Ep 104: Sleeping Beauties: The mystery of dormant innovations in nature and culture (with Andreas Wagner)
Where does biological innovation come from? Why do some innovations wait millions of years for their spotlight?
Life must constantly innovate for evolution to occur, but many forms of biological innovation often lie dormant, sometimes for millions of years. In this episode, we speak to Andreas Wagner about his recent book, Sleeping Beauties: The Mystery of Dormant Innovations in Nature and Culture. Andreas is a professor at the Institute of Evolutionary Biology and Environmental Studies at the University of Zurich in Switzerland. In the book, Andreas explains how novel traits sometimes have to wait until the environment changes to become useful, leading to speciation or offering novel solutions to ecological problems. These long fuses are also evident in our own history, namely the life-changing technologies that we invent but don’t fully exploit until the right social or economic context arises.
📚🐛 Big Biology Bookshelf
Cover photo: Keating Shahmehri
-
SPEAKERS
Andreas Wagner, Ruth Demree, Art Woods, Dayna De La Cruz, Kyle Smith, Marty Martin
Ruth Demree 00:05
Hi everyone. My name is Ruth, and I’m the producer of the Big Biology podcast. I’m here with two of our interns, Dayna and Kyle.
Kyle Smith 00:12
This is the last episode of season 5. Before we start the show today, we have a special request for you to help us continue sharing the biggest and most exciting ideas in biology.
Dayna De La Cruz 00:21
As an independent podcast, we rely on your support to - run the show, to pay our interns and artists, and to help us reach even more listeners.
Ruth Demree 00:29
We know that times might be tough and many of you are students, and we are so grateful for all that you have done for us over the years. Any contribution, big or small, has had a huge impact on our ability to deliver the episodes that you love.
Kyle Smith 00:41
And, as we gear up for season 6, we would like to ask for your continued support. Whether it’s a one-time donation at www.bigbiology.org or a recurring monthly donation through Patreon at patreon.com/bigbio,
Dayna De La Cruz 00:55
Your support will help us continue to share the best of biology for years to come.
Kyle Smith 01:00
Thank you for listening to our show. I'm Kyle Smith.
Dayna De La Cruz 01:03
I'm Dayna De La Cruz.
Ruth Demree 01:04
And I'm Ruth Demree. Here's the show.
Marty Martin 01:14
Hey Art?
Art Woods 01:15
Yes?
Marty Martin 01:16
Good news.
Art Woods 01:17
Hit me with it.
Marty Martin 01:18
Today we're talking about the very long fuse that some ideas and some biological lineages need before they blow up into something important.
Art Woods 01:27
That sounds interesting Marty but why frame this as good news?
Marty Martin 01:30
Because there's hope that your scientific output will eventually catch on, even if just posthumously.
Art Woods 01:38
Good one but your attempts at comedy are as always already post-humorous.
Marty Martin 01:43
Haha. Where do we go from here?
Art Woods 01:45
How about an example?
Marty Martin 01:46
No problem. In fact, the history of human culture is replete with examples of artists and thinkers whose work and ideas went nowhere during their lifetimes, but came to great prominence later, sometimes much later.
Art Woods 01:58
Take Johannes Vermeer, the 17th century Dutch painter now known as ‘the master of light.’ His sublime paintings include the now ultra-famous Girl With A Pearl Earring, which is housed in a museum in The Hague. A lesser known Vermeer, which was the last one to come onto the market, in 2004, sold for about $30 million.
Marty Martin 02:17
But during his lifetime, Vermeer painted few works just 34 to 36 are known today, and he gained only local recognition in Delft.
Art Woods 02:26
And due to this anemic output, lack of recognition, and series of economic shocks arising from invasions of The Netherlands by France and England and Germany, Vermeer and his family slid into crushing debt. He died a few years later, in 1675, which his wife attributed to stress from the debts.
Marty Martin 02:45
Vermeer could easily have slid into permanent obscurity. But his works were surveyed in a published catalog by a French art critic in the mid 1860s – nearly 200 years after his death! – and this exposure alerted the rest of the art world to the masterpieces. From there, his stature has grown into the worldwide acclaim that he has today.
Art Woods 03:03
So, why the long fuse?
Marty Martin 03:05
We talk about long fuses with today’s guest, Andreas Wagner, in the context of his recently published book titled Sleeping Beauties: The Mystery of Dormant Innovations in Nature and Culture. Andreas is a professor at the Institute of Evolutionary Biology and Environmental Studies at the University of Zurich in Switzerland.
Art Woods 03:20
In the book, Andreas proposes a kind of unified vision for long fuses,
Marty Martin 03:25
Or sleeping beauties, choose your metaphor,
Art Woods 03:27
In two vastly interesting domains, the evolutionary diversification of different taxonomic groups,
Marty Martin 03:32
And the origins and spread of new scientific and technological ideas in human societies.
Art Woods 03:38
Although these two domains obviously differ profoundly in timescales and underlying dynamics, Andreas identifies two key processes that operate in both.
Marty Martin 03:46
The first is the enormous and ongoing production of novelty occurring in the background.
Art Woods 03:51
In biology, this means that individual organisms, their populations and species, and the larger taxonomic groups that contain them are producing novel traits all the time.
Marty Martin 04:00
One process is the promiscuity of individual enzymes in relation to possible food substrates and external chemical threats.
Art Woods 04:07
It turns out, for example, that even bacteria from the microbiomes of uncontacted hunter gatherer groups contain at least some latent capacity to resist modern antibiotics, even though they have no evolutionary history with them.
Marty Martin 04:19
In human culture, this means that particular ideas and technologies often arise many times in many places, but sometimes they just take hold locally.
Art Woods 04:27
In other words, there's enormous background creativity, much of which disappears before the ideas become widespread. We of course, don't remember the folks who came up with these ideas.
Marty Martin 04:36
The second process is radical change in the context that suddenly makes the novelty successful.
Art Woods 04:41
In biology, this means that particular lineages may evolve novel traits that lay dormant for long periods of time before suddenly becoming useful.
Marty Martin 04:49
Here we talk about grasses, which seem to have evolved key traits tens of millions of years before the lineages had their viral moment and underwent massive expansion and diversification.
Art Woods 04:59
Likewise in human culture, good ideas in the arts and politics and science often arises long before they come to be adopted widely.
Marty Martin 05:07
Think of Mendel's inferences about genetics from his work on peas. These were ignored for at least 30 years after his initial publication.
Art Woods 05:13
And often those ideas arise in multiple places from multiple individuals or groups
Marty Martin 05:18
Think of the origins of the wheel, or of agriculture, each of which occurred at least 10 different times in 10 different places in the world before really catching on.
Art Woods 05:26
What leads ultimately to the success of these ideas? It's clearly something about the cultural context in which they're occurring. Something changes so that finally other people and society at large are ready to accept and leverage them.
Marty Martin 05:38
Toward the end of today's chat, we talk about how a key Sleeping Beauty in our biology may be driving sleeping beauties in our culture.
Art Woods 05:44
That conversation centers on metaphors, the unsung linchpin of so many intellectual and cultural discoveries.
Marty Martin 05:51
We talk with Andreas about how our brains have evolved to be metaphor machines, how modern cultural evolution has shifted the context so that this ancient human ability takes on profoundly new importance.
