RuBisCO Engineering for Climate-Resilient Agriculture with Robbie Wilson

On this episode of The Climate Biotech Podcast, Paul Reginato is joined by Robbie Wilson, a research scientist in the Department of Chemistry at MIT, where he leads the RuBisCO engineering effort in Matt Shoulders' lab. Robbie has spent his career working on RuBisCO – the enzyme that fixes CO2 –across labs in Australia, Germany, and the United States. He co-leads the Enhanced Photosynthesis in Crops (EPiC) Project, which applies directed evolution to RuBisCO with the goal of moving improved variants from the lab into field-relevant crops.

RuBisCO is present in all autotrophic organisms, including plants, where it catalyzes the conversion of CO2 into molecules that can be used throughout metabolism.  Almost every carbon atom in your body passed through it at some point. It is also famously slow, and the version inside plants is so structurally fragile it needs a dedicated team of helper proteins just to fold correctly. During his postdoc in Germany, Robbie and a colleague figured out exactly which helpers were needed, making it possible for the first time to grow and engineer plant RuBisCO inside ordinary lab bacteria. That unlocked the door to testing thousands of variants quickly.

Robbie's work also points to why a faster RuBisCO is not necessarily a better one. The fastest versions found in nature live in places like hydrothermal vents where CO2 is abundant, and they perform poorly inside a leaf where CO2 is scarce. The more useful questions are about CO2 affinity, the trade-off between activity and stability that keeps the enzyme from being pushed too far, and oxygen interference. Oxygen interference is when RuBisCO mistakenly catalyzes a reaction with oxygen instead of CO2, an inefficiency which wastes the plant's energy and gets worse as temperatures climb. Reducing that oxygen sensitivity is a central focus of MIT’s EPiC Project, and is a high-leverage opportunity to protect crop yields in a warming world.

Listen to learn why heat waves are increasingly dangerous for staple crops like wheat, how machine learning is reshaping the way RuBisCO engineers choose which mutations to test, and what stands between today's lab results and a real improvement in the field. 

Transcript

Carbon Fixation Before Fossils

SPEAKER_01 0:03

The activity of Rabisco actually predates the fossil record in a way. Because some of the oldest examples of biogenic carbon have been found in 4.1 billion year old graphite inside of Zircon rocks. So before you actually have established a really established fossil record on the planet.

Welcome And Guest Introduction

SPEAKER_00 0:27

Welcome to the Climate Biotech Podcast, where we explore the most important problems in climate and environmental biotechnology and how we can solve them. I'm Paul Reginado, co-founder of Homeworld Collective. Together, we have agency to build technologies that enable a brighter future for all life on Earth. Alright, hello everybody. I'm thrilled to welcome Robbie Wilson for a discussion about climate biotech. Robbie is a research scientist in the Department of Chemistry at MIT, where he leads the Rubisco engineering effort in the Shoulders Lab. His work focuses on improving the central enzyme of photosynthesis with the goal of boosting crop productivity and food security. Robbie earned his PhD in plant science from the Australian National University, where he became the first to stably integrate engineered rubisco enzymes into the chloroplasts of a C3 crop, and developed new bacterial screening systems for engineered rubisco variants in E. coli. As a postdoc at the Max Planck Institute for Biochemistry, he worked out the molecular requirements for producing plant rubisco recombinantly, opening the door to engineering the enzyme in tractable hosts. He also determined the structure and function of bundle sheath defective 2, or BSD2, an essential chaperone for rubisco biogenesis, and characterized its role in assembly. At MIT, Robbie co-leads Enhanced Photosynthesis in Crops, or EPIC, the inaugural JWAFS Grand Challenge project, which applies directed evolution to Rubisco with the aim of moving improved variants from the lab into field-relevant crops. Welcome,

From West Auckland To Directed Evolution

SPEAKER_00 2:04

Robbie.

SPEAKER_01 2:04

Great to be here.

SPEAKER_00 2:05

So let's start just by learning a bit about your background and how you came to be where you are. So maybe you could tell us where you grew up and a little bit about your journey getting here. Did you always know that you'd work on photosynthesis and rubiscoenzymes?

SPEAKER_01 2:22

I'm originally from West Auckland, New Zealand. And no, I didn't think I'd always be working on photosynthesis. In fact, I didn't even know if I'd be going to university when I was in high school. Growing up in West Auckland, the working class area of the city, and I had a pretty working class upbringing as well. And I didn't do very well in high school, actually. I I was actually even asked by the dean of my year when I was 16 whether I should consider dropping up and going into a trade work. So my dad was a builder, and I'd helped him out on the weekends and be doing work with my dad doing construction. And I knew I hated it. So that was a wake-up call for me. I thought I better actually start applying myself in high school and try and get him to university, go and do something with a bit more using my brain rather than my hands. Yep. I succeeded in doing that. When I got to university, I didn't really know what I wanted to do. I thought it would be fun to do something involved with forestry. So I like the idea of driving a truck around in the forest and like checking on the trees. And when I got accepted into university, I basically just picked the coolest sounding degree that they had in biology, which was the subject I ended up being strongest in in high school. And that was molecular biosciences. Even now, when I talk to people, they're like, ah, it's a fancy sounding degree name. So I picked that, went in there, had a chat with some of the professors, and they've encouraged me to take that on as my major. And I was much better at university than I was in high school. So yeah, it really thrived in an environment where I was leading my own learning, and rather than being told what to do, I think that was a big shift for me. That was a discipline where we were learning a lot about molecular biology, biochemistry, cell biology. And during that time is when I learned about directed evolution. So directed evolution, I thought was just the coolest technique in molecular biology at the time. I thought it was incredible that you could use natural selection to direct changes in enzymes towards what we call a user-defined goal. And I was like, wow, that's what I want to do. I want to go from here and I want to do something that uses directed evolution. So at the time, which is back in 2010, there wasn't anybody in the southern hemisphere that was really doing directed evolution. My my honors supervisor at the time had been doing some Wayne Patrick. He's now at Victoria University in Wellington. They were doing some directed evolution. I'd worked on DNA ligase at the time during my honors project, but I could only find one directed evolution lab, and that was Spencer Whitney's lab at the Australian National University in Canberra. And so I thought this is my opportunity to go and do directed evolution. They're doing it on this enzyme Rubisco. Didn't know anything about Rubisco. And I also had an opportunity to go there to do a summer scholarship, which is essentially just some time over the Australian summer, November and December, where you live on campus and you do a small project and you spend most of the time drinking with other recently graduated students from around the country. And that was great fun. I had a really good time. And I thought, hell yeah, I'll stick around. And so yeah, that's how I got working on Rubisco. And I have been working on the enzyme ever since.

