannosamine. Sorry, I'm having a little cursor dyslexia. There we go. N-acetylmannosamine is a sugar that our cells build, and they use it in the biosynthetic pathway. About six enzymatic steps take place to convert that N-acetylmannosamine to the corresponding sialic acid, which is here. Then eventually, the sialic acid gets modified again. It goes into the Golgi compartment, and it ends up getting transferred onto those glycans as they're being biosynthesized in the secretory pathway, and it ends up on those cell surface glycoproteins and glycolipids.

So what Werner's group had discovered is that you can modify the side chain of N-acetylmannosamine. So the natural sugar has a methyl group. Where that R group is, that's an acetyl group. You know this because you study acetyllysine and epigenetic modifications, so that's the same acetyl group. And what Werner's group showed is that the R group could be extended from a methyl group to an ethyl group to a propyl group all the way to a butyl group. So those are just one carbon extensions. And as long as you kept the chain short between one and four carbons, all of these enzymes, every single one in this pathway, and there's six of them, would kind of ignore that modification and just happily convert the mannosamine to the sialic acid. And a few hours after you fed the cells that synthetic precursor, you would start to see sialic acids on the cell surface bearing that modified side chain. So the cell is like, "You are what you eat," right? You feed it the sugar, it just puts it into this biosynthetic product without any apparent toxic side effect.

And he presented this, and I looked at that slide and it occurred to me, "That's how you image the sialic acid. You put some R group into mannosamine that when it appears on the cell surface can be used as a chemical handle to attach an imaging probe."

And I came back from that conference and I was writing up my job applications. So I sketched out a proposal for what I would work on if somebody would please hire me to do it. And UC Berkeley hired me. Okay? And this is a graphic that literally came out of my job proposals. So this was made in PowerPoint circa 1994, and I still can open this slide. It just says something about PowerPoint, I guess. And so, here was the concept. If you want to image a cell surface sugar, maybe you can do it in a two-step process. You feed the cell a chemically altered precursor with a group X on it, that's a new chemical group that you put on. The cells do the hard work, they display that sugar in the context of cell surface glycans. And now in a second step, you introduce a probe, and the probe has a complementary chemical group called Y here, where X and Y will react with each other. And now there's a probe on the sugar.

The problem was that X and Y have to react with each other without interfering or interacting with the biological system. In other words, they have to be bioorthogonal. And this was the use case that drove us to invent bioorthogonal chemistry. It was really purposeful at the time. And the funny thing is, is when I applied for jobs I, at that time, didn't yet have an idea for what X and Y could be. And nobody asked me in my job interview. I was thinking, "If somebody asked me what's the X and what's the Y, I'd be like, 'Oh, I got to figure that out.'"

And so, they hired me anyways. And so, when I started recruiting students, I sat them... I say, "All right, we got to figure this out now. They hired me and I promised I would do this."

And so, the first chemistry that we explored to invent a bioorthogonal reaction was kind of an oldie but goodie that I had used myself many times in my work as a synthetic carbohydrate chemist when I was a grad student. And it's a chemistry called the classic Staudinger reduction based on work that was published by Hermann Staudinger, who was a famous German chemist, first reported on this reaction in 1919. This is more than a hundred years ago. He won the Nobel Prize in chemistry in the 1950s, not for this work, but actually because he was a polymer chemist.

And he was the first person to come up with the idea of the macromolecule, the idea that you could take a simple molecule and stitch it together over and over and over and make a polymer or a macromolecule from it. And the reason that became important, because the concept of the macromolecule was the framework that Watson and Crick relied on in proposing the structure of DNA. It wasn't obvious that DNA was actually a macromolecule until Watson and Crick understood the physical properties and the composition and all that. And Staudinger's framework was critical for that.
But regardless of that, what he was also doing was playing around with the reaction between triphenylphosphine and a functional group called an azide. Azides are remarkable. An azide is a small functional group, three atoms, three nitrogens in a row. But for just three little atoms, it really packs a punch because azides can undergo all kinds of interesting chemical transformations. And I had used them in grad school as a precursor of an amine. And you can convert an azide to an amine with the Staudinger reduction.

