Joyce Bischoff, PhD - Endothelial Abnormalities in Vascular Tumors and Vascular Malformations

Speak. But um today we have the pleasure of welcoming Professor Joyce Ellen Bischoff uh to speak to us. She's a native of Detroit, Michigan, but she's been in Boston and in the Harvard community and had children since the early 1990s. She's a full professor since 2010, and her research has been continuously funded by the NIH and other entities since the early 191990s as well. Her laboratory is a part of the vascular biology program and studies focus on the role of endothelial colony. Uh, forming cells and endothelioplasticity and diseases and repair of the vasculature. She's the editor in chief of the Journal Angiogenesis, and she's served as a lab mentor for many undergraduate and graduate students, many countless national and international presentations, and over 100 high quality publications. Today she will speak to us on the topic of endothelial abnormalities and vascular tumors and vascular malformations. Please welcome her. Thank you. OK. Well, good morning, everyone, and thank you for the invitation to present our work this morning. I'm gonna talk, as Ben mentioned about endothelial uh defects and abnormalities in vascular tumors and vascular malformations. So, um, first of all, I just want to show a few pictures of these. I, I, I know it's very familiar to all of you, but one topic we've worked on extensively is infantile hemangioma. Here's a particularly devastating hemangioma. There's also a very rare case. This is a single patient that John Mulligan took care of in the uh late 19 almost 20 years ago now. This child has multiple tumors that were first thought to be hemangioma but later shown to be distinct, and there's hemangio endotheliomas. I'm not gonna present any data on this, but it's a project now we've really revved up in the era of Easy or not easy, but readily uh accessible uh sequencing. So we're trying to figure out what the genetic basis of this. And then also capillary malformations and Sturge-Weber syndrome, and here's a characteristic port wine stain that's um puts the patient at this man did have Sturge-Weber syndrome and puts him at high risk. And I wanna just say um, at the outset that none of this, uh, all of our work has really been a collaboration with Doctor John Mulligan, who I'm sure all of you know in plastic surgery. He takes care of these patients, you know, helping to restore function and, and, um, Tissue integrity and, uh, and when he removes the, the tissue for clinical reasons, we're able to bring it to the laboratory and then isolate cells and do all kinds of basic science research. So, we'll start with the vascular malformations. Over the last several years, there's been a huge progress made in identifying uh genetic um mutations, and these are all in the ones, in some cases, these are germline, but all of these are, have been shown to be somatic mutations in endothelial cells. So, Um, uh, a mutation occurs sometime during development in a single cell that gives rise to endothelial cells. These cells can proliferate and migrate and form the lesion, but it's important to No, this isn't, these aren't germline, they're somatic mutations. So for venous malformations, they're driven by mutations in type 2 or PI3 kinase, the catalytic uh uh domain. Lymphatic malformations are also driven by PIC 3CA. Capillary malformations were recently shown to be due to a a very specific point mutation in GNAQ, which is a G protein. And then more recently, our AVMs, uh, there are two different mutations. Aaron Green's lab showed that Uh, endothelial mutations in the endothelium and MAP 2K1 can cause extracranial uh AVMs, and a recent paper in New England Journal showed that KRAS mutations can cause brain AVMs. So there's been huge progress here and it's really pretty exciting that these are all in the endothelial cells since we're very endothelial centric. Um, so the vascular tumors, it's a little not so clear. There's, for a long time, people have postulated, we ourselves included, that infantile hemangioma is caused by a somatic mutation, and we're actively sequencing, um, working in collaboration with Mika Voulis. But we don't have, we have some variants and we're, we're looking at those, but this is still an open question is whether there is a genetic basis to infantile hemangioma and also the, the really rare and aggressive hemangio endothelioma. We're working on that, but no, no genes are identified yet. So the, the outline of my talk, I'm gonna tell you a little bit of, a little bit of, sort of a summary of our work on infantile hemangioma. Then one story about mutant type 2-driven venous malformations. Um, and then the third is a really new project, relatively new project in our lab about the GNAQ-driven capillary malformations. So infantile hemangioma, I know all of you are aware of this tumor. There, there's a wide range in severity. Many of, they occur, they're quite common in 4 to 5% of newborns, and fortunately, most are just very small lesions and In a location that's not going to cause any disfigurement, but again, there are many that are devastating. These are both children at 6 months of age, so they appear 1 to 2 weeks after birth. They're treated now with propranolol, primarily, but older treatments were steroids, and, and sometimes, as I mentioned, they're removed if that's the best course, and then we, that's when we can work on them in the laboratory. A hallmark of hemangiomas is that they will all involute by 5 to 8 years of age. So it's quite an interesting So hemangiomas, they display this vivid, they're a vivid display of endothelial maturation over the life cycle of hemangioma, which is quite interesting. In the proliferating phase, it's highly cellular. There are vessels here. If you stain this with the endothelial markers, there's tons of endothelial cells that are in the process of forming vessels. And then when you get to the involuted phase, now you have well-formed vessels. These are all lined with endothelium surrounded by parasites. And then in the involuted phase, you have what's the vessels have regressed, what you're left with a fibro fatty residuum, a lot of fat, and just a few remaining vessels. So, so there's really this uh vasculargenesis going on and then regression of vessels. So we've isolated again by getting tissue from the OR from patients we've