EP51: Egg Activation: Cell Regulatory Systems with Dr. Jim Ferrell

In this episode of The Future Conceived, host Cam Schmidt sits down with Dr. Jim Ferrell, Professor of Chemical and Systems Biology at Stanford University. Known for his pioneering work in the logic of cell signaling, Dr. Ferrell discusses the "mechanism" of life through the lens of physics, chemistry, and mathematics.

Dr. Ferrell shares his journey from a triple major at Williams College to becoming a leading voice in Systems Biology. The conversation dives deep into:

• Biological Circuits: How evolution uses motifs like negative feedback and relaxation oscillators to create "all-or-none" switches and rhythmic pulses in cells.
• The "Blender" Experiment: A fascinating look at how frog egg extracts can self-organize from "homogenized garbage" back into complex, cell-like structures.
• Quantitative Reasoning: Why thinking like a physicist—using ordinary differential equations and reaction-diffusion models—is essential for moving biology beyond "stamp collecting" and toward a unifying theory of how life builds and repairs itself.

Whether you are a trainee or an established researcher, this episode offers a profound perspective on how the integration of physical forces and biochemical activities brings about the events of life.

Transcript

One of the important things about life that enables life to be alive is its ability to build itself, also to repair itself. But you have lots of examples on different scales in biology of things that you can break apart, and they will just put themselves back together again. Hello listeners, and welcome to another episode of The Future Conceived, the official podcast of the society for the Study of Reproduction. I'm Cam Schmidt, assistant professor in the Department of Biology at East Carolina University. Today is the third episode in our mini series on egg activation. I have the pleasure of chatting with Doctor Jim Farrell, professor of chemical and systems biology at Stanford University. Doctor Farrell has made incredible contributions to our understanding of cell cycle regulation and Xenopus oocytes, eggs and cell free extracts. His approach to research is a perfect synthesis of empirical data and quantitative reasoning about the logic of cell signaling circuits. His book and many wonderful review articles have had a huge influence on the way I think about and teach cell biology. I really enjoyed learning from our discussion and I hope you will too. Thank you for joining us on The Future Conceived. Thank you for having me. When I go to conferences, almost every presentation I see is kind of claiming to have fully outlined the mechanism of some sort of control process. My research is is somewhat interdisciplinary. I work with physicists and computer scientists. And so I started asking people that question like, what? What do we mean by mechanism when we say mechanism in biology? And what struck me was that I was getting very different answers from different people. That seemed to kind of depend on the backgrounds of the people that I asked. So I thought it would be fun to find people who are working on similar problems, or at least using the same kind of system, but probably have different perspectives about how we think about mechanisms and how we should quantify mechanisms or model mechanisms. What do you think about when you think about mechanisms, or how would you define mechanism in a biological context? I'd say mechanism is sort of how things work, and you can understand how things work at all sorts of different levels. But I think the maybe the reason that the journals are so interested in having their papers be mechanistic these days is because, um, biology is, I think, or at least historically, has been, um, looked upon as a, an unusually descriptive science, um, especially by people from like, physics. There's a famous quote from Ernest Rutherford, the, um, the physicist, great experimental physicist who, um, kind of conceived of the idea that the the nucleus was a tiny proportion of the volume of, of the atom. Um, he said that all science is either, um, um, physics or stamp collecting. And so, um, in that regard, non-mechanistic stuff is stamp collecting and biology gets to collect great stamps. You know, we get to, um, watch things and name things and describe things. And I actually think that that's a really important part of biology. Um, I think descriptive biology is very important, but it's even better when you can get to mechanism and get to understanding how these astonishing phenomena can possibly occur. So that's that's where I would kind of put mechanism. It's, it's an attempt to, to, to Do that part of biology that Rutherford wouldn't have disdained. Do you think that how we approach the problems that we study depends on the assumptions that we make about how it works, sort of from the outset? I do. Um, and also it depends on, um, our, our training and what we're comfortable with thinking about and, and and so on. Um, not everything in mechanistic biology starts from an idea, a guess as to a mechanism, a hypothesis for what the mechanism might be. And, um, you know, attempts to either corroborate the hypothesis or disprove the hypothesis. Um, but, um, but a lot of it does, um, it's, um, we're this is an era of large scale data sets. And um, and um omics kind of research still going from omics data set to an understanding of how things work often, um, really depends on having some sort of a guess as to how to make your way into the morass of information. Rob Phillips, who's a physicist turned biologist at Caltech, sometimes, um, tells the story. Um, he would challenge anyone to take, um, this data set, that, um, tissue bra the astronomer came up with for the positions of the of the planets in the night sky. And, um, come up with anything like the conclusion that the planets are circling the sun and they're taking on elliptical orbits without bringing that into the data set to start with. You know, just to look at the positions and without having that as a, as a guess, come up with the explanation for the for the patterns of the wandering of planets in the nighttime sky. So, so even from that, um, you know, centuries old, um, um, example of of big data. It, it was helpful that that, um, that Kepler tried out a few ideas and came up with, um, an explanation that would, um, would account for the data, a mechanism. And then it was even better that that, um, Newton came up with, um, the law of universal gravitation, which explains why Kepler's laws work. Um, and, um, in, in simpler terms, um, so, you know, going from, um, from big data to hypothesis testing and then to a to a unifying theory can be a great way of making progress toward understanding mechanism. And and, you know, you can get away with just describing things in biology and still make an important, significant contribution. But to the extent that we can understand mechanism and then understand kind of unifying principles that that sit underneath the mechanism, that's I think, where the, the greatest fun is. You seem to have a pretty strong historical sense of where some major scientific ideas came from and how they evolved. Is that something that you started with, or is that something that you've kind of evolved over time? I took a history of science class when I was an undergraduate, and I thought that was really interesting. Um, then down through the years, a lot of the smartest people I