WEBVTT 00:00:03.517 --> 00:00:10.517 This is the Convergent Science Network podcast. Leading researchers in the domain 00:00:10.517 --> 00:00:16.757 of neuroscience, brain theory and technology are interviewed by Paul Vesure and Tony Prescott. 00:00:17.417 --> 00:00:22.197 It's Paul Vesure together with Tony Prescott of the Convergent Science Network 00:00:22.197 --> 00:00:26.317 podcast of the BCBT Summer School. cool. 00:00:26.817 --> 00:00:30.057 And today we're here with our speaker, Zoltan Molnar. 00:00:30.817 --> 00:00:35.797 And Zoltan was speaking about evolution and the development of the neocortex. 00:00:37.637 --> 00:00:44.397 So Zoltan, why the neocortex? Why do you think that's the right target to try 00:00:44.397 --> 00:00:45.377 to understand brain evolution? 00:00:46.817 --> 00:00:50.357 Zoltan Molnar So I think this issue has been around for quite a while. 00:00:50.417 --> 00:00:57.597 And in 1664, Thomas Willis started dissecting some brains, and then he noticed 00:00:57.597 --> 00:01:03.177 that when you compare a sheep brain with a human, the biggest difference was in the cerebral cortex. 00:01:03.397 --> 00:01:06.757 Brainstem, all the other regions were similar proportions. 00:01:07.657 --> 00:01:13.257 And he noticed that the cerebral cortex was much bigger in human in proportion, 00:01:13.577 --> 00:01:16.017 and also it was more convoluted. 00:01:16.097 --> 00:01:22.937 So again, in 1664, he looked at some patients with cognitive abnormalities, 00:01:23.277 --> 00:01:27.397 learning disabilities, and epilepsy. 00:01:27.737 --> 00:01:33.437 And he noticed that the cerebral cortex had extra foldings and a different shape. 00:01:33.537 --> 00:01:37.977 So he concluded that it must be the cerebral cortex, which is the substrate 00:01:37.977 --> 00:01:39.877 of the higher cognitive functions. 00:01:40.077 --> 00:01:44.377 So I believe that Willis got it absolutely right. 00:01:44.597 --> 00:01:49.157 So I think he put his finger on this. 00:01:49.337 --> 00:01:53.217 So we have a structure which really increased during evolution, 00:01:53.497 --> 00:01:58.097 and also even minor abnormalities in the cerebral cortex can have a huge, 00:01:58.237 --> 00:02:03.117 major impact on the cognitive abilities. 00:02:04.342 --> 00:02:12.102 Okay, but now just from some more global perspective, before we delve into the specifics of your work, 00:02:12.402 --> 00:02:16.902 you could argue, okay, this is maybe what they saw a few hundred years ago, 00:02:17.102 --> 00:02:19.362 given the tools they had at the time. 00:02:19.882 --> 00:02:24.322 But brain evolution is not necessarily only focusing on cortex. 00:02:24.522 --> 00:02:28.382 Cortex doesn't operate in isolation, right? There's also expansion of other 00:02:28.382 --> 00:02:35.202 structures, like for instance, basal ganglia as an example, your thalamus, and so on. 00:02:35.262 --> 00:02:38.862 So do you see any critical dependencies there? 00:02:39.162 --> 00:02:47.702 Or for your money, you would really emphasize neocortex as the key source of 00:02:47.702 --> 00:02:51.522 what advanced our functionality as humans? Hmm. 00:02:52.042 --> 00:02:57.422 So I think we didn't have enough detailed studies where you, 00:02:57.542 --> 00:03:06.082 for instance, scan in lots of different brains in 3D, and then you allocate names. 00:03:06.082 --> 00:03:16.602 So for instance, parts of basal ganglia or cerebellum or hippocampus or different thalamic nuclei, 00:03:16.822 --> 00:03:21.642 you know, higher order or sensory nuclei and correlate it to cognitive function. 00:03:22.545 --> 00:03:26.685 And I don't think it has been done properly. So I think that would be also interesting 00:03:26.685 --> 00:03:33.785 to extend these studies, not only to cortex or specific regions of the cortex, 00:03:33.925 --> 00:03:35.925 but also to other extracortical areas. 00:03:36.165 --> 00:03:40.245 We all know that, for instance, the cerebellum is responsible for all sorts 00:03:40.245 --> 00:03:46.825 of other functions, cognitive, language, motor automatisms, etc. 00:03:47.105 --> 00:03:51.385 So probably the cerebellum had to evolve as well, or its connections. 00:03:52.025 --> 00:03:57.045 Interestingly, if you look at the hippocampus, I don't think that the size varies too much, no? 00:03:57.625 --> 00:04:00.725 Do you think? If you look at different species. In food-storing animals, 00:04:01.005 --> 00:04:02.365 it's bigger. It's really bigger. 00:04:02.725 --> 00:04:07.865 They have to remember where they put the... Sometimes I wish my hippocampus 00:04:07.865 --> 00:04:10.085 would be a bit bigger to find my keys or phone. 00:04:11.345 --> 00:04:17.465 Right. But now, so the other thing you emphasized at the beginning of your talk 00:04:17.465 --> 00:04:22.645 was sort of the four elements of evolution that most people got wrong, 00:04:22.825 --> 00:04:25.785 or at least you pointed to one popular video on that, right? 00:04:26.225 --> 00:04:30.105 So if you would have to characterize brain evolution, what are then the key 00:04:30.105 --> 00:04:31.685 principles that are driving that? 00:04:34.085 --> 00:04:41.845 So the video I showed really was just helping us to get started on discussions 00:04:41.845 --> 00:04:43.685 cautions about the principles of evolution. 00:04:44.025 --> 00:04:48.645 Because if you think about it, we are sitting here with, you know. 00:04:50.105 --> 00:04:52.445 Eyes in the front with stereo vision. 00:04:52.585 --> 00:04:56.785 We have the brain developed as it is, but it's a miracle. 00:04:56.985 --> 00:05:04.665 Do you think that you could rewind this tape of evolution of a few million years 00:05:04.665 --> 00:05:05.685 and then you start again? 00:05:05.785 --> 00:05:08.345 Do you think we would end up with the same brain shape? 00:05:09.527 --> 00:05:13.347 Or sensory networks, or motor networks, that would be a very interesting test, 00:05:14.247 --> 00:05:18.047 that how efficient these networks are, no? 00:05:18.387 --> 00:05:19.887 Some people believe that you 00:05:19.887 --> 00:05:24.867 would probably end up with a similar organism. I'm not so sure. Why not? 00:05:25.927 --> 00:05:33.527 So, if you look at the cognitive function or some selected functions of a bird, 00:05:33.527 --> 00:05:35.927 it can be superior to us, no? 00:05:36.727 --> 00:05:43.807 So if they would have had some selective advantages, maybe they would dominate now, no? 00:05:44.107 --> 00:05:47.107 I think you could also look at events in history and say, well, 00:05:47.187 --> 00:05:53.727 if it wasn't for instance mass extinctions caused by objects colliding with Earth, 00:05:54.067 --> 00:05:59.187 then mammals maybe would never have succeeded to the extent they have. 00:05:59.187 --> 00:06:04.067 Have, if you go further back, maybe vertebrates would have been less successful, so yes. 00:06:04.107 --> 00:06:08.667 Or we wouldn't be here if, you know, a dinosaur would have sneezed while our 00:06:08.667 --> 00:06:10.367 ancestors were copulating, you know. 00:06:11.707 --> 00:06:17.347 So I think evolution is, it's a miracle that, you know, we are here now, 00:06:17.547 --> 00:06:23.667 and we have this kind of brain and this circuitry, but probably we could have 00:06:23.667 --> 00:06:26.267 evolved into very, very different directions. 00:06:26.747 --> 00:06:31.767 And that's an interesting task for you guys, because you want to look at these 00:06:31.767 --> 00:06:36.207 computational functions, and you look at the brain, and this is one possible 00:06:36.207 --> 00:06:38.987 solution to solve that problem. 00:06:39.127 --> 00:06:43.547 But you could probably solve the same problem with many other circuits, no? 00:06:43.747 --> 00:06:50.187 But with Sultan, I think, also in your presentation, you made a point that might 00:06:50.187 --> 00:06:54.667 contradict you, because you're also showing that many properties or several 00:06:54.667 --> 00:06:55.867 properties are highly conserved. 00:06:55.987 --> 00:07:00.767 Like, for instance, if you look at transcription during development, 00:07:01.167 --> 00:07:05.307 there are phases in this transcription that seem to be highly conserved. 00:07:05.387 --> 00:07:11.247 Or, for instance, the structuring of a cortical sheet in terms of a supergranular 00:07:11.247 --> 00:07:13.667 layer and a subgranular layer, right? 00:07:13.807 --> 00:07:17.887 So given that you have this conservation of certain design principles, 00:07:18.067 --> 00:07:23.707 you could also argue, well, that would suggest that if I would replay play the 00:07:23.707 --> 00:07:27.747 tape of evolution from a few million years back, 00:07:28.007 --> 00:07:30.707 these conserved properties would still be there and they'd still be, 00:07:30.747 --> 00:07:33.867 if you want, forcing a certain form of development. 00:07:34.027 --> 00:07:39.527 Of course, modulo properties of an environment, but I'm not sure if the end 00:07:39.527 --> 00:07:41.867 result would have been then radically different. 