Philip Ball explores the new biology, revealing life to be a far richer, more ingenious affair than we had guessed. There is no unique place to look for an answer to this question: life is a system of many levels—genes, proteins, cells, tissues, and body modules such as the immune system and the nervous system—each with its own rules and principles.

In this talk, discover why some researchers believe that, thanks to incredible scientific advancements, we will be able to regenerate limbs and organs, and perhaps even create new life forms that evolution has never imagined.

Philip Ball is a freelance writer and broadcaster, and was an editor at Nature for more than twenty years. He writes regularly in the scientific and popular media and has written many books on the interactions of the sciences, the arts, and wider culture, including 'H2O: A Biography of Water', 'Bright Earth: The Invention of Colour', 'The Music Instinct', and 'Curiosity: How Science Became Interested in Everything'.

Philip's book 'Critical Mass' won the 2005 Aventis Prize for Science Books. He is also a presenter of Science Stories, the BBC Radio 4 series on the history of science. He trained as a chemist at the University of Oxford and as a physicist at the University of Bristol. He is the author of 'The Modern Myths' and lives in London.

Transcript:

(00:00) (electric music) (audience applauding) - Thank you. Thank you, it is so nice to be back in the finest science lecture theatre in the world. And I've been giving talks here for over 20 years now, I realise, but I don't think I've ever had a crowd quite like this. So I feel very flattered and a little bit nervous tonight.

(00:22) There are times, aren't there, when we could really do with an instruction manual for the human. Just imagine if whenever something goes wrong, we could just look up the correct fix in the manual. A viral infection, an allergy, whatever, we just look it up. But the truth is that after pondering this question for thousands of years, we still only got a very sketchy idea of the answer.

(00:49) Now, you might think that to answer this question, we should really start with another one. And some very smart people have pondered that question over the years, but still, no one agrees on the answer. I'm going to go, for the time being at least, with the answer that the British biologist J.B.S.

(01:12) Haldane gave at the start of his 1947 essay with this title where he simply said this. (audience laughs) So we can park that question for now, I'll come back to it, but the fact is that I know I'm alive and I know you're alive, and so we can ask how do we work? Well, this is where we'll start with that question, down the pub, specifically down the Eagle Pub in Cambridge which was where, according to the American biologist James Watson, he and Francis Crick burst in in 1953 for Crick to exclaim to everyone, "We have discovered the secret of life."

(01:47) Watson recently admitted that actually he just made up this story for his book. Despite having unveiled this plaque to that effect, he made it up (audience laughs) for dramatic effect in his 1968 book, "The Double Helix". For it was that double helix molecule that he and Crick had just deduced the structure for the DNA, the double helix of the DNA molecule.

(02:12) And Watson wanted us to believe clearly that this was the secret of how life works. DNA was then widely thought, not by everyone, but very widely thought to be where the genetic information that is passed between generations is stored. And in this double helical structure, Crick and Watson had seen how it can be that our DNA is able to store this genetic information.

(02:38) Each of the twin strands of the DNA double helix is a string of a succession of just four different chemical building blocks, which we denote by the initial letters of their names. So C, G, T, and A. And this, they said, then acts like a code, just like the strings of binary digits, the ones and zeros that were then being used to encode information on the magnetic tape that was used by digital computers.

(03:11) So, each of our genes then is a short segment of this code, and we have many thousands of these genes in the string of letters, chemical letters, that for we, humans, stretches to 3 billion of these letters long in which we call the genome. So here, then, was apparently how we work. The single fertilised egg from which we all came comes loaded with a set of instructions in the genome and all that needs then to happen is that these instructions are read out to build us.

(03:48) So all we needed to do now was to read the instructions. And that was what was done in the Human Genome Project which began in 1990 and was completed at least as a rough draught by the turn of the millennium. As Bill Clinton said when that draught was first publicly announced, "Today, we are learning the language in which God created life.

(04:09) " So here it is, our instruction booklet as it's often called, or our blueprint, or the letters of the human genome, which you can go and read, should you be inclined, in these 109 volumes held in the Wellcome Collection on Euston Road. Is this then the secret of how life works? Well, some people have questioned whether this blueprint or instruction booklet metaphor is really the right way to think about the genome.

(04:38) The Oxford physiologist Denis Noble has suggested that perhaps a better analogy might be a musical score. And that's certainly better, I think in some respects. (orchestral music) So what's the cause of this music? Well, some might say it's the score composed by Beethoven But is it? That answer reminds me of the comment by the Russian violinist Jascha Heifetz when he was once complimented by an admirer after a concert he gave, who came up and said, "Your violin has such a beautiful tone."

(05:22) And Heifetz picked up his instrument and said, "I don't hear anything." (audience laughs) And his point of course was that that beautiful tone doesn't just sort of happen. It's not inherent in the violin. It takes something more to produce it. By the same token, if you put the human genome, if you just put it in a glass of water it'll just sit there.

(05:43) It'll never make life at all. It won't make a single cell, let alone us. Well, obviously, right? I mean, obviously you need an orchestra to actually get any music. And so one reason why Denis's metaphor, I think, is quite nice is that it helps one see it's not so much that the musical score is the cause of the music, but that different musical scores, when played by an orchestra, account in some sense for the differences between the music they will make.

(06:15) So the problem with saying that the genome is the secret of life is evidently that the genome is not what makes life happen. Sure, life won't happen without it, just like Beethoven's "3rd Symphony" wouldn't happen unless Beethoven had prepared a score. But what we're really saying is that for life to happen, the genome needs to go into an already living system.

(06:39) So it's really, the story is, (orchestral music) oops, sorry, I move it quickly on from there. Didn't know that would... Oh, it was hovering over the, (chuckles) over the music. So this is really the right way to tell the story. Life is, in other words, a genome plus life. (audience laughs) Or to put it another way, this is the picture that we're often given.

(07:03) And we're told what goes on in this black box is so horrendously contemplated that you're probably better off not looking inside it at all. But just rest assured that scientists are working on it, and one day they'll have it all sorted out. However, that advice neglects this inconvenient fact. So, to understand how life works, I'm afraid that, yes, we really do have to take a peek inside this black box.

