WEBVTT 00:00:00.017 --> 00:00:06.797 This is Paul Vachor with Terence Deacon, who in the school was presenting this 00:00:06.797 --> 00:00:14.577 view, which is sometimes called Evo Devo, with respect to understanding of the brain and behavior. 00:00:14.777 --> 00:00:18.637 So what does Evo Devo actually really entail? What does this mean? 00:00:19.097 --> 00:00:22.297 It has different meanings, of course, for different people. 00:00:22.397 --> 00:00:29.577 But the main idea is that development constrains the pathways that evolution can take. 00:00:30.017 --> 00:00:35.317 For the simple reason that the evolution of species is an evolution of changes 00:00:35.317 --> 00:00:37.257 in developmental programs. 00:00:37.797 --> 00:00:43.237 It's not a change of adults. It's a change in the way you build bodies and the 00:00:43.237 --> 00:00:44.037 way that bodies interact. 00:00:44.477 --> 00:00:51.417 And so if development limits the way things can be built or biases the way things 00:00:51.417 --> 00:00:54.977 can be built, or if there's only a few ways that things can be built, 00:00:55.977 --> 00:00:59.917 then that will limit and influence the course of evolution. 00:01:00.477 --> 00:01:03.877 And so it's changed recently because 00:01:03.877 --> 00:01:10.577 much more about the genetic control of development is now known and how particularly 00:01:10.577 --> 00:01:16.237 various kinds of form and repetition of form gets generated in terms of genetics 00:01:16.237 --> 00:01:21.797 has changed the field a great deal so that we can now not just talk about embryogenesis, 00:01:21.917 --> 00:01:23.997 but molecular embryogenesis. 00:01:26.777 --> 00:01:31.977 But couldn't you then argue that this gives us, let's say, more insight in the 00:01:31.977 --> 00:01:38.857 specific implementation of evolutionary processes, but it's not changing the general concept? 00:01:39.057 --> 00:01:41.717 Or does it really change the general concept? No, I think you're right. 00:01:41.977 --> 00:01:46.877 In one sense, Darwin's argument, as I would like to describe it, 00:01:46.877 --> 00:01:53.137 is almost completely agnostic about how things are produced, 00:01:53.437 --> 00:01:58.717 how variations are produced, how forms are produced, or even how organisms are reproduced. 00:01:58.997 --> 00:02:04.517 And because it doesn't depend upon any particular way that you build an organism, 00:02:05.397 --> 00:02:11.317 it is acceptable when we find new ways that organisms are produced, 00:02:11.537 --> 00:02:16.057 when it was discovered that genes, for example, played a role in transmission 00:02:16.057 --> 00:02:18.737 of traits, in the beginning of the 20th century. 00:02:18.937 --> 00:02:23.877 It became a new field, but it didn't change Darwin's insight when we discovered 00:02:23.877 --> 00:02:27.857 that the nature of the gene was a molecule in the early 1950s. 00:02:27.877 --> 00:02:32.357 It changed the way we understand how the process worked, but it didn't change 00:02:32.357 --> 00:02:34.277 the concept of natural selection. 00:02:34.497 --> 00:02:40.037 And similarly, now that we understand much more about how genes produce body parts and functions, 00:02:40.257 --> 00:02:45.137 it still doesn't change natural selection, but it provides us lots of additional 00:02:45.137 --> 00:02:50.897 information about the mechanisms that are, in a sense, exposed to the pressures 00:02:50.897 --> 00:02:52.457 of natural selection. Okay. 00:02:53.647 --> 00:02:58.407 But to introduce that aspect, those of development, you emphasize quite a bit 00:02:58.407 --> 00:03:02.047 this notion of self-organization and self-organizing systems. 00:03:02.207 --> 00:03:04.987 So why is that relevant for this discussion? 00:03:05.627 --> 00:03:12.287 So it's relevant because the crucial problem in life is the second law of thermodynamics. 00:03:12.807 --> 00:03:19.487 That is, organized systems, unless they're frozen in time, will tend to degrade spontaneously. 00:03:19.487 --> 00:03:22.467 Simultaneously if what needs to happen 00:03:22.467 --> 00:03:25.687 is you need to maintain structure and maintain particularly 00:03:25.687 --> 00:03:29.647 dynamic processes within constrained 00:03:29.647 --> 00:03:34.347 limits there needs to be some way of generating those constraints generating 00:03:34.347 --> 00:03:40.147 those forms and what we do know is that self-organizing processes are the only 00:03:40.147 --> 00:03:47.247 ways that forms are generated with thermodynamic input that means that you need to 00:03:47.267 --> 00:03:55.547 continually put energy and materials into a process, to continually perturb it. 00:03:55.587 --> 00:04:00.567 But if you do so in a regular way, it will produce regularities and maintain them. 00:04:01.027 --> 00:04:06.587 And so a number of people, really since the 1950s, have focused more and more 00:04:06.587 --> 00:04:10.307 on the role of various kinds of self-organizing systems. 00:04:10.627 --> 00:04:16.127 Now, the problem is when you get a very complex organism, you get both sides of that. 00:04:16.127 --> 00:04:19.727 You get both things that become structure, a little bit like they're frozen, 00:04:19.947 --> 00:04:24.967 that they become in a sense the foundation upon which you can build further 00:04:24.967 --> 00:04:30.307 processes to produce further forms, which then can be fixed in place and where 00:04:30.307 --> 00:04:31.627 you can build further forms. 00:04:31.947 --> 00:04:39.287 And in effect, a lot of complex organisms depend upon these many stages of differentiating a platform. 00:04:40.261 --> 00:04:45.981 And then building upon it so that when genes of animal bodies divide them into 00:04:45.981 --> 00:04:53.181 segments, those segments are, in a sense, small modules that can work within themselves. 00:04:53.481 --> 00:04:57.701 Within those segments, there are subsequent processes that divide those segments 00:04:57.701 --> 00:05:00.381 into compartments, and now they can work within themselves. 00:05:00.521 --> 00:05:05.161 But in each case, the original process has a self-organizing feature. 00:05:05.921 --> 00:05:09.421 And then once it's stable, it can now become the basis for other kinds of interactions. 00:05:09.421 --> 00:05:12.281 Actions so so clearly clearly you have really thought 00:05:12.281 --> 00:05:15.261 about this a lot but in some sense if we would sort of backtrack 00:05:15.261 --> 00:05:18.321 if if we talk about the construction of form 00:05:18.321 --> 00:05:22.561 in the face of the second law thermodynamics in 00:05:22.561 --> 00:05:25.321 some sense i i think you're talking about a very 00:05:25.321 --> 00:05:30.441 specific way of creating form because surely i can i can take some lego bricks 00:05:30.441 --> 00:05:33.501 and construct a form as i hear you see i've constructed the format having to 00:05:33.501 --> 00:05:38.501 worry about self-organization but but obviously you you're You're limiting this 00:05:38.501 --> 00:05:44.981 to a process of the structuring of form using microscopic elements that themselves are, if you want, 00:05:45.041 --> 00:05:47.741 also the substrate exposed to the second law of thermodynamics. 00:05:47.921 --> 00:05:54.261 That's right. So what's so special about that process of construction that it 00:05:54.261 --> 00:05:58.281 should worry about this generation of disorder? 00:05:58.621 --> 00:06:01.241 And why is then self-organization the solution? 00:06:01.701 --> 00:06:06.541 So in one sense, it follows from Darwin's original insights. 00:06:06.541 --> 00:06:12.121 And what Darwin was recognizing was that you need form, 00:06:12.301 --> 00:06:17.461 you need differences and variations of form in order to have natural selection 00:06:17.461 --> 00:06:19.441 work, in order to choose some of 00:06:19.441 --> 00:06:22.741 them that are consistent with their environments and others that are not. 00:06:24.561 --> 00:06:29.301 Darwin's whole purpose in developing this theory was to some extent to escape 00:06:29.301 --> 00:06:35.181 teleological arguments, purposeful arguments in which the end was prefigured somehow. 00:06:35.581 --> 00:06:41.721 Now, when I as a person modify my world with respect to some desired end, 00:06:42.041 --> 00:06:44.581 I can produce form because of that. 00:06:44.701 --> 00:06:49.801 Because of this, in a sense, I have a representation of a future form and I 00:06:49.801 --> 00:06:53.901 need to produce it. The problem with life, of course, is that it doesn't have 00:06:53.901 --> 00:06:57.581 this kind of forethought until we have animals with brains. 00:06:57.841 --> 00:07:02.001 As soon as we have animals with brains, everything changes. But before that, 00:07:02.101 --> 00:07:06.161 and in building animals with brains, you need to do it without forethought. 00:07:06.561 --> 00:07:10.061 You need to do it, in a sense, post hoc, after the fact. 00:07:10.961 --> 00:07:15.561 The fittedness of a form has to be produced after the fact, by selection. 