Listener Questions #35: Genetics (featuring Dr. Benjamin de Bivort)

00:00:07
Speaker 1: What does it mean to decode DNA? And having done it, what can you really say?

00:00:14
Speaker 2: Is DNA coding for innate behaviors more like a blueprint or a recipe? How does DNA result in a web by a spider or honey from a bee?

00:00:24
Speaker 1: Can instincts be transferred? And how do we know? Could gene manipulation give me a rooster's crow?

00:00:31
Speaker 2: Whatever questions keep you up at night, Daniel and Kelly's answers will make it all right.

00:00:36
Speaker 1: Welcome to another Listener Questions episode of Daniel and Kelly's Extraordinary Universe. This is episode number thirty five, featuring doctor Ben de Bevore.

00:01:00
Speaker 3: Hi.

00:01:00
Speaker 2: I'm Daniel. I'm a particle physicist and I'm not a biologist.

00:01:05
Speaker 1: Hello, I'm Kelly water Smith. I am a biologist, but I am not a geneticist. And I did kind of try to avoid genetics a little bit when I was in grad school.

00:01:16
Speaker 2: Why is that, to meet? Genetics has so much mathematical underpinnings. It's like maybe the mathiest part of biology. Oh wait, am I answering my own question?

00:01:25
Speaker 1: I am insulted. So I actually I went into ecology because I loved that it had math.

00:01:34
Speaker 2: Right, differential equations.

00:01:35
Speaker 1: Yeah, that's right, that's right. And what you know, I was taking stochastic differential equations over in the math department with the math PhD students when I was working on my masters. I like the mathy stuff.

00:01:46
Speaker 2: Awesome. All right, So then what's your objection to genetics?

00:01:49
Speaker 1: You know, I was changing pretty quickly what I was. Here's here. Bottom line, I'm not good at the PCR machine, and so I just it was like magic and I just felt like I couldn't do genetics in the lab, and so I just kind of veered away from it a little bit. I'm not a lab person.

00:02:09
Speaker 2: I think you're more of a farm person, more of a.

00:02:12
Speaker 1: Farm person or like set of experiment like a behavior experiments kind of person. I'm just I'm not good with a pipeet. That's it, as we've discussed before.

00:02:19
Speaker 2: All right, So then my question to you today is are you doing genetic experiments on your farm? Are you breeding some new variety of goats or something similar?

00:02:31
Speaker 1: You know, I have had a lot of discussion with the people I got my goats from about like, oh, you know, dude, which of the goats on your farm, like which ones had mothers that produced a lot of milk and had good teeth placement and had good you know stuff like that.

00:02:46
Speaker 2: And so I'm sorry, I have to ask what is good teeth placement and what is bad teeth placement?

00:02:51
Speaker 1: Well, well, so it's like a combination of like, you know, are they placed like, you know, far from the legs so it's easy to like get a good hand. And then you know, are they like long so you can get your hands around them because you're like your fingers cramp if it's just like your thumb and your pointer finger. You know, you want them to be nice, you know, long, and you want them to have a big hole so a lot of milk comes through with each squeeze.

00:03:14
Speaker 2: And good for human milking then not for kids.

00:03:18
Speaker 1: Suckling human milking. I feel like that gives me a different visual picture. But good, good for easy, easy for the human who is doing the milking.

00:03:30
Speaker 2: Yes, indeed, yes, okay, wow, fascinating. That's right, all right? And is that factoring into your decisions about these ghost features?

00:03:39
Speaker 1: You know, I will tell you that we you know, we've talked about lotties prolapse a number of times on this this podcast, but that tends to be genetic, and so Lotti's kids will not be bred because we don't want to pass that trait on Wow, And so any little dudes that she had will become what they call Weathers, which means we're gonna finally get a goat named Carl Weathers, which we're excited about. And uh, if we have any boys would uh and then if if we have girls, then we're just gonna they'll be They'll all be pets. So anyway, this is this is factoring into some decisions on our farm. But uh, we're getting a little off track because we're.

00:04:20
Speaker 2: No, we're not. We're talking about how genetics influence organisms, which is a topic of today's episode, isn't it.

00:04:25
Speaker 1: Yeah, No, you're you're right, You're right.

00:04:27
Speaker 2: And because of your famous discomfort for genetics, you called in a ringer, didn't you.

00:04:32
Speaker 1: I did, Yes, right, and not just any ringer. We are a big enough podcast that the ringer we called in is a Harvard geneticist. We got doctor Benjamin de Bevore. He's a professor at Harvard in the Organismic and Evolutionary Biology Department and the Center for Brain Science. He studies evolution and behavior where he asked questions like why do genetically identical individuals behave differently from one another? And I am super excited that we got them on the show. And so today we're doing is this supposed to be a listener questions episode? There were some parts where I felt like I can handle this because so, you know, while I said I kind of tried to avoid genetics, I did get a degree in animal behavior, so I didn't like avoid genetics entirely. And so today we're gonna like jump between a conversation that you and I had with Ben, Yeah, and then me answering some questions because I'm not completely incapable. I'm just maybe a little bit nervous about genetics. And so but I'd like to preface this with any mistakes that are made when I'm just talking on my own, those are my mistakes, not Ben's mistakes. But yeah, we're gonna jump between a conversation that we had with Ben and you know, me trying to hold my own on genetics, and we'll just see how it goes.

00:05:54
Speaker 2: All right, Whenever you feel uncomfortable, just mask that Harvard Genetics button.

00:05:57
Speaker 1: That's right, That's what I'm gonna do, and our trusty and amazing audio guy Matt will come up with the sound for us. We love you, Matt. Okay, So let's go ahead and start this episode by bringing Ben on the show. And it turns out that in addition to enjoying puzzles, skiing, sailing, fossils, minerals, mushrooms, and games, Ben enjoys super forecasting, super forecasting, super forecasting, and he was a super forecaster for the Good Judgment Project. So we are going to jump in to our interview with me asking Ben a question about super forecasting. Matt bring us to the conversation with Ben. What does it mean to be a super forecaster?

00:06:41
Speaker 4: Yeah, so this is the idea that with a little bit of reasoning and reading the news, people can make reasonably accurate predictions about future geopolitical events. It's a set of skills that have been monetized of all things like and now they're just like betting markets, where you can make forecasts about elections.

00:07:05
Speaker 1: And were you working on the betting markets or you were what were you doing?

00:07:10
Speaker 4: Now? This was in an earlier stages where this kind of these techniques were the object of academic research, and so there was a research team out of University of Pennsylvania. Feltetlock and Barbara Meller's were the lead investigators, and it was a research project for DARPA of all things. You know, the government certainly has an interest in knowing what will happen politically in the future. And their first result had been that random people do better than so called experts pundits at predicting future events. And then their main finding when I started to be involved, was that you could aggregate the predictions of amateurs the wisdom of the crowds to make more accurate forecasts. And then they ran these studies and I was just a study subject testing variables like prediction markets versus forums as a sort of discussion venue for making predictions, so testing some theories about efficient markets producing the best estimates. And then I ended up being good at it and being sort of categorized as a super forecaster, which meant that they used us as one experimental group for some further studies and put us head to head against people from the intelligence community who had access to classified information. And so there was sort of four or five studies that came after those initial results.

00:08:40
Speaker 2: And so do you think the CIA has like tapped your living room to hear your dinner conversation, so they can tap into your inside about the future of the world.

00:08:48
Speaker 4: Yes, right, they've just activated the microphone on my for everyone.

00:08:57
Speaker 1: All right, well, I know who I'm going to call the next time I need to know how an election turns out.

