I'd like to welcome you all to the hormonal impact on endometriosis. I'm Jessica Chan. I'm one of the reproductive endocrinologists at Cedars-Sinai in Los Angeles. And I'm Rob Taylor at University at Buffalo. I was just going to make a comment. I really like the subtitle of this, sort of the unseen disorder of reproduction. Throughout my training, endometriosis has been sort of recognized based on its visual characteristics. So it's really quite interesting that we're thinking about it as an unseen disorder. So great. So we'll, Dr. Lawrenson, I think, is going to be the first speaker. OK. All right. Good morning, everybody. Thank you so much to the organizers for inviting me to come and to share some results from our single cell profiling of endometriosis. So I don't think I need to educate this audience about endometriosis, which is a condition where we are finding endometrial-like tissue outside of the uterus. Most commonly, we see it in the peritoneal cavity. But it can affect more distant organs as well, as well as abdominal surgical scars. And really, this is a disease that we know very little about. Currently, surgery is the only way to get a definitive diagnosis. And there's no way to reliably diagnose endometriosis with a blood-based biomarker, for example. For many reasons, diagnosis can take many years, six to eight years, or even longer. This is for many different reasons. Symptoms are broad and non-specific for many patients. And also, I think, approaches to women's pain is also an issue that contributes to this long time for diagnosis that many patients will unfortunately suffer. And overall, there are quite a few treatment options. And a lot of treatment for endometriosis occurs by trial and error. And really, we don't have anything akin to the precision medicine that we have for cancers, for example, for endometriosis patients. And so really, on the research side, we've got a lot to do. And so a few years back, myself and a bunch of clinicians at CEDARS set up a study, which we called the BME study. So that stands for Biologic and Epidemiologic Markers of Endometriosis. And we invite patients who are having surgery, so we'll have a confirmation of an endometriosis diagnosis through path review. We invite those patients to contribute any tissue that would otherwise be discarded. We ask if they would donate serum, plasma, and buffy coat through a blood sample. We also collect vaginal swabs and endometrial biopsies. And we also invite patients to complete questionnaires that tell us about their family history and their experience of their disease. And so even in spite of it being a pandemic last few years, we've already got about 450 patients on that study. And it's been a really enthusiastic population to work with. And so I'm just gonna share one of the studies that we've been doing off the back of the BME collection. Also just to say we are really open to new collaborations and we'd love to get more involved in the endometriosis research community. So if you'd like to collaborate with us moving forwards, please do reach out. And so broadly speaking, we were interested in using the tissue specimens to understand better the transcriptional features and long-term the, I'm really interested in gene regulation, how transcription factors turn genes on and off during disease development. So we're very keen to apply single cell genomics to this disease. It's really the ideal technology to use for a disease such as endometriosis where the cellular heterogeneity is part of the disease, which has really precluded as doing bulk sequencing on this tissue type. And so broadly speaking, when we look at endometriosis, we know that most patients who have endometrioma will also have some peritoneal disease, but not vice versa. And so very broadly, we were interested to know how do these two disease subtypes look in terms of their cellular and molecular biology. But of course, off the back of creating this atlas, we've been able to ask many additional questions. And so this gives an example of one of the patients in the study. So here we were able to get multiple specimens from the same patient. This included deep lesions as well as superficial lesions and an endometrioma. But in all, we profiled over 50 specimens across 21 different patients. The majority of the patients did have endometriosis, although we had some controls who we only profiled utopic endometrium and ovary from. And that was four patients in total. And then from our endometriosis patients, we had a mixture of endometriomas, peritoneal endometriosis and utopic endometrium. And all of the peritoneal disease was classified as superficial or deep by a consensus of two mixed surgeons. And so our methodologic workflow was that the specimens would be bisected. One would, half of the specimen would go into permanence and we'll come back to what we use that for later. And half would be brought back to the lab immediately and we'd dissociate those tissues and immediately capture for sequencing using the 10X platform. And then based on the transcriptional profiles of each individual cell, we can cluster based on the similarity between cells and start to ask questions about the disease. And so this shows just the first pass look at the cells. So when we have a look at the Atlas, this was over 400,000 cells after quality filtering. We see a lot of different immune cells as we might expect, a lot of mesenchymal cells. I'm not sure if this is showing up. We could also capture the epithelial cells and the endometrial type stromal cells that are diagnostic for disease, as well as smooth muscle cells, endothelial cells. And we had some erythrocytes that persisted even after we did a lot of RBC depletion. And so the first thing we wanted to ask was just what are the broad characteristics of the data set. So if we look on the right, the mesenchymal cells and T and NKT cells were the most abundant. So this is a number of cells and these are the different cell types. And then on the left, we stratified our tissues into five major classes. So endometrioma, eutopic endometrium, the peritoneal extravarian endometriosis. We had four specimens where there was no endometriosis detected upon path review. They were from endometriosis patients. And actually when we look in our single cell data, we can detect a small number of endometriosis or endometrial type epithelial mostroma suggesting that that specimen actually did have a small amount of endo in it. And then we had some unaffected ovary tissues. The bottom shows the expected distribution under the null that there would be no enrichment. And so we can compare the different cell types to see if there are cell types enriched or depleted across these different classes. As we would expect, kind of our positive control is that we would see epithelial cells and endothelial cells enriched in eutopic endometrium. One thing that we're very interested to see was B in plasma cells enriched in endometrioma as well as extravarian endometriosis. And so one study we're focusing on now is understanding what is the specialized function of these B in plasma cells and how are they contributing to disease pathogenesis. Another important feature we were keen to note is that each different color here in these rainbows denotes a different patient. And so on the whole, we weren't seeing features that were unique to a single patient. We were having multiple individuals contribute to the signals we were looking at. As we got to stratify down into the subpopulations of these major cell types, we were careful to exclude any clusters that were private to a specific patient so that we knew that we were looking at features that were where we were getting information across multiple patients in the cohort. And so this is quite a nice vignette. This is a PCA plot based on the frequency of all the different cell types present in the whole data set. On the right, we can see the green represents our normal ovary and that starts to separate out quite nicely along PC1. Patient 14 was a unique case in that we had both ovary specimens. The left ovary had an endometrioma. Oh no, sorry, the other way around. The left ovary was unaffected. The right ovary had a superficial lesion and they clustered by the tissue type rather than by the fact that they were from the same patient or that they were normal ovary specimens. And this is something that we've seen come out again and again because when we think about doing single cell profiling on endometriosis, the lesion proper is actually only a small part of the specimen and we have the local microenvironment as well as the host tissue coming along with the specimens that we profile because we don't fractionate them. But what we are seeing is that the local microenvironment undergoes quite extensive changes even at some distance from the lesion itself. And I think that's why we are seeing that the ovary here, the superficial endometriosis specimen on an ovary really is looking like endometriosis and not like normal ovary. I actually, when I predicted, I would have thought it would have clustered with a normal ovary, but I got that completely wrong. And the right ovary, and now we're looking at the correlations between all the different specimens to ask if we have any broad groups or broad kind of subgroups within the data. Broadly speaking, we can see that here on the right in cluster three, we had all of the unaffected ovary and about half of the eutopic endometrium. The peritoneal