00:07:15.761 --> 00:07:20.241 That means the form is produced irrespective of its consequence initially. 00:07:20.661 --> 00:07:23.081 And it's only preserved because it produces that cut to it. 00:07:23.601 --> 00:07:27.041 So how should I think then about the development of, for instance, 00:07:27.121 --> 00:07:31.781 the brain from that perspective of self-organization? What does this mean concretely? 00:07:32.081 --> 00:07:37.541 So interestingly enough, if you think about the logic in which you need to build 00:07:37.541 --> 00:07:43.361 form and then you need to select that forms, some forms, with respect to how 00:07:43.361 --> 00:07:45.001 they fit in their environment. 00:07:45.581 --> 00:07:48.441 One of the things that happens with brains is that the early stages of brain 00:07:48.441 --> 00:07:51.741 development in all vertebrates is very, very similar. 00:07:52.201 --> 00:07:55.561 The way that cell groups are produced, the way that the different, 00:07:55.681 --> 00:07:58.081 you might call them even segments of the brain are produced. 00:07:58.381 --> 00:08:02.761 It's a tube that simply gets divided into segments, and then those segments differentiate. 00:08:03.281 --> 00:08:05.961 That's very consistent across most vertebrates. 00:08:06.841 --> 00:08:14.241 In mammals and birds particularly, we have this phase later on in which within these segments, 00:08:14.521 --> 00:08:19.241 there begins to be, once you've organized it into parts that have particular forms, 00:08:19.441 --> 00:08:24.521 and that's a very generic feature so that you can actually look at the brains 00:08:24.521 --> 00:08:29.901 of most vertebrates at an early stage in their development and cannot tell them apart easily. 00:08:30.481 --> 00:08:33.561 It's only later that they become partitioned into components. 00:08:34.101 --> 00:08:39.101 But now you have exactly the same problem. You can produce lots of forms with 00:08:39.101 --> 00:08:43.301 a sort of generic form production system, which is very, very primitive. 00:08:43.881 --> 00:08:49.701 And then later on, as signals come in from the periphery, from your sense organs, 00:08:49.921 --> 00:08:56.121 or when there's interaction between muscular systems and sensory systems in the nervous system, 00:08:56.201 --> 00:09:00.561 signals now passing through the nervous system actually play a secondary role 00:09:00.561 --> 00:09:05.081 in selecting which of those circuits, which of those forms persist. 00:09:05.081 --> 00:09:10.661 But again, like the theory of evolution itself, you need to produce a variety 00:09:10.661 --> 00:09:13.381 of forms and regularities, and then... 00:09:14.614 --> 00:09:18.534 Put them in competition with each other in the context of an environment. 00:09:18.874 --> 00:09:22.774 And that seems to be crucial for the brain. One of the ways that we were able 00:09:22.774 --> 00:09:26.354 to identify that was using transplants across species. 00:09:26.554 --> 00:09:32.074 One of the reasons I went into this work was to find out how different species 00:09:32.074 --> 00:09:35.414 were in terms of how their circuits were built. 00:09:35.654 --> 00:09:39.714 And by using tissue from one species and another species. 00:09:40.394 --> 00:09:43.854 Transplanting it into another species' brain and watching how it grows into 00:09:43.854 --> 00:09:48.654 that brain, we could begin to see if there was many differences in the way the 00:09:48.654 --> 00:09:53.594 axons, the output branches, found their targets, or whether there were similarities. 00:09:53.914 --> 00:09:57.014 And what we found was there were remarkable similarities, similarities that 00:09:57.014 --> 00:10:00.734 were much stronger than we ever expected between species that were quite different. 00:10:00.934 --> 00:10:02.294 They're all species of mammals. 00:10:02.514 --> 00:10:06.974 I suspect that if we could have done this across birds and mammals or even into 00:10:06.974 --> 00:10:09.954 reptiles, that we would have found very conserved features. 00:10:10.334 --> 00:10:13.914 Can you give an example of these conserved features? Well, so, for example, 00:10:14.034 --> 00:10:18.214 we were able to show that you can take cortical cells and put them just about 00:10:18.214 --> 00:10:24.714 anywhere in the brain, and their axons will grow out to only the appropriate 00:10:24.714 --> 00:10:26.274 targets for cortical cells, 00:10:26.454 --> 00:10:29.414 but not targets in the brain that are not cortical. 00:10:29.854 --> 00:10:35.694 But later on, we wouldn't find function until the competition had eliminated 00:10:35.694 --> 00:10:36.794 some of these connections. 00:10:36.794 --> 00:10:41.974 So there was an initial sort of what I would call generic phase in which the 00:10:41.974 --> 00:10:45.454 cells found any and all targets that were appropriate for cortex. 00:10:46.779 --> 00:10:51.939 But then some of those got eliminated and got cut back, in effect, by function. 00:10:52.119 --> 00:10:56.619 We found this particularly in the, we used cells from the midbrain, 00:10:56.619 --> 00:10:57.939 which are dopaminergic cells. 00:10:58.199 --> 00:11:03.399 But in taking those cells from our donors, we couldn't separate them from other 00:11:03.399 --> 00:11:06.339 cells of the midbrain, which are not carrying dopamine. 00:11:06.499 --> 00:11:11.319 Now, dopamine is crucial for the function of most motor behaviors, 00:11:11.539 --> 00:11:15.259 particularly associated with a disease called Parkinsonism. 00:11:15.259 --> 00:11:18.279 And we produced animals that had 00:11:18.279 --> 00:11:20.999 parkinsonism and our attempt was to 00:11:20.999 --> 00:11:24.979 repair the damage by putting cells 00:11:24.979 --> 00:11:28.459 from another animal back in that produced 00:11:28.459 --> 00:11:31.619 this dopamine this neurotransmitter that controlled that 00:11:31.619 --> 00:11:35.479 interestingly enough we were putting two kinds of cells in dopamine cells and 00:11:35.479 --> 00:11:39.459 other cells as well because we couldn't separate them out and what we found 00:11:39.459 --> 00:11:45.399 is that the dopamine cells found their appropriate targets and over time began 00:11:45.399 --> 00:11:50.179 to alleviate the Parkinson effects in our donor animals, our host animals. 00:11:50.419 --> 00:11:54.599 The other cells sent their axons, even though they're placed in the same place, 00:11:54.739 --> 00:11:58.559 they sent their outputs to entirely different places, a place called the thalamus, 00:11:58.599 --> 00:12:03.399 and made connections there and also influenced the thalamic functions. 00:12:03.759 --> 00:12:08.239 So that even though the cells were put in the wrong place, in fact in an adult 00:12:08.239 --> 00:12:13.339 animal, not a baby animal, The output branches made their connections, 00:12:13.539 --> 00:12:16.019 but their function took time. 00:12:16.179 --> 00:12:20.379 So although they made connections and although the connections became useful, 00:12:20.599 --> 00:12:24.699 that is, they became active synapses in which neurotransmitters were crossing 00:12:24.699 --> 00:12:27.399 between the different neurons of different species. 00:12:27.599 --> 00:12:35.219 So we could, for example, see a pig presynaptic axon attached to a rat postsynaptic 00:12:35.219 --> 00:12:38.219 button and functioning appropriately. 00:12:39.499 --> 00:12:42.479 And what happened in this case is that, 00:12:43.413 --> 00:12:47.153 The synapses were formed, but we did not see function immediately. 00:12:48.193 --> 00:12:52.133 But as some synapses were eliminated and the system was sculpted, 00:12:52.133 --> 00:12:57.993 in effect, by having signals pass through it, the animals gradually improved their function. 00:12:58.233 --> 00:13:03.133 They never were perfect, but they improved their function, suggesting that over 00:13:03.133 --> 00:13:05.253 time, it took two processes. 00:13:05.353 --> 00:13:10.053 It took making connections, and then it took a process of shaping those connections, 00:13:10.213 --> 00:13:15.433 probably for functional optimization and so on. But the humans just never really worked. 00:13:15.693 --> 00:13:19.853 It worked partially. So there was a transplantation. 00:13:19.913 --> 00:13:24.133 First of all, there have been transplantations from human fetal tissue into 00:13:24.133 --> 00:13:25.453 adult Parkinson's patients. 00:13:25.673 --> 00:13:29.333 And they did alleviate, in many cases, some of the difficulties. 00:13:29.633 --> 00:13:33.033 The most severe cases were cases where it wasn't Parkinsonism, 00:13:33.053 --> 00:13:38.093 but actually a loss of function because of a drug overdose effect that actually 00:13:38.093 --> 00:13:42.413 damaged the system. We actually used it in our experiments to actually produce 00:13:42.413 --> 00:13:43.653 artificial Parkinsonism. 00:13:43.733 --> 00:13:46.153 Yeah, that's where the MPTP comes from. MPTP, exactly. 