00:09:02
Speaker 2: Wait, I can't believe I didn't ask him what's gonna happen in the midterms.

00:09:06
Speaker 1: Oh well, I guess we're just gonna have to wait and see, or maybe we can bring him back for a midterms episode.

00:09:14
Speaker 2: I also should have asked him whether his fondness from mushrooms helps or hurts his ability to forecast elections. But maybe I held my tongue because that's not the kind of question you ask a Harvard geneticist.

00:09:25
Speaker 1: Yeah maybe, And you know, frankly, that could have been a whole episode of questions about super forecasting. But we need to get back to genetics. Yes we do. Since many of us learned the you know, the basics of genetics and translation and transcription when we were in like sixth grade, I felt like maybe I could tackle that topic. So let's give it a shot.

00:09:49
Speaker 2: All right, So Kelly explained to us, how does this all work. How's information stored in DNA? And then how does that turn into me and you and mosquitoes?

00:09:58
Speaker 1: Okay, right, fortunately it does turn into mosquitos sometimes. All right, but I'm going to try to keep this as like simple and jargon free as possible. So the way information is stored in DNA is that it's stored in a series of letters. These letters are ATG and C. We're not even going to say the names because Daniel's gonna get annoyed if I start using the names.

00:10:21
Speaker 2: Because there's no information in the names. It's just cognitive burden for your listeners.

00:10:25
Speaker 1: And that's why you never say muon or anything on the show. It's what I appreciate your consistency. So, so.

00:10:35
Speaker 2: Muons do sometimes change the letters in DNA? Actually, oh, so it's relevant cosmic ray radiation?

00:10:42
Speaker 1: Wow? Wow? Does it change like an adnine to aguanine Ooh?

00:10:48
Speaker 2: It changes as something to something mils. Okay, But I think of these four letters as sort of like quaternary code, you know, like binary zero or one, trinary zero one or two. This is like a digit that has more possibilities. Is that the right way to think about it? Yes?

00:11:02
Speaker 1: Yeah, yeah, it's a digit that has four possibilities, and then for each location, and those locations are many. You can have any of these four digits. And the way they're sort of set up is kind of like a rope ladder, is how I think about it. And so you've got like a rope on one side, sort of holding everything together, and then you have rungs, and on the left side you've got like you know, an A, A C, A G and A A TTT and stuff like that, and then on the other side you essentially have the exact same information, but it's flipped. And so to hold the ladder together, the rungs snap together. So instead of being just like one letter that goes all the way across, you get two pieces that snap together. And to get them to hold together, you need to have the opposite letter to essentially make them like molecularly strong enough to hold together.

00:12:00
Speaker 2: In this sense, they're not exactly just like four letters, right, because they have this opposite relationship that's right, Like if it was zero, one, two, and three, like I couldn't tell you three is opposite to one, But here they have this additional relationship that's right.

00:12:14
Speaker 1: Yes, yeah, so I should have earlier been like you're wrong, I missed.

00:12:19
Speaker 2: My chance, or you'll get many more.

00:12:22
Speaker 1: Oh yeah that's or no, you're right, plus complications because it's biology, so all right. So, so across from an A, you always get a T, and across from a T, you always get an A, and across from a C you always get a G. And acrossrom a G, you always get a C. So anyway, the a's and the t's always teamed up. The c's and the g's always teamed up on this rope ladder, and so you get this rope ladder system. The way you usually find it configured is that the rope ladder sort of like twists and so like if you imagine you've got like your sibling on the rope ladder and you really want to like get them dizzy, you sort of twist it at the bottom and then you let it go and your sibling sort of goes sort of flying as it untwists. Can you imagine what that looks like when you twist it?

00:13:08
Speaker 3: So?

00:13:08
Speaker 1: Yes, okay, so when it's twisted, it looks like what's called a double helix. If you get your sibling off, so it'll twist the right way, and if you keep twisting and twisting and twisting and twisting and twisting, it gets really really tight. And that's how we store it in your nucleus. Okay, so that's the way it's usually stored. But you can't really do anything with it when it's stored up like that. So now say you want to do something with it, Say, for example, you would like to use some of the information that is stored in one of those genes. Say maybe you want to make a particular protein that is coded for in one of those genes. The next thing that happens is called transcription. So you start unwinding that messy ball of stuff that you've got, and now you unzip and you start opening up that rope ladder. So remember how I was saying, Like, you've got the a's that are matched with the t's, you've got the c's that are matched with the g's. Now you've got something that goes in and it sort of separates all of those things. And so now your rope ladder is in two different pieces. And now you've got RNA that goes in and it also matches up c's with g's and g's with c's. But now instead of matching up a's with t's, it matches up a's with us. This is a point that you really don't need to understand. Biology is weird. I don't necessarily know that biochemists know why biology decided to add a U instead of a T here, But it's fun, right, It depends it had to be something different at this stage.

00:14:43
Speaker 2: Because it was all about to make sense to me. And then you threw in the U.

00:14:46
Speaker 1: That's right, that's right.

00:14:47
Speaker 2: And you're even kind enough to use a U spoon. You just threw it right out of it.

00:14:51
Speaker 1: That's right, that's right. I know, I'm so sorry. I really loved your joke the other day, as long as it's not a me spoon. So anyways, and so now you've got essentially a copy of what you had before, but it's like an inverse copy with us now instead of teas. But anyway, so now you've got this copy, right, And so.

00:15:12
Speaker 2: So far the DNA has been like tightly stored. It's got information in it, though we don't yet know like how to read that information or what it means specifically. It's just like some set of code tightly bound. Then you've unwound it, and now you're using the RNA to somehow read that, and then it's going to get us to building proteins that make up me and mosquitoes. Is that where we are?

00:15:33
Speaker 4: Uh?

00:15:33
Speaker 1: Yeah, that's where we are. Okay, yeah, I mean, like evolution knows what that stuff does, but yeah, but I have yes, Yeah, that's right, that's right. Okay, So all right, so now you've got RNA that has like an inverse copy of the information, and it now is going to travel out of the nucleus and out into the like watery outside of the cell. It's going to find an organelle like a little or I think you called it an organ eito in a previous Listener Questions episode, and I will always think of them as organ eito's.

00:16:07
Speaker 2: Now it's cute.

00:16:08
Speaker 5: I know.

00:16:09
Speaker 1: It's a little organy too, and called a ribosome. And at the ribosome you are going to have what's called translation, where essentially you are going to translate this information from RNA into proteins. And so now each of those three letters are coding for an amino acid, and amino acid is just a little building block for a protein. So each of those three letters is going to become a chunk of a protein. And so at the ribosome, you just get one amino acid chunk connected to another amino acid chunk connected to another amino acid chunk, and you get all of these chunks together and you get this long string that is going to fold up to become a protein. And these proteins are essentially like what makes up all of your body. It makes up the mosquitoes, makes up your hair, it makes up you know, all of the stuff your the hormones that you know that make you do the things that you do and whatnot. Okay, so now you've got this long string of amino acids, and we've now talked about what is usually thought of as like the central dogma, which is that you know, you've got this information, it's stored in DNA and then gets copied with RNA, and then that RNA gets turned into proteins. But we're going to now talk a little bit about how those amino acids get turned into proteins that then get folded up in a very particular configuration, because Daniel and Ben had a really interesting conversation about how that stuff folds up on itself, and so we're going to jump back to our conversation with Daniel and Ben so that we can talk about protein folding. So we're pressing the interview button and we're heading back to our conversation with Ben.