endometriosis formed one big cluster here on the left. And then in this cluster, we had the majority of our endometriomas and peritoneal, sorry, eutopic endometrium. And so broadly speaking across these three classes, we do have differences in the cellular composition that really drive the similarities between the different tissue types. We also see contribution of the hormonal microenvironment as you might expect. So we can see over here, we were getting most of the samples that were in the proliferative phase of the menstrual cycle and some spattering across each of the clusters for patients who are on exogenous hormones. And this is something we're still kind of working through. It's been a complex story, as I'm sure you can imagine. And so now I'm just gonna share a couple of vignettes about the epithelial portion of the data. So when we pull out just the keratin or epicanpositive cells, we now have just under 14,000 cells. And this shows that clustering on the left, we aligned these clusters with some of the reports of single cell profiling of eutopic endometrium. And we're able to find on the whole some similar cell types. So when we look at these top five clusters are our endometrial type epithelium. And we had the SOX9-LGR5 can more progenitor type cell, and then some secretory mucin expressing cells, and then some nice differentiated ciliated cells. And we can see these in endometriosis legions as well. These were all generally enriched for eutopic endometrium, but we could see them both in endometrioma, which is the dark purple, and to different extents in the pale green, which is endometriosis. And so we were interested to look at the different distribution of these cell types across the menstrual cycle. And so again, here this is, we've got the bar up here that shows the distribution under the null, and then we can look for deviations from that across the different clusters. So this SOX9-LGR5 plus population was enriched in the follicular phase, which is yellow. And then the more secretory cell types are enriched during the luteal phase, as you might expect. So this is eutopic endometrium from patients with endometriosis. But when we take now the same cell types that are residing in an endometriosis legion, that relationship with the hormonal microenvironment gets completely skewed. We're also seeing interesting patterns of this Indian hedgehog SPDEF population, which we didn't see in eutopic endometrium of endometriosis patients, but we are seeing in now the lesions. We do also see this in the eutopic endometrium of patients who are on exogenous hormones. And so we're interested to kind of understand what this population is and why it seems to have this endometriosis-specific occurrence in patients. So we used Tenex Visium Spatial Transcriptomics to validate the presence of some of our populations and understand how they would sit in space relative to each other. And so this shows one peritoneal lesion. On the left, there's a H&E with, we had three foci of endometriosis that circled in white. We had here the different epithelial markers, so FOXJ1, our ciliated cells, PAX8, secretory cells, and then a generic gynecologic keratin, keratin VIII. So we can see that we can detect these different epithelial subsets within the lesions. And then on the bottom, these are showing some of our different mesenchymal cell markers. And so what we see is not only the CD10 positive endometrial type stroma within the lesions, as we would expect, and that can be seen by standard pathology, but we can also see more of this activated stroma, which is the host stroma becoming more like an activated fibroblast-type phenotype immediately around the lesion. And they express MMP11. And then we kind of have this stripe of what we think here are the unaffected fibroblasts now at some distance from the lesion proper, and these are expressing CFD. We also noticed anecdotally that when we look across the endometriosis, eutopic endometrium, and endometrioma, we were tending to see some of the genes come up that were the same across the different tissue types. And so we wanted to ask if that was more than just anecdote. And so we compared genes that were upregulated in endometrial-type epithelium or stroma by site. And so each row, first we have epithelium, and then stroma for eutopic, epithelium stroma for endometriosis, epithelium stroma for endometrioma. And particularly in the endometriosis and endometrioma, we were seeing these coordinated transcriptional responses. So red and yellow means that genes are being upregulated across the two different cell types. So like this strip here for endometriosis and this strip here for endometrioma. We had a number of different complement proteins being upregulated and also genes involved in dealing with genotoxic stress and hypoxia, which is perhaps not surprising. But I think what's exciting about this is that it suggests if we want to find new biomarkers for disease, maybe we're gonna have two different cell types spewing something out, so we've got a better chance of seeing it in the peripheral blood. Also, therapeutically, if we want to hit one of these pathways, we're going to not only disarm the epithelium, but also the support network in the stroma with one foul swoop, and that could potentially be a very potent therapeutic approach. And so we're really interested in exploring those two angles off the back of these results. And then just one last analysis that we did with the epithelium. We were really interested to know about the contextualizing these results with the vision of what the somatic mutations are. So we know that these cancer driver mutations are very common in endometriosis, but we don't know what it means for the cells compared to what it means for cancer cells, where KRAS-R1A mutations have been studied quite well. And so we know that KRAS-R1A, PIK3CAOB2, beta-catenin are very commonly altered genes. And this field was kind of kicked off with this paper from Mike Anglizio and David Huntsman, although there's been a number of other beautiful papers that have confirmed these results in different subtypes of endometriosis and eutopic endometrium. And so we worked with Mike Anglizio's team to genotype our specimens. So this, the permanence, we subjected to a set of, we had them cut into a set of slides and we top and tailed with H&Es so we could make sure the lesions were still there. And then we use the slides for microdissection or immunohistochemistry because they've developed some assays up there for ARID1A and PTEN, which unfortunately didn't work, but with ARID1A we can use immunohistochemistry as a mutation surrogate when it's done with the appropriate controls. And so we had a number of our endometriosis tissues and endometriomas had KRAS and or ARID1A mutations. And so we could now stratify our epithelium and the other cell types as well, but today I'll talk about the epithelium and ask whether we have different gene expression if endometriosis epithelium has wild-type ARID1A or mutant ARID1A. And the broad response is that we did see some changes in gene expression accompanying ARID1A loss. And we were particularly interested to see this gene SOX17, which we'd been working on in the context of fallopian tube transformation into serosevere in cancer. And so we first validated that when we had loss of ARID1A, we had increased expression of SOX17 in the epithelium. So we had more pervasive expression and higher level expression in the epithelium. And SOX17 is interesting because it's well known in the context of early development, but also endothelial cell development. And in ovarian cancer, it actually promotes a pro-angiogenic program being secreted by the epithelium. And we think that's partly how it contributes to tuberigenesis. And so we thought perhaps SOX17 overexpression in the epithelium is contributing to endothelial cell development or angiogenesis and endometriosis as well. And so we turned to the endothelial cell compartment of our data, which we sub-clustered into endothelial cell clusters based on the similarity of gene expression across those clusters. And here is our total column. So this is the null under no enrichment or depletion. And we could see that we had two clusters were enriched in the ARID1A mutant. In lesions with mutant ARID1A in epithelium, we had two clusters of endothelial cells over-represented more than you'd expect by chance. And then the rest of the clusters were showing no significant deviation. And when we looked at those two clusters, we had one was very differentiated lymphatic endothelial cells and the other one less differentiated, but still showing these classical LEV cell markers. And then this just shows the remainder of the blue, not enriched really on expressing these markers. So these two clusters that were enriched in our mutant tissues were lymphatic endothelial cells. And when we looked at our ARID1A mutant and wild-type epithelium, we could see that the loss of ARID1A was accompanied by high expression of VEGF-C, which is the main lymphangiogenic regulator. CCB1 is also a promoter of lymphangiogenesis. And that was also upregulated when we had loss of ARID1A. VEGF-C can be either promoting or inhibiting lymphangiogenesis in different contexts. It's also not as potent as VEGF-C. So we think that probably the main culprit here is gonna be VEGF-C secretion by ARID1A mutant tissues is potentially promoting lymphangiogenesis in endometriosis. And I think like many of these single-cell