run into like, like Rob Phillips, who I mentioned, and Stan Schwartzman, who's at Princeton. Um, who made this point about the relative roles of bra versus Kepler versus, um, Newton? Um, to me, for the first time, it's it's sort of, um, uh, taught me along the way, when you think about new scientific problems that you're working on, do you or have you in the past started with a historical kind of approach like that, like starting with maybe simpler forms of the idea that somebody was already working on to see how it evolves? Not not so much. Um, I like telling things in a sort of historical perspective. Um, I find it interesting how things within my career evolved and developed how, you know, sort of for a while people were taking this approach. They were doing, uh, genetics. And then for a while there was a wave of doing this approach, biochemical purification or and now this approach, um, RNA seq experiments to um, to define and characterize things that come and go. And personally, I don't find that students are always all that interested in the historical details of how, um, ideas have come to the fore and retreated. Um, maybe it's just because I'm not that good at teaching that sort of thing, but, um, you know, maybe there's some better way of presenting it. So your your background is in physics, chemistry and mathematics. Yeah. And then, uh, and so you had started at Williams College and out of curiosity, is, is that three separate degrees or is that a single degree? It was a single degree with three majors. Um, I was I was one of those kids who, like, um, knew he wanted to be a scientist from a from a pretty early age. Um, and I liked all sorts of different kinds of science. So it took me a long time to kind of settle on, uh, a field. That's why I ended up with three majors. It wasn't like I'm going to be a triple major kind of an intentional thing. It was just I couldn't decide which which discipline to cast off. Those three subjects are taught in, you know, in, in our biology program at the university that I'm at. But they're not taught in a whole lot of depth. And I it's part of the reason I asked. I was curious if if having more depth early on in those subjects influenced your follow on decision to pursue chemistry and then ultimately kind of moving in a, in a direction toward biology? I think for me, I, I like the stuff. I liked all the things that I was learning in, in physics and chemistry. I felt like the the rigor of math was good too. I thought it was like interesting to do proofs and fun, fun challenges. Um, so I like the stuff. Um, I eventually went when I was an undergraduate, did research one summer, um, on the hyperfine interaction in, uh, as probed by a hydrogen maser. Um, the hyperfine interaction in, in atomic hydrogen, um, and, um, I still can't really explain what was interesting about the work that I was doing that summer. Um, but I had a roommate for part of the summer who was doing biology research, and, and he could tell me, like, in just one sentence or two sentences, what he was doing. And it was obvious that that it was interesting. Um, so I came away from that summer with the the idea that, um, that biologists got to, um, got to grapple with really interesting questions, really interesting issues and, um, and I think that was true. And I think it's still true. I think that you you don't have to get that deep into biology in order to get the, to the frontiers, um, to, to unanswered questions whose importance is really, you know, quite obvious. Um, and, um, you know, we got the tools available now to take on all sorts of different challenges. It's a, it's in, in from a scientific perspective. It's a, it's a great time to be a biologist. And then the stuff that you learn as a physics undergraduate or a math undergraduate, um, maybe you can use to inform the ways you approach those biological problems. That was that was my hope anyway. So you then did a PhD in chemistry at Stanford and followed by an MD. Was that an MD PhD degree program or were those two separate degrees? I started as a medical student and worked on research right from the the start of medical school. Um, originally with a quantum chemist named Gilda Lowe, who was doing, um, like big quantum chemical calculations on, um, molecules of biological interest. And after a couple of years with, with Gilda, a couple of good years with with Gilda, I set out to try to, um, come up with some, um, lab to collaborate with so that I could, um, um, uh, test some of the things that we were predicting with these, um, calculations. and, um, eventually sort of the best, um, situation I could come up with was, um, the finding a person who, who wanted to have me in the, in their lab and, and, um, whose work was quantitative enough that you might imagine, um, these sort of physics kind of things could pertain was, um, was Ray Heustis in the chemistry department at Stanford, um, who had a background in, um, NMR and EPR spectroscopy and their applications to biological problems. Um, so the original thought was that, um, that I do my PhD with Ray and maybe at some point along the way, something would, um, would arise where I could go back and do some quantum mechanical stuff. Um, the latter never really happened, but I learned a lot of stuff, Um, from from Ray in the chemistry department. So I started out as an MD student. I joined the chemistry PhD program in my third year, I guess finished the the chemistry PhD. Then after six years? No, after. Yeah, after six years. Finally after ten years, then, um, um, finished my MD. So I was a graduate student and an MD student for a long time. Did you consider pursuing a clinical career rather than a research career at any point? I did, um, when after I finished my PhD and went back on the wards, I was, um, initially sort of overwhelmed because I had forgotten a lot of the MD stuff that I had learned, um, but eventually got competent and capable, and I liked it. I liked taking care of patients. I like the sort of hospital based medicine that medical students get exposed to in their training. I, I delivered babies. I mean, that's a wonderful thing to get to do. I helped in some operations, did a couple of operations as as the, you know, person who did all the cutting and sewing. And I really did like that a lot. Um, I guess my initial intention was to practice medicine part time and do research part time. Um, and um, decided that either thing separately was was really hard. And, um, the idea of trying to do both of them at once seemed almost impossible to me. There are people who do it and do it well, but I didn't think I was one of them. So I ultimately opted for for one thing to do. And that was the the research. Your dissertation research was, uh, related to shape, control or morphological control in erythrocytes. Can you just give a little bit of physiological and maybe biochemical background into that problem and how you were approaching it? Um, yeah. Ray was a biophysicist. She was working on, um, on red cells, I think in part because they're the simplest cells in a, in a human. They don't have a nucleus, they don't have any organelles, they don't synthesize proteins or they they basically just carry hemoglobin around, and they flow well through the bloodstream. And part of their ability to flow well comes from the shape that they are. And they they also exchange oxygen well with the in the lung to pick up oxygen and then in the tissues to dump off the