The Orthogonal Bet: Unveiling the Complexity of Life: A Conversation with Philip Ball on ‘How Life Works'

Samuel Arbesman:
Hello, Philip Ball, and welcome to the Orthogonal Bet. It's great to be talking to you. I really appreciate it. So last year, you published this fantastic book, How Life Works: A User's Guide to the New Biology. And I have to admit, and the book is mind-blowing and provocative and deeply interesting. And I say this... So I actually have a PhD in computational biology, although PhD is probably a little dusty and old. But even so, I still learned a ton, which probably speaks to the fact that so much is changing in biology.
I want to get into the content of the book and things like that. But first I want to know, what led you to decide to write this book? Was it talking to scientists and realizing that some of the usual metaphors we use for biology are just deeply out of date or just wrong? Or was there something else that led you to this moment?

‍

Philip Ball:
Yeah. Well, it was certainly that, Sam. I mean, you've said it yourself already that so much has changed in biology over the past several decades. And I wasn't really seeing that reflected in the conversation that we have about biology, certainly in the public arena, but sometimes also the way scientists themselves talk about it. What specifically triggered this book? It was something that I'd been thinking about. I guess I'd been having this concern for the past at least three decades. Really it goes way back to when I was an editor at the science journal Nature. And we would have these meetings. I was there as a physical scientist. I trained in physics and chemistry. But we would have these weekly editorial meetings where everyone would talk about what papers they'd accepted.
And it just seemed to me that some of the papers that the biology editors were talking about, they weren't really reflecting the kind of messages that I was getting from the public discussion about biology. It seemed like something else was going on. So I started thinking, "Hang on a minute, are we telling the right story here?" And I had these vague misgivings and I just lived with them for years. And it was really only in 2019 I was lucky enough to be invited as a visitor to the Department of Systems Biology at Harvard Medical School. And I spent the whole summer there, which was a fantastic experience. And I arrived with all these questions about, are we really telling the right story?
And it seemed to me that everyone, pretty much, that I spoke to in the department, I went round and had conversations with people, and they all seemed to be, one way or another, saying to me, "Well, it's even worse than that, actually." And they told me things that I hadn't realized at all about how the message is different from the one that I'd heard and from the one that I'd seen talked about. So I came away from that experience from that summer with this ragbag of notes and ideas and I was thinking, "What am I going to do with all this stuff?" And of course, the answer was, the only way I was going to make sense of it was to sit down and write a book. And this is really not the recommended way to write a book, but it's what I did nevertheless, which is that I figured the only way I'm going to figure out what all this means and what the story should be is to sit down and write the book and see what comes out, really in an act of faith that something would emerge.
And I hope something did. I hope that there is something coherent that ultimately comes out from this book about the kind of message. I don't think it's clear yet quite what the story is. But I do think we're seeing the broad outlines of a new way of talking about biology. And that's really what I mean by talking about the new biology. It's not some fantastic new theory that changes everything. It's an accumulation of little details in every area of biology, from genetics, to cell biology, to how tissues grow and so forth. That, in one way or another, seem to me all to be pointing towards this same direction. It's time to look for new metaphors and tell new stories.

‍

Samuel Arbesman:
And this idea that biology is far less deterministic or more diffusive in how cause and effect operates. And there's all these different kinds of things that bump up against the traditional ideas. And so maybe you can start by just talking about some of the ideas within biology that are in the popular imagination, they're just deeply wrong.