00:07:42.507 --> 00:07:51.507 I think the reason why we conserve developmental mechanisms is because they are there already. 00:07:52.874 --> 00:08:01.794 So we can't start from scratch, very radical changes in how we construct the brain. 00:08:01.914 --> 00:08:05.654 So we can tinker with it a little bit, but not radically new. 00:08:06.054 --> 00:08:12.894 So just to give another example, it's a very general principle that you generate 00:08:12.894 --> 00:08:16.794 the cerebral cortical cells in an inside-first, outside-lost fashion. 00:08:16.794 --> 00:08:22.034 But if you have one gene missing, the rilin expressed in the first layer in 00:08:22.034 --> 00:08:27.254 the Caja-Retche cells and they are not secreted, you reverse this inside-first, 00:08:27.334 --> 00:08:28.734 outside-lost gradient. 00:08:29.174 --> 00:08:32.774 You will end up with an outside-first, inside-lost reversed order. 00:08:33.154 --> 00:08:37.954 Yet this animal is surprisingly normal. It does have some motor dysfunction 00:08:37.954 --> 00:08:41.314 because it reels, that's where the name is coming from. 00:08:41.574 --> 00:08:45.154 But it's probably due to the cerebellar abnormalities rather than the cortex. 00:08:46.794 --> 00:08:50.894 What would be interesting to see, if you push this guy, realer, 00:08:50.914 --> 00:08:59.694 mutant, would that cause major problems in cognitive behavior? 00:08:59.954 --> 00:09:04.834 Unfortunately, in human, these are lethal very early on, intractable epilepsy 00:09:04.834 --> 00:09:08.614 and mental retardation, if you have a similar mutation. 00:09:08.614 --> 00:09:17.154 But what I'm saying is that if you just change or reverse this very conserved program, 00:09:17.554 --> 00:09:22.014 you still have some self-organizing principles, so they can still wire up, 00:09:22.214 --> 00:09:28.114 they can still process some information, maybe not as well as the normal brain, 00:09:28.234 --> 00:09:31.574 but our brain is quite plastic. 00:09:31.574 --> 00:09:37.674 It is also illustrating one of the, I think, fundamental properties that you 00:09:37.674 --> 00:09:41.974 also investigate in your research and described in great detail is that we all 00:09:41.974 --> 00:09:46.314 talk about, let's say, conserved network elements, 00:09:46.574 --> 00:09:53.254 regulatory networks that in turn become modulated in different ways to create 00:09:53.254 --> 00:09:55.234 a certain variability of brains. 00:09:55.234 --> 00:10:02.214 While, let's say, the core elements of these systems are actually the same, 00:10:02.254 --> 00:10:04.794 but it's the regulatory networks that are changing, and as a result, 00:10:04.814 --> 00:10:06.474 expression patterns change, right? 00:10:06.814 --> 00:10:10.374 Isn't that the key underlying feature we're looking at here? 00:10:10.754 --> 00:10:14.534 Yes, but then all these changes at some point... 00:10:15.470 --> 00:10:21.350 They have to provide you with an advantage during selection. 00:10:21.790 --> 00:10:27.350 So you set up a network and you process information, let's say visual information, 00:10:27.590 --> 00:10:31.530 and then you have several streams to process where, what. 00:10:32.590 --> 00:10:40.010 And then you suddenly recognize that, you know, a chief primate is angry with 00:10:40.010 --> 00:10:42.130 you, so you can run away and you're not selected out. 00:10:42.850 --> 00:10:44.570 So you have some advantages. 00:10:45.470 --> 00:10:49.870 If you have a sensory or motor system which operates more effectively. 00:10:50.410 --> 00:10:58.670 So, over millions of years, you somehow have to select the development of programs 00:10:58.670 --> 00:11:04.050 which is producing the best possible hardware on which you can then develop these circuits, no? 00:11:04.730 --> 00:11:09.410 So this is how I see it. I think you also pointed to some of the constraints 00:11:09.410 --> 00:11:11.230 that are due to evolutionary evolutionary history. 00:11:11.470 --> 00:11:15.370 I mean, the example you gave from Rudolf Raff's book, The Shape of Life. 00:11:16.210 --> 00:11:22.450 How at an early stage in development, organisms seem to converge on a similar 00:11:22.450 --> 00:11:27.390 pattern of organization, which in some sense may be a forced move. 00:11:27.550 --> 00:11:32.210 If you're going to be this kind of multi-celled creature, you have to go through 00:11:32.210 --> 00:11:35.190 this, is it the phenotypic stage, or what's it called. 00:11:35.890 --> 00:11:38.770 And maybe there are several of those in evolution. 00:11:39.110 --> 00:11:43.450 I think yesterday in Paul's talk, he was talking about the Cambrian explosion 00:11:43.450 --> 00:11:48.310 and how all the body plants, the major body plants, emerged then. 00:11:48.490 --> 00:11:52.990 And there have been no fundamental innovation in body plants since that time. 00:11:53.030 --> 00:11:56.470 It's all been evolution within certain body plants. 00:11:56.690 --> 00:11:59.390 And I guess what we know now 00:11:59.390 --> 00:12:02.830 about these gene networks is that these are very heavily conserved 00:12:02.830 --> 00:12:05.810 from the very early days of evolution that certain gene 00:12:05.810 --> 00:12:09.530 networks exist in all these different species vertebrate 00:12:09.530 --> 00:12:12.350 and invertebrate and are co-opted to do different 00:12:12.350 --> 00:12:16.110 tasks so that must be a very strong constraint on 00:12:16.110 --> 00:12:18.970 the parts of design space for brains and bodies 00:12:18.970 --> 00:12:22.350 that evolution can explore that possibly we're 00:12:22.350 --> 00:12:27.170 only exploring one very small area of design space and if you replay the tape 00:12:27.170 --> 00:12:32.150 and you change some essentially accidental events that have happened maybe we 00:12:32.150 --> 00:12:36.950 could have gone in another direction or maybe physics just requires uh that 00:12:36.950 --> 00:12:39.210 some of these things have to happen the way they have happened, 00:12:39.890 --> 00:12:45.150 no it also would mean then from that perspective that issue would like to develop 00:12:45.150 --> 00:12:53.330 a completely different set of animal species you would have to rewind up till the cambrian explosion. 00:12:55.584 --> 00:12:59.624 Because you have to go back to the point prior to just the definition of these 00:12:59.624 --> 00:13:01.384 body plans, because that's the big constraint. 00:13:01.864 --> 00:13:07.384 Perhaps even earlier, because the gene networks, even the basic components of 00:13:07.384 --> 00:13:08.464 the nervous system are there. 00:13:08.984 --> 00:13:12.064 Right, exactly right. In radial creatures before the cambria. 00:13:12.564 --> 00:13:13.724 What do you say about that, Sultan? 00:13:14.944 --> 00:13:19.004 I was thinking in more advanced stages of development, when you already have 00:13:19.004 --> 00:13:22.604 some kind of tenencephalic vesicle, and then you have sub-partitioning. 00:13:22.604 --> 00:13:28.244 That's very conserved in all vertebrates and mammals. 00:13:28.504 --> 00:13:36.284 But if we go a bit further, then that will be the challenge to identify gene 00:13:36.284 --> 00:13:39.944 networks which will make us different from others. 00:13:41.284 --> 00:13:47.524 But now in your own research, you focus very much on cortex. 00:13:47.524 --> 00:13:52.504 And then what you emphasized very much is also this radial expansion of cortex 00:13:52.504 --> 00:13:58.344 or the growth of cortex and the processes that would drive that as compared 00:13:58.344 --> 00:14:01.144 to, for instance, a lizard or avian brain, right? 00:14:01.884 --> 00:14:08.224 If we just try to turn the question we're discussing now in a bit more data-oriented 00:14:08.224 --> 00:14:16.604 version, what are the most obvious differences between such a lizard, avian brain and... 00:14:17.432 --> 00:14:21.632 The mammalian brain, in which we will find this cortical sheet that you really study. 00:14:23.032 --> 00:14:27.592 I often have discussion with Barbara Finley about this. 00:14:28.392 --> 00:14:32.992 She is telling me that if you plot, 00:14:34.232 --> 00:14:43.452 brain size and other parameters, we have absolutely just the right size of brain as total. 00:14:44.592 --> 00:14:49.512 Other species would fit in just as well, birds included. 00:14:50.372 --> 00:14:56.472 Where we disagree is that within that scheme, I think the relative proportions 00:14:56.472 --> 00:15:00.252 of our brain, they change radically. 00:15:00.792 --> 00:15:08.932 So the sauropsids, as I showed you, decided to enlarge different parts of the brain than from us. 00:15:10.352 --> 00:15:16.532 Actually, we did not decide anything. We were selected out because it was advantageous 00:15:16.532 --> 00:15:18.212 to have the enlarged cortex. 00:15:18.732 --> 00:15:25.592 The dorsal ventricular ridge or the derivatives of these structures, 00:15:25.992 --> 00:15:30.432 whatever they are, you know, lateral amygdala, endoperiform nucleus or claustrum, 00:15:30.532 --> 00:15:33.252 they are not major features of our brain. 00:15:33.392 --> 00:15:35.652 So we decided to enlarge the cortex. 