00:13:47.393 --> 00:13:53.893 So what we did in this process is that we created Parkinsonism and then improved it. 00:13:54.453 --> 00:13:59.713 In humans, we also tried to transplant not just from human fetuses, 00:13:59.953 --> 00:14:04.933 which is a difficult thing to do for a variety of both physical reasons, 00:14:05.073 --> 00:14:07.593 experimental reasons, But also for ethical reasons. 00:14:07.673 --> 00:14:10.933 There were lots of problems with it, particularly at the time we were doing this. 00:14:11.413 --> 00:14:15.233 So we began to test these cross-species transplants. 00:14:15.453 --> 00:14:19.333 The first thing we had to show was that it worked across species. 00:14:19.553 --> 00:14:23.213 So we used pigs as our donors and rats as our hosts initially. 00:14:23.213 --> 00:14:26.353 Showed that the connections were made and that there was improvement. 00:14:26.833 --> 00:14:32.813 We then used pigs and monkeys as recipients and showed that there was some effect in monkeys. 00:14:32.953 --> 00:14:37.513 We had some problem because the monkeys were rejecting the grafts in ways that 00:14:37.513 --> 00:14:43.893 the rats did not reject those grafts. It had already been shown that rat grafts 00:14:43.893 --> 00:14:46.013 in monkeys were not rejected, interestingly enough. 00:14:46.973 --> 00:14:52.673 It turns out that all pig tissue is rejected very quickly in humans. 00:14:53.213 --> 00:14:55.493 And we didn't know this when we first started the process. 00:14:57.484 --> 00:15:01.684 During the process, we found that we were able to show that there was no secondary 00:15:01.684 --> 00:15:06.804 problem, no secondary damage, even if the graft completely failed, 00:15:06.984 --> 00:15:08.504 even if it completely died. 00:15:08.704 --> 00:15:15.024 So it was deemed reasonable to use this in what they called safety trials to test clinically. 00:15:15.324 --> 00:15:19.444 And so there were 12 volunteers, all who had very advanced stage Parkinsonism, 00:15:20.044 --> 00:15:23.904 and they were transplanted with pig fetal dopamine cells. 00:15:24.144 --> 00:15:26.664 Actually, as they say, dopamine and a few other kinds of cells. 00:15:27.844 --> 00:15:34.904 The first two patients improved considerably, eventually rejected their grafts 00:15:34.904 --> 00:15:36.244 and their improvement went down. 00:15:36.364 --> 00:15:40.364 In fact, the first patient was our very most successful case. 00:15:40.704 --> 00:15:45.844 He was transplanted. He had lived much of his life in the past five years before 00:15:45.844 --> 00:15:48.384 this, almost completely immobile. 00:15:48.544 --> 00:15:51.364 He would have these periods of mobility. 00:15:52.044 --> 00:15:55.264 He was on dopamine, I mean, dopamine replacement, L-dopa. 00:15:55.484 --> 00:16:00.684 But it turns out that in advanced stages, the L-dopa produces, 00:16:00.864 --> 00:16:06.124 when you're able to move, it produces lots of uncontrollable movement and then 00:16:06.124 --> 00:16:09.064 oscillates with periods of complete inability to move. 00:16:09.264 --> 00:16:11.444 So there's what we call an on and off state. 00:16:13.144 --> 00:16:18.804 And at the stage that he was advanced, he was in much more time in the off stage than the on stage. 00:16:19.044 --> 00:16:22.384 And even when he was on, his movements were almost uncontrollable. 00:16:22.384 --> 00:16:30.864 But after the transplant, he had many more periods of time of on and was not 00:16:30.864 --> 00:16:34.524 having any of these secondary movement problems, so much so that he decreased 00:16:34.524 --> 00:16:36.724 his L-DOPA levels down considerably. 00:16:37.024 --> 00:16:40.324 And that also minimized the secondary effect. 00:16:40.724 --> 00:16:43.984 So that, in fact, he was actually doing better. This is a man, 00:16:44.044 --> 00:16:46.924 as I said, who was pretty much immobilized for about five years. 00:16:46.924 --> 00:16:54.024 He subsequently painted his fence, and he had played golf as a younger man and 00:16:54.024 --> 00:16:55.704 was practicing a golf swing. 00:16:55.784 --> 00:16:59.244 And I have a wonderful picture of him after the transplant, swinging a golf 00:16:59.244 --> 00:17:00.364 club, hitting a golf ball. 00:17:00.424 --> 00:17:05.044 Not very well, but this is something that would have been completely impossible before this. 00:17:05.224 --> 00:17:09.144 But now this works, as far as I understood it. 00:17:10.267 --> 00:17:15.847 You're saying there's a scaffold within the nervous system that coordinates 00:17:15.847 --> 00:17:19.167 the outgrowth of processes, right? 00:17:19.267 --> 00:17:25.707 And this scaffold actually maintains itself throughout the lifetime of the organism. 00:17:25.887 --> 00:17:29.827 And that's what, in these cases, is actually being exploited by the tissue. That's right. 00:17:30.147 --> 00:17:33.047 How should I think exactly about that? Well, it was actually a surprise. 00:17:34.307 --> 00:17:37.927 We didn't know. One of the reasons we did this is we didn't know if they would 00:17:37.927 --> 00:17:43.787 find their targets. So early on, all the transplants we did were in the very 00:17:43.787 --> 00:17:46.567 target, the very same tissue that was the normal target. 00:17:46.767 --> 00:17:51.727 And it was only after we found by transplanting in different places that the 00:17:51.727 --> 00:17:56.847 axons did find their target, even in adult brains, which surprised us. 00:17:56.947 --> 00:18:02.387 It said, yes, these signaling molecules, guidance molecules must be present. 00:18:02.567 --> 00:18:04.367 But here's the surprising thing. 00:18:05.487 --> 00:18:09.627 Adult forebrain, that is, this is anywhere in the central nervous system. 00:18:09.907 --> 00:18:15.267 After you mature, there are lots of molecules that appear that stop axons from growing. 00:18:15.507 --> 00:18:19.547 They literally block the growth of axons. Now, we don't know. 00:18:19.627 --> 00:18:21.427 There's different ways that this can happen. 00:18:21.527 --> 00:18:25.127 It can happen because the axons, they actually have to pull themselves forward. 00:18:25.327 --> 00:18:28.027 And if they don't have anything they can hold on to, so to speak, 00:18:28.127 --> 00:18:31.847 they can't pull themselves forward. There are also inhibitory molecules that 00:18:31.847 --> 00:18:34.027 send the axons in the other direction or just keep. 00:18:35.185 --> 00:18:40.325 Cause these sort of finger-like growth features called growth cones to collapse. 00:18:41.245 --> 00:18:46.165 This is true in adult brains. And so when we transplanted fetal tissue into 00:18:46.165 --> 00:18:52.045 adult brains, the axons grew very, very much slower than they would have in 00:18:52.045 --> 00:18:53.725 a fetus or in a young organism. 00:18:54.405 --> 00:19:00.245 So it took a long time to grow there. But by transplanting, and we only did 00:19:00.245 --> 00:19:03.205 this, we're only able to find this by transplanting from pig brains, 00:19:03.465 --> 00:19:07.185 which are very large and take a long time to mature, into rat brains, 00:19:07.385 --> 00:19:09.585 which are small and take a short time to mature. 00:19:09.805 --> 00:19:16.845 The rat axons could not grow long enough to find their targets if they were placed far away. 00:19:17.085 --> 00:19:22.985 But pigs, because they took a long time to mature, those cells actually did find their target. 00:19:23.245 --> 00:19:29.685 And that proved to us that even though adult brains, for some reason, 00:19:29.725 --> 00:19:31.165 don't want axons to grow, 00:19:31.445 --> 00:19:35.705 nevertheless, keep the information about how to find targets, 00:19:35.865 --> 00:19:39.325 even though, as far as we could tell, it's not useful for that purpose. 00:19:39.945 --> 00:19:43.225 However, so this is great. So here we have this nervous system, 00:19:43.285 --> 00:19:47.125 and it's as if throughout this nervous system, all sorts of paths are being 00:19:47.125 --> 00:19:51.045 labeled with different kinds of markers to still guide growth, 00:19:51.365 --> 00:19:55.565 even though, as you're saying now, the processes that should make use of this 00:19:55.565 --> 00:20:00.405 information stop to appear at some point in life. 00:20:00.745 --> 00:20:05.665 But now, if you look at lesions to the brain, as in the case of stroke, 00:20:05.905 --> 00:20:09.485 then again, you might see this kind of sprouting of processes. 00:20:10.785 --> 00:20:15.345 So, do you see that as a reason why these markers are still there? 00:20:15.445 --> 00:20:19.005 Or do you really see this as just a leftover of some developmental program and 00:20:19.005 --> 00:20:21.765 not to be used again in the future? That's a great question. 