00:17:53
Speaker 4: We can say that, you know, amino acids are compounds. They have nitrogens and carbons. If you take protein powder, it's going to be rich in amino acids. Monosodium glutamate is a salt of glutamate which is an amino acid. And there are about twenty of these amino acids which get pieced together in different sequences to make proteins, and they have different chemical properties, and by assembling them in the right combinations in the right sequences, you can get that protein to fold into the correct shape.

00:18:31
Speaker 3: This is a bit like.

00:18:32
Speaker 4: Origami where imagine like origami, where you don't have an instruction guide with the paper. It's just that the paper has certain properties of like you know, one side has a certain kind of stickiness and it wants to fold up to the next part of the origami. And these are the so called like electrostatic properties of these amino acids, where the ones which like to hang out with water will tend to hang out together and the ones that try to avoid water get pushed together. And so by developing the right sequence of amino acids in the protein, it ends up adopting a particular three D shape, and its function within the cell comes from that shape that it has.

00:19:12
Speaker 2: And so is this the process we here refer to as protein folding that you've put together these amino acids and then they somehow make themselves into this three D machine? And how much of that is understood? Like if you start from a DNA sequence, how hard is it to predict what protein it's going to turn into?

00:19:32
Speaker 4: Yeah, well, I would say for a few decades we've had a good conceptual understanding of it. So there's different levels of protein folding organization. So one thing that happens is, you know, there's interactions between the mo one amino acid and the next one in its sequence. You know, these would be like the individual folds maybe you'd make in the Origami analogy, and you know, so there are certain shapes that proteins have within them, you know, the helices, beta sheets, you know, gamma turns all these things, and this has to do with kind of local interactions. That's sort of the lowest level of protein folding, then there's a question of how do those local structures fold together to form domains within the protein. So maybe if we're making an origami crane, like the head and the neck is the domain and the wing is a different domain. So how do we get the head and the neck out of individual folds? And then how do the head and the neck relate to each other to give the shape of the overall body to the protein. And then there's interactions beyond that. So proteins interact with each other to produce complexes, you know, in the same way that the different parts you'd put into an engine, you know, like the carburetor and the manifold or whatever. These are different elements, but they work together to produce the grand function of the object. Proteins are the same. You might make one protein. This is mixing metaphors, but you have you know, different origami pieces come together to make a you know, or agami scene, and that's Those are the sort of bioactive complexes that run the cell.

00:21:06
Speaker 2: So does that mean that we can predict it or we roughly can predict it?

00:21:10
Speaker 4: A sorry, yes, So we had this nice conceptual framework about these hierarchical levels, and we could in the past do some predictions about like when you would see an alpha helix emerge. This could be done computationally, and then it was considered a great mystery how you produce the overall shape, that sort of grand confirmation of a whole protein molecule. But then the great insight four or five years ago was alpha fold. This is a machine learning based prediction tool for protein folding, which is trained on data from you know, thousands of protein sequences that were figured out by experimental methods, and now for certain categories of proteins, particularly ones that are freely floating within the watery portions of a cell, we can go I had a quite good prediction of its confirmation just from its sequence. It required the tools of machine learning to get there. So, you know, like all kind of AI and machine learning methods, you can have an effective tool at making that prediction without understanding all of the relationships that the model has learned in order to make that prediction.

00:22:23
Speaker 2: So my understanding is that these techniques rely on a bunch of examples. As you say, we have an example of here's the DNA and we know what the protein is. Here's another example where we know the DNA, we know the protein, and machine learning essentially learns a mapping from the DNA to the folded protein. But as you say, knowing how to do it or knowing how to generate an answer, it doesn't always yield an insight. Have we been able to pick apart alpha fold in any way and learn something about the fundamental biochemistry or is it completely a useful black box?

00:22:53
Speaker 4: My understanding, it's more in the useful black box, because you know, it's fairly recent and when we got these mL objects, you know, I would say, like the one I encounter more in my own research is simulations of neural networks and brains more than these biochemical simulations. These objects have a comparable complexity to the biological object you're trying to study, So you can switch from studying the cell to switching the simulation of the cell. But in some ways use the same empirical approaches. You know, you snip off certain parts of the network, or you perturb it in different ways and see how it alters the prediction. But it can be similarly hard to gain insight about how these tools make their predictions as it can be to understand how the biology happens. So I would say that's ongoing. The extraction of principles out of these tools is ongoing.

00:23:52
Speaker 1: All right, Daniel and I are back from our fascinating discussion about protein folding, and we're ready for our first listener question.

00:24:01
Speaker 2: All right, Then this question comes from Mattias and it's about the dire Wolf episode.

00:24:10
Speaker 1: Kelly here, we didn't hear back from Matthias in time. That's okay. We all get super busy sometimes, so I'm going to read their question here. Hi, Kelly, and Daniel, I have a biology question. I'm listening to the dire Wolf episode and I have a question about DNA. What does it mean to decode DNA? I've read, for example, that the human genome is decoded. What exactly does this mean? What do we know when it's decoded? Do we know to change the hair color? You'd need to change those X parts in this specific way? I'm guessing not, Thanks Matthias. Okay, right, so this is a great question. And so I looked through a bunch of different sites to try to figure out what de coded means, and frankly, it meant different things depending on what site I was looking at. I feel like Yeah, decoded is just kind of a word that means like we kind of figured it out.

00:25:13
Speaker 2: I think the fusziness in my mind is does d coode just mean we did that first step the part that RNA is doing, where we said like, oh, it's an A and then a G and then a T and then a C, which already is cool because like we didn't know that, right, we didn't know what the sequence was for a particular human Or does it mean we took the second step of saying and we know what this T is for, and we know what this G is for, and we know how it all bills together to make your eyeball, Which part of it have we done? What do we mean by decoded?

00:25:42
Speaker 1: Yeah? So, as far as I could tell, dcoded is not really actually a technical term. Like when I was looking, I saw it used to mean something like we now know the string of AC's, t's and g's, like we figured out the sequences of those letters that make up the human genome. Or we know that the DNA has been transcribed into RNA and we know what the sequence is in the RNA, so that's the stage where it was copied, or that the RNA has been translated into amino acids, and now we know what the sequence of amino acids is, and so we decoded it into the protein and so now we kind of know what the protein sequence is, and so I saw it used for each one of those three steps.

00:26:24
Speaker 2: And as Ben has explained to us so nicely, those are big differences, right, Yeah, just knowing the sequence is very different from knowing the proteins. Right, there's a huge.

00:26:32
Speaker 1: Leap there, Yeah, huge leap. But then the second part of the question was when something has been decoded, what do you know? And what do you know depends a lot on what you knew before you did the decoding. So, for example, if you got somebody's genetic sequence and they wanted to know, you know, hey, my mom had Huntington's disease? Do I have Huntington's disease? And you decoded the sequence of their genome and you had all this background information, you could tell them if they were likely to get Huntingtons or not, because we know all this background information. And so if you have the sequence, then you know a lot of information. But if you went to an alien planet and they happened to use the same ATGC system, that we use and you got the sequence, you might not know anything. I mean, maybe you'd get lucky and they were they had a lot of the same genes that we had. But like you'd have to do a lot of work to figure out what those sequences mean and what those genes are doing. And so just having the sequences doesn't really tell you where the genes are, what those genes do, what the proteins do, and so you need a lot of background information. So with a dire wolf example, because we know what a lot of those genes do in wolves, just having the sequences meant we could find a lot of genes and we knew what a lot of the proteins do because we know what they do in wolves, or we know what they do in closely related mammals. So we can use a lot of information from other organisms to infer.

