 Okay, so I'm going to tell you about work that's been done by three of us from the University of Colorado. I was at the School of Medicine along with Trevor Williams and David Cluthier who were in the Department of Dentistry. I was in Pediatrics and about two years ago I moved up to University of Colorado Boulder. But this has been a project that's really, I'm going to be telling you a couple of little stories and we've kind of fallen more on the side of wanting to go with our interests in embryonic development and then find new genes and identify how those genes act to regulate, in this case, neural crest development and neural tube closure. And so it's a bit different, we're kind of putting on our science hat and just trying to understand how these processes are evolved. And Trevor Williams was more interested in craniofacial development, David Cluthier in heart development in their neural crest drive and I'm interested in the neural tube and early neural progenitor cells. So I'm going to tell you just a few little stories. I'm going to start with a story from Trevor's lab. This is a published story that he did along with Eric Van Adelue in his lab. And originally, again, we were all interested in these embryonic processes and so I was doing forward genetic screens with enumetogenesis and that's how we originally identified this line. But what was really helpful was to be able to get one of the comp knockout lines and be able to use that as a conditional allele in order to try to better understand how this gene MIMO1 is acting. And so now using the comp allele for this, we can see that there's craniofacial defects and I don't really see a pointer, but there's cleft palate and then there's also some head defects that occur. And so if you do skeletal staining on that, you can see that there's problems in the cranial base here and then also you can see the cleft palate and how there's problems with the skull forming in this region. And so then Trevor's lab went on and did RNA-seq and that was really informative in indicating that actually the cells do progress and they start to undergo contogenesis but they don't make the transition to forming bone. And so they dissected these regions and did RNA-seq from that and what you can see is that there's a large number of genes that are reduced in expression that are reflective of kind of this later osteoblast and E.C.M. degradation as well as then an increase in genes that are more similar to being actually at a contogenic stage. So again, this kind of stall between the two. And then using the laxie knock-in to be able to look at where the gene's expressed, you can see that it's expressed in all of these craniofacial bones including in the region that's most disrupted. And so again, the comp alleles allowed them to be able to do conditional mutants and so they used a number of different drivers to knock out the function of MIMO in specific tissues. And so when they used neural crests derived, went when Cree to knock out MIMO function in the neural crest cells, you can see that again there's this problem within these bones indicating that it's a cell-autonomous defect to the neural crest derived structures. And they went on to study a number of other Cree drivers to try to delve more deeply into this and so they used ones that were for the contioblast stage or the osteoblast stage and as well as one that's more expressed within the vasculature. And perhaps surprisingly, you can see that the one the Cree that drives that would knock out MIMO expression within the vasculature has the strongest defect indicating that MIMO function is required in the vasculature in order to allow these cells to ossify. And then another aspect that Trevor's lab did was they also changed the genetic background. So they wanted to see how much genetic background would alter the phenotype. And so when they took this from originally the C57 Black 6 background and then crossed it on to 129 SVJA background, they found that they had caused an earlier embryonic lethality and again kind of highlights the problem that's occurring within the vascular system and the role of MIMO in vascular function. And I just want to highlight Eric Van Otterloo was trained as a postdoc in Trevor's lab and did a lot of this work with the compilils and he's now taken this forward to his own lab at the University of Iowa including some looking now at some of the roles of MIMO and other tissues including in this case the teeth. So I then kind of wanted to transition just a little bit and tell you about some of the challenges that we've also found, all three of us have found in getting the knockout lines and trying to study those further. So for all of us we found that sometimes we'll get the het lines in and we are trying to generate het animals to be able to look at the homozygous phenotype. Oftentimes they'll be poor breeders and so you know you kind of really struggle in order to be able to get the line going and be able to do the characterization that we want to do. Another thing that has been highlighted from the comp2 projects as well is that there can be incomplete penetrance and incomplete expressivity. So for instance you know I've become, when you're working on a gene that you have no idea what its function may be you're trying to understand the role in these particular tissues. If you only have let's say in the case of a neural tube defect, if you only have like a 20% or a 30% penetrant phenotype it makes it really pretty difficult to go from an unknown gene and to try to understand the function of that protein. And