marianne bronner receivedher degree at johns hopkins university in biophysics and i have known her since she was a graduate student there. then she became an assistant professor at uc-irvine, and after 16 years in uc-irvine and rising up to become full professor, she then moved to cal
tech in 1996, and became the chair of the faculty at cal tech, and she was the first woman to hold that position, and she held that position for two years. so dr. bronner has done seminal work in the field of developmental biology, and her
lab has had a long standing interest in neural crestlineage specification, migration and differentiation, early work pioneered the use of single cell lineage labeling approaching for crest cells in vivo and showing the cells are multi-potent. she then went on to define
several of the signals underlying the induction such as segmental migration and recently her lab exploited this knowledge with systems level transcriptional profiling, genomic analysis and in vivo perturbation experiments to reveal connections responsible
for neural crest formation, evolutionary origin. her work led to the understanding of a number of birth defects related to neural crest development such as autonomic neuropathy and craniofacial defect and has published more than 300 papers
in top journals, she's the editor in chief of developmental biology, and on the editorial board of many journals. and for her outstanding achievements, she's won numerous awards, among them this year she was elected member of the national academy of science.
in 2013 she won the cochran medal from the developmental society for developmental biology, received the american society for cell biology women's award in cell biology as the senior awardee in 2012. and she was elected fellow of the american academy of arts and
sciences in 2009. so the title of her talk today you will all enjoy, and without any further ado i present to you dr. marianne bronner. [applause] >> thank you very much for the lovely introduction.
it's a pleasure to be at the nih i love coming to washington to see rain which we don't have in california. please ship us some. i'm going to tell you about our work on the neural crest which i've beening on for a long time, it's always fascinated me giving
rise to different cell types in the embryo and adult. one reason i like to study them, embryos are very beautiful. if you see this -- you can turn the lights down a bit please? this is an image viewed from above, nervous system, the brain, going down to the spinal
cord, neural crest cells arise from the developing nervous system but migrate out, you see them migrating in this beautiful stream. we study how the cells form and follow the migration inive willing embryos and we can learn by studying how they leave the
neural tube and migrate. they differentiate into the vast number of derivatives, forming all of the pigmentation of our skin, much of the craniofacial skeleton portions of cranial ganglia and other elements of the peripheral nervous system including enteric, sensory and
derivatives. we're interested because we want to understand how a single stem cell population can give rise to such a diversity of derivatives. we're at nih, one reason i tell my grantee agencies i want to understand neural crest development is because defects
in neural crest development are one of the most common birth defects, so about a third of all congenital defects involve the neural crest in one way or the other. in addition, the neural crest cells, when they go bad in the adult, give rise to many types
of cancers, and because they are such an invasive cell type in the embryo we think some programs they go through normally during development may have a role in causing neural crest derived cancers like melanoma and neuroblastoma. most recently, we've been very
interested in the possible regenerative potential of neural crest cells and would like to know if we can take the early populations of cells and by understanding their regulatory programs cause them to differentiate into particular types of derivatives or give
rise to self renewing progenitors for the purpose of regenerative medicine. these are many good reasons to study the neural crest but the reason i'm most interested is because they are such an interesting and beautiful cell type.
