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 Verschure and Tony Prescott. 00:00:17.297 --> 00:00:22.357 So Paul Verschure with the Convergent Science Network podcast here at BCPD15 00:00:22.357 --> 00:00:27.897 together with Tony Prescott. Scott, and we're talking with Stefan Noctor, 00:00:28.057 --> 00:00:31.517 who is a great expert in development of the brain. 00:00:32.037 --> 00:00:37.717 And so, Stephen, you started out by laying out this whole problem of, 00:00:37.857 --> 00:00:43.137 okay, how can we actually control from an evolutionary perspective the size of a brain? 00:00:43.257 --> 00:00:48.237 What are the mechanisms underlying that, right? So, why are you concerned with 00:00:48.237 --> 00:00:49.057 that specific question? 00:00:49.177 --> 00:00:52.457 Why do you think that's the most important one to answer? Well, 00:00:52.557 --> 00:00:55.917 I wouldn't say it's the most important one to answer, but it's certainly interesting to me. 00:00:56.857 --> 00:01:04.157 And my initial interest, and it still is my main interest, is understanding 00:01:04.157 --> 00:01:10.017 the precursor cells that are producing, they're the engine for growth. 00:01:10.017 --> 00:01:13.957 So they're producing all the cells, identifying them, 00:01:14.057 --> 00:01:18.077 characterizing them, and then figuring out what cell types they produce under 00:01:18.077 --> 00:01:24.697 what conditions and defining their functioning in a normal brain, 00:01:24.817 --> 00:01:25.797 normal developing brain. 00:01:26.137 --> 00:01:30.597 And hoping that that will establish a firm foundation for understanding disease 00:01:30.597 --> 00:01:33.617 processes and how they influence development of the brain. 00:01:33.977 --> 00:01:38.757 Right. So what are the different preparations that you look at or have looked 00:01:38.757 --> 00:01:43.957 at? I use a variety of animal models. 00:01:44.057 --> 00:01:46.757 So the bulk of my work has been in rat. 00:01:47.057 --> 00:01:55.537 And we do culture models where we will inject different vectors into the prenatal 00:01:55.537 --> 00:02:01.757 brain and make cultured slices and observe the cultured slices under a microscope 00:02:01.757 --> 00:02:03.857 for four or five days at a time. 00:02:04.477 --> 00:02:09.597 And then to make sure that what we're seeing is something that's physiological 00:02:09.597 --> 00:02:11.157 and not a culture artifact. 00:02:11.217 --> 00:02:15.937 We'll inject vectors in utero in animals and allow them to survive the same 00:02:15.937 --> 00:02:20.017 amount of time and then compare the endpoints to make sure we're ending up with 00:02:20.017 --> 00:02:26.117 the same thing to make sure that what we see is real and physiologically relevant. 00:02:26.537 --> 00:02:29.637 Right. So the standard model, if you want, 00:02:29.757 --> 00:02:35.057 of brain development is this whole idea that we have sort of zones where cells 00:02:35.057 --> 00:02:40.437 are being generated that then sort of swim out radially from these zones where 00:02:40.437 --> 00:02:41.657 we generate these cells. 00:02:42.297 --> 00:02:46.197 Is that sort of the standard model we should still think about with any brain 00:02:46.197 --> 00:02:48.277 we look at or any vertebrate brain we look at? 00:02:48.777 --> 00:02:53.657 It's true mostly for the excitatory cells. So we spoke a little bit about that 00:02:53.657 --> 00:02:56.097 earlier today. But the excitatory cells. 00:02:57.599 --> 00:03:03.559 For the most part, have a radial trajectory. They migrate out along these radial 00:03:03.559 --> 00:03:08.219 fibers that are sort of like the spokes on a bicycle wheel, and they're deployed 00:03:08.219 --> 00:03:10.179 in the growing cortical gray matter. 00:03:10.539 --> 00:03:15.099 But the cortical interneurons have a different origin, and they have much longer 00:03:15.099 --> 00:03:17.919 roots that are migrating. 00:03:17.939 --> 00:03:20.399 They're migrating perpendicular to the radial deployment. 00:03:21.239 --> 00:03:27.179 And yeah, that's a whole other field. I'm looking primarily at the excitatory cells. 00:03:27.679 --> 00:03:34.819 But now the excitatory cells that you talked about actually are showing a very 00:03:34.819 --> 00:03:37.759 rapid growth in terms of numbers, not the sheer numbers. 00:03:38.099 --> 00:03:41.659 I mean, you mentioned something like five billion cells in seven weeks, 00:03:41.759 --> 00:03:42.879 which sounds astonishing. 00:03:43.699 --> 00:03:46.779 In the human brain, yeah, in the human brain, yeah. So in what period is that 00:03:46.779 --> 00:03:49.279 growth explosion really taking place? 00:03:50.979 --> 00:03:53.699 Well, we're really only now starting to get 00:03:53.699 --> 00:03:56.939 a handle on human brain development because there 00:03:56.939 --> 00:04:01.259 have not been a lot of experiments and uh there's studies 00:04:01.259 --> 00:04:05.139 now showing that um cortical excitatory 00:04:05.139 --> 00:04:08.639 neurons in human are being 00:04:08.639 --> 00:04:11.699 generated uh up until birth and 00:04:11.699 --> 00:04:14.759 then the interneurons may be um just a 00:04:14.759 --> 00:04:17.799 little bit longer so throughout pregnancy 00:04:17.799 --> 00:04:20.819 but starting at maybe six weeks 00:04:20.819 --> 00:04:23.659 and throughout the whole time we don't have 00:04:23.659 --> 00:04:26.679 a good rate yet so we don't know if it's tailing 00:04:26.679 --> 00:04:29.679 off you know do we reach a maximum at 20 weeks and then it starts tailing 00:04:29.679 --> 00:04:32.559 off those sorts of data we don't have yet on the 00:04:32.559 --> 00:04:35.879 human brain so it's been well mapped out in mouse and in 00:04:35.879 --> 00:04:38.879 rat and to you know to some degree and ferret and 00:04:38.879 --> 00:04:41.739 and we're starting to get a better handle on on 00:04:41.739 --> 00:04:44.659 primates like a rhesus macaque so now 00:04:44.659 --> 00:04:47.519 you um so in 00:04:47.519 --> 00:04:51.159 this process of of the the construction 00:04:51.159 --> 00:04:54.539 if you want of a neocortex um there 00:04:54.539 --> 00:04:58.679 are a number of principles at work and what's really fascinating in your talk 00:04:58.679 --> 00:05:01.819 is also how you look at this from a more historical perspective to say look 00:05:01.819 --> 00:05:07.199 actually people start to have some ideas about that process early on So what 00:05:07.199 --> 00:05:10.299 do you see really as sort of the highlights or maybe the forgotten highlights 00:05:10.299 --> 00:05:13.399 of that history that we should bear in mind here? 00:05:13.839 --> 00:05:18.239 The highlights of the forgotten history. Well, it was known well over 40 years 00:05:18.239 --> 00:05:24.239 ago, actually 50 years ago, that there was a generation of new neurons in the postnatal brain. 00:05:25.259 --> 00:05:29.379 And there was some hints, some evidence that it might be occurring in the adult brain too. 00:05:29.959 --> 00:05:36.419 And that's something that we began to rediscover in the 90s and early 2000s. Mm-hmm. 00:05:37.272 --> 00:05:45.532 But now, so in your work, what you have been sort of also pioneering is to bring 00:05:45.532 --> 00:05:49.872 different technologies together to really get a handle on the movement of cells 00:05:49.872 --> 00:05:54.412 in the ventricular zone and the subventricular zone that builds the cortex, 00:05:54.772 --> 00:05:57.552 but also their specialization and differentiation. 00:05:58.072 --> 00:06:01.752 So what are the key technologies that you use for that? up? 00:06:01.932 --> 00:06:08.812 So my approach has been, I think, relatively simple, and it just requires patience. 00:06:09.032 --> 00:06:13.412 So labeling cells with fluorescent tags, which became available in the late 00:06:13.412 --> 00:06:17.292 90s, so that you can label fluorescent cells, and then using an appropriate 00:06:17.292 --> 00:06:19.592 vector that will label some of them but not too many. 00:06:19.692 --> 00:06:24.552 Because if you have a field of similarly labeled cells, you can't follow any single one of them. 00:06:24.672 --> 00:06:28.192 So you need a small population of labeled cells. 00:06:28.752 --> 00:06:33.932 And then preparing a a culture condition that allows the cells to move. 00:06:34.052 --> 00:06:38.672 So it probably took me about a year to get the culture conditions right so that 00:06:38.672 --> 00:06:42.792 the cells would proliferate, divide, and then once the new neurons were born 00:06:42.792 --> 00:06:43.452 that they would migrate. 00:06:45.312 --> 00:06:49.852 Once that was working, then it was just a matter of being patient enough to 00:06:49.852 --> 00:06:53.392 sit there and take pictures every hour, 00:06:53.472 --> 00:06:56.692 every half hour, depending on what's going on until you couldn't 00:06:56.692 --> 00:06:59.572 take it any longer um but to watch 00:06:59.572 --> 00:07:02.732 it happen over like five days these are beautiful trajectories that 00:07:02.732 --> 00:07:05.472 you you showed in the in the movies that 00:07:05.472 --> 00:07:10.292 you put together sort of time lapse and you really see cells sort of moving 00:07:10.292 --> 00:07:15.092 inside a brain and for people that normally think about adult brains or think 00:07:15.092 --> 00:07:18.792 about neural network models the thought that these that these actually get up 00:07:18.792 --> 00:07:24.512 and move from one side to another and then change direction that's quite astonishing. 