(07:29) And I'm going to try to give you a little glimpse of it inside tonight. And you know, I won't deny that in all its glory, it really is horrendously complicated. But I hope all the same to persuade you that we can make some sense of it. And what's more, that over the two decades or so since the Human Genome Project was completed, the sense we've made tells a rather different picture from the one that we have traditionally been told in the genomic era.

(07:58) Well, Francis Crick opened up that black box a little bit, and this is what he saw, This is what, in the late 1950s, he called this central dogma of molecular biology. Now, that's an odd name, isn't it? Because science isn't meant to have any dogmas. It's meant to be provisional and subjective revision, right? But Crick later admitted that he called it that because he didn't really know what dogma meant.

(08:22) (audience laughs) But all the same, calling something a dogma in science is a bit like a red flag to a bull because it has goaded a lot of people to devote a lot of time and energy to trying to argue and to show that Crick's central dogma is wrong. I'd say that a better way to look at it is that it's not so much that it's wrong, but it's a bit like the Holy Roman Empire, which was once famously said to be not holy, not Roman, and not really an empire.

(08:48) So the central dogma is not so much wrong as that it's not a dogma and it's not really so central. But let's take a look at it. It sums up what was already thought to be the case and what was soon proved to be the case that what the genes really encode is instructions for making these molecules called proteins.

(09:08) And you probably remember the usual cliche about proteins. They are the workhorses of the cell, which means that they're the molecules that make the cell's biochemical reactions happen generally by acting as catalysts that help convert one biological molecule into another. And the way the code in the genes is read out and turned into proteins is a two-step process.

(09:34) So the first step is that the code in a gene, a given gene, is copied, and the technical word is transcribed, into a very similar molecule to DNA called RNA. So just that little bit of the genome is transcribed into this so-called messenger, RNA molecule, and then this mRNA moves off and is seized by another piece of molecular machinery in the cell called the ribosome, which uses the information that the RNA encodes to put together a particular string of amino acids in a so-called polypeptide chain, which then folds up into this compact shape,

(10:13) and that's the protein molecule. So the protein has a particular shape, and this is the process called translation. And there are thousands and thousands of these proteins in every one of our cells. Here's a snapshot of the inside of one of those cells. And this isn't just a sort of computer cartoon of what it might look like.

(10:33) This is a computer generated image using actual data from inside a cell. So this is what it's like inside every cell of our bodies and you can see it is insanely crowded. But no matter, the usual story goes because each of those proteins has, as I say, a very particular shape which fits together like a lock and key with the molecule that it's meant to transform.

(10:58) So, it will only go to work on that molecule and it will ignore all the others. And through this series of highly specific molecular interactions, proteins somehow put together us. Well, here's how that will be expressed within the context of Crick's central dogma. So this fancy word here, phenotype, basically it just means all the traits we have.

(11:22) What forms we take, how tall we are, what colour our skin and our eyes are, even what behaviours we have. The phenotype is basically our individual selves. And in this picture, it all comes from what's encoded in the genes, in the so-called genotype. But since the Human Genome Project was completed, more and more complications and problems with this story have emerged, and I want to tell you about some of them.

(11:51) So I said that we have thousands of genes and they make thousands of different proteins. How many genes exactly do we have? Well, a banana has 36,000 genes. Now, think very carefully before you answer this question. Which organism is the more complicated? (audience laughs) Okay, so it stands to reason that I'm going to need more than 36,000 genes, right? When the Genome Project started, a common figure that scientists would give for the estimated number of human genes was around 80 to a 100,000.

(12:30) So I've shown it in a dotted line here 'cause it's just an estimate. But we pretty soon got a sobering wake up call, because as we started getting into the project those numbers fell dramatically. And when we actually started getting real data, the solid line here, we found that it is much lower, that it was something like 20,000 genes that encode proteins.

(12:50) And now, some scientists think it might be even slightly lower, might be as many as 19,000. So that's about as many as a tiny soil dwelling worm called the nematode has, and it's scarcely half what a banana has. And these numbers are often paraded now as a comic example of how wrong expert opinion can be.

(13:11) But I think the really important question is why they were wrong? Did we perhaps have the wrong idea about what role genes were playing? Well here's another thing. In the 1990s, there were one or two genes for which biologists didn't seem able to find the corresponding protein. And in the end, they had to conclude that that's because there aren't any.

(13:35) These genes simply make RNA, and not the messenger RNA that gets translated by the ribosome into a protein. The RNA itself is the end. It has some biochemical function, it does the kind of things that we thought proteins do. And these RNA encoding genes, and I've just marked them in red here, they're called non-coding genes.

(13:59) Not clearly because they don't encode anything, but because their genes don't encode proteins, which is what we thought all genes did. Well, okay, but biology is full of weird exceptions, right? Except that these weren't exceptions. Over the past two decades, the number of these non-coding genes kept creeping up and eventually, just a few years back, that number exceeded the number of protein-coding genes.

(14:29) And what's more, current estimates are that it's just going to continue that way. That actually it's going to turn out that there are lots more of these non-coding genes, that they vastly outnumber the protein-coding genes. And the picture is actually even more transformed than that because these are just the pieces of the genome that encode these relatively long, non-coding RNA molecules that qualify as genes.

(14:54) But it's been discovered that there are actually lots of other bits of the genome, or our genome, but also those of other large animals like us, called metazoans, that encode lots of smaller RNA molecules. And there are all these different families with these fancy names that do all sorts of tasks in the cell, protein-like tasks.

(15:16) So, you know, the genome is full of things that don't encode proteins. So you can see, the genome isn't really what we thought it was about, and yet somehow we still seem to be telling, certainly hearing, the same story about it that was being told back in the 1990s or even the 1970s. That can't be right.

(15:40) It's rather as if cosmologists were to have said after discovering that four-fifths of the matter in the universe is made up of this so-called dark matter that we can't see and about which we know nothing, as if they just sort of shrugged and said, "This change is nothing." Fortunately they didn't say that.

(15:57) Well, it's actually even worse than this. Crick's central dogma said that genes encode proteins, and by that we mean that the genes actually programme the proteins with particular shapes so that they can go and do their specific jobs. But, here's one way in which this picture is now modified. It doesn't mean that each gene encodes a particular protein.