00:20:21.845 --> 00:20:23.025 First of all, we don't know for sure. 00:20:24.305 --> 00:20:26.965 But I think there's also a third option. than the one that i 00:20:26.965 --> 00:20:30.365 like though i think it could well be a partial 00:20:30.365 --> 00:20:36.125 repair the problem is after damage and even though there are all kinds of uh 00:20:36.125 --> 00:20:41.445 irritation and damage related effects it doesn't seem to eliminate these inhibitory 00:20:41.445 --> 00:20:45.285 molecules they are there and this is one of the reasons that stroke there's 00:20:45.285 --> 00:20:49.505 minimal recovery after stroke or minimal recovery after spinal cord injury. 00:20:50.517 --> 00:20:54.317 Growing across the injured site is very difficult. Now, in the peripheral nervous 00:20:54.317 --> 00:20:55.257 system, that's not true. 00:20:55.597 --> 00:20:58.297 But for stroke, it's also not necessarily true, right? After stroke, 00:20:58.517 --> 00:21:01.617 you have a rather dramatic, also spontaneous recovery. 00:21:02.417 --> 00:21:05.017 But then, of course, there's some ceiling effect there. Yes, 00:21:05.017 --> 00:21:09.177 right, right. And so what we find is that there is local growth. 00:21:09.797 --> 00:21:14.497 And there's even some neurogenesis locally, as far as we can tell. 00:21:14.617 --> 00:21:17.257 But it's relatively short distance. 00:21:17.957 --> 00:21:21.597 The really important connections, of course, in the brain are crossing long distances. 00:21:21.777 --> 00:21:25.437 By long distances in the brain, I mean a few centimeters, but that's a very 00:21:25.437 --> 00:21:27.477 long distance for a tiny little cell. 00:21:27.737 --> 00:21:31.297 And that kind of growth just simply doesn't happen. 00:21:31.397 --> 00:21:36.817 And so in the attempt to find ways to help people recover from brain damage, 00:21:36.977 --> 00:21:39.937 one of the things that we've been trying to do and many people have tried to 00:21:39.937 --> 00:21:45.177 do is to come up with ways to either aid the growth of axons by giving them 00:21:45.177 --> 00:21:46.977 a new pathway to grow through. 00:21:47.197 --> 00:21:51.877 And one of the ways to do that is, for example, to implant a peripheral nerve 00:21:51.877 --> 00:21:56.657 into the brain, because the peripheral nerves have a kind of a tube made up 00:21:56.657 --> 00:21:58.897 of myelin that the nerves can grow through. 00:21:59.017 --> 00:22:02.757 And so when you get damage to your periphery, you often can regrow your nerves 00:22:02.757 --> 00:22:04.257 as they find this pathway. 00:22:04.537 --> 00:22:09.417 And some people have shown that if you put in these kinds of extra cables, 00:22:09.557 --> 00:22:13.557 you can get some growth across a damage But generally, it doesn't occur. 00:22:14.117 --> 00:22:18.717 Now, what I think is going on is that, first of all, I don't think it's just leftover. 00:22:19.137 --> 00:22:22.837 I don't think it aids damage in the way we're thinking about it. 00:22:23.397 --> 00:22:28.597 But one of the things that happens is that locally, there is lots of movement 00:22:28.597 --> 00:22:31.677 that is within very short distances, within neuronal distances, 00:22:31.817 --> 00:22:32.977 neuronal cell body distances. 00:22:33.217 --> 00:22:37.977 There's lots of changes of axons and dendrites. Dendrites are reabsorbed. 00:22:39.357 --> 00:22:45.177 Synapses are pulled off. off or dissociate in various ways and move around. 00:22:45.417 --> 00:22:49.937 We now know that in learning, there can be lots of movement and change in synaptic 00:22:49.937 --> 00:22:55.437 density and dendrites can increase the number of synapses on them and so on and so forth. 00:22:55.557 --> 00:22:58.717 So we know that there is plasticity at this very local level. 00:23:00.176 --> 00:23:03.496 To make those connections, there had to be target information. 00:23:04.076 --> 00:23:09.516 Now, the target information may be just local, but everywhere in the brain, it's going to be local. 00:23:09.856 --> 00:23:14.676 And so if that target information is everywhere, they are still now used just 00:23:14.676 --> 00:23:18.616 simply to keep synapses in place to say, look, if you break up, 00:23:18.656 --> 00:23:19.736 don't go wandering elsewhere. 00:23:19.936 --> 00:23:23.416 Stick around because there's likely to be another option here. 00:23:23.576 --> 00:23:26.416 Right. So my sense is that it's actually playing a 00:23:26.416 --> 00:23:29.116 role in the plasticity and the 00:23:29.116 --> 00:23:32.216 learning capacity of brains but now is there is there 00:23:32.216 --> 00:23:34.956 also a possibility that um this has 00:23:34.956 --> 00:23:38.256 to do with let's say neurogenesis uh during in 00:23:38.256 --> 00:23:41.076 the adult brain yes you know as you know right the textbook 00:23:41.076 --> 00:23:45.936 knowledge was like well it all stops after a few months yes in humans but there's 00:23:45.936 --> 00:23:49.536 more and more evidence that we still see neurogenesis also in adult brains and 00:23:49.536 --> 00:23:54.336 is that maybe the reason why you still want to have these different guidance 00:23:54.336 --> 00:23:59.256 cues around to make sure that these things embed themselves in surrounding tissues correctly. 00:23:59.516 --> 00:24:02.356 Is that a possible interpretation of this as well? It's possible, 00:24:02.596 --> 00:24:05.416 but I'm less convinced of it for the following reason. 00:24:05.896 --> 00:24:12.496 First of all, the mouse neurogenesis occurs in just a few places where we really see it extensively. 00:24:12.816 --> 00:24:17.076 So in a part called the dentate gyrus of the hippocampus, we can actually see 00:24:17.076 --> 00:24:20.296 cells decline and then be replaced quite regularly. 00:24:20.696 --> 00:24:27.536 Also in an area that lines the ventricles, these open spaces within the brain, there is an area, 00:24:27.736 --> 00:24:34.256 the subventricular zone, in which you actually do find new cells being produced 00:24:34.256 --> 00:24:36.256 and extends out into the olfactory bulb, in fact. 00:24:37.756 --> 00:24:43.316 As far as I know, there's very little migration out of those zones into long distances. 00:24:43.676 --> 00:24:48.396 And for short distance axon connections, and the dentate gyrus, 00:24:48.436 --> 00:24:50.276 for example, are all short distance connections. 00:24:50.676 --> 00:24:53.896 They don't go very long distance. These are granule cells, basically. 00:24:54.116 --> 00:24:56.816 And they only connect to the nearest pyramidal cells. 00:24:58.061 --> 00:25:02.601 Those short distances probably are not going to be inhibited by these growth-inhibiting molecules. 00:25:03.041 --> 00:25:06.001 It's only the long-distance connections that are. Right, okay. 00:25:06.261 --> 00:25:12.181 So I think the issue is that mature brains don't want large-scale chain, 00:25:12.381 --> 00:25:16.541 but they want to allow and even make possible short-distance chain. 00:25:16.781 --> 00:25:21.801 So I kind of think about it as an asymptotic process, that early on long-distance 00:25:21.801 --> 00:25:23.301 reorganization is possible. 00:25:23.301 --> 00:25:28.761 And as you mature, the length of distance in which plasticity can occur gets 00:25:28.761 --> 00:25:32.361 shorter and shorter until it becomes within a few millimeters. 00:25:32.761 --> 00:25:37.721 Very good. But then how about to link that again to this earlier discussion 00:25:37.721 --> 00:25:41.001 about, let's say, the self-organization of development? 00:25:41.281 --> 00:25:45.701 Because in some sense, you could argue that this more global organization of 00:25:45.701 --> 00:25:50.041 the body and the nervous system and its parts is, in that sense, 00:25:50.061 --> 00:25:51.881 not fully self-organizing. 00:25:51.881 --> 00:25:56.021 I mean, a lot of regulatory genes that really try to control and lay down very 00:25:56.021 --> 00:26:00.961 specific aspects of body and brain and organs and so on. 00:26:01.121 --> 00:26:05.441 And so in the description you just gave, that would mean that these more global 00:26:05.441 --> 00:26:08.901 structures are also under more genetic control. 00:26:09.201 --> 00:26:13.521 Right. They create specific forms, if you want, while at the local level, 00:26:13.561 --> 00:26:17.661 you might then allow more of these self-organizing processes to fill in the details. 00:26:17.881 --> 00:26:21.061 Would you accept that cartoon? I would actually reverse it. 00:26:21.201 --> 00:26:24.441 Okay. And here's how I would say it. First of all, a lot of the organization 00:26:24.441 --> 00:26:30.501 that homeotic genes produce is the result of self-organizing processes within genes. 00:26:30.781 --> 00:26:33.921 That is, these genes are in networks with respect to each other, 00:26:33.981 --> 00:26:36.981 turning each other on and off in interesting patterns. 00:26:37.181 --> 00:26:42.281 And it's the pattern of genes turning each other on and off that produces regularity, 00:26:42.281 --> 00:26:47.461 basically a recursion kind of relationship that you could model even in a neural net model. 