00:28:12
Speaker 2: A lot I see. So it's like somebody gives me a map of a brand new house. I've never seen before, but I know what a bathroom looks like, and I know what a living room looks like. I'm not a complete alien, and so it helps me navigate to the bits I need.

00:28:25
Speaker 1: That's right, yes, But just getting a genetic sequence from an organism that we've never seen before, totally unlike anything else. You'd really know nothing. So when we think, yeah, the first genetic sequence we ever got, we were probably like, oh, okay, we can do this, but what does it mean.

00:28:42
Speaker 2: I remember this sense when we were working on the human genome project that once we had this, so many things were going to be unlocked. We would finally have a map to this incredibly powerful territory. Do you think that that came to fruition afterwards or do you think that was a little bit of a disappointment.

00:28:58
Speaker 1: So when Zach and I wrote the chapter on precision medicine for Sunish, my sense was that there was quite a bit of disappointment right after we got the human genome, that there weren't a lot more problems that were solved immediately after that. But I do feel like over time, as we've learned a lot more things and put a lot more pieces together, we have been able to solve more things. And I was talking to someone the other day about like the human connectome, where they're trying to like find the connections between all of the nerves in the human brain, and I was like, come, I mean, do we ever learn anything helpful from all of this stuff? And their take was that I was being a wet blanket surprise, and they were like, no, I do feel like out of all of these pieces of information, more, you know, a more complete picture is emerging and we are starting to be able to solve more problems. And yeah, and they felt like I was being negative, which I probably was, because come on, so I.

00:29:56
Speaker 2: Think A Matias is asking me, essentially, how useful is this? What does knowing this information mean? Does it just mean we have the letters and who cares? Or does it mean we now understand everything? And I think you're saying it's somewhere in between. It's more than we just have the letters, but it's definitely not that we understand everything about what it turns into.

00:30:15
Speaker 1: Yeah, yeah, I mean I think it's context dependent. With the dire Wolf, we probably knew quite a bit given the context we had for wolves, and if it's about humans, we probably also know quite a bit. But yes, it is probably somewhere in between. You don't know everything. It depends on what you knew before you got the sequence. Right, Okay, so let's take a break and when we come back, we are going to talk to Ben about what percent of our genes are associated with behavior? Okay, we're back, and in the third segments of today's episode, we're gonna be talking about innate behaviors. And so we wanted to start this second segment by asking Ben about what percent of our DNA is associated with behavior? And then I sort of hijacked the conversation because I've always been really interested in what percent of our DNA does nothing at all, And so I went ahead and asked him that as well. And so let's dig back into our conversation with Ben. Ben button, So, how much of our DNA is associated with behaviors? This is probably like an impossible thing to know for sure, Yeah, and maybe you have to be kind of handwavy, but like, yeah, what percent of the human genome is associated with what ends up being a behavior?

00:31:51
Speaker 4: A huge amount. And part of this is that genes have many many functions each, So a gene might be involved in coordinating heart rhythms and blood circulation. You might say, well, that doesn't have to do with behavior, but we also know that cardiac health has a huge amount to do with like the behaviors and organism exhibits that gene could also be used to pattern the heart during development, and used in a different context to pattern neural development. So genes get redeployed in different contexts and different organs at different times, all across the body, and so they have many phenotypes and depending on sort of how sensitively you measure behavior. I honestly think if you picked a random gene and disrupted it, there would be a high chance that you could observe some effect on probably any behavior you look at. Right, it's sort of this all to all mapping of genes and phenotypes when you have a sensitive enough instrument to measure it.

00:32:55
Speaker 1: How are we not getting more Nobel prizes? Been? Everything depends on this down.

00:33:00
Speaker 2: The particle physicists or pushing you guys out.

00:33:02
Speaker 1: That's right.

00:33:02
Speaker 4: There's a lot to figure out, and in other fields too, so they can they can get some prices.

00:33:10
Speaker 2: Such a nice answer. Oh my god.

00:33:13
Speaker 1: So when I was in undergrad I remember there was this headline that some really high percent of our DNA does nothing And is that do we still think that's true? Or does it have a function now? And if it does nothing, why do we still have it?

00:33:29
Speaker 4: It's complicated that's going to be the theme of these answers. So you know, a lot of that comes from experiments where you so there's kind of theoretical arguments of when we take a genome and we look at what kind of annotations we've given to it, segment by segment.

00:33:47
Speaker 1: What does annotations mean?

00:33:48
Speaker 4: Annotations means like, Okay, so you have a genome and you put it into a sequencer and you get that raw ATGC sequence and it's billions of base pairs or hundreds of millions, that's the genome sequence. Then there are processes by which you can take that sequence and these days, for the most part, computationally assign some meaning to different regions of that sequence. So imagine if you've got a book written in a language and in a script you're sort of unfamiliar with. You can see what the symbols are and what their sequence is, but to get meaning of, oh, well, this is a chapter and you know this is this part of the plot, and this is referring to a character in the story that's applying meaning to those sequence of symbols. You know, these days this is done drawing on the sort of vast body of molecular biology. Research that's been done already to say, oh, i'm you know, I'm looking at this sequence of symbols and it looks really similar to this one I've already seen in some other organism, and so I can make the prediction that it's referring to a certain biological process. And we know things about you know, how RNA polymerase, which is the enzyme that does transcription that converts DNA into RNA, recognizes the places in the genome where it should start to make a new RNA molecule. Like, we know something about the sequence character that are those transcriptional start sites, so we can the computer can go through and look at all those symbols and say, you know, I have found twenty thousand transcription start sites, and I know something about the sequences based on homology. So this is like shared characteristics by common dessent. So the fact that the sequence looks the same in one organism as another allows us to kind of transfer our knowledge from the one organism into the one that we're trying to characterize. So you get predictions about function that come from this annotation process. And part of the notion that there's a lot of functionless DNA is that much of this of the genome is not a signed of function through the sanitation process. The fraction of the genome which encodes for genes and proteins is small compared to the total genome. It's very species by species. But you know, in humans, I think it's like two and a half percent or something like that is genic and protein coding.

00:36:18
Speaker 1: In the rest over ninety percent, we don't know what it does.

00:36:22
Speaker 4: Well, we know many things that it does, right, So you know, I've talked about proteins being redeployed in different contexts during development and in adulthood, and the rules for how you reuse a protein in different contexts in what cells do you turn it on or turn it off? Those are given in the regulatory regions. So regulatory regions are a large part of that remaining ninety percent, and they provide the kind of context information about when do you deploy a gene. But then there's a whole bunch of intergenic stuff which is probably non regulatory, maybe or maybe not. And this is where the sort of crux of that debate is. And you know, some people will say that a segment of DNA has a function if I can assign any kind of biochemical response or characteristic to that sequence. So you know, DNA, in addition to just kind of encoding abstract information, is this physical object, right it You know, it has a certain length and it gets folded up to fit into a cell. And one of the ways in which you can like measure a biochemical impact or effect of a certain segment is to say, like, if we got rid of it, can we measure something different about the cell exactly? And so you know, it turns out in many cases you can get rid of it and you can measure that something has changed, like that physical molecule now packs differently into the cell. As an analogy, like imagine I've got two hundred feet of yarn and I need to kind of like squeeze it into a baseball or something like that. Right, that's probably approximately the density of DNA and the nucleus, And like the exact kind of folds you get. There are some consistencies from cell to cell, and some of it is going to be random, but you know, the sequence matters for how it folds, and so is that a function? Maybe you know, it's a separate question of like how meaningful is it to the organism, and the folks who are on the skeptical side and say there's a lot of junk DNA will look and say, yeah, you measured a biochemical impact of perturbing that sequence, but it didn't change the fitness of the animal. It still continues to live fine in the lab even though you disrupted it. So it's disposable in that sense.