(16:24) In fact, each of our genes can typically be used to make several different proteins. On average, each can make about six different proteins, but some genes can encode dozens or even hundreds of different proteins. So we have many more proteins and no one knows exactly how many more, but many more than we have protein-coding genes.

(16:44) How is that possible? Well, it's because, as was first discovered in the 1970s, the messenger RNA that is transcribed from a gene is typically chopped up and edited before it is translated. So there's another piece of molecular machinery, this thing called the spliceosome, made up of several different proteins that gets hold of the messenger RNA, chopped it into fragments, throws away some pieces called introns, and stitches together the remaining fragments called exons back in various orders.

(17:18) What decides how this editing and splicing occurs is typically information coming from a higher level of the system. Say, for example, from the overall state of the cell in which it's happening. So, a gene in one tissue might produce one type of protein and in a different tissue might produce a different protein.

(17:40) In other words, the information flow here isn't, as the central dogma at least implied, isn't all from the bottom up from DNA to RNA to proteins. Some crucial information for making the proteins is coming from the outside in some sense. And this is just one of the ways in which in order to build us and to keep us alive all these years, information doesn't just flow upwards from the genes to higher levels of organisation, but flows up and down and in between and in all sorts of directions among them.

(18:13) It's an open informational system, not a closed one. And here's another change to the picture. In the analogy of the genome to a musical score, we might say that the score is what prevents the orchestra from just playing a whole load of random notes, right? Making a racket because it tells each musician exactly which notes to play and when.

(18:36) And that's the equivalent of the way a protein's gene encoded shape, like this one, tells it what to do in the cell, which molecules to grab hold of and which to ignore. So I mentioned earlier this lock and key aspect. But we now know that for many of our proteins, including some with some of the most important jobs in the cell, the DNA score isn't like this at all.

(19:02) It's much more open to interpretation. And what I mean by that is that some genes encode proteins without assigning them a structure. It leaves them loose and floppy, or as biochemists say they are intrinsically disordered. And this isn't some failure of the genome to give proteins a proper shape. It's clearly a deliberate feature that evolution has, so to speak, chosen.

(19:28) Because you see there's much less of this intrinsic disorder among the proteins of simpler organisms like bacteria. So evolution clearly was fine giving proteins all of a very specific structure, but it seems that it has found it useful or perhaps even necessary to give proteins disorder in order to make more complex, multicellular, multi-tissue organisms like us.

(19:54) Given that so many of our proteins, but perhaps a third to a half or maybe even more of them, have parts or holes that are disordered in this way, given that, it seems a little odd that we didn't really know about this until the past few decades. But that's because the methods that scientists have in the past used to look at protein structure generally only work well for the kinds of proteins that are ordered, that have a fixed structure.

(20:22) And so they'll pack together and form nice orderly crystals which are what you need for those methods. It's a bit like the way we overlooked non-coding genes for so long. We tend to see only the things that we expect to see and we tend to study only those things that we have techniques for studying.

(20:43) So proteins with this intrinsic disorder, these floppy proteins, are far less choosy about which other molecules they stick to. They're rather indiscriminately sticky. That's to say, they're quite promiscuous in their molecular unions. So, that old idea that the molecular chaos of the cell is somehow kept orderly and tamed because each of the proteins is highly selective about what it interacts with, that idea doesn't really work.

(21:11) And let me show you a particularly important example of this kind of molecular promiscuity in action. So, the standard argument for how we get to be so complex with so few protein-coding genes goes something like this. That we say, well, it arises from the complexity of all the different interactions between those molecules that interact in these really horrible looking networks, where, you know, each of these blobs represents a protein.

(21:36) And it looks pretty horrid, doesn't it? If you open a copy of Nature at random, you're bound to see pictures kind of like this, and it's tempting to think of them as cartoons of what all the various molecules are actually doing, moving around the cell. But once you remember how complicated a cell really is inside, you have to wonder how on earth a complex dance of molecules like this could be orchestrated.

(22:02) All the same, the idea is that this molecular crosstalk explains how it is that, for example, different genes are turned on and off in different types of cell in our bodies. Making heart muscle cells different from skin cells or liver cells even though they all have the same genome. The idea is that proteins made by one gene could, for example, act as a kind of switch to control the transcription of another gene, so that it turns on or off whether that gene is transcribed and translated.

(22:37) And that process is called gene regulation. Now, we've known for a long time, since at least 1960s, that gene regulation happened. It was around that time that the French biochemists Jacques Monod and Francois Jacob showed how this process worked for one particular type of gene regulation in the bacterium E. coli.

(22:59) Now, E. coli can digest two different types of sugar. It can digest glucose and lactose, but it's not very efficient if the bacterium is constantly making both of the two enzymes needed for those two processes when there's only one sugar or the other around. And so there's a switch, which Monod and Jacob called the lac operon, a switch for turning on and off the production of the so-called Lac enzymes that are the ones that digest lactose.

(23:28) So in short, this is what happens. There's this protein, this green thing here, that can recognise and stick to a little patch on the DNA, this yellow patch, just before the lac genes themselves. And if it sticks there, then it blocks this pink blob, the RNA polymerase, that produces transcription that produces RNA.

(23:53) It blocks it and sort of kicks it off so that it can't do its job. And so it stops the transcription of the lac genes. So, it's like a nice, simple digital switch. There's a transparent logic to it. And for a long time, molecular biologists figured that gene regulation in organisms like us follows the same kind of principles, effectively wiring our genes into a network a bit like a digital circuit like we have in microelectronic devices.

(24:24) Well, you can probably guess what I'm gonna say. That's not how it turned out. (audience laughs) Very often, gene regulation in metazoans like us is much more complicated. These proteins that interact with parts of DNA to control gene expression are called transcription factors. And it turns out that many of our transcription factors are intrinsically disordered proteins, which aren't so selective as these bacterial proteins in what they bind to, maybe in which bits of DNA they bind to or which are the molecules.

(24:57) And the bits of DNA that regulate our genes, like this sort of yellow section here, there are lots of them in our genes and they're not all next to the genes they regulate. Some of them, weirdly, are a long way away on the DNA strand, somewhere else entirely. And these are regions called enhancers that somehow control the extent to which the gene is switched on or off.