atlases, we're not finding necessarily lots of new stuff. We're just able to describe it with a slightly more precision. People have known for a long time that we have altered lymphangiogenesis in endometriosis. We see both in eutopic endometrium and proximal to endometriosis lesions and increased lymphatic vessel density. And this is potentially the main conduit for how lymphocytes and antigens are going to and from local lymph nodes and also may potentially be how endometriosis might spread to more distant organ sites. And so potentially ARID1A, I don't think explains all lymphangiogenesis in endometriosis, but it might be an accelerant of that process and potentially with SOX17 is a critical transcriptional regulator in that pathway. And so in summary, our single-cell atlas, I think is supporting what many people are saying, that peritoneal endometriosis and endometrioma really are distinct entities that potentially going to need to be treated differently. I think there's potential clinical insights we can gain from the fact that we do see these coordinated gene expression programs across sites. I didn't show it today, but when we look into deep and superficial peritoneal endometriosis, we don't see as marked differences as when we compare peritoneal and endometrioma. Potentially I think that's largely driven more by transcriptional plasticity rather than being more distinct disease entities. And then we are seeing both with KRAS, ARID1A that we do have changes in epithelial signatures that could potentially be contributing to disease pathogenesis. And so I welcome any questions, but first want to thank the wonderful team of surgeons, Kelly Wright, Matsi Dofmire-Trong, and also many people on the lab side. So Marcella Harrell, Shib and Furu Abasi, and Marcus Zabrau did all of the data generation and analysis for the work you've seen today. And then Fabiola Medeiros is the wonderful pathologist who partners with us in these studies. Thank you. All right. Thank you so much. We'll open up for questions now. So DeMayo from NIEHS. I really like the talk. It's great. So the hedgehog got me excited because we studied that a long time ago. So did you look at the stromal cells in your single cell and then try to look at a cell chat or something, communication between epithelial and thrombic? Because, you know, we established this hedgehog COOP-TF2 and I think Serta Bullen showed that COOP-TF2 was altered in endometriosis. We are working on that now because, yeah, I think that's going to be the important thing we do next, both with the stromal cells and the immune system. Right. Because hedgehog, we initially published as a progesterone-regulated gene, but it's also estrogen-regulated, which would make sense in your situation. Yeah. And then the second, are you thinking about doing single cell ataxic between endometriotic and endogenous to find out what's the difference in the epigenome? Would that be really exciting? Yeah, we started to dabble. I've got a question and I will ask for people who want to ask a question is to come up to the microphone so that our people at home or, you know, on Zoom can hear the presentations and also we're recording these. So one of the questions was actually about, is it the cancer driver mutations that you've identified, are those all limited to the epithelium as it's been kind of described? And do you see any of these sort of cancer driver mutations in stromal cell populations? We didn't look for the reason you've described. Most people have shown that they are pretty much always limited to the epithelium so far. So we only genotyped micro-dissected epithelium here. And I guess another question is, oh, great. No, please. Oh, hi, I'm Diana Montalvis from Baylor College of Medicine. So I had a question about how deep your sequencing was and whether you were able to identify other stem cell markers that have been identified in the eutopic endometrium, such as SSEA1, CDH2, or ALDH1A1 or SUSD2. Yeah, so SSEA, we didn't see it. It was just too low. We didn't see, CDH2, is that the same as N-cadherin? It's CDH2, it's N-cadherin. Yeah. That one we didn't find specificity. And that's sometimes what we see. A lot of the markers that we people have done by protein-immunized chemistry or flow for years don't always translate that well to this technology for the reasons you're alluding to. It's kind of shallow. And also transcription and protein expression don't always correlate super well. Okay. All right. Thank you. Thank you. I was just going to make the comment that one of the three classic theories of endometriosis for the last hundred years has been, you know, sort of Samson's theory, Meyer's theory of metaplasia, somatic metaplasia, and Halban's theory of sort of lymphatic or hematogenous spread. And this is really quite nice to see this kind of reinvigorating that old idea. So thank you. Well, great. Thank you all. So Hugh Taylor will be the next speaker from Yale. Great. Well, thank you. Today I want to talk to you about not just the lesions in the pelvis, not just the cells that we heard in our last talk, but how endometriosis is also a systemic disease, that this disease really is a lot more than just what we see in the pelvis. And I hope to convince you today that endometriosis truly affects the entire body. And what we see in the pelvis is just a portion of the disease. Again, just by way of introduction, I think this audience knows that it affects up to about 10% of reproductive age women, common in women with infertility and pelvic pain. It's the second leading cause of hysterectomy in the U.S. and about $22 billion in healthcare costs. So a common disease. As we know, it's defined by the ectopic endometrial glands and stroma. And we just learned a lot more about the cellular makeup of those lesions. But I think this simple definition belies the complexity of the disease. Oh, you told me it was very sensitive. So we got a whole instruction about how the microphone is very sensitive. So I didn't want to get too close to it. Truly this simple definition, again, as our last speaker spoke to, of the simple definition of ectopic glands and stroma, really belies the complexity of the disease. We see very different looking lesions. They're blue, they're red, they're white, they're clear, or the big endometrioma. So it is a more complex disease than we'd normally think about. The clinical presentation is also quite different. I mean, sometimes we see women with extreme pain. Sometimes we see women with no pain and they come in with infertility. And sometimes they may have neither pain or infertility and endometriosis is just found incidentally during some other procedure or imaging. So again, complex presentation. And the disease stage does not explain the pain symptoms. Whether we look at dysmenorrhea or non-menstrual pelvic pain or dyspareunia, the amount of pain someone has is the same whether you have extensive disease or just a couple of little dots of disease in the pelvis. I think what we see in the pelvis is really just part of the disease. And again, I hope to convince you by the end of this talk that it is a systemic disease with manifestations outside the pelvis. One of the other things that points to that is the vast number of symptoms that are often associated with endometriosis. Again, we talked about the pain and infertility as the most common symptoms. But we see affected surrounding organs, the bladder, the bowel are affected. There's this whole body inflammation. We can see elevation inflammatory markers. There's also an effect on mood. Depression is much more common in women with endometriosis. Anxiety is much more common. Women with endometriosis have on average a lower BMI. So there's change in their body weight and fat distribution. And now we're learning that even in the long run these women can be at higher risk for cardiovascular disease. So again, this is not just about what we see in the pelvis. It is truly a systemic disease. We tried to get at understanding better the etiology of endometriosis to see if that could explain potentially the systemic nature of disease. And we've always thought that endometriosis arises primarily through retrograde menstruation, Sampson's theory. And again, we did some simple experiments to look at that, as others have. We just injected GFP endometrial cells into the peritoneal cavity of mice. And we can find these GFP cells incorporating into the lesions, either forming new lesions of endometriosis or even incorporating into pre-existing lesions. And I won't go through this completely, but what this looks at is the different types of cells that can come from endometrium. And it's not just the epithelium and stroma. It includes vasculature. It includes multiple different CD45 positive leukocytes of various types. So again, confirming what our last speaker presented in much more detail, we know that there's heterogeneous types of cells that end up in the endometrium. But Sampson's theory really can't explain all endometriosis. As we talked about a moment ago, there may be endometriosis in areas remote from the pelvis, in the lungs, in the brain. We do see that clinically. And certainly that isn't simply by retrograde menstruation. Although it could be hematogenous or lymphatic spread. But it doesn't explain, none of that explains endometriosis that was found in the old days in men, when men were given high doses of estrogen to treat prostate cancer, some of them would develop endometriosis. That