Thank you for attending the session. And those of you watching online, we have a series of very exciting talks today in this late-breaking abstract session. The first speaker is going to be Joshua Gross from Duke University. And he's going to be telling about a very interesting story involving the ghrelin receptor. So Joshua. Start. There you go. Thank you very much. Can everybody hear me well out there? I'm plenty loud, so I don't think that'd be an issue. So thank you for the introduction. I'm a postdoctoral researcher at Duke University. And I work in the Department of Cell Biology there in Mark Carone's lab. And today, I'm going to be telling a short story that we actually just recently published. But ultimately, what our lab is interested in is functional selectivity, or another term for it is biased signaling of G-protein coupled receptors, or GPCRs. And one particular GPCR we're focused on in the lab is the ghrelin receptor, because we think it's quite an interesting and potentially quite useful pharmacotherapeutic target to normalize diseases related to metabolism. I have nothing to disclose today. So I first just wanted to start kind of a bird's eye view of some basic ghrelin physiology 101. I'm sure many of you are already familiar with this, being at the endo conference. But just briefly, you probably have heard ghrelin referred to as the hunger hormone. And it was discovered as recently as 1999. So actually, in scientific years, relatively recent. And very quickly after that discovery, it was identified as a potent pre-prandial driver of food intake, or in other words, it's orexigenic. And if you track ghrelin levels across a day's time in the plasma, you can see these very well-defined peaks and troughs. And if we dissect a little bit more deeply what's happening in those rise phases, we see that the ghrelin peptide is synthesized and secreted from the gastric fundus of the stomach and enter endocrine cells. And it can communicate orexigenic signals to the brain via two distinct mechanisms, one of which is via its receptor, called GHSR1A, on vagal nerve afferents, that's synapse in the brainstem. The other means is by activating its receptor in the hypothalamus. Interestingly, both the brainstem nuclei and hypothalamus are lined by fenestrated capillaries and thus have a very leaky blood-brain barrier. Ghrelin as a bulky peptide can readily get into that area. But in other areas of the brain protected by the blood-brain barrier, such as midbrain dopaminergic nuclei, where it regulates the hedonic aspects of food intake, ghrelin peptide doesn't get in there, but the receptor is very highly expressed. There are two predominant circulating species of ghrelin, one that is acylated and one that is not acylated, referred to as acyl ghrelin. And acyl ghrelin is the one that's active at the receptor, which will be the focus today. So the ghrelin receptor is a canonical GPCR, seven-transmembrane receptor that's expressed on the cell surface. When ghrelin binds the receptor, it induces a conformational change that drives guanine nucleotide exchange and subsequent signaling downstream. The ghrelin receptor predominantly couples to the G alpha Q pathway, which through the activation of phospholipase C can mobilize intracellular calcium and also produce diacylglycerol to activate various metabolic and nutrient-sensitive kinases. But like other GPCRs, it also very readily recruits beta-arrestin, which is a scaffolding protein that initially can desensitize the receptor through steric hindrance of the G protein, can also scaffold endocytosis to internalize the receptor and activate also spatially and temporally distinct signaling relative to G proteins. And so you can classify ligands or drugs that activate the receptors either unbiased or balanced if they activate both pathways. It can be G protein biased if it preferentially activates G protein, or it can be beta-arrestin biased. In our lab several years ago now, I actually discovered two mutations that are contiguous or immediately adjacent to one another in intracellular loop to the receptor, which is the G protein and beta-arrestin coupling interface. And this can actually produce a completely G protein or barrestin biased receptor, which supported that maybe we could discover drugs to recapitulate that behavior. So just very briefly, there is quite a bit of existing literature evidence to support that functional selectivity or biased signaling at the receptor occurs in vivo. And the canonical physiological effects of the ghrelin receptor, such as hyperphagia and secondarily weight gain, growth hormone release, and insulin ostasis, there's quite good evidence to support that this is all G protein mediated through the GQ pathway. Whereas some evidence from our group supports that some of the addictive behaviors or hedonic reward-seeking behaviors is mediated through the beta-arrestin pathway. Some weaker indirect evidence or hypotheses support that things like gastric motility and perhaps the cardioprotection conferred by ghrelin receptor activation could be beta-arrestin mediated as well. But what we decided to do was to screen a library of small molecule compounds to try and find biased ligands or biased drugs at the receptor to develop both research tool compounds to decipher exactly the physiological effects that bias are mediating, as well as potentially some new drug candidates. So we did this in collaboration with NCATS, part of the NIH. And we identified a lead compound called N8279. And we verified the structure of this through x-ray diffraction. We then did a functional screen of this compound across the entire GPCR using the UNC Chapel Hill PDSP core facility. And we found that the ghrelin receptor was the only significant hit for this compound, supporting complete target specificity. And then some work in our lab supported that 8279 could displace the endogenous ligand ghrelin, albeit with a relative affinity of about 500-fold less than ghrelin, which we'll see has an explanation in the subsequent slide. So then when I joined the lab, my role was to characterize this compound with respect to the molecular pharmacology and biochemistry. So I started with a simple calcium mobilization assay, which is downstream of GFQ activation. What we found here is that 8279 was 10-fold more potent than ghrelin, and only about 3- to 5-fold less potent than very high affinity small molecule unbiased ligands MK0677 and L585. Additionally, if you treat cells concomitantly with 8279 and a concentration responsive ghrelin, you see that there's this additive augmentation effect of 8279, supporting that it can actually increase ghrelin-mediated G-protein signaling. We also tested this more proximately to the receptor using a