isolated stem cells, endothelial cells, and parasites from hemangioma, and that's shown just depicted here. We pulled out the stem cells using a marker, a stem cell marker CD 133. These cells are highly proliferative. They can differentiate into endothelial cells, parasites and adipocytes in vitro and in vivo in nude mice. They make a lot of vegFA. The endothelial cells are, are pretty normal, but they, they do show um high VEGF receptor 2 phosphorylation, E selectin, angiopotent 2, jagged 1, all signs of kind of endothelial activation. And the parasites, which we've also isolated, are, are proliferative when you compare them to normal human parasites. They have low angiopoietin 1, low contractility, and high VEGF. So, all of these cells are abnormal, and we're sequence and Um, OK, so just to summarize a lot of our studies, we've taken these hemangioma stem cells which we call hemaCs, and as I mentioned, they, we've determined or shown that they need VEGF receptor 1 to differentiate into glut1 positive endothelial cells, and this is important because glut1 is a really a hallmark of hemangioma. It's how you can distinguish hemangioma from all other vascular tumors and vascular malformations. And these stem cells, they use a notched jagged one pathway to differentiate into parasites, and they can also differentiate into fat, and we haven't really looked at any molecular players there. And then we've, uh, over many years looked at um how drugs that can affect the stem cells, and it turns out corticosteroids suppress the VEGFA being made by the stem cells, and that blocks vessel formation. Um, we did a library, a chemical library screen and found that rapamycin could slow down this high proliferative and self-renewal capability of the stem cells. And then more recently, and those are all, um, this has all been shown in a whole series of papers. And then more recently we've been looking at propranolol. Um, in collaboration with Mathias Francois in Australia, and he's a SOX 18 expert. And we have, so first of all, we've shown very nicely that propranolol blocks the stem cell to endothelial differentiation, and it may be working through a SOX 18, which is a transcription factor, um, well known to be involved in uh vascular development and also lymphatic development. And some, I'm gonna show you some of this is really new data um carried out by my research technician, Jill Wiley Sears. And so here, what we're doing, this is uh uh measuring RNA levels by QPCR and we induce the differentiation of the stem cells with VEGFB. You can see that here, a number of markers, endothelial markers, and, and actually these are hemangioma endothelial markers. They all go up when you treat with VEGFB and they're all suppressed by propranolol. And they're also suppressed by this SM4, which is a SOX 18 inhibitor. It disrupts the dimer formation of the transcription factor. So we're seeing very similar effects between propranolol and SM4, which is interesting. And then we expanded this study, that was just with one hemangioma cell stem cell line and 11 experiment. So we expanded this. Now, Jill did a fantastic job using different stem cells from different patients, and then she, um, and that's what show each, each, uh, Um, circle is a different, uh, data point. So, and, again, here's our control, um, CD31, which is a typical and a very well-used, um, endothelial marker. It's in the control, it's low, goes up when you treat with VEGFB. If we treat with aspirin, which is similar in backbone to the SM4, there's really no effect. The SOX 18 inhibitor inhibits very nicely, propranolol inhibits very nicely, and then propranolol is a mixture of R two antamers, R, R+ and. And the beta-blocker activity is, is really associated with the form. And the R+ form is like a hundredfold less effective in terms of blocking um beta adrenergic receptors. So it's very interesting that the antamer without beta-blocker activity is showing a nice result in this assay. And then we went on and repeated and also analyzed notch one, which we know is high. June Wu's work showed that this is very high in hemangioma endothelial cells, plexin D1. We know from our collaboration with Mike Klagman's lab it's high in hemangioma. Endothelial cells and also VEGFRUL1. And you see the same pattern in all of these. We see nice induction with VEGFB, no effect of aspirin. The um the SOX 18 inhibitor is working well. Propranolol is giving a nice effect, and then it's the R of antimer, which is not a beta blocker that's blocking this endothelial differentiation. So we're pursuing this quite, quite a bit. Um, so future questions are sort of ongoing questions. As I showed you, we, we have all these stem cells, parasites, we've isolated the glut1 positive endothelial cells and glut1 negative endothelial cells, and we're sequencing these and, and working with Mika Vukula's lab to see is there a genetic component to infantile hemangioma. Um, And then the second ongoing question is the propranolol mechanism of action. We're really looking at the SOX 18, if that's a player, and whether you really need the beta adrenergic receptor. Is it acting as a beta blocker or by some other mechanism? And then we're also looking at, we know very that rapamycin targets the hemangioma stem cells very nicely. It targets their self-renewal and their high proliferative capacity. And we've long hypothesized that maybe combining rapamycin with propranolol would really augment the propranolol and um shorten the duration of treatment time. OK, so now I'm gonna, that's our man Joma story. Now I'm gonna tell you about a project on Taitu-driven. Venous malformations. And this is a project done in collaboration with Mika Voakkulis Laboratory in Belgium. So Ti2 is a tyrosine kinase receptor. It's shown schematically here. And um there are, there's a, the, the most uh 50% of um sporadic venous malformations are caused by this mutation L914F, which is in the, is in the catalytic domain of the tyrosine kinase right here. So, Um, and what this does is it activates the kinase activity. So now the receptor is constitutively phosphorylated and, um, and active, and this is a germline mutation that causes venous malformations. It's much weaker and requires a second hit. These are some examples of patients. And in Mica's lab, in his previous publications, what they showed is that this mutation, a downstream consequence of this is