certainly doesn't come from the uterus by any mechanism. So we asked, could there be a stem cell origin of endometriosis? Could cells come from someplace other than the uterus and give rise to endometrium in ectopic locations? And we looked first at some years ago at bone marrow. Bone marrow has these mesenchymal multipotent cells in addition to the hematopoietic cells that can give rise to various cells of solid organs. And we asked, could stem cells travel and differentiate into endometrium in ectopic locations? Could this be another source of endometriosis? We did some very simple experiments many years ago where we transplanted male bone marrow into female mice and then looked by Y-chromosome fish just to see if we could detect the Y-chromosome in endometrium or endometriosis. Here, the results of that transplant, the red is the signal from the Y-chromosome fish. The A in the upper left shows a typical male as a positive control. B shows the uterus of a mouse with a female-to-female bone marrow transplant, negative control. C and D show some bone marrow-derived cells, again containing the Y-chromosome, incorporated into the endometrium. In C, you can see an epithelial cell. The dark area in the center is the lumen. That's an epithelial cell. D is a stromal cell. We looked at a lot of other markers to verify those indeed were the types of cells I'm claiming they are here, but I won't show you all those details here today. And we also confirmed this in humans. We looked at women who had bone marrow transplant after radiation, chemotherapy, bone marrow transplant. In the old days when they actually took bone marrow aspirates rather than peripheral cells. And we looked at women who had a single HLA antigen mismatch between their donor and themselves. And we could use the mismatch HLA antigen to identify the origin of any cell. The brown here stands for the HLA antigen of the bone marrow donor. You can see some epithelial cells on the left and some stromal cells on the right that are a bone marrow donor origin showing that bone marrow cells can contribute to endometrium and indeed endometriosis as well. This is a study where we used a DS red fluorescence and we looked at incorporation of bone marrow cells into endometriosis. We could see these cells from bone marrow incorporating not only into the endometriosis but also in the vasculature surrounding the endometriosis. So they did more than just contribute to those endometrial epithelial and stromal cells that we've classically defined as the definition of endometriosis. So this is a novel origin of endometriosis. And the stem cells can contribute to the disease. Probably accounts for some of the disease we see outside the peritoneal cavity. And although endometriosis is the only so far described disease resulting from ectopic differentiation of stem cells, I wouldn't be surprised if we'll identify some other diseases that are similarly caused by this mechanism. But not only could we find the cells there, but we could also find cells from the endometriosis we could detect in the circulation. So not only did these stem cells contribute to the endometriosis, the endometriosis continued to give off cells. And when we characterized these cells, we found that most of them actually were stem-like cells. So these endometriosis cells gave off stem cells. We started to look at where these cells landed. Where in the circulation could they end up? And what we call micrometastasis of endometriosis. Again, explaining the systemic nature of the disease. When we looked for endometriosis in multiple different organs, these are cells from our endometriosis implants in multiple different organs, we were able to find them in mice in small numbers. In the lung, in the spleen, in the liver, the brain. Every organ we looked at, and in every mouse we looked at, there were at least a few cells in these various organs. Again, the counts were very low. You can see a small percentage of 1%. They weren't anything that would ever be recognizable or clinically detectable for occurring in a patient. But could stem cells incorporating in other organs have an effect? We've always thought of endometriosis as a disease of cell trafficking. Cells moving from one organ to another. We've based our theory of endometriosis on retrograde menstruation, cells leaving the uterus, implanting in the peritoneal cavity. But I've just shown you cells from bone marrow can go to the uterus or to endometriosis and contribute to that disease. And then cells leave the endometriosis can become identified in other organs. And could this be having an effect in leading to some of those systemic symptoms that I told you about earlier? The other thing we looked at was micro RNAs. Micro RNAs are well known to influence gene expression. There are small RNA molecules, most of you are probably familiar with them, that are transcribed but not translated and extensively processed and basically bind the messenger RNA of other genes with a generally to decreased translation or degradation of that mRNA. And we've identified some time ago many micro RNA alterations in endometriosis itself. I won't go into this in detail but this is the let 7 family of micro RNAs which are decreased in endometriosis and more decreased in severe disease. But we know from several other models that micro RNAs can be secreted from a cell, can end up in the circulation either in exosomes or as free RNA and incorporated into other cells very remote from where they're made and influence gene expression in the recipient cells. So could the same thing be going on in endometriosis? We looked at circulating micro RNAs in endometriosis and found there are many micro RNAs that are either expressed at higher levels or increased abundance in the circulation of women with endometriosis or are decreased. And we asked could we even use this to make a serum biomarker? We did look at that where we identified six micro RNAs that at least in a training set as shown in this receiver operator characteristic curve in a training set we could really get a perfect correlation with who had endometriosis and who didn't. When we went back to look at an independent set of patients, the area under the curve wasn't perfect, but it's still pretty darn good. So potentially could be developed as a biomarker for endometriosis. But of course these micro RNAs aren't in the circulation waiting for us to develop a biomarker. We wanted to look at their function. So we looked at their effect on various leukocytes. Here I'm showing macrophages where we recreated the micro RNA concentrations and seen in the serum of women with endometriosis to ask what that would do to expression of inflammatory cytokines from macrophages. The two black bars show mimicking the micro RNA concentration seen in endometriosis, either an increase of micro RNA 125 or a decrease in let 7B, the two black bars. So we see an increase in TNF alpha, IL-1 beta, IL-6, IL-8, many of the inflammatory cytokines that we've seen associated with endometriosis. So through production of these micro RNAs we may very well be seeing a reason for the systemic whole body inflammation we see in endometriosis. The other thing we know about women with endometriosis is they tend to have a lower BMI. Now in the old textbooks you'd often see that being thin was a risk factor for endometriosis. Well in the animal models we can clearly show that it's actually the other way around, that when we create endometriosis these mice fail to gain weight and end up having a lower BMI, lower body fat. So endometriosis causes a lower BMI. Some of that is an effect on the liver, and I won't go through all this detail with you, but many of the enzymes in the liver that mediate metabolism are affected when we create endometriosis in the mouse model. So there is a metabolic phenotype of endometriosis. We also look at adipose tissue. On the left I'm showing that many of the micro RNAs that we found, again apparently expressed in endometriosis here, let 7B and micro RNA 342, change leptin, change adiponectin, change IL-8, hormone sensitive lipase in adipose tissue. So there is a metabolic phenotype that we can see in liver and adipose tissue directed by some of these changes in circulating micro RNAs. On the right I show you some changes in stem cell, adipose stem cell abundance in the adipose tissue of mice after creating endometriosis. So again there is a metabolic phenotype of endometriosis. We also looked at the brain and behavior. I mentioned earlier that women with endometriosis often have higher levels of anxiety, depression, behavioral sensitization of pain, they are more sensitive to pain. We have even seen it questioned in the past whether women with endometriosis are just more anxious and complaining more about the pain. No. What we show in our animal model, again the cause and effect relationship is that when we create endometriosis we cause changes in the brain which lead to changes in mood and behavior. This looks at animals, orange shows endometriosis in the brains of these animals. When we create endometriosis versus a sham surgery, I won't go through it all, but we see changes in gene expression in many areas of the brain, one shown here. We see changes in brain electrophysiology, so the synaptic potentials are different. But in the bottom we look at some of the changes in behavior. Anxiety, for example. In the mice where we create endometriosis they have increased anxiety. To measure anxiety in a mouse you