00:26:48.807 --> 00:26:53.887 The subsequent level is in which cells in different regions now are producing 00:26:53.887 --> 00:26:57.167 diffusible processes produced by their genes, 00:26:57.387 --> 00:27:02.247 which are binding to the genome in neighboring genes, and in neighboring cells, 00:27:02.407 --> 00:27:04.427 excuse me, that have diffused away. 00:27:04.627 --> 00:27:06.807 But the diffusion has pattern to it. 00:27:07.047 --> 00:27:11.927 And they then turn each other on and off, turn other cells on and off by turning 00:27:11.927 --> 00:27:14.167 their genes on and modifying their genes. 00:27:14.327 --> 00:27:17.567 But now here's the interesting thing about diffusionary processes. 00:27:17.567 --> 00:27:19.947 They overlap in complicated ways. 00:27:20.187 --> 00:27:24.607 There's not just one molecule diffusing from one site. There are dozens of molecules 00:27:24.607 --> 00:27:26.187 diffusing from different sites. 00:27:26.407 --> 00:27:31.647 And so a particular cell, in a sense, knows where it is by virtue of having 00:27:31.647 --> 00:27:37.287 genes that are responsive to the relative concentration of a few of these diffusible 00:27:37.287 --> 00:27:39.067 membrane, these diffusible proteins. 00:27:39.387 --> 00:27:42.147 Or diffused, they're even smaller than proteins mostly. 00:27:42.147 --> 00:27:47.267 And that then activates and inactivates different genes in those cells. 00:27:47.507 --> 00:27:53.707 That's a process that's also self-organizing. So it's not fair to say that genes 00:27:53.707 --> 00:27:58.967 are like control and everything else is self-organizing. 00:27:59.087 --> 00:28:05.547 That the genes are self-organizing amongst themselves, produce cellular patterning 00:28:05.547 --> 00:28:11.887 that then produces diffusible relationships and cell migration relationships that self-organize. 00:28:12.147 --> 00:28:17.867 That they then set up a pattern. So in a sense, it's each time you're setting up a new platform. 00:28:19.807 --> 00:28:23.607 The selection part actually is always after that. 00:28:23.787 --> 00:28:27.327 So, for example, even as simple as building your fingers, one of the things 00:28:27.327 --> 00:28:31.967 that happens is there's a diffusion of these diffusible processes across the 00:28:31.967 --> 00:28:36.347 hand, which at early stages is just a sheet of tissue. issue. 00:28:36.587 --> 00:28:42.687 And at different points along this sheet, there are interacting molecules that 00:28:42.687 --> 00:28:44.727 turn on and off certain genes. 00:28:44.907 --> 00:28:49.207 And one of the sets of genes that they turn off are genes that are called calf 00:28:49.207 --> 00:28:50.447 space genes and other genes. 00:28:50.547 --> 00:28:53.427 But what they do is that they cause cells to kill themselves, 00:28:53.747 --> 00:28:55.647 suicide kind of messages. 00:28:55.967 --> 00:28:59.607 And they tell the cells to stop reproducing and to, in fact, 00:28:59.607 --> 00:29:01.267 engage in a kind of suicide. 00:29:01.587 --> 00:29:07.567 And so the spaces between your fingers are actually generated by signals that 00:29:07.567 --> 00:29:10.827 are, in a sense, the result of these interacting diffusions. 00:29:11.988 --> 00:29:16.228 And the cells responding to those interacting diffusions. If you change the 00:29:16.228 --> 00:29:22.048 concentration, for example, of diffusion, you can produce a hand with two thumbs, one on each end. 00:29:22.628 --> 00:29:26.628 What's happened is that it's not that the genes have said, build a thumb over 00:29:26.628 --> 00:29:28.008 here and a little finger over here. 00:29:28.228 --> 00:29:35.128 What they've said is that if you're in a particular position in this diffusion space, act this way. 00:29:35.688 --> 00:29:39.648 So it's a complicated combination. Yes, but this is interesting, 00:29:39.748 --> 00:29:45.028 right? Because this diffusion or the molecule that sets up these gradients, 00:29:45.148 --> 00:29:49.708 in some sense, it's been posing a very strong constraint on the self-organizing process. 00:29:49.988 --> 00:29:54.568 That's right. So it's not completely bottom-up, self-organizing local interaction. 00:29:54.688 --> 00:29:56.808 There's a very strong top-down constraint now. 00:29:57.568 --> 00:30:01.808 For example, the size of the tissue matters because diffusion, 00:30:02.088 --> 00:30:04.608 for example, can't go certain distances. 00:30:04.748 --> 00:30:08.208 It only works certain distances. And that's one of the interesting things about 00:30:08.208 --> 00:30:12.748 brains, in fact, because brains and mammalian bodies and mammalian brains are 00:30:12.748 --> 00:30:15.588 many orders of magnitude different in scale. 00:30:16.908 --> 00:30:22.148 It's hard to imagine that the same diffusible process would generate a brain 00:30:22.148 --> 00:30:24.328 that's very big and a brain that's very small. 00:30:24.508 --> 00:30:28.928 Absolutely. That brings me to the next issue that I really would like to inspect 00:30:28.928 --> 00:30:33.148 in a bit more detail, because you have spent quite some time on a more or less 00:30:33.148 --> 00:30:39.288 a comparative analysis of brains. What I would like to know is what's the common design, right? 00:30:39.328 --> 00:30:43.488 Because we could argue, look, of all possible brains that ever existed on this 00:30:43.488 --> 00:30:45.068 planet, there must be a common design. 00:30:45.368 --> 00:30:48.548 If there's no common design behind these, we're lost as scientists. 00:30:48.748 --> 00:30:53.068 No? No, that's right. So if there's anyone now, at least in this room, 00:30:53.148 --> 00:30:56.388 who would know what the common design is, it's you. So I would like to know what it is. 00:30:58.341 --> 00:31:01.501 There's a couple of stages of commonality, of course. 00:31:02.541 --> 00:31:08.501 When you have organisms that have to move and have to move in one direction, 00:31:08.601 --> 00:31:11.241 typically you have to have a head end and a tail end. 00:31:11.401 --> 00:31:15.161 And almost always you have bilateral symmetry. 00:31:15.381 --> 00:31:19.461 You have two sides to it. So you have left and right, front and back, top and bottom. 00:31:19.981 --> 00:31:25.501 Those dimensions get set up very early in life. And it's not just vertebrates like us. 00:31:26.061 --> 00:31:31.301 Insects and worms all have this feature. And in fact, some organisms like sponges 00:31:31.301 --> 00:31:33.821 that don't have this feature when they're mature, 00:31:34.081 --> 00:31:38.121 when their embryos actually do have this feature, and they end up effectively 00:31:38.121 --> 00:31:41.921 digesting their nervous systems after they've stopped moving and they found 00:31:41.921 --> 00:31:43.281 a place to set themselves up. 00:31:43.421 --> 00:31:48.581 But to move, you need to put sense organs towards the front where you're moving 00:31:48.581 --> 00:31:51.921 towards, because you need to get information about something that's distant. 00:31:52.521 --> 00:31:56.801 You need to predict, basically. And I think one of the things that brains are about is predicting. 00:31:57.241 --> 00:32:01.081 If you're moving, you have to predict what are the consequences of your movement. 00:32:01.301 --> 00:32:05.081 So it looks as though brains have been put up front for that reason. 00:32:05.181 --> 00:32:08.821 Now, it turns out that's a good position because up front is also where you 00:32:08.821 --> 00:32:12.461 want your mouth, where you're going to target food. 00:32:12.741 --> 00:32:17.001 And so you now have to have a bunch of organs to bring things into your digestive 00:32:17.001 --> 00:32:22.241 system. So there's lots of reasons to concentrate the nervous system up front if you're moving. 00:32:22.917 --> 00:32:25.197 In one direction, if you have a head and a tail, so to speak. 00:32:25.517 --> 00:32:29.677 Now, that's true for insects, it's true for worms, and it's true for vertebrates. 00:32:30.297 --> 00:32:35.237 But in all of this, you referred back to this fairly, let's say, 00:32:35.297 --> 00:32:38.457 classic idea of McLean about the tree on the brain. Yeah, that's right. 00:32:38.697 --> 00:32:42.517 So is that where we actually would say, look, we have these three levels of 00:32:42.517 --> 00:32:49.237 the reptilian brain moving forward towards, let's say, the simple vertebrate 00:32:49.237 --> 00:32:51.197 brain to, let's say, the human brain, right? 00:32:51.197 --> 00:32:54.337 Is there any truth to that still? 00:32:54.497 --> 00:32:56.997 Is this a useful heuristic to think about the brain? 00:32:57.177 --> 00:33:02.537 I think it's a useful heuristic only in a very vague functional sense. 00:33:02.697 --> 00:33:07.397 In terms of the anatomy and evolution of the brains, I think it's completely false. 00:33:07.917 --> 00:33:14.937 And here's why. When we look at almost any fish brain, including jawless fish, 00:33:15.197 --> 00:33:20.937 very primitive fish, all the major what we call encephalon regions are there. 00:33:21.197 --> 00:33:26.197 So the most forward part, the teal encephalon, that in mammals like you and 00:33:26.197 --> 00:33:29.937 I, is quite enlarged and quite exaggerated compared to the rest, 00:33:30.037 --> 00:33:31.637 is there in all vertebrates. 