that there's high phospho AKT which will become important in a little bit. And there's also low PDGF receptor B and PDGF receptor B is important for attracting parasites to the vessels. So what we didn't, so we were talking with Mika's lab a long time about this, and we said, well, maybe we can try to put the mutant cells into our in vivo model. And this is the, the human blood vessel formation model that we, we've used in my lab. It was developed by Juan Millero Martin. Um, and it's really quite handy. Um, OK. So what we do is, is, um, we're using the mouse as a recipient, as a host, and we use immune deficient mice, and then we put in our human cells, endothelial cells, and any kind you want. So here we have endothelial cells. Actually, it turns out for normal vessels, you need to add a smooth muscle or a parasite or a smooth muscle precursor. We mix the two cells together in an extracellular matrix, so we have our single-cell suspension, and we use Matrogel, which will gel at 37 degrees when you implant it into a mouse. And about 7 days later, you have a nice vascularized implant. So this is kind of an ectopic vascular network. And this just schematically shows what's going on. They go in as single cells. We know that they start to form sort of proto tubes that are not connected. And then by day 7, they're connected with the host circulation, and we can see that by labeling with, um, I don't know if this is hard to see. Uh, OK, um. Yeah, so you can do a tail vein injection with these fluorescent lectins, one that's human-specific and one that's mouse, and you can see the human vessels and you can even see connections between the mouse circulation. So this is a really handy in vivo model and so what we use to do basically all of our studies, this is our in vivo approach. So, Alisa Boscola, we got these cells from Mika Vukula's lab that are expressing the wild type T2 or the mutant T2, and you can see the really dramatic um constitutive phosphorylation of T2. And Alisa, uh, these are the, the genes were put into Huvex, human umbilical vein endothelial cells, which is sort of the workhorse, uh, endothelial cell in the vascular research community. So she implanted the wild type cells on one side of the mouth and the mutant cells on the other, and Right away, you could see a difference. I hope you can see this is kind of purple and this is not. And when we opened them up, or Lisa opened them up, you could see that the wild type, there's really nothing happening cause we actually didn't add smooth muscle cells or parasites. So they're just nothing and then we get these really red vascular lesions, and here are the eggplants from all the mice and there's clearly a difference. So this is very super exciting for us. Um, and really it was, uh, so, and then by histology, if you look at the wild type, the, the cells are there, they're stained for a human endothelial marker. They're there, they're just not forming any vessels. The mutant forms these huge channels. They're lined with human endothelial cells because they have this Eulex Europus marker. And they're very similar to patient venous malformations shown here. I don't know this thing is not working. Um, Yeah, so they're very similar, these large channels, and they're clearly distinct from infantile hemangioma. And by uh Doppler ultrasound, these have a non-pulsatile flow, so they're venous-like, and by many measures, they're venous. So, at least I looked at the growth of these, they, they expand over time versus the wild type cell implants don't do anything and in fact become absorbed over time. This is the weight of the animals, and again, you just see this amazingly, I mean, stark difference. So clearly the, the, the um type 2 mutation is causing these vascular lesions. Um, and so, Alisa set out to do a number of experiments. The first was to we started testing drugs, of course, and the two that we tested are rapamycin, and then a, a commercially available, it's not a, uh, inhibitor of the type 2 tyrosine kinase domain, so it's a tyrosine kinase inhibitor. And so what we did, or what Lisa did was to pre-treat the cells for 48 hours in vitro in a dish. So the animal's never gonna see the drug. It's just pre-treatment of the cells, and then she implants the cells and watches what happens. And the control, the, the lesions are formed and they get bigger, um. The T2 TKI had a modest effect on the size, but rapamycin had a better effect, and we can see this when we measure vascular volume. We can see this when you measure vascular area in the uh histological sections, and they also have more smooth muscle cells around them. So this was exciting and we asked though, why is the type 2 TKI less effective? And the reason is, is remember that the mutation is in the tyrosine kinase domain, and this is a drug targeted at the tyrosine kinase domain. There's a mutation there, so the small molecule inhibitor may not be binding as well. And that is what we looks like what's happening. So here, this, this is a Western blot for different molecules, and we're testing the cells, we're treating the cells with increasing concentrations of the type 2 TKI and it's in the presence of angiopoietin to really get high phosphorylation. So you can see that the wild type type 2 is inhibited by the drug, but the mutant isn't. And when we look downstream, remember that phospho AKT 473 and there's another uh um phosphorylation site. These are both also inhibited by the type 2 TKI but the mutant is resistant. So, in, really the phosphoT2 and phosphoAKT levels in the mutant cells are just really not, they're not as sensitive and so we think that's why it was less effective. And then when we look more closely at rapamycin, it, it's really great at reducing the fossil AKT uh levels to near wild type. And then we think that's, it actually bypasses the mutant receptor. It's not targeting that, so you don't have to worry about the altered confirmation or anything. And that's, so here we see that this is the key here. This is phospho AKT and you see in the wild type levels are dramatically increased in the mutant cells. Going from here to here. And if you treat with a TKI inhibitor, there's a modest inhibition, but there's a break here in the scale, but the um rapamycin in light gray really effectively brings the, the level of the phospho AKT 473 down to wild type levels. And when we look at another