SPEAKER_00 5:58

So you caught the bug. Something about Rubisco has really captivated you for many years now.

Rubisco Basics And Why It Matters

SPEAKER_00 6:05

Maybe you can just tell us the basics. What is Rubisco? What does it do? What's its role in photosynthesis?

SPEAKER_01 6:11

Yeah, so Rubisco is an enzyme. So it's a catalyst that speeds up the addition of carbon dioxide to a five-carbon substrate called a rubulose 1-5-based. And this process is how most of the biomass on carbon-based biomass on Earth is generated. And yeah, it has an essential role in photosynthesis. Most well-known is in the Calvin cycle, where it performs this essential catalytic role. In reality, though, Rabisco is actually an enolase, so it binds to this RUVP molecule and it analyzes it. And it that it's that intermediate state that actually allows for carbon dioxide addition. So this process would be incredibly slow if it was purely a chemical reaction. And so Rubisco speeds it up. Even though Rubisco is really really slow, the acceleration of this addition of carbon dioxide is enormously greater than it would take in a non-catalyzed enzyme-catalyzed reaction.

SPEAKER_00 7:18

And Rubisco really is a carbon-fixing enzyme, right? Like we tend to, when we think of photosynthesis, we think about Rubisco, but Rubisco is also fixing carbon in non-photosynthetic organisms as well, right? That's right. Is that like chemosynthetic organisms and that's right, yeah.

SPEAKER_01 7:38

Yes. So everywhere you can imagine on the planet you'll find rubisco doing some job and some role, whether it's in deep sea ocean vents, whether it's in lakes and rivers, whether it's in the oceans, whether it's in the leaves of plants, and whether or not it's I know, probably in the sky as well. There's cyber bacteria flying around through the air. So yeah, it's everywhere.

SPEAKER_00 8:06

Anywhere there's primary production, right? So it's so this so this is at the heart of we are carbon-based life forms. All of the carbon in us was at some point fixed by Rubisco from CO2, which is pretty amazing.

SPEAKER_01 8:23

Although I think very ancient enzymes.

SPEAKER_00 8:26

Yeah. I was about to say, I guess we we eat hydrocarbons, we eat petroleum now, like a lot of the there the inputs to agriculture, some of it does come from uh petroleum, but that also was fixed by Rubisco, so that that still holds.

SPEAKER_01 8:42

And so I learned like an interesting thing recently. I would say that the activity of Rubisco actually predates the fossil record in a way, because some of the oldest examples of biogenic carbon have been found in 4.1 billion year old graphite inside of like zircon rocks. So before you actually have established a really established fossil record on the planet, and they real they know that it's biogenic carbon because it contains a carbon-12, carbon-13 isotope ratio that's a signature of life, and that carbon isotope ratio comes from the carbon discrimination activity of the enzyme rubisco. It must have been around very early on for it to have such a long history of the life on Earth.

SPEAKER_00 9:33

Wow. So cool. We've been at this for a while.

SPEAKER_01 9:36

Yes, makes my contribution look trivial.

The Efficiency Puzzle And CO2 Levels

SPEAKER_00 9:41

Maybe now is a good time to talk about the evolutionary history of Rubisco and this kind of big question about you're working on engineering it, right? But nature, through evolution, has engineered, you could say, Rubisco to fix carbon. And the most common thing you'll hear, the first sound bite you'll hear about Rubisco is this is an amazingly slow enzyme and it's inefficient, and we're not sure why it is inefficient, and people have had a hard time engineering it to be more efficient. Why do you think that is? How can we think about that? It's is that a mystery to you as well?