‍

Philip Ball:
Yeah. Well, I mean, there are some ideas that are deeply wrong. But I think actually, more generally, the ideas are just starting at the wrong point of the narrative or they're looking at it from the wrong direction. So again, it's not as though we have found some fantastic new discovery in the past 10 or 20 years that changes everything. It's, we've drifted really from the old message. I'm going to put it very simplistically because one has to. There's this question of, how are we put together? How does life work? It's this incredibly complex thing. These papers in nature and science and all the other journals make it look incredibly complicated and it's full of details.
But the message that we get to make sense of that is, never mind that. The fact is, we have this program, this informational program encoded in our genomes, in our DNA, that does everything. From birth, from the moment of conception really, we start off as this single cell, this fertilized egg that gets the chromosomes, the DNA, that contains all the instructions for building the organism. And that's the message. It's still up there on the site of the US National Institutes of Health, the department that looks after the human genome project. They still talk about this instruction book in our genome. And that's almost the universal metaphor that's used to talk about the genome.
So everything we are is supposed to be encoded in DNA. That I do think is wrong. And I think it's wrong in ways that are not just misleading, but that are potentially dangerously so. Because what it encourages, and people who work in this area, geneticists, they lament the fact that it encourages this, but nonetheless, it's out there. This message is out there and continually being reinforced. It encourages the view that what we are is our genome. So that defines us. If you look at, for example, a company like 23andMe that offers to sequence your genome by mail order. You send off a sample of saliva or whatever and they sequence it. And their advertising says something like, welcome to you.
You can find leading biologists who again and again come out with this idea, "Here's your sequence on this compact disc. That is you." And that, I think, is wrong. And that, I think, we really need to challenge. The role that genes play in building us, of course, is absolutely central and important, but it's not that deterministic idea that somehow it's all encoded in the genome, in DNA. In fact, we can only really trace that informational paper trail up to the point of making the proteins in our cells. Proteins are enzymes, they're molecules that make our biochemistry happen. They facilitate it happening. So they act as catalysts. Other proteins do other things, but that's really what some of the key proteins do.
That information for building a protein is, in one way or another, encoded in DNA, although even that isn't quite as simple as we're often told. That translation has a lot of contextual information about it. That information, which protein is made depends on what type of cell it is, which depends on where it is in the body and what developmental stage you're at. So there's top-down information coming from the whole organism or the developing organism that even determines how the information in DNA is read out and converted into proteins. But nonetheless, there's still a story you can tell about the structure of a protein being, in some real sense, encoded in DNA. So that's fine.
Once you get beyond that, once you start to think, what are all these proteins doing and other molecules? And there are all sorts of other molecules in our cells. How are they working together? What else is in our genome? Because it's not just instructions for proteins plus a load of junk. This was often the old idea that a tiny fraction of our genome is making proteins, all the rest is junk. The story we now need to tell is much more contextual than that. And I think one of the most important things is the way that the molecules interact with each other, the conversations they have. The old story is that there are just these chains of interaction where this molecule, this protein talks to this one and talks to this one. And that way, they pass on information down these chains, and that information ultimately creates us.
That's not how it happens, and it's certainly not how it happens in the cells of complex organisms like us. That's a story that's a bit closer to what happens in bacteria. And one of the other messages I wanted to get across in this book is, what goes on in bacteria isn't always a very good guide, and sometimes it's a very bad guide to what goes on in us. What happens in us is much messier. The molecules are often having group conversations. They get together literally in groups, in little committees, all of them talking, sticking to each other for a little bit and going off and sticking to another one without a great deal of selectivity. It looks like just a milling crowd, each having little conversations.
But out of those conversations somehow comes a decision, turn into this cell, like a liver cell, or turn into that cell, like a heart cell. How does that happen? Those are the things that we're now starting to try to get to grips with. But the logic of those conversations isn't the digital logic that we see in computer circuits. And that's why I think the computational metaphor for biology isn't necessarily the right one. It's much messier, it's much more analog, and it's much less specific than we thought. So the challenge now is to figure out how on earth, from this apparent molecular chaos to something an organism like us arise in quite a predictable manner as we develop in the womb. That's really the new biology.

‍

Samuel Arbesman:
I love this. Yeah. And one of the themes, and as you note, was, because single-celled bacteria are like the low-hanging fruit in terms of understanding, we've kind of thought, "Okay, that's the way everything in biology operates." And then we are slowly but surely realizing eukaryotic cells as well as multicellular organisms, these things have a very, very different way in which they operate. This idea also just that what is happening within a cell or even between cells, it's deeply physical and deeply spatial as opposed to being able to say we can always abstract away the spatial features and just look at it as some sort of informational component.
I think the informational stuff is interesting, but it's certainly only a very small portion or some fraction of the entire story. Do you think that one of the reasons why we've spent so much time trying to shoehorn all these other things and these more traditional metaphors is because of the widespread metaphor of computing? Because of that, there has been this sense of like, oh, we should try to explain everything in terms of circuitry and those kinds of diagrams.