00:07:24.972 --> 00:07:28.932 I mean, obviously it has to happen, but to, and you talked about the distance, 00:07:29.732 --> 00:07:33.532 that neurons having to travel being the equivalent of several skyscrapers. 00:07:33.552 --> 00:07:39.392 And so that also is an amazing thing to know about brains that we don't often think about. 00:07:41.012 --> 00:07:45.012 And there's a special offer, the precursor cells that you're talking about. 00:07:45.172 --> 00:07:48.872 And are these cells that are really just there, 00:07:50.032 --> 00:07:53.052 at the earlier stages of development or do we 00:07:53.052 --> 00:07:57.152 lose these i mean are they are they very kind of so it's a great question so 00:07:57.152 --> 00:08:02.012 i think the we don't really have a good handle on the lifespan of an individual 00:08:02.012 --> 00:08:06.692 precursor so right so are they always giving rise to somebody else that takes 00:08:06.692 --> 00:08:12.052 on the job and continues on and then they pass it on i don't think we really have a good handle so um, 00:08:13.030 --> 00:08:17.230 Does a single precursor cell generate neurons dedicated for each of the cortical layers? 00:08:17.330 --> 00:08:21.810 There's some hints that maybe they do, but we don't really know that well enough. 00:08:21.990 --> 00:08:24.070 The experiments that I did were overlapping. 00:08:24.270 --> 00:08:28.910 They would last four or five days, but they wouldn't continue for the entire period. 00:08:29.390 --> 00:08:34.750 Right. And then at the end, there's still some left. And where do they go? 00:08:35.630 --> 00:08:42.130 Okay. And in hippocampus, presumably, the embryonic cells give rise to precursors 00:08:42.130 --> 00:08:45.490 that remain in the hippocampus, because we know it's now well known that you 00:08:45.490 --> 00:08:49.030 have continued neurogenesis in the hippocampus throughout life. 00:08:49.350 --> 00:08:52.790 Well, dentagyrus, you know? So what makes it a precursor cell? 00:08:52.870 --> 00:08:55.290 Is it just that it gives rise to other cells? Yeah. 00:08:56.330 --> 00:09:01.110 And to me, that's a crucial point because there's one term that people use, 00:09:01.230 --> 00:09:05.330 neuroblast, and some people use it to mean a young immature neuron that's migrating. 00:09:05.390 --> 00:09:07.390 Other people use it to mean something that can divide. 00:09:09.130 --> 00:09:13.390 But yeah, the fact that it can undergo division. Would it not be possible to 00:09:13.390 --> 00:09:18.470 take a cell, a neuron, and turn it into a precursor? I mean, you could... 00:09:18.930 --> 00:09:22.110 There are some people who believe that can happen under pathological conditions. 00:09:22.110 --> 00:09:26.910 Okay, but under normal conditions, there are the precursors, 00:09:26.910 --> 00:09:30.790 and then they give rise to these other neurons which don't proliferate further. 00:09:30.950 --> 00:09:35.670 Right, yeah. So, yeah, the century-old dogma that once a neuron is a neuron, 00:09:35.770 --> 00:09:36.870 it's not going to divide anymore. 00:09:37.370 --> 00:09:42.590 And does each neuronal type have its own precursor, or how does that work? 00:09:43.190 --> 00:09:48.190 I think we're still working that out, so what different cell types precursors 00:09:48.190 --> 00:09:50.470 can give rise to. So, um, yeah. 00:09:52.334 --> 00:09:58.454 One thing that is clear from rodent, and I think it's being worked out in different 00:09:58.454 --> 00:10:02.614 models, but for example, the two basic types, the excitatory and the inhibitory, 00:10:02.614 --> 00:10:04.054 come from different regions of the brain. 00:10:04.974 --> 00:10:10.154 So you have in the basal forebrain in a structure called the medial ganglionic eminence. 00:10:10.214 --> 00:10:12.974 There's also another one called the caudal ganglionic eminence that gives rise 00:10:12.974 --> 00:10:14.734 to interneurons in the rodent. 00:10:14.994 --> 00:10:17.794 And yeah, it's disputed. Some people 00:10:17.794 --> 00:10:20.734 feel that the dorsal cortex which gives rise 00:10:20.734 --> 00:10:23.294 to excitatory cells may also give rise to some of the 00:10:23.294 --> 00:10:26.414 interneurons right so that's being that is being argued 00:10:26.414 --> 00:10:29.734 so when you're identifying these 00:10:29.734 --> 00:10:33.174 neurons and tracking them you're doing that primarily 00:10:33.174 --> 00:10:36.054 based on their response to i mean some 00:10:36.054 --> 00:10:39.274 markers inside the cell that you can use to 00:10:39.274 --> 00:10:42.034 stain them primarily but then you're also able to 00:10:42.034 --> 00:10:45.114 isolate some of these cells later and record from them right and 00:10:45.114 --> 00:10:48.014 that that that's the proof that that really 00:10:48.014 --> 00:10:50.814 is a neuron yeah if they were mature enough then you 00:10:50.814 --> 00:10:53.794 you know or i would be able to distinguish an 00:10:53.794 --> 00:10:57.734 interneuron from uh from an excitatory cell they have different firing patterns 00:10:57.734 --> 00:11:02.174 right when you stimulate them but at the stages i'm looking at they have a very 00:11:02.174 --> 00:11:05.414 immature action potential it's just a little blip it's not a full-blown action 00:11:05.414 --> 00:11:11.114 potential um so it's really actually quite difficult to distinguish a neuron 00:11:11.114 --> 00:11:13.574 from some of the other cell types that are floating around, 00:11:14.094 --> 00:11:17.914 Yeah, you can, you know, with the approach I used, you could just say neuron 00:11:17.914 --> 00:11:20.554 or not neuron. The glial cells have a very different response. 00:11:22.294 --> 00:11:26.134 But that's something that I always believe. You see those inward voltage-gated 00:11:26.134 --> 00:11:30.994 sodium channels, and that's believable, whereas if you stop and you fix the 00:11:30.994 --> 00:11:33.314 tissue and you try to do some immunostaining, 00:11:33.514 --> 00:11:36.074 first you got to hope that your antibodies are going to penetrate and reach 00:11:36.074 --> 00:11:39.814 the cell that you're looking for, and that's not always the case. 00:11:39.814 --> 00:11:44.794 And then you may have people who might dispute, well, that marker under certain 00:11:44.794 --> 00:11:46.894 conditions can be expressed by different cells. 00:11:48.394 --> 00:11:54.154 But now if you look at the taxonomy of precursor cells, so that would mean over 00:11:54.154 --> 00:11:56.374 time, they also will differentiate initially. 00:11:56.994 --> 00:12:01.934 I assume it would start with just a very few precursor cells initially for neurons. 00:12:02.214 --> 00:12:03.634 And these would also differentiate. 00:12:03.954 --> 00:12:09.514 So what do we know about that process? So, when do you see the first. 00:12:11.474 --> 00:12:17.154 The initial precursor cell emerge, and how is that precursor cell then differentiating 00:12:17.154 --> 00:12:21.794 other precursor cells that then form the plates which really start to generate our neurons? Mm-hmm. 00:12:22.354 --> 00:12:27.954 Well, the terminology that's used in my field is that before cortical neurons 00:12:27.954 --> 00:12:31.154 are being generated, people call them neuropathelial cells. 00:12:31.414 --> 00:12:35.914 And then once neurons are being generated, the term switches to radial glia. 00:12:36.794 --> 00:12:40.414 But there really aren't any good methods for distinguishing one from the other, 00:12:40.454 --> 00:12:43.994 other than the fact that you have neurogenesis going on. Yeah. 00:12:45.645 --> 00:12:50.365 So if you would compare the precursor cells for the inhibitory neurons that 00:12:50.365 --> 00:12:55.185 are located somewhere else, versus those excitatory neurons on their own. 00:12:55.185 --> 00:12:58.565 Yeah, those are different though. Okay, so already with two classes at least. 00:12:58.705 --> 00:13:02.745 Yeah, yeah. Once neurogenesis starts, there are markers that exist today that you can put on. 00:13:02.925 --> 00:13:08.505 So for example, the TBR2 marker that I used and showed in several of the figures, 00:13:08.685 --> 00:13:12.705 that is not expressed in the ganglionic eminence where the interneurons are being generated. 00:13:12.705 --> 00:13:18.145 And there are some markers that are more specific for interneurons. 00:13:18.625 --> 00:13:22.745 But so once the developmental program really starts and I will have to force 00:13:22.745 --> 00:13:27.185 you, I force you to give me a number, how many different types of precursor 00:13:27.185 --> 00:13:30.445 cells would we have? Yeah, it's a great question. 00:13:31.805 --> 00:13:35.845 You give me a number, you won't leave the room. Okay. 00:13:38.765 --> 00:13:43.225 That's a great question. And so for all the different cell types, 00:13:43.465 --> 00:13:47.925 I would speculate, and I'm just pulling a number out of the air, 00:13:48.065 --> 00:13:50.925 but I would put it at less than 10. 