signaling pathway, the phospho-Ic pathway, there's really no effect. So, um, yeah, so we think this is the explanation, and here's shown schematically, um, well, so the type 2 TKI is working up here at the receptor, at the mutant receptor and blocking downstream signaling. Rappamycin, we all know is an MTOR inhibitor, and that's what, it inhibits MTORC1 in the acute and chronic phase. But long-term treatment, it also inhibits MTOR-C2, which is responsible. One thing it does is phosphorylate AKT at this 473 site. So the long-term rapamycin is helping to decrease these AKT levels and then block the MTOR pathway. So the next experiment we asked is can, can rapamycin, before we pre-treat it, and that would never be a relevant clinical scenario, but here we let the lesions form, so at least it implanted the mutant cells or wild type, the mutant cells, let them form for 12 days so that they're big lesions, and then treated with the drug by IP injection. And here you can see vehicle versus rapamycin treated. They do look a little smaller. And measuring their, the lesion area, the, the untreated are getting bigger. rapamycin's really plateaued or stopped the growth. And looking at vascular volume by ultrasound, you can see that the, um, at 19 days, there was a statistically significant decrease in vascular volume in the rapamycin treated. So all of this data led our colleagues in Belgium, so Mika Vakula and Laurence Spoon to start a A clinical trial, um, Laurence treated 6 patients who had failed all of their therapies and had really difficult to treat venous malformations, so she put them on, on rapamycin. And, and the results were, are shown here. So first, the D-dimer levels went down in all of the patients. The pain levels went down, their quality of life improved, and their, the MRI measurements of their lesions, although it's not super striking here, that, there was a statistically significant decrease in the um lesion size. So all of this data, we put all of it together, the, the animal. Model, the signaling in vitro experiments, and the and the uh clinical trial into one paper, which is quite remarkable, um, and published in JCI and Alisa Boscola is the first author on this paper. She's now on the faculty at Cincinnati Children's Hospital, and, um, she has her own, she got her own RO1 for this work, so I'm really proud of her, how well she's done. And, you know, so this is a project that now Elisa's taking, is going full blast on and also Mika's lab. But here I'll just, to summarize what I've shown you is that, first of all, the type 2 mutation L914F in human endothelial cells is sufficient to create venous malformations in mice. And that might seem like, oh, we know T2 is important for venous malformations, and it's true from a genetic standpoint. It was known, but type 2 is actually expressed in hematopoietic cells, sometimes been reports in parasites. Here, we really show it's causative. You can take this mutation, put it in an endothelial cell, and get pretty dramatic venous malformations in mice. So this is maybe a subtle point, but I think important and um we showed that rapamycin blocks the constitutively active phospho AKT that's downstream of type 2. It blocks expansion of these lesions in mice, and it reduced lesion size and decreased pain in 6 patients with difficult to treat venous malformations. And then just some follow-up. So in that study, 3 of the patients had known, had documented type 2 mutations. And the other 3, they didn't know what they had. But since then, Mica's lab has shown that they had, they have PIC3CA mutations and actually, which actually result in the have activated phospho AKT cause PIC 3CA is just upstream of that. And there are actually two other papers in 2016 in Science Translational Medicine, also showing in their cohorts of patients with the PCA mutations. So what's, it's really kind of satisfying that there's this type 2 tyrosine kinase, you can have mutations there, you can have mutations in the signaling molecule downstream, and you get the same result. You what happens is there's activated phospho AKT and you get a venous malformation. So, very satisfying project. Um. OK, so the last one, last little snapshot I'm gonna tell you about is capillary malformations and Sturge-Weber syndrome. And this is a, again, a relatively new project in the lab. So the capillary malformations can form in Sturge-Weber, they form on the face, and they're called port wine stains, and they can also form in the brain and the lepto meninges, and they're caused Um, can cause in some cases, strokes and seizures and cognitive delays. And here are some of the features of these. This is a, a Port wine stain, and this is from the brain. What you see are enlarged vessel lumens, and really pretty disorganized perivascular cells around here, and in the brain, there's low blood flow, which is thought could be causing some ischemia or hypoxia. And then in 2013, a group at Hopkins identified or or showed that that identified a somatic activating point mutation in GNAQ. It's a single amino acid change. And this is, they found it in 92% of the port wine stains and 88% of the brain CMs. Since then, at least two or three groups have found the same thing. So this is really remarkable. And what I'll just tell you a little bit, DNAQ encodes a heterotrimeric G protein called G alpha Q. There's lots of G proteins. This one's G alpha Q. And this is just a schematic of G-protein coupled receptors. So the heterotrimeric G proteins are here, alpha, beta, gamma subunits. In their inactive state, they have GDP bound, and they're, they're coupled to G-protein coupled receptors. So when they're activated, what happens is you get an exchange of GDP for GTP. And now this is the active state. So that R183 mutation, and the alpha subunit has inherent GTPAC activity to return to the inactive state. And that R 183Q mutation affects the GTPAC activity, sort of inhibiting it. So what you have is, is GTP bound G alpha Q which is active, so you have constitutive activation. And one of the main downstream targets is phospholipaseC beta. Um, so what we set out to do is figure out which cells, you know, you have this capillary malformation, which cells have the mutation, uh, because it could be, you know, endothelial cell is a good bet, but it could be parasite, it could be some other cell. So Lan Wang, a postdoc in the lab, did these