In an article published in PNAS in 1998. Marc Kirschner and John Gerhart identified four properties of conserved cellular and developmental processes – versatile proteins, weak regulatory linkage, exploratory mechanisms, and genomic and spatial compartmentation – which, they argue, make these processes flexible and robust, and therefore increase nonlethal variation and evolvability. Twenty-seven years after the paper was published, I spoke with Marc Kirschner about the motivation behind this paper, his collaboration with John Gerhart, and his views today on the ideas around evolvability discussed in the paper.
Date and place of interview: Via Zoom on 11 February 2025; Marc Kirschner in Harvard, Hari Sridhar in Klosterneuburg.
Hari Sridhar: I’d like to start by asking you to tell me a little bit about the motivation to write this paper, in relation to the research that you had done up till this point.
Marc Kirschner: Well, it was long in coming, in a sense. I was a physical chemist-biochemist and so was John Gerhart. When I came to Berkeley to do my PhD, John was on sabbatical, and so I didn’t get a chance to meet him. But when he came back, he read my thesis, and I spent a lot of time talking to him. I was really confused about what I wanted to do, and so I spent a lot of time talking to him a lot about things. The two of us decided to begin working on frog early development. He hadn’t done that, and I hadn’t done that, but John had spent three months in John Gurdon‘s lab just learning the ropes about these things. That was the beginning of a very long and continuing scientific relationship with him. In the year that I spent with John in Berkeley, I started working on frog eggs and the cell cycle a little bit, and then I got the idea to work on mitosis and specifically on microtubules. We used to joke about mitosis: EB Wilson wrote a classic book on cell division and early stages of embryonic development which he published in 1896. We used to say that it read like a recent review article in the field. So little had been learned since 1896 about how the whole process worked. Of course, mitosis is a beautiful process, and you can learn a lot about it just by histology. It didn’t really become biochemical, until much later. I played an important role in workings things out. Anyway, John and I spent a lot of time just talking. I think talking is undervalued today but almost everything that followed sprung from those conversations. Despite no publications from my abbreviated postdoc and a seemingly aimless year of talking and playing I went to Princeton and took a job as an Assistant Professor. After that I would come out to California each summer, and John and I would play around with crazy experiments on Xenopus eggs. Finally, on sabbatical together in Holland, we actually did produce a very important experiment. Of course, it seemed more important at the time than it does today. It was based on much older experiments about what in the egg determines the spatial organization of the embryo. It was widely believed that there was something on the surface of the membrane or the cortex of the egg that carried information about the polarity, whether it was dorsal, ventral, or whatever it was anterior or posterior or left or right. A lot of speculation was in all the textbooks. Beautiful experiments were beginning to appear on the fly from Nüsslein-Volhard, based on pre-localized materials, but how did those materials become localized? John and I demonstrated how that actually happened in the frog egg. The egg starts out to be multi-potential, the sperm entry creates an asymmetry in the egg, and things happened internally, and something called the gray crescent (a depigmented region) appeared in the cortex as the future dorsal side. This much was known. Naturally everybody talked about the gray crescent as having immense importance, but we showed that it really wasn’t important at all. It was an epiphenomenon, just an indication of internal things going on in the egg. And we published a paper in Nature – a full article in Nature – demonstrating how the axis of the egg was generated by more or less random processes, and it could happen anywhere. It was sort of basically triggered by where the sperm penetrated. So, the egg was multi-potential, but just a slight little asymmetry was enough to generate the entire embryonic axis. So, anyway, that was a beginning of a wonderful relationship that John and I had.
Then John and I both were invited to a meeting of leading cell and developmental biologists – maybe about 15 or 20 of us – to write a textbook on developmental biology. These were really the leading people in the field. I don’t know why they picked us, but I guess both of us had independent reputations, and we were now in this area. At the meeting, we went around the table and everybody picked a topic, and then it came to us. The only topic that no one had chosen was evolution and cell biology, or evolution and developmental biology. So, we decided It might be fun to write on it. That would mean both of us would have to read about evolution. The textbook never ever got off the ground, but that was the beginning of John and I thinking seriously about evolution, and we were more and more convinced that embryonic systems like the frog egg are really quite capable of responding in novel ways, and that capacity was not trivial. That was kind of the beginning of this interest in putting cell biology and evolution together. In the end, none of the prestigious people around the table really wanted to write the textbook, but this nudge put me in the frame of mind to think about it.