00:13:52.465 --> 00:13:59.425 And I could be entirely wrong, but easily five or six. 00:13:59.685 --> 00:14:03.365 Do precursor cells have sort of stem cell-like properties that, 00:14:03.365 --> 00:14:05.985 you know, there's some sort of equipotentiality there? 00:14:07.145 --> 00:14:10.685 Yeah, depending on the time. Time. So as development proceeds, 00:14:11.025 --> 00:14:14.305 they become more and more restricted in what they can produce. Right. 00:14:14.465 --> 00:14:19.185 So the earlier you get them, they'll have a wider potential for producing different cell types. 00:14:19.325 --> 00:14:24.945 So perhaps some of the precursor cells can produce oligodendrocytes and some types of neurons. 00:14:25.205 --> 00:14:27.805 But the further along you get, the more restricted they become. 00:14:28.045 --> 00:14:30.305 So your interest in the precursor cells 00:14:30.305 --> 00:14:36.325 is potentially that they could be reprogrammed to do different things. 00:14:36.325 --> 00:14:42.945 So a long way down the line maybe from here, but you started your talk setting 00:14:42.945 --> 00:14:47.885 up all these problems with brain diseases which involve development. 00:14:48.325 --> 00:14:54.545 And we could think about reprogramming these cells perhaps, how some cells might 00:14:54.545 --> 00:14:58.565 be behaving as they shouldn't, and thinking about the programming that's caused that. 00:14:58.565 --> 00:15:01.705 It's a challenging problem, because once you have a fully mature brain, 00:15:01.905 --> 00:15:08.465 you know, bringing precursor cells into that equation, it's challenging. Right, yeah. 00:15:08.865 --> 00:15:14.525 So now we are generating, so okay, let's say we have five, six different kinds 00:15:14.525 --> 00:15:19.465 of precursor cells, and we start to generate our neurons, we start to build 00:15:19.465 --> 00:15:21.305 this layered cortex. Mm-hmm. 00:15:22.393 --> 00:15:25.093 What you started out by showing was something actually extremely weird, 00:15:25.233 --> 00:15:30.573 which is that these neurons start to migrate, or the future neurons start to 00:15:30.573 --> 00:15:34.093 migrate, and then they bounce about a little bit between the ventricular zone 00:15:34.093 --> 00:15:35.333 and the subventricular zone. 00:15:35.493 --> 00:15:40.053 And in this bouncing around, sort of differentiation might happen, 00:15:40.093 --> 00:15:45.873 or they might divide, and two cells, you get a sister cell going off, and so on. 00:15:46.593 --> 00:15:49.973 So what's your interpretation of this whole process of just bouncing around? 00:15:50.133 --> 00:15:55.093 How long does this happen? How much movement do we have between ventricular, subventricular? 00:15:55.433 --> 00:15:58.313 What's the speed of this movement? How coordinated is it? 00:15:58.653 --> 00:16:03.433 So the speed of the movement has been fairly well characterized. 00:16:03.713 --> 00:16:08.593 And it's most rapid during G2 phase. So when they're dropping down to the ventricle, 00:16:08.613 --> 00:16:13.913 they can travel 70 to 100 microns in about two hours. 00:16:13.913 --> 00:16:17.173 So that's the most rapid speed of transit and then they'll undergo division 00:16:17.173 --> 00:16:21.593 at the surface of the ventricle and then in G1 they start moving away traversing 00:16:21.593 --> 00:16:28.113 that same 70 to 100 microns roughly but that's very slow that can take 8, 10, 12 hours. 00:16:31.013 --> 00:16:37.253 And the main idea right now is that it's a passive movement that those cells 00:16:37.253 --> 00:16:42.113 are being forced away from the ventricle as others come down via an active process, 00:16:42.673 --> 00:16:45.893 But you would expect that it might then stabilize at some point, 00:16:45.933 --> 00:16:47.353 not if it's a passive process. 00:16:47.873 --> 00:16:52.013 So why does it not stop? Yeah, I think we need to understand the process better. 00:16:52.353 --> 00:16:57.653 But so far, molecular motors have not been found that if you knock them out, they don't move up. 00:16:57.753 --> 00:17:02.093 So that under whatever conditions the cell bodies continue moving away. 00:17:02.253 --> 00:17:05.513 They can stop movement down, but they haven't been able to stop it going up. 00:17:05.593 --> 00:17:08.073 So that's the evidence pointing towards it being passive. 00:17:08.553 --> 00:17:11.313 But the ones that are moving down and moving down quickly are sort 00:17:11.313 --> 00:17:16.313 of actively tunneling their way to the bottom and they're changing their cell 00:17:16.313 --> 00:17:22.173 properties as they migrate is that right they um they're entering prophase so 00:17:22.173 --> 00:17:26.813 they they have a cell process which is tethered to the ventricular surface and 00:17:26.813 --> 00:17:29.173 that nucleus is being pulled down within that process. 00:17:29.833 --> 00:17:35.813 Right and they do a number of interesting things so there was this uh this the 00:17:35.813 --> 00:17:40.233 first piece of work pointing it out from 1935, the gentleman, 00:17:40.433 --> 00:17:42.173 Frederick Sauer, who discovered that, 00:17:42.813 --> 00:17:46.433 pointed out that it wasn't a smooth even movement, but that they would move 00:17:46.433 --> 00:17:49.033 down towards the ventricle, and then there'd be a little hitch up, 00:17:49.093 --> 00:17:50.853 they'd go up a little bit, and then come back down. 00:17:51.093 --> 00:17:54.693 And I noticed that in my time-lapse movies, that that's what they do. 00:17:54.853 --> 00:17:57.573 They start coming down, then they seem to stop, they bounce back up, 00:17:57.673 --> 00:17:58.633 and then they come down and divide. 00:17:58.893 --> 00:18:01.793 So they're always tethered, are they? It's kind of like a yo-yo. 00:18:02.993 --> 00:18:08.833 They're constantly tethered, yeah. But then there must be also a form of repellent. 00:18:09.593 --> 00:18:14.133 The cells that move towards subventricular zone must be repelling those traveling 00:18:14.133 --> 00:18:15.413 towards the ventricular zone. 00:18:16.524 --> 00:18:20.344 Otherwise, you would believe the system would never stay in this dynamical state. 00:18:20.524 --> 00:18:24.444 Do you have any kind of evidence for this of interaction between these cells? 00:18:24.864 --> 00:18:26.484 No, but there's got to be something there. 00:18:26.944 --> 00:18:29.784 Okay. Yeah, there's definitely got to be something there. 00:18:30.364 --> 00:18:34.904 But then if you look at the migration, so if I'm… How do they establish, 00:18:35.104 --> 00:18:38.504 you know, for S-phase and G1, when they're climbing up, they stop at a certain 00:18:38.504 --> 00:18:41.224 point, and it's almost always the same point. 00:18:41.264 --> 00:18:45.284 So they know where that is, and how they know that is an interesting question. 00:18:45.284 --> 00:18:50.624 But do you see this in sort of standard morphogenesis terms, 00:18:50.904 --> 00:18:54.844 like there is some sort of gradient maybe of RNAs or some other signaling molecule? 00:18:55.384 --> 00:18:56.504 Not that I'm aware of, yes. 00:18:57.384 --> 00:19:00.424 But that would be roughly the way to think about it. 00:19:01.064 --> 00:19:05.064 Okay. But now, if I'm one of these cells, so here I am, I'm at the ventricular 00:19:05.064 --> 00:19:06.084 zone, I'm going to move up. 00:19:06.584 --> 00:19:09.264 How many trips do I make up and down? 00:19:10.324 --> 00:19:14.184 So that would tell you how many cell cycles they've gone through. 00:19:15.204 --> 00:19:19.104 In the time-lapse movies that I did, I've seen them go through two or three times. 00:19:19.704 --> 00:19:23.044 On the average, it's two or three. In a time-lapse movie, right. 00:19:23.144 --> 00:19:30.964 But in vivo, it could be six, seven, perhaps more. 00:19:31.624 --> 00:19:33.744 That's what I was saying earlier. We don't know the life cycle. 00:19:33.904 --> 00:19:37.544 So as long as that cell remains, that primary precursor cell remains in the 00:19:37.544 --> 00:19:40.704 mitotic cell cycle, it's going to continue that bobbing up and down. 00:19:40.704 --> 00:19:48.024 And we don't know if each precursor cell will maintain that movement throughout 00:19:48.024 --> 00:19:53.584 the generation of all the cortical gray matter, or if you have overlapping subsets. 00:19:53.604 --> 00:19:58.084 So there are people who believe that you have one subset that generates the 00:19:58.084 --> 00:20:04.024 upper layers and a different subset that generates lower layers, but that's disputed. 00:20:04.084 --> 00:20:07.784 So it's a problem that the field is still working out. 00:20:08.444 --> 00:20:09.544 But now, in some sense, we often 00:20:09.544 --> 00:20:12.884 think about this also in the cartoon you presented about this migration. 00:20:13.464 --> 00:20:17.404 We think about it in terms of a bunch of ping pong balls that are sort of a 00:20:17.404 --> 00:20:19.484 big aquarium as it is moving about. 00:20:20.364 --> 00:20:24.604 But actually, what's going on really, it's not that these balls, 00:20:24.704 --> 00:20:27.264 the somas or whatever you want to call them, are moving about. 