what i'd like to do today is start at the very beginning of neural crest development and tell you about how the neural crest cell forms and how we're trying to study the molecular aspects of their formation. so this is a schematic diagram showing the early embryo at the
time the embryo has three germ layers. this is the top, ectoderm, giving rise to the nervous system and neural crest. midline is the neural plate which will fold to form the central nervous system. so this tissue invaginates into
a cylinder neural tube. lateral is the non-neural ectoderm in white forming the epidermis and between the two is the neural plate border, and these contain the presumptive neural crest cells. during the process of neural tube closure this green zone,
the neural plate border, forms the leading edges of the closing neural tube. after the neural tube closes, the neural crest precursors are initially contained within the central nervous system and they are neuroepithelial cells, they lose epithelial connection and
they undergo something called an epithelial to mesenchymal translation to become migratory neural crest cells to move to distant locations in the embryo. here is just a section through the embryo where the cells start life in the dorsal neural tube and migrate out to along
distinct pathways to many sites in the periphery. there they differentiate into many different cell types, so neural crest precursors give rise to much of the cartilage and bone of the face, many neurons and glia of the peripheral ganglia, and all of
the melanocytes in our skin. however, the neural crest population along the body axis is not homogenous. cranial neural crest cells are the only ones to give rise to cartilage and bone of the face. those from the trunk region give rise to many derivatives
including ganglia and melanocytes but fail to give rise to cartilage and bone even if you take pieces of neural tube and transplant into the head region. there are interesting differences along the body axis with respect to type of
derivatives that neural crest cells form and this is something that has intrigued us for many years and will be part of what i'm going to tell you about today. so we've been studying the neural crest for many years now, and about a decade ago we tried
to understand this, what we would today call a systems biology approach, to try to understand the molecular building blocks that are functioning at different stages, and so we called this assembly of a gene regulatory network, it started just by looking at
individual genes that are expressed in this neural plate border region, and we found that in many different vertebrates the neural plate border genes like pax 7 and zic and msx were in the border. signals like wnts and bmps
are here. somehow genes are turned on that became expressed as the neural tube was beginning to close, neural crest specifier genes, including sox10, snail and foxd and others expressing as the nervous system is forming, the first markers that distinguish
this population from other cell types in the central nervous system. we speculated these genes in turn controlled other factors that are important for causing these cells to undergo an epithelial mesenchymal transition to start to migrate
and ultimately differentiate into many derivatives. i'm going to use this platform as a scaffold to tell you how we're attempting to study the neural crest system and i'll start by describing some old experiments in which we looked at one of these neural plate
border genes, pax7, to try to understand what it might be doing in the processer of neural crest formation. this work was done by a former postdoc of mine, martin garcia castro, looking for a marker that would describe this early neural plate border in a nice
and early fashion. so what he found was pax7 shown in red was expressed very early on in this neural plate border region and seemed to beautifully define and outline where the presumptive neural crest cells would come from. in order to analyze what this
might be doing in the embryo we want to understand the function of pax 7, speculating it was upstream of the neural crest specifier genes and did a knockout experiment. i should mention the first experiments i'm going to tell you about are done in bird
embryos. a lot of people here work on mice. mice are beautiful and wonderful animals, but it turns out nor neural crest development they are more difficult to work on and a lot of the data we've gotten to help us understand
early neural crest development has come from studies of other vertebrates like chick, zebrafish and frog. we wanted to knocks down pox7, we have a chick to knock it down on one side of the embryo. this is the primitive streak, and we know from mapping that
the neural crest forming region is about here. we can put an inhibitory molecule here, an antisense morpholeno, and we know this will be inherited by the neural crest cells on one side of the embryo, and this is nice because we can use the other side as an
internal control. what we found was when we knocked down pax7 that we in turn lost expression of neural crest specifier genes, including snail2 shown here, sox9, sox10 and foxd3, pax7 is upstream of the regulators, a way to validate this presumptive
network. so that's nice because it tells us this order in our network and that these neural plate border genes are upstream of our neural crest specifier genes. what it didn't tell us was if pax7 was regulating these downstream factors, in a direct
fashion, or in an indirect fashion, so the next step was to try to look at direct interactions in our network and the way we did that is by taking a view from the genome up, and i'll just show you one example of how we can do this with a neural crest specifier gene
called foxd3. we chose foxd3 because it's one of the earliest expressed as the neural tube is about to close. about the time the chick genome was sequenced, optimum distance from the mammalian to allow conserved regioned. the coding gene would be highly
conserved but we found in addition to the coding region, when we lined up many different genomes from human, to mouse, to rat, to opossum and chick there were peaks of conservation shown here. if they were conserved in the region they might be conserved
regions. in the check embryo, the nice thing about using the embryo is we can test these regions to see if they have enhancing activity in the neural crest. basically what we do is we take these conserved regions, putative enhanners, hook them up
and go back to the embryo, like we did with morpholinos, use electroporation, and it looks like this. this is from above, forebrain, mid-brain and hind brain and cranial region. neural crest with beautiful
neural crest crest expressing gpf, the pattern overlapped with the neural crest. so this is a way we can identify putative enhancers and i wanted to point out that for those who do work on mammalian embryos this is fantastic assay for chick enhancers, mouse
enhancers, a fast and easy assay in comparison to transient transgenesis in the mouse. the reason we did the experiments was to looks at foxd3, coding region is here, this is the next adjacent gene so we tested the blocks of putative regulatory regions and
found that two regions which we called nc12 and nc2 with enhancing activity and our assay worked but interestingly the enhanced activity was not the same for the two different enhancers. the first enhancer nc1
recapitulated endogenous foxd3, it exactly matches the pattern we got with electroporation of the enhancer. but the other region nc2 mediated expression not in the head but rather in the trunk region. so this was very interesting and
also very exciting to us because it suggested that encoded in the genome was information that was cranial specific or trunk specific regarding the expression of this transcription factor, foxd3, expressed along the entire axis. so the next thing we did once we
had the enhancers was found a minimal enhancer and then we wanted to look to see what might be mediating expression of that enhancer. so i'm going to cut through lots of hard work and just show you what the minimal enhancer looked like because then we can just
look at the sequence by bioatics and look for putative enhancer binding sites, and what you see here is all the different binding sites that we found using programs like jasper and others, and we test each site by mutating them individually. and many of these sites don't
have any effect at all, but what we found is that when we mutated a pax site here or ets or msx, all were necessary. if we mutate any one of these sites, we basically lose the that activity. so then that is very nice
because it suggests binding to these regions may be pax7, what i had predicted might be an input into neural crest specifier genes. but many things can bind to a transcription factor binding site so this does not prove pax7 is the binding site.