00:33:31.797 --> 00:33:35.077 And even, in fact, in insects, there's corresponding structures. 00:33:35.457 --> 00:33:38.097 They're not doing the same thing. They're a very different kind of structure. 00:33:38.237 --> 00:33:43.697 But, in fact, some of the same genes are involved in producing that furthest 00:33:43.697 --> 00:33:45.777 forward most structure in the brain. 00:33:45.917 --> 00:33:49.377 So that the layout of genes from front to back is very similar. 00:33:49.377 --> 00:33:53.077 And the kinds of functions that are going front to back are very similar. 00:33:53.237 --> 00:33:59.297 So the most forward feature is in almost all brains, sensing chemicals, 00:33:59.517 --> 00:34:03.877 which is the nose, basically, as far forward as you can. 00:34:04.297 --> 00:34:08.277 Behind that, you need, in part, I think about it because chemicals are one of 00:34:08.277 --> 00:34:12.057 those things that you can get a sense of from a very long distance because you 00:34:12.057 --> 00:34:13.817 get a gradient of concentration. 00:34:13.817 --> 00:34:16.737 Concentration and that can actually predict not only 00:34:16.737 --> 00:34:19.797 into the future where you're going but actually what was there 00:34:19.797 --> 00:34:22.497 uh it is a wonderful predictive even into the 00:34:22.497 --> 00:34:26.777 past you know right in a sense retro addictive kind of capability and so i like 00:34:26.777 --> 00:34:30.617 to think about the furthest forward parts of the brain as being the most predictive 00:34:30.617 --> 00:34:35.437 parts and as you move back they're more and more proximate to the body so for 00:34:35.437 --> 00:34:42.177 example the next one back um typically is vision vision can predict things in far distance. 00:34:43.597 --> 00:34:47.257 But as you move back from that, you get things that have to do with touch, 00:34:47.397 --> 00:34:51.777 whether it's the touch of sound, which is basically a modified touch system, 00:34:52.497 --> 00:34:56.017 or the touch of the surface of your body. Now it's very proximate. 00:34:56.737 --> 00:34:59.977 So then, in effect, you've got this sort of projection system out front. 00:35:00.197 --> 00:35:05.737 I think one of the reasons that there's so much similarity in brains of animals 00:35:05.737 --> 00:35:09.997 that have a brain and a front end is precisely because of that. 00:35:10.782 --> 00:35:14.782 Precisely because of that need of movement. And what's interesting is that recently 00:35:14.782 --> 00:35:18.662 it's been discovered, now that we understand the genes that are generating, 00:35:18.862 --> 00:35:21.382 laying out that pattern in the first place. 00:35:22.222 --> 00:35:28.382 It's a very primitive pattern. So primitive that it's now possible that you 00:35:28.382 --> 00:35:32.942 can take the gene that's responsible for producing much of the forebrain in humans. 00:35:32.982 --> 00:35:38.902 It's called OTX2 is the name given to it. It was actually named for a homologous 00:35:38.902 --> 00:35:41.402 gene in the fly called orthodentical. 00:35:42.762 --> 00:35:46.902 And the fly gene produces the dorsal part of the head, the front part of the 00:35:46.902 --> 00:35:49.822 head, actually, and much of the brain. 00:35:50.202 --> 00:35:55.122 Now, it's not the gene that codes for the brain. It's what codes for the very front part. 00:35:56.542 --> 00:35:59.482 In vertebrates like you and I, it also codes for the front part, 00:35:59.622 --> 00:36:01.782 which turns out to be much of our forebrain. 00:36:03.822 --> 00:36:09.582 Otx2 and orthodentical are very similar to each other so scientists not in our 00:36:09.582 --> 00:36:14.382 group but other scientists have have eliminated the ot the orthodentical gene 00:36:14.382 --> 00:36:19.822 in flies so that the fly if it develops does not develop a normal head it's 00:36:19.822 --> 00:36:21.322 lacking a big part of its brain, 00:36:22.242 --> 00:36:28.202 but the human gene the human otx2 gene which is related to this gene can be 00:36:28.202 --> 00:36:30.262 spliced back into the fly genome. 00:36:30.762 --> 00:36:35.402 And a fly can develop, and its forebrain will develop. 00:36:35.742 --> 00:36:39.162 It's not absolutely normal, but it's remarkably normal. 00:36:39.502 --> 00:36:44.742 And what that tells you, first of all, is that that gene is not about building our brain or a fly brain. 00:36:44.982 --> 00:36:48.542 It's about starting out. There's going to be a partition up here. 00:36:49.422 --> 00:36:50.882 And that tells you that, first of 00:36:50.882 --> 00:36:54.942 all, the gene is carrying information in context, in a sense like a word. 00:36:55.062 --> 00:36:57.622 Word can mean something very different in a different context. 00:36:57.882 --> 00:37:00.782 Well, Well, this gene can mean something very different in a different context, 00:37:00.942 --> 00:37:04.502 but only in a context that's up in the front of the head. Right, exactly. 00:37:05.762 --> 00:37:09.182 So one of the things that's been important about this is that we've discovered 00:37:09.182 --> 00:37:12.182 that that initial plan is very, 00:37:12.282 --> 00:37:17.642 very primitive because flies and humans carry a common ancestor that was probably 00:37:17.642 --> 00:37:20.802 nearly a half a billion years back in time. 00:37:20.982 --> 00:37:25.622 That tells you that that conservatism has been around a very, very long time. 00:37:26.382 --> 00:37:29.302 Now, I think the reason it has is because so much else depends upon it. 00:37:29.642 --> 00:37:34.382 Once everything else depends upon it, like it's like the trunk of a tree, 00:37:34.482 --> 00:37:36.002 everything else depends upon it. 00:37:36.082 --> 00:37:41.562 And if you damage that or change it in any fundamental way, then lots of things will fail. 00:37:41.822 --> 00:37:46.702 And so I think one of the reasons that these early stages are conserved is for exactly that reason. 00:37:47.791 --> 00:37:52.351 And now tell me, so how would you call this principle? 00:37:52.531 --> 00:37:56.471 Would you still say like, okay, there's like the front, the middle and the back 00:37:56.471 --> 00:38:01.751 and the front is more focused towards, let's say, the environment and predicting 00:38:01.751 --> 00:38:04.031 your interaction with environment. 00:38:04.311 --> 00:38:06.431 And the more we go to the back, the more we start to deal with, 00:38:06.491 --> 00:38:10.911 let's say, the body and control of the body and the bit in the middle interfaces 00:38:10.911 --> 00:38:14.031 between these two systems. Is this roughly a reasonable summary? 00:38:14.131 --> 00:38:19.631 It's not bad. And this is where the McLean idea is, you might say, 00:38:19.651 --> 00:38:21.571 an interesting intuition to get started. 00:38:21.691 --> 00:38:25.511 Because what he claims is that, of course, what he calls the reptilian brain 00:38:25.511 --> 00:38:30.191 is mostly involved in automatic behaviors and regulating the body. 00:38:31.531 --> 00:38:37.231 What he called the paleomammalian brain is mostly involved in an interface between 00:38:37.231 --> 00:38:45.371 the exteroceptive senses and motor control systems. systems and these automatic drive systems. 00:38:45.651 --> 00:38:48.951 So it's the system that he associated with arousal emotions. 00:38:49.291 --> 00:38:52.511 And emotion, if you think about the nature of emotion, it's sort of, 00:38:52.551 --> 00:38:55.971 I like to think about it as the accelerator and the brake pedal problem. 00:38:56.191 --> 00:39:00.411 That what you, even though you want to go fast and you're revving your motor, 00:39:00.711 --> 00:39:04.891 sometimes you have to hold the brake because conditions aren't quite right. 00:39:05.131 --> 00:39:11.491 Right. So he basically considered the, what he called it, the limbic system. 00:39:12.051 --> 00:39:16.731 It turns out to be almost entirely four brain systems, telencephalic systems, 00:39:17.091 --> 00:39:20.671 that are very far forward in the brain. 00:39:21.091 --> 00:39:25.851 His argument was, however, that in evolution, one was added on top of the other. 00:39:25.931 --> 00:39:31.431 We now know that all three of those layers in various forms were there in all 00:39:31.431 --> 00:39:33.611 vertebrates, even in the simplest ones we look at. 00:39:34.131 --> 00:39:39.491 What's another similarity, however, is that as brains have gotten bigger, 00:39:39.631 --> 00:39:41.631 and particularly as we move to birds and mammoths, 00:39:42.051 --> 00:39:44.831 um there is a progressive enlargement of 00:39:44.831 --> 00:39:47.851 those more forward than those more behind 00:39:47.851 --> 00:39:52.751 so that the teal encephalon enlarges much more rapidly than the rest of the 00:39:52.751 --> 00:39:56.811 brain and this is true even going from small mammal brains to large mammal brains 00:39:56.811 --> 00:40:01.751 that the proportion of the teal encephalon to the rest of the brain behind it 00:40:01.751 --> 00:40:05.971 teal encephalon being the part most far far forward um is more 00:40:06.071 --> 00:40:08.171 equivalent in, say, a mouse or a rat. 00:40:08.731 --> 00:40:14.511 But when you get to a human, it is completely overshadowing the rest of the brain. 