Art Woods 06:02
It's our own long fuse, our sleeping beauty that now drives such rapid cultural and technological diversification.
Marty Martin 06:08
I'm Marty Martin.
Art Woods 06:09
And I'm Art Woods.
Marty Martin 06:10
And this is Big Biology.
Art Woods 06:23
Thanks so much for joining us on Big Biology to talk about your new book, sleeping beauties, which was published this year by one world press, we just read the book over the last couple of weeks, thought it was a beautiful exposition of how often sort of very important discoveries both in in biology and in culture, and how you know, human culture has developed over the past 1000s to hundreds of 1000s of years, how those discoveries lay dormant for long periods of time before suddenly bursting forth. And we want to just sort of step through the book and talk about some of the key examples and really dig into some of the mechanisms that underlie these long fuses that happen both in evolution and then culture. And we thought, you know, a good place to start would be with some of the macro evolutionary examples that you describe early on in the book, and especially this spectacular radiation of the world's grasses. So grasses dominate temperate grasslands all over the world. And what's amazing is how recently in evolutionary time that diversification has happened, but how old grasses are, so can you just sort of lay out the grass issue for us?
Andreas Wagner 07:34
Yeah, so you know, most people wouldn't think of grasses as spectacularly successful organisms, but you know, they really are. And by at least two standards, first of all, as you just said, they cover large areas on most continents, you know, millions and millions of square miles. So they're extremely widespread. And they also are extremely diverse in terms of the numbers of species, they have radiated into about 10,000 or so. Grasses are also phylogenetically, that is, evolutionarily very old, they originated in the during the age of the dinosaurs. And we know this, because we can actually find signatures of grass pollen, for example, in fossilized dinosaur dung. So we know that grasses are at least 65 million years old. But it turns out when they first originated, during the age of the dinosaurs, they were actually not very successful, that is to basically barely eked out a living at the margins of the biosphere they were not widespread geographically. And they did not radiate into a lot of species, at least initially. And in fact, it turns out, they had to wait for some 40 million years before those two developments happened, this radiation and the spreading of the grasses. So they are I think, a periodicmatic example, although perhaps not very well known outside circles of specialists of what paleontologists have called macro evolutionary lags. That is a long delay between the origin of a group of species or of taxon to their eventual success, either in terms of becoming widespread geographically or inter radiate, radiating into many species.
Art Woods 09:16
Yeah. So for grasses, what what was the context, the evolutionary change that allowed the sudden radiation to happen?
Andreas Wagner 09:23
There was actually no evolutionary change. And I think that's the key point, there was nothing that as far as we know, that happened inside the grasses themselves. It was actually something that had to do with a change in the environment. That is, at the time the grasses originated the Earth, the planet Earth was much wetter than it is today. But about 40 million years later, it dried up. And when it dried up, there were several innovations that the grasses had since their origins, or shortly thereafter that helped them survive a drier planet.
Marty Martin 09:57
So let's let's talk a little bit more about these macro evolutionary lags, the first few chapters of your book have other examples besides the grasses in a sort of long period of time, from the original origin of the group, to the point that diversification happens. Is there any consistency? What what do we know about the patterns? And macroevolutionary lags? Do they tend to be about the same duration for most lineages? I imagine the answer is no. But the duration of the lags in general, is there anything about those lineages that sort of retrospectively, we can see as pattern? Big things evolve slower, you know, small things, sizes of genomes and things like that?
Andreas Wagner 10:34
It's a great question. I don't think we know the answer at all. I think that would be a very interesting research program. And, you know, as a matter of fact, you know, this is not really a hot research area in paleontology, or in evolutionary biology in general. So there is not really the hundreds of papers that you would actually need to, to characterize some sort of pattern. So I, you know, for all we know, there are no rules, there are no patterns, but that may just be because nobody has looked for them.
Marty Martin 11:00
Yeah. Okay. So the conclusion. I mean, I think the reason that you entitled the book Sleeping Beauty, is that there's this idea that meaningful variation in a macro evolutionary sense in terms of diversification of lineages, and then in a minute, we're going to talk about sort of the experimental bacterial evolution work that you've done, which becomes more micro evolutionary, and the origins of genes and those kinds of things. But the basic idea is that these biological phenomenon, you don't see flourishing or diversification until context changes, until suddenly conditions are right, like you said, for the grasses, the drying out of the habitat. So is another way of saying that, that there's a massive contingency in the biological past. And another way of asking that question, Should we hope for a massive radiation of coelacanths, some 30 million years from now?
Andreas Wagner 11:53
About the coelacanths? I don't know.
Art Woods 11:57
One can only hope.
Andreas Wagner 11:58
Exactly I would have just made a stand here and said, yes, of course, you know, this will happen, right?
Marty Martin 12:03
35 million years, two weeks? And yeah.
Andreas Wagner 12:05
As we all know, a lot. You know, a lot of evolutionary biologists talk about predicting evolution, but usually predicting evolution is something you do only in hindsight,
Art Woods 12:14
Ironically enough.
Andreas Wagner 12:18
I'm sorry now I forgot the first question.
Marty Martin 12:20
So is, is the sleeping beauties idea another way of saying that? Is it evolution is contingent life on Earth? There's a lot of contingency in what's around now.
Andreas Wagner 12:29
It's a particular kind of contingency yes, I would, I would say. It's a process where the contingency comes not from something inside the organism, because those kinds of contingencies people have studied too, but it's contingency that comes from outside the organism, right? External versus internal, right? So internal contingencies are a, you know, an organism has to wait for the right kind of mutation to happen that brings forth a new feature that allows it to be successful, but these are different kinds of contingencies. They are, they're contingencies that happen in the environment, they are not within the control, so to speak of the genome.
Marty Martin 13:03
So I mean, just to drive this home, because I have, I had a different read. And maybe this is where I'm getting intermixed earlier chapters with the later chapters of the book, it's not so much that there is a latent potential as a trait of that evolutionary lineage. Well, I guess I'm gonna let me say that differently, I'm having a hard time understanding how it's not a latent potential of the organism, because there's something that's getting released when the environment changes, but you want to represent that as it's not. It's not something that's internal to that lineage. It's an externality.
Andreas Wagner 13:38
All I'm saying is that, you know, the trigger, of becoming successful is external. But there's already a latent potential for that success inside the organism that, for example, in the case of grasses, right, regards a number of adaptations that made grasses rather drought resistant, including, you know, innovations like C4 photosynthesis, that, you know, are sort of water saving innovations, if you will, right. So a special kind of photosynthesis, that helps grasses survive in the dry world. And, yeah, those were in place for a long time. But they perhaps didn't matter a whole lot in a world that was really quite wet, right? They really started to make a difference when the world you know, became dry, just as we'll probably talk about later, we know of a lot of bacteria that are antibiotic resistant, without ever having encountered antibiotics, right. So that's a latent potential that starts to matter once those bacteria proliferate in a patient that gets into a clinic where doctors apply antibiotics.
Art Woods 14:41
So here's another thought experiment. If we could go take a time machine back 30 million years and look at grass lineages pre explosion, is there any way to predict that they have these capacities for rapid evolution and diversification in the face of climate change? Is there? Is there some some characteristic we could define about them that would allow us to do that?