‍

Philip Ball:
Yeah, absolutely that. We can see the reason for that because there was this extraordinary coincidence really, and it did seem to be a coincidence that we started to understand how it is that information is encoded in DNA, and it really is in a very meaningful sense. We started to understand that in the 1950s at the same time as computers were starting to take off. Not just that, but it seemed to be encoded in DNA as a linear tape of little letters, just like the ones and zeros that were encoded on the magnetic tape that, back then, was what information was held on for computers. So it made perfect sense. It was almost like the technological metaphor we needed to understand biology arrived just as we started to understand the biology itself, the molecular biology itself. That seemed to be the way.
I think looking back now, we can see that we were seduced by that, because there's a real sense in which there are those parallels, but there's a real sense in which that's not the right way to look at it at all. And Sam, as you say, this idea of the spatial is really, really becoming important. The old idea of systems biology was kind of to say, again, to put it crudely, all right, we're going to figure out which molecules are interacting with each other and we're going to draw these big networks, these big circuit diagrams of this molecule speaks to that one, speaks to that one. And there was no spatial component in it. There was no structure in it at all. It was all a conceptual idea.
But we now know that space is absolutely crucial. I mean, things move around in the cell for one thing. There are these compartments. But also, as I say, molecules seem all the time in our cells to form these clusters, literally these blobs, they're kind of like dense liquid-like blobs where a certain number of molecules get together and have these kind of conversations. It's totally a spatial thing. What seems to be really important for which genes are being read and which ones are turned off is the three-dimensional structure of this tangled stuff called chromatin, which makes up our chromosomes, which is basically the DNA with proteins that help to bind it.
This chromatin is unpacked by some enzymes, so the bits of it are physically accessible to the enzymes that are going to read out the DNA instructions and other bits can't be reached. So there's a whole language of the physical structure, the three-dimensional structure of our chromosomes. It is utterly central to the way our genes work, and that just isn't there when you think about it just in terms of a sequence. But it's not only space, it's also time. That how these things happen in time is really important. Just last week, I wrote about a paper where it turned out that the fate of cells in very early embryo, whether they become one type of cell or another type of cell, you could say controlled by a particular gene, but not by just whether that gene is switched on or off. It depended on how long the gene was being turned into protein for.
If you've got a sudden pulse of it, then the cells became one type. If you've got a sustained amount of it, they became another type. And we see this again and again that there's this temporal aspect to it as well. So we really have to think about things happening in biology in space and time rather than in this abstract informational space. And there's information in those dimensions, in the spatial component and the temporal component. And that includes information that's coming from the whole environment of the organism, but also from the other cells around. So that whether one gene gets switched on at a particular point in development depends on what the other cells around that one are like and what they're doing and which signals they send.
So we can only understand that process by looking at the whole, looking at the whole structure rather than just some self-contained program within an individual cell.

‍

Samuel Arbesman:
Okay. And so part of writing this book was to force people to realize that the old metaphors are not necessarily good fits for what we actually have in terms of biology. Beyond just recognizing the spatial and temporal complexity, what do you think might become the new metaphors? You hinted at this idea of looking at the whole organism. And is it these ideas around emergence or complexity science, are those the kinds of metaphors that we should be reaching for when we think about these kinds of things?