so then the technique that we next used to identify the direct input is going back to our electroporation experiments, we have our embryo and we take our wild-type enhancer and put a morpholino, co-electroporated with a morphino on one side of the embryo, on the other side
enhancer with control morpholino, the embryo looks like this, if the targeted morpholino knocks out the right thing we should see a pattern with control activity here and diminished activity on the side that inherited the morpholino. so what we found then when we
did this analysis with nc1, cranial enhancer is that pax7, msx and ets1 were critical binding sites and critical inputs into this enhancer. we did exactly the same type of analysis for the trunk enhancer and interestingly we found some common inputs from pax7 and msx.
however, ets1 had no expression on the trunk enhancer and in contrast another neural plate border gene zic1 seemed to be important. so just to summarize this kind of enhancer analysis, which we've now done for many different genes, what we found
is along the body axis, compression of the neural crest specifier gene foxd 3 was controlled by nc 1 and nc 2, the cranial enhancer had an additional input that was required from ets 1, another transcription factor, whereas the trunk enhancer didn't
require ets 1 but instead required zic1. this made sense to us when we looked at the distribution pattern of these different transcription factors. what we found is pax7 and msx expressed along the body axis. ets1 is a cranial-specific
transcription factor whereas zic1 is specific to the trunk. so that suggesting this regulation had both common and different inputs, and then we went on to verify this using chip analysis and all the things i talked about are direct inputs into the different enhancers.
the reason i show you this is because this allows us to elaborate the gene regulatory network working in the head and the trunk region and we've looks at inputs into foxd3 allowing us to understand the inputs are actually direct in our network. the other nice thing is that
using this kind of analysis we found transcription factors that we didn't really know were part of our network. for example ets1 will had not been previously put in as a member of our network, ets1 was also a cranial specific factor. so that's a nice way to
demonstrate direct interaction and identify new things. so this is nice and we've done a lot of analysis of this sort with many different enhancers, and so this has provided a way to really construct this network in a much more detailed form. what i'm really excited about is
none of this but rather the fact that now this gives us a great tool to further study the neural crest. the cells start in the dorsal neural tube but move all over the embryo, a problem if you want to understand what factors are present in this isolated
population. by using these enhancers as a tool, however, we can now sort discrete populations of neural crest cells pre-migratory, migrating stages or even very late in differentiation, and this gives us a handle on further exploring the regulatory
landscapes present in these cells. basically what we do now is we can use these enhancers and sort the cells at many stages, either early in the neural plate border, when neural crest cells are pre-migratory or actually migrating through the embryo.