00:20:27.404 --> 00:20:30.784 It's much more that the processes are sort of feeling their way around. 00:20:30.784 --> 00:20:35.724 They attach at different points and then they start to exert a mechanical force 00:20:35.724 --> 00:20:40.244 to pull that soma in one direction or the other. So, so. 00:20:41.388 --> 00:20:46.368 If we now start to rethink this whole process in terms of neural processes or 00:20:46.368 --> 00:20:53.028 cell processes sort of feeling around in that space and making different points of adhesion. 00:20:53.828 --> 00:20:59.688 So how do you imagine that? There's like a whole, a huge amount of spider webs 00:20:59.688 --> 00:21:01.448 that sort of are being developed in parallel. 00:21:01.928 --> 00:21:05.688 So how do I have to imagine that? Why do you think it don't get entangled, 00:21:05.768 --> 00:21:08.828 for instance? Well, they start from the very beginning. 00:21:09.008 --> 00:21:13.848 The precursor cells have attachments at the ventricular surface and at the peel 00:21:13.848 --> 00:21:16.208 surface or the outer dorsal surface of the brain. 00:21:16.408 --> 00:21:21.708 And in the beginning, you have a large number of symmetric divisions that expands 00:21:21.708 --> 00:21:23.168 that precursor cell pool. 00:21:23.168 --> 00:21:26.288 The existing cells maintain their contacts and 00:21:26.288 --> 00:21:29.068 the their new daughter cells will establish new ones through 00:21:29.068 --> 00:21:32.108 a mechanism that we don't understand fully but the 00:21:32.108 --> 00:21:35.048 processes seem to grow in a radial direction and perhaps they use 00:21:35.048 --> 00:21:39.348 neighboring processes as guides so they'll establish contacts and then they 00:21:39.348 --> 00:21:42.968 maintain those contacts it seems once they have them they they maintain them 00:21:42.968 --> 00:21:52.228 for the most part um then the there's signals external signals which i i believe are involved in, 00:21:52.748 --> 00:21:55.408 attracting the cells to the ventricle and away from the ventricle. 00:21:55.488 --> 00:22:00.908 And one piece of evidence that supports that is once the precursor cells begin 00:22:00.908 --> 00:22:02.808 generating neuronal daughter cells. 00:22:03.848 --> 00:22:08.748 If you're watching a whole clone of cells, what you'll see is when the mother 00:22:08.748 --> 00:22:11.748 cell, when that nucleus starts moving to the ventricle, the daughter cells are 00:22:11.748 --> 00:22:13.868 also making some downward movements too. 00:22:13.988 --> 00:22:18.108 So they're perhaps responding to that same signaling factor. 00:22:18.388 --> 00:22:21.228 So it's kind of interesting if you look 00:22:21.228 --> 00:22:24.108 at an individual cell by itself you don't realize that it's 00:22:24.108 --> 00:22:26.868 moving in concert with something else but but for you that's 00:22:26.868 --> 00:22:30.268 a signaling system and not mechanical so it's not just a 00:22:30.268 --> 00:22:33.268 combination of the two yeah combination yeah okay because you 00:22:33.268 --> 00:22:37.388 need the microtubules of course you're fortunate to ask you what kind of receptors 00:22:37.388 --> 00:22:42.388 these cells would have and what they would be responsive to yeah they um they 00:22:42.388 --> 00:22:47.088 start expressing some of the standard neurotransmitter receptors in the rat 00:22:47.088 --> 00:22:52.528 for example at e16 which is not too long after neurogenesis has become so um, 00:22:53.028 --> 00:22:56.448 but that would not be on time to explain the basic movement no no, 00:22:58.228 --> 00:23:02.688 well um that i don't think that's been worked out so um, 00:23:06.451 --> 00:23:10.611 Yeah, starting around 15 or 16, you'll have some glutamate receptor expression. 00:23:12.571 --> 00:23:16.651 What's dictating that earlier, yeah. But now in your time-lapse movies, 00:23:17.091 --> 00:23:21.631 what I found really astonishing is that you had this very rapid emergence of 00:23:21.631 --> 00:23:25.911 different kind of processes that were labeled by these specific cells, 00:23:26.071 --> 00:23:28.851 but then also deformations of these cells themselves. 00:23:29.851 --> 00:23:32.851 So do you see this being meaningful? like 00:23:32.851 --> 00:23:35.571 for you also you saw one is really really squeezed down on 00:23:35.571 --> 00:23:38.571 the ventricular zone it was deforming itself in some way there 00:23:38.571 --> 00:23:41.611 was some process sticking out laterally yeah which seems completely orthogonal 00:23:41.611 --> 00:23:44.631 to the direction which you want to move so how do you interpret these 00:23:44.631 --> 00:23:47.851 kinds of variations um well when 00:23:47.851 --> 00:23:51.011 i first saw that i thought perhaps this was a culture artifact but 00:23:51.011 --> 00:23:54.871 and um just taking fixed 00:23:54.871 --> 00:23:57.771 sections of animals that were you know 00:23:57.771 --> 00:24:01.451 you're just looking at fixed time points we see some of the same things um 00:24:01.451 --> 00:24:04.251 but we don't have we don't 00:24:04.251 --> 00:24:08.071 really have an idea for what's guiding that right so 00:24:08.071 --> 00:24:11.071 but now from the subventricular zone we 00:24:11.071 --> 00:24:15.771 have to migrate further up to actually build the cortex yep right so is that 00:24:15.771 --> 00:24:20.931 process under the same control as the one we had just been discussing no i probably 00:24:20.931 --> 00:24:26.131 not and so that uh you know what's guiding the migration of newborn cortical 00:24:26.131 --> 00:24:28.231 neurons, that's fairly well studied. 00:24:28.431 --> 00:24:34.231 And yeah, there's a lot of good ideas for what's regulating that, 00:24:34.311 --> 00:24:36.851 but I would assume that that's a different signaling system. 00:24:37.231 --> 00:24:40.491 But it's also an inside-out construction system. Yeah. 00:24:40.651 --> 00:24:45.291 While the ventricular-subventricular one is maybe not an inside-out construction 00:24:45.291 --> 00:24:49.131 system because you keep on cycling between the bottom and the top. 00:24:49.331 --> 00:24:53.991 Yeah. Whereas the inside-out, the newborn cells Cells are always migrating to 00:24:53.991 --> 00:24:56.431 a position just under a structure called the marginal zone. 00:24:56.611 --> 00:25:00.191 And there are a number of cells that are expressing important proteins, 00:25:00.331 --> 00:25:06.131 including rilin, that are thought to play an instrumental role in the inside-out lamination. 00:25:06.151 --> 00:25:10.171 Because in animals that lack rilin, you end up with a cortex that's almost inverted. 00:25:11.011 --> 00:25:14.311 So rather than inside-out, it's outside-in. Right. 00:25:14.511 --> 00:25:19.391 I was interested that you described this process where the cells actually head 00:25:19.391 --> 00:25:22.591 towards the ventricular zone and then turn back. Initially, yeah. 00:25:22.751 --> 00:25:27.271 So, I mean, do we have an explanation of what's going on there? Is it orienting itself? 00:25:27.831 --> 00:25:33.171 So what I believe is that first process that emits that's being descended down 00:25:33.171 --> 00:25:38.831 is a transient or a vestigial axon. 00:25:39.291 --> 00:25:45.771 So it emits that process and the nucleus of the cell starts moving in that direction. 00:25:46.311 --> 00:25:50.851 And then that process actually stays there. the cell then develops a new leading 00:25:50.851 --> 00:25:54.531 process oriented towards the dorsal surface of the brain and it migrates away 00:25:54.531 --> 00:25:59.311 and that initial process seems to stay there and it stays there for several days. 00:26:01.716 --> 00:26:05.776 The trailing process ends up becoming the axon, but the mature axonal processes, 00:26:05.936 --> 00:26:09.736 it'll be sending collaterals in the white matter that will be going tangential to that. 00:26:09.836 --> 00:26:15.416 And then by the time the animal's maturing, this vestigial process in the ventricular zone disappears. 00:26:15.836 --> 00:26:19.636 But I believe that that's a signaling mechanism, that's a feedback signal. 00:26:19.636 --> 00:26:23.616 There's some kind of anchor that it leaves and then it heads upwards. 00:26:23.936 --> 00:26:28.736 As it's heading upwards. And I believe that it's actually a feedback conduit 00:26:28.736 --> 00:26:34.316 so but now in part of this process also what you what you illustrated is that 00:26:34.316 --> 00:26:39.456 in order to migrate now out of of this whole ventricular subventricular zone, 00:26:40.356 --> 00:26:45.616 um we have cell division right because we have to build more more cells but 00:26:45.616 --> 00:26:50.596 the the orientation of this division again seems to be very systematic like 00:26:50.596 --> 00:26:55.856 at the ventricular zone along a vertical plane and the subventricular zone along a horizontal plane. 00:26:56.356 --> 00:26:58.956 Does that make any sense to you? Why would it have a difference there? 00:26:59.476 --> 00:27:04.016 So the horizontal orientation for the cells in the subventricular zone, 00:27:04.096 --> 00:27:07.496 I believe that that's dictated by the radial glial fibers. 