Patterning Development in the Early Embryo: The Role of Bicoid
Transcript of Part 1: Patterning Development in the Embryo
00:00:05.18 I'm Eric Wieschaus and I'm a HHMI investigator and a 00:00:11.22 professor at Princeton University. My laboratory works on the way 00:00:15.24 embryos develop and I'm particularly interested in the gene activities 00:00:20.28 and the cell biologically processes that operate in the very early embryo 00:00:26.00 to transform what seems to us like a simple freshly fertilized egg into 00:00:31.25 a complex organism with cells with particular morphologies and 00:00:39.17 particular organs all in the right place at the right time. Now these processes 00:00:45.01 are interesting to all of us I think especially with respect to human development, 00:00:50.21 because we are all interested in where it is that we come from, how is it that 00:00:54.09 a single cell is able to give rise to something as complex as you or I. 00:00:59.12 But over the past 20 years we have also learned that all these process 00:01:06.12 that occur in embryonic development and all the gene activities 00:01:08.28 and all the cell shape changes and the controls of cell adhesion 00:01:13.04 and cytoskeleton cell structure that one sees happening and can follow 00:01:20.02 and study in embryos are the same kinds of processes that involve 00:01:23.10 the same kind of molecules that operate throughout the life of almost all organisms. 00:01:29.26 So one of the goals of our work is not just to understand how embryos develop, 00:01:35.14 but to understand how cells and living organisms control their gene activity, 00:01:41.01 how they control their cell shape, how they control all these processes 00:01:45.03 that are essential for life. Not only of the embryo and for embryonic 00:01:48.29 development but for everything that we see today. So what I'm going 00:01:53.03 to talk to you about in my lectures, is my basic plan, or what I'd like to do is to 00:01:57.08 first describe to you a little bit about how the fruit fly embryo, 00:02:01.23 which is the embryo that we work on in the lab develops, and point out to you 00:02:05.24 some of the really interesting features that have made this embryo 00:02:09.12 really helpful, a really powerful tool for understanding development. 00:02:14.12 And then in the second part of my talk, what I'll do is focus in on one particular question 00:02:23.17 which is how is it that in an embryo cells come to achieve different fates, 00:02:30.14 different patterns of gene activity. How do cells come to be different from 00:02:33.25 each other, where does that pattern arise? And as you'll see, we'll start 00:02:37.24 through a basic description of the process and then work our way through 00:02:43.02 gene activities and basic cellular functions but what we'd really like 00:02:49.13 to end up with is a deeper almost biophysical understanding of these 00:02:55.15 processes, and that's my goal. And then what I'd like to do is also expand back 00:02:59.05 out and talk about embryonic development, talk about it in the context 00:03:03.05 of not just fruit fly embryos but more broadly of all organisms 00:03:08.13 on this earth and how processes evolve and have changed during 00:03:12.09 the history of life on this earth. Okay, so the very first slide though and to begin 00:03:16.09 what I wanted to do was just to show you an image of a Drosophila embryo. 00:03:20.29 This is a scanning EM, it's a fruit fly egg, or a Drosophila egg 00:03:25.04 almost immediately after fertilization. It's still a single cell. It's about 00:03:30.08 maybe half a millimeter or 500 microns long. If you look at the surface you 00:03:36.10 don't see anything very much. If you were to look inside you'd see that 00:03:39.28 in this zone here there's the nucleus of the single cell, which is the product 00:03:45.19 of the fusion of the female pronucleus and the male pronucleus. If you look 00:03:51.05 at it and it doesn't really look very interesting. Amazing thing though 00:03:56.05 is what this embryo has to do to become interesting is to convert from being 00:04:01.21 a single cell into a multicellular organism where individual cells 00:04:08.01 can assume different fates and begin to do different things. 00:04:11.03 Now one of the interesting things about fly embryonic development in 00:04:15.16 Drosophila is that the embryo is able to do this extraordinarily rapidly. 