experiments. Oh this is not working. But what we did is use self fractionation um to separate the, um Capillary malformation tissue into specific cell types. So using either port wine stain tissue or we've been able to get, we're now up to 6 different little small pieces of brain when the neurosurgeons are operating on these children because of severe seizures. And we uh land as a collagease digestion to make single cells, and then labels them for for fact sorting. And for endothelial cells, she actually throws in a pool of three different um endothelial markers because we really want to capture all the endothelial cells. Um, she used anti-PDGF receptor beta to get the parasites, and then actually also collects and sort of removes the uh hematopoietic cells using anti-CD45 and glycofluorin. A. So when, um, so we've done this on both port wine stains and on brain lesions, and the results are similar to what I'm gonna show you here for brain. And here's two different specimens. When you look at the total cell, your single-cell digest, the mutant allele allelic frequency for the GNAQ mutation, R183Q, it's a, you know, 6 to 8% of the total cells have the mutation, and it's a somatic mosaic mutation. In the end, the hematopoietic cells are really depleted. There's really no mutant cells there. The endothelial cells, however, are enriched for the mutation. So this is really exciting. The PDGF receptor beta positive cells, we really didn't get enough cells to be able to test this. And then there are some mutant cells left in the triple negative population, which is, so we don't know if this is, we didn't capture all the endothelial cells or is there another cell that's harboring the mutation. But we can clearly see, and again, we've done multiple Port Weinsteins in collaboration with Aaron Greene's lab, and we have also published that data. It's in the endothelial cells certainly have it. So this tells us where to start and where to look for um Try to understand how this mutation is causing capillary malformation. So here Lan has, here's a primary culture, you, you know, besides doing the fact sorting, she takes some of the cells and puts them in culture to start growing them out. And you can see nice little endo what looks like endothelial cells appearing. Um, and then she uses CD31 to select out the endothelial cells, and you can see that works very well. The CD31 selected positive cells are, are CD31 positive, the CD31 negative are negative, and then she rechecks the mutant allelic frequency. And again, the, the CD31 positive cells have the mutation, the negative cells don't. And what's interesting, so here are an important thing, this is the allele frequency, you know, there's one allele, so basically you double this. So these cells are 15, 15% mutant allele frequency, that means 30% of the population has the mutation. And here are 42%, and here is 64%. So, in our brain cult endothelial cultures, it's a mix of mutant and non-mutant cells, which is important just to think about when or for interpretation of data. So we, we, we're trying to get lots of different cellular models. So here, as I just showed you, we isolate endothelial cells from the lesions. We have skin, CMs, and brain. And these turn out to be a little more difficult. Um, So we have patient derived cells and then our other approach is to just like we did or with the type 2 mutation is to express the mutant GNAQCDNA in normal endothelial cells. We actually use a lentivirus. Um, the other studies were done with the retrovirus. And we have two recipients. One is endothelial colony forming cells. These are very robust, uh, young endothelial cells that we isolate from cord blood, or we also have done this with brain ECs. Um, and we have a third model that Colette Bichell is, is, has used CRISPR Cas9 to introduce the mutation on a single allele. It's actually pretty difficult to do to get the editing on one allele and no screw-ups on the other allele, but she's, she has a few, um, clones of that now. So we have a third model. Um, so one thing we, of course, look at endothelial proliferation in response to VEGF. This is kind of your first step in any kind of angiogenesis study. And what we see is that the mutant, the sturge of our brain endothelial cells with, you know, pretty decent mutinnoial frequency, they just grow more slowly, even in response to VEGF. Compared to normal brain endothelial cells. So this, it's not, it's a little bit of a conundrum because you get all these vessels in, in the malformation, but they don't, growing is not their thing. And we also Colette the the cell, who's a post-doc in the lab, she did this. The same experiment with our engineered lentivviral engineered cells where we have the wild type expressed versus the mutants. We actually even threw in an inhibitor, a specific inhibitor of G alpha Q thinking if we can dampen down the constitutive activation, maybe they grow better, but that didn't work. Um, so here in blue is the control, these cells actually grow very robustly in culture, uh, but the mutant in orange is just slow. So, We looked in uh land and looked in vivo to see, there are, there is one paper saying there's increased proliferation and capillary malformation. But we looked, OK, you can see there's, so these are sections from the brain, George Weber brain, stained with Eulex Europus, which are our endothelial marker in red. So you see all the vessels. And then KI 67, which is a proliferating cell marker. And you see the occasional Positive cell, but not many, and especially if you compare it to something like Infantile hemangioma. Here we have lots of KI-67 positive cells, many of them in the endothelium. So Lan Wang in the lab, uh, quantified this, looking at several, uh, sections from each, and you can see that compared to hemangioma, which is truly a proliferative vascular tumor, these, um, the capillary malformations in the brain or in the skin just really aren't proliferating. Um, OK, so, um, let me go back to this, uh, Tell you a little bit more about G alpha Q and we're gonna look at downstream signaling. So here's a lot of information that in endothelial cells, there's some potential receptors, G-protein-coupled receptors that can activate G alpha-Q have been identified. It, as I mentioned, it's known to activate phospholibase C beta, which leads to calcium mobilization, also activation of the MA kinase