00:27:07.696 --> 00:27:12.536 I think the dividing cells are affiliated with a radial glial fiber and they 00:27:12.536 --> 00:27:15.216 anchor on that and are kind of pulling themselves apart along that. 00:27:15.356 --> 00:27:18.116 And the evidence for that is that in some regions of the brain, 00:27:18.256 --> 00:27:23.456 the radial glial cells are not oriented radially, but have more of an S-shaped curve. 00:27:23.636 --> 00:27:26.876 So the division, wherever it occurs along that S-shaped curve, 00:27:26.976 --> 00:27:29.416 it retains that orientation along the fiber. 00:27:31.016 --> 00:27:37.056 And then a further piece of evidence supporting that idea is that when the radial 00:27:37.056 --> 00:27:40.016 glial cells have translocated away from the ventricle and they're all gone, 00:27:40.136 --> 00:27:44.356 then the orientation of division in the submatricular zone becomes completely random. 00:27:45.036 --> 00:27:48.556 So while there are radial glial fibers, they seem to have a preference. 00:27:48.896 --> 00:27:53.456 And in our experiments, it was perhaps as many as 75% of the divisions would 00:27:53.456 --> 00:27:54.456 be horizontally oriented. 00:27:54.936 --> 00:27:59.376 And then as that percentage goes down, and by the time the radial glial cells 00:27:59.376 --> 00:28:02.356 are gone, it's a complete random process. 00:28:02.596 --> 00:28:06.296 It can also be an expression of just a mechanical bias that you want to have in the system. 00:28:06.416 --> 00:28:09.716 Like if I'm at a ventricular zone and I'm dividing, I 00:28:09.716 --> 00:28:12.896 have to do the packing and covering the space laterally so 00:28:12.896 --> 00:28:15.796 if i if i divide across a vertical plane 00:28:15.796 --> 00:28:18.516 at least i create a mechanical force that helps me to 00:28:18.516 --> 00:28:21.676 optimize the packing of of my cells along this 00:28:21.676 --> 00:28:24.776 lateral plane because i'm expanding radially so 00:28:24.776 --> 00:28:27.516 if i have a small deviation in my lateral plane i'm at a whole 00:28:27.516 --> 00:28:31.736 gap in my cortex that's pretty significant so sort of to optimize the packing 00:28:31.736 --> 00:28:36.236 while in the subventricular zone i want to make sure my cells migrate outward 00:28:36.236 --> 00:28:41.016 so now i i initiate the division in a way that already implies a mechanical 00:28:41.016 --> 00:28:45.716 force that is biasing my movement in the direction where I want to have myself to go. 00:28:45.876 --> 00:28:48.736 Would that be a reasonable interpretation or speculation, let's say? 00:28:48.896 --> 00:28:50.856 Yeah, it's one of them. So this has been a. 00:28:51.994 --> 00:28:55.554 A question that's been around for a long time. I don't know who first raised the question, 00:28:56.274 --> 00:29:04.594 but the most common interpretation has been that divisions along what I would 00:29:04.594 --> 00:29:08.754 call a vertical plane at the surface of the ventricle, where both daughter cells 00:29:08.754 --> 00:29:10.974 remain at the ventricle, 00:29:11.614 --> 00:29:14.014 that would be a vertical plane where they're sitting side by side, 00:29:14.074 --> 00:29:18.714 that was initially thought thought to be a symmetric division that doesn't produce neurons. 00:29:19.594 --> 00:29:24.214 And occasionally you have a division that's perpendicular to that, 00:29:24.294 --> 00:29:27.834 in which one daughter cell remains at the ventricle and the other one is sitting on top of it. 00:29:28.054 --> 00:29:32.554 And those were initially thought to be neurogenic because the idea was, 00:29:32.614 --> 00:29:36.174 well, that top one is free to leave and migrate towards the cortex. 00:29:36.734 --> 00:29:41.974 But I don't know who first proposed that idea, but one of my favorite researchers, 00:29:42.174 --> 00:29:46.694 this Frederick Sauer guy, He mentioned it in passing in his article saying, 00:29:46.854 --> 00:29:51.334 this is from 1935, saying, well, if this was the case, if this horizontal orientation 00:29:51.334 --> 00:29:53.174 actually produced neurons, 00:29:53.374 --> 00:29:56.714 there's just not enough of them to produce the millions and billions of cells 00:29:56.714 --> 00:29:59.054 because it's an infrequent thing. 00:29:59.274 --> 00:30:03.494 So we looked at the incidence of these divisions in rat and a number of other 00:30:03.494 --> 00:30:07.974 species and it's, you know, maybe 5% or so. It's a very small percentage. 00:30:08.854 --> 00:30:13.754 Um, my belief is, so I actually did a study where I tried to correlate that with outcomes. 00:30:13.854 --> 00:30:17.654 That's why in the movies you could see I was measuring the angle of the plane 00:30:17.654 --> 00:30:22.574 and, and averaging it across a number of different cells. And it didn't seem to correlate at all. 00:30:22.914 --> 00:30:28.214 What correlated with cell fate was the time of development at a certain time 00:30:28.214 --> 00:30:33.094 in development, regardless of orientation, you would have a specific outcome. Um. 00:30:35.286 --> 00:30:40.366 But still, these horizontal divisions at the ventricular zone are very infrequent. 00:30:40.566 --> 00:30:43.046 Yeah. Right? It would not be the standard pattern. 00:30:43.266 --> 00:30:48.926 Right. So, the other thing we've seen now is that as I'm migrating out, 00:30:49.006 --> 00:30:49.826 so now I'm going to climb. 00:30:50.186 --> 00:30:53.486 Your comparison for the human case was seven Empire State Buildings, 00:30:53.506 --> 00:30:56.766 I think, no? Four, but… Oh, four, sorry. Still, that's a lot. 00:30:57.166 --> 00:31:01.306 Right. It's a tall thing. Okay. So, the ventricular zone and the subventricular 00:31:01.306 --> 00:31:02.726 zone will get me to which floor? 00:31:03.686 --> 00:31:07.946 Oh, that's a good point. they would still be within the first building, 00:31:08.046 --> 00:31:10.586 maybe halfway up the first building. Okay, halfway up the, okay. 00:31:10.846 --> 00:31:14.466 So, okay, here we are. And now I'm hitting the subventricular zone. 00:31:14.526 --> 00:31:16.106 I'm going to be pushed out. I'm going in neuron. 00:31:16.366 --> 00:31:22.546 And then you showed actually in substance gradually as I move along this distance, 00:31:23.326 --> 00:31:29.166 I slowly start to express neural-like property, like sodium-dependent responses. 00:31:29.926 --> 00:31:33.586 So how should we think about this? Yeah, so it's a matriarchal gradient. 00:31:33.586 --> 00:31:38.686 So the earliest that I recorded a newborn neuron is maybe eight hours after 00:31:38.686 --> 00:31:42.226 it was generated, and there were already detectable sodium currents in those 00:31:42.226 --> 00:31:46.786 cells within eight hours after being born. 00:31:48.066 --> 00:31:53.306 As they migrate further along, those responses become stronger and stronger. 00:31:54.606 --> 00:31:59.206 Still, in many of the experiments that I did, within four days, 00:31:59.246 --> 00:32:00.546 it's not a mature neuron. 00:32:00.646 --> 00:32:04.986 It it doesn't have a full blown action potential and it can't fire repeatedly yet. 00:32:05.186 --> 00:32:09.426 So they still have a number of days until they're much more mature. 00:32:09.906 --> 00:32:13.246 So in some sense, if I know what day we are. 00:32:15.906 --> 00:32:21.346 Of the animal, what day of development, and also how far I am from a ventricular 00:32:21.346 --> 00:32:25.506 zone, you can sort of predict what kind of physiological properties that cell would have. 00:32:25.666 --> 00:32:29.046 Yeah. So that's a pretty deterministic system then. Yeah. Okay. 00:32:30.209 --> 00:32:36.889 So, now we have a bit of an idea how we sort of generate, right? 00:32:36.929 --> 00:32:41.269 How we generate these many billions of cells that might form a cortex ultimately. 00:32:42.209 --> 00:32:48.649 But what was really astonishing was your atom. 00:32:48.749 --> 00:32:53.329 Because you said, look, to generate is one thing. We just have to sort of constrain. 00:32:53.509 --> 00:32:55.749 You have to put boundaries on that as well. 00:32:56.049 --> 00:33:01.029 And you pointed out to a very specific class of glia cells that you think play 00:33:01.029 --> 00:33:04.569 an important role in also limiting growth. 00:33:05.089 --> 00:33:08.869 So what kind of cell type is that exactly? And how did you discover it? 00:33:09.269 --> 00:33:14.509 So the cell type you're talking about, they're called microglia and they're 00:33:14.509 --> 00:33:17.329 the immune component cell in the brain. 00:33:17.869 --> 00:33:24.389 And we discovered those in the laboratory because we were classifying all of the dividing cells. 00:33:24.669 --> 00:33:27.649 We wanted to know what were the different types of precursor cells. 00:33:28.469 --> 00:33:32.869 And these microglia are mitotic, and they represent about 5% of the dividing 00:33:32.869 --> 00:33:35.649 cells at the stages of development we were looking for. 00:33:36.209 --> 00:33:42.349 That's how we stumbled across them. And we're really excited to see that they 00:33:42.349 --> 00:33:49.229 populate and they specifically colonize the proliferative zones in early stages of fetal development. 