BRET-based assay for G-protein dissociation. Here, 8279 had a potency that was almost exactly equivalent to its calcium mobilization and was about equipotent to ghrelin and L585. And similarly, we could recapitulate ghrelin-induced augmentation of G-protein signaling with concomitant treatment. When we tested beta-arrestin activity, we found exactly the opposite, which is that 8279 had a considerably weaker potency and moderately lower efficacy than ghrelin or L585. It also, upon pretreatment of cells that are stimulated with ghrelin and measuring for arrestin recruitment, we could see that 8279 could actually inhibit this effect, which is contrary to what we saw in G-protein pathways. And then using a titration assay, we found that 8279, at least relative to ghrelin, could induce a rightward shift of beta-arrestin coupling to the ghrelin receptor. And this would support that 8279 is stabilizing a confirmation of the receptor that decreases the affinity or the ability of beta-arrestin to couple to the receptor. So this altogether supports that 8279 is a truly biased compound. And this can be quantitated using the relative intrinsic activity model of signaling bias. And what we found here is when you use assay standardized conditions and you measure actual transducer coupling of G-protein versus beta-arrestin using those proximal assays, that 8279 has anywhere from about a 14 to 23-fold G-protein bias relative to barrestin. And this is in comparison to ghrelin. If you use other reference ligands, such as L585 or MK0677, you get about a similar full change here. And importantly, 8279 is very readily brain penetrant in mice. And so we did this in collaboration with NCATS. We found that at a relatively low to moderate dose of 8279 into C57 black 6J mice, that 8279 very rapidly entered the brain at pharmacologically relevant levels and maintained itself there with a half-life of about seven hours, which is generally advantageous with respect to a clinical drug candidate. And this is important because, as I alluded to earlier in the talk, many ghrelin receptor agonists are actually very poor at activating ghrelin receptor in blood-brain barrier-protected areas, such as these mesolimbic dopamine circuits. So with respect to modulating the receptor in these hedonic reward circuits, agonists, let alone biased agonists, have been difficult to develop. So we have a candidate here to try and address that limitation. And lastly, I just wanted to summarize the characteristics of this particular compound. So it has target selectivity for the ghrelin receptor across the whole GPCRM. It's functionally selective for G-alpha-Q relative to many other G-alphas, which also will couple to the ghrelin receptor, albeit less than GQ. And this is on the order of potency and efficacy differences anywhere from about 30 to 100-fold. So it's quite selective for GQ. As I showed, it's G-protein-biased over arrestin when you compare it relative to the endogenous ghrelin. This is about 20-fold on average for proximal transducer coupling measurements. And when you measure downstream or compare downstream data, it gets up to about 100-fold. It has some biased ago-allosteric modulator-like properties in that it can potentiate ghrelin-mediated GQ signaling and actually inhibit ghrelin-mediated barrestin coupling. But there's additional biochemical and biophysical studies we're doing now to confirm that mechanism. It also, which I couldn't show today, has a very distinct receptor binding profile, at least based on mutagenesis and molecular docking studies, supporting that 8279 preferentially binds at the very superficial parts of the receptor and requires ECL2 in particular in a manner that's quite distinct from ghrelin and other unbiased ligands. It has excellent brain penetrance in mice. And I didn't show it today, but it is published in the recent paper. It has efficacy upon systemic administration in multiple models of hyperdoveminergy in mice. And lastly, I just wanted to summarize, this is only but one story in which we're engaged in the lab. We're mainly interested in developing multiple tools. So we use ligands to study the ghrelin receptor. And 8279 is one of those, but we're doing considerable SAR around this scaffold in order to find a better compound. We've also designed mice that have biased receptors from birth. And so we can use this as a parallel strategy is to discern how does bias contribute to ghrelin physiology. And we're also studying receptor oligomerization, particularly ghrelin receptor homodimers and heterodimers and how that may affect bias. But ultimately, what we're interested in is using these tools to study ghrelin receptor bias in order to advance drugs towards IND. And we think that this really has a multitude of potential applications, as I alluded to earlier in the talk. And I just want to thank everybody in the lab, particularly my mentor Mark Carone, who actually just recently passed, but was an absolutely wonderful mentor, as well as Larry Barrett, who's played a major role in this project. So with the time I have left, I would love to take any questions if you have them. So this talk is open for questions. Please, if you go to the microphone, identify yourself, which I neglected to do at the beginning. I'm Joel Elmquist. And I'm the director of the Center for Hypolemic Research at UT Southwestern. So I'll ask the first question. Is there any evidence for heterodimerization of the receptor with other GPCRs? Absolutely, yeah. So on the last slide there. So the one that we're most interested in is the D2 dopamine receptor, but the ghrelin receptor in particular has been shown to be a promiscuous heterodimerizer. So there's many, many GPCRs that have been shown to dimerize it. This is almost exclusively restricted to in vitro studies. But yeah, this is something we're very much engaged in studying. And what about the effect of your compound on the constitutive activity of the receptor? Yeah, so it does not appear to have any effect on constitutive activity in either pathway. It seems to be only in agonists or ligand-stimulated conditions in which we see changes. Great question. And then another final question for me would be, what indication or disease do you see this compound potentially being effective in? Yeah, so we originally developed the compound specifically for dopaminergic brain disease, so addiction and also Parkinson's disease. And so we still think there's utility there. But as we've discovered more, particularly with all the signaling assays, we think it could potentially be useful for things like diabetes, some peripheral metabolic diseases. So we're very much engaged in repurposing the compound as we discover more about how it works. Thank you. Any other