Episode 6 - Giulia Palermo: Reimagining scientific creativity, the RNA revolution, and truly multiscale systems

In the sixth episode, Giulia Palermo and I discuss the challenges of studying truly multiscale biomolecular systems, such as the CRISPR/Cas9 complex she famously got involved with. While such problems can take us way out of our comfort zones, they also provide us with clear pathways to innovative and transformative science, something very much needed in the new revolutionary era of RNA biology. In this vein, Giulia also reflects on the nature of scientific creativity, the component that goes beyond our formal training but that can eventually spell the difference between a groundbreaking discovery and run-of-the-mill research.

Transcript

Milosz: 0:01

Welcome to the phase space invaders podcast, where we explore the future of computational biology and biophysics by interviewing researchers working on exciting transformative ideas. Today I'm joined by Giulia Palermo, a prominent group leader at the university of California Riverside, renowned for her pivotal contributions to understanding the molecular intricacies of the CRISPR Cas9 system. I remember meeting Giulia at the conference several years ago, where she was just establishing her lab, and a very good friend of mine decided to join her group as a postdoc. Since then, it was a real pleasure to observe the great research that came out of that interaction. So our conversation revolves around the fascinating challenges that are specific to such multi scale systems as Cas9, a protein where allostery, chemical reactions, large conformational changes, sequence recognition, and mutagenesis all contribute very important roles to the biological and biomedical function of the complex. We also venture into the broader revolution in RNA science, Where new discoveries are constantly shifting the paradigms of cell biology. Finally, Giulia is making a point that for truly transformative science, we need to rethink creativity and sometimes go against our formal training and scientific habits to arrive at real breakthroughs even if that means keeping a pet project in a drawer for a decade. So let's go. I hope you like our discussion. Giulia Palermo, welcome to the podcast. Good to have you here.

Giulia Palermo: 1:51

for inviting me.

Milosz: 1:54

So in previous episodes, we've already talked about the perils of idolizing Nobel prizes, but your research after starting in the De Vivo group, and then with Ursula Roethlisberger, took a very interesting turn when you went on to work with the team that started the CRISPR Cas9 revolution. And the nobles aside, I was an undergraduate myself back then, but I remember very clearly that when that science paper dropped in 2012, it was promising to change everything in biotechnology. So do you mind sharing the story of how you became involved in this collaboration?

Giulia Palermo: 2:31

Uh, yes. So, um, I was a postdoc at EPFL in the group of Ursula Rettlisberger. And, um, I had, I was writing a fellowship to join the group of Andy McCammon at UCSD. So I was in the process of developing a project, a computational project to study with the, um, with the tools developed by Andy McCammon. And I was reading the literature about, interesting things to study. So Ursula gave me a lot of time to actually develop my own ideas and she was very generous. She suggested to study something that was related with viruses, but I was not completely sure I wanted to study the dynamics of a virus. It was not something that excited me. I was very excited about the interplay between protein, RNA, DNA, and, uh, so I started reading the literature better, and once I found this, uh, uh, this system, I found this paper, and what was very intriguing of this paper is that everybody was interested in it, but nobody could understand how it works. Because at that time, it was very early, in 2015, 14, nobody knew how it worked, and I was very interested in understanding literally how it works, right? How it binds nucleic acids, how it cleaves. And I thought, this is going to be my project. So I sent an email to Andy McCammon and I said, I want to write a fellowship on this project. And he said, Oh, that's fantastic. Yeah, we should do it. We should do it. And yeah, it was very encouraging. So I wrote this fellowship which had a lot of, uh, initially people were not really convinced that it would be successful. I remember that reviewers say that the system is too big, you're not going to get reliable, reasonable dynamics, system is too big, structures are too poor, this project is going to fail. And, uh, yeah, this was a little bit disappointing, but on the other side. I knew it would have been successful because I knew that, uh, yeah, the system is big, but we will initiate starting looking at the dynamics of the system. Then we will look at the open questions and maybe we will use free energy methods to study each open questions. There is still a lot to do with refining the cryo maps. So I knew it would've been as successful. You know, I knew it for, from the first day I saw, uh, a cover on Cell about it. I saw that cover and I saw a protein, health resolved, half solved, uh, then D-N-A-R-N-A, you know, one, once I saw that picture, I knew that would've been my project. So