SPEAKER_01 10:21

Yeah, nature has been at it for a long time, and it's clear that there are there have been there there is a very small window of diversity in terms of rubisco speed. So you have very slow rubiscoenzymes, which have reaction rates of 0.1 reactions per second. Then you have things like plants, which have a reaction rate between two to maybe six reactions per second. And then at the upper limit, now we're starting to find robiscoenzymes that are doing around 30 reactions per second. So that's really the kind of window that you're looking at in terms of speed. That's changing all the time. When I was in grad school, in my story from earlier, the fastest rubiscoenzymes known had a maximum rate of 12. Within that time period, what's that now, 15, 16 years? We've more than doubled the known rats. What actually is the fastest enzyme out there in nature could still be found. We get we have a lot more resources now to find from metagenomic sampling and advanced analyses of these data sets. But I would actually say that's one thing we've been really trying to focus on is it's not all about speed. It doesn't really matter if you're a really fast Rubisco. What matters is if you're an efficient enzyme. So those really fast Rubisco that have reactions of 30 per second, they have what they have a very ore affinity for carbon dioxide. So they're what's called enzymatic efficiency, which is speed over affinity or VMAX over Km, is much, much worse than most slower enzymes that have a better affinity for CO2. And it's that affinity for CO2 that is what really makes Rubisco an exceptional enzyme. When we think about that affinity in particular, one thing that I've noticed is that enzymes that have a really high affinity for CO2 tend to have more of these molecular chaperones that they're that they need to both fold and assemble the enzyme. So plants are almost the zenith of rubisco evolution. They have a very high enzymatic efficiency. They're reasonably resistant to oxygen, which is the other major issue with rubiscocatalysis, fixation of oxygen. And they have some of the strongest requirements for assembly chaperones. So it's a very unstable protein until that rubisco large subunit, which is rubisco and plants, it's a there's an assemblage of large and small subunits. Until that large subunit is actually assembled with that small subunit, it's super unstable, right? So it basically wants to collapse in and of itself and aggregate. So why is that? Is it because recently in the atmospheric geological history of the planet, we have actually have quite low CO2 levels in the atmosphere. And that's really driven the evolution of Rubisco to have this high affinity for CO2, so it can actually operate at a meaningful rate inside of its photosynthetic host.

SPEAKER_00 13:29

So there's a few things in there that I want to try and break down. The first we talked you were talking about efficiency versus its maximum rate. Yeah. And so you said V max over Km, which is this measure of an enzyme's efficiency, Vmax being the fastest cat catalysis that you can get from that enzyme when you have a large amount of substrate.

SPEAKER_01 13:54

Infinite substrate, yeah. That's about kind of an imaginary number.

SPEAKER_00 13:57

And Km being the rate at at half saturation.

SPEAKER_01 14:03

And so Km is the substrate concentration at half of the maximum rate.

SPEAKER_00 14:09

Yes. Uh and a low Km means you're reaching a very fast rate with a small amount of substrate.

SPEAKER_01 14:17

Relative to something that has a higher km.

SPEAKER_00 14:19

And so when we talk about an efficient rubisco, this is something that is able to catalyze carbon fixation at low CO2 concentrations. And so what is what is the spectrum of CO2 concentrations that these rubiscos operate under? Are some of the ones that that are not particularly efficient, are they operating at very high CO2, like at hydrothermal vents or something like that?

SPEAKER_01 14:47

Yeah, that's exactly right. So some of the fastest guys are down in these hydrothermal vents. They're also in underground aquifers. So you have a lot of carbonate minerals that get dissolved into the water, so that elevates the CO2 level there. Even in soils as well, you have percentages of CO2, so tens of thousands parts per million CO2. So, you know, you're thinking 30 to 200 micromolar dissolved CO2 gas. So Rubisco reacts with a gas, which makes it quite challenging to work with as a as an enzymologist. And the chloroplast of a C3 plant. So they're basically just taking in CO2 from the atmosphere, it's getting and it's getting shuttled to the Rubisco and the chloroplast. The operating concentration of CO2 is around eight micromolar. So the Km for, let's say, uh plant Rubisco is usually around 10 to 20 micromolar. So it's operating below its Km and most of the time inside. So that V max of let's say three, the Rubisco in the chloroplast isn't actually even operating close to that. It's much lower. And you can imagine that if you have one of these deep sea guys in there who they have a Km of 300 to the 500 micromolar, they're barely able to do anything at uh at eight micromolar. That's why if you put these guys that are fast into a plant and you try and grow it in air, that plant will perish pretty quickly.

SPEAKER_00 16:20

Because it can go fast under conditions that it's never going to see.

SPEAKER_01 16:24

Yep.

SPEAKER_00 16:25

Okay, so let's zoom out for a sec. We've been getting into the weeds a little bit, and I want to center on some of the motivations for this work in the big picture.

Climate Stress Photorespiration And Yields

SPEAKER_00 16:35

So what are the big challenges in the world? What are the big challenges in the world that Rubisco engineering might help us solve? And and why is Rubisco so important for addressing them?

SPEAKER_01 16:50

Yeah, so we live in a warming world. And in a warming world with a changing climate, we have many challenges for agriculture. Heat waves can decrease yield and directly on the enzyme Rubisco. With those declining yields, we have a still have a growing population. So we need more food in the future, despite the threat of potentially decreasing yields centered largely around the efficiency of photosynthesis. So improvements to photosynthesis are really essential. We just published in in Sal kind of an overview of different symbiostrategies that have been useful in improving the productivity of photosynthesis and where the future of that might go. And so Rubisco's role in that is really enzyme engineering of both Rubisco and features around it to try and make it more resilient to increases in temperature. Benefits to Rubisco also relate to water use efficiency. So if you have increasing drought, you can have more resilient crops to against drought as well. And it's also tied in with nitrogen efficiency as well. Ribisco is the most highly abundant protein in leaves. And so a more efficient Rubisco could translate to a lower nitrogen usage as well. You'll often find that everything kind of revolves around this enzyme in one way or another when it comes to agriculture. So the future of our food will always have rubisco will always have a spotlight shining on it.

SPEAKER_00 18:21

And is Rubisco the is rubisco the limiting factor for productivity in a lot of cases? Do we know that for a fact that if we get a faster rubisco, that's going to result in faster plant growth or more efficient plant growth? No.