And what happens is, you add triphenylphosphine, it's a soft nucleophile. Azides can be soft electrophiles. If you ever learned about hards and softs and nucleophiles, electrophiles. Everyone here took organic chemistry, because it's a requirement if you're even a bio major. You have to take O chem. But you forgot it all. Oh, aren't we lucky that we're here with the intro O chem professor? So the soft nucleophile matches up with a soft electrophile, and they form an intermediate called an aza-ylide. And the aza-ylide is interesting in its own right, but in the presence of water, so if you add water to the reaction, it will spontaneously hydrolyze. Water cleaves the phosphorus nitrogen bond, and that's how you get to your amine. And the other product is triphenylphosphine oxide.

So it was well-known in the chemical synthesis world that you could put an azide into your molecule, carry it through a synthesis, and at the end, do a Staudinger reduction and convert it to an amine. So an azide was just nothing more than a cryptic amine. But unlike an amine, which is a nucleophile and a base, and it can react with lots of other things, azides are actually quite inert until you put triphenylphosphine in there. They'll survive a lot of conditions of acids and bases and high temps and low temps and hard neucleo... They'll ignore a lot of other chemically reactive groups, but triphenylphosphine and water, boom.

The other thing that attracted me to this chemistry was the fact that azides don't exist in biology. As far as we know, azides, were by humans for humans. Nature doesn't make azides that we know of. Same with triphenylphosphines. These are not in nature. There are phosphines in exotic organisms, but not this type and certainly not in humans. So for all intents and purposes, phosphines and azines are bioorthogonal. They don't interact or interfere with the biological chemistries, and they react with each other, and they form a product. The problem is the product that they form initially is not stable in water. And if you're doing chemistry in a biological system, what's your solvent? It's no choice. It's water.

So we thought if we could somehow manipulate the aza-ylide to be stable in water, maybe we would have something bioorthogonal that could be useful for imaging sugars, which led us to develop this very simple modified form of the Staudinger reduction that we call the Staudinger ligation. This was our first invented at home chemistry that actually we could perform in living systems. And the only modification we made was to take triphenylphosphine and introduce this ester. That's it. And that might be a simple modification, but it's got huge consequences, because when you make an aza-ylide, now, that nitrogen atom wants to cyclize into the ester to form very rapidly an amide bond, and then water can mosey along and cleave the phosphorus-nitrogen bond. But that's fine because the amide keeps the two things together.

And this is, I think, a real page out of O chem. When you took organic chemistry, did you learn about how fast reactions occur when you're making a ring? Do you kind of remember that? Well, they're fast. And five-membered rings are the fastest of all, and that's what you're doing here. Nitrogen attacking this carbonyl is a five-membered ring. So It's faster even than the hydrolysis. Even the water is the solvent, which means the concentration of water here is 55 molar. You get a PhD in chemistry. All right?

So once we had the chemistry worked out, we kind of went to town with it. Putting azides into sugars turned out to be the simplest part of all, because we had that work from Werner Reuter's lab to use as a precedent. And so we said, "All right, we know that extended side chains on N-acetylmannosamine are tolerated biosynthetically. So, let's put the azide right there."

We put the azide right on the position where that R group was, and we fed this derivative to cells. And lo and behold, a few hours later, out would pop the corresponding azide-labeled sialic acid on the cell surface. And you could do this with pretty much any of the sugar building blocks. So we did it with a sugar called fucose. We did it with N-acetylgalactosamine and so on and so forth. Once you had the azide there, all you had to do was a reaction with a phosphine probe, and you'd have an imaging probe on your sugar. And you could do this on cells. We published several papers on that, but even more importantly, you could do it in the body of living animals.

This was the first paper we published where we showed we could inject animals with azido and acetylmannosamine and then load them up with that sugar and then wait a couple days, and then harvest organs and tissues. You could find the azide-labeled sialic acids on their cells in the body. And then knowing that they were there, you could also just inject the live animals in a second step with that Staudinger ligation-armed phosphine probe on your probe of interest. It could be a fluorescent probe, a biotin, a Flag-tag depending on the experiment. And then, when you took the cells out of the animal, you could detect that product of the Staudinger ligation. So the chemistry had worked in the body of the animal.