‍

Philip Ball:
Well, there are people who are working very hard and having some time to turn this notion of emergence from being the vague metaphor into being a real science, to really understand how emergence comes about. So I'm hopeful that there we'll actually have a real scientific language to understand that process. But we will need metaphors. We always need metaphors. And it's understandable why we use the old ones. I think one thing I say in the book is that the problem with metaphors is that it's very hard to get rid of them. It's very hard to know when to stop using them. We have a process crudely called the scientific method for figuring out when a theory is obsolete, that if experiments start to contradict it, then it's time to rethink the theory perhaps.
But there's no process like that for chucking out metaphors, so they just stick around because they're convenient. And people like me write about science, but also people... Scientists themselves just take it off the shelf without really thinking about the metaphor. But one thing I do say and I really strongly believe this, is that there is no technology that we have ever made that will serve as a good metaphor in general for how life works. That's what we've tended to do. And obviously we've talked about the technology of computing that we've tended to use. We've also used metaphors like that to understand how the brain works.
So once we thought that it was a kind of hydraulic system, and that now, we often think of that as a computer as well. But the more we understand about how life works, the more we see that those sorts of technological metaphors are, at best, extremely limited, at worst, extremely misleading. And so what I say in the book is that, it seems to me that the metaphors we need to understand life are ones that have to come from life itself. So for example, I think that it's meaningful to talk about life all the way down, down to the level of single cells in cognitive terms, to think of it as not a computational process reading out an algorithm, but to think of it as something more like the kind of cognition that happens in beings like us with brains.
And that's of course not... That is a metaphor. It's not to say that individual cells have any sort of consciousness or they're doing any sort of real thinking. But what they are doing is integrating information from many different sources in order to come up with a decision about what to do, how to develop, where to go next. And that to me seems to be a cognitive process and one that operates often on analog principles of somehow doing this integration and figuring out from all these different kinds of informational sources how to reach decision. That's not a strictly computational process.
That metaphor, if it is only a metaphor, and some people think it's more than that, of cognition, I think is a better way to understand life than computation.

‍

Samuel Arbesman:
And it reminds me also just the idea of agency. Agency and cognition in terms of individual action based on this combination of inputs and outputs. But I also love this idea, there is no technology that is quite like biology that we have thus yet. And therefore, we need to either grasp for biological analogies for biology itself or almost have this certain amount of humility. I mean, the way I think about this is, I guess it's not quite as common anymore, but several years ago, there was this whole trend of biohacking. Find this specific chemical or technique that will overclock your body or your brain in some specific way.
And I felt like it often came from this very engineering or computational or technological perspective of like, "Oh, we have engineered systems. We can understand them. The human body is a machine of some sort. And so therefore, let's just find this specific thing." And of course, we are these highly evolved, very messy, optimized in multi-dimensional way systems. And so therefore, it doesn't really work. Related to that though, what is the evolutionary story in all of this? And we have evolutionary computation where we can evolve within computers certain solutions to problems.
Obviously, evolution by natural selection has constructed these systems that are incredibly messy and complex. Do you have a sense of what evolution might be optimizing? I feel like I'm struggling with the right exact terms. What is the evolutionary story of how this very different system results in robust agency or cognition?

Philip Ball:
Right. Well, I'm going to start off by saying, the answer to that is we don't know.

‍

Samuel Arbesman:
Okay.