Quite separately there was a meeting at Princeton to mark the writer and biologist John Bonner’s retirement. I had known John Bonner when I was a junior faculty at Princeton but I had since moved to San Francisco. John Bonner had invited the people he thought he’d like to see to help him celebrate retirement. He invited me to talk about evolution! John Bonner had a quirky but endearing sense of humour. There was no reason that he would have known that, for a brief moment, I had been tasked with writing a chapter on evolution and developmental biology. So, in my mind, I simply imagined that John Bonner had picked people he thought were interesting and asked them to talk about things they knew nothing about! In the end, it turned out I had misjudged the invitation – John probably thought I would talk about microtubules, something I worked on. All of the other speakers, it turned out, were distinguished evolutionary biologists and they talked about their life’s work. When I realized this after the first day, I stayed up late and put thoughts together that John Gerhart and I discussed and tried to link it to my emerging but still pre-conceptual work on microtubules. Nearly everybody at this symposium was thinking about genetic variation, and I was thinking it was as important to understand something about phenotypic variation. Not surprisingly, when I gave my talk, nobody seemed interested at all, which was understandable. But at the dinner afterwards, I met one of the speakers, Mary Jane West-Eberhard. She cornered me and said, Marc, you’ve said really important things, you have to write a book on this subject. But, she continued, “You have to send me every chapter so that I can make sure that you don’t offend people, and that you don’t make mistakes”. That was the start of writing Cells, Embryos and Evolution with John Gerhart. which was about what understanding cell biological, biochemical and developmental mechanisms could say about the nature of phenotypic variation. Mary Jane fulfilled her offer and translated things we said into something an evolutionary biologist would understand. I still have her notes. I think it was 100 – 200 pages of writing, mostly trying to correct our language and avoid inadvertently offending people, and, most importantly, adding thoughts of her own.
John and I would write together, typing next to each other, taking long walks in the Berkeley hills. It was our first book on cell biology and evolution. When we reached the last chapter, we talked about our ideas of evolvability: how might the nature of phenotypic variation facilitate evolution in part by reducing lethality of change. It is a simple set of thoughts: you can change things genetically, but unless the system is able to incorporate it and not die, it’s not going to contribute to evolution. After that, we thought we would write an article more generally about evolvability. It seemed worthwhile since nothing much was written about phenotypic variation. And so, we wrote an article on evolvability, completing the evolutionary thoughts that we failed to do in our book. We sent this article to Nature and it got horrible reviews. One reviewer very patronizingly said that everything we said had already been said by some 19th century Russian embryologists. The other two reviews were just as crazy. These people were clearly interested in evolution and knowledgeable but despite their familiarity with genetics they knew nothing about modern cell biology, physiology, and developmental biology. he reviewers’ complaints made no sense to us, as our examples were about microtubules and transcription and allosteric proteins and all sorts of things that were not known in the 19th century. So, at that point, we said, well, we will write the paper we want to write. Since John and I were both members of the National Academy of Sciences, it meant that it was much easier to get this thing in to the Proceedings of the National Academy of Sciences. This is often seen as unfair but our experience of peer review at Nature was chilling. PNAS would send it out to review it to make sure that we didn’t say anything horrible, but there was more freedom for us. That was how it ended up being published in PNAS in 1998.
It was fair to say that it got some notice and there were some people who really understood that our approach was novel and the insights were important. We then later wrote The Plausibility of Life to make it easier for others with an open mind to understand. There were definitely people in various branches of biology who were deeply affected by our article and that was reassuring. One very famous immunologist – I won’t mention his name – in an interview with The Scientist, when asked what was the most influential article in his life, mentioned this article. He said that he had the habit of reading scientific papers when he took the train from New York to New Haven for work and underlining all the important sentences. With our paper, he realized he had underlined every sentence! Most evolutionary biologists, I don’t think found what we said in the paper disturbing or wrong. Simply, many of them felt that what we wrote doesn’t affect anything directly about their work. Fair enough. Typically, population geneticists just felt that, although what we said was not wrong, it was just beside the point. It’s not something they needed to know. I think that it’s since calmed down. Nobody’s angry with us anymore, but it’s just that it doesn’t affect their work.