questions? If not, thanks again to Josh. Thank you. OK. Sure. OK, our next speaker is Junichi Okada from Albert Einstein College of Medicine in New York. And he's going to be telling us about his work on temporal and spatial cellular heterogeneity of a parasite metabolism. Jun. Good morning, everyone. My name is Junichi Okada. I'm a research fellow at the Albert Einstein College of Medicine. And today, I would like to talk about the temporal and spatial cellular heterogeneity of hepatocyte metabolism. I would like to start off with talking about the structure of the liver. As you can see, the liver is made out of hexagonal lobules. And each lobule has a central vein and a portal triad, which is the portal artery, the portal vein, and the bile duct. And the hepatocytes are aligned to these veins. And it is told that the hepatocytes surrounding the central vein are called the pericentral hepatocytes. And the hepatocytes surrounding the portal vein are called the periportal hepatocytes. Now, studies have been shown that between these two, pericentral and periportal hepatocytes, that they have different metabolic functions. For example, pericentral hepatocytes are known to be good at glycolysis and lipogenesis, whereas the periportal hepatocytes are known to be good at gluconeogenesis and beta-oxidation. However, when we talk about metabolic function, you also have to consider what the feeding and fasting conditions are. And there hasn't been any study about how these functions changes throughout the feeding and fasting, the spatial and temporal metabolic function differences. And so we wanted to do a fed and fasting time course to analyze the temporal and spatial heterogeneity of the hepatocytes. So we used a C56 black 6J mice, 12-year-old male. And we fed them. And we took the food out and then started fasting them. So zero hours would be the fed state. And then we take out the food. And we have a fasting four hours, eight hours, 16 hours, 24 hours, and 30 hours. Now, how we would discover these differences, we focused on gene expression, because we think that gene expression is very important in terms of long-term regulation in metabolic functions. And also, another thing is nothing is more precise than single-cell techniques. So when you have qPCR data, when something is high, you don't know whether the whole cells are expressing high or whether a subset of gene cells are expressing high. So what we wanted to do was first do a single-cell RNA sequencing, so a single-cell RNA seq, which we have the liver tissue, isolate individual cells into hepatocytes, isolate the RNA, and then sequence so that we get thousands of gene information per each cell. And you get these cell clusters. And the cell clusters are aligned based on their gene expression profile. So it's more of like when you think of a fruit salad, you see that the raspberries are next to the strawberries. The grapes are next to the blueberries. They just are aligned together based on how they look, the characters and stuff. So they don't give us the spatial information that we wanted. So what we also did was we also did a single-molecule fluorescence in situ hybridization as in fish, which we also have the liver tissue. We fix them, hybridize them, we wash, and then we image them. So this gives us the, although this is only a few genes, this gives us the spatial context of the tissue as well as we see these transcription sites so that we can also analyze. So this is compared to the fruit salad. It's more like a fruit tart, where now you see the strawberries are lying next to the blueberries, and those two are making the edge of the tart. So we use these two techniques. So these are complementary techniques. So combining these techniques, we tried to explore our questions. So I like to start off with the single-cell RNA-seq data. So this is the UMAP of our data. We had five libraries we sequenced, starting from zero hours, which I said was the fed. And we had four fasting time points. So as I said, that the spatial part we don't know, which we can't get from the single cell. So we have to use specific zonation marker genes that are known to be high in the pericentral or the periportal. For example, CYP2U1, GULO, are known to be highly expressed in pericentral. So as you can see in the UMAP, the upper part of the UMAP are highly expressed with these markers. So we designated them as pericentral cells, whereas CYP2F2 and HSD17B13, which are periportal markers, are expressed in the lower part of the UMAP. So we called them the periportal cells. So now we can compare the pericentral and periportal markers, the hepatocytes. And I want to start off with showing you the gluconeogenic genes. So as you can see, PCK1, in the fed state, you would assume that the gluconeogenic genes would be off. But in a subset of the periportal hepatocytes, you see that the PCK1 is already positive. And when you have a short-term fasting, you see the periportal cells first express gluconeogenic genes. And then in the starvation mode, you see the pericentral cells now express the PCK1. So in the previous, in the first slides, I showed you that the periportal cells are good at gluconeogenesis. But actually, depending on which fasting time point you look at, both periportal and pericentral cells are responsible for gluconeogenesis. It's G6PC, you can see the similar trends. Fed state, you already see some of the subset of periportal cells expressing G6PC. And then the portal cells come up. Compared to that, when you look at the lipogenic genes, in the fed state, both the pericentral and periportal cells are expressing, you can see a significant amount of positive cells in both regions. But as soon as the fasting starts, you see almost no expression of both ACLY and FasN. So depending on what metabolic genes you look at, they have different characteristics. Now I'd like to go to the SMFISH data. So this is the lobule. And on the left, you see the central vein. Right is the portal vein. And these, when you look at the magnified images, these blue circles are the dapisanes, so these are the nuclei. And each pink dot you see are the transcription sites. So you can see that each dot, when you see a dot, it is actively transcribing PCK1, for example, in this image. In the FET state, you can see, it's hard to see maybe, but in both the periportal and pericentral hepatocytes have somewhat of transcription sites. When you have a fasting four hours, you see the transcription sites getting brighter, you see more transcription sites per nuclei, and you see the majority in the portal. And then in the fast 24 hours, you see both periportal and pericentral hepatocytes have a significant amount of transcription sites. When we quantify them, based on the