SPEAKER_01 18:40

There are many features of so you have the light reactions and you have the dark reactions as they were formerly called, or the light independent reactions. And so they both work in concert with each other. So Rubisco is often rate limiting when you have a lot of light. So you've got an abundance of chemical energy. So that means Rabisco is always going to have its substrate. A lot of the time, there isn't enough light being absorbed by the leaves of agricultural plants for Rubisco to make use of that. So in those cases, Rabisco is not limiting. Despite that, though, Rubisco is still limiting under those kind of low light conditions because of these oxygenase activity and the effect of photorespiration, which is a problem that comes from Rubisco's activity in a way. Also, yes, improving by reducing oxygenase activity of rubisco, I think it's now pretty conclusive that you can improve plant growth and agricultural yields.

SPEAKER_00 19:38

And what is the oxygenase activity of rubisco?

SPEAKER_01 19:42

Yes, when Rubisco fixes oxygen to its substrate instead of CO2, it leads to the production of a compound called 2-phosphogly. And the plant basically recycles this in a process called photorespiration. And you can think about it as the opposite of carbon fixation because photorespiration leads to energy loss and it leads to carbon dioxide emission. The cellular metabolism of the plants actually losing CO2. And there's been a lot of really interesting work being done on what are called photorespiratory bypasses. So they're using synthetic biology to minimize the amount of energy that's utilized by photorespiration. And these have been and are continuing to show me really meaningful improvements and growth.

SPEAKER_00 20:35

And so Rubisco will some fraction of the time, instead of putting a CO2 onto the substrate, it puts an O2.

SPEAKER_01 20:43

And yeah, and this is up to 25% of reactions in uh C3 Plum. Interesting.

SPEAKER_00 20:51

And so does that also correlate with the lower efficiency? Because you were saying that the enzymes that some of the Rubisco enzymes they have a lower affinity for CO2, even if they have a high V max or a high maximum speed. If they have a low affinity for CO2, does that mean they might be more likely to be fixing more oxygen? Or would those also fix more oxygen? And I guess why is it? What is it about oxygen that makes it step in instead of CO2?

SPEAKER_01 21:22

I always understood that it was because they're quite electronegatively similar. So it's just hard for an enzyme to discriminate between a CO2 molecule and an O2 molecule. It would also depend it depends on the chemistry of the catalytic mechanism of Rubisco. And that one important thing to re to that I should bring up as well is that Rubisco doesn't actually bind to carbon dioxide. It doesn't have, as far as we understand, what a traditional enzymatic binding site for any of these gases, it simply manipulates its five, its IEBP substrate, into a configuration where. Where gas can attack it. That's really a key thing to think about. If you're doing that, whether there's a CO2 molecule or an O2 molecule around, it can then just automatically a transition energy barrier to catalyze that reaction to allow that reaction to proceed. Got it. But yeah, so those guys that are really fast, they are typically very sensitive to oxygen. They will have a hard time discriminating.

SPEAKER_00 22:30

The Climate Biotech Podcast is powered by Homeworld Collective, a 501c3 nonprofit unlocking biotech solutions for planetary health by fostering community, building knowledge, and directly supporting early stage research. We are always looking to connect with scientists, innovators, and funders who want to accelerate progress in this space, whether in our current program areas of critical minerals and greenhouse gas removal, or in other applications areas. If that's you, reach out to us at hello at homeworld.bio. Interesting. So there's a lot of trade-offs here in this enzyme, it seems. So, okay, so let's so you've gotten some detail on Rubisco now, some detail on why we're trying to engineer Rubisco.

Engineering Rubisco From Bench To AI

SPEAKER_00 23:15

Let's talk about your research arc. You started at the Australian National University, then you were at Max Planck, now you're at MIT. And it's worth noting that now at MIT you're working in the chemistry department, not the biology department. And so that's uh an interesting aspect here too. Maybe you can share how your research identity has evolved across those phases and what the main focus was for each of them and what kept you focused on Rubisco the whole time.

SPEAKER_01 23:44

Yeah, working in a chemistry department was probably one of Oli's my greatest horrors, but because then people are gonna expect that I'm a chemist where I'm I am much more comfortable in a biology role. And biologists are way more chill, people, especially plant biologists, some of the most relaxing people to work with. So that has been an evolution, but the people at MIT that I've met, including yourself, have been all very bright and talented individuals. So it's been a real privilege to work with them and to have them share the same passion that I do. And yeah, I guess I've evolved as a scientist as well over that time. So I've gone from being someone who's working at the bench to doing project management and leading the goals and directions of the research together with the professor of the lab, Matt Shoulders. And I think one of the biggest that there's been these kinds of stages during my career. At first, when you're in grad school, you learn a lot of new things, you learn how to actually perform experiments and control them, and then become more independent as a scientist. Then when I was working as a postdoc in Germany, I became a lot more analytical. So I learned the value of doing, really taking your time and getting the experiments done in a complete way rather than just get something done and then go, oh wow, that was really cool. What's the next thing? And then move a jump onto the next thing. And then sometimes if you do that, you have to you find you can get lost quite easily and you have to come back. Whereas if you take a really methodical, almost German mindset to that approach, you can really get things done in a clinical way. And it's very satisfying when you've done something in biochemistry and you've done it, then it's it's so clear that it's the only that there's really very few other outcomes that so you can really sure of the result.

