Enrico Coen: designing flowers and tracking development

Today's Titans of Science features Enrico Coen in conversation with Chris Smith. We pick up the conversation with them talking about repetitive sequences...

Enrico - A lot of our DNA is made up of sequences that move around, so-called repetitive sequences. There are multiple copies of them. So yes, they're ubiquitous. They're found in all sorts of organisms. Humans are no exception. As I said earlier, the rules of life don't treat humans special in a special way and we have transposed on. So do mice, so do bacteria. Pretty much every organism you can think of will carry these elements. But they were first discovered, again, in this bizarre way by Barbara McClintock being curious about what was causing these spots on these kernels of corn. That to me is amazing that so many really fundamental discoveries are made in this bizarre, almost sort of idiosyncratic way. And it turns out to be just by following your curiosity, you uncover something that's radical.

Chris - So what were you doing with them, with the antirrhinums?

Enrico - Well, initially, I was trying to understand how these pieces of DNA moved about and then I started to get interested in using these pieces of DNA to identify genes. So if you take your bus analogy, if it causes a bus crash, then it helps you identify the bus. This is one way of thinking about it. So these pieces of DNA that move around, once you have got some way of identifying the piece of DNA that moves, when it lands next to a gene that's interesting, such as your bus crash, and you say, oh, I'm really interested in the bus because the bus now has disappeared. Because this piece of DNA has landed in this position, it kind of labels the bus for you in a way. It sort of carries with it a kind of tag. And that means that you can use that to identify the bus gene. So that's what we started to do. We started to use this piece of DNA that was moving around to identify the genes it was landing next to. So if we go back to the idea of the book and a sentence is moving about, if that sentence lands next to a very interesting sentence, it creates a change your attention is drawn to. And that's what we started to do. We started to identify genes using this method, using these pieces of DNA as tags to find particular genes.

Chris - And what genes jumped out, if you excuse that pun?

Enrico - I was interested already at the time in these genetic switches. We were talking earlier about the switches between different types of primrose, the switches between male and female. Another type of switch is the switch between different types of repeated structures. For example, your arms and your legs, they're very similar in some ways, jointed and they have digits at one end, but they're also different. An arm and a leg are distinct. And so this genetic switch that determines whether you make an arm or a leg. Similarly, if you look at a fruit fly, you'll see there are different segments in the fly. Some have wings, some have legs. So again, it's almost variations on a theme in a fruit fly and in any insect for that matter. In plants, we also find variations on a theme. And the most striking example is the flower. So the flower is the reproductive part of the plant, the sexual part of the plant. And it has various types of organs. It has the male organs, they're the stamens. They produce the pollen, which generates the sperm cells. And then there are female organs called the carpels, that they house the eggs that are then fertilised to produce the seed. And then there are the petals. Petals are the attractive organs usually. And they serve to advertise the flower, to attract pollinators. And so they have these often very showy colours. And then finally, on the outside of the flower, there's another set of organs called the sepals or sepals. And they enclose the bud. So a flower has these four types of organ, sepals, petals, stamens and carpels. And they're arranged in this concentric arrangement. And the question is, how are these different types of organs being produced? What are the genetic switches that control this? And we didn't have any idea of how that was done back then. We're talking again, the 80s, late 80s. And so we started to identify mutations using these pieces of DNA that moved around. We look for mutations that changed these types of organs so that you would end up, for example, instead of stamens, you would end up with petals. Or instead of sepals, you would end up with carpels. These mutations, by the way, we see all the time in gardens. So on Valentine's Day, if you get a rose, it won't contain stamens. It's a mutant. It's a mutant in which the male organs have been replaced by petals. So, and also the female organs. So a rose, a wild rose has stamens. It has carpels. But we've selected mutants which have lots and lots of petals because we find those attractive. The plant, of course, can't reproduce normally. It can reproduce vegetatively. You can propagate these roses. But they can't produce sexually because they've lost their sex organs. But we find those attractive. So every time you give somebody a flower on Valentine's Day, what you're really doing is giving them a mutant.

Chris - It doesn't have quite the same romantic sound to it, does it?