Andreas Wagner 15:04
Well, again, hindsight is 2020. Right. And I think you've perhaps guessed from my previous remarks that I'm rather skeptical about our ability to predict much about evolution, at least the interesting stuff, you know, that the novelties, the new features that have come about in evolution. And so here too, I think, a we know very little, especially about this whole, you know, complex with the grasses and all the traits, and when exactly some of those originated. And I suspect that even if we knew, we could only predict these kinds of things in hindsight.
Art Woods 15:49
Let's move on to talking about bacterial metabolism, and then origins of new genes. So and so in the book, you describe a sort of astonishing set of things about the rapidity with which bacterial metabolisms can evolve. And the sort of story that I gathered from that chapter was that bacterial metabolisms in all their complexity have a lot of latent capacities to use new molecules as potential foods. So maybe let's just tell us about bacterial metabolism and where do those latent capacities come from?
Andreas Wagner 16:25
The first thing to know about metabolism is that it's a very complicated chemical reaction network, each of these reactions or most of these reactions are catalyzed by enzymes, which are encoded by genes. And so if you look, for example, at a supposedly simple bacterium, like E. coli, its metabolism catalyzes more than 1000 of these reactions. And what's also remarkable is that for about the last 20 years, we've had very good computational tools to predict what a bacterial metabolism can do. If you give it a set of nutrients in a particular kind of environment, a particular set of growth conditions. Specifically, we can predict whether the metabolism can sustain life in that environment. And what that typically means is that the cell, like a bacterial cell can synthesize all the building blocks of DNA of RNA of protein, and the number of lipids. That's something that was not possible before. And these predictions are actually in pretty good agreement with experimental results is at least for well studied organisms, and for these predictions we use so called metabolic models, or genome scale metabolic models. So these are metabolic models that represent most or all of the reactions that take place in such a metabolism computationally, and that allow us to say what a metabolism can produce, given a set of nutrients as input into the metabolism. And what you can do then computationally is a lot of really interesting things. So for example, you could ask whether a particular metabolism might be able to sustain life on very exotic nutrients that the cell might not normally encounter its environment, right? What you can also do is you can simulate evolution of a metabolism, right, you can say, okay, I started with a metabolism of E coli, and I basically altered its metabolism by either adding reactions to it that we know occur in the biosphere, perhaps encoded by some other organism or delete reactions from it, such as might happen through what we call a loss of function mutation in the genome of a bacterium and to do this, you know, 1000s and 1000s of times, until you have changed starting metabolism to something unrecognizable, that may actually never have existed in the biosphere. And so, when you do that, you can say, Okay, can I produce a metabolism that can survive only, let's say, on a single carbon source? So we asked, we required the metabolism survive only on glucose it's provided as the sole source of carbon and energy, you can do that. And then you can ask, Well, does this metabolism really only have that capacity, right. And what you find out more often than not is even though you selected the metabolism to be viable only on glucose, that is to produce all the building blocks out of the carbon in glucose molecules. It's usually also viable on several multiple up to dozens of other carbon sources, that your simulated selection never actually required it to be viable on. Right. And so you can repeat that 1000s of ways you know, with different nutrients, metabolisms of different complexity, many reactions, fewer reactions. And you find typically, always the same thing, that metabolisms that may have been selected for survival in specific environments are often also viable in other environments.
Marty Martin 19:57
When you get lineages like that, can you dissect those lineages and figure out the biochemical pathways by which they come to metabolize those, is that is that understandable? In the same way? Okay, so you can decompose it in the same way that you could build up from the, yeah?
Andreas Wagner 20:12
Exactly. That's kind of trivial, almost, you know, and what basically happens, you know, a lot of the more exotic reactions are converted into a very, very small number of chemical reactions into more conventional small molecules that are then fed into sort of the main highways of metabolism, if you will. And I think this brings me now to a metaphor, and it's kind of makes this metabolic reaction networks a bit easier to understand a little bit like highway systems, right? Like road networks, in a large and old, historically grown city, like Paris or London, right? The jumble of roads, there's some, you know, major arteries, through which a lot of metabolic traffic goes. So those would be the, you know, the highways, you know, seems like glycolysis, sort of citric acid cycle, from which lots of smaller roles branch off, those would be biosynthetic pathways that are necessary for building amino acids or nucleotide precursors, DNA precursors, and so forth. So one thing that becomes kind of obvious when you think about metabolism as a road network, as a highway network is that often when you build a road, in the real world, your goal may be to actually connect, let's say, you know, a settlement a community to a major city, right. And so that's, you know, it may that community may have funded the road, you know, because they want to be connected to the city. But when that road is built, there may be other settlements, you know, other houses or whatever, that happened to be close to that road, and you can connect them by that road to the, to the major city as well. So, by analogy, in metabolism, we see something similar. So there is lots of these metabolic pathways that represent the chemical conversions of molecules into other molecules. And some of these intermediate steps then can act as nutrients, so to speak, sort of inroads into the major metabolic highways of an organism.
Art Woods 22:06
Yeah, without having to evolve them specifically as target, target molecules.
Andreas Wagner 22:11
Exactly.
Marty Martin 22:12
So Andreas, can I,
Art Woods 22:13
I have some follow ups, but you go first.
Marty Martin 22:15
Yeah. Okay, I just want to try to pull this back to the sleeping beauties, title of the book, and maybe also push you to connect it to the macroevolutionary conversation we're having about C4 grasses. So do you think that part of the reason that these networks evolved to focus on glucose are able to use other substrates is their size and or complexity, like the concept of epistasis for geneticist is often used in this space. And if that is so, when we were talking about C4 grasses and Art asked about the time machine, can we go point at something that gave them this potential? It would seem to be something like that might have been a trait that could have predicted, you know, diversification later on. So how do you think about those things? And especially on the bacterial level that as the sort of where does the variation reside, that allows them to metabolize things that you didn't evolve them on?
Andreas Wagner 23:05
Yeah, so complexity is kind of important here. So I mentioned in passing, you can do these computational analyses in many different ways, with different sizes of metabolisms. And it makes no difference. But it turns out, it does, I've glossed over that. So thanks for catching me on that. It turns out that, you know, for just survival on a, you know, very simple chemical environment and metabolism like that of E. coli is way too complex, you can actually create metabolisms in silico that are much simpler, that contain not more than 1000 reactions, but maybe only three or 400 reactions that will also produce everything the cell needs in a simple environment, right. And when you ask whether such a simple metabolism is also able to survive in multiple other environments, that is typically not the case. So you basically lose with the complexity of metabolism this gratuitous ability of survival in other environments. So part of this ability of E coli comes from the fact that it is more complex than is necessary to survive in a simple environment. And the reason for that is likely that it has had to survive in its evolutionary history in many, many different environments. And those different environments, gave it a metabolism that is more complex than necessary. And this metabolism buys it the ability to survive in even more environments.
Art Woods 24:37
So that's, that's great. That touches on a conversation that we had a couple of months ago with John Glass about the minimal cells that that he's been building at the Venter Institute. And I wonder, have you modeled the metabolisms of these these minimal cells that they've been working on and, and you know, like, my sense is that those minimal cells are like they've stripped away all the peripheral roads, right? And they've just not the main arteries left and they've gotten rid of everything else.