SPEAKER_00 25:39

And so from what I can tell, your work has spanned the science of Rubisco and also development of novel methods for engineering and evolving Rubisco. You can share how that how that interplay has worked. Like you're you first started with directed evolution, and so maybe you can share what that early directed evolution work was, and then we can talk about your work with chaperones and your this new wave of directed evolution work that you're doing now. Yeah, is that the right way to think about it? Have I framed it properly?

SPEAKER_01 26:19

Yeah, I think so. There's an interesting story there in terms of I noticed in your introduction you said that I made modifications of Rubisco and then put them into a plant chloroplast and showed them improved photosynthesis. We actually that that is true, but we cheated there because the way we started and the target that we used for directed evolution was a very bad enzyme. So I'm I'm talking like the worst Rubisco you can really imagine. Maybe couldn't even really call it a Rubisco. So this enzyme from it was actually from an archaea from an undersea, underwater lake in Antarctica called embotoni. It has very low reaction rate, as super poor KM for CO2. It's really sensitive to oxygen. So we use that as our starting point for directed evolution, and that made it a lot easier to make universal improvements to the enzyme. Because we made those universal improvements, we then put it into a tobacco chloroplast, and the evolved mutants supported improved photosynthesis relative to the starting point. But these plants took 18 months to grow for the wild type being the wild type archaeal enzyme. Got to have the record for probably slowest growing bona fide photosynthetic plant on the plan on the planet. Where have we come from there? So now we're doing directed, we've now we've done directed evolution on plant Ribisco. So my former supervisor had a paper come out last year during the directed evolution of tobacco Ribisco. That was only possible because we'd figured out the assembly pathway of plant Rubisco in E. coli, so we can use some of these directed evolution tools on these really advanced forms of the enzyme. And we're discovering new things all the time, right? Every time we do work with directed evolution, we tend to find both changes that pop up really frequently. Usually these are involved with improving the amount of the rubisco that's in the system, but then also catalytically important residues as well. And these often track back to the natural evolutionary history that we observe in nature for the enzyme too. So from directed evolution, we're now moving towards an age where we have very good computational systems through which to make mutational predictions. And I think one of the focuses of the work at MIT has been how can we use can we use artificial intelligence or machine learning to better predict sites that could be leveraged in a directed evolution context? So when you do all my previous directed evolution work has done random mutagenesis. So you're randomly making changes to the coding sequence of the enzyme, and then you have a big library, maybe depends on where you're from, but from me, big libraries were the hundreds of thousands of mutants that we were screening in these platforms. The question though is because Rubisco is so uh conserved, how many of those residue changes are unlikely to have any real benefit at all in a random mutagenesis context? So we can use things like machine learning to see what are the best sites to make changes to the enzyme so that we can have smaller, more compact and streamlined diversity that we can screen in these systems and better analyze the outcomes. And so that's been well, we've been focused on that at MIT and we found some really interesting things with plant rubisco. So we've found changes that improve the catalysis of the enzyme, and we've found changes that are not observed in the evolutionary history of land plants. So, yeah, there's some interesting kind of questions there that we're still digging into, but that work should be published pretty soon.

E Coli Screens Chaperones And Assembly

SPEAKER_00 30:15

And can you give us more detail about these? You know, you talked a little bit about the chaperones, and there's also in here there's the screening system, which you developed this screening system in E. coli, which is a really big deal for doing directed evolution. Yeah. Why is directed evolution in E. coli a big deal? And what was hard about making that happen? And how did the how did the chaperones kind of change that picture?

SPEAKER_01 30:44

Yeah, so the first directed evolution systems for Rubisco in E. coli, done long before my time. So they were done, I think it was Chira Metsumura's lab in at Emory University. They had just what they did was express Rubisco's kind of the enzyme that comes just before Rubisco in the Kalton cycle is called phosphorobulokinase, and that makes Rubisco's substrate. So it just so happens that Rubisco's substrate is pretty toxic to bacteria. They really don't like it. It's a dead end substrate for it's a dead-end metabolite for the bacteria. So its accumulation is a toxic effect. So having a rubisco in there kind of detoxifies the system and allows the bacteria to achieve a faster growth rate. And they hate this. So they just because natural this is one of the beauties of natural selection, right? You often just find the easiest pathway to the outcome that you're looking for. And for the bacteria growth context, the easiest pathway is just to get rid of this PRK enzyme that it doesn't like. So you'll get transposons being integrated into the coding sequence of PRK or into the promoter region to kill the expression of it. And of course, this makes the bacteria achieve a wild-type growth rate again. So when I was in grad school, I was doing this work and there were like just heaps of false positives in the experiment. And I kind of got sick of that. So I thought maybe we could just fuse an antibiotic resistance marker to this gene so that if the bacteria try to get rid of it, they'll be killed by the antibiotic. And that was successful in massively reducing the number of false positives during that screening process. Since then, there have been many different forms and systems of doing directive evolution in E. coli that are far more advanced than what I've cobbled together. And these are leveraging changes in the metabolism, the central metabolism of the bacteria, so that it's basically dependent on the Rubisco for carbon assimilation. Or it's not it's not so much about detoxifying REBP, it's like using the activity of Rubisco to assimilate or repair breaks in the metabolic pathway of the bacteria. And these are much more robust than the screening systems that are developed in at the Australian National University. So I think these are really the kind of future of screening systems for Rubisco. And yeah, it's very exciting. I think the combination of a lot of these features, more advanced selection systems, more advanced directed evolution design, and the use of like machine learning, put putting that all together is going to be really powerful for surveying the catalytic potential of the enzyme.