00:40:14.651 --> 00:40:17.191 And so we see, when we look at the surface of a human brain, 00:40:17.331 --> 00:40:20.771 all we really see is telencephalon, and then at the back, the cerebellum. 00:40:21.511 --> 00:40:25.891 Cerebellum is probably the exception to this, because the cerebellum has also 00:40:25.891 --> 00:40:28.671 enlarged, but it's partway down the system. 00:40:29.453 --> 00:40:33.713 But that means the way you're describing it would suggest that it would be maybe 00:40:33.713 --> 00:40:40.713 not reasonable to treat the components of this brain as independent modules. 00:40:41.653 --> 00:40:46.233 Because apparently from the beginning, there's a plan that really tries to keep 00:40:46.233 --> 00:40:50.233 all these three structures together and also as a co-evolving system. 00:40:50.493 --> 00:40:54.433 Right. So would you agree with that? Or do you believe that we can look at this 00:40:54.433 --> 00:40:56.873 nervous system from this more modular perspective as, okay, 00:40:56.913 --> 00:41:03.653 let's just understand what this cortical column does in isolation from its connections 00:41:03.653 --> 00:41:07.473 to subcortical structures and surrounding tissue and so on. So do you believe 00:41:07.473 --> 00:41:08.373 in these kinds of modules? 00:41:08.853 --> 00:41:14.133 Or should we really think more about, let's say, an integrated system that we 00:41:14.133 --> 00:41:19.673 cannot just decompose in those terms? Well, certainly I strongly am leaning 00:41:19.673 --> 00:41:22.393 towards the latter, a more holistic view of the brain. 00:41:23.033 --> 00:41:28.033 But I think it's important to recognize that as brains develop, 00:41:28.213 --> 00:41:29.493 they become more modular. 00:41:30.333 --> 00:41:34.833 And even thinking about function, a skill ultimately becomes modular. 00:41:34.953 --> 00:41:40.993 That is, it becomes more unconscious, more automatic, more invariant over time as we develop the skill. 00:41:41.153 --> 00:41:44.973 So in one sense, there's functional modularity. One of the things that brains 00:41:44.973 --> 00:41:48.893 are about is generating functional modularity to produce this, 00:41:48.953 --> 00:41:51.153 because when you have something that's extremely predictable, 00:41:51.433 --> 00:41:53.913 it's ideal if you don't have to think about it. 00:41:53.993 --> 00:42:01.413 It's ideal if you can run it off in its most optimal form, but it doesn't start out that way usually. 00:42:01.673 --> 00:42:04.493 Now, I think in general, that's true about development of the brain. 00:42:04.613 --> 00:42:09.493 The brain develops and is relatively undifferentiated, and these so-called modules, 00:42:09.713 --> 00:42:11.333 modular parts, develop. 00:42:11.333 --> 00:42:14.813 Developed so columns are a fairly late developing feature the 00:42:14.813 --> 00:42:18.053 columns are something that are the result of axons coursing 00:42:18.053 --> 00:42:22.753 into the cortex and competing with each other and chasing each other out so 00:42:22.753 --> 00:42:27.873 to speak from different sources and creating this kind of sculpting into modules 00:42:27.873 --> 00:42:32.553 so in that sense the module is a consequence not a cause so to speak of this 00:42:32.553 --> 00:42:35.953 process right so not to sort of. 00:42:36.989 --> 00:42:43.649 Get me to the conclusions of this discussion. So you've been very deeply involved 00:42:43.649 --> 00:42:47.189 in this research. You have a deep understanding of the brain. 00:42:47.849 --> 00:42:51.009 We're trying to catch up, which will take a lot of time. 00:42:51.229 --> 00:42:57.209 So what's this one law of Terence Deacon you want us to adhere to in understanding mind and brain? One law? 00:42:57.489 --> 00:43:00.209 Yeah, only one. Only one? Yeah, yeah, yeah. 00:43:00.729 --> 00:43:05.569 I have been quoted for a law. Some people even call it Deacon. 00:43:05.569 --> 00:43:07.709 You're right people, but you're wrong people. Oh, by the right people. Oh, okay. 00:43:08.909 --> 00:43:13.749 And it was called the displacement hypothesis. Now, it's not a law about brain 00:43:13.749 --> 00:43:16.889 function. It's a law about brain development. 00:43:17.669 --> 00:43:20.869 And the displacement hypothesis has basically— It's not a law. 00:43:20.969 --> 00:43:23.909 It's not a hypothesis. I know, but it is a law. I think it is a law. Okay. 00:43:24.209 --> 00:43:27.289 Now, there are probably many laws. I'm not sure which is my favorite one, 00:43:27.329 --> 00:43:30.669 but this is the one that I'm known for, so I'll keep it going. on. 00:43:31.689 --> 00:43:37.129 The displacement rule, if you want to think about it, is that because connections 00:43:37.129 --> 00:43:42.949 in the brain in the final stages of development are competing with each other 00:43:42.949 --> 00:43:45.009 for space, competing for targets, 00:43:45.769 --> 00:43:48.829 they compete on the basis of signals passing through them. 00:43:48.969 --> 00:43:51.289 We call it activity-dependent competition. 00:43:52.369 --> 00:43:57.289 And what happens is that there's a standard rule, and it's not my rule, this is an. 00:44:00.549 --> 00:44:04.909 The neurons that fire together wire together. It's not wrong, 00:44:04.969 --> 00:44:07.669 and it really dates back to the work of Donald Hebb, of course. 00:44:08.249 --> 00:44:11.869 But it turns out that size matters. 00:44:12.809 --> 00:44:18.129 Because if you've got many axons coming into a target that fire together, 00:44:18.789 --> 00:44:24.249 they're going to have more control over that target than smaller numbers of axons. 00:44:24.789 --> 00:44:31.049 And so one of my arguments about the development of brains is that as brains 00:44:31.049 --> 00:44:32.529 have developed and evolved, 00:44:32.929 --> 00:44:37.069 one of the ways that they've changed function and biased their function one 00:44:37.069 --> 00:44:42.149 way or another is to cause more cells to be produced somewhere as opposed to somewhere else. 00:44:42.429 --> 00:44:47.769 And in doing that, you cause those that are enlarged to have a better chance 00:44:47.769 --> 00:44:49.709 of making connections in some place. 00:44:49.869 --> 00:44:53.449 And those that have been shrunk down will tend to lose their connections. 00:44:53.709 --> 00:44:59.189 And so instead of having specific information about where your wires should go. 00:45:00.422 --> 00:45:06.142 It allows the brain to send the wires many places and then let the competition decide. 00:45:06.522 --> 00:45:10.402 So a fairly simple process of just changing relative numbers, 00:45:10.562 --> 00:45:14.382 changing how much, how many cell divisions take place here and there, 00:45:14.502 --> 00:45:17.142 you actually can control the wiring of the brain. 00:45:17.302 --> 00:45:20.302 Right. So Deacon's law is size matters. Size matters. 00:45:20.602 --> 00:45:24.962 Cool. And the important thing about that is, of course, human brains are unusually 00:45:24.962 --> 00:45:29.842 large, unusually large for our bodies. and we have parts of our brains that 00:45:29.842 --> 00:45:32.602 are unusually enlarged with respect to other parts. 00:45:32.762 --> 00:45:38.662 And what that means is that that's a clue about what makes our brains different than other brains. 00:45:38.962 --> 00:45:44.302 Very good. Then to finish up, if five years from now I'm going to go visit you, 00:45:44.822 --> 00:45:48.922 and I'm going to ask you, look, Terence, five years back you made this one prediction 00:45:48.922 --> 00:45:50.782 and today I'm going to check whether it came out. 00:45:51.042 --> 00:45:53.842 What's this one prediction which you really commit yourself to today? 00:45:54.422 --> 00:45:56.482 My one prediction is not about the brain. 00:45:57.182 --> 00:45:59.262 That's fine. It's about the origins of life. 00:46:00.362 --> 00:46:03.702 And I have argued, and it's now a big part of my work, 00:46:03.822 --> 00:46:07.462 that the simplest reproductive molecular 00:46:07.462 --> 00:46:14.402 process that reproduces itself and maintains the constraints to maintain its 00:46:14.402 --> 00:46:20.102 own structure and reproductive capacity is made up of two reciprocal self-organizing 00:46:20.102 --> 00:46:25.502 processes that happen all the time in living processes, not very often in non-living processes. 00:46:26.393 --> 00:46:30.333 One is called self-assembly. It's a process by which molecules, 00:46:30.513 --> 00:46:32.973 for example, form the coats of viruses. 00:46:33.233 --> 00:46:37.273 And with similar molecules, because they have a geometric relationship to each 00:46:37.273 --> 00:46:42.333 other, they tend to form polyhedrons or tubes spontaneously because they stick 00:46:42.333 --> 00:46:45.173 together edge to edge. They form containers spontaneously. 