Enrico - I find it romantic, but I love mutants.

Chris - So did this enable you then to track down the genes that do that so that we're in a position to now understand a lot more about the genetic machinery that is guiding how flowers develop?

Enrico - Exactly. That's exactly what we could do. By looking for these weird, weird flowers where one type of organ has been replaced by another, we could identify the genes. We're using this transposable, this piece of DNA that moved. We could find the genes and not just our group, there are other groups as well that also identified these genes and started to come up with a model for how the combination of these genes or the activity of these genes was controlling the different types of switches, making different types of organs in the flower. So it was an example in the plant world of a fundamental phenomenon, which is how genes switch between different types of related structures.

Chris - Are the roses that you use, rose as an example, are roses and antirrhinums, are they analogous in the sense that the gene that does that job in a rose, you've got an equivalent gene in an antirrhinum. You haven't got to go completely hunting from scratch in a different species of plant.

Enrico - Yes, these genes turned out to be highly conserved. One of the most exciting things was that while we were studying this phenomenon in snapdragons, in antirrhinums, another group, Elliot Meyerowitz and his colleagues, was studying in the United States. They were studying a similar phenomenon in what's called Arabidopsis, a weed, a very distantly related weed to snapdragon. And they came up with essentially the same idea and it turned out to be the same genes. So two very different plants. It's a bit like comparing, say, a human with a frog or a fish. You're taking plants that are very distantly related. You find the same genes are involved in those switches. And so, not surprisingly, roses also have those, turned out to have those similar genes. I mean, each species maybe has a slight tweak, but fundamentally those genes are conserved right across the board among all the flowering plants. And one of the, presumably the origin of flowering plants, relates to deploying these different types of genes.

Chris - Did I read somewhere that you've branched out since into carnivorous plants? They've got to be one of my favourite, in fact, the world's favourite types of plants. Is that true?

Enrico - Yes, yes. We got very interested in the formation of the trap, the carnivorous traps. So one of my interests in the flower, as we started to understand how these genes were controlling these switches, we then started to think about, well, how do these switches actually produce an organ of a particular shape? In a sense, you can think of it like in the 1980s, 90s, people were very interested in these switches, which you can think of like, how do you create a pattern or like a painting, different patterns of regions. If you look at a painting, it has different regions of colour. And these regions of colour are produced by putting paint in certain places. In the same way, what we were discovering through these switches is that as an organism develops, whether it's a human, a mouse or a snapdragon, genes were coming on and off in different places, just like a picture having different colours, genes were coming on and off in different places. But the question was, how did the activity of these genes lead to a particular shape? It's the difference between painting and sculpture. So with painting, you have a pattern. With sculpture, you have a three-dimensional shape. And that started to interest me, how you move between one situation and the other. How does the pattern influence the shape? There's a nice example from Alan Turing. Alan Turing, the guy who was key in terms of inventing computation, he was very interested in this phenomenon, in this problem. And he came up with what's still a very fundamental idea for how this patterning process, the painting process happens. But somebody asked him about this and he said, and you could use this patterning to explain things like the stripes on a zebra. And he said about the zebra, the stripes is the easy part, the hard part is explaining the horse. In other words, the pattern was easy, relatively, although even that isn't so easy. The question is, how do you produce the shape of a zebra, this horse shape? And that's what we became interested in. And that's why we looked at carnivorous plants, because they have these amazing shapes. They have these cups that essentially trap animals.

Chris - And what you're interested in thinking, how does it deploy its genetic arsenal to produce that particular configuration?

Enrico - Exactly. Not only that, this configuration has been produced four times independently in evolution. So carnivorous plants, there are four independently evolved cases where these cup-like leaves have been produced. Saracenia, for example, is one type. Another type, which was the one that we worked on, was eutricularia. Eutricularia has these tiny little cups. It's an aquatic plant. And it has these little trap doors. And if a little crustacean happens to wander by, it triggers one of these traps. It gets sucked in to one of these tiny little cups and then gets digested. So a much bigger version of that is Nepenthes, the pitcher plant. That's a much bigger example. But this is tiny. And we were interested in using that as a system to try and understand, how do genes produce this form? How are genes leading to the production of this amazing cup shape? So all these cups are derived from leaves. They're kind of modified leaves. It's as if somehow somebody has taken a leaf and decided to shape it in a completely new way to make a cup. But of course, that's not the way that evolution works. It doesn't take a leaf and then, as it were, cut it up into little pieces to make a cup. These cups are starting off cup-shaped from a very early stage of the development of these structures. So they start off, when you're talking about a fraction of a millimetre big, they're already starting to form this cup shape.