oxygen. And that comes in part from the shape of the cell. So, um, The hope was that, um, that given how simple it was, um, maybe we could understand how the cell comes to be shaped the way it is shaped. And, um, how in different pathological circumstances, the shape of the cell can, um, can change. It's an interesting problem that's still not really completely solved. It's one of these things where, um, you know, you've got physical forces and things like membrane tension and stiffness and so on, acting at one level. And then you've got the tools that biochemists have to interrogate systems which are things like O, purifying proteins or knocking out genes or knocking down the expression of genes and so on, where the the physics and the biochemistry don't always mesh. I actually think that's still an interesting, um, kind of interface, the integration of physical forces and biochemical activities or genetic activities to bring about the events of life. Can you kind of summarize what the kind of model of shape control looks like? It was motivated by the fact that that the cell is normally this biconcave disc and it's smooth. But when the cell runs out of energy, it turns into a bumpy cell. So how is it that running out of cellular energy you grow these bumps on your on your surface, these spikes, they're called echinocytes because like echinoderms, they're they're spiny. Um, and the prevailing hypothesis was something that, um. Mike Sheetz, um, great cell biologist who died fairly recently, um, came up with he, um, thought that, um, anytime The cell membrane. The cell membrane actually operates almost as two independent monolayers, the constitute a bilayer um, where the head groups of the phospholipids that are part of the membrane face the outside of the cell, and then another monolayer of phospholipids have their head groups face the inside of the cell. And his idea was that, um, if the outside monolayer got to be too big relative to the inside monolayer, the whole cell would change shape to become a new geometry that could accommodate this imbalance. And if the inside got too big compared to the outside, um, then you'd get almost the opposite kind of shape change. Instead of little spikes sticking out of the cell, you'd have invaginations in pouching going into the cell. Um, and so guided by that hypothesis, we set out to see if we could figure out what made the the bilayer balance change in shape when a cell ran out of ATP. And our evidence was that it had to do with, um, a couple of phospholipids that can exist in phosphorylated forms or dephosphorylated forms. One was phosphatidic acid, which can be dephosphorylated to diacylglycerol, and one was phosphatidylinositol bisphosphate, which can go to monophosphate or non phosphate. Um, and our idea was that that when they got dephosphorylated in the inner leaflet of the plasma membrane, that would make the inner leaflet shrink and the cell would have to change shape. It's probably true. It's it's results that have stood up, I think, pretty well down through the year. But still, you know, we don't know everything about that. Uh, after after you'd worked on on those problems for a while, you ended up working on Xenopus oocytes among some other systems. Or maybe a better way to ask that question is following your graduate work. What happened next? Okay, so some people will finish their PhD work and go on and work on something very closely related as a postdoc and then get a permanent position after being a postdoc. But a lot of people take that opportunity to change direction because they've, I don't know, gotten interested in something else or because maybe having a two things that they're, uh, a world's expert on will, um, be better than just having one thing that they're a world's expert on. Um, I thought at the end of my PhD time that the most interesting stuff happening in biology had to do with oncogenes and proteins that regulated other proteins, as opposed to proteins that directly did things like, um, Glycolytic enzymes or motor proteins or proteins that were part of the the control networks that orchestrated the events of cells. And so that's what I did as a as a postdoctoral fellow, I was in Steve Martin's lab at UC Berkeley and working on protein kinases that phosphorylate other proteins on tyrosine residues. And that was involved in some contexts anyway, in both normal cell growth control and in the derangement of cell growth control that happens in cancer. And late in my time in Steve's lab, um, a postdoc from UC Berkeley came over to, um, from UCSF, uh, came over to Berkeley to give a seminar, um, a guy named Andrew Murray, who's now on the faculty at Harvard. And, um, he had done experiments in these, um, frog egg extracts, cell free, undiluted cytoplasm that you, um, that you got by just basically taking frog eggs and putting them in a centrifuge tube, getting rid of all the extracellular water or buffer that might be around, and then squeezing them so hard that their cytoplasm would, would squeeze out of the cells without getting diluted. And he discovered that this cytoplasm would like, do very life like things, including doing the cell division cycle, which most people were studying, um, in mammalian cells that maybe divide once a day and divide in a very poorly synchronized fashion and kind of unreliable fashion. And his cell cycle extracts would do a cell cycle You know, once every half hour or an hour and you could do things to the extract like, um, deplete components or add components of interest to the extract and assess the consequences in ways that you could never do in an intact cell. Um, and he discovered really important things about the cell cycle. He discovered the, um, the importance of the cyclin protein for driving the cell cycle, and that you could drive a frog cell cycle, um, where you don't synthesize any proteins at all except one cyclin protein, a B-type cyclin protein, and get every aspect of the cell cycle to to happen. And I just thought that was unbelievably, um, wonderful, inspiring. And I set out to see if I could learn how to do experiments on frog eggs. Um, and I had one idea for an experiment that would be maybe a worthwhile experiment to do. It had to do with Map kinase activation during frog oocyte maturation, oocytes or the eggs before they've been laid by the frog. Eggs are in a slightly different part of the cell cycle, but, um, they're basically almost the same cell as the as the oocyte. But after it's sitting in the pond, um, and um, Berkeley was one of those places where even, like a postdoc like me could make an appointment to chat with, like the Xenopus expert who was, um, a guy named John Gearhart and tell him my ideas. And he thought it seemed like a reasonable idea for experiments and and offered to let me collaborate with his technician, who was a guy named Mike Wu, who was better at doing experiments with frog eggs and frog oocytes and probably anyone in the world. So that got me into the field. The kind of the combination of seeing this inspiring seminar from Andrew Murray and then the generosity of John Gearhart to, um, you know, uh, train me to do stuff, um, and getting trained by this person, Mike Wu, who was just so good at doing the experiments. Um, so that's how I got into the, uh, the experimental system of frog oocytes, um, which is a good system for addressing questions that pertain to anything the oocytes happen to do. And one thing that the oocytes do is