Scientist Stories: Ophir Klein, Genetics of early tooth development and dental disorders

I want to thank Brunella for the organizing of this session. And I'm going to start with a basic science part of the session, telling you about genetics of early tooth development and then later into how teeth mineralize. And then Agnès is going to give the second part talking about the more clinical aspects. So I'll kind of segue into the clinical stuff in the second part of my talk. So our lab is generally interested in how organs form in the embryo and how they renew in the adult, and we look at a number of different organs, particularly focusing on epithelial biology, including the tooth, which is what I'm going to focus on today, as well as other craniofacial structures like the tongue and the oral mucosa, and we also study the intestine, but I will focus today on our, on our tooth work. One of the concepts that I want to convey throughout the talk is how fascinating teeth are from a biological perspective. I'll touch a little bit on the different shapes of teeth and the genetic regulation of this shape, and also from an evolutionary perspective, how interesting teeth are for biologists. Then, as I said, I borrowed this image from Agnès. In the second part of my talk, I'll move a little bit more into human dentition and models of human dental disorders. And then Agnès will tell you about a whole host of clinical syndromes. So the organization for the talk with a general theme of how teeth are a great model for organogenesis is kind of marching through the various stages of tooth development, starting with the early events of how teeth are positioned in the jaw,, and then how signaling centers form and how they regulate morphogenesis. And then I'll switch into using the rodent incisor, which is a model that we and other labs have been developing as a system for the study of adult stem cells, to discuss a syndrome, Costello syndrome, which is one of the RASopathies, and some dental findings that we made in that syndrome. And then I'll finish with talking about ectodysplasies in the gene responsible for X-linked hypohydrotic ectoderm, dermal dysplasia. So teeth, like many if not most other organs, undergo a lot of epithelial-mesenchymal interactions to guide their organogenesis. And throughout my talk, I'll have the epithelium in blue and the mesenchyme in yellow. And these kinds of epithelial-mesenchymal signaling interactions that we'll discuss are mediated by members of the major growth factor signaling pathways, including FGFs and BMPs and hedgehogs and Wnts. Specifically, what we're going to be talking about is the tooth, but these kinds of interactions are important for the ectodermal organs in terms of the types of tissues that are involved being the ectoderm. But these types of interactions also can teach us much about what happens in the development of other organs such as the kidney and the lung and others. So as I said, I'm going to use the stages of tooth development as the guiding principle for the talk. And teeth form at their early stages from two ectodermally derived tissues, the oral epithelium, which is derived from the ectoderm, as well as the neural crest-derived mesenchyme, which starts off as an epithelium and then, as I'll show you in a minute, migrates into the branchial arches and forms the mesenchyme. And the earliest events in development of the tooth are the placodes. So I'll start off with some examples of this. And what I've tried to do in the talk is both interweave work from our lab as well as work from other labs in the field. So the placode stage is one in which the oral epithelium forms this little invagination into the underlying mesenchyme. And then this invagination process continues into the next stage, which is called the bud stage, because there's a little bud that forms. And at this point, the mesenchyme begins to be a distinct structure, which is seen visibly condensing around the epithelium. The third stage is called the cap stage, because the epithelium forms what looks like a cap around the mesenchyme. And here, there's a very important structure called the enamel knot. This actually doesn't have anything to do with enamel. It's a historical term. This is an organizing center, which I'll discuss in a little while. And this secretes members of all of these major signaling pathways. And then the final stage is called the bell stage because it kind of looks like an inverted bell. This is also known as the cytodifferentiation stage because it's when the tooth-specific mineralizing cell types like the ameloblasts that make the enamel, the odontoblasts that make the dentin, and the cementoblasts that make cementum form. And it's also when a lot of morphogenesis happens, so you can see the cusps forming. And then this will then erupt out into the oral cavity, and that's what gives the shape of the adult tooth. And as I'll also discuss in a little while, in human teeth and in many mammalian teeth, the