00:46:45.733 --> 00:46:50.833 The other process that's ubiquitous in the cell biology, the molecular biology 00:46:50.833 --> 00:46:53.053 of the cell, is what we call autocatalysis. 00:46:53.053 --> 00:46:56.113 Processes catalysts that have a circular relationship 00:46:56.113 --> 00:46:59.213 where one catalyst produces another produces another which produces 00:46:59.213 --> 00:47:02.633 the first but that means it will amplify that process 00:47:02.633 --> 00:47:05.633 it turns out if you mix those two processes 00:47:05.633 --> 00:47:11.513 together each of them have demands that are required in order to work and they 00:47:11.513 --> 00:47:16.513 produce form as a consequence they produce a regularity as a consequence and 00:47:16.513 --> 00:47:21.433 it turns out that the boundary conditions that allow each of them are produced 00:47:21.433 --> 00:47:23.873 by the other. So let me describe what I mean. 00:47:24.053 --> 00:47:29.953 And that is that the boundary conditions for autocatalysis, the crucial one 00:47:29.953 --> 00:47:35.373 is having all the reciprocal catalysts that depend on each other have to be 00:47:35.373 --> 00:47:37.733 proximate to each other. They have to stay close together. 00:47:38.173 --> 00:47:42.633 But typically what will happen in a solution, of course, is they will diffuse away from each other. 00:47:42.793 --> 00:47:48.213 So autocatalytic processes in an open solution basically undermine themselves. 00:47:48.893 --> 00:47:52.433 But how do I make a textual prediction Prediction, which can be wrong. 00:47:52.693 --> 00:47:57.093 So that's what I'm going to do here. So the second thing, if an autocatalytic 00:47:57.093 --> 00:48:03.353 process produces as a side product a molecule that self-assembles into a container. 00:48:04.513 --> 00:48:08.933 Then that container will tend to grow where the autocatalysis is fastest. 00:48:09.433 --> 00:48:14.393 And become an amplifier. Well, it will enclose, actually. So what will happen 00:48:14.393 --> 00:48:16.393 is it will tend to capture those catalysts. 00:48:16.913 --> 00:48:23.613 Now, catalysis will therefore stop. it will become non-dynamic because you'll run out of substrate. 00:48:24.273 --> 00:48:29.733 But now you have a complex, I call it an autocell, that has enclosed, 00:48:30.233 --> 00:48:35.293 but it encloses the catalysts that are necessary to make it again if it's broken. 00:48:35.793 --> 00:48:40.553 So that if it's broken in any kind of environment, say heated up and it breaks 00:48:40.553 --> 00:48:45.693 open, what will happen is now the catalysts can begin to make more catalysts 00:48:45.693 --> 00:48:47.393 and make more or shell molecules, 00:48:47.853 --> 00:48:51.273 they will tend to, shell will tend to form where there's the most catalysts. 00:48:51.273 --> 00:48:54.913 They will enclose it again and will now have a capacity to reproduce. 00:48:55.633 --> 00:49:00.593 So what I've argued is that, in fact, life doesn't start with DNA or RNA. 00:49:00.813 --> 00:49:03.293 It starts out as simple as this. 00:49:04.230 --> 00:49:07.730 And that this is something that we can produce in the laboratory. 00:49:07.850 --> 00:49:12.050 Now, we have not succeeded yet, but I think we can. And we've done so. 00:49:12.110 --> 00:49:16.470 We begin to approach it by simulations, trying to get more and more chemically 00:49:16.470 --> 00:49:18.770 accurate simulations of this process. 00:49:19.010 --> 00:49:26.070 So five years from now, I can see this self-replicating autocatalytic process at work in your lab. 00:49:26.070 --> 00:49:34.350 And I think it will be very useful for nanotechnology because it's the key feature 00:49:34.350 --> 00:49:37.390 to nanotechnology to make it work is self-replication. 00:49:37.450 --> 00:49:41.530 If you can get self-replication, now nanotechnology becomes very, 00:49:41.590 --> 00:49:42.570 very powerful and robust. 00:49:42.950 --> 00:49:46.870 Well, this is a system much simpler than a biological system, 00:49:46.930 --> 00:49:53.950 no DNA, no RNA, that replicates itself in a very simple way and could be controlled. 00:49:53.950 --> 00:49:56.930 Interestingly enough the process is also 00:49:56.930 --> 00:50:00.110 not chemically specific any molecules 00:50:00.110 --> 00:50:02.850 that can act as catalysts and any molecules that can 00:50:02.850 --> 00:50:08.590 act as self-assembling molecules are capable of doing that that may not just 00:50:08.590 --> 00:50:13.630 be organic molecules this is a general chemical process that i think is possibly 00:50:13.630 --> 00:50:19.190 capable and even mineral like processes i happen to think that the beginnings 00:50:19.190 --> 00:50:22.810 of this are the beginnings of life, even though this is not really alive. 00:50:23.010 --> 00:50:27.470 There's no constant metabolism here. There's no information molecules. 00:50:27.870 --> 00:50:29.950 But I think this is how life begins. 00:50:30.390 --> 00:50:32.710 Now, there's one extra step to this. 00:50:33.430 --> 00:50:38.090 To start this process, you need catalysts. And catalysts are typically molecules 00:50:38.090 --> 00:50:39.670 with large reactive surfaces. 00:50:40.630 --> 00:50:46.150 In the primitive Earth, any place where there's water, You tend to break big 00:50:46.150 --> 00:50:48.750 molecules up. They dissolve into smaller molecules. 00:50:49.030 --> 00:50:53.850 One of the problems with origins of life stories is that we don't really have 00:50:53.850 --> 00:50:58.490 ways that these molecules can be built up unless they get dried out on clay or something like that. 00:50:59.702 --> 00:51:03.522 I'm actually convinced that this process begins in the outer solar system. 00:51:04.282 --> 00:51:09.502 It begins with molecules made up of hydrogen cyanide. 00:51:09.662 --> 00:51:16.082 It turns out that probably the most abundant large molecules in our solar system 00:51:16.082 --> 00:51:18.782 are hydrogen cyanide polymers, 00:51:18.902 --> 00:51:24.862 in which hydrogen cyanide is a carbon-nitrogen-hydrogen link. 00:51:24.862 --> 00:51:27.782 And it makes a polymer by getting 00:51:27.782 --> 00:51:30.642 rid of the hydrogen and leaking another nitrogen to the carbon 00:51:30.642 --> 00:51:33.742 the result is you get a polymer that is 00:51:33.742 --> 00:51:36.602 carbon nitrogen carbon nitrogen carbon nitrogen carbon nitrogen 00:51:36.602 --> 00:51:39.182 and it then can fold up and 00:51:39.182 --> 00:51:43.282 make a complicated three-dimensional shape it can only form in the absence of 00:51:43.282 --> 00:51:48.882 liquid water but in the outer solar system where water is all frozen it can 00:51:48.882 --> 00:51:52.802 form easily and there are many people who now think that the dark bands on some 00:51:52.802 --> 00:51:56.622 of these planets are very rich hydrogen cyanide polymer zones. 00:51:57.162 --> 00:52:01.142 What's interesting about hydrogen cyanide polymers is that backbone I just described 00:52:01.142 --> 00:52:03.162 is the backbone of a protein. 00:52:03.362 --> 00:52:06.462 A protein has a backbone that's carbon-nitrogen, carbon-nitrogen, 00:52:06.502 --> 00:52:08.562 carbon-nitrogen, with side chains on it. 00:52:09.002 --> 00:52:13.642 It turns out when you put hydrogen cyanide polymer molecules, 00:52:13.842 --> 00:52:16.982 they're called polyamidines, in water, their side 00:52:16.982 --> 00:52:20.022 chains get eliminated and replaced by 00:52:20.022 --> 00:52:23.302 carbohydrates and what results is peptides 00:52:23.302 --> 00:52:26.542 that is not amino acids first 00:52:26.542 --> 00:52:32.122 but proteins first but beginning from these molecules from from the outer planets 00:52:32.122 --> 00:52:37.102 so i my view of the origins of life and i think it's testable because i think 00:52:37.102 --> 00:52:42.062 first of all we're going to be able to send probes out and identify the amount 00:52:42.062 --> 00:52:44.122 of hydrogen cyanide polymers that are out there. 00:52:44.322 --> 00:52:48.582 And second of all, we're going to be able to find, I think, some of these primitive 00:52:48.582 --> 00:52:51.802 forms that I just described, these what I call autocels. 00:52:52.562 --> 00:52:55.382 I think we're going to find fossils of autocels on Mars. 00:52:56.042 --> 00:52:59.882 And I think so because the first place that they would have likely formed will 00:52:59.882 --> 00:53:03.502 have been on Mars early on when there was some liquid water on Mars. 00:53:03.622 --> 00:53:08.522 But it's closest to the outer planets where it's going to get a strong rain of these molecules. 00:53:08.842 --> 00:53:12.722 So I have a series of predictions about the origins of life that I think are 00:53:12.722 --> 00:53:14.962 going to turn out to be very good. 00:53:15.322 --> 00:53:18.922 Terence Deacon, thank you very much for this conversation. I'll see you in five years. All right.