HS: Say a little more about the sabbatical in Holland – which year was this, and where in Holland were you based?
MK: It was in 1978. John and I took this sabbatical together. And it was in Utrecht at the Hubrecht lab, which was totally devoted to frog, particularly Xenopus, embryology. There was a famous aging professor there by the name of Peter Nieuwkoop. We were attracted by his classic work on early Xenopus development. And when we got there the lab people had kind of rebelled against him. He was old-fashioned and they didn’t want to do that kind of stuff. They wanted to do modern molecular biology and cell biology. We, in fact were really modern molecular cell biologists but we felt otherwise. Peter Nieuwkoop was an interesting guy and so at lunch we would end up spending time talking to him. In the Hubrecht lab we got very interested in the gray crescent, that distinct pigmentation change formed during the first cleavage on the future dorsal side of the frog egg, was like the holy ghost: arguably very important but also, like the holy ghost, very inscrutable.
HS: Can you talk a little bit about the nature of your collaboration? Were your contributions to this paper complementary in some way?
MK: Yes and no. Our backgrounds were similar. We both were biochemists. We both were, in fact, trying to learn biology of organisms. We both started to use Xenopus to understand the cell cycle and morphogenesis but we could not help being entranced by development and evolution. After less than a year as a postdoc with John in 1972, and several summers working together on both microtubules and Xenopus development, we took the sabbatical together in Holland, where we showed how the symmetry of the egg becomes asymmetric and how the slight asymmetries can define the whole body axis. John followed up on this in his lab when he returned to California. And we published papers together showing how cytoplasmic rearrangements stimulated by sperm entry led to the movement of signalling molecules that generated a dorsal ventral asymmetry that was converted into the dorsal ventral asymmetry in the tadpole and adult. It was impressive on a molecular level that such a system should be so robust. You never got no axis or double axes except under very strong intentional interventions. And so, yeah, I think that those ideas, starting in 1978, got us thinking more about phenotypic variation, and what kind of systems are easily modified and yet come up with functional outcomes. At the time, nobody understood really the underlying molecular and cellular events that were going on. We had great work on genetics, but we didn’t know how the phenotype was generated. And we felt that the ease of generating the phenotype, and the kinds of processes that take place, said a lot about how easy it is to evolve something. If you take a novel and start changing words at random you don’t get another novel – you get gibberish. How is it that, in the case of the phenotype, we get functional outcomes? So, anyway, coming back to your question of what we each brought to our work: we were working with the cell biology of the organism. He stayed more with the developmental circuits as represented by physical chemistry and not informatics. Later I delved more deeply into molecular cell biology. But the focus on biophysical mechanism has continued that way to this day. As far as how we wrote together, at least when it comes to the books, it’s fair to say that we wrote almost everything in each other’s presence – like two little kids playing in parallel. John would come and visit me, I would go and visit him, and we’d be sitting there typing away and talking to each other and sharing manuscripts. It was a very intimate collaboration. You couldn’t say, in the books we wrote, who wrote what. But we are different. I can best see the elements of John’s superiority. John is a deep thinker. He Is very artistic and a consummate experimentalist. He is very careful. He is a terrific reader and can keep many subjects in his head at once. His understanding of physical chemistry and organic chemistry is exquisite. It is hard for me in a matchup to see where I outperform him. Yet our work together was his best and my work with him was my best. I can imagine that there are some things that I brought to the table. I had a habit of coming up with many ideas. He was a good filter – variation and selection all over again. Yet, it is objectively true that, for us, the whole was greater than the sum of the parts. We inspired each other, stimulated each other, and motivated each other. I don’t think we could accomplish as much on our own.