epithelium will die away after the enamel is formed. But there are some very interesting mammalian species that we'll discuss in which this doesn't happen and that can potentially teach us much about how nature normally renews dental tissues. As I said, there's, in addition to the medical and dental implications of understanding tooth development, there's very important evolutionary consequences of this understanding. And I think it's also helpful to think about teeth from an evolutionary perspective to understand how we have the teeth that are present in mammals. So teeth are a vertebrate innovation and they arose in animals that were our ancestors, kind of like this fish in which within any given species, all of the teeth are the same. So this is called a homodont dentition. And also all of the teeth are continuously replaced place. If you think of, for example, a shark. So this is called polyphyodonty. Now, when mammals arose, there were a number of changes that happened as represented by this dog skull here. And dogs and humans have a relatively primitive mammalian dentition, meaning our teeth are relatively similar to the mammalian ancestors in many ways. And there were a couple of important changes that happened. One of them was the switch from polyphyodonty to diphyodonty. So we and dogs just have two generations of teeth, the milk teeth and the permanent teeth. Also, there was a switch from homodonty where all the teeth are the same morphologically within a given species to heterodonty, in which you can see here that there are molars and premolars and canines and incisors. Now in the lab, we use the mouse as the main model for understanding tooth development. Obviously, these skulls are not to scale because a mouse skull is much smaller than a dog skull, but I've just put them at the same side so that you can compare the changes that have happened. Mice have a very derived dentition, quite different from the ancestors. One of the changes that was important is that mice are monophyodonts, so they never replace their teeth. They just have a single generation of teeth. There also have been a lot of changes in the number and the shape of teeth, and I'll review some of these as I'm going through my talk, including mice have no premolars or canines. They just have this diastema in this region. Then another topic that I'll discuss in a little while is the ability of mice to grow their front teeth continuously, and this is a stem cell-driven process that we've become very interested in. So as I said, humans have dentition which is similar to the ancestors, but it's also a little bit reduced. So for example, we have fewer incisors than the ancestor, but again, just to compare, this is here a mouse jaw now laid on its side where you can see the three 3 molars present here, and the incisor, which is here, is this long tooth that goes all the way back under the molars. And that's an ever-growing tooth, which is in contrast to the human incisors that have roots. So these are some of the general ideas that I want to discuss. Organogenesis involves a number of fundamental processes, including initiation, which is the positional information that's provided to place the organ in the proper place in the jaw in this case. Then morphogenesis, which is how the shape of the organ is built up. And then differentiation, which in the case of the tooth is when the mineralized tissues are formed. And as I'll try to show you in the rest of the talk, teeth are really a great model to study all of these different processes. And so the types of ideas that we'll talk about are How does a tooth form? How are the different tooth types determined? How is the position of the tooth in the dental arch determined? And how do these mineralized tissues form? And the reason that we're interested in the tooth is as a paradigm for organogenesis to understand all of these various concepts, including positional information, shape determination, size control, cell proliferation and migration, terminal differentiation and matrix formation, concepts like how programmed cell death sculpts the organs. And we kind of think about it like this, which is where and when and how does the tooth start its development, what regulates its shape, how is it then— the shape continue to develop, and then how is this process stopped and the terminal differentiation occurring. So the starting materials, as I mentioned a few slides ago, are all derived from the ectoderm, including the oral epithelium, as well as the neural crest-derived mesenchyme, which I'm sure this audience is quite familiar with. But just to review for you, the neural crest begins at the border of the neural and non-neural ectoderm, and then it migrates out to form a very impressive variety of tissues throughout the embryo. For our sake, it's these ectomesenchymal cells which are going to be the ones that produce the neural crest-derived mesenchyme. So these will migrate into the first pharyngeal arch, which is where