‍

Philip Ball:
At least I don't know. But I don't think anyone really knows. And I hope it's not a cop-out that actually, although I had to talk about some evolutionary aspects of all of this as I went through the book, actually I end it with an extended box section saying, "Well, what does all this mean for evolution?" Because I felt like, for one thing, that needs another book, for another thing, it needs to be written by someone who has a far more profound understanding of evolutionary theory than I do. Because it's really subtle. It's really much harder than we think. But also because I think that it's a book that is going to have to be written in some years time. We are not ready to really say yet whether there are implications for evolution.
Having said that, I think I'm quite a conservative from that point of view. There are some people who say this new understanding, and in particular, once we start to think more of living things in terms of agency rather than as kind of automata that are just being impelled by our genes, that means we need to profoundly change some things in the standard neo-Darwinian evolutionary model. Now, that may turn out to be true. But at the moment, I haven't seen compelling arguments for why that has to happen. I understand why that is being suggested, but I'm taking the position at the moment that we just don't know. Let's wait and see and let's ask that question in a quite cautionary way because it's very tempting. And we see this again and again to say, "Hey, Darwin was wrong. We've got to reinvent the theory." But actually, I don't see any need for that yet.
Having said that, one can see that there could well be some important implications of once we start to think of living organisms really as agents, they're in a very real sense cognitive. They're responding in an intelligent way to their environment rather than acting as automata, as programmed machines with the genes in control. Once we recognize that actually the genes aren't in control of this, to understand this notion of agency that every living thing has, and in fact is probably, I think, the defining characteristic, the most defining characteristic of everything alive, once we understand that that is not coming from the genes, that is a property of the whole organism, then perhaps there might be some ways in which we have to rethink how that process of natural selection is being reflected back into the genome.
There's one area in particular where I think there's going to be, and already is, an interesting discussion that's going on around what this means for evolution, which is, what's crucial to a lot of this new biology is not what genes we have and how genes are mutated, but how genes are regulated, how they are actually used. And what we actually find is that if we go way back to organisms that are single-celled but that seem to be the ones today that are most like the kind of organisms that were probably the single-celled organisms that became multicellular way back in time, whenever that was, we find that actually they have most of the genetic resources that we have. They're kind of ready to go. What they don't have is the control mechanisms, the regulatory mechanisms.
And in fact, we can see, and people have done this, of various stages in evolutionary history, when something profound has changed in the kind of structures that we see, for example, the sudden appearance of complex multicellular creatures, what seems to change, what seems to matter is not that they got a whole load of new genes. What they got was new regulatory processes, parts of DNA that aren't turned into proteins, but that are involved in switching genes on and off or turning their expression up and down. And so the story of how the genome evolved was regulatory. And there are some people who think that the way that happened was perhaps not actually through natural selection, but just through random drift, which is something that happens in genomes.
You get little changes that because they don't really make much difference to the organisms as a whole because they're not deleterious, they just get stuck there, but they change the expression of the organism. There is certainly a very serious school of thought and very good arguments that at the molecular level, it may be that a lot of the evolution that happened in the more recent past in our own evolution as complex organisms wasn't so much due to natural selection as it was due to random drift. None of that is at all incompatible with standard neo-Darwinian theory. It just means we need to recognize that it's grown quite a bit over the past 150 years or so.

‍

Samuel Arbesman:
I am also inclined to agree that we do not need to throw out all of the evolutionary biology that we have developed thus yet to explain some of these things. But related to this idea of almost like a repurposing of the regulatory processes and the actual pre-existing genes for these new functionalities or new structures, one of the things you discussed in your book and you go through different levels of hierarchy and complexity from within the cell to single cells, and then eventually you're going all the way up to the structure and shape and body plan of these organisms. This stuff, it was also just visually stunning to see how certain chemical gradients and things caused certain structural and morphological features to arise.
I don't know if you can talk about how these new approaches and then some of the ideas of understanding temporal and spatial changes, how that has given us a better understanding of how organisms are often shunted into specific body shapes and sizes and whether or not this is dictated by physics, biology, a complex interplay between all these things. Yeah, I would love to hear more about that.