(25:23) So gene regulation in us tends to involve a whole bunch of different components, transcription factors and other molecules, often including some of those non-coding RNA molecules. Most of them are regulatory. They have functions in gene regulation. And then there are other things. There are all these bits of DNA that are controlling the process somehow, some of which because they are so far away they are brought close by pulling out these big loops of DNA and sort of looping them back around, like tugging out a piece of wool from a tangled ball.

(25:58) And all of these components fit together into this kind of gigantic regulatory assembly that isn't something that just kind of clips together neatly. It's a loose disorderly blob, a dense cluster sometimes called a condensate, which forms like a kind of liquid droplet, a bit like a blob of vinegar in the oil of salad dressing.

(26:24) So, you know, it looks like a really messy way to do this job. It's as though gene regulation in us is done by these loose committees of molecules all talking to each other rather indiscriminately. And it really, it's pretty amazing that somehow all of these components still manage to make a reliable decision about whether to switch gene expression up or down despite all this fuzziness amongst their interactions.

(26:52) This sort of fuzziness and the way it produces collective decisions rather than simple, digital logic that we see in bacteria based on precise molecular unions, this fuzziness is a characteristic feature of our molecular biology. We don't quite understand how it works, but I think we can start to see why it is that for us life does work in this fuzzy analogue and rather open-ended way.

(27:24) You see, once you start to think about it, the more complex an organism is, the more a blueprint or an instruction book approach to controlling how it works starts to look like a terrible idea. If the organism working correctly depending on each of those instructions being executed perfectly at just the right place and just the right time in some sort of complex clockwork manner, then it's just not going to happen.

(27:52) Not least because the molecular world isn't like clockwork machinery, it's full of randomness and noise. It would be like trying to, expecting a mechanism like this to go on working perfectly if you were to immerse it in the bath and shake it around. It's not going to happen. I mean, sure, you can, and this was often the story that was told before, you can build incontingency plans.

(28:18) So if one bit fails, there's another way for the same thing to happen. So the idea was that in these complex networks, there's more than one route to the same end, so that if this route fails, there's always this one. But when you think about it, that's not a great way to solve the problem.

(28:34) The answer to the fragility that comes from something that's very, very complex If it has to all work perfectly, the answer to that problem can't be to just give it more complexity. Instead, you need to use totally different design principles for making it. And that's surely why we have these fuzzy molecular mechanisms, where the details often don't really matter.

(29:02) The committees can still come to good decisions even if some of the members are absent or asleep. What these principles really amount to, what this idea really amounts to is taking the responsibility for the correct functioning of the whole thing off the lower levels of the system and handing it up to some higher level.

(29:23) In other words, we work in a way that is designed to take the pressure off our genes. To make them, in general, no longer the real cause of our traits and behaviours of our phenotype. So, mistakes and malfunctions at the lower levels can be compensated for higher up. And this happens actually at other sort of levels in the stratum in the hierarchy of the way complex organisms like us work.

(29:52) So that, for example, if cells during the development of an embryo are sort of don't quite end up where they're supposed to, often there's a way to compensate for that further down the developmental line so that you still end up with a perfectly viable organism. And let me give you briefly a couple of examples of how these higher level principles that help organise our tissues and bodies reveal this kind of dispersal of responsibility so that they involve genes without in any sense being blueprinted by them.

(30:29) So, the surfaces of our intestines are covered with these little finger-like protrusions called villi, which hugely increase its surface area so that it can absorb nutrients sufficiently into the bloodstream. And these are basically bulges in the surface layer of tissue called the epithelium, and their growth is triggered by a protein that rejoices for roundabout reasons in the slightly silly name of Sonic Hedgehog.

(30:55) (audience laughs) Don't ask. But this doesn't by any means imply that this Sonic Hedgehog protein, or the gene that encodes it, is a gene for villous growth. In fact, Sonic Hedgehog is a general purpose, embarrassingly to some extent, it's a general purpose ingredient that keeps cropping up again and again in development.

(31:15) And in this instance, what happens is that in effect it has the effect of switching the cell types in the epithelial layer so that some can keep on growing while others stop. And so this is basically what happens. You have this layer of tissue, and if some of the cells start emitting Sonic Hedgehog protein, if by chance a little bulge develops in that tissue as it grows, then this concentrates the Sonic Hedgehog protein to a point where it can trigger this switch and stop some of the cells growing so that they just continue growing at the base of this unit

(31:54) and the rest of it just then gets sort of pushed up further and further into this sort of finger-like protrusion. And it's a self-amplifying process. The more it gets kind of constrained, the better a trap it is for the Sonic Hedgehog protein. So the real cause of these villi is therefore really a mechanical one.

(32:14) It involves changes to the rigidity of the epithelial layer. Here's another example. We generally have five fingers on each hand, right? But there's no gene that specifies that. Certainly, no gene that specifies this number five. The way our fingers are now thought to grow is that in the paddle-like bud of the developing limb in the embryo, there's a small set of, again, general purpose developmental proteins that interact with one another in a particular complex way, a way that was first talked about

(32:50) by the British mathematician Alan Turing in 1952. Turing showed how it's possible for a soup of reacting chemicals to spontaneously segregate into stripes of different composition, different concentration of the ingredients. And that's what seems to happen in the development of the fingers. That stripes, radiating stripes develop in this bud of the growing limb, and these concentrations, these stripes in turn trigger the growth of bone that becomes the finger bones.

(33:26) And the reason that there are five of them is that the stripes have a just an intrinsic width that depends on the properties of the proteins. And they happen to grow at just the stage where five of these stripes will fit within the embryonic limb bud. Now, the whole growth process is more complex than this.

(33:44) It always is in biology, but it seems that only a little tweak in the growth conditions or in the timing of it might be enough to generate more or fewer stripes. And that's possibly what we see in the way for ray-finned fish, they have more of these stripes that develop in their fins. So you can see here that genes are providing resources for the body plan, but that plan doesn't exist in any meaningful way within the genome itself.