they switch from being immature oocytes to being mature oocytes in a way that depends on the activation of Map kinase and its biology. That's sort of all or none in character. But if you think about it, um, and oocytes got, um, tens of billions of Map kinase molecules in it. So you would think even if you were to, um, activate the Map kinase, you could go from having one molecule, the two molecules up to tens of billions of Map kinase molecules. You would think that you'd have all sorts of different graded outcomes depending on how you adjusted the slider for how much activation of Map kinase you wanted out of it, but the biology was really all or none in character. So it seemed like this might be an easy to examine example of how nature takes graded control elements signal transducers, and arranges them in some way as to as to make the biological outcome be either this or that. And you're only in between when you're on your way from one state to another state, just like with a wall switch, um, turning the lights on. You know, if you push the wall switch up a little tiny bit, the lights won't turn on. And if you let your pressure off, you'll go back to off. But once you push it past halfway, the lights turn on. And they stay on forever with. Even if you're not continuing to push so that the the switch like behavior there is kind of like a, a dynamical motif, like a, a common pattern that you might see and how a system responds to an input. Um, and you wrote a really wonderful book called Systems Biology of Cell Signaling. Recurring themes and in quantitative models and the the tact that the book kind of takes is to compartmentalize these motifs, um, among different forms or common forms of cell signaling processes that we see conserved across a lot of different, um, um, systems and processes. I'm curious to, to know, I mean, you just kind of mentioned that that the intuition and these very complex systems is that the behavior shouldn't be so deterministic, but it kind of is, or at least it is in many cases. So I'm curious to know, uh, why do you think that that approach works so well? That's a good question. Um. Why why does this approach of kind of boiling things down to a few kinds of behaviors, um, that do seem to recur over and over again in the regulation of biological systems. Um, you know, you'll you can have, um, responses that are graded responses. You see them over and over again, but then sometimes you see responses that are sharper than graded responses. And, um, sometimes you see responses that have a memory built into them. And sometimes when you hit a cell with a stimulus, instead of getting just a response that stays on for a while, maybe you'll get a response that's a an explosion and then another explosion, and sometimes they'll be periodic. Um, so those are things that, um, that, um, once, um, once we as a field have studied enough, um, different individual signaling processes you can see crop up again and again and again And, um. The question then arises, is there some commonality to the control, um, systems that make these things happen? Um, and, um, the answer is, yeah, probably there is. I mean, like if you get adaptation, you don't necessarily know, like and by adaptation, I mean, you put a stimulus on, you get a response, and then the response fades away even though the stimulus isn't removed. And that happens over and over again in, in biology. Um, if you see that happening, you don't know for sure what the circuit is that underpins that, but you can bet that it probably either includes negative feedback, where the response causes the upstream stuff that made the response happen to no longer be sensitive to the stimulus. Um, or. Um. What? In control theory is called, um, incoherent feedforward regulation, where what you have is the stimulus comes in and it makes two different responses happen. One response is positive, and the second response, which has a time delay built into it, is negative. And so what you have at the bottom of it all is a response that sees the positive things first and then the negative things, and comes back to to where it started from as a result of that, um, the slower negative regulation. So you can be pretty sure that it's going to be one or the other of those. And, and why? Um, and my bet is that, um, part of it is because, um, these circuits that work in cells have to have evolved and as a result, they can't be ridiculously complicated. They can be complicated, but, um. But you've you've got to get sort of some function every step of the way during evolution. Or at least maybe you don't get some positive function, but you get at least you don't break things in the in the course of building your ultimate complicated machine. So I think that evolution, you know, if it, it doesn't necessarily, um, pitch things toward the simplest possible solution, but it, but it um, biases things toward simpler rather than more complicated solutions. And so that might be why, you know, so often we can boil down adaptation processes to these, um, these two things. Um, sometimes it's not exactly those two things. Sometimes it's, it's a different kind of thing. But So that's, um, that's that's when, you know, it's amazing and, uh, that that these approaches, um, work. It's also amazing that they work because as you, as you mentioned, um, the the numbers of agents involved in biological regulatory systems aren't infinitely large. And sometimes, like, if you're talking about, like, the expression of genes, you're talking about, you know, one locus or, or two loci on a genome that are being turned on and turned off. And so you could expect responses to be very noisy. You could expect responses to be stochastic. And, um, the, the textbook that I, um, that I wrote, um, starts with the convenient lie that you can describe things, um, in biological regulation as though they were deterministic and are described by ordinary differential equations, which are valid for large numbers of actors. And differential equations are built on infinitesimal time elements and continuous stuff. And um, uh, it's it it doesn't have to be that way. And there are certainly aspects of biology that are not that way, but still, um, you can you can get a long way. Oftentimes I think in biology you can get a long way with, um, starting with a fib. Um, that is a convenient fib that that allows you to, to at least go so far. Um, and then the third thing that is sort of amazing, um, that these types of approaches assume that, that things are spatially heterogeneous, spatially homogeneous, um, And well, mixed. Um, and that's not true in biological systems. Anybody who's ever looked at a cell, um, can appreciate that the, the, um, there's structure at every scale you look at, there's, there's, um, you know, beautiful structure at the level of, of, um, an individual protein molecule or a supramolecular complex of protein molecules, like a, like a microtubule or the level of an organelle or a level of a cell or the level of a tissue. So there's, there's, there's lots of spatial organization built into it. So, um, and yet, um, you can get somewhere by starting with the assumption of reactions being approximated by the odd behavior, the well-mixed system behavior. And that maybe operates on some tiny distance scale within the cell, and then you can work from there to understanding how, um, when you introduce, um, space into the picture and the movements of molecules from one region