And that was really wonderful and exciting, and we felt like the door was going to open to us to do all kinds of really interesting imaging experiments. And we were kind of focused at that time on trying to detect these cancers with their tropical jungle of sialic acids. But we ran into a roadblock, because as clean as this chemistry was, and as well-tolerated as it was in the mouse, it was kind of sluggish. Kinetics matter in chemical reactions, but they matter less when you're doing the reaction in the round-bottom flask. Because if your reaction is going too slowly for your taste in the round-bottom flask, you can heat it up, but you can't just boil your mouse and reflux in toluene at 110 degrees. That's not going to work for the mouse anyways. It's not going to work. Also, when you do reactions in the round-bottom flask, you can achieve an equilibrium state, right? Nothing leaves that flask unless you take it out. Everything stays in there until the reaction is done.

But that's not true in a live animal because the minute you put reagents into an animal, they're getting metabolized in the liver, they're getting cleared in the urine. They're not just going to stay there until your reaction's done. So if the reaction is too slow on the time scale of metabolic clearance, well, it's not going to work, right? And that's what we discovered. The reaction that Staudinger ligation chemistry was about two to three orders of magnitude slower than we really needed it to be, and we couldn't just heat up the mouse to compensate.

So it was around this time that we in parallel started thinking about other bioorthogonal chemistries that we might develop that would be faster. And we really love the azide, because it was so small and so inert in a biological system. And the cool thing about the azide is that not only do azides react with soft nucleophiles like a phosphine, they also react with alkynes in a process called a 1,3-dipolar cycloaddition. And that's another "ancient German chemistry", as my students might've put it. This time, not quite as ancient, but it was Rolf Huisgen from University of Munich who did a lot of mechanistic work on this. And it became known in the middle of the previous century as the Huisgen cycloaddition.

And what happens is the azide reacts with the alkyne to form a product called a triazole. Now, in its published forum, at the time, this Huisgen cycloaddition was also considered slow. In fact, really kind of side by side, it would be even slower than the Staudinger ligation chemistry. So typically in the laboratory, in the round-bottom flask, you would heat up these reactions to get them to go. And so we knew that the standard version of this reaction was probably not going to work for us because it was also too slow.

So we started thinking about ways we could speed up this cycloaddition chemistry. And that was around the time that Barry Sharpless and Morton Meldal discovered that you could dramatically accelerate that cycloaddition with the use of a copper catalyst. It's kind of a fun story about how they ran into this, but the long and the short of it is, they both published papers right around the same time, on the same discovery. And it was total serendipity, but serendipity in totally different ways in the two different labs, which is, I guess, the nature of serendipity when you get right down to it.
But they discovered this accidentally, but it's quite remarkable. The caveats are that the alkyne has to be terminal, which means that there's an R group on one side, but a hydrogen on the other. You have to add the copper with this ligand around it. But if you do it properly, it's super-fast. This was plenty fast enough for what we needed at the time. So we got really excited when these papers came out until we took a step back and realized that the copper catalyst is cytotoxic. Oh, they won the Nobel Prize, too, by the way. We shared it. But this is a catalyst. Copper(I) is toxic to cells of all types. And we tried. We thought, "Well, maybe if we could use just a tiny bit of copper or something else..." We were never able to get this chemistry to work in a way that wasn't harmful to the living system, so that wasn't going to solve our problem either.

Now, I teach O chem. Who loved their O chem class? Oh, good, all right. Usually it's like... They're like, love... Okay. I really did. And I remember some of the lectures that really got me hooked. And one of the lectures that I remember vividly from my undergraduate days was the one on ring strain. And now as an O chem teacher, I also always give my annual ring strain lecture. And what we teach our students is that there were certain geometries, certain rings, that carbons like to sit in like cyclohexane ring, the six-membered ring. And they like those rings because there's not much strain. All the bond angles in the ring match the preferred bond angles of the carbon.

