Scientist Stories: Emmanuel Levy, Self-Organization in the Proteome

So we're going back in time, way, way back. Emmanuel was, I think, my third PhD student. Um, so this is really towards the beginning of the lab. Yeah, 2001, well, or fourth, third or fourth. And, um, and he's gonna go back kind of to talk about protein complexes, protein biophysics. And I really look forward to the talk, Emmanuel. Over to you.

Hi.

So cool background. I should also say that my font comes from Emmanuel. You can see he's really talented with art, and the Avenir font that I use, that was Emmanuel's innovation. But I can see you're using a different font now. You've moved on.

So students of mine taught me another font. But Avenir is coming next, because this is not my slide, and after it's my slide with Avenir. But yeah, so I would like to first thank you, Sarah, because as you've said, mentors are important, and I think you've been an amazing mentor for me. And hi to everyone who's here and with whom I've interacted— it's been a privilege to be part of your lab. To link to the title of this talk about chance and necessity, I think that me ending up in your lab was actually a lot of chance. I don't have time to tell the whole story, but as you know, it was basically a chance interaction with José in a different location.

Oh yeah, yeah, now I remember.

And I think that really, like, drove me to come working with you, and not only because of me, but like, I think José makes the interaction. And so, yeah, so all of this to say that in the context of me and the Teichmann lab, chance was at play. Now, what about proteins? So for those of you who don't know me, you actually might know me through that picture which Sarah showed in the slides earlier. This was the first time that I used Blender to render protein structures. That was the peak of my career as an artist. I think it was the best picture I ever produced. I could never reach that again. However, some people could, and this is another beautiful picture from Gael MacGill, and this is what I think Actually, I think, Sarah, you would be best placed to get there, because this would be single-cell structure. Single-cell structure. I think this is— if we continue the thinking of Sydney Brenner, I think this is where we need to get if we are to truly understand what's going on. And with Christine, we wrote a perspective earlier this year, and we coined this structuromics. So I hope that structuromics will come of age sometime soon. And this is somehow also what motivates the research in my lab. So I will first, like, just give you an overview of what we do in my lab in the Weizmann Institute, and then I'll go back in time and really go back to this idea of interactions and protein complexes to answer the question in the title. So in my lab, we're motivated by really this idea that we know a lot about the building blocks And if we think of AlphaFold, we even know the structure for many, if not most, of the proteins in a proteome, yet we don't know how they interact at the structural level. We don't know the quaternary structure, and we know very little about the higher-order organization of these proteins in cells. So a lot of our work is motivated by this, in that we try to characterize the quaternary structure of proteins, and so, so Sharita has been doing amazing work in this respect. And we're trying to understand principles of the organization of proteins in cells using different approaches, protein evolution, but also in vivo experiments where Hector showed that point mutations can actually trigger proteins to form from complexes to self-assemble into these gigantic structures. And we also use synthetic biology to verify that we do understand the principles of formation for these mesoscale assemblies. And I would just like to highlight quickly Meta's work, because it's a work that mixes assembly with RNA. So I think it was well suited for this audience. So what Meta did is to engineer a two-component protein system. You can see the two building blocks here. It's a dimer and a tetramer. The dimer has a binding site in orange that can bind to the blue site on the tetramer, and it can only link two tetramers intermolecularly. It cannot— the two sites cannot bind on the same tetramer. As a result, when we express those two proteins, they form a mesh inside the cell. And if I go back to the movie, this mesh phase separates to form this puncta, and we can actually characterize the phase diagram of those puncta, but I'm not going to show that. What I would like to show you is a simple idea that we had, which is that there could be a link between translation and RNA localization and organization in cell. What is this link? It's just that when a protein is being translated, the ribosome links the protein to the RNA, so you could use a ribosome as a bridge. And here we were wondering whether when the dimer is synthesized, it might— protein folding is way faster than translation, So, this binding domain could bind to, could recognize its targets onto the condensate, which is very large, way bigger than a ribosome, and as a result, localize the mRNA that is being translated to the condensate. So, in other words, the co-translational binding of the dimer could drive the localization of an mRNA, and we know many mechanisms that can help or that have evolved to localize mRNAs, I would like to think that this is actually the simplest of all because it's just encoded within the RNA, within the proteins encoded by the RNA. So to test that, what Meta did was to fuse MS2 loops to the mRNA, which are recognized by the Cp protein which is fused to GFP, so that now we can actually express this in cells. And we have the condensate in red and the mRNA in green, and we can ask whether they colocalize. So here you have a yeast cell, and you can see the condensate in red and the two mRNAs bound to it. And if we take a movie, you can see that they move along with it, so they clearly colocalize. Not only this, in this case we could see the mRNAs before the condensate. So that really tells you that in the cell, if you think of mRNA