so this is work that was done by marcos coasta, a wonderful post doc in my lab doing transcriptome analysis of the populations, the technique is to do exactly what i showed you, electroporate into embryos, performing rna-seq, and this was standard, many of you are doing
similar types of things, but i was excited about it because when i -- every time a new technique came about to somehow be able to isolate the transcriptome of cells, we tried it. for those who are older in the audience, we tried differential
display which is probably like 20 years old. we've made macro arrayed libraries and found lots of interesting genes, but very few transcription factors came up in these analyses. what i was heartened by, by doing rna-seq, we found 1300
genes upregulated in the cranial neural crest compared to the whole embryo, and of these 50 were novel transcription factors that we had not previously put into our network. that was good news on two counts. we got a lot of transcription
factors and secondly we didn't have so many that we couldn't possibly analyze them all. 50 is a reasonable number to study. so then we did this nice transcriptome analysis, we wanted to get the paper published, hopefully in a good
journal, so we did all the things to make it look pretty and one of the things we did was this biologic pathway analysis, which isn't really that interesting, but i wanted to point out one thing. that is that we found lots of genes that were upregulated in
win signaling, important in northerly crest formation, so i was very interested to see this large upregulation in factors that were involved in the win pathway. this took me back to some experiments that we did many years ago, to look to see what
kind of signaling molecules were important neural crest cells form here, the juxtaposition of the neural and non-neural ectoderm, it didn't take a rocket scientist to think that perhaps there's some kind of interaction going on between the
two tissues, the presumptive epidetermine is and presumptive we took a region. tube and putit together with a region of ectoderm that's also far away from where neural crest cells form and when we put them in culture or let them stay in the embryo, what we find is that
juxtaposing the two tissues, neural and non-neural, gives us a new cell type, the neural crest, so this is like a classic inductive interaction. we did a candidate search to look to see what might be involved and what we found is is that wnt was very nicely
expressed in the ectoderm, just the right place and time to be responsible for the signaling. we then blocked wnt signaling here and found we no longer got neural crest formation, and we further took a piece of the neural tube from here which normally would never form neural
crest, add wns and get neural crest cells, conclusion wnt is what you need to get neural crest, wnt is important in neural crest formation. going back to the regulatory network that, for why we thought wnt signaling was important. when i saw the rna-seq data, one
of the genes that was upregulated in our screen was actually one that was thought to be wnt effector gene, we wanted to analyze whether the gene in the next slide might be somehow involved in linking this signaling module to the neural crest specifier module.
and so ets 1 is expressed early in the neural crest as shown here, we found it was knocked out in mice and had a neural crest defect in the palate, making us wonder if it might be important for expression of some of these neural crest specifier genes and in fact we found axud
was earlier expressed, in a pattern that nicely overlapped with it, double in situs, we found fox d3 came up in the axud domain, shortly therafter. the same experiment that i've shown you many times before you now, electroporatting, into one
side, the other side the when we lost axud expression we lost foxd3, the opposite experiment expressing it early in neural fold we get premature foxd3 expression. interestingly, shown here, we not only had an effect on endogenous foxd3 but the foxd3
enhancer shown previously was also expression mediated enhancer eliminated when we knocked down axud. so i showed you this before, where we looked at the sequence of this enhancer, for putative binding sites. the we really didn't know that
axud existed when we first did this but now we went back and found that in fact there was an axud binding site. when we mutated that site we lost foxd3. that was intriguing to us and made us think that axud may
really be a direct effector gene of wnt signaling. to verify that we looked for closely and when we looked through sections through the embryo, here is the neural tube, this is the dorsal neural tube where neural crest cells alies, looking at wnt1 expression, axud
is in that domain. we knocked down wnt, lose axud, and the same thing happens with beta-catenin. what this showed then is that axud was downstream of wnt, but if in fact axud is the wnt effector that's responsible for turning on neural crest
formation, then we would expect it to be able to rescue. so this is the critical experiment. because we know that wnt is important for neural crest formation, in fact when we knock down wnt we lose neural crest formation as shown here but what
we found is that when we knocked down wnt but added back axud that was sufficient to rescue the phenotype. that took axud as the downstream effector that mediates expression of genes like foxd3 and it works with others as well showing axud is important for
turning on the neural crest specification module. that was nice and sort of validates the fact that these rna-seq screens can give you nice new genes you can put into your network but there was also one other puzzle we were hoping this would solve, and when we
formulated this gene regulatory network, we knew that there were transcription factors at the neural plate border and signaling molecules and both were somehow important but what we didn't know is how these modules interacted with one another.