percentage of transcription positive nuclei, initially you see the portal cells, but somewhat of the central cells too, express transcription sites. Short-term fasting, you see the periportal hepatocytes express transcription sites. The long-term, the pericentral also catch up. When you look at FasN, which is a lipogenic gene, as I showed, these are the transcription sites. You can see both the pericentral and periportal cells are expressing transcription sites. But as soon as the fasting starts, at fast four hours, you see no transcription sites, which is consistent in the long-term fasting too. And when you quantify that, you see that at the FET state, you see both periportal and pericentral hepatocytes expressing transcription sites, but as soon as fasting starts, you see no transcription sites. So both the single-cell RNA-seq data and the FISH data suggests that in the FAST state, lipogenesis genes turn off fully, as you see here, whereas in the FET state, gluconeogenic genes remain in idling mode, meaning like even in the FET state, you see a subset of periportal cells expressing gluconeogenic genes. And so there are differences between these two types of metabolic genes. So in conclusion, I would say that hepatocyte donation of metabolic gene is dynamic and can vary substantially depending upon the facet and FET state. And these findings provide the molecular basis of determining the dysregulation and insulin resistance states. Last, I'd like to thank William Kosaku, as well as the Genomics Corps for the single-cell RNA-seq data, Austin Carroll for the FISH, and as well as my colleagues in my lab. Thank you. So this talk is open for questions. So I have a question about what do you think in the physiologic significance or context in which this zonation response is active? Thank you. Thank you for that important question. I think in terms of comparing when you compare lipogenic genes and gluconeogenic genes, for example, gluconeogenesis is producing glucose. When you have a fasting state, initially I think the glycogen breakdown starts providing glucose to the body, but as soon as the long-term fasting starts, you need the gluconeogenic genes to start going and make the glucose. So I think keeping the gluconeogenic genes in idling mode and also initially having the portal cells express it, but when you have to really, when things get serious, you want to start immediately, plus you want to have both hepatocytes activating it, which is why I think the gluconeogenic genes have this more interesting dynamic response, whereas the lipogenic genes is more of like an on-and-off switch. Okay, thank you. Yes? Manuel Gadu from Dresden, Germany. So I have a technical question. So when you are talking about the FISH staining and in-situ hybridization, how did you determine which portal and the different... Portal and central, yes. Thank you for that question. So there, as well as the single serotonin sequence where I showed you the gluconeogenic, excuse me, the periportal and pericentral markers, there are also known FISH staining markers, for example, GLU-L, which is expressed specifically in the pericentral region. So we added those markers to our staining so that when we look at the microscope first, we can know which are the central cell, pericentral, the central vein. Portal vein also, when you look at, when you make an observation of the portal area, you can also see the bile duct and the portal, you see the triad. So that also gives us an idea of whether it's the portal, periportal area. Thank you. I have a technical question as well. Okay. What are the inherent limitations of the sensitivity of your single cell FISH? Okay, oh, single cell. Not the single cell, the, yes, single. RNA-seq, yeah. So in terms of- My question is, are you missing things? Yes, so when you have a single cell RNA-seq data, you submit it to and sequence it, but you only have, you have, based on how much money you have, I guess, you can sequence a certain amount of genes or data. So you can't really sequence the whole gene and the whole amount. So it is, so when you look at like data like this, it is hard to see the quantitative, how you can quantitatively analyze the single cell data. But so the technique that we use is called a special technique called the targeted gene expression, which we choose select amount of genes that we have interest in, and just we saturate those genes so that in terms of the genes that we are looking for, we know that those genes are saturated so we can actually quantify the genes. Okay. Any other questions? Thank you very much, June. That was great. Thank you. Thank you. Okay, the next talk is from Diana Monsivius from Baylor College of Medicine, and the title of her talk is Uterine TGF-Beta Signaling Controls Endometrial Cell Homeostasis and Regeneration. Okay, good morning, everyone, and thank you to the organizers for the opportunity to present my work. So the endometrium is a highly regenerative tissue with the potential to undergo hundreds of cycles of proliferation, regeneration, and breakdown throughout a woman's reproductive lifespan. It's under the cyclical control of the steroid hormones, estrogen and progesterone. Estrogen, which is released during the proliferative phase of the endometrium, which stimulates the mitogenic potential of the tissue, and progesterone, which is secreted following ovulation by the corpus luteum, which stimulates the differentiation of the tissue. In the absence of a pregnancy, the withdrawal of progesterone stimulates a pro-inflammatory reaction, which then stimulates matrix metalloproteinase expression, leading to the breakdown of the endometrium and to the subsequent release of the tissue through menstruation. So it's long been hypothesized that the regenerative potential of the endometrium is due to the presence of stem cells residing mainly in the basalis endometrium, which is a portion of the endometrium that is retained following menstruation. However, recent studies have indicated that the actual location, as well as the signals that are driving the regenerative potential of the endometrium, are still unknown. And so this is still an active area of investigation. So the mouse endometrium has a similar profile to that of the human. So it's still under the cyclical control of the steroid hormones, estrogen and progesterone, but some key differences are the length of the cycle. So they have a five-day estrous cycle. And the mouse endometrium also does not go menstrual breakdown. Instead, it's resorbed in each cycle due to a process that involves endometrial resorption. And this process involves apoptosis of the cells and possibly even autophagy. And so another example of the regenerative potential of the mouse endometrium can be