Andreas Wagner 25:04
Yeah. So, you know, I've not looked at the metabolism of their specific constructs, but we have excellent, you know, people in my lab have tried to build and study minimal metabolisms that are no, you cannot make them any smaller, you cannot move remove any chemical reactions anymore, or the whole thing will collapse, you know, the metabolism will not produce what is needed for survival anymore. And, you know, one of the interesting things you find is that, while you might think that such minimal metabolisms are unique in the sense that, you know, you'd converge on something, some unique minimal metabolism, that is the one and only the smallest possible one. And it turns out, that is actually not necessarily the case, so if such a metabolism exists, it's extremely hard to find. It turns out there is a myriad minimal metabolisms, each different from one another, from each other, that have this property of minimality. And they're in the sense that you cannot remove anything from them without destroying the ability to sustain life. So I think that's one of the most, you know, fundamental observations we made that not even in this world of minimal metabolisms, things are unique.
Art Woods 26:15
Yeah, that's super cool. We asked Glass that question, but in the sort of very different context of asking, so he, they started with mycoplasma, and were getting rid of genes out of mycoplasma. And our question was, well, if you started with something else, would you have arrived at the same minimal metabolism. And the sort of related thing here is we asked him multiple times about plasticity and sort of ability to thrive in diverse sets of environments, you know, these minimal cells that they have require a very specific set of environmental circumstances in order to thrive. And our question was, well, are you throwing away all of this capacity to be able to deal with interesting and complex variation in the environment? And I think the answer was yes. And that's part of what you give up when you go minimal, right?
Andreas Wagner 27:00
Yeah, I strongly believe that, you know, minimal organisms will be in a sense, very brittle. That is, you know, if you change the environment, they can't handle a lot of that change. We may have designed them either in the lab or in a computer to be to just survive, in a particular circumstance in a particular environment. And if you change anything about that, they'll just fall apart. And so from a practical point of view, you know, they may not be very useful, right, such organisms, although as an intellectual exercise, I think it's very important to study them.
Marty Martin 27:35
So back back to the E. coli, how much bigger are their networks than these minimal networks? And is that difference in size, like understandable in terms of the other challenges they face? And, to say how many environments they appear in, I mean, that's effectively everything, so that's not a great question, but.
Andreas Wagner 27:50
There's a very concrete example of that, and it's not actually mentioned in the book. There is a species of bacteria called the buchneras, that are fairly closely related to E. Coli. And it turns out there are what are called endosymbionts, that is they live inside other organisms more specifically, they live inside cells of aphids, and their metabolism has been very well studied. And it contains of the order of 280 to 300 chemical reactions that is catalyzed about 300 chemical reactions. They also have very small genomes. So basically, this is a reduction in size of a factor four to five in terms of metabolism. It's actually relatively easy to understand the structure of their metabolism, once you look at what they needed to, what they need to do. In fact, their host cell provides pretty much most of the nutrients they need, right. So all of the chemical reactions that E coli needs to synthesize, for example, certain amino acids that are provided by the aphid host cell, well, the buchneras don't need them anymore. We also know that these symbioses are often fairly old, you know, 50 million years or more. So we can think of the buchnera as having persisted in a nearly constant intracellular environment for 50 million years. And as a result, that could shed almost everything in their metabolism.
Art Woods 29:08
Before we move on from bacterial metabolism. I want to link this back to the idea of sleeping beauties. And let me just articulate what I gathered to be part of the sleeping beauty within E. Coli. And that is that many of the enzymes that are running these metabolic networks are unexpectedly promiscuous in the sense that they they don't just do one reaction, right? They have this capacity to also do other things, maybe not as well or as rapidly, but that those other things represent a kind of capacitance for rapid evolutionary change. So is that is that the sleeping beauty that's inside these complex metabolic networks?
Andreas Wagner 29:48
So let me unpack this because you're bringing in a new idea here, I think that is worth talking about. So even if there were no promiscuous enzymes that is in any enzyme, every enzyme can only catalyze a single reaction, this property of metabolism that I just described, would still exist. And where sleeping beauties comes in here is that a metabolism that may have adapted to a particular environment through the evolutionary history of the organism that it resides in, may actually be able to survive in environments that it has the organism has never encountered before, merely because of these network-like properties of the metabolic highway network. So sleeping means here, this is a dormant a latent potential to survive in a new environment that an organism may have never encountered before. And that may become important if that environment arises, for whatever reason, just like you know, the earth dried up 40 million years after the grasses originated. Now, promiscuity is another phenomenon that exists on top of that. And people have studied that too. That is how promiscuity of enzymes might enable organisms to survive in new environments. And there have been some estimates, for example, for E. coli based on what we know of how promiscuous E. coli enzymes are and about what we know about the structure of its metabolic network in which these enzymes are embedded, they've estimated that E coli might be able to survive in about 20 additional environments that its metabolism would not be able to survive if enzymes were not promiscuous. And this promiscuity also plays an important role in another phenomenon, namely antibiotic resistance.
Art Woods 31:28
So do you think that this promiscuity, is that just an inherent trait of enzymes in metabolic networks? Or is that kind of promiscuity evolved?
Andreas Wagner 31:38
That's, you know, a simple question with, and those are the worst ones, because they have, you know, the most complicated answers.
Art Woods 31:46
Or the best.
Andreas Wagner 31:50
So that's a matter of debate. Okay. So there is one school of thought that says ancestral enzymes, old enzymes, you know that exists, perhaps at the time of the most recent common ancestor of all extant life, we're probably highly promiscuous. Part of the reason is that the kind of machinery that was necessary to produce these, for example, transcription and translation was a whole lot more error prone. Enzymes themselves were not, you know, enzymes expressed in a cell, may themselves have not been just single polypeptides, but entire pools of polypeptides. And if you have that kind of problem, it's much harder to select for enzymes that catalyze reactions with high precisions. So the idea according to the school of thought, is ancestral enzymes were highly promiscuous. And then when it was necessary for enzymes to really become highly efficient, which is often contradiction to promiscuity, evolution holds them to make them highly efficient. So from this point of view, promiscuity was ancestral, enzymes were generalists, that's how people express this too, and have become specialists. But we don't actually really know. And there is others who say, Well, no, perhaps promiscuity is an inevitable property of all enzymes. It's just has to do with molecular motions in the three dimensional fold of an enzyme, and how that catalyzes chemical reactions and that will inevitably lead to some promiscuity that you just cannot get rid of right.
Art Woods 33:17
Seems like one last possibility is that there's actually sometimes positive selection for promiscuity, if it confers some sort of, you know, robustness or resilience on these these networks, right? Or is that not possible?
Andreas Wagner 33:30
Yes. So I think that's definitely a possibility. And you know, a candidate example, may be plant secondary metabolism. So there, we know, there's lots of promiscuous enzymes. And these secondary metabolites that plants produce are essentially often defense chemicals. And plants may benefit from producing a wide array of chemically similar defense chemicals. And that's exactly what a promiscuous enzyme achieves, right. So yes, there is definitely situations where chemical diversity is important in evolution. And in those kinds of cases, promiscuity of enzymes can be highly advantageous.
Marty Martin 34:13
Yeah. So one more quick question, I was going to move on to the origin of genes. But I can't resist this because this is such a cool area and something that I think about a lot. You know, you use metaphors, the thing that we want to make sure we save time to talk about is metaphors and the use of metaphors by humans. It's fascinating, but to use a metaphor for what you just said, with plant secondary defenses. In principle, I think you can imagine that different plants have different toolkits, you know, screwdrivers and hammers and saws and these kinds of things to solve the different problems that they have. But how do we think about the evolution of toolkits when many of those tools will never even get used? How does natural selection operate on a diverse toolkit, if you only ever need the screwdriver? This is a I mean, this is a really bizarre thing that we could argue that you know, these sort of promisciuous factors would be useful, but what if you never get the chance to use them?