so because axud was important for turning on the same genes pax7 and msx were responsible for we wonder if there was a link with the neural plate transcription factors and wanted to test to see if they might be interacting with one another and to do that we used a technique
called proximity ligation, which allows us to look in vivo at interactions between different pairs of transcription factors and other molecules. so this just shows the results, here we're look at dorsal neural tubes in these pictures. in this case we looked at
interactions between axud1 and pax 7, when you see interaction you see the red dot. pax7 and axud did in fact interact as did axud and msx, as negative controls we used ets1 and snail and they did not interact above what background levels you would see in other
parts of the neural tube where transcription factors are not present. this suggested that in vivo axud was interacting with pax 7 and msx. our model is axud is a missing link that allows this neural plate border module to integrate
with the signal module and turn on neural crest specifier genes. and our results show that these then go on to the enhancer region of genes like foxd3 and mediate expression and each component is required and we found if we eliminate axud we no longer have binding of pax7 or
msx to the enhancer region. what i like about the story, we took an approach where we looked for genes expressed in the neural crest by transcriptome analysis but by ticking candidate genes like axud and looking in depth we were ableto solve a puzzle, how you mediate
interaction between signals like wnt and transcription factors at the neural plate border like pax7 and msx and what data show is that cannonnical wnt signaling mediate axud expression turning on foxd3 to turn on neural expression, it gave us a great in road into our
regulatory network and shows how we can take the information gained from transcriptome analysis and really get a better understand of the system. so emboldened by this, we went back and we wanted to really try to understand other components of the neural crest network, and
everything i've shown you so far has been looking at the cranial neural crest and we've focused on that basically because it's easier. so one of the things i'm very intrigued by is why are there differences in the potential of neural crest cells at different
levels of the neural axis, giving rise to cartilage and bone of the face, whereas trunk cells ordinarily cannot do so. so if transcriptome analysis works so nicely for cranial crest we thought it would would be a nice way to compare cranial versus trunk.
so marcos went back and repeated this same type of transcriptome analysis looking at genes that were expressed in the pre-migratory cranial versus pre-migratory trunk neural crest, and we focused on the cranial neurocrest genes because we want to understand why they
can make cartilage whereas trunk cells cannot. and we were very happy to find, again, novel transcription factors in the pre-migratory neural crest and found 15 transcription factors that were uniquely expressed at this axial level, and this is really great
because our experiments are largely done with morpholinos, which with pretty expensive, so if we found even 50 would be hard to work with, but 15 was manageable within our budget. so what i'm going to show you then is how we're taking this kind of analysis and trying to
use these new genes that we're finding and try to order them in our regulatory network. so we found a lot of the beautifully expressed genes, brn3, dmbx1, lhx and a xud, cry which is cranial specific. we want to use analysis to put them in a network.
with axud, it's a mediator of foxd gene expression, ets1 was also an input specifically in the cranial region. we have novel factors, dmbx1 is expressed slightly after brn3, expressed nicely in the early neural folds. then we test what it does, the
same way i've shown by doing a knockout on one side and control on the other and what we found is that when we looked, using ets1 as readout, we lost ets expression when 2003 knocked down this gene putting dmbx1 upstream. we looked at another factor, myb
is an interesting gene, because when we knocked down myb we also lost ets1, suggesting myb was an input into ets1, we sent to review, a leftover of a graduate student's project, the reviewer said, well, you looked at a few genes but what about other interesting genes like zic1. we
did the experiment, as one has to do for reviewers, and what we found was really interesting, although myb 1 seems to be an input, it inhibited zic1 expression. ets1 is an input in cranial region, zic1 in the trunk. early on zic is expressed in the
head but when myb comes on it gets turned off. we're grateful to the reviewer because by proposing the experiment we have a better understanding of how the network works. we looked at other genes, one of the earliest expressed was brn 3
and when we knocked it out we lost dmbx1, temporally and spatially, the earliest that comes up in the cranial specific we looked downstream at later genes expressed and found alx seems to be important and alx is downstream of ets1, important for cartilage formation.