00:38:44
Speaker 3: It's chunk in that sense.

00:38:46
Speaker 2: The way I think about it is I look at like somebody who has a messy office, like Kelly's husband Zach, and I wonder like, is it necessary that he has, like, you know, a Dorita's bag on the floor and a pile of you know, random stuff in the corner. Maybe that's essential for him to make his hilarious comics. If you cleaned it up, maybe he wouldn't be able to do with job. It's like participating in some indirect way, right. Is that a fair way to think about it?

00:39:09
Speaker 3: Yeah?

00:39:09
Speaker 4: I think that is fair.

00:39:12
Speaker 1: On so many levels. That's fair.

00:39:14
Speaker 5: Looking up the do.

00:39:15
Speaker 4: Reno bag is going to have maybe have a measurable impact, like maybe you know, a different snack bag replaces it because.

00:39:23
Speaker 2: You changed it.

00:39:25
Speaker 4: But again, like the overall fitness of the operation may not be impacted that much. There's there is kind of another category, and these are like the selfish DNA elements, and you know, so a substantial portion of our DNA, and I may need a fact check on this, maybe it's like ten percent something like that. Is these repetitive sequences that are these sort of ancient viruses that are still present in our DNA. They copy and paste themselves through their own active and passive processes, producing more copies of each other of themselves within the DNA. It's very likely these are not advantageous to people, but are the result of their own evolutionary process that plays out within our genome. It probably adds a little bit of a fitness drag, so it's like a little bit costly for each cell to replicate those selfish DNA elements. But they also get co opted through evolution to you know, offer useful value to the genome. So it's not obvious that you could get rid of them without impacting fitness.

00:40:29
Speaker 1: Can you give us an example of a useful thing, like a way we've co opted a virus?

00:40:33
Speaker 4: Yeah, like that information could be used for scaffolding. There's this phenomenon of telomeres, so more jargon.

00:40:41
Speaker 3: I'm sorry.

00:40:43
Speaker 4: Okay, So DNA, we haven't talked about chromosomes, but these are the like when you take a single molecule of DNA from end to end, that is a chromosome, right, and in deploid organism, So in organisms that have two copies of all their chromosomes, which include humans, we have forty six of these chromosomes in each nucleus, so it represents twenty three different chromosomes present in two copies each. And when we reproduce a chromosome, imagine proteins kind of marching along that sequence of DNA making another copy. They have this kind of awkward property that the machinery that replicates the DNA has a particular size. Let's just say it's like measured in the length of DNA nuclear bases. Let's say it's like fifty bases wide. So the core enzyme might be kind of in the middle of that fifty base wide object, so it has trouble like getting to the end of the DNA because like it'll fall off before the sort of core enzyme can copy the final trailing sequences of at the end of that chromosome molecule.

00:42:00
Speaker 2: Ok.

00:42:00
Speaker 4: Right, it's like we're into analogy. So it's like, if a road is a mile long and a car is eight feet long, the last like three or four feet of the road can't be used by the car because it's like it's going to fall off and ground itself when that first tire goes off the end of the road, right. And so if the car was making more road, it would produce a road that's like six feet shorter than the last road. And the next time you copy it, it's going to be six feet shorter than that, and the next time it's six feet shorter than that. So one of these kind of ancient viral mechanisms, it's thought, has provided this set of sequences which just get grown onto the end of chromosomes so that the copying machinery has some distance. It can run past the useful protein coding region of the genome, but without missing the end. And these are called tea and it's thought that they're derived from ancient viruses, so you know, there's a sort of a self replicating dynamic there. Telomeres tend to make more telomeres in some sense, it's wasteful, but it has been co opted in the sense of now it provides this like buffer to the replication machinery.

00:43:17
Speaker 1: I didn't realize that telomeres were co opted viruses. That's pretty cool. Daniel and I are back for a moment. So, yes, I got us pretty far off track there talking about.

00:43:28
Speaker 2: Following your curiosity about this extraordinary universe.

00:43:31
Speaker 1: Yeah, okay, thank you, thank you. I appreciate your support. But bored, let's let's get a little bit back on track here. One of the main things we brought Ben on the show to talk about was genetics and behavior, and Ben runs a fruitfly lab looking at genetics and behavior in fruitflies, and so let's jump back into our conversation with Ben and ask him about what he studies in his lab. Ben Interview, button pressed. So you study fruit flies. What kind of behaviors in the fruit flies do you study?

00:44:04
Speaker 4: Well, we're interested abstractly in variation and behavior, right, So why do individuals behave differently? Why do species behave differently? And we pick and choose behaviors based on their utility and answering other kinds of questions. So we study a lot of different behaviors in the lab. We look at spontaneous walking behaviors. Do animals tend to turn left or right? For example? We look at temperature preferences. Do they like to hang out where it's warm or where it's cold. We look at odor preferences. Do they like the smell of bananas versus apples? We look at learned behaviors. So if we give them training associations to say that a particular odor proceeds getting shocked, and then they will learn to avoid that odor. We look at how that varies across individuals. So the choice of the behavior is motivated by the kind of question we're asking. When we ask questions about the genetic basis of behavior, we love to take advantage of the experimental tools that are available in fruit flies. These are animals that folks have been studying in the lab for over a century, and there's this great legacy of tens of thousands of different kinds of flies that have different genetic variations and alterations. And it's basically a lego toolkit for designing experiments like if I want to manipulate the activity of a single neuron to see how does that impact a behavior? Odds are good that there's a tool sitting out there to target that individual neuron and do a specific manipulation to it, so you can kind of mix and match these genetic reagents.

00:45:46
Speaker 2: Sorry, how many neurons does the fruitfly have? It's not as simple as a worm that has like one hundred, right, a.

00:45:51
Speaker 4: Fruit flying has about one hundred and fifty thousand neurons in its head and another fifty to seventy thousand in its ventral nerve cord, which is like it's spinal cord.

00:46:02
Speaker 2: And so out of that huge number of neurons you can target an individual one. They have like names. This is like this is Jack, and there's a similar one in every fruit fly that like does the same thing.

00:46:12
Speaker 4: Yes, essentially, maybe the slight caveat on that is much of that count of one hundred and fifty central one hundred and fifty thousand central neurons is in the optic lobes where you know, the animals have a retina which has like eight hundred facets. It's a compound eye, so it sees the world through eight hundred different lenses arranged in this hemispheric array. Each of those like points of visual orientation in the visual world has a whole sort of computational circuit that's downstream of it. So there are sort of computational units that are repeated eight hundred times, and it might be hard to target, you know, omatidium seven thirty five versus seven thirty four with a differentic reagent, because those are just set up through the development of the eye to be repeating units. And there's not much genetically that distinguishes, you know, a neuron performing one computational element at one of those positions from the next position over. But within the central brain, where there's less kind of just sort of copy paste of the neural circuit template the way there is in the visual system, then yes, it's more or less like you're saying that there's a neuron you can find across individuals, and we have ways to access it genetically.