00:04:20.09 Such that after two and a half hours this single cell has now transformed itself 00:04:25.16 into a multicellular organism with now 6,000 cells, about a hundred cells 00:04:31.26 along the anterior with the future head and future tail anterior/posterior axis 00:04:36.25 of the embryo. And the way that its able to do this transformation so rapidly 00:04:42.17 is that what it's done is that unlike most other organisms where when 00:04:48.00 cells divide and replicate you have a DNA replication and then a mitosis and 00:04:52.01 that's followed by cell division. In Drosophila during the early stages 00:04:56.23 of development these mitotic divisions, these replications of nuclei 00:05:02.17 occur without cell division such that an individual fertilized egg which starts 00:05:08.00 out with a fused of a single nucleus goes after one mitotic cycle goes from 00:05:12.17 one nuclei to two nuclei, the cycles are synchronous so the nuclei divide 00:05:16.10 again without cell division so that you have from two nuclei to four nuclei, eight nuclei. 00:05:22.05 And what happens after about an hour and a half to two hours is that 00:05:27.00 through a sequence of thirteen of these synchronous rounds of nuclear replication 00:05:32.20 the embryo is now still a single cell, but it's a single cell with 6,000 nuclei on 00:05:40.10 the surface. And amazingly then, at that point these mitotic divisions 00:05:45.01 temporarily stop and it's only at that point that the cytoskeleton 00:05:51.02 and the membrane synthesis is reorganized in this embryo to now make 00:05:56.02 new membrane such that membrane can be pulled down, plasma membrane 00:06:00.16 can be pulled down between individual nuclei to separate them 00:06:03.20 or partition them into individual cells. And it's after that process that's called 00:06:08.09 cellularization that the embryo has now converted itself from one cell 00:06:12.27 into an embryo with multiple cells, 6,000 cells. And it's only at that 00:06:19.17 point that those cells can begin, that when you have individual cells, that those cells 00:06:24.23 can begin to become different from each other and show distinct behaviors 00:06:28.14 that are ultimately related to their fates as skin or muscle. Now 00:06:33.02 there's one other thing that's really interesting about this phase 00:06:37.04 and that's that if you look at the early stages when the embryo is undergoing 00:06:42.13 these rapid mitotic divisions all the gene products that it needs, all the proteins 00:06:47.06 and all the RNAs are supplied by the embryo's mother, put into 00:06:52.14 the unfertilized egg before fertilization and what the embryo 00:06:56.28 is doing is just going through a cyclic pattern of DNA replication and mitosis 00:07:02.16 doing the same thing over and over again using the same gene products, 00:07:07.00 these gene products that are supplied by the mother. What happens 00:07:12.05 once those repetitive cycles are done and the embryo wants to do 00:07:17.15 something new or different, is that it begins to transcribe its own genome. 00:07:22.29 So it begins to make what we call zygotic RNAs and zygotic proteins. And so at 00:07:27.27 this point, it becomes a very interesting point for us in development, 00:07:32.29 because initially for this first two and a half hours, the embryo has been 00:07:37.19 doing something repetitive over and over using only maternally supplied gene 00:07:42.25 products and then the sense is that's when it begins to transcribe its own 00:07:47.14 genes and transcribe specific genes that are required ultimately to do new 00:07:53.18 things, to go on to the next step in development. So the sense is this stage 00:07:59.10 becomes really important for us to look at because it marks a stage 00:08:04.26 not only where something new begins to happen, you stop mitosis, 00:08:10.18 you begin to change and cells begin to become distinct from each other 00:08:14.01 but it's also associated with adding new gene products. Now I'd like to 00:08:20.11 though just continue on with the description of development, 00:08:23.09 for you to keep in mind though that now what we're going to be looking at 00:08:26.20 begins to reflect the active contribution of genes in the embryo itself. 