HS: At the time when you were writing this paper, were you already familiar with the literature on evolvability, or was this something that you had to discover as you were writing this paper?
MK: I think it’s fair to say that, as we were thinking about our own ideas, we became more and more familiar with other people’s ideas. So, by the end, we were pretty scholarly about what other people had thought about. Even the failures that people had – which we considered failures – were useful to learn about, to try to understand what they were trying to get at.
HS: After this paper, is this a topic that you continued to stay in touch with? Has this become an area of research for you over the years?
MK: I would say it continued through the writing of The Plausibility of Life. That was written in a much simpler way than the first book. The ideas stimulate me today. The PNAS paper was compact and pretty heavy going. I mean, you had to really stay with it, because the topic isn’t very familiar to many people. I know a number of people read it, and raised questions, but, in general, I have no idea really how much of an impact the paper has had. All I know is that nobody seems to be criticizing it very much. It may be that, to many people, it isn’t important. It has become part of the of the field of evolutionary biology now. But I think most evolutionary biologists – at least those who focus on genetics of evolution – don’t really feel they need to understand this. They’re not opposed to it. It’s just not as important as, for example, the genetic basis of evolution. As a scientist, you often have to ask yourself: is this a paper that you need to know to be a practitioner in your own area? I don’t think our papers and book are felt to be critical for most biologists, and that includes most evolutionary biologists. But I could be wrong, I keep running into people who say that it was important to them. However, there are also people who are standing back a little bit and trying to understand, you know, something fundamental, something broader, about the field of evolution; they would be attracted to it. But if somebody is studying the evolutionary divergence of beetles or something like that, it’s not necessarily helping them. They’re not interested in the cell biology so much. They’re interested in the phenotype and they’re interested in the genotype and they don’t need to know the underlying process. I don’t get the feeling that this is like essential reading, but I do, at the same time, feel it has had some impact. And many people feel it is important, and I’m really satisfied with that. And some people working very distantly, even in theoretical physics, see what we wrote as important.
HS: It’s been cited over 1700, times. I was wondering if you have a sense of in which fields it has attracted the most attention? Is it from other cell and molecular biologists? Is it in evo-devo?
MK: It certainly is in evo-devo. Both John and I had recognition in that field. And also, because our first book is where to go when you want the details about evolution and development that are hard to find. Mostly, it is the place to put evolution in the larger frame of chemistry. So, I suspect, it’s often useful in that context. In the area of population genetics and evolution, I know there are some people who think it’s really important, but my suspicion is that 99% of the people working on that don’t feel that they need to know this information. I think that is sad and I hope I am wrong. The strange thing is that the books and the article appeal to physicists. To some degree, we approach biology from a quantitative and abstract perspective, and that is appealing to some and repulsive to others.
HS: I noticed that, in your online profiles, you don’t describe yourself as an evo-devo biologist. Do you see yourself as an evo-devo biologist?
MK: I do. Probably not so much now in terms of what I’m doing, but I never give up that perspective. After we published these two books and the paper, I remember saying to John, you know, if we’re not to be considered frauds, we should actually be working in this area specifically. So, we did a lot of reading and thinking about it and we became familiar with the work that was done on hemichordates, studied in the late 19th century by William Bateson and TH Morgan. And so, I thought we should actually be doing some evo-devo work on hemichordates, which are interesting because there is current speculation about whether hemichordates say something about how chordates branched off from some stem organism multicellular group. So, to prove to myself that I was a serious student of evolution, John and I went down to Woods Hole and I dug up some hemichordates. For several years we would come out in September and work out methods to study their development and eventually their anatomy and gene expression. So, that was kind of like me paying my dues, that I’m not just sitting out here being a cell biologist and saying, you guys should be thinking about cell biology and evolvability. Look, I’m really here making a contribution in trying to understand phylogeny of an obscure but closely related phylum. John and I still have one more paper coming out on the kinds of processes and cell types in hemichordates. It is an incredibly long paper and will be our last contribution to actually working on evolution and development.
HS: Can you say a little more about digging up some hemichordates at Woods Hole. When did this happen?