SPEAKER_00 33:37

And where do the chaperones come in? And maybe you can actually just tell us, we can take a little segue into the work that you did on chaperones. You had a big paper come out on the discovery of some key chaperones and then expression of a plant rubisco in E. coli. And so that was a big shift from like the this the microbial rubiscos that you've been working with before, right?

SPEAKER_01 34:04

Yeah, big shift. Yeah. So myself and another postdoc in the lab at the time, Harold Eigner. Harold was actually the one that put all the pieces together that achieved the production of Arabidopsis rubisco, which is a C3 plant E. coli. And then together we worked to really define the assembly pathway of Hlant Rubisco N. coli. And that requires four dedicated assembly chaperones. So these are the guys that once the Rubisco is correctly folded, it all gets assembled up into its final hollow enzyme form. And another really strange thing that we found was that the folding of the Rubisco large subunit, so it's catalytic subunit, requires very strictly the protein folding machinery of the chloroplast. So the protein folding machinery of the coli, which is called Groel, it doesn't substitute for this. So these are called chaperonin, and they're basically these little folding cages. They're like a little bowl that the premature protein goes into, and then it is uses ATP hydrolysis to arrange it into its final three-dimensional shape. These are really promiscuous proteins. So the chaperones then help the folding of many different proteins inside of the cell. And all rubisco require to fold. So there is no example of a rubisco that does not need these. So it's probably it, I would say it is the most fundamental chapel rubisco, and probably the first one to evolve as well.

SPEAKER_00 35:43

Wow.

SPEAKER_01 35:43

Yeah, so the that was really interesting for me that there has been this coevolution of these protein fundamental protein folding machineries to support plant rubisco biogenesis. And then after that, you have all of these other guys as well. And they're RAF1, RAF2, RBCX, and BSD2. Two of those are found in the cyanobacteria, which are the ancient progenitor of the chloroplast, and the other two are unique to eukaryotes. So green eukaryotes.

SPEAKER_00 36:18

So I'm curious about these chaperones. You were saying they use ATP hydrolysis to help the protein fold. So how can they be so promiscuous if they are, or to put it another way, what are they using the ATP for to in in this folding process? I normally think of a folding process as reaching reaching a thermodynamic minimum by going through this kind of landscape of walking down a thermodynamic gradient, and eventually you have this exquisitely folded protein that started out as an unconformed chain. Where do we need chaperones in that process?

SPEAKER_01 37:08

Yeah, so the ATP, whether it whether the chaperonin is bound to ATP or ATP, causes the cage to expand and grow. It's almost kind of a little slinky. So these conformational changes while the protein is inside of the cage must have some, must exert some influence on that folding landscape. So I think I guess it's believed that the addition of the ATP energy in some way allows it to move out of maybe one of those funnel kind of areas, so the valleys in the folding pathway, and move towards a new maxima where it can then explore different regions of that landscape. And how it actually occurs for Rabisco, I really couldn't say. My hypothesis or my prediction would be that there are residues on the sit on the interior surface of a cage that interact with a wide variety of different, what are called client proteins, very sexually charged like language for protein, but it pushes and pulls on this on this encapsulated client to manipulate it in a way that is favorable to its more thermodynamically stable final configuration and not some congealed, malformed thing.

SPEAKER_00 38:32

I see. So it's so like a misfolded protein is caught in some energy minimum that isn't the true minimum, but it's it's it doesn't have enough energy to get out. And so it so it sounds like the chaperones are making sure that it gets jiggered around in the right way to get kicked out of those wrong minimums. That's how I'm thinking of it based on your explanation. That's really cool. I hadn't ever thought about that before.

SPEAKER_01 39:01

Yeah, that yeah, that's how I think about it. And also the chaperone needs to be able to recognize in some way when it has achieved a correct, correctly folded state and when it is in a intermediately intermediate state. So I think it's very rare that you'll get like a fully disordered protein going into a chaperun. And usually as soon as these things come off the ribosome, you have some secondary structure forming a little alpha helix, and then these all have to arrange it to a correct three-dimensional configuration as the folded protein, and these maybe just be slightly aligned or not quite in the right space. They recognize often hydrophobic regions that are exposed on proteins and then make sure they get protected while they fold and to prevent those being exposed exposed to the solvent. And so maybe it just cycles in the chaperone until it can bury these hydrophobic residues and continue its journey.

SPEAKER_00 39:56

Very cool. Okay, so you've been able to improve Rubisco in a few different ways. We've we we've learned how to use these chaperones in E. coli to do directed evolution and engineer plant rubisco. You've done a bunch of engineering on you, and others have done engineering on microbial rubiscos. Are we seeing translation of these into applications? Or are there barriers still to translating that? And do we still need even better rubisco?