the jaws are derived from, both the maxilla and the mandible, and provide the mesenchymal starting material for the tooth. Establishment of the tooth field begins quite early in the mouse, around embryonic day 10.5, and And at this stage, there are a number of very important interactions involving signals from the epithelium into the mesenchyme, particularly members of the FGF and BAP families, which induce the formation of a number of transcription factors. And some of these are important for human disease and will be discussed in the next talk. I just want to point out here that when you read these reviews of dental development, you often see FGF8 as being reported to be a critical initiator of tooth development. And this is a molecule that I'll discuss in the next few slides. So for the first story from our lab, I'm going to tell you about the role of these FGF8-expressing cells in organogenesis. So when we began this project, we saw many of the cartoons similar to the one that I showed you on the previous slide in which FGF8 is reported to signal directly to the underlying mesenchyme and thus initiate tooth formation. But actually, if you look at the literature, that was all based on experiments in which FGF8 was knocked out and led to severe jaw defects. And so that suggested to us there was perhaps more to look at in this, in this field. So when we began the project, and the idea was that there was a single population of cells that invaginated into the underlying mesenchyme and in a single place form the tooth. And that's represented, for example, in a cartoon from this review from Irma Teslaff and Jukka Jyrenval, in which the epithelium is represented signaling back and forth to the mesenchyme, but in a very static fashion in terms of its position. So, we were interested in understanding this process in further detail, and this is all the work of Honza Prochaska, former postdoc in the lab who now is back in Prague and has his own lab there. And again, what Honza was interested in doing is interrogating further this idea of whether these FGF8 expressing cells function in this static fashion. And so the first thing that he did was just review the expression of FGF8 using an FGF8 lacZ knock-in allele. And what he saw was consistent with the previous in situ hybridization studies that the FGFA was expressed at the early stages of tooth formation, and then by the cap stage had gone away. And I should just mention that in these cartoons, what you're looking at here is you're looking down at the jaw. And this is the molar field, and this is the incisor field, and this is the tongue here in the middle. So the next question that Honza wanted to answer was, what are the descendants of these FGF8-expressing cells? And for this, he used a lineage tracing approach in which a Cre recombinase was knocked into the FGF8 locus. So when this was then crossed with a lacZ reporter, it enabled us to trace the fate of the FGF8 expressing cells. And here we're using a constitutively expressed Cre. I'm also going to show you some experiments with an inducible Cre, but the results in this case were quite similar.. And so, what this shows you is when FGF8 is first expressed, it's labeling these cells in the molar field. I should mention again, these are the whole mount views, so looking down at the jaw, and these are the section views in this plane. And then, when we look at subsequent stages, this is now embryonic day 12.5 and then 14.5, so this is the bud stage and the cap stage. And what you can see is that at all of these stages, the epithelium is labeled from these early FGF8 expressing cells. So, what this tells us is that the entire epithelium of the tooth is derived from these early FGF8-expressing cells. And then to ask whether these cells are required for tooth formation, what Honza did was cross this FGF8 now inducible Cre with a diphtheria toxin. So what this does is whenever the Cre activity is induced, this kills the FGF8-expressing cells. And what he saw was that compared to the control, which at embryonic day 14.5 has this nicely formed cap stage tooth. When he killed the early FGF8 expressing cells, these ones, that led to the absence of tooth formation. So what that tells us is that the FGF8 expressing cells are required for tooth formation because when we kill them, there's no tooth that forms. So then what Honza did was a very clever experiment, which was to watch what the cells are doing. And again, this gets back to this question of of whether the populations are static in the early jaw. And here what he saw was a very fascinating finding, which is if you look at the red asterisk here, this is where the FGF8 expressing cells are, and this is about a 48-hour movie. And what you'll see is that there's this stream of cells migrating anteriorly from the FGF8 expressing region. And here I'm just showing you this in higher magnification.