so that's also I would say probably made us a little bit hesitant with some of the lines have really nice phenotypes but again it can be at such a low penetrance that it just makes it hard for us to study. And then in particular for David and Trevor for a lot of these things to be able to do bone and cartilage staining you need to have the embryos grow long enough at least to eat 14.5 or 16.5 and oftentimes they die prior to that. And so again it makes it difficult to characterize the phenotype when they're failing to survive. So I just kind of highlighted, I'm going to be telling you about the stories that we've been doing in just a minute about looking at neural tube defects lines but this set of lines here ultimately we imported but we just because of the challenges that I just talked about we're really never able to kind of pursue those further. And so we don't really understand their function in neural tube closure. And then there's just a set of lines that Trevor Williams lab is still working on but again it's been slow in kind of getting to the point of being able to characterize it in terms of their craniofacial and other neural crest phenotypes and so I don't have any data to show on that. I wanted to show a little bit just to again highlight some of the struggles that David Cluthier's lab has had as well in terms of this incomplete expressivity and so this was just a line that he had imported, this LGL4. And so it had been originally characterized to have a short and lower jaw and when they continued to study that using both laxie expression as well as then characterizing the phenotype of the knockout embryos they weren't able to recapitulate at least at any high enough level those phenotypes in the craniofacial structures are in the heart and so he discontinued that line. And here's another one that David's lab looked at ATP11A and it looks really quite nice but again because of kind of the low penetrance of the phenotype again made it a bit of a struggle to study so by laxie expression it's expressed in the heart nicely as well as in the ventricular septum and the work by the DMDD group had done some really nice HRM analysis and had found that there was this ventricle septal defect and thinning of the heart myocardium and so he was really interested in this. He got the allele in order to be able to make a conditional and then be able to knock it out and so he knocked it out with NKX2.5 which is a CRE that will then drive expression in the myocardium in the secondary heart field and in that case they saw a hypertrophy of the heart muscle but again at a relatively low penetrance that made it difficult to kind of continue with that and again he was interested in the neural crest derived aspects of this and was thinking that this could be responsible for the ventricle septal defect and so he knocked it out again with WITWIN CRE to knock it out in the neural crest but he failed to see a phenotype in those embryos. So I wanted to move on and tell you about some of the work that we've done to try to understand the process of neural tube defects. So I would say first of all it's really obvious if there is a neural tube defect it's a really obvious phenotype that heads kind of blown open or they have spina bifida and so I'll tell you three little stories about animals that we've been working on to try to characterize the function of genes that hadn't been previously identified to be involved in neural tube closure. And so SNCC3 came from Steve Murray and Jackson Labs and so they had some really nice micro CT which was not shown here but showing the exencephaly or the cranial neural tube defect. If we trace that back we can see even at like E8.25 we can already see defects in these embryos. And SNCC3 is a sorting nexone gene. It falls within a family but it's really quite different than all the other ones and it's thought to be involved in the early endosome. So really we didn't know much about what SNCC3 did not certainly not in mammals but in flies and in worms it had been found to regulate the trafficking of wintless which is a receptor that binds the wint ligand and helps to so SNCC3 is thought to bring that wint receptor back into the cell so that it can then go back into the ER or the Golgi pickup wint again and be able to secrete that. And so that was Heather Brown's first hypothesis that it's involved in recycling of wintless so that you can get this cycle and increase wint signaling which we know is really important in neural tube closure. And so that's indeed what she's found. She's found that it's involved in both canonical and non-canonical wint signaling. So this is just looking at the expression of some genes that are markers of wint signaling so left one here you can see that it's greatly reduced in these mutant embryos the SNCC3 knockout embryos. And you can also recognize that these little mutant embryos are short in squat which is reflection of another problem with wint signaling that's required for convergent extension to allow the cells to kind of integrate to extend the embryonic axis and bring the neural folds close to one another. And so these have a convergent extension defect. And so we always try to kind of move back and forth between the mouse embryos and in vitro systems if we can. And so since the idea was that maybe SNCC3 is binding wintless and recycling that from the membrane back to the early endosome and back to pick up more wint and continue the strong signaling essentially without going into the data. That's what she found is that in wild type SNCC3 and wintless co-localize within