pathway. Uh, VEGF stimulation leads to PL phospholiase C beta 3 activation via VEGFR2 and G alpha-Q, and shear stress can also activate the G-alpha-Q. Independently of the receptor signal. So schematically here you see GTP bound to G alpha Q that makes it active. The phospholi AC can stimulate calcium release and also dacylglycerol and lead to lots of signaling. So obviously the first thing we're going to look at is phospholiase C, the first downstream modulator. And Lan uh did this, we, we're looking in, in this case in quiescent cells. So they've been serum starved and we took away all the growth factors that we use to grow them. So they're quiescent. Here's normal brain endothelial cells. Brain endothelial cells with the wild type GNAQ expression. And brain endothelial cells with the mutant overexpressed. So here you see G alpha Q. It's overexpressed, so it's a lot easier to detect. And there's two different phospho phospholipase C antibodies. The 537 here is the activating phosphoserine, so it's clearly much stronger in the mutant cells. The inhibitory fossil sharing, there's really no difference. Here's the total fossil light they see. So we have active, we clearly have activation of this immediate downstream target, which is good, nice to know, and lands on several Western blots and quantified them, and you can see we have a statistically significant increase in the activating phosphoserine. So that's all good. But then when we looked at, again, and this is a lot of data here, but the red bars just shows the first line, these are the quiescent cells, and then we stimulate with VEGF. Sorry, it's not showing up here. But if you look at the red bar, OK, there's no, in, in the normal endothelial cells or in the mutant, there's no phosphorrelation of VEGFR2. There's really no phosphorylation of earth, phospho-er, which is interesting cause it's kind of the assumption is that this is gonna activate phospho er basically from a lot of studies on uveal melanoma where there's a similar but stronger mutation. Actually, the control had a little phosphorus, but there's really none in the mutant, and there's no phosphor P38. If you stimulate these cells with VEGF, which denoted here for different periods of time, 5 minutes to 120 minutes, clearly these cells have, they respond to VEGF. You get phosphorylation of the VEGF receptor 2, rapid phosphorylation and defhosphorylation as expected, Earth's activated, so everything's working fine, but there's just, there's no constitutive activation here. And so we, uh, we also looked in the engineered cells where, again, these cells are expressing either wild type or the mutant R2 and the, the first lane, the zero is The quiescent state. And again, there's, there's no activation of phos the VEGF receptor, there's no activation of phospho-EC. And ran ran this many times again, because the expectation in the literature is that, is the ERK pathway should be activated. And you may see a little, in red is the mutant cells, a little higher expression of irk, but it's really not significant. So, um, To summarize this, and here we have a schematic, here's the cell membrane, we have the G alpha Q is um Here and in the mutant cells, the, the mutation causes a constitutively on alpha Q subunit which can activate phospholibase beta 3. We're pretty convinced of that, so we have a nice bright arrow here. Well, we've looked at ER, we've actually looked at phospho AKT. It's really modest if there's any upregulation. So we're, now we're actually really focused on the NETC1 pathway, looking at downstream targets, one of which is thrombospondin one, and that's sort of our ongoing direction right now. So the next question we ask is, do these mutant cells, the, the end of the mutations in the endothelial cells, are they disrupting the neighboring cells? And it really looks like by histology, there's just all this kind of chaos going on around the vessels. So we, and you can see that also by immunostaining, it's really the, the sort of chaos. And so we looked at one experiment we did in collaboration with the neuroscientists at um Right in CLS is to, uh, we, we did this proteomic screen of the conditioned media from either the mutant cells or control cells. And we, we compared our engineered cells versus wild mutant cells versus wild type and the sturge rubber brain endothelial cells versus normal brain. And you could see this should be proteins actually. Um, there are a number of proteins that are up regulated. And then we, we looked for ones that were upgraded in both situations in the mute, well, altered, either up or down regulated in the mutant in both, both cell models. And we've got a big long list, which is quite interesting. And the one I'm gonna focus on is CXCL5, which is a chemine, it's a proangiogenic chemoine. It's not a lot known about it, but it's clearly when we then landed uh Um, QPCR and we see it up regulated in both cell models. So that's exciting. And then, She sustained some sturge rubber brain sections, and I hope you can, here's our endothelial marker, you see all the vessels, and they're also co-expressing CXCL-5 very nicely and, and when you merge them, you see kind of a yellow orange signal, which means they're co-localized. So, and when uh we don't have normal brain yet, but if we look at normal skin, I hope you can see, here's some red vessels, there's a couple of tiny ones down here, and they're just not expressing CXCL-5 at all. So this seems to be a marker perhaps for the Sturge-Webber or the capillary malformation vessels. So again, we turn to our in our in vivo model, which I told you about in the beginning where we're gonna, we can implant mutant cells, either normal or mutant cells combined with a Smooth muscle precursor. And we implant them into nude mice and watch vessel formation. And what we see here, we, so we implanted our surge Rover brain endothelial cells versus human brain endothelial cells, and our control is the human ECFCs. Now obviously get lots of vessels here. The brain endothelial cells did not do very well in this assay. We only had a few tiny vessels, but the. Webber brain, they also were, we got kind of immature nascent vessels, and I'm just showing you the, the immunostaining for just the Sturge Webber brain endothelial cells. We see human endothelial cell vessels, and they're CXCL1 positive. So this is a nice result. We think these patient-derived