SPEAKER_01 40:29

Yeah, I think a lot of the time in science, you have this applied goal, and it can be like a really big, challenging biochemistry, biotechnology applied goal. And you try really hard, you get and you work towards it, and along the way, you learn a tremendous amount about how nature works. And we've learned a tremendous amount about how nature builds one of the most important enzymes in life for life. And the we've learned a lot about the catalytic mechanism and the biochemistry of Rubisco as well. And we have found both my both us and other labs that have been working with the Rubisco have found many residues that offer improvements to the enzyme's catalysis. But how much of an improvement do you actually need? What's considered a success? We've largely the changes that we've found through techniques like directed evolution have still been within the spectrum of known catalytic diversity. These engineering approaches have not succeeded in generating something that is not in one way or another currently or already present in one natural enzyme or the other. The question is, can you actually you mentioned trade-offs before, can you balance these trade-offs in a way to try and change rubisco that will have a meaningful translational impact in a photosynthetic crop? Can you change rubisco to improve crop yields? And the we have these really good mathematic predictions for photosynthetic CO2 assimilation. And you can always plug into those models what would be the effect for CO2 assimilation in a plant if the rubisco had these phonetic parameters. A 20% improvement to its rate, a 20% improvement to its affinity, a 20% improvement to its specificity for CO2 over oxygen. I think the biggest benefits that would be observed are improvements to its ability to discriminate CO2 from oxygen. And uh yeah, so that's that I think that's the bit the biggest goal. Also the hardest. So improving that feature of the enzyme is without a doubt the hardest engineering challenge. And I'm not even sure that the screening systems we have in E. coli are designed to operate on selection on that parameter. You're just selecting for how quickly can you get rid of the substrate or how quickly can you turn over its substrate from REBP into something that it can use. But yeah, I think we're still we're learning lots of things every day. I keep pushing on it because it's really hard. It's a scientific challenge that I think is always going to have new things to find out. And the more we find out about it, the easier it becomes as we continue those engineering efforts.

SPEAKER_00 43:37

So it sounds like so an assay for oxygen sensitivity sounds like a or like a directed evolution assay screen for oxygen sensitivity sounds like a would be a big open problem. Yeah.

SPEAKER_01 43:52

Is that really that yeah, that would be really useful if you had get a system through which you could get a good readout of how much oxygen is activity. Rubisco was exerting inside of a cell or a directed evolution system, yeah, that would be powerful.

SPEAKER_00 44:07

And but I'm still curious, like we've even if you're even if the rubisco that or even if the mutations that you're discovering are present somewhere in nature, if you combine them into one rubisco, can you make a rubisco that is that would perform better than what's in a given crop? Are we ready to do we have anything we want to put in crops yet?

SPEAKER_01 44:34

What tends to happen is if you combine these substitutions, the enzyme will start, will just fall apart. So we talk about trade-offs between these kinetic parameters. The bigger trade-off is actually, in my opinion, between activity and stability of the enzyme. When you look at something like a plant river visco, we already talked about them having all of these chaperones, it's a very unstable thing. It's really on the edge of evolution between how much instability can a protein tolerate and still even assemble it all into it or to a meaningful level the amounts of it's found inside a plant where it's going to have X activity. So I think that's a real trade-off. I think when you make these modifications and you try and combine them all together, what you typically find is that you don't get any assembly of the enzyme. So now you've made it so those mutations might, in a theoretical landscape, make for a super rubiscoat, but it's simply too unstable to exist in that region of protein space.

Stability Tradeoffs And Crop Translation

SPEAKER_01 45:34

What do you need then? Do you need more kind of chaperones in order to further stabilize it through that assembly into its final form? Maybe. Maybe that could be one pathway. Whatever. Or maybe it's just fundamentally constrained in that way. Or maybe you need to start from a completely different kind of starting place. There was the rationale behind using this really crappy Rubisco enzyme in the first place. It's like we'll just go right back to the beginning and we'll use directed evolution, just rebuild a new embrace from scratch, essentially. And although attractive, that that process is going to eventually hit this stability wall, right? Because what also has happened through evolution, and you notice this probably following the great oxygenation event, is that Ribisco has developed a small subunit. So originally Ribisco doesn't have a small subunit, somebody got it picked up along the way. That small subunit stabilizes the large subunit. And typically, Ribisca that have a small subunit are just better enzymes than ones that don't in the vast majority of cases. So this additional subunit has allowed it to kind of, again, get into a region of protein fitness space that allows for a trade-off against the stability of the active site. You see the stability of the active site and all kinds of little things in the Ribisco's mechanism. Ribisco has all these floppy bits on it, they're kind of like flap around. And these are intrinsically tied in with the catalytic mechanism of the enzyme. So much so that when the rubis another problem with rubisco is that it gets inhibited, it inhibits itself, so it gets like stuck and gummed up. And it's got a completely other protein that has evolved to get it out of that state. And how this protein works is that it pulls on some of those floppy bits, and that actually loosens up the active site such that the enzyme spits out the inhibitor. I think that this is probably intrinsically tied to the stability activity trade-off. The active site has to be at some have a certain degree of conformational flexibility, and that's tied in with the catalytic mechanism, and then that's also tied in with the overall ultra structure and stability of the enzyme complex. So you can only push it around so far falls apart. Evolution, natural evolution, has to navigate this like really fine kind of tightrope in terms of stability and activity in order to make meaningful advancements on the enzyme. And that that is really in a bad position at the moment because of the increasing temperature on the planet, which is occurring very quickly, and agriculture and natural evolution and plant breeding, it doesn't have the rubber scopes can't change quickly enough to accommodate for that.

SPEAKER_00 48:20

I see. So that so I was I was about to move into a question about your current work and the future of the Epic project. And is heat sensitivity part of what you're working on there with the Epic project?

SPEAKER_01 48:35

Yes, the Epic project is now moving, it's now funded by the Grantham Foundation for the protection of the environment. And that will continue for the next few years. And one of our singular focuses for that research is how can we improve Rubisco's resistance to oxygen, essentially, or reduce its sensitivity to oxygen. So that sensitivity to oxygen actually increases with increasing temperature. So as temperature goes up, Rabisco is less able to discriminate between CO2 and oxygen. Other researchers and collaborators in the field are focusing on internet helper, chaperone, it's called Rubisco Activase, it's the one that helps Rabisco spit up its inhibitor. That is a real failure point for thermal stress on plants. So it's very thermolabile. So you can get heat waves, especially in wheat, and they can wipe out the productivity in those wheat plants if they're exposed to heat waves early in their vegetative stage. So our focuses kind of on the rubrisco side of that. So how can we achieve improvements to the specificity factor of the enzyme and translate these two improvements as well? So trying to create more resistant, more resilient forms of agriculture that are very likely going to be needed in the future.