MK: It was maybe 10 years after this paper. I was at University of California in San Francisco, John was at Berkeley, and then I moved to Harvard, and so I was closer physically to Woods Hole. I’d spent some time there before. And so, yeah, I mean, I literally went out with a shovel and dug them up! There had been a couple (the Colwins) who had studied them at Woods Hole but their papers did not explain in detail how to get the females to ovulate. That set us back on our first summer’s effort. The first real work on hemichordates was done by Bateson and Morgan. And they had the same motivation – they were interested in something about the divergence of phyla. TH Morgan came out to Woods Hole, and Bateson came out to a place further south, in the Atlantic coast. They dug up the Saccoglossus kowalevskii and studied and described its development and speculated on the origin of the notochord, and its relationship to a structure called the stomochord, which I don’t think has anything to do with the notochord. They didn’t have the molecular tools that we now have, with molecular expression and stuff. So, John and I worked for a number of years, but we stopped working on it. At this point, I’m still very interested in the subjects around evolvability – the conceptual things. It’s really influenced how I think. But I’m not actively working on divergence of organisms, or evolution and development.
HS: I wanted to read a couple of extracts from the paper, just to get a sense of how you think about these things today. One of the things you do in the paper is to identify four properties of cellular and development processes that, you say, circumvent or reduce constraint and increase evolvability. These are flexible versatile proteins, weak linkage, exploratory systems and compartmentation. Today, would you think about this in the same way?
MK: Very much so. I think it is our most important contribution to the overlap of biochemistry and evolution. Yeah, I mean, I think for me, these ideas have great explanatory value about how things work in biology. Take, say, weak linkage. We give a lot of examples, but the most obvious one is allostery, you know, the whole question of how do you change the structure of the molecule. You have a pre-equilibrium between two different states, and all you have to do is bind rather weakly to one of those states, and it drives the equilibrium to that configuration. So that’s, for example, how haemoglobin works. You know that haemoglobin has got four hemes on it, each which binds a molecule of oxygen. But what happens is, if the oxygen tension is low, it just binds with one bond, on average; it doesn’t do very much. But when oxygen tension goes up, there are more oxygens bound, and this pre-equilibrium is shifted to a situation where it binds with more affinity. So, you get this kind of sigmoidal curve, which explains how we so easily take up oxygen in our lungs where there is a lot of oxygen and release it efficiently in our tissues. It’s not that the oxygen brings any information to the molecule. The molecule already exists in these two states. That flexibility can be stabilized by oxygen binding or pH, and all these things have physiological consequences. So, I think, that was definitely an idea, that the functional connections in biology are weak, which is linked to evolvability. Many protein molecules pre-exist in two states, and it doesn’t take a lot of information, or a lot of energy, to shift some huge molecule from one state to the other and you don’t have to achieve a lot of engineering in evolution to get that to happen and to adapt to different conditions.
Exploratory systems: now that is actually something that I am continuing to be interested in, so much so that I’m engaged in writing a book on this. For example, in the case of animal foraging, you have no idea where the food is. So, if you go out and march out for a while, and if you don’t find any food, you go back to where you started, and go up and back, and up and back. But then when you find the food, you need to communicate that. I mean, that’s what happens with ants. Ants are basically blind; they can tell light from dark. If there’s some tasty dead cockroach sitting out there in the field, they don’t know it’s there. So, what do they do? They come out from the nest, and they march out in some random path, and they leave a pheromone trail behind them, like Hansel and Gretel. After a while, if they don’t find anything, they follow the pheromone trail back to the nest. But if they find this tasty cockroach out there, it means a meal for a month for the nest. Then, they merely reinforce the pheromone trail going back, because a single ant can’t carry the cockroach back on his own. A key fact is that these pheromones are volatile, so the trails don’t last very long, but when they’re reinforced, they last longer, and ants tend to follow the existing tracks. So, you don’t have to know much, to get a phenotype, to get a direction, in which to go. And these kind of exploratory process explains so much of developmental biology. I mean, for example, the bones may evolve in their form, but how do the muscles know where to go to support them? How do the blood vessels know where to go to support bones and muscles? How do the nerves know how to connect to muscles, which have connected to bones.? It’s by random exploration and then reinforcement. So, very adaptable! If you’re going to generate a new limb or a modified limb, you don’t have to change the mechanisms for making blood vessels, for instance, or nerves or muscles. These properties are the kinds of things that reduce the difficulty of generating new phenotypes. And this is something that’s come up more in our thinking more recently; it’s really the importance of randomness in biology. You kind of think noise is something you want to avoid, but noisiness and randomness are processes that makes things adaptable.