the early endosome but when there's a mutation in SNCC3 this endocytic recycling is disrupted and instead the wintless goes to the lysosome and gets degraded and she can rescue that phenotype. So another angle that we've taken all three little stories that I'll tell you about is that we also are interested in how we can take what we're learning in the mouse embryo and be able to try to understand human development. And so we've established a nice relationship with Hope Northrup and Paul Awe at the University of Texas Health Sciences Center at Houston where they've sequenced a number of patients that have spina bifida or myeloma ninja seal and done whole exome sequencing. And so in this case they found a mutation in SNCC3. It ends up that the mouse SNCC3 and the human SNCC3 are 100% conserved across the whole domain. And so we can just create this patient-specific point mutation, put it back into our in vitro assays to test the functionality. It's one of the big problems that we find with trying to understand any of the human genetic data is how causative is that point mutation that you may find in the actual phenotype that we see. So we're going to model this one. Okay, so moving on to a different mutant that we got from Baylor University and Mary Dickinson's group is a mutation in TMEM-132A. So these animals, this is the micro-CTs work that came from the Baylor group. And so you can see that they have spina bifida, which is here in this caudal part of the neural tube. They have heart defects and they also have limb defects. So you start out again with TMEM-132A. All it stands for is transmembrane protein-132A. And so where do you go from there? And so what Binbin Lee in my lab did was mass spec analysis to try to identify interacting proteins with this transmembrane is a single-pass transmembrane protein to try to understand how it's working. And surprisingly, one of the biggest hits we had took us back to Wintlas, which takes us back to that SNCC3 story that I just told you about. And so the three major pathways that he found again relate to Wint signaling and to cell adhesion. So he found that it interacts strongly with Wintlas, also beta-catenin, which is part of the Wint signaling pathway, as well as many of the caedherans and catenins that are involved in cell adhesion. So where do we go from there? He's been doing a lot of biochemistry with this, but essentially just showing you just a few little bits of information. Again, this unknown transmembrane protein positively regulates the secretion of Wint, and I'll come back to give you a little model in a minute, but essentially in the homozygous cells, there's less Wint that gets secreted into the media, and we can do co-culture to see whether Wint signaling is occurring, and that's disrupted, which is also shown by disruption of genes that are downstream of Wint. And the other piece of data that he was just showing me yesterday, so I don't have to show you, but he's also found that this transmembrane protein also perhaps is like a chaperone or somehow in complex, really greatly increases the interaction between Wintlas and the Wint ligand. And so this transmembrane protein perhaps is this kind of trimolecular structure increases those interactions. So to kind of circle back between the SNCC3 and the WintLis and the transmembrane TMEM132A, we think that the transmembrane protein is expressed here in the Golgi, where it interacts with WintLis to pick up Wint in order to secrete this through vesicles out to the surface of the cell. And then the SNCC3 acts to bring the WintLis receptor back into the cell in order to be able to recycle this and allow signaling to continue. So just to give you a little... so it's complicated, you know, all of these genes and proteins probably act in multiple pathways. So our first kind of favorite hypothesis, there's some really kind of interesting work that's come out that really started from working with embryonic... or IPS cells and trying to generate neural tissue and found that much of the spinal cord is generated from a kind of a dual progenitor population that are called neuromesodermal progenitors. And depending on the level of FGF and Wint signaling, these cells can either pick a mesodermal fate or pick a neural fate. And we had done RNA-seq analysis, and again it highlighted that there was problems with Wint signaling, and it also showed that there was less mesodermal markers that were expressed. So we really thought that we might be sitting at this border and that this might be a regulator of this transmembrane protein, or a regulator of this bipotential choice between neural or mesoderm. But that's probably part of the story, but I think that's probably a subset of the story. And so we're still kind of trying to figure this one out, but we also do see that CDX, which is another important regulator downstream of this pathway that's needed to keep the tail growing and to give... and to your posterior identity, that we see these CDX genes get dysregulated in the mutant. So we're kind of getting close to kind of finishing off a small story on that, but now we've kind of gone back to thinking that a lot of the spina bifida may also be involved through changes in cell adhesion and this transmembrane protein in regulating the adhesion of these cells within the spinal region, and if that doesn't occur properly, a spina bifida. And again, taking this to the human side, there's multiple mutations within this transmembrane protein in humans with myeloma ninja seal, or spina bifida. So again, we have these nice in vitro assays that we