of the cell to another region of a cell, um, what additional behaviors you can get on top of the, the, um, the kind of deterministic Ode sorts of behaviors that got a little technical, but but that's that I think takes us up to what's really one of the, um, one of the frontiers of understanding biology, how chemical reactions splay out in space. And there's a huge, rich field in physics and mathematical physics, um, um, on, um, partial differential equations and reaction diffusion systems that, that is still yielding, um, active theoretical, um, new insights and stuff and is is there for us biologists to to learn from. I saw a talk that you gave us. It's recorded online, and you started out with a interesting thought experiment where you posed the question, if you put a computing machine like a digital computer into a blender, what would you get out on the other side? Would it still have computing capability? And if you put a biological system like frogs eggs into a blender, would you still get the the biological self-organizing qualities on the other end? Can you talk about that thought experiment a little bit? Yeah, I used to do this thought experiment for an undergraduate class that I, um, that I taught, and the point of giving them the thought experiment was to show them that a daughter cell Doesn't just inherit the genetic code, the genetic information from its mother cell. It also inherits, um, spatial organization, um, from the from the mother cell. So I would, I would, um, you know, say to them, yeah, if you, if you took frog eggs or bacteria or something and put it in a waring blender and then blended it up, let it sit around for a while would have come back to life. And that all say, and why not? And it's the same ingredients as the ingredients in a in a cell. And that work out that, you know, really the, the way, um, it would take too long to randomly reform the structures that allow life to, to happen without having the pre-existing structure that came from the mother. You know, all cells come from cells and they're, um, their, um, inheriting cell like organization as well as, um, genetic blueprint. So, yeah, I used to tell him that, and then, um, uh, really smart postdoc in my lab did an experiment to basically test that. And, um, we were gobsmacked to find out that if you take frog eggs and make an extract from them and just let them sit around for about twenty minutes, they turn themselves from being pretty random, not totally homogeneous, but homogenized, highly disrupted garbage into things that look like cells. And that put their organelles in the right places and that don't have quite the boundaries between repetitive units the way cells do, but in appearance look like a almost a sheet of cells organized by cytoskeletal elements. And if you allow those reorganized cells to do the cell cycle, which they will miraculously do if you just, um, don't kill them, um, then, um, these little cell like compartments will divide and divide and divide. They'll undergo successive rounds of DNA replication and reductive division. So that to us was, um, first of all, something that was like a gobsmacking curiosity. Um, but second, a lesson that, you know, one of the important things about life that enables life to be alive is its ability to build itself and, um, also to repair itself. Um, but you have lots of examples in, on different scales in biology of things that you can break apart and they will just put themselves back together again. And that must have been selected for by evolution. I don't think that's something that you'll get by, you know, the random synthesis of, of um, of polymers. Um, that's something that, that, you know, it must be that, um, that the ability of proteins to form dimers or trimers and the, the abilities of, of these um, complexes to form higher order structures and so on is something that has been selected for by evolution and allows life to, to, to work. That was one of my favorite things. The lab has come upon in a long time. And that was discovered by a, um, a postdoc named Shawn Ray Chung, who's now, um, got his own lab at USC. He's an assistant professor at USC. And, um, you know, occasionally you just, um, you know, run into something that you hadn't anticipated. And that was one of those things. You noted that you had to watch a video of cell lysates kind of repeatedly to notice that the nuclei were aggregating, I think. Is that kind of and if you hadn't made that connection, then it didn't seem like the the you would have arrived at that conclusion. Yeah. It was the experiment was aimed at learning something totally different. It was we were testing a hypothesis that, um, that apoptosis, this, um, this kind of programmed cell death, a very well organized mode of cell death, um, would spread through the cytoplasm by a self-regenerating trigger wave. I won't go into too much detail about that, but a trigger. Trigger waves show up in physics and biology. Action potentials or trigger waves. The spread of a fire through a forest is a trigger wave. The spread of a joke on the internet is a is a trigger wave. Um, and um Shunrei was testing a hypothesis that that, um, apoptosis would spread as a trigger wave. And the reporter he was using in the cytoplasm, um, was um, uh, cell nuclei, as the cell nuclei would pop when they underwent apoptosis. And he could just watch this happening and see if the spread of apoptosis down a thin tube of cytoplasm would happen in the way trigger wave propagation happens with constant speed and never stopping, no matter how far you went. And the answer was yes. And we looked at the movie that sort of first showed us that the answer was yes. Again, I'd say I probably looked at that movie fifty times without noticing, because it wasn't what the experiment was aimed at, like discovering without noticing something that he had noticed, which was that every time he did the experiment, the the nuclei that were reporters would start out everywhere in the cytoplasm. And by the end of the experiment, there'd be organized into little clusters. They were self-organizing into the middle of what were these cell like compartments. But we weren't looking for cell like compartments. We weren't thinking about the self-organization of the cytoplasm. And so, you know, I didn't see them until, um, Shawn Ray pointed it out to me. And then it was like, of course. And Shawn Ray said, you know, do you do you think this would be an interesting phenomenon to investigate further? And I just sort of, of course. And he went on, did fantastic work. Maybe we could swing back to that circuit analogy that you mentioned earlier, that I think I was calling these kind of dynamical patterns motifs, but in the book and kind of elsewhere in some of the other papers that you've written, you've also kind of referred to those as circuits. Can you talk a little bit about that analogy, and how you use that analogy to help you understand the differences between these different types of responses that you see? Yeah, you can call them motifs. Um, these sort of recurring themes, like a leitmotif in a, in a piece of music, um, early on uses the term motif a lot. He's a he's a really great systems biologist, physicist turned biologist, and it's a good term. Um, you can also use the term circuits. um. And I think that gives you a little bit more of the idea that it's, it's a it's an entity. It's a level of organization that does something and that does something