put them in a cage and the anxious mice will huddle in the corners of the cage whereas the unanxious mice will walk freely around the cage and you can see that the time in the center of the cage is greatly decreased statistically significantly so within six weeks of creating endometriosis, yet the total travel time has not changed. They're spending more time huddled in the corners of the cage. They're more anxious as a result of the endometriosis. We look at depression. To measure depression we hold a mouse up by its tail and the depressed mice will hang there listlessly for a longer period of time before trying to get away and you can see the immobility time in the bottom right there is increased very quickly within four weeks of creating endometriosis. Then pain sensitization. To measure this we have a warm surface. We have the mice put the paw on the warm surface and see how long it takes them to withdraw the paw so that they sense that as an unpleasant sensation and you can see that duration before they move their paw within 12 weeks is significantly decreased in the mice where we've created endometriosis. So the cause and effect relationship here is that the endometriosis changes things in the brain, changes gene expression, changes electrophysiology which leads to changes in behavior and pain sensitization and we're working now showing that some of these at least but not all of them are mediated by those circulating micro RNAs. And finally something we just published a few months ago looking at endometriosis and atherosclerosis, again epidemiologically there's been an increase in cardiovascular disease in women with endometriosis reported. Women with endometriosis of course undergo many surgical and medical therapies, could that be a consequence of hormonal changes or is it really due to the endometriosis? Well here we created endometriosis again in our mouse model. Here we use the APOE knockout mice because mice just don't develop atherosclerosis like humans do, but in this model when we gave mice endometriosis compared to a sham surgery you can see that oil red o-staining on the left is increased in the mice in the aorta of the mice with endometriosis and on the right you can see the luminal diameter is decreased in the aorta of mice with endometriosis and the wall thickness is increased. So again the endometriosis is really leading to this increase in atherosclerosis. Mechanism for this we thought was there an effect on lipids? No they're not, lipid levels are the same whether we created endometriosis in these mice or not, but many of the inflammatory markers that are associated with cardiovascular disease IL-1 alpha, IL-8, interferon gamma, and VEGF were all increased in the mice where we created the endometriosis again leading to the atherosclerosis. So I hope I've convinced you during this talk that when we think about endometriosis, when we think about those little blue lesions in the pelvis, we're really only seeing the tip of the iceberg. That there are many systemic manifestations endometriosis really is a disease that affects the whole body and not limited to just the just the lesions in the pelvis. So concluding endometriosis is a systemic disease. The systemic nature of the disease explains the extensive symptoms that we often see in patients with endometriosis. We focus mostly on the pain and infertility but there are a whole constellation of symptoms that are very very common in these patients and I've showed you that ectopic differentiation of stem cells, micro RNAs, whole body inflammation are some of the mechanisms that mediate these long-range effects and it has pretty profound clinical implications. The varied presentations and these diffuse symptoms have made it very difficult for practitioners to diagnose endometriosis. Far too often patients complain about other symptoms and are sent to a urologist for their bladder symptoms, are sent to a psychiatrist for their depression, are sent to a gastroenterologist for a colonoscopy before they get their endometriosis treated. These are all extremely common and it's led to this delayed diagnosis of endometriosis. As our first speaker mentioned it's six to ten years from when someone has classic endometriosis symptoms till they're typically diagnosed. So I think really recognizing these constellation of symptoms that are directly caused by endometriosis may be very helpful in understanding the disease. The other clinical implication is that surgical therapy which removes the lesions may be great in treating that local disease but if we really want to get all these whole body manifestations, medical therapies may be necessary and indeed I think future therapies and some that we're looking into now may specifically target the effects, these disseminated effects of disease in other organs. So I'll just leave you with this one quote that focus exclusively on pain I think underestimates the true extent of all these consequences of endometriosis and our patient's full disability. I think understanding these extensive symptoms of endometriosis will help us really better understand what our patients are going through and give us the possibility to start to think about treating all of the manifestations of the disease. I'll end here by thanking my clinical collaborators in our endometriosis center and the people in my laboratory have done much of this work and are many collaborators. Thank you. Any questions for Dr. Taylor? I actually have one so thank you so much for that very extensive discussion. So usually our first line treatment for endometriosis is medical management with combined hormonal contraception. Have you looked in any of your work how some of these outcomes are modified if you put them on birth control pills for example? Yes it's very interesting because some of them as you might expect are much improved but not all of them. As a matter of fact we think specifically some of the brain and some of the metabolic phenotypes are not and we're finding some of these changes in gene expression are epigenetically programmed so they may be very well lasting even after the lesions in the pelvis are treated. So again I think the need for treatment of these individual manifestations of endometriosis in these remote organs is probably the future. Sam Ball from Emory University. Is there any relationship between the microbiome and endometriosis? There is yes but not well characterized to date but there are people who are looking at that and absolutely I think you'll find that that plays a role. Thank you. Hugh I've got a question for you. In the mouse model with the DS red where you're looking at the kind of sort of total body distribution of the lesions through this sort of metastatic or however you want to kind of think about it, the spleen was positive but pretty low and you know there's been this sort of anecdotal thing in clinical endometriosis that somehow the spleen is supposedly you know protected in endometriosis. Now I don't know how well established that is but I remember learning that as a student and I haven't really questioned it but it was kind of interesting to see. It was. It's one of the least populated I guess in your model. Yeah it is and I know that consensus pretty well that the spleen is somehow protected but I don't know the data really that underlies that. I'm not sure it's too strong but this data would be supportive of that absolutely. Yeah. Hi Margarita Psarska, Cedars-Sinai and it's the unseen disease because we don't see it unless we do surgery okay so just for clarification of the title but great talk by both of the first two speakers and obviously one of the common themes we're seeing is an inflammatory process that's going on and in both of the presentations and even some of the particular manifestations and associated disease processes such as cardiovascular events and the like are also inflammatory driven. Have you had a chance to look at some of the additional other inflammatory markers that might be contributing to the manifestations associated with endometriosis as well as in some of these other disease processes? Yeah absolutely first it's unseen because we don't look until we do surgery but also I'd suggest that it's unseen because even when we do surgery we miss all these systemic manifestations so in both ways unseen but yes we have and I showed with the atherosclerosis some of the inflammatory markers that we measured I showed in the micro RNAs affecting macrophages and their inflammatory cytokines that they produce other leukocytes similar effect which I didn't show and yes we're looking at other inflammatory markers in addition to the ones I showed here today. Are you also seeing them in the endometriosis itself or around endometriosis as well so it's in all areas? Yes. Okay great. Absolutely. Gary Shangold in Terrace Biopharma. Thanks for a nice talk Hugh. I have a question your underlying thesis for which you've presented lots of intriguing information about the systemic nature of the disease and the very unusual places that endometriosis can be contributing to systemic endometriosis can be contributing to systemic pathology invites the question how frequently is is there data to suggest how frequent is it that someone might exhibit evidence of unusual anatomic locations of the disease that are not visible as well in other words I guess what I'm trying to ask is how frequently do we look for endometriosis in the pelvis or abdominal cavity not see it and yet have evidence