Milosz: 5:29

So there was a

Giulia Palermo: 5:30

that's how we started. Yes, and then I, I contacted Martin Jinek at University of Zurich. And, um he helped me with, supporting the experiments that we would do during the fellowship. And then I went, uh, I got this fellowship from the Swiss National Science Foundation. I went in the lab of Andy McCammon. We started refining the structures with cryo electron microscopy, uh, refinement methods. We, uh, studied the large scale conformational dynamics with, a new Gaussian accelerated molecular dynamics, uh, method. And then once I met Jennifer Doudna because she visited UCSD, she visited UCSD and Andy introduced me to her and she, um, in turn introduced me to her, her now former student Janice Chen, who is now a very success, very successful founder of companies of, on gene editing. And so I started working with Janice and, uh, yeah, so that's how it, it, uh, it went, yeah.

Milosz: 6:36

Great. I think, I know a lot of cases where this kind of vision or a dream of addressing some big problem in science drove scientists for many many years in, you know, crossing the frontier.

Giulia Palermo: 6:49

Yes, you know and sometimes you also get a little bit upset, right, when you get this review and say, this is impossible, this is impossible, and you think, you know, what I thought is, you know, I had this, uh, the way how I reacted, you know, I thought, what is he talking about? This is possible. This, he says that this is not possible because he or she or they want to do it. No, I'm going to do it first.

Milosz: 7:16

That's a

Giulia Palermo: 7:16

what I thought actually.

Milosz: 7:18

That's a good point.

Giulia Palermo: 7:19

Yeah.

Milosz: 7:20

But that brings us to the, well, to the multi scale aspect, right we were kind of used to doing things in just a single way, which is you take the atomistic structure of something, you run an atomistic simulation, maybe with some restraints, And then you try to answer questions, but systems like Cas9 require a completely different approach. As you mentioned, as you alluded to already, you have to integrate experimental data. You have to connect multiple levels of description because there's chemistry involved. So you need a quantum level. There is huge conformational movements, binding and binding events. So that's a perfect case for advancing the field for many, many years to come.

Giulia Palermo: 8:06

Yes. And that's why I also was fascinated by the system because when I started, structures were not complete, right? So we had to work using cryo electron microscopy tools to refine the structures, add the pieces of the structure that were not present. And these actually brings to a point that I think is very important. The fact that cryo electron microscopy now is giving structures of systems in their native environment, which are very big, very large. And molecular simulations now have to cope not only with the time scale problem, but also with the size of the, of the system, right? We cannot anymore think that we are gonna do some transformative work, thinking that a small ligand binding to a small protein, which is something that has been done for so many years, right? But if we want to do something transformative, we really need to go toward what experiments are doing, be, be. Sometimes even be better than experiments if we can, right? We need to be very bold and very courageous in under this point of view, studying larger systems, large scale, conformational changes, and also how These very large systems bind with each other, under long range effects. For instance, um, something that I think now would be very timely, is, uh, Brownian dynamics. Uh, this idea was very futuristic when they implemented it, I don't know if 20, 30, or 40 years ago, when they first had the idea of Brownian dynamics. Now, It's very important because you have this very large system, you want to know how they associate with each other, and, um, I think this is one of the frontiers that we want to pursue. But another frontier is certainly, studying these systems holistically from the very small conformational change or catalytic aspect that is happening at the active site level nucleic acid binding that happens on larger scales, for instance, for which we have to integrate, cryo and refining methods, quantum mechanics, free energy methods, uh, altogether to establish the mechanistic function.

Milosz: 10:26

Right. I think for me, the biggest transformative moment in my understanding of molecular biology came when I understood how small or very tiny energy preferences can give rise to those huge movements, right? That some contacts become established that are very weak initially and they can propagate. And, uh, we have this tendency to think in terms of pre conceived movies, that things know how to move, things know how to rotate, how to open and close. But in reality, there are those very tiny, very subtle forces at play. And I think this is the biggest challenge for this multi scale methods, right? Because you want to retain the detail, but also gain the large picture,

Giulia Palermo: 11:15

yes, exactly. That's why I think in computational chemistry and biophysics, every contribution is important. So the contribution of people that develop. Methods to reach higher accuracy is very important, at the same time is important the contribution of people that want to push the frontiers of what really computational biophysics can do. Right? So of where we can go in solving real problems and this also brings us to another aspect that I think with our job with computations, we need to give, Results and give answers that are real so that can be really confirmed with experiments. But not only that, that can predict experiments. That's where when we can be really successful when we can predict something that couldn't have been done with experiments or couldn't have been done with such a timely, fashion or, uh, uh, maybe would have required, much more expensive, uh, experiments. Equipment, experimentally.