But as you make those rings smaller and smaller, now the bond angles are kind of out of whack compared to what the carbon would prefer. And so, energy builds up in these strain systems. And when you get down to these three-membered rings, you've got 29 kilocalories per mole of strain locked in the ring. And if you can subject that ring system to a reaction where the strain is released, usually those reactions go a lot faster, because there's a driving force. The ground state is destabilized compared to the transition state. And then, if you put double bonds into these little rings, it's even more strained, because the bond angles are even more out of whack compared to the preferred bond angles. And these chemistries on these ring systems can be used. They can be harnessed to get things to react in ways that they normally wouldn't react.

So I was writing that lecture back in around the year 2002-ish, on a plane coming from, I think, Boston, actually, to San Francisco. And I had a layover in Chicago. And as I was writing that lecture, the thought popped into my head, "I wonder if anybody had ever used ring strain to accelerate a Huisgen cycloaddition. And so, I had the cell phone number of one of my grad students on me, and I called him from Chicago. And I said, "Hey, can you go to the library," because in 2002, you had to go to the library, "and see if you can dig up anything on strained alkynes."

He goes, "All right." And so by the time I landed in San Francisco, he had left a message on my phone and he says, "You won't believe what I found." In 1961, Wittig and Krebs had reported this paper in German on the unusual, unexpected reactivity of this strained ring system cyclooctyne with phenyl azide to form the triazole product.

So cyclooctyne, I learned, is the smallest cycloalkyne that is stable, that can be isolated at room temp. It's got a lot of strain, almost 20 kcals per mole. And the reason is that when you have an alkyne, the bonds really want to be linear, but in a ring, they're bent, and that's where the strain comes from. And in this reaction, when the alkyne and the azide react to form the triazole, this triple bond becomes a double bond. And the bond angles of the double bond are a nice match for that ring. Whereas in the triple bond, they're not. So you go from strained to unstrained. And this was in German. There was no Duolingo at the time or Google Translate or anything. So we had to fight our way through this paper.

But it didn't matter. You don't have to speak German because this one sentence kind of said it all, which is, "Blah, blah, blah, phenyl azide, blah, cyclooctyne, and then, explosion."

And we're like, "Yeah, that sounds about right. Let's try that."

Oh, another little factoid. Georg Wittig also won the Nobel Prize in the previous century for a reaction called the Wittig reaction. But by the way, Wittig had done a postdoctoral position in Staudinger's lab. So it's a small little world, isn't it?

So we kind of went to town over the next few years figuring out how to build cyclooctynes, how to chemically modify them and adjust their physical properties and tune their reactivities. And here is a quickie overview of some of the compounds that we made and the lessons that we learned. So first of all, up in the upper left, that was one of the kind of parent cyclooctynes that we made early on, and we studied its reactivity with azides, and this is the second order rate constant of its reaction. I won't go through the details, but suffice it to say, this was on par with the Staudinger ligation, which we already knew was too slow, but that was where we started. That was the floor. And then we tried to tune the reactivity up from there.

We discovered that introducing two fluorine atoms next to the alkyne would boost the reactivity almost two orders of magnitude. So that was a pretty big step function to go from this compound to this one. We call this one diflo for short, the difluorinated cyclooctyne. But you can kick it up another notch by fusing benzene rings to the west coast and the east coast of the cyclooctyne. So now you've got a molecule that's almost a thousand times more reactive than the parent. And if you're willing to do a pretty tough synthesis, you can even kick it up another notch by shrinking the ring from eight atoms to seven atoms. But you have to put a large sulfur atom in the ring, because it's too strained if it's all carbon.

But anyways, once we had this worked out, we thought, "Okay, we're kind of between diflo and this molecule." We called it Barack. It was very clever. It was 2008. The postdoc who worked on this came from Chicago. There were all these reasons behind that.

But regardless, these two compounds sort of are now the standard use compounds for people who like to practice this chemistry. And by the way, since at that time, Barry's wife had coined the term, click chemistry, and this was a version of that type of chemistry, but without the copper, we started calling this copper, free click chemistry. And those of you who have noticed my earrings, we can talk about that later.