Andreas Wagner 35:04
So I think it's actually a really interesting question that I think can be embedded in a much broader question. And that is, when we see a trait of an organism, how do we actually know it's an adaptation? So we have this reflexive tendency to, you know, when we see a trait, say, Oh, this is an adaptation, it's got to exist for some reason, right. And I think a lot of these latent traits that, you know, we see both in computational analysis we predict through computational analysis, of lab experiments, they're not like that. They are traits that exist as byproducts of something else, that may have never served a purpose. They may, they may never will, and they're still there. Right. So I think this whole area raises a huge challenge that is completely unmet, I think, to distinguish adaptive traits from other kinds of traits, no matter how you'd like to call them. And in the case of plant chemical diversity, if, the way I'd like to think about it is that, you know, maybe a plant needs in its lifetime, you know, 50 different chemicals to defend itself effectively against herbivores, and the ability to produce those chemicals may be subject to natural selection over multiple generations. But as a byproduct of getting these 50 chemicals, you get, you know, 300 others. Right. And, and I think, you know, most chemical ecologists would probably agree that, you know, out of the hundreds of 1000s of plant secondary metabolites that plants are known to produce, we're often not sure which of the ones are, are really the important ones. We know certain classes of metabolites. And we may know for specific plants and metabolites that are important, but plants produce many more that we don't know anything about.
Marty Martin 36:44
Yeah, yeah, I think it's really interesting, because I in the way that I would think about it, a lot of my lab is sort of immunology, if you push back the sort of definition of a trait to a slightly different place, it's the propensity to be able to generate that variation, whether it's that set of 50 that solves the known problems, or that entire constellation of 300, that, you know, occasionally, you might need to bring out I mean, your point about where is the adaptation? Is it the one that does the work, or the machine that generates that variation? So let's switch to talk about the origins of another innovation and its new genes. I think historically, you know, the idea is sort of, the traditional view is traced back to Francois Jacob about duplication, a lot of new genes come about by duplication of existing genes. And that still definitely happens, massive evidence of that, consequential evidence. But in the book, you talk about an unexpected discovery in the last six to eight years about lineages coming up with completely new genes from scratch. So tell us about these de novo origins of genes.
Andreas Wagner 38:01
Yeah, so let's let's begin with Francois Jacob again, who said basically, in one of his essays about evolution, that the novel creation or generation of new genes in biological evolution would not occur. He basically emphasized that new genes would come about through duplication. And he had a point, you know, if you think about it, a gene is not really a simple object, right? So and we're talking about protein coding gene. So what you need is you need a start codon, you need a stop codon, you need a sequence of nucleotide triplets in between, right? So that needs to be in place, or you can't actually produce the proteins, you need an open reading frame, then you need transcriptional start signals, you know, that allow the gene to be transcribed and send the stop signal, then you need, you know, for prokaryotes, you need a ribosome binding site to have that thing translated. So it's a complicated option, it's kind of not so easy to see how it could originate de novo. But it turns out that, you know, once a large enough number of genomes or mostly eukaryotic genomes whatever you wanted to compare to against one another. It became clear that this de novo origination of genes does actually occur, it occurs seems to occur quite frequently. And basically, what people realize is that, you know, since our genomes are so large, even by chance alone, for example, you'll find lots of open reading frames, that is start codons stop codons with integer multiples of nucleotide triplets in between, for example, in the human genome alone, there's about 13 million of those, and they are not necessarily very short or can actually be quite long. And then it turns out, you know, something that Jacob, although he, you know, discovered the lac operon and was one of the fathers of gene regulation, didn't know at the time is that especially in eukaryotes transcriptional regulation signals are often very short DNA words. So they arise by chance alone, often in random stretches of our genome. And these observations help rationalize this experimental finding that new genes seem to pop up all the time in eukaryotic genomes. And most of these genes, you know, eventually just disappear again, or not actually never have any useful functions, perhaps a very small minority of them that become useful and sometimes essential to an organism.
Art Woods 40:28
So it's like we have this sort of background churn of origin and disappearance of de novo genes all the time. And at some point, some of those may become useful. Do we have examples of those sorts of de novo genes, either in humans or other eukaryotes?
Andreas Wagner 40:44
There is, no, well, you know, for example, in humans, it's been estimated by genome comparison in a supplemented by some experimental work that about 800 new genes have originated since the, since our split from chimps. And by some estimates, only about six of those have come under natural selection, which is a hallmark of being you know, having become useful. That is, their nucleotide sequence does not evolve neutrally anymore. And some mutations must be eliminated in those genes. That's comparative data. There's also experimental data from Drosophila I believe, and other organisms that basically knocks down or knocks out some of these de novo genes and shows that the fruit flies are not happy when that happens. There is a selection for maintaining those genes. The problem is that, you know, a lot of these genes are not very well studied. So you know, I can't really tell you, there is this one gene that has a very obvious function, and well, it's kind of obvious that, you know, should be essential. And indeed, it is essential, and it's a de novo gene, and this may be because simply, we haven't had enough time to study their function. And in fact, when I wrote the book, I desperately looked through the literature to find good examples, you know, that are easy to explain. And there is not really a whole lot out there of genes with, you know, simple functions that we know a lot about and says, okay, these are de novo genes.
Art Woods 42:05
I don't know if six new genes seems like a lot or not. I mean, you know, it's a small number, but at the same time, it seems like a lot since our split with chimpanzees really to have that that many de novo genes in us.
Andreas Wagner 42:16
Yeah, I actually, I mean, if you want to come back to the Sleeping Beauty angle, you know, I think the more almost the more striking piece of information, is that six out of 800, right. That means the other ones, you know, they are solutions in search of a problem that never occurred, right, at least since we split from chimps.
Marty Martin 42:37
Or hasn't occurred yet. Right? Just give it enough time. Yeah.
Andreas Wagner 42:39
Exactly, exactly. If they hang out for long enough, you know, maybe for another 20 million years, maybe one or the other will become useful, right? What's nice about the molecular world, and this is why you know, my lab is studying the molecular world is that we can get mechanistic insights and quantify the importance of, you know, sleeping beauties. A de novo gene that is not useful for anything is a Sleeping Beauty, it could become useful at some point in the right kind of environment. And what these data on gene origination tell us is that the number of these dormant innovations, solutions in search of a problem, may vastly outnumber those that eventually become useful.
Art Woods 43:19
I want to ask a sort of levels of biology question here. So we've, we've talked about bacterial metabolic networks, so we can say that's sort of the enzyme level, we've just talked about the origins of the new genes. Marty and I think a lot about physiological systems that are occurring at sort of higher levels of organization. So you know, physiological regulatory networks and homeostasis and the organs and tissues that are involved in in that, you know, have you thought about applying this sleeping beauties idea to those higher levels of regulation and higher levels of organization? And are there sort of obvious statements to make about those?
Andreas Wagner 43:57
Well, you know, in our, in my labs research, we haven't done much of that. Because, I guess I have a simple mind and like simple things. And a lot of these physiological system, they're just daunting, right?