HS: I wanted to also read the concluding lines of your paper, to ask you how you see this today almost 27 years after you wrote the paper. You say, “Today, we see the survivors of lineages that underwent multiple radiations. These lineages have diversified by maintaining a core of highly conserved processes and modifying others. The core processes have unusual capacities to deconstrain change in other processes and components. This has proven to be a powerful strategy for the variation side of Darwin’s variation and selection principle of evolution.” I wanted to ask you to reflect on these lines, and, more generally, whether your conceptualisation of evolvability has changed in any ways, or remains more-or-less the same?
MK: I think it remains more or less the same. Certainly, Darwin had no idea how things change, especially the underlying anatomy. Can’t blame him for that! In fact, we can only admire him for thinking there must be a way in which things change and avoid lethality. Because if you take a motor or a computer or something, and start just randomly switching wires around, it’s just not going to work, let alone work in a new and powerful way. What has to happen, in evolution, is not just that the final product is better than the initial product, but that all the intermediate steps going from one to the other are viable and functional. And that’s non-trivial. And that’s where, weak linkage and exploratory processes and all these things that we talked about, these are all features of living systems which are very close to what’s essential for life. And this has been understandably ignored by Darwin, who had no idea of what any of this was, and even in recent years, the geneticists, who would explain how a genetic circuit worked, you know, but really, who gave very little attention to how you evolve such circuits. I think that the population people who thought that evolution is just change in gene frequencies, – not that what they were doing wasn’t very important – weren’t really thinking about the underlying physiology, biochemistry, cell biology. How can something be so adaptable and yet so highly adapted and functional in new circumstances? Just think about a tumour, for example. The tumour grows in your body, but it wouldn’t grow unless it had a vascular system, and there’s no evolution for a vascular system that will provide blood to a tumour. That’s not exactly something you would want, anyways; it’s certainly not something that’s an advantage to the organisms. But the very process of these blood vessels just growing towards anoxic conditions and being stabilized by secretion of angiogenic factors by tissues, makes it ,unfortunately, easy for a tumour to grow and be supplied by the vascular system, starting from a single cell to a large mass. But it also means that new anatomical specializations would automatically be supplied by blood. If that were not to be so possible, without new mutations, then most innovations would just “die on the vine.” So, there’s that side of things. I feel that, if I have any kind of sadness about this work and this thinking, it’s that people don’t teach this fascinating aspect of biology, which not only demystifies evolution but also demystifies pathology. To leave out the evolution of the capacity to change from the story of evolution is making evolution, even as Darwin understood it, still implausible.
The other thing I would emphasize today is the importance of random variation. Randomness seems like the antithesis of carefully structured anatomy and physiology. Yet, Darwin had the audacity to imagine that random variation could be at the heart of evolution. The geneticists gave us a more concrete picture of variation on the genetic level but they missed the important role of properties like the growth of neurons and blood vessels that achieve their ultimate structure by a separate non-genetic process of stabilization of random structures.
And I think these things are more general than just biology. I remember, there was a professor from the Harvard Business School, who was very excited about our book, The Plausibility of Life, which was an expansion of the ideas in our first paper. She wondered, how do corporations and companies evolve? And I think the principles of evolvability are quite general. I think people who have read and thought about evolvability in biology will find it worthwhile to be thinking about in other fields. So, I’m very pleased that we found our way to explore the principles of physiological and anatomical variation and interestingly discovered that for their normal developmental function they depended on variation and selection, weak linkage and other very general principles of biological variation.