So now, that we had the chemistry, we thought, "Let's go back in vivo and try and do some imaging experiments."

But by this time, the students in my lab working on this thought, "Well, instead of going into mice," which is kind of usually where you go when you want to model human disease, "let's pick an organism that's a little bit more amenable to optical imaging at least, which is the zebrafish." Also, a vertebrate. Has a developmental program that's very similar to those of other organisms like humans and mice. And the great thing about zebrafish development is that it unfolds in a pretty fast time scale. So people who work with zebrafish, what you do is you take the oocytes and you fertilize them in vitro and you make embryos and the embryos grow. Five days later, you have a young adult fish with all of the body parts. And these fish are translucent, so you can literally watch all of this unfold in real time in a fluorescence microscope. And there's a lot of similarities between the development of tumors and the development of organs during embryogenesis, including this hypersialylation phenotype.

So we started working with zebrafish, and we would do experiments where we would make embryos, we'd fertilize these embryos, and then we would just bathe them in a solution of azido sugar. And the sugar would get taken up and metabolized, and now the cells have azido-glycans on the surface. And then, we could come in with a copper-free chemistry reagent like diflo connected to Alexa Fluor 488, pick your color, and image those sugars and visualize their distribution. And you could do this in a temporally controlled fashion. For example, in some experiments we would do this once, do a little quick wash, and then take those labeled fish and add more azido sugar, label a different population of glycans, and then tag them with a different color probe. And you could figure out which glycans were made at what time point during development in which tissues.

WHich led us to collect images like these. This is the head of a 5-day-old zebrafish that was labeled with three different sugars, three different colors, three different time points. And you can see the distribution of these things. For example, the epithelium of the nasal cavity or the upper respiratory tract has a different glycoprofile than the epithelium that constitutes the skin of these fish. And we could do imaging on pretty short time scales, even in early embryos. So this embryo is only 24 hours old, which means that a lot of the cells are still making decisions about their fate and what tissues they'll belong to.

We noticed that during rounds of mitosis, there was a transient redistribution of certain types of sugars on the plasma membrane, and we studied that for a while. Sorry, somebody's trying to get me to update my software. I'm sorry, I'm not going to do that. So for example, these arrows, if you can see this, are pointing to cells that are in the late stage of mitosis where you can kind of see an intensity building up right at the junction between the daughter cells. And if you blow that up and do kind of time-lapsed imaging, this is what you see.

So all of these panels are the same exact frame with the same mother cell that is undergoing mitosis to two daughter cells. What's different is that different molecules are labeled. So in this frame, we have a probe on the DNA, so you can see how the chromosomes duplicate and then segregate, remember that? In this frame, you're looking at a membrane protein. So it's a membrane marker. And you can see that it starts out that the mother cell is surrounded by a membrane, but as the cell starts to form two daughter cells, that membrane fills in the gap.

And then for the first time ever, we were able to look at the same process through the lens of the glycans. And you can see first of all, the cell surface glycans look like fuzzy, hairy things, which they are. That's kind of what it looks like there. But you can see how late in mitosis there is this accumulation of this particular type of glycoprotein that we were studying at the junction. And it turns out it contributes to the biophysics of segregation of the two daughter cells. So those are the kinds of things we were doing at this period of time. And it was a lot of fun because for the first time, you could actually take a look at sugars on live cells, and there were all kinds of interesting applications.

But as soon as we started publishing on these experiments, lots of other people got ideas. Because if you can label and probe sugars in live cells this way, why can't you use the same strategy to label and image proteins or nucleic acids? There's lots of molecules you'd like to look at in live cells, and this is a platform that allows you to do that. So other people started putting azides into pretty much all the building blocks of life. People started putting azides into nucleic acids, and now you buy alkyne labeled EDU. That's now a standard reagent to look at de novo RNA biosynthesis. And people were putting azides into lipids to do lipid profiling and imaging, into covalent enzyme inhibitors. Tag your enzyme and now you can go fishing for it with a click chemistry handle. And labeling amino acids with azides and so on and so forth. So now, there's dozens of clickable reagents that you buy. There's kits that you buy that uses this chemistry in a lot of different ways. It's kind of baked into the fabric of life sciences research now, and that was a lot of fun to watch.