‍

Philip Ball:
Right. Well, the way I think about it is that, for an organism like us, the genetic resources that we have, and that's really what our genes are and everything else that's in our DNA that is functional, they're resources that the cells use rather than programs that make the cells. So they're the resources that cells use to build these complex organisms. And the ways in which that happens, they're often expressed at a much, much higher level. And one very striking way in which that happens, like most people have got five fingers. So why have we got five fingers? Is there a gene for five-ness that gives us these five fingers? Well, there isn't.
It turns out that the way we think that happens, and there's good evidence for this now, it's due to a mechanism that was first described by the mathematician, British mathematician Alan Turing in 1952 in a very now famous paper in which he talked about basically the chemical basis of morphogenesis. Morphogenesis being how we get our body shape. He was thinking about a very simple chemical system in which the chemical agents interact with one another and react with one another in quite unusual ways. One of them is an autocatalyst. It just makes more of itself. The other one is an inhibitor. So it stops the autocatalytic one from doing that.
Under the right circumstances, those two components, out of a very homogeneous mixture, you can think of it as just a well-mixed beaker of stuff, they cause that system to spontaneously separate into patches with different chemical composition. And Turing came up with this model because he was interested in the way early embryos, as he would call it, break symmetry. They start off as what seem to be a uniform ball of cells. So how you break that uniformity and start getting things that become limbs and organs and so on. And he suggested that this might be a way that it happens, that you get particular biochemicals that he called morphogens that just diffuse throughout the system and that interact with one another in this way of autocatalysis and inhibition that create these patches that are then different from one another.
So one patch might then develop into a limb and another might develop something else. And the idea was that once you have those differences in composition, perhaps that throws a genetic switch, switches on certain genes that then drive the development towards a limb. Now, Turing came up with this idea and it just sat there for decades because it seemed really neat, but biologists didn't know what to do with it. They didn't have any reason to believe that it actually was the way embryos developed. It's only been over the past really two or three decades that we started to find biological systems where something like this seems to operate, and the development of our hands and our fingers is one of them.
So what seems to happen, and this has been imaged and studied, and we can start to get some idea of what the actual proteins involved are, that what seems to happen is that in the developing organism, in the developing embryo, the hand starts off as just a round bud, the blunt end of a limb. But then you get this diffusion that's happening with these morphogens. And one of the characteristic patterns that emerges from these Turing systems is stripes. So basically our fingers, I mean that's what they are. They're just stripes of more or less equal width. So within this limb bud, you get a system of radiating stripes that happen. And then within some of those stripes, bones start to form and they start to become the fingers. And in the bits in between, the cells die away so we don't have the webbed connections between them.
And sometimes when that doesn't happen perfectly, you do get a bit of webbing that's retained, and some people have that in fingers and toes. So that's really what's going on. And it just so happens that we have five fingers because this process happens at just the right time during the growth of the embryo that precisely five stripes fit into the available space for the bud. And if something goes a little bit awry there, and it can be because of some regulatory process actually that happens with the genes that are involved in this process, then sometimes you can get more than that. And so you get polydactyly, so people born with more than five fingers.
And the way in which that happens is exactly what Turing's model would predict in that often it's the bifurcation of just one of these fingers into two that happens midway, perhaps down the finger. That's exactly what Turing predicts. So that seems to be what's going on. Is that process encoded in the genome? I don't think it's meaningful to think about that. You'd never deduce that, no matter how you actually studied the genome. It's really something that has its own rules that operate at this much higher level of the way that the tissues themselves are developing. And that's one of the key messages I wanted to get across in this book. There is no privileged level at which we must look in biology. No privileged level of magnification.
Whether it's genes, gene networks, or cells, or tissues, all of them matter. All of them seem to have their own, more or less, autonomous rules. And we need to understand all of them to figure out literally how life works and how an organism like us grows and functions.

‍

Samuel Arbesman:
And two of the researchers you mentioned quite a bit in the book, Josh Bongard and Michael Levin, and they're thinking at some of the even higher levels of abstraction in terms of like electrical gradients and things like that. What is your sense of... And well, maybe you can talk about some of the research, kind of what they're building and exploring there. What you think might be the implications for thinking at this yet other level of abstraction in terms of hierarchy?