00:15:36.092 --> 00:15:41.432 Now, the sauropsids decided to enlarge that bowl of cells, and they have huge 00:15:41.432 --> 00:15:45.592 circuits, as we discussed during the talk, inside that bowl. 00:15:45.812 --> 00:15:49.312 And some of them, they look very similar to the mammalian circuits. 00:15:49.792 --> 00:15:54.132 And after the talk, I think Tony's comment was a very good one. 00:15:54.252 --> 00:16:00.132 So what is the selective advantage to build the cortex, like we do, memos. 00:16:00.212 --> 00:16:05.412 I think the major advantage is that you can produce some embryonic transient 00:16:05.412 --> 00:16:09.032 circuits, and the cortex is. 00:16:10.830 --> 00:16:15.130 Receiving and delivering information at the time when the elements are still 00:16:15.130 --> 00:16:18.450 constructed and still added to it. 00:16:18.610 --> 00:16:23.490 So I think this is where I see a huge advantage in having the kind of structure we have. 00:16:23.570 --> 00:16:28.690 So we have a transient platform below the cerebral cortex and things begin to 00:16:28.690 --> 00:16:33.410 line up just as we generate other cortical cells. 00:16:33.550 --> 00:16:36.670 So I think maybe this is a big, big, big advantage. 00:16:37.470 --> 00:16:39.850 I don't think we can push that point too far. I mean, I mean, 00:16:39.850 --> 00:16:43.170 if at an early stage in hominid evolution, 00:16:43.430 --> 00:16:47.570 our line had been wiped out, it might be today crows that are recording interviews 00:16:47.570 --> 00:16:53.850 about why their brains are better than those mammals that just scurry around. 00:16:55.610 --> 00:17:00.850 And it's interesting that birds and mammals both have enlarged forebrains, 00:17:00.950 --> 00:17:02.550 but they've enlarged in different ways. 00:17:02.810 --> 00:17:09.450 And from my understanding, the big debate is how are those similar and different? 00:17:09.830 --> 00:17:15.110 And what I got from your talk today actually was a very interesting idea that 00:17:15.110 --> 00:17:20.330 maybe there's more flexibility in the evolution of the brain than we realize 00:17:20.330 --> 00:17:25.030 because starting from very different underlying architectures, 00:17:25.030 --> 00:17:30.990 the avian forebrain and the mammalian forebrain, although they look very different, 00:17:31.130 --> 00:17:36.310 are perhaps converging on something similar in terms of functional architecture. 00:17:36.310 --> 00:17:39.670 Is that a fair reading, or am I exaggerating your position? 00:17:39.910 --> 00:17:44.590 Very much so, because we talked specifically about these thalamic recipient cells, 00:17:44.870 --> 00:17:49.330 and these have very similar transcriptomic networks, 00:17:49.450 --> 00:17:54.890 and we have demonstrated that you have a significant overlap between these networks 00:17:54.890 --> 00:18:00.170 in layer 4 and also in the nidopallium in the birds, although they come from different parts. 00:18:01.958 --> 00:18:09.718 I think thalamic influence can elicit all sorts of differentiation on these cells. 00:18:09.938 --> 00:18:13.958 So that would be already very interesting to see. Is it activity? 00:18:14.318 --> 00:18:21.998 Is it the pattern of activity? Or you might even deliver some amino acids maybe 00:18:21.998 --> 00:18:24.178 transneuronally from the retina. 00:18:24.418 --> 00:18:27.558 So all these should be explored a bit further. 00:18:27.718 --> 00:18:33.058 So there is an influence from the thalamic fibers and we don't know what it 00:18:33.058 --> 00:18:35.898 is exactly that you have these convergent networks. 00:18:36.158 --> 00:18:39.358 And this is just the first step of the circuit. 00:18:39.518 --> 00:18:47.178 So we don't even know where to take the next step, no, in the avian and the mammalian brain. 00:18:47.378 --> 00:18:51.078 But now in the comparison of these four brains, can you say something about 00:18:51.078 --> 00:18:58.078 the kinds of cell types you find in mammals and birds and how similar, dissimilar these are? 00:18:58.078 --> 00:19:07.398 So all our work I presented today was on regions, dissected pieces of our brain, not single cell. 00:19:07.678 --> 00:19:14.238 So I completely agree. What would be really good is to go into single cell level. 00:19:14.978 --> 00:19:21.498 And then cluster these cells according to the expressed gene, 00:19:21.658 --> 00:19:26.298 and that's one way of identifying cell types and relate this to somatodendritic 00:19:26.298 --> 00:19:28.518 morphology, physiological properties, connectivity, 00:19:28.858 --> 00:19:36.358 and then you come up with ideas whether these evolved together or separately. 00:19:36.878 --> 00:19:43.278 Yeah, because what you mentioned in this discussion is in some sense there was 00:19:43.278 --> 00:19:49.098 this idea of Harvey Carton expressed in the 60s about the equivalent circuit hypothesis. 00:19:49.658 --> 00:19:53.618 And And you have tried to target that by doing, if you want, 00:19:53.718 --> 00:19:59.738 gene fingerprinting of these areas to allow you to establish whether we look 00:19:59.738 --> 00:20:04.218 at homologues or not between bird and mammals. 00:20:06.858 --> 00:20:13.158 So in your opinion, if you look at these genetic markers, how similar, 00:20:13.218 --> 00:20:17.598 dissimilar do these structures look? would you declare them being homologues 00:20:17.598 --> 00:20:19.538 and that Harvey Carton had it right? 00:20:19.798 --> 00:20:22.838 Or do you think it's a bit more complicated story? 00:20:24.278 --> 00:20:27.478 So as I mentioned, if you compare, 00:20:28.418 --> 00:20:33.258 gene expression patterns between these structures, let's say hippocampus and 00:20:33.258 --> 00:20:37.918 striatum, you notice that you have huge overlaps in gene expression pattern 00:20:37.918 --> 00:20:40.378 and also these networks have common elements. 00:20:42.063 --> 00:20:49.463 If you then, and also I mentioned oligodendrocytes, and then the next one was 00:20:49.463 --> 00:20:55.023 very interestingly layer 4, and parts of the nidopallium where you have the thalamic targeting. 00:20:55.423 --> 00:20:59.263 Now, just because they have a similar gene expression pattern, 00:20:59.463 --> 00:21:02.663 and just in the light of that particular information, 00:21:02.843 --> 00:21:06.163 you can't say whether they are homologous or not, because we developmental mental 00:21:06.163 --> 00:21:11.743 biologists, we like to talk about homology when they come from the same piece 00:21:11.743 --> 00:21:15.203 of the neuroepithelium, and it's not the case. We know for sure. 00:21:15.463 --> 00:21:18.743 So this is more of a convergent gene expression pattern. 00:21:19.143 --> 00:21:23.143 Now, you could argue that, okay, so why do I say that the hippocampus is then 00:21:23.143 --> 00:21:25.783 homologous and the striatum is homologous, but not the layer 4? 00:21:26.063 --> 00:21:29.823 So you're absolutely right. Just by gene expression pattern, 00:21:29.943 --> 00:21:37.083 I can't say that that piece of neuroepithelium is homologous to the avian hippocampus. 00:21:37.083 --> 00:21:43.283 It's more of the origin, lineage, chrono relationship, and the gene expression 00:21:43.283 --> 00:21:45.983 prove that they are similar. 00:21:46.243 --> 00:21:50.123 But just by gene expression, you can't really say that these are homoerogens. 00:21:50.363 --> 00:21:55.663 But your definition of homology seems to be a little bit biased there because 00:21:55.663 --> 00:21:57.263 when we talk about homology, 00:21:57.443 --> 00:22:01.643 what we're really trying to get at, surely, is that a common ancestor had a 00:22:01.643 --> 00:22:07.743 structure and that these two species we're looking at now have inherited that 00:22:07.743 --> 00:22:09.923 structure and adapted it from that common ancestor. 00:22:10.183 --> 00:22:14.963 That's what we really want to mean by homology. And then your target... 00:22:14.963 --> 00:22:17.883 My definition is developmental rather than evolutionary. 00:22:17.903 --> 00:22:23.803 So that is the problem because, you know, I can't... 00:22:24.905 --> 00:22:28.225 Uh use that kind of definition in developmental 00:22:28.225 --> 00:22:31.005 terms because they are meaningless yeah but but for 00:22:31.005 --> 00:22:34.305 us so homology to you isn't an evolutionary concept 00:22:34.305 --> 00:22:41.405 at all it's a it's uh so developmentally if you talk about homology of of structures 00:22:41.405 --> 00:22:47.625 it means that they arrived from the same or progenitors from the same part relative 00:22:47.625 --> 00:22:51.805 position of the tissue and that's not the case for these structures. 00:22:51.985 --> 00:22:56.505 If you talk about that, yes, you had a thalamic recipient cell group, 00:22:58.165 --> 00:23:05.985 which was targeted by thalamic fibers and they then initiated some sensory experience, 00:23:06.285 --> 00:23:09.465 yes, they perform similar function. 00:23:10.185 --> 00:23:13.085 But would that be homology? 