(34:16) The key genes and proteins involved in both these processes and in many others are, they're just general purpose developmental proteins. As I say, they're not proteins for developing any particular body type. The story isn't really about them as such. It's about the cells and tissues of the developing organism being triggered to make those proteins in just the right time and place and sequence.

(34:44) And this shift in the location of true causes in biology, it's not just a metaphorical way that I'm using to talk about what's going on, it's something we can measure. When the primary causes of the behaviour of some complex system arise, not at the lowest levels as they do for clockwork with all the cogs having to fit in the right way, but that if they appear at higher levels of organisation, that's something that scientists call causal emergence.

(35:15) And we can see it all the time in our social structures. For example, in the way a company can still operate if some of the workers are off sick because there are usually ways that the others can kind of compensate for their absence. And we can see it in the flocking of birds, for example. So that if there's a kind of acentric or a tired bird somewhere in this flock that's doing something different, it doesn't cause chaos and confusion among the whole.

(35:39) The larger scale organisation is robust against any little disturbances at the lower levels like that. And there are ways of measuring the amount of causal emergence in complex systems. And when a team of scientists applied those methods to look at the mechanisms that are used by simple organisms like bacteria, so-called prokaryote, and at more complex so-called eukaryotic organisms like us, they could see a clear difference.

(36:08) We eukaryotes have more causal emergence, and I call this causal spreading. And it's spreading rather than just a shift in where the causation is happening because the cause really is spread across a range of levels. So there are still some traits, like diseases like cystic fibrosis, that can really be pinned to a single gene.

(36:32) There are some traits that are, in a meaningful way, caused by that gene. But most of our traits are caused at higher levels, above the genes, even though the genes can still influence them. Perhaps the ultimate expression of this causal spreading is the brain. In "The Selfish Gene", Richard Dawkins sounds almost affronted that our behaviour sometimes seems to go against what a selfish gene picture should make us expect.

(37:03) But I think this is precisely the point of a brain. You see, the challenges faced by bacteria and the decisions they have to make really aren't that diverse. You know, where's the food? Where's the moisture? How do I have to move to get there? They tend to live in single environments and they tend to die if they go outside them.

(37:23) But we get everywhere, and every day we face situations and challenges that we have never encountered before in quite that way in our lives, or that our ancestors have ever encountered before. So, no genetic programme is going to tell us what we should do in the face of every eventuality. And the responsibility for that decision must be passed up to some higher level to our brains, which don't have a kind of programme that is able to compute exactly what we should do in any given circumstance, but is able crucially to improvise,

(38:00) in the face of the unexpected to improvise using fuzzy rules of thumb, not some precise, digital computation. And this way of behaving, so not through totally automated and predictable machine-like stimulus response, but by genuine cognitive processing, this isn't just a good analogy for how life works at all levels, even down to the level of single cells.

(38:30) Some biologists argue that actually it literally is like that. That all living things should be genuinely considered cognitive systems. As the biologist Mike Levin and the philosopher Dan Dennett have put it, life is cognition all the way down. And this doesn't mean that the bacteria have some kind of mind worthy of the name, let alone any kind of awareness.

(38:53) Cognition doesn't have to require consciousness to be genuine cognition. Well, however you feel about this way of thinking about life, I think it does capture one important truth about the way life works and that is, that the best metaphors for talking about it aren't ones that come from technologies, clockwork, or computers, but are metaphors that are drawn from life itself.

(39:21) What really distinguishes living things from any machines that we've yet made is that they're not automata, but they have real agency. And what I mean by that is that they're able to manipulate and alter themselves and their environment in order to try to attain some self-determined goals. Now, when we recognise that organisms have goals, we usually sort of say, "Well, the goals for all organisms are to survive and reproduce.

(39:50) " Right, but while an awful lot of behaviour can of course be explained that way, I don't think it's enough. I'll hazard the guess that the goal you set yourself in coming here tonight wasn't about eating and reproducing. And if it was, I'm afraid you're probably gonna be disappointed, (audience laughs) although who am I to say? But I think that we're not the only animals in having agendas and purposes of our own that we decide that are not obviously linked to evolutionary imperatives

(40:24) and aren't wholly predictable. Even single celled organisms and single cells of our body set their agendas to some degree so that for example, what can seem like identical cells might behave in different ways to an identical stimulus because there is some internal setting that they have that determines that.

(40:46) If you like, they've sort of made up their own minds. They have their own goals. I think that displaying agency is actually a more fruitful and more general way to think about what living organisms are than to try to come up with some kind of checklist of, you know, what life is. Like say, reproduction, metabolism, homeostasis, and so on.

(41:08) That's why to have an overarching view of how life works, I think that biology needs an understanding and ideally, really, a theory of agency. What are the basic ingredients that would require? We don't really know, although this new book by the neuroscientist Kevin Mitchell makes a superb stab at starting that conversation and along the way shows how it is that what we call our own free will is really just an aspect of the kind of agency that we, as complex, cognitive, and conscious beings, possess.

(41:47) But I do think that we might be able to see some of the things that agents are probably going to require. For example, they need to make predictions about their environment so that they're not constantly wasting energy coping with things that they might have anticipated and avoided. And in order to do that, an agent needs some kind of memory in which it can build up and store information about its environment, which amounts in the end to a kind of representation of a crude model of its environment.

(42:22) And we all have that. You know, if you leave here tonight and you were going back to Green Park tube and you start heading north along Albemarle Street, then, you haven't really stored a good internal representation of your environment because it's the other way, it's south. So you've wasted energy, which in that case is not a matter of life and death, but sometimes making the right prediction could be.

(42:45) Okay, this new view of how life works is I think it particularly matters when life isn't working so well and when we want to put it right. It matters for medicine. This issue of causal spreading matters for medicine because if we want to affect some change to a system, we do best to intervene at the place where that outcome is caused.

(43:11) Is the cause of such and such a disease some gene that we should be targeting? You see, what we tend to hear about in discussions of gene-based medicine are the exceptional cases, where cause really is situated in a given identified gene. For example, in the recent announcements of the use of gene editing, gene therapy to treat sickle cell disease, which is primarily caused by mutations to a single gene so that we can, in principle, use genome editing to go into that gene and to put it right.