SPEAKER_00 49:58

And so part of the goal of the Epic project, it's enhanced photosynthesis in crops, is translating some of these new rubiscos into crops. And so what are the challenges to doing that outside of getting the rubisco working on its own, kind of in the lab?

SPEAKER_01 50:20

The single most challenging aspect to delivery is what can you what crops can you actually put these changes into? So chloroplast transformation, which is necessary to make alterations to the rubisco-large subunit. It's one of the few genes that are encoded by the chloroplast genome. This is a technique that has very limited spectrum of crops that it can be applied to. So you can modify the nuclear genome of a much wider variety of crops than you can the chloroplast genome. So that's a limitation. Another limitation is can you even grow a certain crop on a tissue culture medium? So these are the like a synthetic growth medium in the lab. It's basically sucrose and like a gel that and that has hormones in it that allows you to form a callus and basically manipulate the plant and then regenerate it from a single cell once you have identified a successful transformant. This is also highly limited. So advancements in those that forms a fundamental block on what can be achieved. You've got advancements that are happening all the time. So you have CRISPR-based genetic modifications in plants and increasingly also in editing the chloroplast genome. There's some very uh interesting work happening in Japan on that front as well. So making changes to chloroplast DNA using CRISPR. We have these limitations, but then we also can, with the advancement of technology, overcome them in certain ways as well. And who knows what happened, what might happen in the future. We might be able to maybe we can do away with tissue culture. If that was possible, that would be a huge advancement.

SPEAKER_00 52:00

Sounds like there's some pretty important work on the horizon for getting all of this to come together. Yeah. We got smart people working on it.

Rapid Fire Takeaways And Advice

SPEAKER_00 52:09

So we have reached the point in the discussion where we go into these rapid-fire questions that we ask every every guest on the show. And these are just to get some nice nuggets from you about what has inspired you and and little ideas about provocative ideas and uh and advice for others in the field. So the first question is could you tell us about a single book or paper or piece of art or idea that blew your mind and shaped your development as a scientist?

SPEAKER_01 52:42

Yeah, I think the I actually remember the first time I knew I wanted to work on directed evolution that was when I saw George Church's, I think it was Mage, the Mage paper, when I was an undergrad. And I was like, I was like, oh my god, this is incredible. And that just yeah, completely inspired me and as a scientist. And also like neuromansor, I think that's pretty cool as well. And maybe with the future of AI, we're gonna have more horrors to keep us entertained.

SPEAKER_00 53:12

I think that's guaranteed.

SPEAKER_01 53:14

Yeah.

SPEAKER_00 53:15

Okay, what is the best line of advice that a mentor gave you?

SPEAKER_01 53:20

Don't burn the lab down, which I've been successful at, but I've almost burned down every lab I've worked in.

SPEAKER_00 53:27

Love it. If you had a magic wand to get more attention or resources into one part of biology, and it can't be specifically what you're working on, what would it be?

SPEAKER_01 53:37

One part of biology. That's rapid fire, right? I don't know. Probably whatever my friends are working on, because I want them to be happy.

SPEAKER_00 53:47

Good call. Okay, give us a hot take. What is one view that you hold related to climate and environmental biotechnology that you think some others in the field might disagree with?

SPEAKER_01 54:01

Yeah, this is probably the one I'm quickly becoming a zealot for at the moment. I think there is no solution to climate change unless we leverage agricultural biomass. So permanent burial and storage of agricultural carbon.

SPEAKER_00 54:21

And finally, what is one aspect of personal development that you think that what is one aspect of personal development that you would encourage biotechnologists to spend more time on?

SPEAKER_01 54:35

I think this would be for any scientist is to draw boundaries. If you're unhappy with how much you're giving to or how much time is being taken away from with your interests, it could be anything, could be your job, could be anything in your life. Yeah, draw boundaries around that and then respect them. And you'll find that you become a lot happier. We had a graduate student in the lab, Julia McDonnell, very talented scientist now working at Rockefeller. I really respected her, I really learned a lot from her because I respected how she could work really hard, but she also took her own personal time and respected her own personal time. And I think she was uh one of the most productive scientists that I've ever worked with, despite that. So it just goes to show that if you have personal boundaries, you can still do great things. You don't have to be a workaholic like me. You don't have to spend 24-7 thinking about scientists, science, and challenges.

SPEAKER_00 55:37

I love that one.

SPEAKER_01 55:38

Gotta touch grass, man.

SPEAKER_00 55:39

Yeah, I'm working on that right now, too.

Closing And How To Connect

SPEAKER_00 55:42

Okay, this brings us to the end, Robbie. I'm really grateful for you taking the time to talk to me today. This has been fascinating to take a deep dive on Rubisco and what it what engineering of Rubisco can do for the world. So thank you so much. Thanks for tuning in. I hope this has been educational and inspirational for you as you navigate your own journey to bring the best of biology into planet scale solutions. I'll be back soon with another conversation. In the meantime, you can stay in touch with Homeworld on LinkedIn, X, or Blue Sky. Huge thanks to our producer Dave Clark, along with Paul Himmelstein, Kayla Sims Austin, and Mikario Sarsoza for making these episodes possible. Until next time, I'm Paul Reginato, and this is the Climate Biotech Podcast.