00:47:35
Speaker 2: I think that's really fascinating because it already suggests that like the structure of the brain is determined by the genome. I know a little bit about the fly brain randomly because one of my oldest friends worked on the connectum Mala Murthy at Princeton. Oh yeah, yeah, and so it's incredible to me that like so much of the structure of the brain is determined at the gene level. I think a lot of people imagine that, like your brain just sort of like turns out this way or that way. Obviously my brain is different from Kelly's brain. Fruit Flies must have individuality also, since essentially sounds like you're telling me they're basically robots.

00:48:10
Speaker 4: Well, I'll start by doing a shout out Tomala Andmala and I are recently co authors on a new CONNECTO, which is the first example of a nervous system where we have the brain and the venom nerve cords still attached to it tool and we've done the electron microscopy reconstruction of all of the central nervous system from a single animal.

00:48:31
Speaker 1: Awesome.

00:48:31
Speaker 4: Well, yes, so the notion that you can find the same neuron across individuals definitely supports this intuition that they are robot like and that they're all identical. And that's part of the compelling paradox for us of why do genetically identical flies, raised and standardized environments behave differently from each other, especially when the components of their brains are the same.

00:48:57
Speaker 3: In this way, you know.

00:48:59
Speaker 4: We think that the details matter and that there is individual variation in how the neurons connect to each other. There also is some variation animal to animal in the cell type composition. So, for example, as part of the olfactory system, which is how the animals smell, you have a layer of cells out in their nose. Essentially it's their antenna, and they all send their axons so that like neurons, convey information from their input side, which is their dendrite side, to the output side of the neuron, which is the axon, So that's how they send information to other parts of the brain. The olfactory cells in the antenna send their axons into the brain to deliver odor information into the brain. And the sort of the neuron that comes second in that hierarchy, after those neurons that are in the antenna that directly sense the molecules. The next neuron that comes in the next layer is a thing called a projection neuron, and projection neurons pool the information from all of the olfactory neurons that express a single odorant receptor, so they are sort of signals of how much signal am I getting from the environment in a particular signaling channel.

00:50:16
Speaker 5: For the presence of voters, Let's.

00:50:17
Speaker 4: Say generally there might be two projection neurons for a particular category of olfactory receptor neurons. But we've seen in these connectomic studies where you take a brain and you dissect it and reconstruct neurons very precisely with electron microscopy, that sometimes there's three or sometimes there's one. So you can address them all genetically and you can find that same corresponding projection neuron from animal to animal, but sometimes there's a variation in exactly how many there are, and then there's the variation in how they connect differently into the neural network of the brain. And this is where we think a lot of the magic happens of individual is that it's in these variations in the connections. Another way to think about this is that like DNA is providing this set of blueprints for the animal to build itself during development. Right, you know, flies lay eggs and then you know those eggs are allotted a certain amount of nutrient to use to grow, and they use the energy in that nutrient to produce cells and pat in their body by implementing the instructions that come in the DNA. But just the same way that if you had some Ikea instructions on how to build a house. If you implemented it twice, you'd probably get slightly different outcomes. You know, you would lose some bolts and you'd make do with some other fastener, or the board is slightly warped that you install and so everything is shifted by a few millimeters. You know, you just wouldn't expect that the process of going through all those mechanical steps produces the same object in the end. And that's where we think the sort of detailed difference is the wiring of the brain are producing this variation in behavior.

00:52:03
Speaker 1: So the projector neurons, sometimes you have one, sometimes you have three, and we think that's not because of like a mutation that results in three. It's because that's right, mistakes are made along the way. A fastener was missing, the board was warped.

00:52:17
Speaker 4: It's very reasonable to think it might reflect a mutation. So animals, in all creatures, in all circumstances, acquire mutations. It's inevitable. It happens even in the lab. But you can test this, so you can you know, you can take that individual which happened to only have two or had two instead of three, let's say, or at three instead of two and made it to another individual that had three instead of two, and ask do their progeny show this, Do they show that they also have three projection neurons?

00:52:49
Speaker 5: And they don't.

00:52:51
Speaker 4: So traits in this kind of category seem to have low heritability, at least in the context of our lab studies. So that's saying that there was not a genetic difference between the animal that had three projection neurons and the one that had two, because otherwise it would have passed that trade onto its prochny.

00:53:11
Speaker 1: And Daniel and I are back, okay. So that perfectly sets up the questions that we have from the listeners for the last segment of the show, and so when we get back, we will jump into those listener questions and we're back. We've been talking about genes and behavior with Ben de Bevore from Harvard. We've chatted about how genes are associated with behavior in lots of ways, and I remember when I was in grad school. We used to hope that we were going to like a gene that was associated with a behavior and that they would be like this one gene that would break everything open. But over time it actually looks like in lots of cases, it's like tens or maybe hundreds of genes that are influencing behavior, like and each one has like a small effect on the behavior, and it not quite as easy as we were hoping. And that might also be why I ran away screaming from genetics a little bit, because it was more complicated than I thought. But one of the nice things about studying fruit flies is that you can actually kind of start to get a handle on things. And so Ben was talking about how in fruit flies we know where essentially the neurons are going to end up, and we know like what genes are influencing where the neurons are going to end up. And because he works on this model organism, they can kind of get they can kind of pin down the links between genes and neurons and behavior. And so with that background information in mind, let's go ahead and listen to questions that we got from listeners. And we actually got these questions after the initial interview we did with Ben, and so I sort of pieced this episode together after that interview because I was like, oh, these need to all go together. So anyway, let's listen to the question from Jim, and the question from Mark about innate behaviors.

00:55:16
Speaker 5: Here we go, Hi, Daniel and Kelly. I've always read that DNA isn't so much a blueprint for an organism as it is a recipe. By analogy, you make chalk chip cookies. The directions don't say exactly where or place each chip. Instead, they just say, add so much of this, add so much of that. Stuart, so many times put an oven at a certain temperature for so long and then a cookie like thing comes out at the end. That makes sense to me, for how the body plan unfolds and how self structures can self assemble. I can even understand how brain can form. Mostly is a blank slate, and that subsequent experience will prune the network to develop a human brain. But the question I've always had is this innate behaviors require something more like an exacting blueprint than a recipe. Do we know how DNA can code the development of these innate behaviors with such precision. Looking at spiders, it's wired to know how to navigate the world, avoid predators. It select a good place to build a web, build a web, wrap up the food, find a bait. All of small amount of DNA. From an information theory point of view, I'm a programmer type. It seems like the amount of information which is encoded in that network must take a large part of the spider's DNA. So what's the trick? How does nature pull that one off? Do developmental biologists know? And thanks for all you do.

00:56:29
Speaker 3: Hello, Kelly and Daniel. This is Mark in Honolulu. I would like to learn about instinctive versus learned behavior. My questions include how much of a human's behavior is instinctive compared to that of say, a babboon or a goose or a frog. Exactly where in an organism are the instruction stored in individual genes? Are they similar to a computer programs? If then commands, could gene manipulation be used to transfer instincts between organisms in sort of rooster gene and me and a crow in the morning. I'm curious is my curiosity instinctive or learned? Thank you?

00:57:18
Speaker 2: Oh wow? These are great questions, and I imagine the answer is we don't know, or it's complicated, or it depends or some mixture of those.

00:57:26
Speaker 1: Yeah, it was a mixture of them. And also the answer involves me banging my head on the table a lot, and it was like, oh, these are big ones, all right, so where to start. Let's let's start by but actually, let me back up and say, these are great questions. Yeah, they're just questions that are like not super straightforward to answer.

00:57:47
Speaker 2: So let's start by defining exactly what we mean by an innate behavior and why is it a useful distinction?