Milosz: 12:21

Right? This seems to be now the holy grail that we're approaching.

Giulia Palermo: 12:25

Yes.

Milosz: 12:25

Having, predictive power in

Giulia Palermo: 12:28

Yes, I completely

Milosz: 12:29

modeling future results and guiding the experiments. We've also had a few comments in previous episodes of the podcast, exactly on this point. So definitely this is the right moment to start thinking about this very deeply And that brings us to the second point that you mentioned initially, which is how we need to get creative about our scientific ventures and how the future of biophysics or computational biology will not just rely on doing more of what we're doing today, but also on coming up with completely new ways of handling the problems, right?

Giulia Palermo: 13:08

Yes, because the way how we are pursuing science in general, not only as a computational biophysicist, but also, uh, people in structural biology, synthesis. What people do, people like to work on things where they, when they are comfortable. It's like your mind gets comfortable when you do things, when you read things, or when you pursue things where you are fine, you're comfortable in. And this is how. We, we pursue our science, right? We are comfortable in what we are doing, but when we really get an improvement is when we get out of our comfort zone and when we put our brain in a situation of instability, when you basically try to reach a peak out of your plateau of your comfort zone, you try to push hard to get to another level and ideally to get a paradigm shift in what you are doing. So this requires, I think, not only, formal training, an excellent formal training in terms of math, physics, uh, statistics, but also creativity. And, uh, creativity, I think, goes beyond formal training. Sometimes formal training is completely against creativity. And you have to be able to embrace creativity to be transformative in science, I think. Mm

Milosz: 14:28

It is interesting because it's Kind of reminds me of the sparse rewards problem in, in reinforcement learning because you need to put in a lot of work that goes against the typical rewards of the field, right? So you're not getting grants by doing something that's completely out of the box but you need to invest a lot to get the reward at the end, which might be getting to something completely distinct. Okay.

Giulia Palermo: 14:55

Yeah, absolutely. I agree. Well, I think that, uh, clearly somebody that, is supporting your work, Financially and financial agency would like to not throw away money, you know, in something that is not bringing to any relevant, uh, work, any relevant, publication. So that's why I think for the funding agency, it's very important that you do solid work, work that, uh, will lead to publications, but at the same time you have to think as a scientist. I think you can have a side you can work on your main grant, on your main papers, things that will give you the papers, right? You want to have your papers. First of all, you want to have your papers to move forward, but at the same time you want to think at something that is transformative, that can lead you to new discoveries, right? So I think that, what my lab does is to work in a very solid way on things that will give to my group members publications, which are also important for their future career. If they want to be they want to continue in academia or industry, in any case, they need an excellent CV. And for this, they need publications. But at the same time, we need to do something that is new, is transformative, that can go against what people believe, people think. Something that can be, um, Yeah, transformative is the right word, which I came to know is a word invented by NSF.

Milosz: 16:28

I know that's what

Giulia Palermo: 16:28

Yeah.

Milosz: 16:29

aiming for also within this podcast.

Giulia Palermo: 16:31

Yes.

Milosz: 16:32

very well yes it's quite important to have your pet projects that sometimes can take 10 years to, to yield something that's publishable or even showable to the people around. Right. But

Giulia Palermo: 16:44

Exactly.

Milosz: 16:45

the amount of thinking you put over your, I don't know, morning routine in those 10 years accumulates to something that nobody else has thought. It is really hard. I mean, I appreciate a lot how hard it is to come up with something that nobody in the scientific community has come up before. So very often you think of something and then you realize, you know, Oh, someone in the eighties already thought of that. And it's, um,

Giulia Palermo: 17:12

true. Yeah.

Milosz: 17:12

we're really competing with a lot of smart people. in the field, but what are your thoughts

Giulia Palermo: 17:17

don't think that we really.

Milosz: 17:18

creative? Yeah,

Giulia Palermo: 17:23

because, uh, so it's not an egoistic effort. So so you are creative in everything you do. You put your creativity in, uh, doing your paper, you have to be creative in studying a specific mechanism, you can be creative in developing a DFT functional, right? You can be creative in anything, and this effort overall in the evolution of science at a certain point will bring to a paradigm shift. So I think we have to see science a little bit more. broadly than with our self, that just us, right? Our selfish effort.