— and this was very surprising to us because it really contravened the previous idea that the cells were just signaling from this static population and invaginating directly into the mesenchyme. But rather, what we saw was that there was this dramatic anterior movement of the cells. And then what he was able to do— and I'll show you a couple of other examples— is actually trace the forward movement of individual cells. So each of these colored dots here represents a cell, and then he's able to trace it migrating anteriorly over time. And what was interesting about this was that they were all moving in a similar direction. So it's not exactly a collective cell behavior, but it's some sort of coordinated or organized anterior migration. So then the next question was, what regulates the anterior migration of these cells? And here, what Honza did was then go back to the literature. And as I showed you in that previous slide with a cartoon from the review article, in every review that you look at, what you see is that there is the idea that these signaling molecules that are important are coexpressed in the early developing tooth. So that FGFs and SONICS and BMPs are sort of expressed in the same region. So Honza decided to look at that again. And the way he did this was by crossing a sonic hedgehog reporter— this with a GFP knock-in— together with the FGF8 LACZ knock-in that I showed you before. And what he observed was that rather than having Sonic and FGF8 coexpress, as had been previously thought, that these were spatially distinct. So the Sonic-expressing cells are actually anterior to the FGF8-expressing cells. So this then led us to the hypothesis that perhaps Sonic and FGF— signaling co-regulate this anterior movement. So, in order to ask whether this was the case, Honza then cultured developing teeth in the presence of either cyclopamine, which is an antagonist of sonic hedgehog signaling, or SU5402, which is an antagonist of FGF signaling. And what he saw was that these two antagonists had different effects on the movement of the cells in the culture. So, in the control, all of the cells which start off here will move anteriorly over this couple-day culture experiment. Whereas in the cyclopamine, the cells are still moving, but rather than moving in a focal way, they're spreading out much more broadly. And in the SU5402, the FGF blockade, all the cells stay posterior. And I'll explain to you in a couple of slides what we think is going on in terms of a model. So, you can also see in the cyclopamine this sort of spreading out of the cells, and in SU-5402 they're still in an invagination, but it just remains very posterior. And that's also reflected in these reconstructions where in the control the cells have moved anteriorly, in the cyclopamine they're spreading out and not really invaginating deeply, and in the FGF blockade they're just staying posteriorly. And another way to visualize this is with this cell-tracking method where each vector here represents the movement of an individual cell. And you can just see visually, and I'll show you the summary on the next slide, that in the control, it looks like all of the cells are migrating at a similar— the length of the vector, length of the arrow is the distance they're migrating. Looks like they're all migrating at a relatively similar rate and in a coordinated fashion. In cyclopamine, they're not moving in the correct direction, and in the FGF blockade, the arrows are much shorter and. And it looks like the cells are just kind of hovering in place. And so then Honza quantified the movement of each individual cell. And what he saw was that in the controls— excuse me, each of these blue lines is an individual experiment. This is about 3 or 4 experiments. In the controls, all of the cells are moving again in the same anterior direction. In the cyclopamine, they're still moving as indicated by the dots being further out from the center, but they're not moving in the proper direction. And SU-5402, the dots are much closer to the middle, and also they're not moving in the proper direction. And so, we then combined all of these various lines of evidence, including gene expression, lineage analysis, and live imaging analysis, blocking the migration in culture. And then we did a bunch of genetic experiments, which I didn't get into for the sake of time to put together a model in which the idea is that the anterior sonic hedgehog expressing cells serve as a guide and attract the FGF8 expressing cells so the tooth plants itself in the proper, proper place. And what the FGF signaling does is function in a cell autonomous fashion in order to propel this forward movement. So that's why when we block sonic hedgehog signaling, the cells don't really know where to go. And when we block FGF signaling, the cells get stuck stuck in place because this cell autonomous FGF signaling is required to propel the cells forward. Now another important point that I want to make, which is a very classical experiment about the early events that happen in dental epithelial development, is that when— this