seen in the postpartum phase, where the entirety of the endometrium is regenerated within a 24 to 48 hour period in the postpartum phase. And so similar to the situation in the human, the actual identity and factors that are leading to the stem cells repopulation of the endometrium are still being investigated. However, recently, lineage tracing experiments have identified that LGR5, Xn2 and Pax8 positive cells reside in the crypts of the glands and are potential stem cells that are driving the regeneration of the endometrium, along with signaling molecules such as Wnts and possibly Notch being important factors. So in our previous studies, we found that signaling of TGF-beta through the type one receptor ALK5 was important for the endometrial regeneration in the postpartum phase. And we found this by performing a unilateral oviductomy and controlling ALK5 conditional knockout mice, and then allowing the mice to undergo repeated cycles of regeneration by exposing only one uterine horn to pregnancy. At the end of the extended mating period, we found that only the uterine horn that had been exposed to multiple cycles of regeneration developed this endometrial defect, where the endometrial glands were mixed in with the endometrial stroma, suggesting that TGF-beta signaling through its type one receptor ALK5 is important for driving the endometrial regeneration of the mouse in the postpartum phase. So to study this further, we decided to generate a mouse model that contained conditional deletion of the downstream factors of the TGF-beta signaling pathway, the SMAD2 and the SMAD3 transcription factors. And doing this would allow us to uncover the cellular mechanisms that are driving endometrial regeneration. So to do this, we obtained mice with SMAD2 and SMAD3-bloxed alleles, and mated them to a lactoferrin-CRE expressing male. This allowed us to generate our control mice, which had floxed alleles of SMAD2 and SMAD3, and then our conditional knockout mice, which had the two SMAD2 and SMAD3-floxed alleles plus one copy of the lactoferrin-CRE. Here we showed that we could effectively obtain a deletion of SMAD2 and SMAD3, both at the mRNA and at the protein level in the purified uterine epithelium of these mice, of the conditional knockout mice. A long-term analysis of the conditional knockouts identified that the SMAD2-3 conditional knockout mice developed endometrial tumors with metastases to the lungs, and that most of the SMAD2-3 conditional knockout mice perished by about nine months of age. When we analyzed the tumors histologically, we found that both the endometrium from the control mice and from the tumors of the SMAD2-3 conditional knockout mice expressed estrogen receptor. However, when we looked at progesterone receptor, we could see that the control expressed, the control mice expressed progesterone receptor both in the stroma and the luminal epithelium. However, the SMAD2-3 conditional knockout mice lost progesterone receptor expression in the epithelium. However, they still retained it in the underlying stroma. This suggested that the endometrial tumors in the conditional knockout mice responded to the signals of estrogen through ER, but were unable to respond to the anti-proliferative effects of estrogen because they lacked PR in the epithelium. So to study this further mechanistically, we decided to culture the endometrial epithelium from the control endoconditional knockout mice as 3D epithelial organoids. So we purified the uterine epithelium and then encapsulated it in matrigel and grew them as 3D organoids in the presence or absence of the TGF-beta-L457 inhibitor in the endometrial epithelium. So this is the TGF-beta-L457 inhibitor, A8301. This allowed us then to examine how the epithelial organoids would grow when they had either a genetic inactivation or a pharmacological inactivation of TGF-beta signaling. And so what we found by looking just at the morphology of the endometrial organoids was that the control organoids grown under the vehicle conditions or a genetic or the pharmacological inhibition of TGF-beta signaling resulted in this altered endometrial morphology, organoid morphology, where the organoids became more dense and glandular looking in shape. When we analyzed the organoids histologically at the cellular level, we could see that in fact the organoids that had either the pharmacological or the genetic inhibition of TGF-beta signaling did have more of a secretory-like appearance here. And when we stained the organoids with Mucin-1, we could see that in fact the organoids from cultured with pharmacological or genetic inhibition of TGF-beta secreted more Mucin-1. And then they also expressed higher levels of FOXA2, indicating that suppression of TGF-beta pharmacologically or genetically led to more of a glandular differentiation in the endometrial organoids. So to get a better idea of the signaling pathways that were being altered in response to the inhibition of TGF-beta, we performed RNA-seq of the organoids from the three different groups. And we identified by doing gene ontology analysis of the overexpressed genes and the down-regulated genes that we had an over-representation of genes involved in retinoic acid signaling, BMP signaling, as well as in Wnt-beta-catenin signaling. This here is a volcano plot that shows in red the overexpression of the genes involved in retinoic acid biosynthesis that were overexpressed in the SMAG2-3 conditional, in the organoids from the SMAG2-3 conditional knockout mice. We also saw increase in BMP signaling. So these are genes that are activated downstream of BMP signaling, like ID-1, ID-3, and ID-4. And then a down-regulation of genes involved in Wnt-beta-catenin signaling, such as Frizzled and Wnt-9A. We validated the fact that some of these genes involved in retinoic acid signaling were, in fact, overexpressed in the endometrial organoids that were cultured with inhibition of TGF-beta with the pharmacological inhibitor, and also with the genetic inhibition of TGF-beta signaling. So, again, confirming the fact that we had an activation of the retinoic acid signaling pathway when we suppressed TGF-beta. So here we show, so it's known that the aldehyde dehydrogenase enzymes are important in converting retinol to retinoic acid, and that retinoic acid binds to the RAR and RXR transcription factors to lead gene expression. And so the gene expression of this complex has been shown to be important in maintaining stem cells, as well as driving the regeneration and differentiation of many tissue types, such as the breast, the testis, and the kidney. And when we performed immunostaining of these enzymes of ALDH1A1 and 1A3 in the uterus of wild-type mice, we actually saw that the expression of these proteins was