Milosz: 18:08

hopefully there is a back and forth when someone gets inspired. Again, that's also what what I want to do here to cross pollinate those ideas whenever someone has a pet theory or pet idea that they want to share with the community. I hope that's makes other people, interested or inspired that would be a great outcome

Giulia Palermo: 18:30

yes

Milosz: 18:31

so, um this out,

Giulia Palermo: 18:32

that I'm also very interested in, if I can say, so we say the, revolution of structural biology with cryo EM and how we have to cope with that and be able to, this is a challenge of course At the same time, I think another revolution that really, really intrigues me is, what RNA can do. I think this was shown through, of course, the vaccine, right? But if you read the papers in RNA biology, you figure out that there are so many new mechanisms that can lead to therapeutic strategies. And these mechanisms are so exotic with respect to what we were thinking is therapy. So we were thinking of therapy as computational chemists, as drug binding on the receptor, right? So every type of drugs binds the receptor, but now the field of RNA is bringing out so many new mechanisms. So many new mechanisms that are unexplored, that are unknown, where we can really say something. Where we can, we can even invent the mechanism. It's so interesting and another thing that makes me excited is, uh, the role that microbiology is having. The increasing role of microbiology, which has always been important. But, we are figuring out that in microbes, in bacteria, there are things that, uh, we did not believe. They, they're in, like we thought bacteria and viruses are less complex systems with respect to the human cell, for instance. But in reality, there are so many things that bacteria can do. We have not yet explored them. Microbiology is becoming ever increasingly important to discover these biological functions and lead to new studies,

Milosz: 20:25

Yeah, of course, it's true that Cas9 came from a bacterial system and, uh, Well, they had billions of years to evolve with a very very high reproduction rate. So that's a great playground for evolution. On RNA, I agree very much because, right now I'm switching to focus more on RNA myself. And sometimes even figuring out the abbreviations of all the types of RNA that exist is you know, illuminating. You didn't know that such a thing existed, that such a mechanism existed until you read that there is an RNA classified as something and then you have to study really the whole mechanism of how it works to understand what it does. Yeah, it looks like really the next boundary of what biology and biomedicine will try to use to to push the boundary of therapy.

Giulia Palermo: 21:19

yes, using biology to cure biology, so. So,

Milosz: 21:25

again, you know much better than me that the Cas system is much richer than just Cas9, right? So, there are many, many functions or many, many tools that are RNA based that you can find inspiration for from these bacterial systems

Giulia Palermo: 21:41

yes, now, nucleic imaging, uh, these are all, fields where, uh, CRISPR systems can be applied, not only gene editing, uh, but also, uh, as I say, the nucleic acid detection, imaging is, uh, so, uh, these are also very important fields. Yes, Casual.

Milosz: 22:06

Cas12, right, was uh, involved in, I think, shredding RNA, uh, what was the function.

Giulia Palermo: 22:13

so, we started Cas12. Cas12 is another very interesting system because this system paved the way for nucleic acid detection and was actually instrumental during pandemic to develop detection tools that were very rapid SARS CoV 2 virus. Uh, another very interesting protein that, uh, differs from Cas9 because of the presence of a single, catalytic site, and there was a bargaining question. How could the system cleave both DNA strands if one was so far from the active site, which is something we studied using ANTON2 and, free energy simulations. There Akash Shaha from my lab. Um, just, uh, published, his paper on that. Another system is called Cas13, and Cas13 is very interesting because is a system that targets RNA and is used for RNA imaging and, detection. So this system had a very intriguing allosteric mechanism that we were interested in studying. But while we were studying this allosteric mechanism, We found that specific residues could help the selection of the target nucleic acid against off target sequences. So we proposed to Mitchell O'Connell at the University of Rochester to study The off target effects, uh, mediated by these mutations and we found that, uh, mutations can improve the selectivity and we also patented the, the variants. Yeah, I think the field

Milosz: 23:51

that sounds amazing. so many tools, so many applications, and for every combination of tool and application, there's also so many modifications, um, you can study.

24:04

Okay, Giulia. It was a pleasure having you on the podcast. Thank you for the amazing conversation and for sharing your thoughts for sharing your ideas and, uh, have a great day. Thank you, Milos. Thank you for inviting me. Thank you for listening. See you in the next episode of Face Space Invaders.