Philip Ball:
Well, one of the general questions that comes about from thinking this way is when you start to realize cells have their own rules for the kinds of ways they build and the kinds of things they build. And those rules are in some sense derived from things that are happening at the genetic level and higher levels within the cell, but in general, probably not the fine details of that. Cells communicate with each other to form these structures. So normal development. They turn for us into something that looks pretty much like us. And we know that there's a whole range of different bodies that they can form. And we often think of, for example, if a limb fails to develop to its full length, we think of that as a mistake during development, but it's not a mistake. It's one of the possible outcomes of what cells can build.
And of course, it can. That's not to deny the complications and the difficulties that it can cause for someone who has a condition like that. It's simply a possible outcome of those rules that cells have. Now, we can then ask, "Well, what else can they build? Have we seen the full palette of those possibilities?" And I think it's inconceivable that we've seen the full palette. There are all sorts of things I suspect that cells could build that evolution doesn't build because the conditions don't allow for it. There's this work by Mike Levin at Tufts and Josh Bongard who is I think still at Vermont, and they're collaborators, thought, "Well, what if we liberate cells from the normal constraints that determine development and just see what they'll make?"
And they did this with frog cells because frog cells are big and they're quite easy to work with. And they took frog cells from a frog embryo, but they developed to the point where they'd become skin cells already. And they just separated them from the embryo and left them in solution to see what they would do. And the cells clumped together, not just at random, they seemed to clump into clumps of a particular size of several hundred or several thousand cells, and then they stopped clumping. They decided, "Okay, this is what we want to make."
The thing with frog cells also is that they have little protein appendages on the surface of the cells, like little hairs really, that can wave about. Because in the frog itself, these appendages wave and they move mucus about on the surface of the frog skin. But in solution, it turned out that these clumps of cells, they develop these little appendages that start waving about, and they allow these things to swim around, and they seem to be able to swim around in a coordinated fashion as though they're rowers in a boat, pulling on one way on one side and another way on the other side, and off this thing goes.
So they look like organisms. They look like little microorganisms, but not single-celled ones. They're clumps of cells. And so Mike and Josh called these things xenobots. Xeno here coming from xenopus laevis, which is the Latin name for the kind of frog that they used. But I like to think the ultimate root of xeno is from the Greek strange or stranger. So these are certainly strange robots and they're living things. They're absolutely alive. Are they organisms? Mike thinks they are. And lots of people think they're just a epiphenomenon. They're not really living in any real sense.
But they do seem to be structures that have, in some sense, an agenda. Some of them move in very particular ways. Some go in straight lines. They seem to navigate themselves round, bends in a tube. Some of them even seem to marshal other cells together so that they form new xenobots. So they have a behavioral repertoire. It's work like that that is starting to open up this broader question of, what else can cells make? And this is a question that has all sorts of implications for biomedicine and beyond, for biorobotics. What can we make if we understand the rules that cells use to assemble and to differentiate into different cell types? Can we use and manipulate those rules to build new kinds of living entity, possibly even new kinds of living organisms.

‍

Samuel Arbesman:
Even when there's something that's a non-traditional outcome, which we might consider a disease, it's the outcome of these rules maybe operating in a different way than we might've expected. I would love to hear more about how this cognition-based understanding or these different metaphors might inform how we should think about disease or medicine going forward.

‍

Philip Ball:
One of the ways we need to think about that is, when we think about disease, what is the real causal level of that disease? We've tended to think... And again, we hear this, I think, sometimes distorted message that once we've read our genome, we are going to find out everything we need to know about disease, because they must all stem from genetics. And it's now no secret that in trying to understand and develop cures and treatments for something like cancer, that approach really hasn't been very successful. We understand quite a bit about the genetic roots of cancers. But in terms of treatments, it's given us very little. And we're still treating cancers basically by hitting them with stuff that's going to kill off everything. And we hope that it'll kill them more than anything else, or just cutting them out. And we're doing it now with robots instead of just a surgeon's scalpel.
But that's still the way we're dealing with it, that we haven't really been able to attack cancer at its genetic roots. And I think the part of the reason for that is that those aren't the roots of cancer. And in fact, we know that they're not the roots because this idea that cancer is due to mutations in our genes, very often, that seems to be some aspect of what comes about. But so many different mutations can lead to cancers, and they all seem to be, in the words of the developmental biologist, Connor Wollington, they seem to be canalized into this one mode. So cancers, they all pretty much do the same thing, which is to proliferate. And they don't just proliferate madly. They actually proliferate in, some sense, quite intelligently. But that's what they're doing.
All these different mutations can cause that. Because ultimately, what they do is cause some dysfunction in the way the cell cycle, the replication cycle of cells is regulated. But we now know that there are some cancers that don't even have genetic, they're not even related to genetic mutations. Quite recently, people have reported cancers that seem to come about... There's no change in the genes, there's no mutations. It's in the way the genes are controlled by what are called epigenetic markings. Basically chemical groups that are stuck onto them that regulate them. That in itself can cause cancers in some cases. So we need to think about what's really causing these cancers.
And I think that's already starting to happen. That when we see it as a cell level phenomenon rather than a gene level phenomenon, then, for example, one relatively new approach is to start saying, well, what's really happening is that our cells have gone into a different cell state. We can think of it as a defective one if we like, but actually it's just one of the possible states that our cells have. It just happens to be one that can kill us. So if we understand those transitions between different cell states and what controls them, can we switch them back again instead of killing them off? Can we reverse that process and reverse the cancer cells back into a healthy cell state, or perhaps direct them into a state where instead of proliferating, they just sit there and don't do anything and eventually die?
And this is an approach called differentiation therapy. Differentiation being this transition from one cell type to another. There, we're not doing anything to horribly try to kill as many cells as possible. It could potentially be a gentler way that is just guiding the cells back into a non-pathological state. That comes about from thinking of cancer at the causal level of cells and a cellular dysfunction rather than genes.