(43:44) But I think it's a fairly well kept secret that most of the regions of the genome that are found to be associated with most common diseases aren't even within genes at all. They're found within the non-coding regions that are presumably involved somehow in gene regulation. And it's often very hard to make effective interventions at that level anyway for one reason, because the genetic effects tend to involve lots of very tiny effects spread across lots of different regions of the genome.

(44:19) But I think ultimately, the reason why it's hard is because the real causes of these conditions operate at some higher level of organisation than genes. For example, in the functioning of the immune system. That's a common one. So we might still see associations of the condition with particular bits of the genome, but in a sense all that we are seeing there are weak echoes of the genuine causation sort of coming down from higher levels.

(44:47) That may be why genetic-based approaches to cancer treatments in particular have been so disappointing. You see, because cancers, we know that cancers can arise from mutations that can happen to our genes, perhaps just because we get old or because we've been exposed to something in the environment that causes them.

(45:07) And so it used to be thought that we could find cures by looking at the genetic roots of cancer. But it seems increasingly that the most effective levels of intervention are higher ones, and again, in particular, intervening in the immune system to help the body itself fight cancer. As the cancer biologist Michael Yaffe said in 2013, we spent fruitless years looking for cancelling genes, not because we ever really had any reason to think they were the key to developing new treatments, but because we have the techniques for looking for them.

(45:47) As Yaffe said, "Like data junkies we continue to look to genome sequencing when the really clinically useful information for cancer therapies may lie someplace else." This broader view of what governs the behaviour of our cells and tissues and bodies, I think also matters for understanding the kinds of things that they can make.

(46:12) For example, we've discovered in the past two decades that our bodies, that our cells can be tweaked to switch between different states so that cells in our bodies that have already developed into a mature tissue type, a particular tissue type can be turned back and switched to a stem cell-like state from which they can develop into any tissue type.

(46:34) And I've experienced this myself directly. I've had cells of my skin taken from my shoulder reprogrammed into that stem cell-like state and then developed into neurons that grew into structures a bit like this called brain organoids, which look a little bit, not just sort of in colour and size and so on, but actually in anatomy look a little bit like developing embryonic brains.

(47:01) And this, looks like a normal embryo. Actually, this is like a normal mouse embryo. They're made from mouse cells. But this is something a structure that has assembled spontaneously from stem cells. No egg or sperm was involved in creating this so-called embryo model. So what our cells can make is not prescribed or pre-ordained.

(47:27) I think it's better to think of them as being imbued with the potential to assemble into forms. And if we can understand what those higher level principles of assembly are, then who knows what new forms we might be able to make. But there's surely also good reason to understand these questions, to understand how life works simply to make us appreciate all the more how astonishing life is.

(47:57) The idea that what makes living matter alive and different from a rock is simply that it's being programmed to be alive is not just incomplete, but it's a bit boring. One of the sobering things about a gene-centered view of evolution and of life is that it turns out to make a puzzle of why organisms exist at all.

(48:21) Richard Dawkins has called this the paradox of the organism. Now, I don't know about you, but if my theory of to explain the profusion of wondrous life forms ended up implying that those life forms shouldn't really even exist, then I wouldn't so much call it a paradox as go back and think, "Where have I gone wrong?" There are ways to think about this paradox within the selfish gene view of life, and it's interesting to do that, but I think it's not too hard to see ultimately why this paradox arises.

(48:54) If you have taken all of the genuine agency that exists in real organisms and squeezed it into genes to make them look like little organisms in their own right, all existing and competing with each other in some in undifferentiated pool with an agency they don't possess, then it's not really any surprise that you no longer seem to need real agents at all.

(49:21) The idea that an account of life is to be found by this gradual increasing of the magnification until we arrive at molecules is a fantasy, and that's something that the Nobel laureate biochemist Albert Szent-Gyorgyi understood very well when he said this. He said, "My own scientific career was a descent from higher to lower dimension, led by a desire to understand life.

(49:45) I went from animals to cells, from cells to bacteria, from bacteria to molecules. On my way, life ran out between my fingers." What it this comes down to then is looking for an explanation of how life works that really does justice to the truly amazing nature of life itself. Saying that life is just a machine made by genes or a computer running a programme is not only wrong from the perspective of what modern biology is now telling us.

(50:18) I think it banishes life altogether from biology. It sterilises it. The eminent physicist Michael Berry told me recently that he was once asked the question, "What is the biggest unsolved problem in physics?" And he figured that the question, it was probably expecting him to come out with some standard answer like dark matter or quantum gravity.

(50:39) But he found himself saying that, "If, as we think, all matter is described by quantum mechanics, then where does the aliveness of living matter come from?" Now he didn't mean, thank God, that it must have some quantum explanation, but he meant that living matter is profoundly different from other kinds of matter and we don't really know why.

(51:05) Parts of the universe, all of these parts have become aware of themselves and their place within it. Biologists should never forget that what they're really trying to do is to understand and to explain that. No wonder, it's really hard. But I believe that we should refuse to accept too cheap an answer to this question, the most profound question in science.

(51:34) Thank you. (audience applauding)

It isn't an animal, a plant, or a fungus. The slime mold (Physarum polycephalum) is a strange, creeping, bloblike organism made up of one giant cell. Though it has no brain, it can learn from experience, as biologists at the Research Centre on Animal Cognition (CNRS, Université Toulouse III—Paul Sabatier) previously demonstrated. Now the same team of scientists has gone a step further, proving that a slime mold can transmit what it has learned to a fellow slime mold when the two combine. These new findings are published in the December 21, 2016, issue of the Proceedings of the Royal Society B.

Imagine you could temporarily fuse with someone, acquire that person's knowledge, and then split off to become your separate self again. With slime molds, that really happens! The slime mold—Physarum polycephalum for scientists—is a unicellular organism whose natural habitat is forest litter. But it can also be cultured in a laboratory petri dish. Audrey Dussutour and David Vogel had already trained slime molds to move past repellent but harmless substances (e.g. coffee, quinine, or salt) to reach their food.1 They now reveal that a slime mold that has learned to ignore salt can transmit this acquired behavior to another simply by fusing with it. 