i went through this quickly and i did that intensionalally. by doing simple experiments and taking advantage of expression patters you can begin to assemble a decent regulatory why do i want to do this? one of our goals in terms of regenerative medicine is to see
if we can program cells like naive neural crest cells to make cell types like cartilage and other cranial-specific cell types. what we'd like to do is test whether by taking components of this network, we might be able to convert cells from naive
cells or from ips cells into particular lineages. i'm just going to show you briefly how we're starting to try to do that. we call this reprogramming, the interesting word right now, but basically here is our embryo. i've shown you that we have
cranial specific enhancers like the foxd3 enhancers, this normary -- normally comes on inthe head but fails to come in the trunk region. as our first step to reprogramming we're trying to see if we can take that network i showed you and misexpress it
in the trunk region and perhaps turn on this cranial-specific program. so we've made attempts to do that and the first thing we did was tried to express this very early factor brn3 and see if it would turn on the cranial enhancers in the trunk, and the
bad news was it didn't work at all. but when we took this construct and then knocked -- simultaneously did a loss of function for zic1 we found the cranial enhancer now was expressed ectopically in the these are preliminary
experiments and we're continuing them. these cells do not go on to make cartilage yet. we know we need to add other components but we think this is sort of an inroad into trying to use our regulatory information to then reprogram the neural
so everything i've shown you so far has been using the bird embryo which i think is very good for looking at early steps in neural crest formation, and so i've shown you sort of three different stories, one where we did in depth regulatory analysis to look to see how genes in our
network were regulated and i showed you how we found that foxd3 and cranial crest was regulated by pox7 and msx, in the trunk regulated by paxand msx together with a different transcription factor zic. using these enhancers we can then do a lot of nice
transcriptome analysis giving my lab basically everything we need to do for the next hundred years or so, if i last that long, and what we found by then taking some of that transcriptome data and looking in depth we found a new player axud 1 giving us an exciting result to turn on
neural crest specification working withrd boor -- with border genes. using the information we gleaned from earlier experiments to do reprogramming, basically we're trying to assemble a gene regulatory network that tells us what's different between cranial
and trunk neural crest cells and going forward we would like to do this for other neural crest populations as well. so this is what we've done to look at early events in neural one of the things i like to do going forward is go later in time and look at how the cells
are migrating and how they are differentiating. and to understand cell migration and differentiation, the chick system is good in some ways but it's also problematic in that it's hard to visualize cells as they are migrating, you can do it for short periods of time but
later on it's hard to see them. so for the next step in this analysis, we've actually transitioned to largely using the zebrafish model and i just wanted to show you some rather quick experiments to show you what we're trying to do going forward.
and these are experiments done by another postdoc in my lab, and he got very interested in migrating neural crest cells and their contribution to different structures and found an unexpected result and that is that neural crest cells not only form the derivatives that we
knew about that i showed you before like pigment cells and neurons but contributes to elements of the node to specific population of olfactory neurons. this is a zebrafish embryo, head here, neural tube here, green cells are migrating neural crest cells and they migrate and
populate almost the entire head region and surround the nose at later stages, this is a slightly older embryo, here is the nose in red with the neurogenin problem. if you look into the nodes, this is an embryo viewed face on, here is the neural tube, here is
the olfactory epithelium, on the left, and the right side, there are ciliary neurons shown here in red, but he found with time these green cells that are neural crest derived began to come in from the nasal cavity and invade the olfactory epithelium and become another
population of neurons which were called microbillus neurons. we're interested in studying how they move from the neural crest into the nose and differentiate into the specific and beautiful population of neurons, so this is shown here and the reason i show you this in zebrafish, it's
so much easier to visualize. this is another one at a later stage, olfactory epithelium, and we were interested in following many aspects of development. we wanted to look at how they got into the olfactory epithelium and how they differentiated into neurons and
finally how they put axons to the olfactory bulb but you get a nice face-on view of the embryo but we really wanted to look in 3d and so when you look at the olfactory bulb you see the neurons, if you rotate 90 degrees it looks awful and you can't see very much.
so we were stuck here and one of the things i was talking to lot of students about this and postdocs about how interesting the technology development has been and how that's helped our science, so here we thought perhaps by using new technology, we could actually do better and
see deeper. so at about this time, a talk from eric's lab, he developed a lattice light sheet microscope, we decided to see if this would help us in image. he went to genelia to see if this would work on a bird embryo, previously applied to
drosophila. it has high resolution in four dimensions and has very low phototoxicity. so this is actually one of the early im images he got, eachframe shows the same olfactory epithelium with crest cells that form neurons, you can see from
four angles it looks almost identical, really very beautiful. this shows the comparison between what we were seeing with traditional confocal microscopy and lattice light sheet, which is just beautiful. so this is i think a clear
example of where modern technology can be addressed to classical embryological experiments and although we're working on this trying to understand different processes we think this will be a great way to understand neurogenesis of this population of cells.