00:23:14.825 --> 00:23:21.505 Let's turn it around. In your definition, which is a more developmental of homology, 00:23:22.185 --> 00:23:26.065 aren't you actually also importing a Trojan horse? 00:23:26.185 --> 00:23:30.305 Because you said yourself that this 00:23:30.305 --> 00:23:40.545 19th century idea of development recapitulating evolution was not fully correct 00:23:40.545 --> 00:23:46.965 because you see divergence in development that you will not see if you recapitulate evolution. 00:23:47.325 --> 00:23:54.265 But now, if you insist that homology must be defined as originating in the same 00:23:54.265 --> 00:24:02.025 part of the neural tube, then you seem to imply that development does recapitulate phylogeny. 00:24:03.965 --> 00:24:09.425 No I don't imply that but I think you. 00:24:11.771 --> 00:24:18.951 This is just not the case, that these two cell groups are derived from the same 00:24:18.951 --> 00:24:22.111 part of neuroepithelium, and they were rejuggled. 00:24:22.431 --> 00:24:29.211 Yeah, but you cannot say that. Why not? I have all the lineage information. I can track the cells. 00:24:29.451 --> 00:24:34.211 I know where they come from, where they go, and they are completely separate. 00:24:34.531 --> 00:24:38.691 Now, let me clarify. Of course, you have freedom of speech. You can say whatever 00:24:38.691 --> 00:24:42.071 you want, okay? Okay, in that sense, we will respect whatever you say. 00:24:42.191 --> 00:24:47.691 But the point is that if you compare a developing bird brain and a developing 00:24:47.691 --> 00:24:54.031 mammalian brain, the timing of the developmental process will be rather different. 00:24:54.151 --> 00:24:58.771 So for instance, the neuroepithelium in which you will find your progenitor 00:24:58.771 --> 00:25:03.691 cells might start to divide at a much later stage in the mammalian brain than 00:25:03.691 --> 00:25:05.071 it does in the avian brain. 00:25:05.291 --> 00:25:09.171 And as a result, the actual layout of the neuroepithelium will have a very different shape. 00:25:09.951 --> 00:25:16.371 So to just then use that location at that point in time as your benchmark is 00:25:16.371 --> 00:25:20.711 a sure road to failure because the developmental program is radically different. 00:25:21.271 --> 00:25:28.251 Yes, but that's exactly what I'm saying, that if you follow the entire course of development, 00:25:28.471 --> 00:25:35.331 so you look at these different proportions of the mammalian or avian or reptilian 00:25:35.331 --> 00:25:37.311 telencephalic physicals. 00:25:37.311 --> 00:25:41.571 If you look at all the segments, all the neurogenesis, all the migration, 00:25:41.931 --> 00:25:47.111 all the clones you get from these and you trace the migration of these cells... 00:25:47.978 --> 00:25:53.178 I know that this is not just stage-dependent. They never, ever arrive from the same. 00:25:54.538 --> 00:25:58.298 So basically, that's why I'm so radically against what you are saying, 00:25:58.458 --> 00:26:01.498 that you know, oh, it might be just a timing effect. 00:26:02.218 --> 00:26:06.838 And if you look at different time, yes, they were coming from there. No. 00:26:07.198 --> 00:26:14.678 There is no evidence whatsoever that they ever arrive from the same progenitors 00:26:14.678 --> 00:26:19.298 and they just have a different migration. So they come from in very different relative positions. 00:26:19.518 --> 00:26:27.598 We have now markers, some basic homeobox genes which mark some of the segments 00:26:27.598 --> 00:26:32.398 of these and also we have some morphological features which help us and they 00:26:32.398 --> 00:26:34.318 are coming from different parts. 00:26:34.538 --> 00:26:41.478 And the avian brain adopted a strategy to amplify neurogenesis and change the 00:26:41.478 --> 00:26:45.118 migration of those regions whereas the mammalian is radically different. 00:26:45.118 --> 00:26:50.378 And that's why I think that's not a big problem. I don't care whether they are homologous or not. 00:26:51.458 --> 00:26:55.418 We now sorted that they are not. Well, you're making a strong claim here, 00:26:55.438 --> 00:27:01.518 which goes back, I think you were talking about Haeckel and his idea that very 00:27:01.518 --> 00:27:04.258 early on in embryology, we all look the same. 00:27:04.458 --> 00:27:10.618 And you're saying that there's a starting point in embryology where every vertebrate 00:27:10.618 --> 00:27:12.638 has essentially the same embryo. 00:27:12.658 --> 00:27:16.738 And from that point, you can track what happens to each bit of the embryo as 00:27:16.738 --> 00:27:19.958 it moves around. What I'm saying is that the telencephalic vesicle, 00:27:20.198 --> 00:27:25.218 so let's say I give you two or three sections. 00:27:26.355 --> 00:27:31.635 Matching stages from a turtle, chick, or a mouse. 00:27:31.875 --> 00:27:38.615 And I stain it with the same four representative Homo box genes, 00:27:38.895 --> 00:27:43.355 and I cover slip it, give you a microscope, and then I ask you, 00:27:43.395 --> 00:27:44.475 okay, tell me which one is which. 00:27:44.595 --> 00:27:48.195 You couldn't be able to tell me at an early stage. 00:27:48.455 --> 00:27:54.215 So in that sense, yes, Heckel was right that these early stages are highly conserved. 00:27:54.735 --> 00:27:59.055 But you know, they don't go through the, you know, later or especially the adult 00:27:59.055 --> 00:28:02.495 developmental stages to go to the next stage. That's where we disagree. 00:28:02.855 --> 00:28:06.335 But it seems a bit circular here because you're now using gene expression to 00:28:06.335 --> 00:28:09.435 say, this is my standard. 00:28:09.515 --> 00:28:14.535 This is the common pattern across all these creatures. And I know it because of the gene expression. 00:28:14.815 --> 00:28:19.235 But you're saying- Tony, you can use many other things. Gene expression is just one. 00:28:19.455 --> 00:28:26.575 But people usually accept gene expression quite well, because these are highly conserved. 00:28:26.755 --> 00:28:32.655 You never have a Hox gene changing its gene expression at these early stages. 00:28:33.855 --> 00:28:39.135 You're now saying when we come to the forebrain that gene expression seems to 00:28:39.135 --> 00:28:44.775 suggest that part of the dorsal ventricular ridge is similarly organized to 00:28:44.775 --> 00:28:47.635 neocortex, but you don't want to accept homology. 00:28:47.635 --> 00:28:50.535 No, no, no. The gene expression is not supporting that. 00:28:50.955 --> 00:28:57.655 These are very early studies from several groups, but basically if you look 00:28:57.655 --> 00:29:04.015 at these, you have a couple of key genes which mark a cortex, for instance. 00:29:04.175 --> 00:29:06.375 So that's the EMX genes. 00:29:06.735 --> 00:29:12.435 Then you go to this intermediate part, that's the Pax6. 00:29:12.995 --> 00:29:17.695 Then you have the DLX genes and so on. So So the order of these and the segments 00:29:17.695 --> 00:29:21.815 are highly conserved early on. I'm not saying that's the only criteria. 00:29:22.035 --> 00:29:28.235 You can use many other things probably, but this is clearly indicating that unfortunately. 00:29:30.758 --> 00:29:36.478 And cortex, developmentally, they come from different parts of the brain. 00:29:36.698 --> 00:29:39.478 But they converge to have the same pattern of gene expression. 00:29:39.798 --> 00:29:41.818 Exactly. Yeah, so finally! 00:29:42.878 --> 00:29:49.178 And that was a very interesting… And that's why I mentioned it in the talk, 00:29:49.358 --> 00:29:53.498 that that's why it was the thorniest question, the thorniest problem of. 00:29:54.798 --> 00:30:00.198 Evolutionary biology because the gene expression, hodology, and physiological 00:30:00.198 --> 00:30:03.478 properties, they all supported that they are homologous. 00:30:03.538 --> 00:30:06.598 But the developmental data is clear that they are not. 00:30:06.718 --> 00:30:10.038 But this is just as interesting. We need to be more flexible about what we mean by homology. 00:30:10.338 --> 00:30:12.798 Yeah, I think that's a problem. Because you're saying that at an early stage 00:30:12.798 --> 00:30:18.438 in the embryo, all embryos are essentially the same. And then you can see what goes on from there. 00:30:18.578 --> 00:30:22.318 But what we're saying now is that in the adult, although these have a very different 00:30:22.318 --> 00:30:26.378 developmental paths, they've converged on a very similar structure, 00:30:26.478 --> 00:30:27.758 at least in terms of genomes expression. 00:30:28.118 --> 00:30:31.718 So you could say that in the adult, they are homologous. Right. 00:30:31.898 --> 00:30:38.098 But in my talk, I mentioned that when I use similar terms, then I was warned 00:30:38.098 --> 00:30:43.818 by others, including Neuwenhuis, who is an expert in this, that he said, 00:30:43.918 --> 00:30:46.138 you should stick to one meaning of homology. 