Marty Martin 44:11
Yes, they are.
Andreas Wagner 44:12
There's just so many interactions, and you know, so much going on, you know, you just want to tear your hair out.
Art Woods 44:18
But it also feels like they might have just a vast capacity for promiscuity in terms of being able to deal with new new situations and new insults and opportunities.
Andreas Wagner 44:28
You know, so if I had to venture a guess, then I'd say you're 100% right, you know, I think there is, you know, if we see that much, you know, hidden potential on the level of an individual enzyme, right, that we thought you know, catalyzes one chemical reaction but catalyzes five or 10. Now, you actually exponentiate that with, you know, all the enzymes that are out there and all the proteins that interact in some, you know, complex way with other proteins or small molecules, and it just boggles the mind. You know, what could be out there that we have not, you know, even touched or scratched the surface of.
Art Woods 45:01
I agree at some intuitive level. I mean, if you exponentiate this stuff up to physiological systems, it's got to be massive. But it would be nice to, you know, figure out a sort of entry point into understanding that complexity.
Andreas Wagner 45:12
Yes, I think that'd be great. I think you know, there's multiple research programs in that area. Latent research programs.
Art Woods 45:21
The meta metaphor. Yeah, I like it.
Marty Martin 45:25
Well, I think we wanted to sort of majorly switch gears here and move into the human sort of psychological dimension, in a sense, and talk about your chapter eight, on analogies and metaphors. I just think it was, it was, it was phenomenal. I mean, I've skipped over a bunch of really cool sections. We've talked a lot about genetics and molecular biology, but tell us the story about how you understand metaphors as either sleeping beauties, or the products of sleeping beauties, maybe is the is the better way to think about them.
Andreas Wagner 46:02
Yeah, so I hope I can articulate this at all, because it was always one of the shortest chapters in the book, it was the most difficult to write, and I almost deleted the whole thing. Because I said, you know, I can't just explain this, explain this, clearly enough. Right. But, well, let's just begin with analogy, which is sort of an elaborate form of metaphor, if you will, that's important in science, we often think of this analogies as didactic devices, you know, that, you know, we explain to a student, how I don't know homeostasis works in blood sugar, by comparing it to the water level, and how its controlled in a toilet bowl, that sort of thing. Right? Tiolet tank, right. But history of science shows that analogies are much deeper than that they go much beyond didactic devices. So you know, in some cases, there are important discoveries in and by themselves that holds both in science and engineering. So in science, for example, I think the analogy that atoms behave like strings that are vibrate, did a lot to explain atomic properties, during the early days of, of quantum quantum theory, quantum physics. In engineering, we know that there is many, many, if you will, analogical discoveries. So for example, the invention of velcro took place when a Swiss inventor I forgot the name right now took walks with his dog in the woods and found that the burrs of some plants got attached to the dog's fur. And he thought, well, maybe this could be used technologically, right? So there's many examples like that, then in literature, you know, the analog of an analogy is the metaphor, as a sort of a very compact analogy. And there's a lot of psychological and psycho linguistic work that suggests that, you know, our minds are fundamentally built to think in metaphors. And that metaphor is not just something superficial, to our minds, it's just really foundational, there's really, really excellent work in that area that, you know, I'm not an expert on at all, I've just sort of read this all this material in connection to the topics of the book. And the way people who study these mental structures, also often called conceptual spaces, think about it is that you know, there's some latent connections in our brain, that we have this latent ability to map relationships in one domain of life, let's say the vibration of strings, to another domain, certain objects in physics, for example, and it is this ability, this latent ability that leads to discoveries through analogies. And that's really all there is to it. Right? Our ability to discover such hidden relationships between objects is a fundamental ability of our brain that facilitates discovery.
Art Woods 48:57
Yeah, that's super beautiful. You know, when I read this chapter, I think I came into it with a mindset that metaphors and analogies obviously are important. And they can lead to discovery, but I didn't quite grasp how central they are, to the way we think and the way we speak. And that was a real revelation to me. And let me just bring up another sort of topic that you cover in this chapter. And that's this idea that we use metaphors all the time, and that many of them map on to sort of archetypal underlying metaphors. So you get this interesting example of mapping language onto water, right? So we say things like, you know, scalding insults, words flowed from a pen, somebody is showered with praise, and that there's this sort of underlying mapping of language and water somehow in our brains. And there's a similar sort of archetypal mapping of time and space, which is super interesting. And so the way I think may be to link this back to the idea of sleeping beauties, is it your idea that these sort of archetypes are providing this underlying structure that allow us to understand really complex information in the world, and to make linkages that we wouldn't otherwise make when we're confronted with new information.
Andreas Wagner 50:14
Yeah, I think that's correct. And before I say anything further, you know, the people who have really, you know, made the best case for this are, you know, Steven Pinker who has, I think, has popularized this, our metaphorical way of thinking as in some really fantastic books, and George Lakoff, linguist, a psycho linguist who, I think, first, make the wider world aware of how important metaphor is. And the way I think about it, and this is, again, you know, taken from the literature largely, is that this ability is sort of a fortuitous byproduct of how our brains process information, and how our brains are able to map existing relationships on new relationships, right. So there is a very interesting experiment that was done a psychology experiment that was published in Science magazine a few years ago, where people were shown images of a bird on a computer screen, and they were told that they could manipulate two aspects of the morphology of that bird, it's a cartoon bird, really, that is the length of the neck and the length of the legs with two independent knobs. And they basically were trained to, you know, to use these knobs to create birds with all kinds of new morphologies. Right? Then, the people who did this experiment studied, which kind of neural circuits are being activated while the volunteers did this. And it turns out that the circuits that were activated are responsible for spatial navigation in the real world. The reason why this is interesting is because it tells us that, you know, the brains of these volunteers mapped an abstract relationship, namely, a neck length and leg length of a cartoon bird, onto something a circuitry that is ancient that we use to, that we use to navigate space? Right? There's very few examples like that, because this has not been studied extensively. But I think it's very telling. It shows that circuits in our brain can be used for a lot of tasks that they have not necessarily been evolved for. So they have this huge latent potential to to discover or map new relationships.
Marty Martin 52:30
So in one sense, this sort of conflicts, it seems with promiscuity, as we talked about before, but maybe that's because I'm sort of putting the Sleeping Beauty part of these ideas in different places, if some enzymes may have evolved because they provide this promiscuity or at least it's the case that some enzymes are promiscuous, and sometimes that can help fitness, for metaphors, we're mostly talking about the reduction of complexity, and then the sort of interlinking of somewhat disparate ideas. Are these things related? Or how am I misunderstanding this?
Andreas Wagner 53:01
So you know, the way I would think about it is that we have a latent potential of, you know, circuitry in our brain that's somewhat analogous to the promiscuity of enzymes that is, just as these circuits you know, can be used for things that didn't evolve for us. So can enzymes catalyze reactions that did not evolve for.
Marty Martin 53:21
Yeah, and if you turn it, you use the metaphor to understand promiscuity. Does it mean that somehow promiscuity becomes limited, like tractable in the same way that metaphors make complexity tractable? Is promiscuity somehow limited? Because the you know, the metaphor has evolved to simplify, is promiscuity actually more simplified than it first seems?