is work that was done by Mina Mina and Ed Kohler about 30 years ago— that the early information for odontogenesis relies in the epithelium. And the way that they showed this is by recombining early dental epithelium and mesenchyme. And when they recombined the epithelium before about day 11.5 or 12 with non-dental mesenchyme, this was able to induce the formation of teeth. However, when they combined the epithelium at later stages with non-dental mesenchyme, this was no longer able to do so. So again, these are a very large series of experiments, but the take-home point from them was that the early information for tooth development resides in the epithelium, and then later on it switches to the mesenchyme. So now moving on from this initial initiation and placode formation stages, we'll move on to the bud and cap stages, which is when a lot of the morphogenesis happens. And as I mentioned, the enamel knot, this organizing center, is a very rich source of many growth factors, including some of the ones shown here, such as Wnts and FGFs and hedgehogs, as well as BMPs and others. And the other important point I want to convey with this cartoon is that there is intensive epithelial-mesenchymal signaling happening. So FGFs are signaling both from the epithelium to the mesenchyme, as well as from the mesenchyme back up into the epithelium. And I'll I'll give you some more examples of this in a little while. And also, there are a large number of transcription factors again, which some of which Agnès will discuss in terms of their role in human tooth development. So, the next topic I want to talk about is this bud to cap transition and how a signaling center is formed. And I also want to mention that the development of the molar and the incisor in the mouse are slightly different because the incisor becomes this ever-growing tooth. So the initial events are quite similar, but then around the formation of this signaling center, there's a rotation event in the incisor where it begins to grow parallel to the long axis of the jaw. And then rather than having roots forming as you see in the molar, in the incisor there are these structures which will house the adult stem cells that we'll talk about in a minute. This project that I'm going to talk about next was done by Chenying Li, a former postdoc in the lab, and Jimmy Hu, who's a current postdoc and about to leave to start his own lab at UCLA. And it was to answer this question of what regulates the formation of this enamel knot signaling structure in the incisor. And it was based on an observation that Chenying made, which is that when a protein called alpha-catenin, which is an important cell adhesion molecule, is knocked out, that leads to an arrest of tooth development. And so you can see here, this is in the control, in red is alpha-catenin staining, and when we knock it out, already by E13.5, there is a smaller tooth germ. And now in histological sections, you can see that by embryonic day 14.5 in the alpha-catenin knockout, the tooth is essentially gone. This structure here is called the vestibular lamina, and this is what's going to form the space between the cheek and the jaw. And this is showing you the same thing here just in 3D reconstructions that you can see there's an initial invagination here in the control and in the alpha-catenin knockout, but this doesn't grow and by 14.5 it's regressing. So, the important observation that Chen Ying made is that what alpha-catenin is doing and the reason that the tooth goes away in the alpha-catenin mutant is regulating the formation of the enamel knot.. And what she observed was that compared to the control in which the epithelium is very proliferative, but there's a small region here— in green I should mention is Ki-67 staining to mark proliferation— there's a small region here which stops proliferating, and this is the region which is going to form the enamel knot. In the alpha-catenin knockout, there is continued proliferation in this region, and that's quantified here.. Now, why is this important? Well, it turns out that a critical step in formation of this enamel knot is for the cells to stop proliferating and for them to turn on p21, which is a cell cycle inhibitor. And so you can see that happens in the control, but it's not happening in this alpha-catenin knockout. And I'll just mention as an aside that we also showed, but I won't get into this in detail, that this is a generalized principle that happens in many other organizing centers in the embryo. For example, in the apical ectodermal ridge of the developing limb bud. So coincident with this exit from the cell cycle is that the enamelot turns on many of these signaling factors. So for example, sonic hedgehog and BMP4 are turned on in the control, but this is not happening in the alpha-catenin knockout. So why might this be? Well, there is a body of evidence that shows that one of the functions of alpha-ecatinin, in addition to regulating cell adhesion, is also to regulate a