00:08:33.03 What happens between these two stages you go from uniform behaviors 00:08:37.03 to distinct cellular behaviors you can see that again looking at the scanning EMs 00:08:41.11 of the egg. Here, this is again the embryo that I showed you before 00:08:44.24 with 6,000 cells, 100 cells all arranged along the anterior/posterior axis 00:08:50.11 all the cells look pretty much the same. If I'd fixed this embryo for scanning 00:08:54.18 EM about five minutes later, what you would have seen is that now all of the 00:09:00.00 cells in the embryo are no longer the same shape. Their clearly distinct 00:09:04.28 things happening, there's an area here that's ultimately going to form the 00:09:09.06 head of the embryo that's marked off by this fold, it's called the cephalic or 00:09:15.00 head fold from the rest of the embryo. There are other things beginning 00:09:18.06 to happen in the embryo clearly by this stage at this point, 5-10 minutes 00:09:23.08 later, the cells are marking, they're showing distinct behaviors 00:09:29.15 and showing how different they are. We can actually watch these behaviors 00:09:33.03 in living embryos by tagging the surfaces of cells with fusion proteins between 00:09:41.05 GFP and various membrane proteins so we can follow individual cell shapes. 00:09:45.12 And when you watch this movie again what you can see, this is again 00:09:47.29 the blastoderm stage that we talked about and watch in this area here 00:09:51.21 all the cells are pretty much the same, but right about there you can begin 00:09:54.15 to see this fold happening and then you begin to see remarkable changes 00:09:58.28 in the behaviors of cells, you can see, this movement of cells as they sweep 00:10:03.28 and move around the end of the embryo. Obviously as you watch 00:10:10.18 the process, we'll watch it one more time, all the cells are the same, 00:10:14.08 individual cells begin to become distinct, and with the extraordinary 00:10:20.12 reproducible pattern, so embryos always make a head and they always 00:10:24.11 make a head right here, they're always separated by folds. This behavior 00:10:29.07 here, these cells here that are moving into this invagination 00:10:31.29 here are ultimately going to form the endodermal or gut regions 00:10:36.07 of the embryo. We can roll this embryo over and now we can watch from 00:10:40.09 the ventral side, and we can see that other things are going on. You can 00:10:43.23 actually see a little bit of this head fold here but what's more striking 00:10:48.11 is this movement of cells, and watch right here you'll see a fold forming. 00:10:53.11 These are future muscle cells that are going to be brought into the interior 00:10:57.13 of the embryo, because obviously that's where the embryo wants to have muscles. 00:11:01.25 Now, by looking at embryos and characterizing the behaviors of individual cells 00:11:09.09 and the overall changes in morphology in the behaviors of individual cells. 00:11:13.18 What we've learned, is that all major morphological changes, all growth, 00:11:20.13 all changes in the visible appearance of the embryos involve local changes in cell 00:11:27.14 shape. The initial changes occur without any cell division. There is 00:11:31.05 no growth. There's no mitosis anymore after these first initial cell divisions. 00:11:42.22 The embryo will begin mitosis later, but at this stage these major changes 00:11:48.09 that we saw in the movie are all happening because individual cells in specific 00:11:52.19 places change their shape say from being long and square, to being rounded up. 00:11:58.19 And it's that kind of cell shape change that produces ultimately 00:12:03.12 the changes that we see here that will say separate the future head 00:12:07.07 from the tail of the embryo. So what we want to know is why is it 00:12:10.27 that certain cells, and only certain cells in certain places change their shape, 00:12:16.20 and others not? We know that those cell shape changes correlate 00:12:26.16 with patterns of gene expression. So that if we go back to this head fold 00:12:31.22 here that separates the head from the rest of the embryo, we can look 00:12:38.16 and see there is actually a single row of cells that are making the fold and 00:12:44.04 there are genes that are expressed exactly in stripes in this embryo that have 00:12:49.01 just begun to be transcribed at this stage that we talked about before. 