that there is systemic disease elsewhere yeah good question I don't think we know the answer to that because we don't have good ways to identify none of the stuff we saw in mice would be clinically detectable there's a single you know rare cell in these organs is it is it having an effect and what's the mechanism of that effect we're still working on trying to discern that right now but I think you're right there may be men if there may be cells identified in these other organs or an effect even when we don't see it in the pelvis hard to know my my bias has always been that we the lesions in the pelvis are there and then spread to other organs and create some of these systemic manifestations that then take on a life of their own but we don't know that for sure it may very well be that that there are cases of endometriosis where the disease is primarily affecting these in a clinically undetectable way affecting these other organs without having the lesions in the pelvis we do as surgeons occasionally find women who have pain that we think is classic for endometriosis and we look in the pelvis and nothing's there and is that really because they don't have endometriosis or is it because they have predominantly systemic manifestations without pelvic disease again at this point I don't have any way of knowing that for sure but I I think it may very well be true thank you okay thank you thanks you thank you and we'd like to invite our final speaker Dr. Kim professor at Northwestern I just want to thank the organizers for inviting me to give this talk and to put together such an excellent symposium I did learn a lot from the first two speakers today I'm going to talk about something a little bit different in terms of modeling the endometrium as well as its associated diseases and here I'll focus a little bit about it on the endometriosis project that we've been doing I have nothing to disclose and so the common theme in this talk is basically about modeling models do matter on how we experiment in the lab every model has strengths as well as its limitations and so we have to consider what is the research question that we're asking so I'm going to go over a little bit about some of the more advanced technologies that we've done in modeling the endometrium as well as endometriosis and then I'm going to talk about our most newest and unpublished data using induced pluripotent stem cells to study endometriosis and then I'm going to start a conversation about how we can recreate endometriosis in vitro and I'm hoping that you guys will help contribute to this conversation as we move forward so we can model and study various different tissues of the reproductive tract in vitro and people have done this in in many ways taking whole ovary from mice as you can see in this video and and watching follicles grow and ovulate etc we've also made 2d versus 3d renditions of the endometrium we always have to take into consideration as well what the hormones are doing and so trying to recapitulate a hormone cycle outside of the body in a dish can become a real challenge as well over a period of 28 days and then we have these microfluidic platforms now that can actually help us study the paracrine interactions between cell types between tissues and also be able to study tissues for over a long period of time without compromising longevity and so all we do these modeling to understand biology but we also do it to eventually test compounds and drugs outside of the body first to screen what would be the most effective drug compound to use and so we've gone through many renditions of model systems to study the endometrium specifically you can see in vivo all the way from humans to primates and even to the lower mammals and in vitro and we've we've we've learned so much from each of these models and in vitro we can use various different cell types tissue explants monolayers we've done that a lot with a new emerging technology of organoids now people are using 3d structures we can study the interaction between two different cell types depending on our research question and then we can have these platforms where we can study multiple interactions at the same time kind of trying to mimic what Hugh and Kate were showing about the more systemic effect of endometriosis but we all know that the more complex we get the more difficult the experiment and so there's also that yin yang that we need to consider oops and so we've done the simple monolayer cultures we've done them as 2d 3d we put them together and really the the goal the ideal goal would be to recreate an endometrium in vitro with all the different cell types of the endometrium and and so you know having the vasculature there as well as immune cells there i think would tell us a lot about the biology of the endometrium outside of a woman's body and so in my lab we've created a two cell endometrial organoid of epithelial and stromal cells in which we take endometrial stromal cells isolated from a tissue and then we put them in culture in in in certain ratios and it's amazing how these cells know how to organize by themselves in such a structure where the epithelial cells are lining around the internal stromal cells and so there is an architecture that forms in vitro and and then we can study the paracrine actions between these two cell types and so we've used cell different kinds of cell cell models but i also want to now tell you a little bit about the bioengineering that we've done to create some different platforms and this work started about 10 years ago when the tissue chip consortium was developed and this this project led by Teresa Woodruff who was the PI back then we and we put together a study a proof of concept that we can make a platform where different tissues of the reproductive tract can be cultured as well as they can communicate in a microfluidic system outside of a woman's body and so this was a really cool group project that we did and in in collaboration with engineers at Draper Labs who made us this platform we were able to do just that and so the platform that they made us was basically different compartments that you can put on this platform that are connected by microfluidic channels and so you can they can share media depending on what tissue you have and so it was very versatile very flexible and there was a as you can see on the right there's there's a compartment where you put fresh media it flows through and then you can collect the media at the end and so this was the first multi-organ platform that was designed for the reproductive tract and just to summarize a whole bunch of work because it's already been published we were able to put a mouse ovary fallopian tube tissue from human and we were able to create a 3D rendition of the human endometrium and we also created a 3D rendition of the ectocervix and then we put in a liver there for metabolic studies and so the cool thing here is that we can program a mouse ovary by adding LH or HCG and FSH in a pattern of a human menstrual cycle and so over the period of 28 days this mouse ovary released estrogen and progesterone in the pattern of a human menstrual cycle so that was what was driving the hormones for this platform and long story short we found at the end that even though each of these different tissues require specific media when you put them together they were happy in one media also tissue longevity increased because we had the constant replenishment of media and elimination of waste and we were able to go for a period of 28 days and these tissues responded to these hormones in a physiologic manner but and we also found that this functional response to hormones was greater when we put them all together kind of like they liked communicating with each other and so that was the first proof of concept that the female reproductive tract could be studied in vitro in a systemic manner and so we moved on to the next stage of this project in that we had to model a disease a disease that we could use our microfluidic platform and because it was a multi-organ platform we thought PCOS was the perfect disease to study as we all know PCOS involves multiple organs in the body not just of the female reproductive tract but in fat liver pancreas etc and so this is what we're currently doing but I'm not going to talk about PCOS today but we set out to model a disease but the first thing that we wanted to do was make a new platform to make it simpler to use as well as a cheaper to make and so that we can give it out to scientists that so that we can really study the biology and so this was kind of like the process of finding that cheaper more efficient solution we wanted to make sure that the material was good that it didn't absorb hormones a lot of the microfluidics in the field are made of PDMS which absorb hormones and other hydrophobic molecules and so that's that wasn't so we had to make sure that it was a good material we had to make sure that it was familiar to us as scientists so we looked at and modeled it after like a 12 well plate a 24 well plate the first thing we tried was let's 3d print something because that's really a cheap solution but we found out that the materials that we used to 3d print eventually became toxic to the ovary and the follicles so that was scrapped and so we decided to go back to polystyrene which is what all cell culture plates are made out of and we made a simple base station to create that pump and pump and flow and so I just want to show you a little