To achieve this, the researchers taught more than 2,000 slime molds that salt posed no threat. In order to reach their food, these slime molds had to cross a bridge covered with salt. This experience made them habituated slime molds. Meanwhile, another 2,000 slime molds had to cross a bridge bare of any substance. They made up the group of naive slime molds. After this training period, the scientists grouped slime molds into habituated, naive, and mixed pairs. Paired slime molds fused together where they came into contact.2 The new, fused slime molds then had to cross salt-covered bridges. To the researchers' surprise, the mixed slime molds moved just as fast as habituated pairs, and much faster than naive ones, suggesting that knowledge of the harmless nature of salt had been shared. This held true for slime molds formed from 3 or 4 individuals. No matter how many fused, only 1 habituated slime mold was needed to transfer the information. 

To check that transfer had indeed taken place, the scientists separated the slime molds 1 hour and 3 hours after fusion and repeated the bridge experiment. Only naive slime molds that had been fused with habituated slime molds for 3 hours ignored the salt; all others were repulsed by it. This was proof of learning. When viewing the slime molds through a microscope, the scientists noticed that, after 3 hours, a vein formed at the point of fusion. This vein is undoubtedly the channel through which information is shared. The next challenges facing the researchers are to elucidate the form this information takes, and to test whether more than one behavior can be transmitted simultaneously. If Slime Mold A learns how to ignore quinine and Slime Mold B to ignore salt, the biologists wonder whether both behaviors can be transmitted and retained through fusion. 

Habituated (H) P. polycephalum individuals fusing with a naive (N) individual. The pseudopodium crossing the bridge belongs to the naive slime mold. A vein may be seen forming between two slime molds at their point of contact (see magnified region).

Bibliography

Direct transfer of learned behavior via cell fusion in non-neural organisms, David Vogel & Audrey Dussutour. Proceedings of the Royal Society B, 21 December 2016. View web site

Notes

1. See the following press release dated April 27, 2016: A single-celled organism capable of learning. http://www2.cnrs.fr/en/2751.htm.

2. Physarum polycephalum, a cell with thousands of nuclei that can cover an area on the order of 1 m2, has the natural ability to fragment when encountering an obstacle and to fuse with other slime molds.

Bacteria make up more than 10% of all living things but until recently we had little realization that, as in humans, soil bacteria have internal clocks that synchronize their activities with the 24-hour cycles of day and night on Earth.

New research shows just how complex and sophisticated these bacterial circadian clocks are, clearing the way for an exciting new phase of study….

An international collaboration from Ludwig Maximillian University Munich (LMU Munich), The John Innes Centre, The Technical University of Denmark, and Leiden University, made the discovery by probing gene expression as evidence of clock activity in the widespread soil bacterium Bacillus subtilis. [Emphasis added.]

“Pervasive” Clock Activity

The authors published a paper about this in Science Advances, announcing that the bacterial clock “evokes properties of complex, multicellular circadian systems.” The lead author, Francesca Sartor, noted that the clock activity is “pervasive” in this tiny microbe. It regulates multiple genes and behaviors.

Professor Antony Dodd from the John Innes Centre added, “It is astonishing that a unicellular organism with such a small genome has a circadian clock with some properties that evoke clocks in more complex organisms.”

Moreover, the researchers believe that clocks are widespread in bacteria. What happened to the notion of simple to complex evolution by gradual steps? Would a “blind watchmaker” start with a Rolex?

Professor Ákos T. Kovács, from Leiden University and Technical University of Denmark said… “it is amazing that the circadian clock in Bacillus subtilis — a bacterium with just four thousand genes — has a complex circadian system that is reminiscent of circadian clocks in complex organisms such as flies, mammals, and plants”.

“Just four thousand genes” sounds flippant. Try counting to four thousand out loud; it will take over two hours at two seconds per integer. As you count, think of a molecular machine, regulatory element, or purposeful role represented by every one of those digits. Each bacterial gene, moreover, is composed of 900 base pairs on average. That’s a lot of functional information packed into an organism a micron in diameter. Even so, evolutionary biologists did not expect to find circadian clocks in bacteria that match the functional sophistication of those in flies, mammals, and plants.

Are Genes Blind Watchmakers?

Audrey Mat, a marine biologist at the University of Vienna, says that genes are “The Great Clockmakers.” Writing in The Conversation, she gives the ho-hum response to the existence of timekeepers in living organisms. “The rotations of the Earth, Moon and Sun generate environmental cycles that have favoured the selection of biological clocks.” Under this reasoning, pressure waves favor the selection of ears. Photons favor the selection of eyes. Planet rotations and orbits favor the selection of clocks. Environments can favor things all they can, but complex sensors to detect and use them do not logically follow.

The circadian clock mechanism was first discovered in the fruit fly, also known as Drosophila, in the 1970s. It is based on feedback loops in the transcription and translation of several genes — gene A promotes the expression of gene B, which in turn inhibits the expression of gene A — creating an oscillation. During the day, light induces the diminution of specific factors of the loop via a photoreceptor called cryptochrome. Interestingly, the key factors in the mechanism essentially only comprise a few genes named period, timeless, clock and cycle. However, the fine-tuning and regulation of the clock is based on a complex molecular and neuronal network that ensures its timing and precision.

According to Mat, physical forces not only drive the emergence of devices to sense them; they also tune them and maintain them. They even adjust their responses to the changing seasons. How does Darwinism explain this? It doesn’t:

The circadian clock is not the only clock mechanism that exists in nature. Many biological processes are seasonal, such as the migration of a host of birds and insects, the reproduction and hibernation of many animal species and the flowering of plants. This seasonality is generally dictated by several factors, including by what is known as a circannual clock in the case of many species. The mechanism of this clock has not yet been determined.

Can Clocks Be Darwinized?

The paper in Science Advances makes no claim for Darwinism, either. The authors put evolutionary explanations in future tense: 

Discovering mechanisms by which this memory of entrainment conditions during development of a circadian system occurs, in diverse systems, will inform on convergent and divergent evolutionary processes.

That’s all they say about evolution. Don’t hold your breath, though, for answers. Faced with complex functional timekeeping in the most primitive organisms, evolutionary biologists have their storytelling work cut out for them.