that maybe the individual components that make up the circuit would not do. So it gets at the issue of function emerging at a level of complexity, um, in the um, in, in the whole scheme of things, not the level of complexity of like a whole cell, but the level of complexity of, say, this protein activates this protein, which feeds back to inactivate this protein, handfuls of, of um, of proteins. And, um, it turns out that in the same way, um. Um, the you might see, um, you know, some protein or some gene being important in this biological process or this biological process or this biological process. Sometimes you'll see a particular motif or a particular circuit of proteins that are involved in a whole bunch of different biological processes, so that it seems like, you know, nature has, um, ended up, um, uh, maintaining that organization that's bigger than the organization of an individual protein. And so then that gets at questions of, okay, what could what can that organization do that, um, might make it worth keeping around. And if you're unlucky, the answer will be, oh, I don't know. It could do a million things. But sometimes the answer is, well, it could do amplification, or it could do adaptation and then you can do experiments. Um, the, the test, um, whether that, um, kind of provocative answer might be correct or not. You mentioned earlier that, um, it's not only the genes and sequences that are inherited, but also a pre-existing structure. Do you think of the circuits as being kind of manifest in, in both the genes and the pre-existing structure? Yeah, that's a good question. Um, and I don't really have a, have a favorite opinion on that. As to the answer to that question, how much, um, uh, life gets out of the pre-existing structure and how much of life, how much life can put together from disorder. Um, so I would bet that it that, you know, Sometimes you can put things together, and sometimes it needs the pre-existing structure in order to, to, to make the circuit work. So nowadays, I think we're all pretty familiar with digital computers, and digital computers solve problems essentially by making symbolic versions of the problem, and then doing some operations on it and then converting it back into the original language of the problem. Um, but, uh, before digital computers were were everywhere, uh, scientists were using electrical circuits to create analogues of physical systems that they wanted to study. And I'm curious to know if the circuit analogy maybe works so well here because it seems, well, I guess one, because these systems are being analyzed with ordinary differential equations. And that's what those original circuit analogies were good for. I guess my question is, are is the circuit analogy maybe related to that, that same sort of relationship that there's sort of a pre-existing precedent for using electrical circuits as analogs for Odes and systems of Odes and other systems? Yeah, that that may well be true. Um, for example, when when the first electrical circuits were being built, um, you can go back into the early twentieth century when people were like building vacuum tube oscillators and stuff like that. There's a there's a classical simple circuit, um, called the van der Pol oscillator that will, um, produce oscillations. Boom boom, boom boom. But the oscillations aren't like sine wave oscillations. They're more like what I just said boom, boom, boom, boom. And van der Pol, in analyzing this circuit that he had built out of electronic tubes right away, thought, you know, it may be that something similar to this is happening to make the heart go boom, boom, boom, boom. And it actually is true. He's he's right that, um, what we know now about how the heart is driven, it's driven by an electrical oscillator that starts in one kind of location, the sinoatrial node, and spreads throughout the heart and makes the whole heart muscle contract. Um, but the type of pacemaker oscillator circuit that starts up there at the sinoatrial node is what physicists call a relaxation oscillator. And that's what the van der Pol oscillator is. It's a relaxation oscillator. And, um, the important thing there is every relaxation, every relaxation oscillator is different from every other one in its details, but every relaxation oscillator is the same as every other relaxation oscillator in terms of some fundamental aspects of how how it works. It has this kind of nothing happens, nothing happens explosion and nothing happens. This kind of fundamental rhythm to it, and it gets built out of positive feedback and negative feedback combined in circuits. So learning about one relaxation oscillator teaches you something about all other relaxation oscillators. And it turns out that oscillations are common in biology. Um, we have not only the heartbeats like once a second, but we have circadian rhythms once once a day, and we have seasonal rhythms once a year. And, um, human women have, um, have monthly menstrual oscillations. Um, flowering bamboo can flower once every one hundred and thirty years. And that's a that's a long scale oscillation. Cicadas have an oscillation where they come out of the ground every thirteen years or every seventeen years. And, um, look for, uh, somebody to get married to. It's not always the case, but it but it seems like it's often the case that, um, biological oscillators are relaxation oscillators. They're all built kind of the way this circuit that Vanderpol studied in the nineteen twenties works. And, um, they're built in the same way that people who study cardiac physiology understand, um, and, um, they're pretty different from the oscillators that most of us learn about in physics. Like, they're pretty different from a pendulum. They're pretty different from a, um, um, um, a harmonic oscillator where the the path that the oscillations take depends on where you start out. You know, if you if you start out with the spring pulled a lot, you'll get big amplitude oscillations and biological oscillations. Generally, no matter where you start out, you'll end up in the same groove over and over again. Um, and, um, relaxation oscillators are a kind of oscillator that will do that sort of function. Um, so that's an example of how, um, yeah. Right. From the first circuits that electronics people were learning about. Um, uh, biologists have gotten some inspiration into how control works in biological systems. Another I think kind of interesting aspect about circuit analysis is, um, at least in electrical circuit analysis Often what's being measured is the is actually the energy potential in different portions of the circuit. And do you think using circuit analogies to describe the dynamics in biological systems could also be easily, uh, or could could easily incorporate energy potentials into the analysis? That's an interesting question. Um, thermodynamics, people might say yes, that you can start with chemical potentials, um, as a way of understanding, especially equilibrium systems. Um, I think from my experience actually in teaching biologists, I think the answer is no. Um, I think that unless you go into a biology with a, with a really strong intuition based on having previously studied electrical circuits, that that whole framework is not too useful. Um, I, um, I actually think that it's, um, easier to think in terms of, um, chemical velocities, which are the things that you deal with in ordinary differential equation models rather than the