HS: What do you see as the place of this paper in the literature on the topic today? For what reasons do you think this paper should be read today?
MK: Well, you know, I’m so close to the subject that I would be just trying to argue for why I think it’s important. But I think in talking to other people, for example, that immunologist who said it’s the most important paper he has ever read – and he is not even an evolutionary biologist – I just think that it would be useful for everybody to think of this unique aspect of biology and how it applies to what they study. I’m probably more passionately attached to the writing that John and I did in our two books, this PNAS article on evolvability, and a couple of other things, than anything else I’ve done, because it speaks to a wider subject principle then simply explaining the steps in a particular biological circuit. I think we just scratched the surface in, you know, say microtubules or the cell cycle, areas that I have worked on in my career..
I will give you an example of something I am very proud of, that actually preceded and inspired the evolutionary ideas. My first independent scientific problem my nascent lab investigated was about how microtubules grow. Microtubules were first seen clearly by electron microscopy and found to be in all eukaryotic cells. To me, they were fascinating structural elements. I am sure that many people were curious about how microtubules were organized spatially in the cell. Were there special machines that moved them there, like those that place railroad tracks? Were they enticed to grow in certain places like animal preserves? Nobody suspected that they really got to be “placed” where they are in the cell because of their intrinsic dynamics and because of their localized stabilization. Before we wrote these books on evolution, a student (Tim Mitchison) and I discovered how microtubules might be placed in the appropriate locations of cells. It was a property that we now call exploratory dynamics, which for microtubules was based on their mechanism of growth, which we named dynamic instability. This was mathematically similar to the dynamics of ant foraging. Microtubules grow randomly from a point of initiation of growth in the mitotic spindle or the nerve axon and are stabilized if they encounter a capture site. If they do not encounter a capture site (i.e. period went to the “wrong” place) site they would spontaneously depolymerize (disappear). This mechanism was general and important in biology. It was just an absolutely wonderful example of how ultimately, anatomy, at least on the cellular level, was facilitated by this rapid turnover and localized stabilization. In this way, collision with the centromeric region of a chromosome stabilized those few microtubules that were nucleated at the spindle poles and stabilized at the chromosome’s centromere. Variation and selection are not just for evolution but are really important parts of all biology.
HS: I guess you’ve sort of partly answered the question I wanted to end with, which is whether you would see this piece of work as one of your favourites?
MK: I think, personally, it’s my favourite because it is the most general. I have been fortunate to have students in my lab who have done other very important things. So, like a good parent, I try not to pick favourites. As I said, the growth and shrinkage of microtubules reinforced my belief in variation and selection as fundamental ideas in biology. Perhaps, learning is another powerful process in life, also involving selection. Song birds are capable of learning very different songs. You put a song bird next to a bird of a different species, it will learn that new song. If you play music to it, it’ll learn that song. So, it has this capacity to learn. And then, after internalizing this knowledge, it is rigid. Surprisingly, this too seems to be a lesson of variation and selection. The details of how a birds learns is similar, conceptually but not mechanistically, to how ants learn where food is and how microtubules learn where the chromosomes are. The muscles in the birds throat initially contract in random ways and produces random songs. However, the brain reinforces the muscle patterns that generate the right sound. “Right” in this case is a copy of the tutor’s song. This may be the way human beings learn to sing and dance. In the case of microtubules, stabilization is by specific interaction with a structure, which can be anywhere in the cell; in the case of mitosis, it is the centromeres of the chromosome. In evolution, it is simply the survival of the whole organism. John and I called these mechanisms exploratory processes. There are many homeostatic mechanisms in biology that remind one of a governor on a steam engine. They are machines. But exploratory processes are different. They are often not rigid at all. They depend on random variation. They are very life-like. It is intellectually the most original thing I have ever been associated with. There are some people today who seem to believe that the discovery process in science itself can be automated through massive data collection and machine learning. Both of these approaches are very useful, but, as yet, they only augment and do not displace tinkering, experimental design, and logical analysis. Science is more like ant foraging and microtubule assembly than we care to admit.
0 Comments