‍

Samuel Arbesman:
I love this. Yeah. It also makes me think that there could be potentially, yeah, for just... and scientifically mapping this high dimensional territory of different states and seeing which ones are closer topologically and how can we move between these in a much more natural cell-based understanding. Now, this is super fascinating. So you wrote the book because of all these things that you were surprised by and trying to make sense of. Was there one thing in particular that you were most surprised by in this journey?

‍

Philip Ball:
Well, there were several. But I think one of the ones that really gave me a new perspective on this was, as a chemist, I'd grown up with this idea that cells are so full of different... There are thousands and thousands of different proteins in every one of our cells, and all this other stuff. There's lipids and there's salt ions and all the rest of it. How on earth does any one protein... You see these nice tidy diagrams in the scientific papers where this protein speaks to that protein. It almost just walks over to it and have a little conversation. But in fact, it's this mad crowd. So how does that specificity happen?
And I have this very comfortable picture that I've been told, and you find it in a lot of textbooks that, oh, well, you see, these proteins have very specific shapes. They're like a lock and key. So this protein only fits with this one. It only fits together with it. And so, it just ignores all the others because they're the wrong shape, and it just wanders around until it finds its partner. That's fine. So that way, you get these very specific communication channels. And then I discovered that it's not like that. That actually, so many of our proteins, and in particular, some of the most important proteins in our cells, that being ones that often are at the hubs of disease and of development, are proteins that don't have these nice specific structures at all. They are so-called intrinsically disordered proteins, or bits of them are intrinsically disordered. They're floppy. They're loose.
And because of that, they stick to all kinds of things. This idea of so-called molecular recognition, I mean, it happens in some cases, but in a lot of cases, it doesn't. It goes out of the window. And so you are back to this question of, well then how on earth, what is the logic of all these molecular interactions if they're not that specific? And that's where you have to start thinking in terms of these fuzzy interactions, these communities and collections and committees of molecules that have degrees of preference for one another, but they're not all that choosy about it. And somehow, out of that kind of logic, comes still these very specific and what seem to be quite deterministic developmental trajectories on the whole.
So that opens up a whole new set of questions, not just in terms of how life works, but why is it this way? Because this is the really interesting thing to me. It's not really like that in bacteria. Nature is quite capable of making proteins that fit together very nicely. And you see much more of that. You don't see many of these disordered proteins in bacteria. So if you like, put it anthropomorphically, evolution has chosen this way for us because there must be something useful in it. There must be something very important in this promiscuity of molecular interactions that allows the kind of adaptive and, in some cases, innovative responses that our cells need.
Because that's the thing about complex creatures like us. One day is never the same as the next. We never know what's going to come along. We never know what we are going to encounter. So evolution can't encode every behavior into our genome. It has to build this flexibility into our systems at the behavioral level, sure, but even at the molecular level. And that's again why I think it makes sense to think of this as a cognitive process. So I think there are very clear and sound reasons why we work this way. But that realization that it's not by lovely, neat molecular recognition that really changed my perspective.

‍

Samuel Arbesman:
Oh, that is amazing, and probably the perfect place to end. This is fantastic. Thank you so much for joining me. And yeah, this has been an amazing conversation. Thank you.

‍

Philip Ball:
Oh, thank you, Sam. I've enjoyed it. That's fun.

‍