equivalent of forces or potentials, um, which are sort of accelerations instead of velocities. Um, is is a more straightforward path to understanding things. But that's that might just be me. Often in, uh, I think a kind of common, uh, theme. And some of the research papers that I see is that there tends to be a kind of, uh, very sophisticated investment in the experimental system that's being used. So, you know, there are there are very, very complicated ways of manipulating genes and ways of measuring a lot of different transcripts at once and things. And, and then a lot of, uh, experimenters will, will do the experiments, they'll do the hypothesis tests sort of separately. And then they take the mean differences to be, um, the effects that they're looking for. And then they kind of draw a diagram at the end of the paper where they show how the effects and the experiments translate toward interactions between different things. So if this was a cell signaling paper, for example, they might say, you know, the the signaling input some hormone stimulates calcium influx in the cell which they measured with a fluorescent dye or something like that. But often they kind of stop short of anything else. And and it ends with a qualitative diagram. And I know in your research, uh, you incorporate a lot more into the follow on analysis of what those models entail and how they might behave in the future if we were to observe them again. So I'm kind of curious to know how how you think about those that approach to research that kind of stops short of any additional quantitative analysis and what biologists can do to to maybe improve things a bit. Sometimes you can, um, you can get away with, um, just a sort of textbook style cartoon of a complicated process where this turns on that, that turns on that, that turns on that, that turns on this and off that. And, um, you can get a pretty complete understanding of the process without making anything any more quantitative than than that. It's really, um, you know, it's equivalent to making a Boolean model of a complicated process. I'm not sure why that is. It may be that those types of models are most successful, where the circuit that you're dealing with has something built into it to enforce a digital character to the response so that you can't arrest an intermediate states. Um, um, but um, sometimes understanding the quantitative nature of, of um, of things really does, um, help you, help you to think, um, better about a problem, um, how to understand things more deeply. Um, and you can do it. Um, and it's not that hard to do it. Um, and I actually always encourage, um, people to at least learn how to do it once at some, you know, learn how to do, um, some modeling stuff. And in fact, we require of all the chemistry or chemical and systems biology PhD students who come through our department to, to do a couple of weeks of, of um, of modeling, um, so that they'll have this, um, arrow in their quiver, um, uh, so that if, if something down the road happens where they, that, that kind of cartoon model is just not cutting it, it's, it's not, um, explaining something that they see over and over again. Um, they'll have the ability to go in and, and look at things a little more deeply. Uh, the, the undergrads at the institution that I'm at, they take, uh, cognates as part of the program. So they take general chemistry, organic chemistry, two semesters of introductory physics, two semesters of calculus. And I'm always a bit surprised that they're not able to use any of those tools in a functional way by the time they make it to me. And I think it's interesting that they're learning these things, but it doesn't seem like, um, at least not in our, our program. In fact, that's that's one of the kind of goals that I have for a systems biology class that I teach to more advanced undergraduates is to try to sew those things together and show them how to actually use it. Um, for example, some of them, they take this two semester calculus course and they never learn how to numerically integrate differential equations. So they've seen a few examples where they can come up with analytical solutions. Um, and they work through those homework problems. And they answered those questions on the test. And then they never, uh, they never really learn how to, uh, just generally set up ordinary differential equations and find some sort of a solution. Yeah, that's a that's a problem. My my hope is, um, when I teach somebody something, you know, maybe they'll forget. Or when I motivate them to, to learn about something on their own. Um, um, they'll learn it and maybe they'll forget it, but they'll remember that they once understood it and they once knew how to do it. And so the activation energy for going back and relearning it if, if it's important to them, um, at some point in the future isn't, isn't so high. And that's been true for me personally, uh, like a lot of stages in my career. Um, for example, when I was an undergraduate, I did an honors project on a nonlinear dynamical system that would go from, um, kind of deterministic orbits, uh, like orbits in a little triangular ashtray, um, to chaotic orbits. Um, and, um, doing that work involved writing down differential equations and solving them numerically and, um, plotting trajectories and all sorts of stuff like that. Um, I did that as a senior in college, and then I didn't do anything at all like that when I was a, an MD PhD student for ten years. And I didn't do anything at all like that for four years as a postdoc. And I didn't do anything at all like that for my first four years running my own lab as an assistant professor. Um, but it got to the point where we we, um, you know, new stuff about the Map kinase cascade, the structure of the cascade, the organization of the cascade, and, um, it seemed like, um, running some odes could allow us to test some ideas about amplification and speed of response and steepness of response, things that you could you could think of as being a possible outcome of the organization of the Map kinase cascade into this three tiered, um, succession of regulatory proteins. And the what you have to do to do those simulations is exactly the same thing as I was doing for my undergraduate PhD thesis. So I didn't like, you know, remember exactly how to do it. But I knew that I once could do it and set about trying to do it again as something to supplement what the rest of the lab was busily working on. This was kind of my idea of how I could, um, make a little bit of a contribution that went beyond just managing the graduate students and postdocs in my lab, but actually doing doing something on my own, um, that I could pigeonhole into the little bits of free time that, that I had. Um, and, um, it wasn't that hard to relearn this stuff. Um, and part of it was because there were things like software packages that made it easier to do this time around than it was when I was a, um, an undergraduate. And part of it was there was this theory book by Steven Strogatz that I thought was really great theory book that that, um, taught me some of the stuff about these nonlinear dynamical systems. That wasn't even known when I was an undergraduate. Um, but but a lot of it was just that, you know, I knew that I had been able to do this once so I could do it again. And so I was a lot more willing to, you know, feel like a dummy for a while, knowing that I'd eventually get up to speed. Do you find it