00:30:46.398 --> 00:30:48.538 And developmentally, they are not. 00:30:49.338 --> 00:30:53.658 Okay, so Tony, I'm afraid we will not convince Zoltanov during this podcast, 00:30:53.938 --> 00:30:58.718 but we will twist his arm afterwards. It took 60 years for Harvey, you know. 00:30:59.918 --> 00:31:04.138 No, but look, so we just, okay, so homology, at least I think this discussion 00:31:04.138 --> 00:31:09.138 makes clear, it's actually not completely resolved what we mean with it exactly. 00:31:09.398 --> 00:31:14.598 There is some controversy around its definition dependent on the field in which 00:31:14.598 --> 00:31:18.678 it is used. Exactly, especially your field has a problem because we know exactly. 00:31:19.078 --> 00:31:22.158 Okay. Now, this is exactly what we're talking to you, Zoltan, 00:31:22.158 --> 00:31:24.158 because you know and we don't, you see. 00:31:25.818 --> 00:31:29.658 No, but the point is that then if you talk about this…, 00:31:31.073 --> 00:31:36.973 this sort of equivalent circuit hypothesis now, okay? So the idea would be that 00:31:36.973 --> 00:31:44.553 functionally, this avian forebrain compared to a mammalian forebrain functionally 00:31:44.553 --> 00:31:45.953 might do similar things. 00:31:46.153 --> 00:31:50.853 It might show even, as you showed, right, similar response properties to visual stimuli. 00:31:51.673 --> 00:31:56.393 But in terms of its internal organization, it's different because in terms of 00:31:56.393 --> 00:32:01.313 its layering, it's different. So, how should I interpret that? 00:32:02.633 --> 00:32:07.853 Because, so in your talk, you showed that in cortex, mammalian cortex, 00:32:07.973 --> 00:32:12.253 you would exploit more the six layers of a mammalian cortex to perform certain operations. 00:32:12.253 --> 00:32:18.073 While in this avian brain, you might use a more, at least in this conceptualization 00:32:18.073 --> 00:32:19.633 also of Harvey Carton and others, 00:32:19.833 --> 00:32:24.753 a more, let's say, a sequential operation between different subzones, 00:32:24.753 --> 00:32:28.993 as opposed to exploiting parallel processing within layers. 00:32:29.373 --> 00:32:35.713 So how should I see then exactly this equivalence between these two anatomical structures? 00:32:35.713 --> 00:32:44.393 So, Harvey in the 60s suggested that if we just look at these circuits, 00:32:44.673 --> 00:32:49.313 the thalamic targeting, so the cells are in... 00:32:50.950 --> 00:32:56.750 Different arrangements. So the thalamic targeting cells are equivalent, let's say. 00:32:56.890 --> 00:33:00.730 Then the next step is the supracranial layers. 00:33:01.830 --> 00:33:04.750 That would be the next step. So they are located somewhere else. 00:33:04.950 --> 00:33:10.130 And then the final output layer is in the dorsal part of the dorsal cortex, 00:33:11.750 --> 00:33:14.230 hyperpolyum in sauropsids. 00:33:14.310 --> 00:33:19.850 And then you would have that in layer five and six in the mammalian cortex. 00:33:20.330 --> 00:33:26.730 So I mentioned two studies, one from Suzuki, from Hirata's lab, 00:33:26.890 --> 00:33:31.450 and another one from Jennifer Dugas Ford from Ragsdale's lab, 00:33:31.550 --> 00:33:34.570 when they looked at a handful of gene expressions, 00:33:35.190 --> 00:33:41.150 with the most interesting markers for the supragranular and infragranular cells, 00:33:41.350 --> 00:33:46.030 and that further confirmed this kind of segregation of circuit elements. 00:33:46.030 --> 00:33:49.330 And they supported Harvey's original idea. 00:33:49.570 --> 00:33:51.650 But now explain to me, there's something I don't understand. 00:33:53.270 --> 00:33:56.990 If you look at the mammalian cortex, you would have a continuous interaction 00:33:56.990 --> 00:34:00.490 throughout with thalamus, for instance, right? 00:34:00.710 --> 00:34:07.450 But now in this avian solution, you would only have a subzone that is sensitive to thalamic input. 00:34:07.650 --> 00:34:12.290 So do you see a patchy projection from a thalamus into that area? 00:34:12.590 --> 00:34:13.550 Because that's what it would imply. 00:34:14.250 --> 00:34:20.010 So I'm not an expert, unfortunately, in the thalamic projections in sauropsids, 00:34:20.170 --> 00:34:23.090 but they do have a parallel processing. 00:34:24.490 --> 00:34:29.290 So you have the collicular input is going to the nucleus rotundus, 00:34:29.390 --> 00:34:34.190 and then the nucleus rotundus will project then to a different part of the brain, 00:34:34.290 --> 00:34:36.370 to the dorsal cortex, rather than to DVR. 00:34:37.170 --> 00:34:41.930 Whereas the specific sensory nuclei, they go straight to the DVR. 00:34:43.451 --> 00:34:49.931 In mammals, of course, you have the first-order and higher-order thalamic nuclei, 00:34:49.951 --> 00:34:52.431 and they have completely different layer-specific innovation. 00:34:52.791 --> 00:34:57.331 The first-order target layer, mostly layer 4, although they go to all other 00:34:57.331 --> 00:35:03.271 layers, whereas the higher-order thalamic nuclei, they project mostly to layer 1. 00:35:04.311 --> 00:35:10.211 And also the output from the cortex itself is different to these different thalamic nuclei. 00:35:10.211 --> 00:35:15.931 So, from the primary sensory areas, you have layer 6 projection back to these, 00:35:16.171 --> 00:35:23.991 whereas to the higher order thalamic nuclei, most of the cortical input is coming from layer 5. 00:35:24.211 --> 00:35:30.711 So, if you have a little bit of mismatch between 6 and 5, then you can actually 00:35:30.711 --> 00:35:35.491 communicate cortico-cortical information via the thalamus. 00:35:35.491 --> 00:35:41.931 And that's an idea which Ray Guillory and Murray Sherman developed, and it's very powerful. 00:35:42.871 --> 00:35:49.931 And that's also indicating that you have to coordinate cortical development 00:35:49.931 --> 00:35:51.471 and evolution with the thalamus. 00:35:51.651 --> 00:35:56.671 And this is not really followed up in comparative sense. So that would be interesting to see. 00:35:57.571 --> 00:36:01.471 And I think that's what you asked. Yeah, that's a test of hypothesis then, right? 00:36:01.531 --> 00:36:05.571 There will be a consequence of this equivalent hypothesis. But you're suggesting 00:36:05.571 --> 00:36:11.311 in the talk that DVR neurons were having some very similar functional properties 00:36:11.311 --> 00:36:13.531 to, for instance, visual cortex neurons. 00:36:13.991 --> 00:36:19.051 Electrophysiologically, if you Paul Menger recorded from the iguana dorsal ventricular 00:36:19.051 --> 00:36:23.671 ridge, and it was interesting to see that the receptive field properties were 00:36:23.671 --> 00:36:26.811 very similar to mammalian primary visual cortex. 00:36:26.811 --> 00:36:33.591 And also you had multiple representations within the DVR with mirror reversals, 00:36:33.671 --> 00:36:35.311 which is also very interesting. 00:36:35.491 --> 00:36:40.571 So even if it's a nuclear representation, you still have multiple representations. 00:36:41.948 --> 00:36:47.088 Which is also a feature of cortical visual fields. You have dozens of visual 00:36:47.088 --> 00:36:50.108 areas with multiple representations. 00:36:50.808 --> 00:36:56.048 And is there something like a DVR microcircuit, which might be similar to a 00:36:56.048 --> 00:36:58.828 cortical microcircuit? You're really pushing me here. 00:36:59.268 --> 00:37:06.348 But according to Harvey Carton, yes. So he can see lots of similar elements. What's that confirmed? 00:37:06.768 --> 00:37:12.028 Lots of similar elements. And I think this would be a really good time to tackle 00:37:12.028 --> 00:37:17.868 this in more detail, I think, this comparative circuit analysis. 00:37:18.668 --> 00:37:23.948 But now the other thing, in this comparison between a mammalian brain and an 00:37:23.948 --> 00:37:28.208 avian brain, you also look at the development of these structures, right? 00:37:28.208 --> 00:37:33.148 And the point you were making is that the migration patterns during development 00:37:33.148 --> 00:37:41.308 to build a cortex or to build this DVR structure are actually rather different. 00:37:41.808 --> 00:37:47.308 But so what are the similarities between these migration patterns of the cells 00:37:47.308 --> 00:37:50.008 that form these areas and what are the main differences? Yes. 00:37:52.069 --> 00:37:55.809 So I don't know too much about this corner of the brain, you know, 00:37:55.829 --> 00:37:57.309 where the DVR is coming from. 00:37:57.329 --> 00:38:00.969 I told you that this is one of the most complex regions, so you have a lateral 00:38:00.969 --> 00:38:10.009 stream of migrations, and then some authors like Luis Puelles, who is an expert here, 00:38:10.149 --> 00:38:12.509 is suggesting that as a field, 00:38:12.829 --> 00:38:20.349 they produce the claustrum, amygdala, and endoperiform nucleus, 00:38:20.729 --> 00:38:26.029 these regions as a field, and then they differentiate further from this stream. 00:38:26.289 --> 00:38:35.069 Whereas in sauropsids, you develop some developmental migration patterns or lack of it, 00:38:35.189 --> 00:38:40.689 and then these cells protrude into the lateral ventricle, and they form this 00:38:40.689 --> 00:38:42.049 dorsal ventricular ridge. 