00:33:49.629 --> 00:33:53.949 So they're very well studied and very well characterized in the adult brain. 00:33:54.049 --> 00:33:57.329 And in the adult brain, they have an even distribution throughout the brain, 00:33:57.449 --> 00:34:00.529 white matter and gray matter. They have this property of tiling. 00:34:01.449 --> 00:34:06.989 But in the fetal brain, they specifically colonize the proliferative zones. 00:34:07.149 --> 00:34:09.889 And that was very exciting to us to see that. 00:34:10.309 --> 00:34:14.769 So we've been studying interactions between the microglial cells and the precursor cells. 00:34:15.229 --> 00:34:17.909 And one of the observations that we've made is that, 00:34:18.767 --> 00:34:22.007 They seem to like precursor cells. In fact, they like to eat them. 00:34:23.527 --> 00:34:28.547 That's one of the mechanisms that we think is helping to put a brake on cell 00:34:28.547 --> 00:34:32.267 genesis because through chance or through evolution, 00:34:32.507 --> 00:34:36.027 these cells flood into the brain as cell genesis is going on, 00:34:36.127 --> 00:34:41.067 and they specifically populate the precursor cell zones and begin consuming 00:34:41.067 --> 00:34:43.627 them, and that gets rid of some of the precursor cells. 00:34:43.807 --> 00:34:47.967 So we were not the first to make that observation. 00:34:49.027 --> 00:34:53.727 After noticing that, I was able to dig up some papers by immunologists who had 00:34:53.727 --> 00:34:57.067 mapped out the distribution in fetal human brain, and they saw that there was 00:34:57.067 --> 00:34:58.567 this band in the precursor cell zone. 00:34:58.847 --> 00:35:02.667 And their idea was that, well, they must go there in order to divide and make more of themselves. 00:35:03.247 --> 00:35:06.827 And we followed that up by looking at where the microglia divide, 00:35:07.047 --> 00:35:09.187 and they divide everywhere. 00:35:09.527 --> 00:35:15.647 So it doesn't appear that they go to the proliferative zones just to divide. 00:35:15.867 --> 00:35:17.847 I think they go there to eat, to feed. 00:35:18.287 --> 00:35:22.047 Why are they microglia? I mean, are they like other glia cells? 00:35:22.747 --> 00:35:27.927 Yeah, so in the mature brain and in a healthy brain, the soma is very small. 00:35:28.247 --> 00:35:32.227 Right. And that's where that term comes from. So they have a very small soma 00:35:32.227 --> 00:35:34.127 with these fine ramified processes. 00:35:34.807 --> 00:35:39.887 Okay. So that's called a ramified or a resting cell. 00:35:40.047 --> 00:35:42.927 In the adult brain, if you had some pathological condition, they 00:35:42.927 --> 00:35:46.027 change their appearance they become what people 00:35:46.027 --> 00:35:49.267 might call activated the soma gets much bigger 00:35:49.267 --> 00:35:53.527 and the processes swell and you have fewer processes they have a completely 00:35:53.527 --> 00:35:58.287 different characteristic look and for whatever reasons in the fetal brain they 00:35:58.287 --> 00:36:01.847 are just super activated you don't see any of these so-called resting cells 00:36:01.847 --> 00:36:05.567 they're they're just all jacked up they're they're ready to go so the environment 00:36:05.567 --> 00:36:08.787 it's got a two-stage life cycle this these cells so they've. 00:36:09.549 --> 00:36:15.969 They have this active phase where they're running around eating bacteria and other cells. 00:36:16.269 --> 00:36:21.529 And then they sort of settle down, grow more processes, and become part of the furniture. 00:36:21.909 --> 00:36:28.789 That's what we think, but we don't have a good handle on the life cycle. 00:36:28.989 --> 00:36:33.569 The lifespan of an individual microglia. So the guys that are present in the 00:36:33.569 --> 00:36:37.129 fetal brain, are they still present in the adult? or do they continue dividing 00:36:37.129 --> 00:36:41.089 and producing descendants that will populate the adult brain? We don't know that yet. 00:36:41.389 --> 00:36:44.609 That's something that we would like to try to figure out, but we don't know yet. 00:36:44.789 --> 00:36:50.209 There's something else that's extremely strange, at least from my naive perspective of these neurons. 00:36:50.529 --> 00:36:53.989 They're like space invaders, right? They are not intrinsic to the developing 00:36:53.989 --> 00:36:58.729 brain, but they have invaded that brain from this embryonic sac. 00:37:00.509 --> 00:37:03.329 So what's going on here? It's fascinating. Fascinating. 00:37:04.309 --> 00:37:09.129 We need, you know, for our survival, we need a cell to perform this function. 00:37:09.449 --> 00:37:13.689 And evolutionarily, it's worked out that they're introduced at this time, 00:37:13.729 --> 00:37:15.949 and it doesn't impede development. 00:37:16.029 --> 00:37:19.229 And in some ways, it may be helping it. It's definitely shaping the process. 00:37:19.649 --> 00:37:24.849 But then, do they also infiltrate other organs, or they really are specific to the brain? 00:37:26.849 --> 00:37:30.709 Traditionally, the microglia are classified as those cells which enter the brain, 00:37:30.709 --> 00:37:37.589 And then they have related cell types, which would colonize peripheral or the body. 00:37:37.889 --> 00:37:47.029 But in some way, they're like white blood cells. They're sort of having this role of cleaning up. 00:37:47.329 --> 00:37:50.569 Yeah. So you have macrophages in the body and then the microglia. 00:37:51.249 --> 00:37:56.629 And in general, it's thought that they don't mix very much under normal conditions. 00:37:56.629 --> 00:38:02.109 But under pathological conditions, I think it's been shown that the macrophages can enter the brain. 00:38:03.309 --> 00:38:11.729 But why don't we call them just macrophages or brain-based or brain-specific macrophages? 00:38:11.829 --> 00:38:13.309 Wouldn't that be a more appropriate name for them? 00:38:15.257 --> 00:38:21.117 I guess it's the inherited terminology. So we're stuck with that. 00:38:21.477 --> 00:38:26.397 So these microglia come from the mother. So this might also be potential. From the yolk sac. 00:38:26.637 --> 00:38:33.017 From the yolk sac, right? So this might be also an epigenetic mechanism because 00:38:33.017 --> 00:38:39.037 this might also be a way for the mother during gestation to sort of modulate 00:38:39.037 --> 00:38:40.277 a developmental process. 00:38:40.597 --> 00:38:44.437 Is it also how you think about it? I don't think about it that way. 00:38:46.197 --> 00:38:51.397 I guess I think of the yolk sac as part of the fetal organ. 00:38:52.477 --> 00:38:58.677 But the question of whether maternal cells can enter the fetus is something that's interesting. 00:38:58.917 --> 00:39:03.477 So there have been studies showing bidirectional cell transfer between fetus and mother. 00:39:03.697 --> 00:39:07.557 But what proportion of cells that is, I think, remains to be determined. 00:39:08.197 --> 00:39:11.897 That is a question that we're looking at. at the, are there fetal, 00:39:11.897 --> 00:39:16.397 are there maternal immune cells or maternal cells of any sort in the fetuses? 00:39:16.817 --> 00:39:21.617 But now, how did you really discover the effect of this microglia? 00:39:21.697 --> 00:39:25.297 Because I could imagine that if you see for the first time a microglia engulf 00:39:25.297 --> 00:39:29.717 another cell, that you might think, well, okay, that's some error. 00:39:30.437 --> 00:39:34.237 This cannot really be the case. Well, it was, yeah, no, it wasn't difficult 00:39:34.237 --> 00:39:39.417 to come to it because the first image my student showed me was this dense band 00:39:39.417 --> 00:39:41.957 of microglia colonizing the proliferative zone. 00:39:42.057 --> 00:39:46.117 So right away, we just did some staining for the precursor cell markers and 00:39:46.117 --> 00:39:49.357 the microglia and immediately saw them. 00:39:49.477 --> 00:39:54.057 And lucky for us, we were first working this out in primate because it's much 00:39:54.057 --> 00:39:55.817 more prominent in primate than it is in rat. 00:39:56.157 --> 00:40:00.837 So in rat, we can find the same things happening, but it's going on at a far 00:40:00.837 --> 00:40:02.777 greater pace in the primate brain. 00:40:02.937 --> 00:40:05.137 So it was just happening all over the place. 00:40:06.786 --> 00:40:12.386 But now, is there some sort of ratio then between the propensity of neurons 00:40:12.386 --> 00:40:19.646 to divide and to generate large pools of neurons and the prevalence of these microglia? 00:40:19.706 --> 00:40:25.026 Is there some sort of magic balance between these two? That's a great question. 00:40:25.166 --> 00:40:29.846 We've noticed that there are different proportions in different species. 00:40:30.886 --> 00:40:36.726 There's a higher number of microglia in primates, potentially more in humans 00:40:36.726 --> 00:40:38.026 than there are in monkeys. 00:40:38.206 --> 00:40:43.846 There's more in monkeys than there are in rodents, more in mammals than there 00:40:43.846 --> 00:40:45.446 are in reptiles and birds. 