pathway known as the Hippo signaling pathway. One of the important roles of the Hippo pathway is to regulate the localization of two factors called YAP and TAZ, which are transcriptional co-effectors. These will go into the nucleus and complex with these transcription factors and then turn on target genes that often will lead to proliferation. So, what alpha-catenin will do is keep YAP and TAZ sequestered in the cytoplasm to prevent them from going into the nucleus and triggering proliferation. So, then Chunying and Jimmy looked at what happens to YAP, and this is YAP staining in red. And what you can see is that compared to the control in which there is a relative decrease in this red staining in the forming enamel knot region, YAP is maintained in the nuclei of the alpha-catenin knockouts. And that's shown here with the overlay where you can see in green is YAP and in red is DAPI to stain the nuclei. So there's relatively little green here in the control and still a fair amount of YAP staining in the nuclei of the alpha-catenin knockouts. Alpha-catenin knockouts. So then Chenying did what I think is the most remarkable genetic experiment that we've done in our lab to date, which is to make this quadruple knockout in which she knocked out not only alpha-catenin but also YAP and TAZ, these transcriptional regulators, with the hypothesis being that if what alpha-catenin is doing is keeping YAP and TAZ in the cytoplasm, that if she knocks out YAP and TAZ, that should then rescue the alpha-catenin mutant. Indeed, that's what happened. So, you can see in red here, there is no more alpha-catenin, but because she's also knocked out YAP and TAZ, she has p21 coming on as well as Sonic Hedgehog, the signaling factor that marks the enamel knot. And here, if we look a day later, what we see is that alpha-catenin still has an important role in cell adhesion. So, there's a bunch of rips and holes in the tissue, but yet because we've removed YAP and TAZ,, the enamel knot can still form and the structure can still undergo morphogenesis. I think this is a very dramatic genetic separation of these two functions of alpha-catenin into regulating the signaling as well as the cell adhesion. So, another example of this bud to cap transition is this is work that I did when I was a postdoctoral fellow at UCSF in Gail Martin's lab. Is on the role of sprouty genes, which are negative regulators of growth factor signaling. And what we saw was that compared to the wild-type mouse in which there are these 3 molars, when we knock out one of these negative growth factor signaling pathways, we get the supernumerary tooth which forms anteriorly to the first molar. And it turns out that again, this is because of this regulation of epithelial-mesenchymal signaling. Where what these negative regulators of signaling are doing is repressing epithelial mesenchymal FGF signaling. And so when we knock them out, what happens is that rather than having these little buds that form in this region normally in the embryo but will regress, when we knock out the negative regulators of growth factor signaling, that provides sufficient levels of signal to the buds, and they can then go on and form these supernumerary teeth. Another example of this is work that was done by Cyril Scharl, a former postdoc in the lab who's now a faculty member in Lyon, which is to look at what happens when he knocked out more of these sprouty family members. And here what he saw was when he looked at the incisors that there was actually duplication of these teeth. So rather than having one large incisor like in the wild type, there are two incisors in each dental quadrant.. And Cyril, having evolutionary training, realized that this is somewhat similar to the primates like us as well as to the rodent ancestors in which there are two smaller incisors as opposed to one larger one. So perhaps one hypothesis that we put out is that the levels of growth factor signaling were important for forming this one large incisor in the rodents during evolution. For the last part, I'm going to talk about mineralization and also the stage when the cusps form, which is the bell stage. As I mentioned, one of the main differences between mammalian teeth as well as between mammalian teeth and the ancestral teeth is that we have roots and we only have two generations of teeth. This process of root formation is essentially driven by the disappearance of the epithelium after the mineral begins to form. But that doesn't happen in all mammals. So in many mammalian species, there has been the evolution of these continuously growing teeth such as elephant and walrus tusks, and rabbits, all of their teeth grow continuously. There's even some primates that have ever-growing teeth, and in our lab we work on the mouse, which has ever-growing incisors. I want to show you this short movie which explains how this ever-growing incisor works, which is that as the animal is gnawing on its food all the time, that leads to continuous abrasion