Andreas Wagner 53:41
Now you lost me and I think the reason you lost me, is because I'm not quite sure in which sense metaphor simplifies.
Marty Martin 53:49
Well, if you know, I'm thinking simplification in the sense that water flowing from a pen is a way that we come to understand we come to represent something happening because our brain evolved to understand flowing water. I mean, it's something that was seen and so that that mapping happens, complex kinds of things come to be understood via metaphors because of a historical utility.
Andreas Wagner 54:09
I agree with that. I mean, I have no problem with that. But you know, I think I'm not sure flowing water describing language as flowing water or describing heat, like flowing water, this has also been taught in physics, right, makes things simpler. Rather, it reveals an essential relationship between objects in two different domains.
Marty Martin 54:30
Okay, I think I'm using yeah, the essentiality thing as, I'm sticking in simple when that's probably not the appropriate thing. It's more of the essentiality. Yeah.
Art Woods 54:37
Okay. Here's a follow up. I mean, I certainly think of metaphors and analogies as something that's human specific, it's something we can do. But if we look further abroad, taxonomically, do you think other primates and maybe other mammals also use metaphors in their thinking?
Andreas Wagner 54:54
Wow, that's a good one. I have no idea because if we did a hypothetical experiment, you know, where you did exactly the experiment that I described to you, you know, with the volunteers and the birds changing shape, right, and let's assume we could train monkeys to do this, I would suspect we would find the same neurological correlate that we'd see in humans. Right? That a question, in which sense, is this metaphorical? Right, you know, typically, metaphor is linked to language. Right? And, you know, we don't we know that don't have symbolic communication. That's characteristic of the kind of language that we use. Right? So I don't know.
Art Woods 55:36
Yeah, that's a fun one.
Marty Martin 55:44
Okay, let's do one more fun one, and then we'll start to wrap up. But let me ask another question. Probably push all of us to as biologists realms that we don't usually think a lot about, I was just listening to a podcast this morning with David Krakauer talking about intelligence. And he was particularly talking about AGI artificial intelligence, artificial general intelligence, making the point that I think I should have realized a long time ago that the AGI is that now exist, chat GPT and such, are these massive lookup tables, right, with an insane amount of memory and information that no human brain really has. So it's fundamentally a different way of being intelligent than we are. And yet, maybe it looks like it can do some of the creative kinds of things that we attribute to human intelligence. But do we train? Have we taught AGI? Have we made the attempt to give metaphorical thinking? Or do we have any evidence of metaphorical thinking in chat GPT?
Andreas Wagner 56:42
I really don't know. You know, I think that's going very far out of my comfort zone here. Right?
Marty Martin 56:48
That's what I promised.
Andreas Wagner 56:53
Well, you know, I know a little bit about how these models work, you know, I code very simple versions of neural networks, you know, for our own work. I am mostly, you know, stunned by the public debate right now, you know, people are so shocked that, you know, they what the engineers call they hallucinate, you know, they could actually speak untruths, where, you know, if you know, how these things work, basically, depending on very subtle correlations in word occurrences in billions of documents, right? They make everything up, right, every single thing is made up, it's a surprise that anything is true.
Marty Martin 57:29
So that's how it works.
Andreas Wagner 57:33
So going back to your question, my suspicion is they don't have the underlying, you know, structure that is needed to produce meaningful metaphors.
Art Woods 57:44
Okay, so so it's like giant correlations among words, and they're trying to predict like, the next word in the sequence of words. And that strikes me is like a very low level of imposing a structure and imposing a correlation on, you know, what's coming next. Isn't a metaphor, sort of the same thing, but it's at a more conceptual level, it's not trying to predict relationships among words, but relationships among concepts. And so maybe, maybe this is the path to like, a more intelligent AI is to actually build in explicitly this capacity to, to understand and use metaphors.
Andreas Wagner 58:18
That's quite possible that you're that you're right. And yes, I agree.
Art Woods 58:21
And quite possible that I'm wrong.
Andreas Wagner 58:24
There's nothing, you know, like a concept, the concept of a concept built into these large language models, right? Yeah. I mean, there's two readings of it is first is that, you know, it's all too too low level to actually go anywhere, or, you know, we are at so many times in our human history, overestimating our own abilities.
Marty Martin 58:44
Okay, one more on this space Andreas, and then we can move on, because clearly, we're outside of our wheelhouse here. It's so much of a blackbox right now. Right? So how, do we know that they aren't finding relationships among completely disparate kinds of things? I mean, basically, producing metaphors? Is there anything about the architecture of these lookup tables that would prevent them from using something like a metaphor to come up with an answer?
Andreas Wagner 59:13
It wouldn't. But I think this is actually a research question. You know, there's a lot of people who want interpretable AI, that's what it's called, you know, basically have a neural network that, where you understand what it's doing, and for very good reasons, especially if you apply AI in the legal domain or the medical domain, you know, you you'd like to know whether, you know, this machine fails or you got cancer. Right. And I think this the question that you raise there is a question, right, for researchers who are really into finding interpretations or building networks that are actually interpretable.
Marty Martin 59:47
So to make a bit of a practical turn, first, metaphorical thinking, I mean, we've been arguing is quite powerful, and it's definitely powerful in the sciences. Is there a way that we can become better metaphorical thinkers? Is there, you know, do you do you encourage or coach your graduate students to think more metaphorically? And how so? If you do?
Andreas Wagner 1:00:08
A short answer is that no, we don't talk about that very much. Because, frankly, for the majority of them, you know, I'll be happy if they can write a straight sentence that is declarative and where you don't get lost in all the verbiage. And I think, you know, the, the question goes very much in the direction of what makes a mind, you know, creative? And can you promote that in some way? And, you know, psychologists might have an answer to that. I don't. It is related to something that's called divergent thinking. That is basically, you know, when you ask a person to come up with associations to a particular word, right, and there are some people who come up with, you know, only those associations that everybody else comes up with, and then there's others who come up with more unusual associations, and there's yet others who come up with very bizarre associations, and you know, there's sort of the line between madness and creativity, right? In fact, the earliest word association tests were actually not tests of creative thinking but for madness.
Marty Martin 1:01:14
I'd like to take one of those tests, I might might end up in a
Art Woods 1:01:17
We already know that you're mad, Marty.
Marty Martin 1:01:18
Yes. Well, that's what I'm saying. So another kind of applied question, scaling out to the sort of ecological ramifications of the whole idea of Sleeping Beauty, and its many forms that we've talked about. Climate is changing dramatically. The humans are changing the planet in many, many different ways. How do you feel about the sort of impacts on life on Earth, given what you know what you've come to know about sleeping beauties? I mean, is there less to worry about with anthropogenic change than maybe the the conventional wisdom has?