00:57:52
Speaker 1: Yeah, so these are behaviors that you do without needing to learn them. And so, as the listener pointed out, like, so you know, spiders are born, and then often what happens is they you know, like they hatch out of their egg, they release a string like a web out of their rear end. That web gets picked up by the wind. They move away from their mother, they never see her making a web, and they just know how to make their own web on their own.

00:58:22
Speaker 2: That moment, when spiders like fly using their web, that's so cool. It is so cool, incredible.

00:58:28
Speaker 1: There was a day when the light was just right and I actually saw some of them. I think it's called ballooning. I saw some of them like just moving past me, and they're so tiny, and I was like, you all are starting your lives. Good luck and it was just anyway, it's very cool, very cool that they can do that.

00:58:45
Speaker 2: But your point is they've had no parent answer, and so they're out on their own. They can only rely on their innate behaviors what they come programmed with.

00:58:53
Speaker 1: That's right, yes, And so the question is how the heck do they do that? And that is kind of amazing because different spie make different kinds of webs, you know, like if you've see in a black widow, they make these webs that just kind of look like a mess on the ground. Or then there's orb weaving spiders, and then there's a humped trash line spiders, which are some of my favorite spiders. We've got those on the farm. What a great name for a web. But anyway they have, you know, they all have their own unique web that they know how to make without anyone teaching them how to do it. And so how does that happen is the question? And for a lot of these the answer is, I like, I don't think we know. So I started my search with how do spiders know how to make their webs? And I wasn't able to find a link between like, you know, this gene creates these neurons creates this web building behavior. I mean I was able to find like videos where they watch the different steps that the spiders take to make their webs, and so they were able to break it down into like, Okay, first the spider goes and they like find all of the different attach atachment points in an area, and then they randomly like connect all of the different attachment points, and then they start doing circles to like connect all of those things. And so they're like trying to break it down into steps. And you need to know those steps because each one of those steps might have a different like neural circuit connected to it. Yeah, and so breaking down the steps is the first step.

01:00:21
Speaker 2: But it doesn't tell us how the spider knows the step.

01:00:24
Speaker 1: That's right, right, and so I and maybe I missed it, but I wasn't. I don't think that in spiders we have worked out, like you know, how the genetics of web building at the innate level. But in organisms like fruit flies, we have a little bit more information. And so some of the behaviors that have been looked at in terms of like innate behaviors are like sex differences. So for example, often what happens is that, like, you know, very soon after eggs are fertilized, it's clear if an organist is going to become genetically male or female, and so you get like the production of you know, sex organs, and those sex organs start producing hormones, and those hormones start impacting the development of the brain, and male brains and female brains will in some cases develop differently and they'll start developing different neural circuits, so different connections between different neurons and different parts of the brain. Some brain regions will be bigger or smaller than other brain regions, and then when they reach sexual maturity, these neural circuits will essentially come online. So, for example, in fruit flies, the males have very specific stereotypical like dances they'll do for the females to try to attract them that they don't need to learn. They just know how to do them because they have these neural circuits that get activated when they become sexually mature, and females have very stereotypical responses to the males. Usually they reject them initially, is how I read it, And so the answer is that you know, these these innate behaviors are things that are essentially coded in genes, and then that impacts neural circuits in the brain that then get turned on at the right time when the individuals are older. And so I think it was Mark that had a question that was like, how is the information for these behaviors stored? And I think the answer is that it's stored in the genes and in the connections between nerves in the brain.

01:02:29
Speaker 2: Wow.

01:02:29
Speaker 1: And then Jim, who's a programmer, had a question about is this more like a blueprint or more like a recipe?

01:02:37
Speaker 2: Yeah?

01:02:38
Speaker 1: And you know, metaphors always fall apart when you're talking about biology. You know, they just capture some aspect of the system that you thought was helpful when you developed the metaphor, but when you dig in, maybe you discover the metaphor isn't perfect. But this sounds more blue printy than chocolate chip recipee to me. You know. It's like when we were talking to Ben about where the neuralurons are, it sounded like he knew exactly where the neurons are for the most case, Like it was very specific. So like, yes, sometimes you're missing a screw and it doesn't go exactly the way you thought it was going to be And so he was talking about how sometimes you think you're going to get one neuron in one place, but instead you get to you know, as someone who has tried to follow blueprints and has messed up a lot, has ended up with like things not exactly where they thought they were going to go. I can I can definitely understand that, but but this sounds more blueprinty to me.

01:03:33
Speaker 2: It seems like that's the only way it could work. I mean, if you just have some sort of like general instructions like sprinkling chocolate chips, you're not going to get the same web across spiders that are in the same.

01:03:44
Speaker 1: Species, right, Yes, yeah, that's right. But on the other hand, you do need some flexibility. So like when I was talking about how they were documenting the different steps that were taken in web web building. So like you know, when when a spider goes to build a web, they don't always have the exact same connection points, you know, so they like, if they're building a web in a tree, there's going to be branches at different spots, and so they need to be flexible in where the web is going to start, you know, so it's not always going to be exactly the same, and so they need to be flexible to some extent. So you know, maybe they always need five connection points and that's like five chocolate chips or something, and so like it's somewhere between I think a blueprint and a recipe in that you do need to be a little bit flexible, cause you don't know you're not going to have exactly the conditions you want. Probably. But I've got another example that's a little bit more concrete maybe and might help us think through this a bit. And this example was inspired by the question could gene manipulations be used to transfer instincts between organisms? And I took the word instincts to actually mean something more like innate behaviors.

01:05:00
Speaker 2: So this is like, can you take the spider web instinct out of a spider and put it in a person and then they can crouchet that same web or something?

01:05:08
Speaker 1: Yeah, And so I think that when you are trying to make big jumps like that, the answer is probably no, almost certainly, absolutely no, because a lot of these behaviors are controlled by many, many genes, And so I think if you can't transfer lots of different genes without messing the system up, right. But if you have two closely related species that differ for something that's like on a gradient, maybe you can transfer some things around. And so an example that I like is that there are prairie voles whose Latin name I won't say for Daniel's sake, and they're monogamous, and actually from here on out will just call them the monogamous voles.

01:05:49
Speaker 2: Nice.

01:05:50
Speaker 1: And then there's polygamous mountain voles, and so there are voles where they pair up for a long time, they raise kids together, and if you put them in cages in the lab for twenty four hours and they breed, they will form a strong pair bond. They'll huddle together. And if you give them a choice between some other vole in the lab or the partner that they made it with yesterday, they'll keep choosing the partner they made it with yesterday.

01:06:17
Speaker 2: How long did it take to form that bond?

01:06:19
Speaker 1: Twenty four hours?

01:06:20
Speaker 2: Twenty four hours? Yeah, your first date and done?

01:06:23
Speaker 1: Huh, Well, you know, twenty four hours and they are bonded. But the polygamous voles, you put them together for twenty four hours doesn't really impact how which partner they choose the next day. If you give them a choice between the one that they were with for twenty four hours the day before or a new one. The male's like, oh hey, someone new that's awesome, and like if they're out in the wild, they like diversity of partners.

01:06:51
Speaker 2: So these are two different instinctive behaviors monogamy or polygamy.

01:06:55
Speaker 1: Yes, and in voles that are closely related. What we think is happening here is we think that this has to do with what's happening with vasopressin. So Vasopressin is a neuro peptide. It's like a hormone in their brain. I'm not going to go into too much detail, but it's a hormone that we think it binds the receptors in a certain part of their brain and it gives them a reward for affiliating with a particular partner. Interesting, and the monogamous vowles have more of a receptor and a certain part of their brain. So we think that when they are bonding with this one particular partner, they feel better. So being monogamous just gives them a bigger reward, and so they want to stay with that one partner because just being with that one partner makes them feel better, whereas the polygamous male gets less of a reward hit for being with that one partner, and so they so more wild oats.