But over the last decade, I think the focus has really been in the world of clinical translation. This is always really gratifying for an academic to see something that they develop get deployed in the real world in a way that can directly impact patients. So a number of companies have now launched around therapeutic modalities that really rely on bioorthogonal chemistry for the construction of the drug. And antibody drug conjugates are a big sector in which click chemistries are becoming really foundational. These are just a bunch of companies that on their website brag about how they use click chemistries and so on. But I think every major pharma company that makes ADCs, as we call them, is probably interested in this chemistry. Companies have made vaccine conjugates that are in humans right now using click chemistries. Cell therapies, chemically modified cells that are deployed in the setting of cancer immune therapy. I just came up from a few days at Alnylam. They make oligonucleotide conjugates this way. And at the end, I'll show you a company that's actually doing chemistries in the human body for drug delivery.

Let's talk a little bit about ADCs. I think these are familiar to most people now because they're one of the big hot areas of mostly oncology drug development. And an ADC is exactly what it sounds like. It's basically a chemical conjugate to an antibody where the antibody is able to bring a drug selectively to a target cell, usually a cancer cell. The antibody binds an antigen, gets internalized. The antibody drug conjugate gets deconstructed inside the cell, and the toxin is released and it targets usually either DNA replication or microtubule formation, and induces cell death.

So it's an industry that's had a lot of kind of ups and downs and roller coaster rides. It's kind of on a high right now. It wasn't always that way. In the 1980s, Bristol Myers Squibb had an antibody drug conjugate group that they eventually disbanded five years later thinking that these drugs were never going to make it. And it was only after many decades of solving problems, having to do with making good antibodies, making good linkers and good drugs, and doing good chemistries, that brought us the dozen or so drugs that are approved today.

So I won't go through all the work that people did to make ADCs work, because this is all published work, but I will say that where we have focused with bioorthogonal chemistries is doing a better job connecting the synthetic part, which is the drug linker part, to the protein part. It was literally the chemistry-biology interface in such a way that you can make antibody drug conjugates that are structurally defined and can be engineered and optimized for a particular indication.

So, before click chemistries came along, people were using kind of old-school bioconjugation methods to make ADCs. And these two early successes in the world of ADCs illustrate the chemistry. Seattle Genetics makes this drug, Adcetris. And the way they do it is by alkylating the cysteine side chains of the hinge region with maleimides. That's a chemistry you probably use here even at NEB, because it's old-school bioconjugation. If you want to biotinylate a protein, you could do it this way. Genentech, which makes this drug, Kadcyla, does acylation of lysine side chains with active esters like NHS esters. And these are the chemistries that are right now literally in every single approved ADC. They've all been made with one or the other of these two conjugation chemistries.

What that means is that they're mixtures of heterogeneous molecules. And just to give you the most extreme example, when you do chemistry on lysine to make a drug like Kadcyla, you're trying to pick off maybe 3 or 4 lysines to put a drug on out of maybe 75 that are on the surface of the protein. And what this means in practice is that you're going to end up with a statistical mixture of products. By contrast, if you could put a bioorthogonal chemical handle into the protein at defined locations and with defined copy numbers, you could make chemically defined antibody drug conjugates where you knew exactly where the drug was attached and exactly how many drugs were attached in total. And we thought that would be beneficial because if you have a statistical mixture of many molecules, it's hard to optimize your drug.

So we developed a platform for this. We call it the aldehyde tag platform. And basically what we do is we use an enzyme called the formylglycine generating enzyme that has the natural ability to recognize a 5-amino acid sequence and oxidize a cysteine side chain to an aldehyde. What that means is that if we clone that sequence into a protein of interest, call it the heavy chain of an antibody, that protein will get expressed with aldehyde groups at the site encoded by the cysteine as long as that FGE formylglycine generating enzyme is present in the cells. And once you have an antibody with an aldehyde, you can do a chemical reaction on the aldehyde, which is a unique side chain. It's orthogonal to the other 20 amino acid side chains. And the chemistry we developed for this, we call the Pictet-Spengler ligation. It's another oldie but goodie from the early part of the 1900s that we adapted with some modifications I won't go into.