cells are actually recapitulating a signature of the vessels in, in the patient's specimens. So this is very new data, um. And I'll summarize this part of the talk and the, this final part on our GNAQ mutant endothelial cells. So first, we showed the mutation is enriched in endothelial cells from skin and brain capillary malformations. This is published work. We have, have two papers on that. The mutant cells proliferate slowly even when they're stimulated with VEGF, and the proliferation in vivo is also very low compared to infantile hemangioma. So it's It doesn't seem like proliferation is an ongoing thing. When you take the quiescent, meaning serum-starved mutant endothelial cells, they have constitutive activation of the phospholibase beta 3. So there's something's happening. Again, we didn't see strong activation of, we, we don't know what happens after phospholibase beta 3. And then the proangiogenic chemokine CXCL-5 is up regulated in these mutant cells and in the endothelium of brain, the Sturge-Webber brain and port wine vessels. And then the Sturge-Weber. Um, endothelial cells form CXCL-5 positive vessels in the xenograph model. So that's our, our, uh, OK, ongoing and future studies, we, we're really focusing on the calcium and that C1 signaling downstream of phosphate-base beta 3. We'd like to ask the CXCL5 play a functional role in CM or is it just a marker? Um, do, and then we're doing a lot with the endothelial condition media in collaboration with Ana Pinto and Mustafa Sahin in, in the Translational neuroscience Program. We're looking at does this conditioned media have any ability to stimulate or overstimulate human neurons, and, and at the same time, we're looking at the endothelial condition media on the parasites and the perivascular cells. So you have this mutation, is it having cell non-autonomous effects. OK, and then to summarize, uh, here's a picture of the, my current lab members, and they're listed here. I talked a lot about Lan Wang's work and also Colette Bixe. We also have Charlotte Erpun and Jill Wiley Sears. We have a lot of, fortunately, we have a lot of funding right now, and, and I have a giant list of collaborators and different groups that we collaborate with, and I think that's really a key, this has really enabled our research tremendously. So first, the vascular biology program, and then, of course, the vascular. Anomaly Center here, John Mulligan, Harry Kozakkiewicz, Denise, and Cameron. Um, we're also on the Sturge Weber Project working very closely with the Translational Neuroscience Centers, Robin Kleinman, actually, she's gone on to Biogen, but Anna Pinto, Liz Buttermore, Katherine Kiever, Mustafa Saheen, and Joe Madsen is the surgeon we got the brain specimens from. Um, we work, and then in pathology, Sandra Alexandrescu is a neuropathologist and she's been fantastic for the Sturge Weber Project and Dick Robb on some eye work I didn't show you. Um, we work very closely with Aaron Green's lab and his team, um, on capillary malformations and also helping them with their AVM project. And then, of course, we have the Vascular Anomalies Center in Brussels, Mika Voula and Lawrence Boon. I know Steve knows this, this group very well. And we also work with Matt Warman's lab on all kinds of digital, droplet digital PCR and Genetics. So, thank you and I'm happy to take questions. Oh Start by first thanking you for bringing this exciting presentation to us this morning. I think, uh, identifying the genetic abnormalities uh involved in the tumors and the malformations is, is quite promising. I guess I'm not surprised to some degree that In the, in the tumors and with the um Mutations are able to affect the proliferation of the cells following the malformations, which I think of the cells as being very static, you were talking about at the end with the Sturge Weber. That since they aren't proliferating, then attacking the genetic mutations really doesn't affect proliferation at all. Yeah, but I mean, still in the capital, if you look at it, you know, there's a lot of vessels, there's way too, there were, there must have been too much vessel formation at some point, whether that was due to migration or proliferation. You know, it's, it's sort of the go to, it's like the first thing you always look at because there's, there's too much, too many vessels. So how did they get there? Sure. So yeah, we ruled it, and also there's one, there's one report, uh, on surgery of a brain saying, oh, there is increased proliferation. It's like. You know, maybe It goes from like 0 to 0.001 or something. So tiny, it's not probably not a. It's not a significant mechanism in formation, but Yeah, no, it makes sense. I have to get the data. I mean, the, the one ultimate proliferation that's, that's of significance is in the Sturge-Weber patients that didn't develop a malignant brain tumor. And is there any idea which one of these malformations could be ultimately involved in the malignant degeneration of the malformation? That's a great question, um. I don't know, but it's true. So the other, you know, the surge rubber patients, many of them also have a, a capillary malformation in the choroid of the eye, and, um, you know, so where the, where it originates from, starting in the left meninges and then populating the eye and the skin or some other way, and, and maybe if, if, if we knew the really originating cell, then that might be a candidate for tumor formation for to give rise to a tumor cell. Cause you just need another hit. Additional questions for Doctor Bishop. Doctor Moses. Thank you. Joyce, I'm really interested in what happens when the hemangioma regresses, and so have people looked at that fat pad or that mass of adipose tissue because adipocytes are great regulators of the vasculature, right? So has anyone looked, do they play an active role or no? Um, I, you know, people have done some microarray analysis and, and things like that a little bit. You know, that we actually have a lot of involuting specimens because sometimes when John is removed, well, not a lot, but you know, he's doing sort of a Uh, helping with the cosmetic aspects an older child, and, and so we do have some of those specimens just frozen away, but But not really. I mean, I think the evolution mechanism is really unknown and is extremely interesting, but we have actually focused on what's driving the tumor because then if