00:38:42.489 --> 00:38:47.049 Unfortunately, I don't know too much about the subdivision with this dorsal 00:38:47.049 --> 00:38:51.609 ventricular ridge, but that could be also very interesting to see how they differentiate. 00:38:52.029 --> 00:38:55.149 So this is a key feature which is different. 00:38:55.349 --> 00:39:01.409 Now, the dorsal cortex, as we discussed and argued, is coming from a different 00:39:01.409 --> 00:39:06.229 part of the neuroepithelium, and it has an inside-first, outside-lust pattern. 00:39:07.649 --> 00:39:13.949 And we also mentioned reeling expression, which is actually setting of this major polarity. 00:39:13.969 --> 00:39:15.869 So the inside-first, outside-lust. 00:39:16.129 --> 00:39:20.709 But now you did mention that if you take a mammal of which you knock out PEC 00:39:20.709 --> 00:39:23.209 5. PEC 6, yeah. Or PEC 6, sorry. 00:39:24.689 --> 00:39:31.069 Then you get in the end a brain that seems to follow this idea of a ball of 00:39:31.069 --> 00:39:32.389 cells being a forebrain. 00:39:32.729 --> 00:39:42.469 So, would you then see that as an indication that to go from DVR to a cortex 00:39:42.469 --> 00:39:45.369 really depends on a single regulatory gene? 00:39:45.369 --> 00:39:48.929 So in the talk, I mentioned a study 00:39:48.929 --> 00:39:53.229 which was done in collaboration with Anastasia Stojkova in Göttingen. 00:39:53.369 --> 00:40:01.269 And what we noticed was that if you, in the PAK6 knockout, you will have problems 00:40:01.269 --> 00:40:05.029 with this lateral stream of migration and differentiation, 00:40:05.289 --> 00:40:10.809 especially these areas, like lateral amygdala, claustrum, endoperiform nucleus. 00:40:10.809 --> 00:40:15.709 Nucleus, and you will begin to see a bowl of cells protruding into the lateral 00:40:15.709 --> 00:40:19.009 ventricle, very similar to what you see in the turtle. 00:40:19.389 --> 00:40:24.289 Now, we also discussed during the talk that, unfortunately, Pax6 is a master 00:40:24.289 --> 00:40:27.609 gene, so it's involved in lots of different other developmental steps, 00:40:27.789 --> 00:40:29.749 including cortical development. 00:40:30.209 --> 00:40:36.909 But nevertheless, this clearly indicates that, you know, just by changing some 00:40:36.909 --> 00:40:40.469 aspects of mammalian development, you can actually regress. 00:40:40.809 --> 00:40:44.349 To producing this bowl of cells. 00:40:44.689 --> 00:40:48.809 So I also emphasized during the talk that it's probably not the... 00:40:52.051 --> 00:40:55.151 I'm not saying that, you know, PAK6 was responsible for this, 00:40:55.271 --> 00:41:00.611 it could have been, but I think there was a very interesting rearrangement of 00:41:00.611 --> 00:41:03.931 that zone during evolution, and 00:41:03.931 --> 00:41:08.171 it's still reflected now in these basic developmental patterns in mammal. 00:41:08.411 --> 00:41:13.331 But in the PAK6 knockout mice, what's the behavioral signature of that? 00:41:14.111 --> 00:41:19.371 If you knock out in all these structures, it's death. So if you have conditional 00:41:19.371 --> 00:41:25.151 knockout, you can have cortex-specific, or you can have other conditional knockout, 00:41:25.251 --> 00:41:26.971 then it's a bit more subtle. 00:41:27.131 --> 00:41:33.151 And what is also interesting is if you change peptic expression at these zones, 00:41:33.411 --> 00:41:42.491 then this intermediate island of cells is blocking off the thalamic fibers of entering to the cortex. 00:41:42.611 --> 00:41:47.071 So that's why I was so interested in this mutant, because you have a thalamic phenotype. 00:41:47.371 --> 00:41:51.691 They can't make it through them, this part. So is it the case that the dorsal 00:41:51.691 --> 00:41:57.731 cortex in turtles is a true homologue in the sense that you want to use that 00:41:57.731 --> 00:42:00.751 word of the neocortex in mammals? 00:42:01.091 --> 00:42:03.531 So the same piece of neuroepithelium, 00:42:04.463 --> 00:42:12.823 is giving rise to the dorsal cortex in turtle and to the dorsal cortex of mammal, but the mammalian. 00:42:14.223 --> 00:42:19.123 Dorsal cortex, that part of the neuroepithelium, was turbocharged and is now 00:42:19.123 --> 00:42:25.303 producing lots of progenitors and it has an enhanced neurogenic capacity. 00:42:25.783 --> 00:42:30.743 And this may be because of genes including PAK6 perhaps, which are involved 00:42:30.743 --> 00:42:35.263 in suppressing the development of the dorsal ventricular ridge, 00:42:35.463 --> 00:42:39.243 which otherwise would take on this perhaps higher cognitive function. 00:42:39.863 --> 00:42:43.143 And is it the case that in birds they have this dorsal cortex. 00:42:44.023 --> 00:42:46.263 But it hasn't really changed much from the turtle? 00:42:46.483 --> 00:42:53.303 That's a very interesting thought, that to be able to promote the dorsal cortex, 00:42:53.463 --> 00:42:57.443 you have to suppress the DVR, which is very interesting. 00:42:57.443 --> 00:43:04.223 I always thought about it that you just cranked up the neurogenesis in the dorsal 00:43:04.223 --> 00:43:11.223 cortex, and then you start of sensory modalities and other motor functions were colonizing it. 00:43:11.423 --> 00:43:15.663 But your suggestion that you also have to suppress the other, 00:43:15.843 --> 00:43:17.703 then it's an interesting one. 00:43:17.863 --> 00:43:23.283 It is a possibility. But whatever happened, 00:43:23.523 --> 00:43:30.823 what I wanted to say at the middle or last part of my talk is that that part 00:43:30.823 --> 00:43:35.003 of neuroepithelium started to generate more neurons. 00:43:35.723 --> 00:43:40.723 The developmental program introduced more intermediate progenitors, 00:43:41.023 --> 00:43:48.563 and probably some of these intermediate progenitors, they are also amenable to regulation. 00:43:49.903 --> 00:43:56.783 Increasing the potential that you produce neurons according to the need of the dorsal cortex. 00:43:56.923 --> 00:43:59.663 So it's an other aspect of self-regulation. 00:44:00.163 --> 00:44:07.163 Unfortunately, if you block thalamic input at these early stages of development, 00:44:07.303 --> 00:44:11.063 you get the same, very similar cortical numbers. 00:44:11.583 --> 00:44:15.763 So it's not the thalamic input. It's probably there's an innate, 00:44:16.663 --> 00:44:20.583 inherent program which is producing these cell types, and the thalamic input 00:44:20.583 --> 00:44:22.243 is probably just modulating it. 00:44:23.134 --> 00:44:26.414 But now, can you tell me, what do we know really about PAK6? 00:44:26.594 --> 00:44:27.694 What's it regulating exactly? 00:44:28.694 --> 00:44:35.594 I mean, it's hundreds of genes. So, when you ask a development in neuromyelitis, 00:44:35.614 --> 00:44:38.474 okay, tell me how this transcription factor is acting. 00:44:38.954 --> 00:44:46.174 So, what they will usually say is, okay, let's get some of these transcription factors, 00:44:46.334 --> 00:44:51.694 produce some antibodies against them, do some chromatin immunoprecipitation 00:44:51.694 --> 00:44:57.254 when you precipitate these DNA parts, smaller fragments, 00:44:57.494 --> 00:45:02.554 then you pull out the parts which were specifically binding by this transcription factor, 00:45:02.814 --> 00:45:05.754 you sequence them, or you identify the genes where they bind, 00:45:05.914 --> 00:45:07.854 you have hundreds of binding partners, 00:45:08.134 --> 00:45:11.974 and then you pick one of them, and if you're lucky, that's important. 00:45:11.974 --> 00:45:16.914 So, this is how the field is looking at Pax6, it has been studied by many, 00:45:16.974 --> 00:45:22.554 many outstanding groups, and they are slowly beginning to understand how it's 00:45:22.554 --> 00:45:25.814 regulating eye development, brain development. 00:45:26.654 --> 00:45:31.534 But another interpretation, which relates to some other data you showed, is that actually, 00:45:31.734 --> 00:45:36.694 you know, there's a very specific temporal patterning in the migration of these 00:45:36.694 --> 00:45:42.534 cells and also the construction of their interconnectivity and that in itself 00:45:42.534 --> 00:45:46.094 can form barriers for the migration of other cells, right? 00:45:46.154 --> 00:45:51.554 So to actually go from a ball of cells more to in the center to a sheet at the 00:45:51.554 --> 00:45:54.454 outside, you must cross these kinds of barriers. 00:45:54.674 --> 00:46:01.214 So could the key regulatory switch be the one that's like a traffic cop helping you to regulate. 00:46:02.402 --> 00:46:09.022 The crossing of multiple migrating populations of cells through these kinds of bottlenecks? 00:46:09.082 --> 00:46:12.142 Do you see that as a possible mechanism or is that irrelevant? 00:46:12.682 --> 00:46:16.282 It's a very interesting question you just wrote. 