difficult to communicate your research at biology conferences or places where you share your work? Due to the, I guess, the kind of quantitative nature of the work that you're doing? Yeah, you'd think that that maybe the answer would be yes, because, um, I mean, you can still do biology perfectly well without ever doing anything quantitative. There are still many qualitative, important qualitative discoveries to be made. Um, and you would think that maybe, um, I'd be received coldly by such Qualitative biologists, but the opposite has been true. Um, and I think that part of that might be, um, when I started out, um, when I went back into the, the, you know, doing modeling and stuff, um, qualitative biology had been so preeminent for so long that people found it refreshing to, to, um, kind of take a little different tack to things. And so, um, you know, there were enough people, there were I'm sure there were people who were sleeping through my talks, too, but there were, um, there there were enough people, um, um, who who'd find it, um, fresh and who would come up and give me encouraging feedback that, um, that it it kept us going. So it's it's not been that hard. And so why hasn't it been that hard? Um. I don't know. I think in all scientific communication, if you can just sort of peel the jargon away and kind of put yourself in the in the mindset of somebody who doesn't think about these things all the time. Um, you can often, um, communicate pretty sophisticated ideas. Um, I mean, I think I can I probably have, in fact, in this podcast already spoken in jargon that's as impenetrable as as you can hope for. But but, you know, just being able to to step back from it and, and, um, think in, in more, um, standard English terms instead of science terms, um, can get you a long way. And this is kind of an out there question. Maybe it might not even be that relevant, but the idea of a mechanism. I think sort of it has a long historical context, and it goes back to systems, um, that once, once there was kind of a general conservation principle for momentum in systems. Um, then you could explain the way that clocks worked, uh, in a reasonable way. And I think, you know, I guess I'm kind of picturing that, that kind of time era of Newton as being the age of clocks. You know, they were they were making very, very sophisticated clocks. And, um, and I, I think, uh, kind of the broad approach to biology that I see, particularly in molecular biology, is sort of this idea that if you can identify all of the parts within the system and dissect the system into its parts, catalog the parts, figure out exactly how those parts are connected to each other. When you reassemble that sort of a system, you should be able to predict its future behavior. And we mentioned earlier that that some of these motifs, these kind of deterministic dynamical systems motifs do actually work, I guess. Is that a fair assumption for biological systems? And are we ignoring the kind of outliers, the the ways of looking at these systems or the biological processes that don't fit the deterministic models that we're kind of imposing? Well, first of all, um, just as, um, physics as mechanics has, um, a few huge kind of overarching theories. I mean, we call them theories like Newton's theory of gravitation, um, or um, um, the Newton's laws of motion, um, that the laws of motion come out of, of principles of equivalence of this frame versus that frame and, um, that, um, that allow us to, um, get into, um, big kind of more complicated, detailed understandings of, of how clocks work and so on. Um, there are a few big principal theories that, um, that sit on top of all the biological processes that, that, um, that we're interested in. Um, the first one and the most powerful one, I think, is the theory of evolution, um, which unifies, in my opinion, more disparate kinds of observations than than even Newton's law of gravitation. Um, uh, does um, so, you know, that's a wonderful thing. Um, the other the next thing is probably, um, some stuff from, um, from physical chemistry, laws of thermodynamics, although, um, living systems, um, are nonequilibrium thermodynamic systems and, um, you know, pump energy into themselves in order to avail themselves of behaviors that you can't get out of, um, thermodynamic systems. Um, uh, but still, it's an important principle. Then maybe these motifs that that we talked about as Uri Alon would, would call them or these, these basic circuits that, um, that, um, are, are present over and over again in um, network diagrams. Um, are another way of um, of of kind of getting, um, um, some overarching unity of disparate observations saying that there's a basic regulatory logic, maybe that that applies to both, um, I don't know, the, the seventeen year, um, reproductive cycle of the cicada versus the, um, the, um, alpha rhythm in brain waves in humans. Um, um, and then for other things, I don't know, there's a, there's, um, I have nothing against, um, coming up with parts lists. I think that it's really important to understand what the parts are that living matter is, is built out of. But there's a lot more to understanding how life works than just, um, Delineating the parts. So it's like it's an important first step, but it doesn't mean that the the game is over for for the rest of us. The kind of final question I'd like to ask is is really not much of a question at all, but I'd just like to turn it over to you if you, uh, is there anything I didn't ask about or any any thoughts that you might have that you'd like to share? Um, I'd say I'd just like to emphasize, um, you know, this is a this is a tough time for American science right now. Um, and yet, um, it's a wonderful time to be a scientist in, in terms of the kinds of questions that are still open, unanswered, and just waiting for, you know, great ideas and hard work in order to, um, um, get some insight into and I mean, that both in terms of things like, um, um, getting answers to questions that will allow us to alleviate suffering, cure disease and do practical good, but also try to understand the how the, the world works. Um, and um, so it's it's a great time to be a scientist, great time to be a biologist in terms of a lot, of a lot of things. Um, and, um, I do think that, um, that people who have a quantitative bent, um, can um, at least sometimes, um, contribute in a distinctive way to our understanding of biological, um, phenomena. And so I encourage any of your listeners who are who who are oriented that way. Go to it, work hard, do something important. And this has been an absolutely wonderful conversation. Thank you so much. I've learned a ton. My pleasure, Cam. Well, listeners, that does it for today's episode. This podcast was sponsored by SSRs Virtual Education Committee, whose mission is to develop virtual programs that aid in education, highlight the lives and careers of society members, and bring updates on the latest scientific advancements in reproductive biology. If you're not a member of SSR, now is the perfect time to join this incredible network of researchers and professionals in shaping the future of reproductive science. For more information, please check out our website at. If you enjoyed this discussion, please like and subscribe wherever you get your podcasts and join us for our next episode. When we begin our three part mini series on embryo development with Ramiro Alberio of the University of Nottingham. Until next time.