00:40:45.746 --> 00:40:48.526 So there are differences across species. 00:40:49.646 --> 00:40:56.566 So there might potentially be a way you could program these microglia to serve useful functions. 00:40:56.766 --> 00:41:01.766 So for example, brain tumors, could you program them to break down cells in 00:41:01.766 --> 00:41:04.046 the brain that were cancerous? That would be exciting. 00:41:06.726 --> 00:41:12.206 But can you believe microglia can be made into something so specific? 00:41:12.926 --> 00:41:17.746 I mean, how specific are they in their sort of... Yeah, I don't think... 00:41:17.746 --> 00:41:20.266 In my system, I don't know. 00:41:20.366 --> 00:41:23.206 I don't have an answer for how specific. It seems like they'll eat just about 00:41:23.206 --> 00:41:27.466 anything because we find them eating young neurons, glial precursor cells, 00:41:27.646 --> 00:41:30.726 neuronal precursor cells. They might even be eating each other. 00:41:32.246 --> 00:41:35.546 Have you ever observed that? No, but I'm going to go look for it now. 00:41:36.566 --> 00:41:40.266 So what stops them sort of running away and eating the whole brain? 00:41:40.386 --> 00:41:44.046 So what are the control mechanisms for these? We don't know the control mechanisms, 00:41:44.126 --> 00:41:47.346 but we know they exist because it's not indiscriminate. 00:41:47.426 --> 00:41:50.746 So there are regions of the developing brain where it does not happen. 00:41:50.826 --> 00:41:55.206 So the pineal gland is one example I brought up in the talk today where you 00:41:55.206 --> 00:41:56.346 have the same cell types. 00:41:56.466 --> 00:42:01.446 You have these vimentin-expressing, PAX6-positive radioglial cell types in the 00:42:01.446 --> 00:42:03.826 developing pineal gland, and they're just... 00:42:05.102 --> 00:42:09.862 Out of bounds. They're just not to be touched. They aren't touched. 00:42:10.442 --> 00:42:14.362 Later on in the adult pineal gland, then this phenomena begins. 00:42:14.922 --> 00:42:20.542 So that's one region. And in the adult brain, if you look at the dentate gyrus 00:42:20.542 --> 00:42:23.022 where you have continuing neurogenesis, it doesn't occur there. 00:42:23.122 --> 00:42:27.422 And in fact, there's some evidence that they might be supporting cell genesis in a way. 00:42:28.242 --> 00:42:31.782 So I think the best approach would be to compare those regions, 00:42:31.842 --> 00:42:37.082 to compare pineal gland in a developing animal with the cortex where it's happening. 00:42:37.162 --> 00:42:41.242 And that may help us find cues that are guiding the process. 00:42:41.682 --> 00:42:45.462 What exists in the pineal gland that tells them, you know, hands off, leave us alone. Right. 00:42:45.782 --> 00:42:49.322 But then they have some sort of specificity because so far you haven't seen 00:42:49.322 --> 00:42:50.462 that friends eat each other. 00:42:51.342 --> 00:42:55.902 Right? Also, what you have not seen, they also limit themselves largely, 00:42:56.022 --> 00:43:00.762 not fully, but largely to the proliferation zone. Yeah, they're attracted to it. 00:43:00.782 --> 00:43:04.542 There's something that attracts them there. So there is some specificity there. 00:43:04.662 --> 00:43:06.902 And for all we know, it could be something as simple as ATP. 00:43:07.102 --> 00:43:11.042 We know they like ATP, and there's evidence that precursor cells might be putting 00:43:11.042 --> 00:43:15.142 out ATP in large quantities, and it might be as simple as that. 00:43:16.077 --> 00:43:20.357 But now, if it's, let's say, ATP-dependent, then you could also argue, 00:43:20.417 --> 00:43:22.957 well, maybe these guys are just out there to clean up the mess. 00:43:23.117 --> 00:43:27.697 And ATP is sort of a signal that something messy is going on because we're leaking 00:43:27.697 --> 00:43:30.917 an intercellular element. 00:43:32.197 --> 00:43:37.377 So how about we just interpret these microglia as sitting in this proliferation 00:43:37.377 --> 00:43:39.877 zone because this is where we do most of the cell division. vision. 00:43:40.097 --> 00:43:44.877 We want to assure that we have no cells coming out of that zone that are sort 00:43:44.877 --> 00:43:48.837 of anomalous because they can create havoc in the rest of the system. 00:43:49.077 --> 00:43:52.617 So I want to have a super conservative system that is as soon as there's even 00:43:52.617 --> 00:43:57.097 the tiniest chance that there's something wrong with any cell, I just destroy it. 00:43:57.777 --> 00:44:01.097 So would that be a reasonable way to think about this? Yeah. Another 00:44:01.097 --> 00:44:04.097 exciting idea to me 00:44:04.097 --> 00:44:06.937 is the potential that these cells could be shaping the 00:44:06.937 --> 00:44:10.217 proliferative zone so the proliferative zones have defined boundaries 00:44:10.217 --> 00:44:13.297 and this could be potentially a 00:44:13.297 --> 00:44:16.197 mechanism for shaping that for keeping the boundaries 00:44:16.197 --> 00:44:21.077 you know you stray too far you're gone okay and there's some evidence for that 00:44:21.077 --> 00:44:24.737 it's that's something that we're following up on you already have some ways 00:44:24.737 --> 00:44:31.237 of uh intervening in in the behavior of these cells uh and to show the effect 00:44:31.237 --> 00:44:34.357 of maybe reducing the population size. 00:44:36.817 --> 00:44:40.637 So one of the effects then is to increase the number of precursors. 00:44:40.817 --> 00:44:43.777 Yeah, when we transiently got rid of them, it would increase that. 00:44:43.937 --> 00:44:48.077 So what's the next step there in terms of the kinds of things you'd like to 00:44:48.077 --> 00:44:49.237 be able to do with the microglia? 00:44:49.997 --> 00:44:55.537 We're developing culture models to put precursor cells with microglia in chambers 00:44:55.537 --> 00:45:01.477 and be able to experimentally control the conditions to try to work out the 00:45:01.477 --> 00:45:05.757 signaling that attracts them to them. That's one thing that we're doing. 00:45:06.885 --> 00:45:11.965 We're also trying to repopulate, so you could delete the microglia from a slice 00:45:11.965 --> 00:45:16.705 and repopulating them with microglia from different aged animals to see if there's 00:45:16.705 --> 00:45:17.865 any different function. 00:45:18.825 --> 00:45:24.385 But to repopulate them, you could isolate them from a postnatal brain, 00:45:24.585 --> 00:45:27.965 and that process in and of itself changes them. 00:45:28.185 --> 00:45:31.985 There's caveats with everything, but those are the sorts of things that we'll 00:45:31.985 --> 00:45:38.625 be trying. But are you able to say already that microglia are a critical part 00:45:38.625 --> 00:45:41.465 of the neurogenesis process? 00:45:41.705 --> 00:45:45.905 If you don't have them, then you have a pathological brain. 00:45:46.145 --> 00:45:50.445 I believe so, but there's other groups that don't. So there's knockout models 00:45:50.445 --> 00:45:54.665 where presumably you have no microglia, and they claim that the brain is completely normal. 00:45:55.025 --> 00:46:00.225 But as we were talking about earlier, there are different types of immune cells, 00:46:00.445 --> 00:46:05.325 and so I think it needs to be looked at whether or not other cells are performing similar functions. 00:46:05.845 --> 00:46:12.885 Or it may be that there's many mechanisms and microglia is one of the mechanisms, 00:46:13.025 --> 00:46:16.985 but if they're not there, something else will serve the job. 00:46:17.345 --> 00:46:20.285 Yeah, I just have a hard time believing that if you remove microglia from the 00:46:20.285 --> 00:46:21.605 brain that it's going to develop normally. 00:46:23.625 --> 00:46:27.545 But now if we would turn this around and say, okay, we're going to give you 00:46:27.545 --> 00:46:30.585 all the tools you want to give us a seven layer cortex, right? 00:46:31.005 --> 00:46:35.525 Do you think you would have the mechanisms at hand to do that? 00:46:38.069 --> 00:46:44.649 Um, probably not. Or just, let's say, a cortex that is twice as thick. 00:46:45.809 --> 00:46:51.549 Twice as thick. We would need to extend the length of neurogenesis, so. Sure. 00:46:51.829 --> 00:46:56.829 Yeah, we need to figure out what's promoting it. They enter a phase where the 00:46:56.829 --> 00:46:59.049 primary precursor cells are in a steady state. 00:46:59.209 --> 00:47:04.689 They continue dividing, and each division produces a copy of themselves and 00:47:04.689 --> 00:47:08.629 a neuronal daughter cell. and they maintain that. 00:47:08.769 --> 00:47:13.789 So they appear in ferrets, they're in that period longer than they are in rats. 00:47:13.869 --> 00:47:17.409 And in primates, they're in that period for even longer than they are in ferrets. 