of the anterior part of the tooth. And so this anterior abrasion is compensated for by growth from the posterior part, which is driven by stem cells that are present in this most proximal part of the tooth. And if we look more closely in here, what we see is that the central part of the tooth is— consists of this mesenchyme, which has the stem cells that make the odontoblasts, whereas the stem cells that make the ameloblasts, which are the cells that produce the enamel, are present in this most proximal part of the incisor. And we've spent a fair amount of time in our lab looking for markers of these epithelial stem cells, and this is just one example of this work done by Brian, a former postdoc in the lab, and Jimmy, the one I mentioned before, looking at a gene called BMI1. Which is present in this most proximal part of the tooth. And lineage tracing studies that Brian and Jimmy did, in which we obtained a Cre recombinase— inducible Cre recombinase knocked into the BMI1 locus from Mario Capecchi's lab, now crossed with this red to green reporter, showed that before we induced the Cre activity, all of the cells are red because the Cre hasn't had a chance to recombine them to green yet.. And then what you can see here is that with just a single induction of Cre activity, we produce progeny for essentially the entire lifetime of the animal. And so we have used this system now to identify a number of different markers for adult stem cells. But what I want to show you for the next few slides is how we've used this system in order to study a syndrome which is called Costello syndrome. And this is work that was done by Alice Goodwin, a former DDS PhD student in the lab. And this audience needs no introduction to the RASopathies, but just to point out that Costello syndrome is caused by gain-of-function of HRAS signaling, and we're now interested in looking at some of the other syndromes as well in terms of their craniofacial and dental findings. And Alice was fortunate to go to the Costello syndrome meetings, which were held every couple of years. She went to the one in Berkeley in 2009 and in Chicago in 2011 and was able to study a number of patients there, which was very fortunate because since these are rare syndromes, we don't see that many in our clinic. And she put together a nice description published a few years ago of the craniofacial findings, including the ones you can see on this slide with macrocephaly, bitemporal narrowing, downslanding fissures, thick lips, and a large-appearing mouth.. And also there were a number of other researchers there who looked at other systems as well. But one of the important findings that Alice had made was to realize that the teeth of these patients had not been studied previously. And so with just an initial look, she already realized that their teeth were different from their control siblings who were also present at this meeting with increased wear and these white lesions and striations. And then actually She asked the families to send us the teeth, the baby teeth after they were shed and was able to do SEM on the teeth and saw that there was decreased mineralization of these teeth. So fortunately, we were able to obtain a mouse model and we saw that indeed the mouse model of these teeth also had decreased mineral. And so then, as you can see that here both in the incisor and in the molar compared to the controls.— So then Alice looked histologically and she saw that the reason for the decreased mineral is that compared to the controls in which the ameloblasts are all nicely polarized, in the aTRAS mutants the nuclei are not all at the basal layer and that led actually to a decrease in the expression of the enamel proteins. So then Alice wanted to ask whether she could use our knowledge of the RAS signaling pathway and block either the MEK or PI3 kinase AKT arms of the pathway to determine which of these are important in amylogenesis. And so when she did this, what she observed— so this is in the control, you can see the decreased mineral in the mutant compared to the wild type. And what she observed was when she blocked PI3 kinase signaling, there was still decreased mineral as well as abnormal structure of the mineral. But when she blocked MEK signaling, this led to a complete rescue, which told us that at least in the dentition, it's really MEK signaling which is critical for the regulation of amelogenesis. So, what Alice found was that the patients have this enamel defect, which is caused by the abnormalities in the cells that I showed you, and that by inhibiting MEK signaling, she was able to rescue the enamel defect. Now, for the last couple of minutes, I'll tell you about our work on one more syndrome, which is on X-linked hypohidrotic ectodermal dysplasia. And again, I'm sure this audience needs no introduction, but just to remind you, the mutations in EDA1, which is on the X chromosome, cause the X-linked form of this syndrome, which— and this is a ligand that signals through