00:46:16.502 --> 00:46:22.122 So this corner of the brain where you have either the dorsal ventricular region 00:46:22.122 --> 00:46:24.742 sauropsids or you have this lateral stream in mammals, 00:46:24.982 --> 00:46:29.862 it's interesting that you have to clear that part by the time thalamic projections 00:46:29.862 --> 00:46:35.182 interactions past that region and also corticofugals, without clearing that in time, 00:46:35.482 --> 00:46:43.242 delaying the development or altering that stream will have serious implications 00:46:43.242 --> 00:46:45.702 on the thalamocortical targeting. 00:46:46.102 --> 00:46:52.442 So lots of thalamic fibers fail to pass through that region if you have any. 00:46:53.182 --> 00:46:55.642 Migration problems of these cells out of the way. 00:46:55.802 --> 00:47:00.242 So Sonia Garrel in Ecole Normale 00:47:00.242 --> 00:47:06.402 in Paris looked at some of these evolutionary steps and basically she identified 00:47:06.402 --> 00:47:12.582 some of the factors which would help the thalamic fibers to go through the region 00:47:12.582 --> 00:47:17.642 and there are some evolutionarily conserved mechanism. 00:47:17.922 --> 00:47:25.782 So probably it's not a surprise why we have all these developmental phenotypes 00:47:25.782 --> 00:47:32.122 in that particular region, because it's so vulnerable for timing and migration and cell patterning. 00:47:33.505 --> 00:47:41.885 So, you mentioned that the neocortex becomes turbocharged in mammals, 00:47:42.125 --> 00:47:46.625 but it doesn't happen just all in one shot, does it? 00:47:46.665 --> 00:47:51.065 There's stages to that. So you mentioned that if you look at an animal like 00:47:51.065 --> 00:47:52.265 Monodelphus domestica, 00:47:52.445 --> 00:47:57.065 which seems to be similar to, in some ways, an early mammal, 00:47:57.225 --> 00:48:04.925 the density of neurons in the organization of the neocortex is much less than in a mouse, you say. 00:48:07.285 --> 00:48:11.405 And it's got something like the six-layer cortex, but there are changes that 00:48:11.405 --> 00:48:16.585 happen between that and later mammals, which make it even more turbocharged. I mean, is that right? 00:48:16.665 --> 00:48:21.125 Do you see two stages here in neocortex, or more than two stages, 00:48:21.425 --> 00:48:23.565 in the way that neocortex has evolved? 00:48:25.665 --> 00:48:28.465 Tony, this is exactly how I imagined this. 00:48:29.205 --> 00:48:38.185 So I think we could argue whether these intermediate progenitors are specifically mammalian or not. 00:48:41.485 --> 00:48:45.125 If you talk to George Streeter, 00:48:45.685 --> 00:48:51.625 He would argue that these intermediate progenitors are present in larger avian 00:48:51.625 --> 00:48:58.365 brains, and they are responsible for producing more cells in the avian brain. 00:48:58.365 --> 00:49:04.825 Brain, where we slightly disagree is that whether in the hyperpolyum, 00:49:04.825 --> 00:49:10.285 so the dorsal cortex equivalent part of the neuroepithelium, 00:49:10.325 --> 00:49:17.365 whether you have these intermediate progenitors in large number in birds or not. 00:49:17.485 --> 00:49:20.365 I would like to suggest that you don't have many of these there. 00:49:21.585 --> 00:49:23.965 But you might have some in the subpolyum. 00:49:26.538 --> 00:49:32.998 So in all mammals studied so far, including monodelphys and wallaby, 00:49:33.118 --> 00:49:36.738 as I showed you, we can see these intermediate progenitor cells, 00:49:36.998 --> 00:49:38.998 but much later in development. 00:49:39.318 --> 00:49:43.198 And I also showed you the cell counts for monodelphys, and it's roughly half 00:49:43.198 --> 00:49:48.478 of the mouse in a unit column area of the brain. 00:49:48.478 --> 00:49:56.378 And I also showed you that the onset of these obventricular proliferative profiles 00:49:56.378 --> 00:50:02.438 in the subventricular zone, they are delayed considerably, but you still have them. 00:50:02.778 --> 00:50:07.718 We have a little bit of an argument whether these are neurogenic divisions or not. 00:50:08.018 --> 00:50:13.078 They are already gliogenic with Antonello Malamaggi in Trieste, 00:50:13.278 --> 00:50:17.238 but we all agree that eventually you will have some divisions there, 00:50:17.458 --> 00:50:21.818 and they begin to produce neurons or gli. 00:50:21.818 --> 00:50:26.278 In my opinion, they start producing a little bit of neurons as well. 00:50:26.378 --> 00:50:29.358 But we have to do some clonal analysis to be sure of that. Now, 00:50:29.518 --> 00:50:32.878 if you look at then the elaboration of these subventricular zones, 00:50:33.278 --> 00:50:35.098 Henry Kennedy will be speaking here soon. 00:50:35.278 --> 00:50:40.198 He will probably touch on this, that if you look at the primate ventricular, 00:50:40.378 --> 00:50:44.898 subventricular zones, you have several subzones with lots of other progenitors. 00:50:44.898 --> 00:50:53.458 And they start to produce, you know, there is an explosion in numbers. 00:50:54.638 --> 00:51:01.698 And these progenitors, they produce neurons in many, many different levels. 00:51:01.858 --> 00:51:05.478 And I've shown you these interkinetic nuclear migration profiles, 00:51:05.758 --> 00:51:09.478 so they have to descend and divide there. These intermediate progenitors, 00:51:09.638 --> 00:51:14.938 they either don't have to move, like the TBR2 positive, or they move to the opposite direction. 00:51:15.278 --> 00:51:20.298 So you have three additional floors where you can produce neurons, 00:51:20.418 --> 00:51:27.318 and suddenly you can have this explosion of neuronal production in the brain, 00:51:27.498 --> 00:51:30.258 which is necessary, I think, to produce. 00:51:30.498 --> 00:51:35.238 And coupled with this, you have a stable platform in the subplate where you 00:51:35.238 --> 00:51:37.298 can start already building the connections. 00:51:37.298 --> 00:51:44.638 I think we have a very powerful developmental program which can not only continue 00:51:44.638 --> 00:51:47.318 to produce neurons, but already start wiring them up. 00:51:48.360 --> 00:51:53.020 So just to be clear, the Monadolphus has all the same progenitor types, 00:51:53.200 --> 00:51:55.520 or is it missing some types that other mammals have? 00:51:55.740 --> 00:52:01.920 The proportions are different. Right. But in my opinion, you have the radial 00:52:01.920 --> 00:52:08.620 glia progenitor, and you also have the TBR2 positive intermediate progenitors. 00:52:08.720 --> 00:52:11.100 But these progenitors are in smaller number. 00:52:11.380 --> 00:52:17.440 Whether they have these outer radial glia progenitors or not in very low numbers, I don't know. 00:52:17.440 --> 00:52:23.440 Even in mouse, it's less than 5% of the progenitors, which is actually a very 00:52:23.440 --> 00:52:28.020 prominent feature of the primate brain. Henry will show this. 00:52:28.360 --> 00:52:33.320 But now with these intermediate progenitors, does it mean some sort of modular 00:52:33.320 --> 00:52:35.280 construction process during development? 00:52:35.500 --> 00:52:40.020 That means that you first can produce a small set of layers, 00:52:40.120 --> 00:52:44.940 if you want, of a cortical sheet. but then you have to inject an intermediate 00:52:44.940 --> 00:52:49.200 progenitor on top of those layers to build your next layer, etc. Is that how you see it? 00:52:49.660 --> 00:52:53.220 I see it more of an amplification machinery. 00:52:54.200 --> 00:52:59.220 So I believe that these intermediate progenitors, they contribute to all layers, 00:52:59.460 --> 00:53:06.200 as I showed you with this Navnit Vashistha and Fernando Garcia Moreno paper 00:53:06.200 --> 00:53:09.300 where they trace the origin of these cells. 00:53:09.300 --> 00:53:17.160 Cells, and I think we concluded that 25 to 50% of the mouse brain cells come 00:53:17.160 --> 00:53:18.120 through these progenitors. 00:53:19.940 --> 00:53:26.540 But I don't think similar lineage analysis has been done for the outer radioglia 00:53:26.540 --> 00:53:31.820 cells, and that will be very important, and also for the short radioglia progenitors. 00:53:32.120 --> 00:53:35.220 So I think this has to be done in the future. share. 00:53:35.860 --> 00:53:40.620 Okay, so this linear analysis you did by actually um you, 00:53:42.319 --> 00:53:47.819 tagging different cells with different colors in the end, right? 00:53:47.919 --> 00:53:52.359 And then you could see, okay, which cells are clones of which progenitor cells. 00:53:54.259 --> 00:53:58.619 So, but what did you, what were the key observations in that experiment, 00:53:58.979 --> 00:54:00.499 which looks pretty amazing, actually? 00:54:01.019 --> 00:54:06.379 We only have these experiments for the TBR2 positive intermediate progenitors 00:54:06.379 --> 00:54:12.079 because we had access to these TBR2 Cree progenitors progenitors, 00:54:12.239 --> 00:54:14.279 and then we use the clone method on the top of that. 00:54:14.499 --> 00:54:19.179 And what we could now tell is what is the average clone size, 00:54:19.399 --> 00:54:23.159 how many cells you have, how many times they divide. 