00:12:53.05 The stage right when the embryo has completed the mitotic divisions that precede 00:12:59.21 the cell behaviors and mark the infolding of these cells such that 00:13:06.01 the cells that change their shape right here are cells for example that 00:13:09.27 are not expressing the green gene here which is a gene called paired, or 00:13:19.16 the orange gene which is a gene called runt. And so there are specific patterns 00:13:24.04 of gene expression, of transcription, that have arisen at this stage, 20 minutes 00:13:31.10 before the cells have begun to change shape that direct the cells 00:13:35.14 and control their cell behaviors. But all that does is of course is just push 00:13:42.13 the question back. We want to know why it is that cells in a given region 00:13:47.26 of an embryo show particular patterns of cell behavior, particular shapes 00:13:52.05 and now I tell you well that's because their expressing different gene 00:13:54.13 products that doesn't really answer the question because if what you 00:13:58.03 really want to know is how is it that spatial and temporal patterns of 00:14:01.09 gene expression are established at the blastoderm stage. And go back again 00:14:08.11 to this central idea that it's at this stage the onset, right before gastrulation, 00:14:15.08 right before these cell shape changes begin to occur that individual genes 00:14:22.13 come to be expressed in the embryo in specific patterns. I indicated that before 00:14:28.25 that stage cell behaviors were uniform and maternal RNAs and proteins, 00:14:33.25 and they depend on maternal RNAs and most of those maternal RNAs 00:14:39.14 and proteins, and actually for a long time we thought all of them were 00:14:43.08 uniformly distributed throughout the egg, but what we've learned now is that 00:14:48.13 we have to put the emphasis on mostly these maternal RNAs and proteins 00:14:54.02 that are supplied and necessary during these early stages are mostly 00:14:57.17 uniform but there's a very small number of proteins and RNAs that are put 00:15:02.05 into the egg by the mother and show distinct patterns of distribution. 00:15:06.09 And one of the most important of these and this will be important 00:15:08.27 for the remainder of my talk is a protein called bicoid. It's a transcription factor 00:15:13.15 supplied by the mother, it's present during these early stages and if you look 00:15:18.24 at its distribution in the early embryo you can see that this protein is localized 00:15:24.14 at high concentrations at the anterior end of the egg, the future head region 00:15:28.24 of the egg, and then grades off in cells as we move more and more posterior 00:15:38.20 in the embryo. And one of the things we've learned and that I'll tell you more 00:15:43.04 about is the controlling role for this protein distribution in establishing 00:15:49.04 the patterns of gene expression and transcription that occur at these stages 00:15:54.10 at this process right before gastrulation and that are responsible for the cell 00:15:59.08 shape changes. Now what's really going to be essential to the problem 00:16:11.07 is that if we have a graded distribution for protein, a maternal protein 00:16:17.13 in the egg. A protein that the mother has directly put into the egg. 00:16:23.11 How does that distribution of maternal bicoid protein established? 00:16:28.19 How is it formed? And from wonderful experiments by 00:16:31.21 Christiane Nüsslein-Volhard and Wolfgang Driever and a number of other laboratories 00:16:35.08 we've learned that what's central here is that this protein gradient that we can 00:16:42.06 see in this embryo here, it's about two hours old, arises not because 00:16:51.03 the mother puts the protein into the egg, and not because she puts 00:16:54.04 the protein in a graded fashion but instead what she does is when she's 00:16:59.18 making the egg back in the ovary back long before fertilization, when she was 00:17:03.28 making the egg, she deposits the RNA that encodes this protein and anchors 00:17:10.21 it to the cytoplasm in the anterior end of the egg. This RNA is not translated 00:17:16.14 during oogenesis, as long as she's holding this egg, as long as egg 00:17:21.01 is not fertilized the RNA sits there in an inactive form. When the egg is 00:17:27.15 fertilized, a consequence of fertilization or activation of this egg is that this 00:17:35.17 RNA is released from its anchor and begins to be translated because 00:17:42.16 the protein is not anchored, the protein is thought to diffuse from this source of 00:17:46.26 synthesis, continue to make protein constantly here from the RNA, 00:17:52.16 that's localized here, but the protein diffuses and what's established over time 00:17:59.12 in these first two hours, is a gradient of this transcription factor bicoid. 