Chris - Is this where the computer modelling comes in? Because once you get to a situation where we know what the genes are, but we know that we've got enormous numbers of interactions and degrees of freedom and permutations to consider, it becomes very, very difficult to think of this as a human. And you begin to need some way of seeing how this can evolve on a computer in order to test, is this how it could be happening in nature?

Enrico - Sometimes computer modelling helps you test ideas when it's too complicated to work things out intuitively. And that was the challenge. How do you figure out that process? And since most of the things that we're familiar with do not grow, we had to develop methods that would allow us to simulate how growing material would behave. OK, that kind of was the challenge. We had to say, right, we wanted to create, in a sense, a sandpit where we could say, well, now in this sandpit, if you do something to this material, it's going to start to grow. And if you do something else to it, it'll grow more in this way or that way. And then we can test ideas out and see, could this be one way that the carnivorous leaf, these cups, could have formed? And it was really a way of us exploring different hypotheses and seeing what would work in principle. And then once you've done that, you can then say, can we experimentally test that hypothesis and see if it's correct?

Chris - Where does all this leave us then? Because you began by looking at primroses, decided you didn't like them because it was a hard problem, and it needed 30 years more science, to your credit, to solve it, which it has been solved since. Then antirrhinums and discovering a lot of the genes that help things look the way they do, and then carnivorous plants, similar sort of problem. Where are we now? Are we at a position where we have a really thorough understanding so we can almost do for plants what Build-A-Bear does for teddy bears? You can sort of almost design a plant from the ground up now.

Enrico - I don't think we're there yet. Another feature of science is when you answer one question, you know, 10 more sprout. And so although we have some understanding, certainly much more understanding than we did have when I was beginning these questions, we're still in the dark in all sorts of areas. And this is not just true, incidentally, of plants. It's also true of humans or animals. We still don't understand the principles of how this growth is actually being controlled. So what you were talking about, the designer, are we in a position to make designer organisms? The answer is no. We can certainly have mutations that change organisms. We know about that. But if you were to say to design something new, we're very, very far from that because we still don't understand a lot of the basic principles by which these things are put together. If I may give an example, because it may be hard for people to think, well, why? Surely, surely we've got smartphones. Surely we've got the internet. We've got AI. How can this be a problem? Well, let me give an example. Suppose you went to a shop and you said, I'd like a new smartphone, please. And they give you a tiny little packet. And in this packet, there's a tiny little speck. You take the speck home. You put it into a pot. You water the pot. And gradually, this smartphone emerges. OK, now you'd say, wow, that is incredible. How could that happen? That's a miracle. How did this tiny little speck grow, turn itself into a smartphone? Well, that's what plants and animals are doing all the time. But we're so sort of used to it that we take it for granted. We don't see the miracle. We just think, oh, well, yeah, of course, you put a seed into some soil. It turns into a plant. Yes, a man and a woman reproduce. They have a baby. Yeah, that's what happens. But it's incredible. It's totally incredible. And mind-boggling that this can happen at all. And we don't understand how. Many of the aspects of how that works. How something can kind of construct itself. See, the smartphones that we make, for all their beauty and technology, we assemble them. We make them. There are, in a sense, hands on the outside that are constructing these things for you. A seed doesn't have that. You started off with a tenth of a millimetre sphere. There was nothing that was going. It's all in that sphere. That tiny package of information is enough to somehow assemble itself and turn itself into a human being having a conversation with me now. So that's just incredible. And its familiarity, in a sense, breeds contempt. We're so familiar with it that we kind of forget how amazing it is.