00:47:18.069 --> 00:47:23.509 So we need to figure out what the signal is that promotes that and what ends it. 00:47:23.709 --> 00:47:26.589 Right. But when you're comparing primates, ferrets, and rats, 00:47:26.909 --> 00:47:31.869 are you seeing any big differences, for instance, in the primates in the way 00:47:31.869 --> 00:47:37.409 that this process is working, which will explain some of the- We're doing those experiments now. 00:47:37.489 --> 00:47:42.629 So we have 10 monkeys that we've saved up and we're hopefully going to be getting the answer. 00:47:42.709 --> 00:47:45.829 So I think my postdoc is cutting one of the brains today, I hope. 00:47:47.189 --> 00:47:50.649 What's the sort of question you're asking there then? What kind of things might 00:47:50.649 --> 00:47:51.929 be different in the primate? 00:47:52.349 --> 00:47:59.849 So the subventricular zone is a very important structure and neurogenesis is 00:47:59.849 --> 00:48:02.629 beginning before the subventricular zone is present. 00:48:02.789 --> 00:48:06.689 So what do the neural precursor cells look like before that zone? 00:48:06.849 --> 00:48:09.229 Are they subventricular zone 00:48:09.229 --> 00:48:12.309 cells that just happen not to have coalesced into a cohesive structure? 00:48:12.629 --> 00:48:16.449 Or is there something different going on at that stage of development when deep 00:48:16.449 --> 00:48:21.169 layer neurons are being generated versus later on when you have a substantial subventricular zone? 00:48:21.229 --> 00:48:24.089 Those are some of the simple questions we're asking, but we're also... 00:48:25.989 --> 00:48:30.429 Looking at how the microglia are interacting with the precursor cells in a more 00:48:30.429 --> 00:48:33.269 defined way in the primate model. 00:48:33.489 --> 00:48:38.889 Any thoughts you might go to Monodelphus, which I know you have in Davis, 00:48:38.969 --> 00:48:44.109 to look at the early origins of cortex or even to reptiles? 00:48:44.949 --> 00:48:49.429 So we've looked at turtle. We have crocodile that we're just starting to look 00:48:49.429 --> 00:48:52.389 at now. So they're present in turtle. all. 00:48:52.689 --> 00:48:57.429 I have some tissue from doves and from chickens and they're present there. 00:48:57.609 --> 00:49:03.889 So it's the population of precursor cells with microglia or microglial-like 00:49:03.889 --> 00:49:05.629 cells is something that's been around for a long time. 00:49:05.729 --> 00:49:08.509 So evolutionarily, it's something that occurred fairly early. 00:49:08.869 --> 00:49:15.149 But we've got very distinctive differences between the reptile cortex and the mammalian cortex. 00:49:15.569 --> 00:49:21.269 So are there clues there as to what's happening with the mammalian cortex or 00:49:21.269 --> 00:49:25.649 what has changed what what other things that stand out for you to me determined 00:49:25.649 --> 00:49:30.389 really okay so there's there's just lots of open questions yeah yeah it's exciting 00:49:30.389 --> 00:49:31.869 it's a really exciting time i think, 00:49:32.509 --> 00:49:36.509 but then to follow up on that and i could also challenge you from another angle 00:49:36.509 --> 00:49:41.629 and say well look if you look at it as a comparative lead and you would expect 00:49:41.629 --> 00:49:46.589 that the mechanisms to lay down a cortex are sort of piggybacking on the way 00:49:46.589 --> 00:49:49.749 i lay down earlier structures in the brain the way i I might develop, 00:49:49.889 --> 00:49:54.349 let's say, a brainstem or cerebellum or basal ganglia, right? Or spinal cord. 00:49:54.529 --> 00:50:00.429 Or spinal cord, right. So how would I have to configure the specific system 00:50:00.429 --> 00:50:03.289 you studied that gives rise to cortex to not give me spinal cord? 00:50:04.449 --> 00:50:08.649 It's a great question. That's probably at the genetic level, you know. There are... 00:50:15.240 --> 00:50:21.140 No, because in some sense, okay, the cell types might be somewhat different, right? 00:50:21.260 --> 00:50:25.740 So there might not be the similar kind of layering, but we also will have precursor 00:50:25.740 --> 00:50:30.880 cells that will be giving rise to these proton neurons that have to link up together. 00:50:31.080 --> 00:50:34.980 You might also have to wash out the detrius from that system in the same way. 00:50:35.060 --> 00:50:39.240 So right now, there's not a clear understanding of what these common principles 00:50:39.240 --> 00:50:41.320 might be across different structures, is that correct? 00:50:41.880 --> 00:50:46.220 There's definitely something different about spinal cord. then there is forebrain. 00:50:46.240 --> 00:50:52.080 So we have looked at this phenomena, the microglia, and they don't appear to 00:50:52.080 --> 00:50:53.640 be in the spinal cord at the same stages. 00:50:53.720 --> 00:50:56.000 So there are key differences. 00:50:56.740 --> 00:51:00.220 But now, do you already see some neuropathologies that you think are linked 00:51:00.220 --> 00:51:03.060 to pathological activity of this microglia? 00:51:04.060 --> 00:51:09.000 So one potential is, and this is something that many, many people are studying, 00:51:09.120 --> 00:51:17.200 is that that if a pregnant woman is exposed to a pathogen at a specific stage of pregnancy, 00:51:17.900 --> 00:51:21.480 that this can actually influence the function of the fetal immune cells. 00:51:22.080 --> 00:51:28.780 And one example that's fairly well known is that if a woman is exposed to influenza 00:51:28.780 --> 00:51:33.260 in the first trimester, then there's a greater likelihood that her child will be schizophrenic. 00:51:34.280 --> 00:51:38.700 And that's a potential example. Yeah, well, there are also links between autism 00:51:38.700 --> 00:51:41.080 and maternal immune responses. 00:51:41.780 --> 00:51:46.340 So the mother is exposed to some pathogen. Her body generates an immune response 00:51:46.340 --> 00:51:47.520 to fight off the pathogen. 00:51:47.960 --> 00:51:51.720 So she develops a full-blown response. Many cytokines are produced, 00:51:51.940 --> 00:51:55.220 which are helpful for her, but they actually get into the fetal compartments 00:51:55.220 --> 00:51:56.360 and they can get into the fetal brain. 00:51:56.580 --> 00:52:00.600 And that changes how these cells function because that's what they're built to do. 00:52:00.820 --> 00:52:04.360 They're built to respond to foreign pathogens, to challenges. 00:52:04.360 --> 00:52:06.700 And that can change the equation. 00:52:06.900 --> 00:52:11.260 So there's hints that in neurodevelopmental disorders, schizophrenia, 00:52:11.520 --> 00:52:15.860 perhaps autism, that this will play a role, that it can change the trajectory 00:52:15.860 --> 00:52:19.740 and influence the outcome for the worse. Right. 00:52:21.047 --> 00:52:29.847 So now, okay, so we made a lot of progress in some sense in our understanding of how a cortex is built. 00:52:29.907 --> 00:52:33.167 And also what we hear now, there's still many, many questions to be answered. 00:52:33.267 --> 00:52:36.127 And hopefully you will find those answers anytime soon. 00:52:36.227 --> 00:52:41.107 But in this whole trajectory that you're on, understanding the developing brain, 00:52:41.287 --> 00:52:45.427 what is Stephen's law that we should follow to understand the brain? 00:52:46.367 --> 00:52:52.187 What is Stephen's law that we should follow? I think we're still working that out. 00:52:52.647 --> 00:52:57.187 We're still, I like to say that what we do in the lab or my approach, 00:52:57.307 --> 00:53:02.607 my way of looking at things is that we're defining normal, understanding very 00:53:02.607 --> 00:53:04.887 well what's going on during normal development. 00:53:05.787 --> 00:53:11.007 Okay. And now Tony likes traveling, so he'll be in Davis soon. Good, yeah. 00:53:11.827 --> 00:53:16.047 It's also five years from now he'll be there. and he's going to come to your 00:53:16.047 --> 00:53:19.267 lab and visit you and he's going to put out this piece of paper that says, 00:53:19.407 --> 00:53:23.607 okay, Stephen, five years ago you made this prediction and you told me that 00:53:23.607 --> 00:53:25.527 now you're going to give me the answer. 00:53:25.767 --> 00:53:29.347 So what's the main prediction you would like to sort of commit yourself to today 00:53:29.347 --> 00:53:31.987 that Tony's going to get the answer to five years from now? 00:53:32.227 --> 00:53:34.287 The signaling that controls the 00:53:34.287 --> 00:53:38.587 colonization and the interaction between microglia and precursor cells. 00:53:38.847 --> 00:53:41.587 I would really hope that we could have that figured out before five years. 00:53:42.387 --> 00:53:45.807 All right, Stephen. That's perhaps ambitious, but that's what we're hoping for. 00:53:45.987 --> 00:53:48.307 Okay. Stephen Nocturne, thank you very much for this conversation. 00:53:48.547 --> 00:53:52.847 Thank you. Thank you. The CSN Podcast was produced by the Convergent Science 00:53:52.847 --> 00:53:59.087 Network of Biometrics and Biohybrid Systems, a project funded by the European 00:53:59.087 --> 00:54:01.347 Sevens Research Framework Program. 00:54:03.087 --> 00:54:08.167 For more interviews, recorded lectures, or upcoming conferences in the field 00:54:08.167 --> 00:54:14.407 of biometrics and biohybrid systems, go to csnnetwork.eu. 00:54:15.280 --> 00:54:23.120 Music.