00:18:05.13 What then happens is that ultimately, and this is a little cartoon diagram 00:18:15.08 that the highest concentrations of the bicoid protein will be at the anterior 00:18:21.27 end of the egg the concentration will fall, this is a transcription factor 00:18:26.27 and there are genes in the embryo that are going to become 00:18:31.23 transcriptionally activate at this time this is the stage where 00:18:36.05 major transcriptional activation occurs in the embryo but 00:18:40.09 those genes are activated by bicoid as a transcription factor 00:18:48.04 in a concentration dependent way. So there are certain genes for example like 00:18:52.00 the hunchback gene that are activated by relatively higher concentrations 00:18:57.03 of bicoid protein and so show expression only in the anterior most 00:19:03.02 48% of the egg. So we can see here, that other genes that can be activated by 00:19:08.04 lower thresholds for example, the Krüppel gene shown here, and that 00:19:12.14 Krüppel gene then and the hunchback gene define domains 00:19:20.24 of gene expression, are the genes in fact that are expressed 00:19:25.07 in the embryo and are involved in establishing those spatial patterns. 00:19:30.09 Now, that understanding of development was quite remarkable, 00:19:39.16 something that the role of maternal RNAs and maternal proteins 00:19:43.24 bicoid was the first such maternal RNA which was functionally demonstrated 00:19:49.21 to provide this gradient like form of information across the whole embryo. 00:19:55.29 Discovered by Christine Nusslein-Volhard and Wolfgang Driever more than 00:19:59.04 15 years ago had an extraordinary impact in developmental biology 00:20:04.27 and our understanding of the processes of embryonic development. 00:20:09.13 This created a great deal of excitement cause you always ask in science 00:20:14.10 why was that result so exciting, what was so important about it. 00:20:17.26 And so I think there are a couple of interesting things that happen. 00:20:22.13 If you look at the model that you have a graded maternal protein that 00:20:29.09 controls transcription defined by having downstream targets, genes activated 00:20:39.15 in a concentration dependent way, a specific threshold. You establish a pattern 00:20:44.17 of gene expression, but what it's really telling us is that in biology information 00:20:50.25 is quantitative. People had even speculated that mothers may 00:20:56.05 put gene products in the egg to establish pattern. No one knew what the nature of 00:21:01.10 that maternal information was. What these experiments are telling us is that 00:21:05.11 information in biology is largely quantitative. The cells make choices 00:21:11.12 based on levels or concentrations of bicoid protein. And what that does 00:21:19.23 is it also tells us that the choice process depends not just 00:21:26.00 on concentration levels, but on the ability of nuclei, or cells, 00:21:32.10 to measure concentration and make permanent cell choices in response 00:21:39.15 to those measured concentrations. So one of the things we'll talk about is 00:21:43.02 how is it? What do we know about cells? What are these measurements? 00:21:46.24 How do they actually work? Is this the right way of thinking about the process? 00:21:51.02 What are the problems thinking about the process, but what the 00:21:53.25 experiments did is that by emphasizing the quantitative nature 00:21:57.15 of the information they changed how a developmental biologist 00:22:01.14 thought about the process of development. Another interesting thing 00:22:05.17 if you go back and look at the process and think about how it is that 00:22:09.05 the bicoid gradient is itself established. How is it you provide information 00:22:16.19 to an embryo. The initial localization, the initial pattern is a localized RNA 00:22:24.04 which is very finely localized to the anterior end. There's not a lot of 00:22:34.00 information in a simple localization of a single molecule in a single place. 00:22:39.04 What's important is that the final, that is we'll say the information rich 00:22:44.05 distribution that the cells are going to use to make their developmental 00:22:47.05 traces is not that initial distribution, but the final distribution is achieved by 00:22:57.06 simple physical parameters. You localize an RNA and then when that RNA 00:23:04.03 begins to make a protein, the protein diffuses. And if you think in terms 00:23:08.18 of what kinds of things can impact on that, things like how fast 00:23:12.18 molecules move, how fast are they degraded, all those things will ultimately 00:23:19.13 set up and define what the pattern of distribution is. All those things 00:23:26.21 are potentially measureable and so what's exciting about having a cartoon 00:23:31.18 picture of a localized RNA and a