Circadian clocks are pervasive throughout nature, yet only recently has this adaptive regulatory program been described in nonphotosynthetic bacteria. Here, we describe an inherent complexity in the Bacillus subtilis circadian clock…. We report that circadian rhythms occur in wild isolates of this prokaryote, thus establishing them as a general property of this species, and that its circadian system responds to the environment in a complex fashion that is consistent with multicellular eukaryotic circadian systems.

The complex abilities of the bacterial species included entrainment, or the following of cues. Like catching a train, entrainment requires sensing environmental cues, called zeitgebers, and getting on board to go somewhere on purpose. This also presupposes a memory of the cues. 

Starting Expectations and Startling Conclusions

One doesn’t always see “surprised” in a stodgy scientific paper, but the word stood out in this one:

Entrainment leads to the establishment of a stable phase relationship between the external (environmental) and the internal (circadian) time. Circadian systems use zeitgebers for entrainment, leading to a set of remarkable phenomena. We were surprised to observe that a prokaryote challenged with chronobiological protocols exhibits a variety of highly complex entrainment properties…. The presence of aftereffects (see table S1) suggests that information regarding zeitgeber exposure is stored, much like a memory.

They didn’t expect this. “It would be naïve to assume that a prokaryotic circadian clock shares these properties with multicellular organisms,” they initially thought, but the observations proved otherwise. Using red and blue light as zeitgebers, and watching responses with fluorescent cues, they were able to entrain the microbes and alter their behaviors by modifying the free-running period (FRP) of the light. The results demonstrated that “this organism shares many circadian characteristics occurring in eukaryotic organisms, some of which have yet to be documented in established clock models in cyanobacteria or fungi.” 

Our observations also underscore that a combination of zeitgebers is used by B. subtilis, which is analogous to the situation for fungal, mammalian, and plant cells. The task of the circadian clock is to “read” the local environment and, for many systems, this means harvesting not just one but many cues. We suggest that by using both blue and red light and temperature as zeitgebers, B. subtilis can fine-tune clock-regulated processes to a greater range of situations.

For this to be true of tiny microbes that live in the soil is indeed surprising. How do they do it without eyes? The “light-sensing mechanisms used by B. subtilis for the purpose of entrainment remain unknown.” Perhaps the microbes respond to the energy levels of different wavelengths of light penetrating the soil. Whatever is involved in the bacterium’s clock led to a second use of the word “remarkable” in the conclusion:

In conclusion, we find it remarkable that a relatively simple prokaryote, which lacks the obvious hierarchy of organization of multicellular organisms, evokes properties of complex circadian systems.

Design advocates would certainly find it remarkable, too. But surprising? For those committed to explaining biology by unguided material causes, surprise is understandable. Those who recognize the hand behind the superb engineering all around us in life are delighted but not surprised.

Does death really mean the end of our existence? Great thinkers from Plato to Blue Öyster Cult have weighed in on the question. Now, a study shows that that at least one aspect of life continues: Genes remain turned on days after animals die. Researchers may be able to parlay this postmortem activity into better ways of preserving donated organs for transplantation and more accurate methods of determining when murder victims were killed. 

Before you ask, microbiologist Peter Noble of the University of Washington, Seattle, and colleagues were not trying to find out what allows zombies to stalk Earth and slurp the brains of the unwary. Instead, the scientists wanted to test a new method they had developed for calibrating gene activity measurements. Their research had already taken a morbid turn—2 years ago they published a paper on the abundance of microbes in different human organs after death—and they decided to apply their method to postmortem samples. “It’s an experiment of curiosity to see what happens when you die,” Noble says.

Although scientists analyzing blood and liver tissue from human cadavers had previously noted the postmortem activity of a few genes, Noble and colleagues systematically evaluated more than 1000. The team measured which of these genes were functioning in tissues from recently deceased mice and zebrafish, tracking changes for 4 days in the fish and 2 days in the rodents.

At first, the researchers assumed that genes would shut down shortly after death, like the parts of a car that has run out of gas. What they found instead was that hundreds of genes ramped up. Although most of these genes upped their activity in the first 24 hours after the animals expired and then tapered off, in the fish some genes remained active 4 days after death. 

Many of these postmortem genes are beneficial in emergencies; they perform tasks such as spurring inflammation, firing up the immune system, and counteracting stress. Other genes were more surprising. “What’s jaw-dropping is that developmental genes are turned on after death,” Noble says. These genes normally help sculpt the embryo, but they aren’t needed after birth. One possible explanation for their postmortem reawakening, the researchers say, is that cellular conditions in newly dead corpses resemble those in embryos. The team also found that several genes that promote cancer became more active. That result could explain why people who receive transplants from the recently deceased have a higher risk of cancer, Noble says. He and his colleagues posted their results on the preprint server bioRxiv last week, and Noble says their paper is undergoing peer review at a journal.

 “This is a rare study,” says molecular pharmacologist Ashim Malhotra of Pacific University, Hillsboro, in Oregon, who wasn’t connected to the research. “It is important to understand what happens to organs after a person dies, especially if we are going to transplant them.” The team’s approach for measuring gene activity could be “used as a diagnostic tool for predicting the quality of a transplant.”

In an accompanying paper on bioRxiv, Noble and two colleagues demonstrated another possible use for gene activity measurements, showing that they can provide accurate estimates of the time of death. Those results impress forensic scientist David Carter of Chaminade University of Honolulu. Although making a time of death estimate is crucial for many criminal investigations, “we are not very good at it,” he says. Such estimates often rely on evidence that isn’t directly connected to the body, such as the last calls or texts on the victim’s cellphone. Noble and his colleagues, Carter says, have “established a technique that has a great deal of potential to help death investigation.”

A mouse or zebrafish doesn’t benefit, no matter which genes turn on after its death. The patterns of gene activity that the researchers observed may represent what happens when the complex network of interacting genes that normally keeps an organism functioning unwinds. Some genes may turn on, for example, because other genes that normally help kept them silent have shut off. By following these changes, researchers might be able to learn more about how these networks evolved, Noble says. “The headline of this study is that we can probably get a lot of information about life by studying death.”