01:07:53
Speaker 2: So they're all just narcissists though They're all just doing.

01:07:55
Speaker 1: What feels good, aren't we all.

01:08:04
Speaker 2: We're all just voles deep down.

01:08:05
Speaker 1: I guess we're all just obeying our instincts. Perhaps, but anyway, So this was an experiment that was done in like two thousand and four by limit All, and what they did was they manipulated the receptors in the polygamous voles, and they increased the number of receptors in their brain in that particular area, and the polygamous voles that had more receptors in this particular part of their brain became more monogamous. And so by tinkering with the receptors, which we think the number of receptors is like genetically coded for by having like these regions in the genome that say, like make more of the receptors, you were able to make a polygamous vole more monogamous. And so I think that you can change innate behaviors, whereas the innate behavior in this case is a tendency towards monogamy by just tinkering with something like the number of receptors in the brain.

01:09:02
Speaker 2: Does that make sense, Yeah, that does make sense.

01:09:05
Speaker 1: So I don't think that you could do something like make a human crow like a rooster at the crack of dawn, because moving a couple genes from a crow to a human is unlikely to have exactly the desired effect because we have a totally different like neuronal background, like totally different connections in our brain. It just wouldn't do the same thing. But if you have close to related speceis and you're just like moving dials up and down a little bit, you can you can probably do some tinkering. Does that make sense.

01:09:34
Speaker 2: Yeah, that's fascinating. I think that's solving this puzzle will really help us understand how the brain works, right, how it's put together, how it's essentially programmed by evolution. I think it'd be fascinating to figure out like which things are hard coded, which things are learned, if there's some fuzziness in between, I mean, like reverse engineering any system is always fascinating.

01:09:55
Speaker 1: Yeah, absolutely well. And then one of the questions was how much of human behavior is instinct relative to a baboon, a goose, a frog, et cetera. And you know, to your point, like, first of all, it's hard to even bin some of these behaviors. So things like aggression in males is to some extent innate, because males are more likely to be aggressive than females, and part of that is from, you know, the programming that happened when they were a fetus. But then also some of it is learned from, like, you know, the outcome of fights that they have when they're younger, and that tinkers with how aggressive they'll become when they're an adult. And so it's hard to bin some of these things as innate versus learned, because sometimes it's a mix of both of them. So if we can figure out, like how much of this stuff is like hardwired versus how much of it is learned, like, I think we're going to find that a lot of it is both, and some of that might help answer this question about like, you know, do humans have more hardwired behaviors relative to a goose or a frog. I don't think anyone has tried to like rank those things, because so many of these behaviors probably fit in more than one box, and it's just kind of hard to do all of that stuff. But you know, in general, I think we'd love to learn more about how behaviors are tied to genetics. Although I have to also say that when you start talking about how behaviors are tied to genetics in humans, it becomes especially fraught, and it, you know, very quickly becomes, yes, a difficult topic to talk about for a variety of reasons, back and off, and that's why I've stuck with fruit flies and bowls. So let's go ahead and see what Jim and Mark thought about those answers.

01:11:34
Speaker 5: Thank you Kelly for doing all that research into what I think is a really fascinating topic. Like most good answers, it raises even more questions than it answered. I guess I'll submit those later for a subsequent episode. Thanks again for all you do, and I look forward to hearing the rest of the episode.

01:11:50
Speaker 3: Thank you, Daniel, Kelly, and Ben. I now have a better understanding of innate behavior. We are fortunate to live on a planet with an enormous for right of life forms and behaviors. Let's continue to explore and learn more about this world, and while we're at it, the universe.

01:12:09
Speaker 1: Okay, and it wouldn't be an episode of Daniel and Kelly's Extraordinary Universe where we have a guest if Daniel didn't have an opportunity to ask our guest the alien question. So let's jump to our interview with Ben and see what he thinks how about aliens? All right? Is it alien question time? Daniel, go for it.

01:12:33
Speaker 2: How universal do you think our system of encoding information in DNA or some DNA like molecule is. If we're about to land on an alien planet and dig into alien life, what do you expect to find as the molecular basis of heredity on an alien planet?

01:12:51
Speaker 4: I would i this is a putting my forecaster hat on. Yeah, I would say I would very much go with like a informative prior here and just say fifty to fifty on DNA. You know, I think there are some first principle reasons that RNA will play a important role in early life. It's relatively easy to make as things go, it's reactive, it can do all these enzymatic functions. You know, then, whether you sort of find a way to stabilize RNA or find a different stable hereditary biopolymer or stumble into DNA. Maybe that's a bit of a contingent matter for each independent origin of life. But I you know, I think it's quite plausible that you would stumble into DNA more than once. I think you'd have a pretty radically different genetic code. I think the mapping of codons onto amino acids feels extremely arbitrary, so maybe you know that sort of second step might be different. But yeah, DNA seems plausible.

01:14:00
Speaker 2: High. That's a big number. I mean, if there's a lot of life out there, then that's that's a lot of DNA.

01:14:05
Speaker 3: Well, that's right.

01:14:06
Speaker 2: I'm not disagreeing. That's awesome. And maybe if I can ask you a more philosophical question, if personality is controlled or strongly influenced by genetics, what does that say about our responsibility for ourselves and our choices and who we are? If I'm a jerk, is that my fault? If it comes from my parents because they were jerks and they passed it on to me, can I take credit for being a nice guy because it's just in my genes? How does this inform our feeling about ourselves and our sense of responsibility for our behavior?

01:14:41
Speaker 3: You know what.

01:14:41
Speaker 4: I think you can have robust ethical frameworks whether the universe is deterministic or random or has free will in some sense, you know, like if we believe that one of the evolved purposes of punishment or positive rewarding is to alter behavior and of the sort of overall fitness of behavior, these things can happen in a deterministic framework, right, and they can happen in a framework that accommodates free will. So I think you can uncouple these questions. I do think that just as in the sort of psychological domain, there's something slightly liberating about not having to explain personality variations in these deterministic, genetic or environmental frameworks. I think it sometimes takes the burden off of people in explaining the way they are, to just say, no, there's not really a good reason. It's just dumb luck, and we don't have to think about it beyond that. We don't have to attribute it to something some quality our parents had or some childhood experience. So I think it sort of lightens the load in that sense. But ethically, I think it's a different access doesn't talk to that deeply.

01:16:00
Speaker 1: Thanks so much for being on the show Ben. We had a lot of fun learning about genetics.

01:16:03
Speaker 2: Happy to do it.

01:16:04
Speaker 3: Thank you all for having me.

01:16:12
Speaker 2: Thanks everybody for listening. Please go and do us a favor and rate the show on whatever podcast app you're using. It really helps people find us.

01:16:20
Speaker 1: Daniel and Kelly's Extraordinary Universe is edited by the amazing Matt Kesselman.

01:16:24
Speaker 2: He really is a wizard. You can also find us online on Blue Sky, Instagram, and x D and K Universe. Come engage with us.

01:16:34
Speaker 1: You can email us at questions at Danielankelly dot org. We really do want to hear from you, and you.

01:16:40
Speaker 2: Can find our website www dot danieland Kelly dot org, where you'll also find an invitation to join our discord where everybody comes and talks about the amazing universe.

01:16:51
Speaker 1: And we also have the most amazing moderators. This is an iHeart podcast. Thanks for joining us.