located or enriched in the crypts of the endometrial glands, suggesting that these are, in fact, perhaps sites that are controlling the regenerative pool of the endometrial epithelium. So our conclusions are that inactivation of SMAD2-3 in uterine epithelium result in endometrial cancer with metastases to the lungs. And we also found that the endometrial tumors express ER but lack PR expression. And our studies in endometrial organoids reveal that TGF-beta signaling through the downstream effectors SMAD2 and SMAD3 are controlling BMP and retinoic acid signaling. And so to put this together with our picture of endometrial regeneration in the mouse, we can place then TGF-beta, as well as retinoic acid and the BMPs, as being important growth factors that are controlling the regenerative pool of the endometrium. And placing the enzymes ALDH1A1 and 1A3 as potential markers of the stem cell pool residing in the crypts of the endometrial glands. And so I'd like to acknowledge the co-authors on this project. The project was started by Maya Kreisman when she was a REI fellow, and also when I was a postdoc in the lab of Marty Macek. Also, I'd like to thank the collaborators and advisors who have helped me, as well as the core facility labs at BCM and our funding sources. Thank you. Thank you. This talk is open for questions. Yes, please. Julie Kim, Northwestern. Beautiful talk, Diana. Could you tell me more about the tumors that arose in terms of their subtype and what kind of tumors they were? So because they express the ER, we think that they're more similar to the type one endometrial carcinomas that are seen in women. We're also doing some experiments to actually test that they are estrogen-dependent, so removing ovaries and putting in a pellet, an estrogen pellet, or testing that independently of the lactoferrin CRE. Are you aware of any companies doing trials or developing drugs for the TGF-beta SMAD for endometrial cancer? So I know that they are looking at it for other cancer subtypes where TGF-beta is overexpressed, but then that would be of concern for us because suppressing TGF-beta would then in turn lead to a tumorigenesis phenotype. So were you surprised when you got that degree of tumorigenesis? Right, right. So yeah, I think it's impressive and it just shows that it's really tissue and cell type dependent. Yes. DeMayo from NIH. That's very, very nice. So my question is when you're typing your tumors, do you think it might be better to just do an RNA-seq or get a transcriptome and compare that to the human data to really go with more than one marker? And what do you think about that? Yeah, I mean, I think that would be, we haven't done the actual tumor phenotyping. And then when you look at, so there's been a genomic analysis of endometrial tumors with the different mutations. Have you looked and seen, is any of these SMADs mutated in any of the tumor types that would fit with your model? So some of the, so the SMADs are okay, but we have found mutations in the receptors. So like ACVR1 is mutated. And so some of those mutations are similar to other tumor types. Okay, it's time to move on, but thank you for that great talk. Thank you. Okay, the next speaker is Arunika Goyal from Morehouse School of Medicine. And the title of her talk is Effect of Melatonin Receptor 1 on Glucose and Lipid Metabolism in Melatonin-Proficient Female Mice. Hello, everyone. Today, before you all, I'm going to present my work, which is in regard to the effect of melatonin receptor 1 removal on glucose and lipid metabolism in melatonin-proficient female mice. I have no financial relationships to disclose, and to start with I would like to say that melatonin is known to be an important player in animal physiology. The melatonin is produced during the dark phase of the day when a signal goes to retina, which in turn signals the SCN, and which in turn signals the pineal gland to produce melatonin from where it goes to various organs to carry out its functions. The functions of melatonin are mainly carried out by its two receptors, MT1 and MT2, and in a third circumstance where both these receptors form a heterodimer. The functions of melatonin or these receptors, they carry out by means of GQGI signaling. Usually the general functions of melatonin are associated with circadian rhythm, immune functions, reproduction, anti-inflammation, and anti-tumor activities. However, pertaining to metabolism, it's involved in insulin secretion, energy in glucose homeostasis, and lipid homeostasis. It is to note that inhibition of melatonin is generally carried out during the night phase of the day or under any other circumstances when during night light is exposed, eyes are exposed to light like, for example, jet lag, shift work, modern devices like screens, mobile phones, et cetera. So in our lab we found that in such a circumstance where eyes are exposed to light at night, melatonin signaling is affected and which in turn affects the glucose metabolism because we saw that that mouse which are exposed to light at night, they had a higher blood glucose level in comparison to the mouse which had a normal daylight cycle, and when these mouse which were exposed to light at night were given melatonin, their blood glucose level fall back to normal. Usually both the receptors of melatonin are known to affect the glucose metabolism, however we found that in case of melatonin-proficient male mice, it was the MT1 receptor knockout mouse which had a higher blood glucose level in case of a glucose tolerance test and insulin tolerance test in comparison to the MT2 knockout and the normal wild-type mice. It was also seen that MT1 knockout mouse have a systemic insulin resistance because during a euglycemic pump, they had a higher, a lower glucose infusion rate, a lower glucose turnover and a higher hepatic glucose production. They also had a lower glucose uptake in skeletal and adipose tissue and a lower glycogen synthesis. Further in human patients, it was seen that although there was no difference between the obese and the lean patients in the expression of MT1 receptor, but when the comparison was done in between the patients which had diabetes which was poorly controlled and the diabetes which was well controlled which is shown here in the yellow bar, there was a downregulation of MT1 receptor in the liver of the patients with type 2 diabetes which was poorly controlled. All these results that I have shown you here were normally from the male mice which were melatonin-proficient and there is no data available for the female mice. From the literature, it has been known that usually the females can resist the effects of diet-induced obesity and certain metabolic parameters are well-resisted in case of females. But