we can inhibit that, you know, it could have a, could be beneficial, um, but it'd be interesting to, to have a look because obviously it's a robust, you know, evolution and it's maintained, and so it's the same old angiogenesis story. There are inhibitors there that might be maintaining that. You know, uh, dormant phenotype or something of that nature, so interesting. That's interesting because with propranolol therapy, if it's stopped too soon there's a rebound. So is that somehow releasing, you know, so is the propranolol effects and making them and that's how it works. Yeah, I'd put money on the fact that they're not just structural, you know, they, they may be having an active role. We can talk back home now. Doctor Fishman. I'd like to make more of a comment than a question. Um, if Doctor Folkman were here, he'd be proud. Um, there's, you know, a lot of people in this room don't even know who Dr. Folkman was, even though his picture is right there in the auditoriums named after him. And, and, uh, in, in the lifetime of everybody in this room, this field has transformed. When I mean Joyce and I really came up in this together on different ends on the scientific side and the clinical side, and um there were these geniuses of John Mulligan on the clinical side and, and Judah Folkman in the laboratory, and some people don't know that Dr. Folkman was actually a pediatric surgeon. He was the chief of surgery here before he devoted his, his full experience to the laboratory, and he's known as the father of angiogenesis and conceived of angiogenesis inhibition, and he has progeny all over the world. Doctor Moses, Doctor Bischoff, uh, many, many others, almost all the people that Joyce mentioned. And, you know, he focused. Um, mostly on cancer. Um, but he partnered with John Mulligan because he was interested in the model of angiogenesis demonstrated in, in vascular tumors and vascular malformations, and he thought that was a hint. And he brought people, um, into his laboratory and encouraged them to find their way. And almost all of them focused on cancer, and Joyce is the one who, who chose to, to go the other way and has really led the world in understanding on the initially the cellular side, how these benign processes were working simultaneously, Mika Akula, who, who, who Joyce referenced a lot, uh, who was here and then in Brussels, focused on the molecular genetics. And I always thought that. The settler stuff made a lot of sense. And I remember for years asking Mika. OK, so you can find the genes, so what? These are malformations that happen before birth. What are you gonna do, change them? Um, and I was challenging him, but also, you know, he, he, he, he, he was challenged to find the answer. It was scientifically interesting, but it wasn't clear that it was ever gonna make a difference. And in our, in our, and now you can see that the molecular and the cellular have come together, and now Mika and Joyce are close, close, um, colleagues, um, and sort of have now their own progeny, Lisa and all the rest who, who, who, um, who are doing this, um. I remember many, many times we'd be sitting around our Wednesday night conference where some people in this room come to where we discuss these difficult patients. There's a bunch of surgeons and interventional radiologists talking about what we could do for the patient, and we have a patient we just couldn't fix. We'd just turned and Dr. Faulkman would say, We need a drug. And after Doctor Faulkman passed, John Mullen and I would look at each other and say, we need a drug. We didn't exactly believe that there'd be a drug within our careers, but in large part due to this kind of work, we have patients on drugs every day. In fact, Denise Ems can't be here right now because she's leading a trial and she's seeing a patient on one of these experimental drug trials. So within truly a quarter century, we've come full circle from You know, when I started this, when we, we, we got here. It was sort of like Virkau, right? John was just making up names with Harry Kazike, which based on what it looked like clinically and what it looked like in the microscope, and they were naming conditions. And now, and we're still doing that, and we're still figuring out the natural history of the conditions and we're figuring out what the complications are and figuring out what interventions are safe and the optimal treatment. We, we've totally changed the way we treat them. At the same time, within the same. Um, world here, in large part because the connection between Doctor Folkman as a surgeon and as a scientist, the whole world of, um, blood vessel development and maldevelopment, um, centered here and, and, and this says a lot about why do you do science at a hospital. Right? Most scientists thought of us at, at a university or in, um, in pharmaceutical. And this is the largest pediatric um scientific institute in the world, uh, and it's a hospital. And, and these scientists here, there are appointments in the department of surgery, some of them, um, and, and we need to, we need to champion that and recognize that and continue that because it is a collaboration between serious world-class PhD scientists and people who see patients and cut them off and. Given the address. I was sending specimens to Brussels for years, you know, you know, and saying, so what? But the, so what has come true and it changed their clinical practice. So I think it's important to realize sitting in this auditorium, um. The, you know, the contributions that Joyce has made, um, it goes beyond that. She mentions Aaron Green. He's a plastic surgeon. He came up sort of the whole thing. Uh, Joyce mentored him scientifically. Marsha Moses sitting here, and, and she, she did the whole cancer world and Doctor Folkman's. lineage but also mentor a surgeon like several surgeons, but like Ed Smith, who's now leads the world in vascular anomalies in the brain. So it is this collaboration between really smart scientists and some competent MDs who try and sort of hang on to their coattails who make this all possible. So I, I think that um that he would be very proud. Now I'm gonna see if I I need to have like this all disappear and then a picture of Doctor Polkman come up as my you know. Thank you. Very nice comment, Steve. Any other questions for Doctor Bischoff? Now, thank you so much for being with us this morning, Joyce. Hm.