localization, before you think of any mechanism that could— orthogonal mechanism to localize this mRNA, you The simplest of all would be that the protein encoded by the mRNA will bind to a specific place. Okay, enough for this. Now I will quickly go over this idea of the chance and necessity in the evolution of protein interaction. So you can see two things in this picture. One is that scientists were real gentlemen a few decades ago. The second thing is that Max Perutz is holding hemoglobin, which has a stoichiometry of A2B2, two alpha chains, two beta chains. John Kendrew is holding myoglobin, which is a monomer, and you can feel that John Kendrew, although he's very proud of this big model, is maybe somewhat jealous of Max's higher stoichiometry. So there are many proteins that have this stoichiometry A2B2.. And what I would like to ask now is, are all of the complexes with that stoichiometry equal in the face of evolution? So I'll give you two very distinct examples. So this one is an ABC transporter. It's composed of one homodimer, and that homodimer binds another homodimer. So you can imagine it as two homodimers forming a hetero interaction. In contrast, here you have a homodimer that's pretty big, and two satellite subunits in green that bind to it. So in this case, the two green subunits don't interact with each other. So here you have a monomer that binds to a homodimer. Both of them have a stoichiometry A:2B:2, but these two are actually, we think, very different in the face of evolution. And so these students actually investigated this question, which came somewhat by chance in their data science class, So these two students were just part of the class, and Hugo, he's a postdoc in my lab, and he was a TA for the class. So we did this all together. So what do I mean by equal in the face of evolution is just if you think about what are the chances to evolve an interaction between any two proteins. So if you think of two monomers, they can interact in many different ways. So they can interact like this, like that, like that, like that, like that. And like hundreds of thousands of ways in which they could potentially interact. So in other words, there are lots of possible interactions that exist between two monomers. So we would like to think that two monomers are highly evolvable— interaction between two monomers to form a heterodimer. Now if you think of a dimer monomer, it's the same thing. The monomer can bind here, by symmetry it should also bind on the other side, but again, it can bind in really in any way it wants, except perhaps at the interface. So there are also lots of possible interactions. Now, if you think of two homodimers, we have very, very different scenarios. And that's when we realized this, we said, huh, that's actually an interesting question, because it cannot bind like that. The reason it cannot bind like that is that it cannot form a stable complex, because another— this dimer will then also bind to this subunit. And also there, and it will continue to infinity. So if two homodimers want to form a finite and stable complex, they cannot interact like that. They have to align their axes of symmetry, and this dramatically reduces the number of ways in which they can productively interact. So just this reduction of the potential space to evolve an interaction might lead to a reduction in the numbers of complexes having this topology. So that's if chance is important. If necessity is important, it doesn't matter. What matters is what's selected, and even if the space was small, it would be found and they should be equal. So by looking at the ratio of the frequency of these two complexes, we could get a feeling for whether chance or necessity is more important. So what we predict is that these should be less frequent than this. Now what Hugo has been doing is really quantifying this in a much more accurate way, but this is work in progress. This is literally from yesterday, where we, instead of showing animated PowerPoint cartoons, we actually try to quantify the number of ways two proteins can interact. Now if you take this dimer of homodimers and there's no restriction on symmetry, it can bind— they can bind to each other in 300,000 different modes. Whereas if you put a constraint on the symmetry— I should move this maybe— then we only have now 300 productive modes. So you have this reduction of 1,000-fold. So is this reflected in the frequency of these two topologies? So to do that, so the students, what they did is to take these topologies here and the topology associated to this evolutionary scenario and just count the number of distinct topologies that we have in each. And so what we see is that it's all about chance, that we have 97% of A to B to having this topology and only 2.5% of those. So it's a huge difference. It's not like, you know, like 20, 60, 40%. It's 2.5% and 97%. So this is really hard to evolve, and it almost does not exist. So, yeah, so it's all about chance, not only for me and the Teichmann Lab, but only for proteins. So now, yeah, so this is now the lab in the Weizmann Institute. It's a group of brilliant and talented people, and I must say, Sarah, that often I think, and it helps me a lot when I'm faced with the difficulties, what would Sarah do in this position? And I get the answer. So thank you again, and thank you all for your attention.

Yeah, I wish it were that easy for me to find the answers. So think about what you would do. I'm glad you're still thinking about the Homo mars, hetero mars, and so on. That's fantastic. And, and I'm also glad that I invested my EMBO Young Investigator Program money into hiring you, Emmanuel. That was complete chance. That was complete chance as well. Yeah, I had it at the time. Yeah. Um, because basically at LMB at that time, if you're non-UK, you, you, you weren't automatically eligible for the, for the PhD stipends and so on for the salary part. So I paid the salary part for Emmanuel out of a prize or award that I had, and that was the best investment ever. That was also— well, not chance, it was strategy. Emmanuel, I could see the genius in you.