we wanted to know what is the effect of removal of melatonin-1 receptor on glucose and lipid metabolism if we have a female melatonin-proficient mouse and we keep it under a high-fat diet which can induce diet-induced obesity. So for this purpose, we had both wild-type and MT1 knockout female mice which were kept under a high-fat diet or a low-fat diet regime for 14 weeks. Weekly measurement of weight, food intake and basal glucose levels were done and after the finishing of 14 weeks, glucose tolerance and insulin tolerance tests were done, mice were euthanized and tissue were collected. First of all, I want to show you what happens to the weight and the food intake parameters. So on the right top, you can see that after 14 weeks, both wild-type and knockout mice have nearly the same weight due to high-fat diet in comparison to their low-fat diet-consuming controls. The food intake was also not much difference between the high-fat diet group and the—between the high-fat diet group. The energy efficiency, however, was higher in the melatonin-1-receptor knockout females in comparison to their wild-type counterparts. When we look at the blood glucose levels, we see that due to high-fat diet, there was no significant increase in the blood glucose level of the MT1-receptor knockout females in comparison to their low-fat diet-consuming counterparts. However, there was a difference between the wild-type consuming—wild-type mice which consume high-fat diet and the melatonin-1-receptor knockout mice which consume the high-fat diet. Further, when we performed the glucose tolerance test, we found that both—irrespective of the diet, both wild-type mice on high-fat and the low-fat diet had a normal glucose metabolism after the glucose injection. MT1 knockout mice which were feeding on the low-fat diet were relatively slow in metabolizing the glucose, but the severe effect of disrupted glucose metabolism was seen in case of MT1 knockout mice which were having a high-fat diet where the glucose metabolism was totally disrupted. After the insulin injection, in the insulin tolerance test, we see that both wild-type mice feeding on high-fat and the low-fat diet respond to glucose—insulin injection, and their blood glucose level falls down, and which comes up to the basal normal level after 100 minutes of the test. However, in case of MT1 knockout mice which were feeding either on the high-fat or the low-fat diet, there was no effect of insulin injection, and their blood glucose levels continued to rise. Next, we looked at the lipid levels in the serum, and we found that due to high-fat diet, there was an increase in the triglyceride levels of both wild-type and the knockout mouse, and they were significantly different from their low-fat diet feeding counterparts. The total cholesterol was different only in the case of MT1 knockout mouse. The phospholipids which come from the dietary food was different between the wild-type and the MT1 knockout mouse in both the cases, whether they were feeding low-fat diet or the high-fat diet. The nephor, which is a measure of free fatty acids, was not different between the four participating groups. Therefore, we could infer that lack of MT1 increases the level of lipids in serum of female mice. The increase in the level of lipids can often cause an accumulation of triglycerides in the liver of the mice. When we wanted to test what is the level of triglyceride accumulation, we found that in case of MT1 knockout mouse feeding on high-fat diet, there was a significant increase in comparison to their low-fat diet feeding counterparts. Although low-fat diet feeding MT1 knockout mouse also had a higher triglyceride level, but this effect became more apparent when they were fed the high-fat diet. The accumulation of lipids in case of liver is often known to be the cause of non-alcoholic fatty liver disease, or it is a marker for injury in the liver. So when we looked at one of the markers of liver injury, that is the alanine aminotransferase or the ALT, we found that it was significantly increased in the case of MT1 knockout mouse feeding the high-fat diet only and not in the wild-type mouse. Therefore, we could say wild-type females have healthy livers, MT1 knockout females show hepatic lipid accumulation, and they have an increase in the injury markers for non-alcoholic fatty liver disease. Next we performed RNA-seq on the liver tissue, and we found that under the low-fat diet conditions when comparison was done between the wild-type and MT1 knockout mouse, only 1.8% genes were differentially expressed. Specifically due to our interest, we looked at the genes which participate in the glucose and the lipid signaling, and we found that all of them on this volcano plot were located in the black region, which is the region for the non-differentially expressed genes. Only one of the genes, that is PI3K, was upregulated in the case of the MT1 knockout mouse. Next we looked at the pathways which were enriched in the case of this comparison between the wild-type and MT1 under low-fat diet conditions, and we found that PI3 kinase pathway, fast signaling, which are known by the literature to be enriched in case of MT1 knockout mouse, were present in the list of our pathways. However, when the comparison was done between the wild-type and the MT1 knockout mouse under high-fat diet conditions, we found that the differential expression increased to 5.82%, and most of the genes which participate in the glucose and lipid metabolism, which were located in the black region in the previous graph, were now moved towards the significant significance. Most of them were downregulated, but also some of them were upregulated. While taking these genes when we did the pathway comparison, we found that the pathways relating to the lipid metabolism and the glucose metabolism were enriched. So to summarize, I would like to say that our model is, or we hypothesize that in wild-type mouse, which are melatonin-proficient females, under high-fat diet they are just having the diet-induced obesity. But when a mouse lacks MT1 signaling also, in addition to diet-induced obesity, it can have disrupted glucose metabolism, insulin resistance, elevated lipids, triglyceride accumulation in liver, and an increased liver injury markers. With transcriptomic or RNA-seq analysis, we saw that there was an enrichment of glucose and lipid metabolism-related pathways in the liver of MT1 knockout mouse under high-fat diet, and they can be a cause of NAFLD or non-alcoholic fatty liver disease. I would like to acknowledge the help of our funding partners and also the University of Cincinnati Core Lab for the work they have done to help us. Thank you so much. This talk is open for questions. Perhaps I'll ask