this TNF-alpha superfamily pathway. And the condition is this classic triad of inability to sweat, missing or absent teeth, and thin or sparse hair. And there's this, I think, very beautiful description by Darwin, which I encourage— I won't read it all, but I encourage you all to read if you haven't read it, in which he really describes every aspect of the condition. And then the mouse mutants for the ectodysplasin ligand were identified in the '50s. And in 2012, there was the beginning of a clinical trial that we participated in to give recombinant ectodysplasin to patients, in this case postnatally. And this is work that was done, that was based on very important studies that were initiated by Pascal Schneider about 15 years ago in Switzerland, in which Pascal reasoned that if he were to take these Tabby mice, which are the ectodysplasin mutants, and give them during midgestation the missing protein, this is the period of odontogenesis, that this would then lead to a rescue of all the phenotypes that you see in the tooth and the skin and the hair. And indeed, this is what happened with the mouse studies, and then it was shown later in dog studies that this could actually be given postnatally. And so the initial studies, as I mentioned, that we did were using postnatal administration of the ectodysplasin ligand, These unfortunately did not work, but fortunately a doctor in Germany also named Schneider, Holm Schneider, continued with these studies. And I think this is an amazing report that was just published a few weeks ago in the New England Journal of Medicine, in which Holm was able to give this ectodysplasin to fetuses prenatally and correct many aspects of the syndrome. And to my knowledge, this is the first this is the first time that it's been possible to correct a structural birth defect with a medicine instead of with a surgery. So I think this is a very important and paradigm-shifting study. And so for my last slide, I just want to show you some work that we did kind of connecting back to the basic science. We had been working with these Tabby mice, the ectodysplasin mice in the lab, to understand some of the clinical aspects of the condition, but also we were interested then in using them for an understanding of evolution. And so this is a collaboration that we had with Jukka Jyränväl's lab in Helsinki that was done by Annie, a former student in Jukka's lab, and Kirsten, who was a postdoc in my lab, with the idea being that although we had found many mutants such as the sprouting mutants that I showed you before that somehow resembled what we hypothesized to be evolutionary transitions, we wanted to really choose a target in fossil record and see if we could actually mimic the ancestral morphologies. And so the one that we chose was this animal called Tribosphenemys minutus, in which there is this very important morphology where the two cusps here are separated by this valley, in contrast to the wild-type mouse in which the two cusps are— extant mouse— in which these are connected by a crest here. So this animal existed about 50 million years ago, and this is the, the extant wild-type mouse. So, through a number of studies, which I won't get into in detail, including both computational and in vitro studies, we came up with the idea that taking this Tabby mouse and during midgestation antagonizing it with a sonic hedgehog blockade, similar to the cyclopamine that I showed you before, would lead to separation of the cusps and maintenance of these two cusps as individuals.. And so indeed that's what happened, and we called this the Intelligent Design Project because we wanted to intelligently design the shape of teeth. And so this is again, to our knowledge, the first time that it's been possible to design a morphology of an organ based on what we know about signaling molecules. And what we're interested in doing now going forward is looking for other changes that we see in the fossil record and seeing if we can, by tinkering with various signaling pathways, also achieve additional morphologies that were existing in ancestors. So with that, I'll summarize that what we were able to do with these studies was to show that these stepwise transitions that we see in the fossil record are reproducible experimentally. And again, this very important point is that this XLHED represents the first structural birth defect syndrome with a medical treatment. And one thing that we're interested in looking at in the future is how we can study these parallels between between human disease and what we see evolutionarily. And so I'll stop there. I mentioned all the folks in the lab who did the work as I was going through, and I particularly want to acknowledge a couple of our close collaborators. Yuka, with whom we do a lot of the tooth work. Gail Martin, who was my mentor and in whose lab I started this. And Fred de Sauvage, a colleague at Genentech with whom we work a lot. As well as our funding agencies, and thank you for your attention.