protein gradient, is that we believe that 00:23:38.20 once you have that cartoon in your head it directs you to say, 00:23:44.00 well what is interesting if we want to test this. Can we actually measure 00:23:47.11 diffusion constants? Can we actually measure degradation? Can we 00:23:52.23 actually test the model, can we as a biologist test the model by defining 00:23:59.09 the biophysical parameters associated with the generation 00:24:02.27 of biological information? And lastly, there is one other interesting point 00:24:11.03 that it took me a very long time to appreciate about this model, 00:24:15.15 but I think is really really essential is that after bicoid was discovered 00:24:19.29 and those remarkable initial characterizations, the expectation was, 00:24:25.15 it was at a time where we were coming to realize as molecular biologists that 00:24:30.15 most of the genes and most of the processes that one sees say in 00:24:35.04 one species or in one organism, most of those same genes are also found 00:24:39.22 in other organisms and they probably also function in much the same way. 00:24:42.29 So there was initially an expectation that the bicoid protein given 00:24:48.18 its predominate and central role, the central role that it plays in 00:24:54.14 governing embryonic development in Drosophila given its importance, 00:25:00.02 that one would quickly identify similar proteins the bicoid gene in the frog 00:25:07.02 or the bicoid gene in humans maybe even. And what's remarkable 00:25:12.09 and what was unexpected at the time was that in contrast to many genes 00:25:17.01 that are conserved to a hide degree of fidelity across all species, bicoid, 00:25:22.05 the protein of the bicoid gene, is a fairly new invention. Meaning that 00:25:29.06 even if you look in other insects, even if you look in other Diptera or other flies, 00:25:33.28 you don't find the bicoid protein. Bicoid evolved at a point in the history 00:25:42.12 or the evolution of the higher flies as a new gene, a new solution 00:25:49.18 to a problem that must have been old and has always existed for all embryos, 00:25:52.28 because all embryos have to be able to establish pattern. But what's interesting 00:25:57.06 then about this particular problem about bicoid, is that it's not conserved. 00:26:03.06 It's a new solution to a fundamental developmental problem. 00:26:07.24 And so on the one hand that's interesting because you can say well how is it 00:26:11.05 that organisms in the course of their evolution establish new problems 00:26:15.26 and new solutions. Why do they do that, and what's the nature of those 00:26:19.23 new solutions. The other interesting thing, the question that then though 00:26:26.13 arises from the evolution of bicoid is that although we've said that bicoid 00:26:33.18 is a fairly recent evolution among the Diptera, there are a number of flies 00:26:38.05 which we'll see have from the point where bicoid was established as a 00:26:43.15 patterning mechanism have a number of fly species that have continued 00:26:50.22 to use bicoid and those species have evolved in other ways. As you'll see 00:26:55.18 they make big eggs and small eggs and so what becomes really interesting 00:26:59.06 about bicoid is how is it, during the course of evolution, when a new solution 00:27:04.24 arises. How is it after an organism or group of organisms chooses 00:27:13.02 that solution. How is it that during the course of further evolution 00:27:17.29 that solution is modified or changed or adapted to make it suitable as the 00:27:24.03 individual organism evolve and radiate and establish themselves into different niches. 00:27:29.26 And so bicoid, I believe, ultimately from the standpoint of evolution really 00:27:33.28 provides an interesting opportunity to study how evolutionary pathways 00:27:39.17 are modified. So basically for the remainder two parts of this lecture what I'd 00:27:47.15 like to do is focus on those questions. I'd like to talk a little bit about 00:27:51.03 how it is that this bicoid gradient is established. What have we been able 00:27:57.24 to learn by applying biophysical techniques to the establishment, 00:28:01.12 to fly embryos to figure how molecules move and how stable are the actual patterns 00:28:08.10 of bicoid in the embryo. Is it stable enough to provide information 00:28:14.22 that cells can make choices on? And then in the very last part, 00:28:17.23 I'd like to speculate a little bit and tell you a little bit 00:28:20.10 more about our experiments on evolution. Thank you.