Andreas Wagner 1:01:49
No, I don't, I don't think so. I think, you know, we should be just as worried as we've been for the last, you know, 20 30 years. However, I think we're, you know, we don't necessarily give life enough credit, in its ability to come up with new solutions that it faces, and in fact, some of these solutions may already be hidden in there somewhere, right? We know, though, that, you know, there's limits to what evolution can achieve. So something that everybody's talking about is evolutionary rescue. Right. And there's important questions there, how often that is going to be possible. Climate change, you know, is essentially something that happens very slowly. But it's also associated with extreme climate events that can drive populations to extinction. And, you know, can some of these latent properties that people have discovered help populations in this evolutionary race? And I think that'd be a really interesting research question to ask, the only thing I can say is that whatever latent properties they have, there's going to be hard limits evolutionarily. So for example, a former postdoc in my lab, did a really neat evolution experiment where she took an Antarctic bacterium that is adapted to very cold temperatures, and basically tried to evolve it to survive at high temperatures. So I think the optimal growth temperature of that bacterium was somewhere around 10 degrees Celsius. And you know, she got it to 30 degrees. But there was just no way it would go beyond 30, you know, 31, everybody was dead. And to try this in multiple replicate population, that just never worked, right. So we don't quite know what the reason is, and we have some hypotheses about what might be going on, it has to do probably with the fact that the proteins become really unstable at these higher temperatures, and this bacterium does not have certain variants of a, of chaperones that protect the proteome against this instability. And if you don't have those variants, you know, you're screwed, you cannot evolve, adapt beyond a certain temperature. So there's going to be limits. But I think it'd be very interesting to find out what role any kind of latent adaptations might play in shaping those limits, or mitigating the the hardness of these limits.
Art Woods 1:04:11
Yeah. Seems like you could imagine that there's a lot of sort of latent capacity for rapid microevolution, but that at some point, you hit the limits, and you need something else, you need horizontal gene transfer something more major to happen in order to provide the variation that allows you to surpass the limits, right?
Andreas Wagner 1:04:29
Yeah, we're gonna need a lot more time.
Art Woods 1:04:31
Yeah, right. It's a long PhD or long postdoc.
Andreas Wagner 1:04:34
Exactly, the 1 million year long postdoc. And that's, I think, one of the big limitations of laboratory evolution experiments that, you know, we basically impose selection pressures on organisms that are so much higher than what they might often experience in the wild, right? And on such compressed timescales, that challenges that they might meet in the wild, they're just not going to meet in the lab.
Art Woods 1:05:00
So here's another climate related question. And that has to do with with plastics, you know, I feel like I just read so much in the last few years about plastic contamination and microplastics everywhere. Do bacterial communities altogether have a lot of latent capacity for metabolizing plastics? And is there hope for much more rapid degradation of plastic in the future decades?
Andreas Wagner 1:05:24
Funny that you mentioned that I just got a request from somebody who wants to do a postdoc with me, and the person who is a marine ecologist, and she actually studied plastic samples from the Great Pacific Garbage Patch, took them to the lab and observed microbial colonization and how fast that occurred on these plastics. And it turns out, yes, you know, there's microbes that grow on these plastics. And you know, one thing that you want would want to pursue, is actually to find out whether a, you can evolve the microbes in the laboratory to digest these plastics faster. And where the microbes that have evolved to digest one kind of plastic can perhaps also digest 1000 others, something similar to what we've observed with antibiotics experiment.
Art Woods 1:06:13
Sleeping Beauties in the plastics world.
Andreas Wagner 1:06:14
Exactly, sleeping beauties exist there as well.
Marty Martin 1:06:17
Wow. That's brilliant. That's brilliant. So one more applied question of a sort in terms of how humans are modifying the planet and the latent plasticities, or the latent traits that may exist. What about the species that seem to be problematic for us? I mean, is there any reason to expect that in some lineages there are, there's more latent innovation, and I'm thinking in particular, of introduced species, especially with pests, like, you know, Phragmites, and many other things that we spend billions of dollars trying to control?
Andreas Wagner 1:06:47
Well, you know, surely cuts both ways. Right, you know, the things that can benefit us can also harm us, you know, and, and we, you know, yes, invasive species are a potential example, antibiotic resistant bacteria, you know, are another example, you know, so that we haven't discussed this explicitly. But there are these interesting and beautiful studies on samples of bacteria that have never been in contact with human civilization that may have been living in isolation for millions of years in subterranean caves. And you study them and you find, you know, some of these are resistant against eight antibiotics, and some of them at clinical concentrations, and even including antibiotics that are synthetic that is manmade, that don't occur in nature, right. So there's a latent potential there, that does not act in our favor.
Marty Martin 1:07:35
Yeah, yeah.
Art Woods 1:07:37
Well, Andreas, this has been a super fun conversation, we think this might be a good place to start wrapping up. And one, one question we always ask for our guests is whether there's anything we didn't cover any points you'd like to make here at the very end, that we didn't already ask you explicitly, or, you know, any, any major points from the book that we didn't, that we didn't cover that you'd like to communicate?
Andreas Wagner 1:07:58
Yeah, perhaps there's one thing that we really didn't touch upon that sort of the human dimension of this all, you know, when it comes to human creativity and the frustrations that human creators often have. So I'm part of the impetus for me writing this book was actually not the biology side that we discussed today. But it's actually the human side. And in fact, there was an earlier version of this book, earlier draft that had two thirds examples from sleeping beauties in human technology and culture, and maybe only 1/3 in biology. And so one of the one of the motivations for me writing this book was that, you know, I know a lot of creative types, you know, artists, musicians, and as I'm an amateur musician myself. Some painters, you know, writers, and although they mostly love what they're doing, they're often extremely frustrated with their lack of success. Right? Even successful creators, you know, they have some real hits, but you know, a lot of other creative products they produce are duds. Right? And I think there isn't a productive scientist out there who, you know, had a paper at some point where he thought this is gonna rock the world, and then nobody even reads it or cites it. Right? The point here is that, you know, I realized, when I talked to these people on the one hand, and read all these literature about sleeping beauties in you know, dormant innovations in technology, artworks that have been ignored for hundreds of years. And then of course, all the biological examples, we discussed, that, you know, it's not actually worth always doubting yourself about whether what you're doing, you know, is good or not, you know, as long as you enjoy creating, because you have no control over the success of your creation, a lot of times and the vast majority of the times, it's going to be determined by some something in the world, some, you know, coincidence, or perhaps some, the right environment that in which your, you know, novel or whatever needs to be born into to determine its success. And I think that's sort of for human creators, perhaps the most important lesson from to draw from the book.
Art Woods 1:10:02
Yeah, and our creations may not find the right context until far after our deaths right?
Andreas Wagner 1:10:07
Or never.
Art Woods 1:10:08
Or never. Yeah.
Andreas Wagner 1:10:11
As long as you're fine creating them, you're gonna be fine.
Marty Martin 1:10:13
Yeah, good. Well, hey Andreas, thank you so much. This has really been a lot of fun. I wish you the best success with the book and hopefully we can catch up again on the when the next one comes out.
Andreas Wagner 1:10:23
Fantastic. Thanks a lot for having me it was a real riot here and you know to talk to you guys was really great.
Marty Martin 1:10:44
Thanks for listening to this episode. If you like what you hear, let us know via Twitter, Facebook, Instagram, or just leave a review wherever you get your podcasts. And if you don't, we'd love to know that too. All feedback is good feedback.
Art Woods 1:10:55
Thanks to Steve Lane who manages website and Ruth Demree for producing the episode.
Marty Martin 1:10:59
Thanks to interns, Dayna De La Cruz and Kyle Smith for helping produce the show. Keating Shahmehri produces the fantastic cover art.
Art Woods 1:11:05
Thanks also to the College of Public Health, the University of South Florida, the College of Humanities and Sciences at the University of Montana and the National Science Foundation for support.
Marty Martin 1:11:15
Music on the episode is from Podington Bear and Tieren Costello.