so basically we're applying this to cell migration from the nasal placode into the neural crest, we want to see how they are guided, how axons make their way to the olfactory bulb, and the nice thing about this technology not only can you look broadly, you can look deep at
cytoskeletal elements, very exciting. i'll end by showing a gratuitous movie, this is the nose of a slightly older fish, and we're just looking with lattice light sheets through different layers, i'm showing this because it's really pretty, and you can see
the cilia beating, we slowed the movie down four times to follow the cilia beating. i'm going to end there and summarize. what i've tried to tell you about today is our neural crest gene regulatory network, and basically we've been working on
this on a number of species, mostly i showed you what we found in the chick embryo and i showed you we can learn about direct regulation of genes within our regulatory network and use enhancers as tools to do transcriptome analysis which gives us lots of new material in
order to try to analyze this network in every deeper depth, but in order to understand what's going on you can't just look broadly. it's great to look at a systems level, and we all sort of say systems biology is the thing of the future but really to
understand what's going on in detail you have to look in depth and so i showed you one example of you who by picking a candidate gene axud, which has been proposed to be an effector of wnt signaling we solved a mystery in the field and found axud integrates wnt signaling
with neural plate border transcription factors to turn on the neural crest regulatory how cells migrate and differentiate into different populations of cells, we've seen in zebrafish unexpectedly that neural crest cells contribute to a population of neurons,
microvilleus neurons in the nodes and trying to understand the role. to end, i want to acknowledge the people who did the work. most of the work i talked about today was done by a fantastic postdoc in my lab, marcos simoes-costa together with son
ya and tatiana. he did the axud experiments and in-depth analysis of this transcription factor and transcriptome analysis and has been assembling the network. i talked about janelia forms to
use lattice light sheet microscopy to understand how neural crest cells migrate and differentiate. i'd like to end there and i'm happy to take any questions. thank you. [applause] >> hello.
>> at the beginning of your lecture you mentioned -- i know you were going more for the reprogramming of cells but you did mention a melanoma neuroblastoma and glioblastoma. melanoma is diverse, that will go to the melanoma and to a million other things.
the neuroblastoma is essentially a pediatric tumor with the property that it can spontaneously regress. and the glioblastoma multi-form is the end of a high end of range of astrocytomas in the brain. i can't find any common
denominator. they are very -- my experience in pathology, they are strange actors at the periphery of everything. in terms of transcriptome have you found anything you could apply actually to tumors that would serve in two manners, with
diagnosis and treatment. >> that's exactly where we're going. i want to an anr meeting last year for neuroblastoma and i got excited about neuroblass tomorrowa struck by the fact people didn't know the normal profile of the cells so one of
the things we want to do is profile the adrenal precursors and see what type of neuroblastoma they may resemble. my hypothesis the more neural crest like they are, the more aggressive the tumor will be will you i -- but i have no data.
that's what we'd like to know. genes that are, you know, always mutating neuroblastoma included, we've been studying myc, i didn't have time to show our data, but myc has an important function in the early normal crest relating to defects and cell migration and could be
involved in metastatic behavior that would correlate to neuroblastoma. we're just beginning these studies but my hope is to -- i think that the tumors that are most aggressive are going to be those that most resemble the least differentiated neural
crest cells, if that's the case by understanding the normal transcriptional program we may be able to find some therapeutic targets, but that's sort of years down the line. >> but the diagnostic aspect would be most important in melanomas.
>> right. >> the other ones are not really diagnostic challenge, just to mention that. >> well, yeah. >> okay. thank you very much. >> you mentioned ets1, is that linked the way you linked wnt
signaling? >> nobody knows the links. the data comes from many vertebrates, and fgs signaling has been best studied in frogs and we don't actually see in our system much, at least in the cranial crest for fgs, we have evidence for dnp signaling, i'm
sure that will come in. there's probably a little bit of notch signaling involved, perhaps at a later stage, but one of the things i'm excited about with these experiments is we don't know how these things link up, so by finding -- by doing this pathway analysis
which is in some ways vacuous because you see the different components, but in the case of wnt we found at least one component that seems to be a, link, maybe others will be similar, that's the long-term hope. >> i'd like to thank marianne
for the spectacular talk and audience for questions and i'd like to invite you to the library for a reception where you can continue to ask her questions. [end of program]
0 comments:
Post a Comment