can create these patient-specific mutations and then look at the functionality of this transmembrane protein with these functional mutations. And I'll just end up on one last story because we're also interested in neuroprogenitor proliferation and differentiation, and if that balance gets disrupted, this can give rise to microcephaly or where the brain is too small. And so this is another allele that came from another gene knockout that came from the Baylor group. So this is Act R5. So you can see obviously the mutant embryos are much smaller than the wild type here at E9.5. And they have clearly major head problems in the neuroprogenitor cells within the head. So again, we don't know much about what Act R5 does. There's work in yeast that indicates that it can do a number of things also through the I-080 complex. But some of the things it's involved in is a regulation of DNA damage and DNA repair and chromatin remodeling. And so we only imported this line not too long ago. So again, we try to play back and forth between cell culture work and the embryo in order to better inform both studies. And so we've been doing some cell lines while we're waiting for the line to get up so that we can start to analyze the embryos. We've been doing some cell line work using neuroprogenitor cells and do you find that indeed the Sector V in mammalian cells is involved in DNA repair and DNA damage response? So we see when we knock it down with S-I-R-N-A's, we see that there's an increase in phospho-H2A-X, which is a marker of DNA damage. There's an increase in P53, another marker of DNA damage, and there's activation of caspase. We also see that there's changes in how the chromosomes segregate and cell survival in these cells. And so we're really looking forward to having some tissue from the embryo because we think that's going to be a large part of the problem and why we get microcephaly and loss of these neuroprogenitor cells. And this has been done by Louisa Gayes-Omi Miranda in my lab. And again, there's mutations, multiple mutations in Act R5 that are also found in the human patient samples. So again, we've got the cell-based assays that we can test functionality. And so with that, I'd really like to thank the COMP2 project. It's been a lot of fun. In particular, I want to thank Mary Dickinson and the whole Baylor group. We've gotten a number of reagents from them. Also, Steve Murray and Jackson Labs and UC Davis has also been really helpful in sending us animals. Thanks. The memo one story is really cool. So the tytoocree should knock out the vasculature as well as some hematopoietic cells. So is it... Do you know anything more about whether they've teased that apart because... I don't think so. I don't know if Eric is going to continue with that in his group. Yeah. Yeah, there are some good ones that you can use. Like the flick one. Well, there's a runks one and there's another. That's escaping me, but... Yeah, I don't know. The other question that I have that I would love to get with you on is when you talk about the penetrance things, how are these reflected in the numbers? So we have, you know, obviously we see variable expressivity when we first take the lines. And I'm curious to see if they're really statistically different. Now we have a small capture, but if there's really a lot of variation compared to when they go out to you or not... I don't think it's probably that big. I mean, you guys have definitely seen that variability as well. I think at the very beginning, we were just like, we were so anxious to get reagents in. It's like, we don't care, we'll take it. And then when you really start to delve into it, again with a novel gene, you realize it's just really tough to do it if it's a lower penetrance. So I must say I've been more hesitant for getting neural tube defect lines if they're like, you know, 20% penetrant, because I just know that it's going to have to just be so much reading to do. And really you want to look at the phenotype before it presents, but then you don't know which is going to be the ones that might have the phenotype. We just want to make sure that we have the data out there so you can make those choices. I mean, as you say, the number that you analyzes relatively small, so obviously we have a lot more numbers, but I would say it's not really that different. I mean, we did know that there would be incomplete penetrance. Okay. Can you use the mic, Rob? Please. It'd be great to know if there's a heterozygous adult phenotype for the TM-132A. It's a very good candidate gene for total brain weight in mice. That's one of the highest QTLs. So is there any kind of adult phenotypes for the hetes? You know, I'd have to go back and look. I mean, because there was some really nice het screening that we're done with all these lines first. At least it's not obvious to us. I mean, we're dependent on having viable fertile het animals in order to be able to continue with our homozygous screen. And so I don't remember for any of these that they have a het phenotype. The only one that we've seen, I did get an addendum to add on to the comp, too, from the I forget which group, but the dietary supplement group, because we've been doing a lot of work on folic acid. And so we've found that actually with the SNCC3, we can see then a het phenotype in embryos based on the level of folic acid that we add. But overall, I think the animals are pretty fine as hetes. Which is not to say there's not a brain phenotype, but they can eat and they can breathe. We're not putting them through any other behavioral tests than that.