video that tells us what this our system we call lattice can do we have a plate on top that has eight wells and each well can house whatever tissue construct that you want and we can run this for about 28 days we can go more and then each tissue secretes some factors that are then shared through these microfluidic channels that go to other wells which and this pump actuation is controlled by a computer and you can tell it where to go where the pump can go etc and so really it's a simple concept in that media is pumped through this this device which then switches into the different well and then shares it with another well and so very very flexible very versatile where media is flowing through different channels into the different wells using this system, we put a mouse ovary in there, and in the bar, in the red curve in the graph, that's the amount of estradiol that the mouse ovary produces. You can see that there is a peak at day eight that is higher than if the mouse ovary were to be cultured in static conditions. And so the dynamic conditions did promote estradiol production. And you can see on the right a video of a mouse ovary with developing follicles. And you can see on the outer edges, ovulation occurring of the follicles. And this is all captured within Lattice by putting a camera that takes pictures at different intervals. And this is basically a stitched model. And then this well, of course, especially if we want to image something every day, and we want to image the exact same place every day, we don't want these cultures to move around. And so we're kind of working on 3D printing some molds and some holders for the cultures so that they stay in place. There are different reasons why we use the molds. For example, here we're making one for the follicle so the follicle doesn't move too much. But we've been having a lot of fun 3D printing some of these molds and creating a cast so that we can have agarose molds in these. So a lot of bioengineering that goes on in creating the perfect system. And the final thing that I want to mention is that we wanted to make sure this microfluidic platform was robotically handled in that we can put this in a system where a robotic arm takes the plate and then transfers it to a microscope, takes a picture, and transfers it back daily or every six hours, et cetera. And so we made it compatible so that four of these units can be housed in this kind of holder, what we call the hotel, that we put in this robotic handler. And so it's never going to be a high throughput system, let's say, if we wanted to test drugs or whatever. But it is more than one. It's like 32 wells that we're studying here. And so that's basically some of the advancements that we've made using microfluidics, using different architecture, 3D renditions of the endometrium, et cetera. And now I'm going to flip to how do we recreate reproductive diseases. And I think this is essential because if you look at all the pathologies here of the female reproductive tract, there really is no permanent cure or treatment other than surgery, which is not always an effective treatment. And in my lab, we study fibroids, endometriosis, endometrial cancer, and polycystic ovarian syndrome. And even though we know a lot about these diseases so far, there's still a lot of gaps in knowledge. And this is because it's quite challenging to study reproductive diseases outside of a woman's body. There are physiological models in vivo and in vitro, but it's not always perfect. The human reproductive tract is quite unique. The thing that really frustrated me when working with endometriosis is the limiting yield of tissue we get from the clinic. We don't really get peritoneal lesions either because they're so small. So we're working with endometriomas. But there's always a challenge in getting those tissues from the OR. And the current in vitro models that we use is very simple. You grow them in monolayer, especially the stromal cells, epithelial cells. You can grow them, but they don't really grow. And there's just so little that we can do with those epithelial cells. And of course, the hormonal response is very complex and dysfunctional. And so this is what we know about endometriosis so far. I'm not going to go through this slide too much in detail because our previous two speakers did a wonderful, fabulous job in telling you what the advancements of endometriosis are. And thanks to the models that we have and thanks to all the work that has been done, we do know quite a bit about endometriosis as a disease. We do know that there are epigenetic factors that are associated with the pathogenesis of endometriosis. But we also do know that genetics matter too. There is a genetic component. And that is where I want to focus on right now. We do know that it is heritable. There have been GWAS studies that have identified loci in endometriosis. And there have been 18 loci identified so far in the current ongoing meta-analysis study. They've identified 27 significant loci so far. But really, the key challenge right now is to identify what the function of that loci is. Does that loci really do anything to the disease? And the challenge there is, how do you test it, except go one by one by one and mutate that? And then what cell types do you use? So we decided to take a completely different approach to understand the genetics of endometriosis. We decided to make some induced pluripotent stem cells from women with endometriosis. And just to go over what this technology is, is you can take any somatic cell, usually skin or peripheral blood cells. You can reprogram them by adding Yamanaka factors, SOX2, OCT4, CALF4, and CMYK. And that promotes pluripotency of that somatic cell. And then with those iPSCs, you can then differentiate them into whatever cell type that you want. And that does work. And so we've decided to take this approach to study the genetics of endometriosis to see, are there any differences right from the very beginning of women that have endometriosis or not? So just to describe this experiment that we did, we worked with clinicians to get blood PMBCs from women with stage 3 and 4 endometriosis, very well characterized, as well as women that have absolutely no endometriosis, no polycystic ovarian syndrome, just blood, so from volunteers. And then we were able to reprogram this at our stem cell core using Yamanaka factors to make iPSCs. And we have been successful in getting iPSCs made from four endometriosis patients and four control lines. We're still working on some two. And of course, the first thing we did with this iPSCs is to make endometrial stromal cells. And it's already been published in the literature by two different groups, the recipe or the protocol to make endometrial stromal cells. And so we decided to use the protocol by Chung et al, both researchers are from Northwestern, and basically is taking those iPSCs and treating them with these factors, as you can see. And at each time, day 4, day 8, day 12, they should be differentiating into the different cell types. And what we did was we verified whether they were differentiating into the proper cell type using PCR and markers. And so here you can see that OCT4 and SOX2 are the Yamanaka factors that are put in to make iPSCs. And we're hoping that they're not going to integrate into the genome. And that's exactly what we see, is when we take cells at the different points of differentiation, that the pluripotency markers are high at day 0, but then they eventually disappear. So that was great to see. We have other markers that we've tried. PAX2 at day 4 is a great marker for malarian duct mesenchyme. HOXA10, as we know, is a great marker of endometrial stromal fibroblasts. We see that PGR does go up as we differentiate them, which was great to see. And then we also took day 12 endometrial stromal fibroblasts and then decidualized them with a decidualization cocktail that everybody uses, MPA, cyclic AMP, for about three days. And we do see decidualization occurring as we see an increase in marker of decidualization, IGFBP1. I also wanted to show you that the purple lines are the lines from control patients, and the blue is the lines from endometriosis patients. And so we do see that they are differentiating both the control as well as the endometriosis lines, but we wanted to compare them at different stages of differentiation. And so we did RNA-seq at day 0, 4, 8, 12, and 15 in the four endometrial control lines, as well as the four endometriosis lines. And this is very new data, and it's unpublished. And so I just want to just highlight At first, you'd think, remember, this is DNA coming from peripheral blood mononuclear cells, right? And they're not coming from the tissue of endometriosis. These are blood cells that you've reprogrammed, and now you're differentiating them into endometrial stromal cells. And you'd think, well, if the genetics is the same, it shouldn't be too different. But we are already seeing differences in the gene expression of endometriosis. We're seeing differences in the gene expression profile of endometriosis versus control lines. It's not huge, but there are differences. And this is just a heat map of showing those differences in all of the time points together, endometriosis versus control. And these are a full change of 2 with a p-value of 0.05. This is a volcano plot. You can see that there are genes that are upregulated and genes that are downregulated in endometriosis versus control. There are a