00:54:24.159 --> 00:54:28.179 We are now sure that they divide at least two, three, maybe even four times. 00:54:29.839 --> 00:54:32.979 But we haven't got the data for all the other progenitors. 00:54:33.259 --> 00:54:39.639 And you will hear from Henry in about two days about their observations, 00:54:39.639 --> 00:54:48.439 simulations where they imaged cortical progenitors in the primate brain in vitro, 00:54:48.579 --> 00:54:50.879 and then they followed these cells to the cortex, 00:54:51.099 --> 00:54:56.619 and then they identified the phenotype and laminar position of these cells. 00:54:56.739 --> 00:55:01.459 So they could have some snapshots, they could put these primate brain slices 00:55:01.459 --> 00:55:06.299 in the dish for, in fact, for weeks, and then they followed them, 00:55:06.339 --> 00:55:07.739 how they divide, where they migrate. 00:55:07.739 --> 00:55:13.139 And what is interesting, and what Henry and Colette are proposing, 00:55:13.299 --> 00:55:18.599 that these progenitors, they transform into one another in different directions. 00:55:18.919 --> 00:55:23.459 If this is the case, then things will become very, very complicated. 00:55:23.539 --> 00:55:28.319 Because if you can transfer a TBR2 positive intermediate progenitor back into 00:55:28.319 --> 00:55:33.699 a radial glia, or you can transfer them to an outer radial glia and vice versa, 00:55:34.019 --> 00:55:37.279 then things will become very, very complicated indeed. 00:55:37.739 --> 00:55:41.659 But do you find that a reasonable interpretation? It is, yeah. 00:55:41.799 --> 00:55:45.939 It is. But I haven't done any time-lapse studies as such. 00:55:46.279 --> 00:55:48.199 Right. Tony, your phone. 00:55:49.999 --> 00:55:54.359 Okay, so… I mean, somebody has to listen to this podcast, and probably they 00:55:54.359 --> 00:55:57.259 are all asleep by now. No, no, no. No more. 00:55:58.619 --> 00:56:02.259 We're just going to do the best bit. No, we're recording. 00:56:03.059 --> 00:56:10.859 So you in the end summarized the main results you're presenting in terms of 00:56:10.859 --> 00:56:12.779 innovation, conservation and convergence. 00:56:13.519 --> 00:56:15.499 So what did you mean with that? 00:56:16.519 --> 00:56:20.599 So I have to come back to this topic where we spent quite a bit of time in discussing that. 00:56:23.079 --> 00:56:28.959 So after all this transcriptomic analysis with Grant Bellegarde and Chris Ponting, 00:56:29.099 --> 00:56:36.119 combining it with the cell lineage and clonal analysis and gene expression, 00:56:36.479 --> 00:56:39.599 I mean, earlier gene expression, developmental gene expression, 00:56:39.599 --> 00:56:45.399 I had to realize that some parts of the brain where you have different developmental 00:56:45.399 --> 00:56:47.779 origin, they can adopt similar gene expression. 00:56:48.039 --> 00:56:54.379 So that's what I mean by, you know, they converged to that particular. 00:56:54.659 --> 00:56:59.999 In the striatum, hippocampus, oligodendrocytes, we emphasize the conservation 00:56:59.999 --> 00:57:01.699 of gene expression networks. 00:57:02.459 --> 00:57:12.059 And what was the third? Innovation. Innovation. So, when you compare brain regions, 00:57:12.179 --> 00:57:13.659 that's the interesting bit. 00:57:13.879 --> 00:57:18.959 What is new? What are the mechanisms which you don't see in other species? 00:57:19.199 --> 00:57:23.599 And especially during development, which we didn't do, that would be very interesting 00:57:23.599 --> 00:57:29.039 to see what is it, how did we get this turbocharge of the cortex? 00:57:29.339 --> 00:57:34.359 I think that would be a very interesting… So Zoltan, you're in this business now for a while. 00:57:35.079 --> 00:57:39.519 You have gained this incredible insight in the development of the neocortex. 00:57:41.199 --> 00:57:45.599 Even though we still have to explain to you about homology, but that's not a discussion. 00:57:46.619 --> 00:57:51.159 But now, given your experience in neuroscience and the study of the brain, 00:57:51.699 --> 00:57:55.939 if we would like to follow in your footsteps, what's Zoltan's law that we should adhere to? 00:57:57.159 --> 00:57:59.499 To Zoltan's law of the study of the brain. 00:58:01.299 --> 00:58:11.259 So here on this course, I emphasized the comparative evolutionary developmental aspect. 00:58:11.499 --> 00:58:20.179 But most of the work I'm doing is understanding how mammalian cortical circuits are put together. 00:58:21.179 --> 00:58:27.199 And I did not really talk about this part, but I'm very interested in this early 00:58:27.199 --> 00:58:32.919 transient platform, which is below the developing cortex, the subplate neurons. 00:58:33.219 --> 00:58:36.299 So these are largely transient cells. 00:58:36.599 --> 00:58:39.019 They are the earliest generated cells in our brain. 00:58:39.359 --> 00:58:42.859 If you look at subplate in the primate brain, Andrew will probably show you 00:58:42.859 --> 00:58:46.479 some, it's bigger than the cortical plate early on during development. 00:58:46.859 --> 00:58:50.699 And then these cells, after they set up all the connectivity, 00:58:50.879 --> 00:58:55.499 integrate into the intra and extracortical circuits, then they will disappear. appear. 00:58:55.979 --> 00:59:02.659 And you only have a couple of scattered interstitial white metal cells left 00:59:02.659 --> 00:59:08.579 over, they are cleared away, and then the final product, the cortex, will remain there. 00:59:08.799 --> 00:59:13.639 So I'm fascinated with the association of these scattered interstitial white metal cells. 00:59:14.098 --> 00:59:20.598 Residual cells to cognitive abnormalities. So it's like when you have sloppy 00:59:20.598 --> 00:59:26.858 builders, they leave some of the building, the scaffold behind when the adult structure is finished. 00:59:26.978 --> 00:59:31.478 And probably they haven't done a good job anyway in the adult structure because 00:59:31.478 --> 00:59:34.638 they left the scaffold inside and then they can't remove it. 00:59:34.718 --> 00:59:36.338 So I'm fascinated with these. 00:59:36.458 --> 00:59:40.698 And that's why I'm so interested in development, because I agree with Willis 00:59:40.698 --> 00:59:47.598 from 1664 that many of the cognitive disorders are brain developmental abnormalities 00:59:47.598 --> 00:59:49.058 and they are related to cortex, 00:59:49.298 --> 00:59:56.378 and these transient scaffold cells or the remnants of them, they tell us quite 00:59:56.378 --> 01:00:00.838 a bit about how the brain is constructed. 01:00:01.258 --> 01:00:05.598 And simply because these are the earliest generated cells, to understand them 01:00:05.598 --> 01:00:10.718 and understand their evolutionary origin, that's why I do quite a bit of comparative work. 01:00:12.078 --> 01:00:16.518 Although it's a bit more clinically driven, what I want to emphasize is that 01:00:16.518 --> 01:00:22.258 sometimes it's good to just step back and look at the bigger overall picture. 01:00:23.758 --> 01:00:30.078 Although it's a simplification, but the subplate is maybe the reptilian framework of our mammalian brain. 01:00:31.098 --> 01:00:35.458 And then if we're going to meet up with you in Oxford four years from now, 01:00:35.558 --> 01:00:39.638 and we're going to remind you of this discussion of today, 01:00:39.898 --> 01:00:43.278 and we're going to tell you, look, four years ago you made this prediction, 01:00:43.398 --> 01:00:46.358 and now we're going to go check and see if it came out. 01:00:46.418 --> 01:00:48.498 What's this one prediction you 01:00:48.498 --> 01:00:52.258 would like to make you feel most passionate about today in your own work? 01:00:52.258 --> 01:01:01.278 So if in the next four years I could monitor and modulate these transient scaffold 01:01:01.278 --> 01:01:08.038 cells in the mammalian brain and show that by modulating their function could 01:01:08.038 --> 01:01:09.918 alter cognitive development, 01:01:10.418 --> 01:01:13.578 then I would be very, very happy. 01:01:13.578 --> 01:01:19.018 And if I could extend it some neuropathology, so histology or even imaging, 01:01:19.338 --> 01:01:25.178 now we have seven Tesla machines, and if I could show that abnormal cortical 01:01:25.178 --> 01:01:29.258 development will have an impact on these subplate cells or vice versa, 01:01:29.498 --> 01:01:30.698 and then I would be very happy. 01:01:31.538 --> 01:01:34.378 Okay, great. Sultan Molnar, thank you very much for this conversation. 01:01:34.778 --> 01:01:37.378 Sultan Molnar Thank you, Paul, and thank you, Tony, for having me. 01:01:40.818 --> 01:01:46.698 The CSN podcast was produced by the Convergent Science Network of Biometrics 01:01:46.698 --> 01:01:53.498 and Biohybrid Systems, a project funded by the European 7th Research Framework Programme. 01:01:54.618 --> 01:01:59.978 For more interviews, recorded lectures or upcoming conferences in the field 01:01:59.978 --> 01:02:06.078 of biometrics and biohybrid systems, go to csnnetwork.org. 01:02:06.160 --> 01:02:14.640 Music.