But that technology got commercialized. We formed a company called Redwood Bioscience, and this is my former grad student who was the co-founder with me. And eventually, that was acquired by this other company called Catalent. So Catalent now has this as a technology offering where if you give them the heavy chain and light chain sequences, they'll express the antibody with aldehyde tags, as many as you want, and then do the Pictet-Spengler chemistry to attach the drug linker, the payload, to make ADCs. So that was our particular contribution to the ADC list. But every company on this list has their own variation on the click chemistry theme.

Finally, let me tell you about this chemistry, Shasqi, the company, Shasqi. These guys have figured out how to do bioorthogonal chemistry in the human body. So back to this early slide, I talked about how I had fantasized about taking chemistry out of the round-bottom flask and onto the cell, and we did that. But what Shasqi did is they went one step further and took it into the human patient. This is their platform. They've published on this. But they use a bioorthogonal chemistry called the tetrazine ligation. It also is kind of a strange cycloaddition. It's a bit of a takeoff on the copper-free click chemistry. But it's even faster than our copper-free click chemistries. The tetrazine ligation, I think right now holds the world's record in terms of kinetics of bioorthogonal chemistries.

One part of that ligation is the tetrazine, and what this group at Shasqi does is they put the tetrazine on a hydrogel polymer, which is hyaluronic acid. It's the stuff that's used in cosmetic filler. And then, they'll inject that chemically modified hyaluronic acid hydrogel into a tumor. Now, nothing else happens. This tetrazine has no reactivity in the human body. It's bioorthogonal. Then what they do is they'll infuse through IV into the same patient a caged or pro-drugged version of a chemotherapy like doxorubicin, used to treat solid tumors. Doxorubicin is a pretty vicious chemotherapy with horrible side effects, like most of them. Hair falls out, people are nauseous, they lose their immune systems. So there's a lot of interest in targeting doxorubicin to the cancer cell, but limiting its exposure in other tissues. And that's exactly what's accomplished here.

So they cage doxorubicin with this trans-Cyclooctene, it's a strained alkene, so that now the molecule is harmless in the human body. It's not toxic. But as soon as that drug happens to encounter the polymer that was injected into the tumor, a bioorthogonal reaction occurs that, for mechanistic reasons I won't go into, causes the release of the free drug which kills the cells. And this is in now a phase 2 clinical study. So I'm really excited to see where this is going. But the long and the short of it is, what we started as a humble tool requirement to do cell surface glycan imaging, kind of grew into a field unto itself, which is now having a big impact in the biopharmaceutical industry.

So the take-home message is this. We started this really as very much a basic science project. We invented new chemistries, we studied their mechanisms, we optimized them, and then we used them in the laboratory to study changes in sugars in live organisms. Eventually, what started as basic science was translated into applied science. So I think you will all resonate with this sort of yin and yang synergistic relationship between the two. And most importantly of all is I couldn't have done any of this if I hadn't had 30 years of continuous funding by the now threatened National Institutes of Health. In life sciences and academia, we live and die by the NIH. NIH right now is being gutted by our administration. This is existential for those of us in academia. And so, I hope that we as a nation come to our senses and reinvest in the federal agencies that support basic sciences and the life science, because otherwise there won't be another bioorthogonal chemistry 30 years from now.

So with that, I'll thank my research group. I just talked about pretty much 30 years of science. I can't mention all the people by name that contributed. But after the Nobel Prize announcement, my alumni kind of all connected online through LinkedIn and so on, and we did a Zoom party, and there are people from every time zone on the planet who came. About 250 of my current and former trainees logged in to toast, and it was really, really wonderful of them. So, thank you once again for this really wonderful invitation. It's a great honor for me. And we can do questions maybe, I don't know. Okay.

Tom:
That was fantastic. So at this point, we're going to end the live stream. We're going to have some questions, but we're going to end the live stream first.