lot of genes that are downregulated, which is really interesting. And so if we compare control day 12 with endometriosis day 12, we do see differences in gene expression. Some of these are associated with pathways in proteasome genes, in inflammatory cytokines, increases in genes associated with platelets and erythrocytes, et cetera, and especially immune cells, neutrophils, et cetera. So already, even during differentiation, we're seeing differences in the endometriosis line versus the control line. And here are some of the differences when we try to decidualize those lines. We see an increase in CXCL genes. There are various pathways of genes associated. It's a decrease in pro-inflammatory cytokines. And so already, again, you manipulate these cells in culture, treating them with things that they should become endometrial stromal cells, and yet they are different. And so this is really telling us that whatever genes that they have at the very start, they are influencing gene expression as they differentiate. And so this last part, I just want to think together about how we can recreate endometriosis in vitro with all the tools that I've just shared with you so far. And so iPSCs, I think they're a really great tool. They're a really great tool to start understanding the genetics of endometriosis and how that influences ultimate differentiation into the cell type, whatever cell type it is that we want. We can manipulate those iPSCs at the very, very beginning too. If we find that, oh, there's something wrong with this particular gene, we can either put it back normally, or we can actually mutate certain genes that we want to see what happens during the differentiation process. Already, we've put in GFP just in the iPSC so we can kind of track where they go and what they become. And there are certain gene candidates that we're looking at that we can CRISPR in and out. So it's a great tool because these cells are also, you're able to freeze them down and thaw them and propagate them. So they're not an endless resource, but they are a resource that you can keep and study for a while. So the exciting thing is that we did endometrial stromal cell differentiation, but now we can differentiate these into different cell lineages. We're really interested in the immune component. Are these macrophages any different? If we differentiate them into macrophages, do they behave differently at all? What about neurons? We can differentiate these iPSC interneurons. Are the neurons from the endometriosis lines behaving differently at all? And then what happens if you put them all together, either in the same construct or in a plate, like in a microfluidic platform? So those are some of the things that we can do to really start understanding endometriosis from the very, very early stages. The other thing that we want to try, and this is just for comfort for us, is that we want to use our lattice system to kind of semi-automate this differentiation process. For those of you who do work with iPSCs, you know how laborious it is, how you have to babysit them every single day, how long it takes. You mess up in one time point, and it's all garbage. And so we want to use the lattice system And so we want to use the lattice system to be able to replenish that media, to be able to do it in a more gradual way instead of just stark, taking off media and then treating them. So there's a lot of things that we can try. Maybe we can speed up the differentiation process or at least just make it easier. And so we're using lattice to do this. And so to summarize the talk, models do matter. If we use novel approaches, we will get novel data. We can start asking novel questions, especially for endometriosis. I think it's time to really look at the biology in the lab, but in ways that really make sense. And so there are different ways of culturing cells, whether it's from primary or whether it's from iPSCs. We can put them in different architectures. That's going to give you different answers as well. We can put them in static culture. We can put them in microfluidics. There have been studies that showed interactions between endometrial stromal cells and endothelial cells in a microfluidic system. And we've learned a lot looking at the effect of flow, for example. And then, of course, we'd love to test drugs first in a system, not just one cell type, but in a system that includes different organs. As we heard from Hugh, it is a systemic disease. And so I think it would make more sense if the liver was there, if other cell types were there to study the efficacy of a drug. And so I want to acknowledge the people that have been involved in this project. It was a real collaborative project, especially for creating the tissue chip. Involved folks from Northwestern, as well as UIC, and the Draper Labs, the engineers, and our collaborators from Rutgers University. And in terms of the iPSCs, it was really a very fantastic graduate student now who took a gap year in my lab before going to graduate school, Hannah McDowell, who was able to work with all four lines. So all eight lines together within a span of a year. And Cassandra is a master's student who's working on lattice in the iPSCs. And Han is Campo, a postdoc, who worked in developing lattice, et cetera. So I want to thank my lab, as well as a stem cell core, as well as a reproductive endocrinologist, of course, that gave us the samples and that I'm able to have a conversation about clinical endometriosis with. And of course, the funding, this was funded by the Friends Apprentice Northwestern Foundation, as well as the Endometriosis Foundation. So with that, I'll take questions. So Julie, really, really nice. So some insight on making endometrial epithelial cells from iPSC cells? That's like the. Yes, I'm glad you asked that. We are definitely working on that. And there is no protocol, as many of you know. So we had to go back to developmental biology. And of course, I looked in the literature, talked to Richard Berenger, as well as Jerry Cunha. And we've decided to just put epithelial, put malarian duct cells, so at day four, with adult stromal cells. And we are actually seeing differentiation of those cells with epithelial markers. So we do have some. I'm not comfortable saying that there are actually epithelial cells. But I'm trying to get some funding so that we can continue on with this work. But yes, yes, we're going through that route. Hey, Julie, wonderful talk. I just was curious, your iPSC cells you showed decreases some of the inflammatory cytokines? Yeah. Sort of surprising. We normally think of endometriosis as inflammatory, making more. When you use some of your modeling to look at effects in other cells, did you see an effect on leukocytes macrophages from the interaction with the endometrial cells? No, no. I haven't done that yet. No, no, no. This is very new. And there's so many things I want to try now, especially after listening to your talk and Kate's talk. There's just so many interactions that could be important and experiments designed that way. Yeah, absolutely. Thank you. Gerilyn Pryor, Vancouver. I was concerned in the original data that you showed us that the mouse ovary produced progesterone, but that the estradiol during the luteal phase was extremely low. And that does occur in some animal like rhesus monkeys, for example, but it's not a human model. Yeah, yeah. Yes, we are recreating that spike in estrogen and the spike in progesterone, but like I was saying, not every model is perfect. And so we have to take into context what we have and interpret it accordingly. Yeah. Hi, Julie. Very nice. So I was wondering if the cells are terminally differentiated, and would you be able to de-differentiate them by inhibiting BMP2 with a BMP receptor antagonist? That's a great question. We just have to try it. Right now, we're just trying to see if we can freeze each of the different stages so that we can resurrect them and study them that way. We're trying to find out if at day 12, when we do get endometrial stromal cells, whether they stay that way, because endometrial stromal cells, as you know, are progenitor-like as well, how long they will stay endometrial stromal cells, whether we can freeze them and thaw them, et cetera. So those are experiments that are ongoing. Thank you. Julie, I had a question about the kind of the steroid spike in the mouse model. So what dictates that is, are you actually losing follicles so that it's actually, what brings the spike down as opposed to just sort of more and more and more production of steroids? Does it come down because it's kind of a rate-limiting, the follicles are dying, essentially, and it creates that type of a spike? Yeah, that's a great question. And as you know, I'm not an ovarian biologist, but you know. Nor am I. I have the same question about COVID. What causes the curves to come down? I understand the upswing, but it's a kinetics question. I think in this particular model, the gonadotropes are really dictating how the follicles are growing. And that down that we're seeing, we're not really maintaining the FSH there. So it could be that. That was a very clumsy answer to your question, but I can refer you to an ovarian biologist. You can talk to Teresa Woodruff. How's that? Thank you. OK, well, thank you very much. Great. Thank you.

endometriosis

diagnosis

treatment

surgery

BME study

gene markers

therapeutic targets

stem cells

circulating microRNAs

models

cell differentiation

research