the first one. Is the model that the melatonin is acting directly at the hepatocyte, and does the hepatocyte express the melatonin receptor? Yes. Melatonin receptor 1 is known to be expressed. All the melatonin receptors are known to be expressed. Has anybody had a site-specific knockout of the receptors yet? I don't know about other labs. I'm sure they are not, because we are under the procedure of generating one. You guys are making those mice. That's almost exactly the question I have. Is the MT1 receptor expressed in pancreatic islets, or is, again, the model that it's through the liver and lipid metabolism there that the insulin sensitivity is being affected? I'm sorry. I could not hear. Is it direct at the hepatocyte, or is it in the beta cell? Is MT1 expressed? Right. Or in any of the pancreatic islets. Yes. MT1 is expressed in the beta cells. Okay. Yeah. And it's obviously in the hepatocyte. And so it inhibits insulin secretion? Does it inhibit insulin secretion at the beta cell, melatonin? No. In fact, MT2 is known to be involved in that action, because MT2 is found in the islets of, in the pancreas, mostly MT2 is expressed. So there is a debatable evidence that, because one paper says that it is the main receptor which controls the glucose and the glucose metabolism. However, there is a contradictory study which says that an increase in the level of MT2 can also inhibit the glucose metabolism, but that evidence is in humans. So in mice, which are melatonin-proficient, we don't see a relation of MT2 with glucose and lipid metabolism. Thank you. Sure. So, as you mentioned, the FAS and FAS ligands are involved in the inducement process of the beta cells, as you were saying before. So, which, I'm sorry? As you mentioned, the FAS and FAS ligands. The FAS ligands. Yeah, so I guess he's asking is the FAS ligands involved in the effects that the beta cell, with MT2? We didn't say involved with that, but the inducement process of the beta cell. I don't know that from my study. Like, we haven't investigated that in MT1 knockout mouse, but usually literature suggests that what you're saying is correct. So it is known to be involved. Okay. Thank you. Thank you very much. That's an excellent talk. Okay. Okay. The next talk is from Shahwaz Aman, from University of Toledo, correct? And the title? No, second. Go to the second presentation. Okay. So, we'll let that come up. Here we go. There we go. Okay. Good morning, everyone. Thanks for giving me the opportunity to present this work. Make sure you talk into the microphone. Good morning, everyone. Thanks for giving me an opportunity to present my work in the Indocrine Society meeting. This is Shahnawaz Imam from the University of Toledo, Ohio. Today we are presenting the technology involved the CART. So don't confuse CART with the cancer. This is the CART we are presenting related with autoimmune diseases, especially in type 1 diabetes. We have a disclosure, conflict of interest with ADAPT process, patent technology with the University of Toledo. And make sure you speak to the microphone. Now, before going into that, first I bring the background of the immunology of type 1 diabetes. So as we know, pancreatic islet pancreas has two parts, exogenous and exocrine endocrine pancreas. Endocrine pancreas is mainly of the islets. And islets have the beta cells that produces insulin. But in case of type 1 diabetes, this insulin produced in beta cells were destroyed by the T-effector cells. These T-effector cells are especially activated against the anchored GAT65 protein, which is present in most of 90% cases of the human type 1 diabetes. And as I said, these T-effector cells kill the beta cells. Unfortunately, we don't have any cure. Insulin injection several times, as well as pancreatic transplant requires surgery and immunosuppression. Next, the role of T-rex cells in type 1. As we know, the immune system has a hemostasis. So first is the T-effector and T-response. T-effector and T-regulatory cells imbalance is one of the reasons for developing on any autoimmune disease. And the same for human type 1 diabetes. So here, T-effector cells need to be regulated. And that's to abrogate autoimmune against the insulin-producing beta cells. So these two downregulated T-effector cells, we need abundantly present of the antigen-specific T-rex cells in and around the pancreatic islets. But unfortunately, we don't have enough native T-rex cells that have the GAT65 specificity. Now here, we are introducing the engineered car T-rex cells. These engineered car T-rex cells are to downregulate the cytotoxic T-effector cells. And you can see here, once the car T-rex cells navigate themselves to the islet, they recognize the GAT65, get activated, proliferate, and produce anti-inflammatory cytokines to silence T-effector cells. So keeping this view, our objectives are to prove of reversal of type 1 diabetes, human car T-rex cell therapy, inhumanized mass model of type 1 diabetes. And second, we are stepping into the human experimentation with the homing of the autologous engineered car T-rex cells to the human pancreatic islets and validation of hypothesis in human ex vivo. So as we know, GAT65 is a dimer protein. And it have various immune-dominant regions. But we have two GAT65 beta cells paratope known to interact with two immuno-dominant epitopes. We're selected for development of the car T-rex cells. First is the N-terminal. So N-terminal is recognizing the anchored part of the GAT65 present in the beta cells. And middle regions is recognizing the solubles in the blood, as well as from the dead and debris cells. These GAT65 car T-rex cells, we are used to prevent or treat type 1 diabetes in our humanized mass model for type 1 diabetes, as well as we are elucidating the ex vivo homing of the car T-rex cells towards autologous human pancreatic islets. These are our car designs. So this car have T-cell receptors with linker and GAT65 antibodies. We have, as you see in the western blot, the CD3, there are two band of the CD3 zeta. First band is for the TCR and another band from the car. Not only that, we look for the expression of CD25 and FOXP3 in the transduce cells. And you can see here the T-rex expression in the nice T-cells as well as in the transduce car T-rex cells. Now coming into the mice. These transduce cells, we are adaptively transferring to the mice. Before that, before going into the experiment, I need to bring a special emphasis on our humanized type 1 diabetic mice. These are not the classical NOD mice. These mice are expressing human GAT65 in pancreatic islets as well as these mice have null for the MHC, mice MHC, and expressing DQ8 in all antigen-presenting cells. As you know, DQ8 is a predisposition factor for type 1 diabetes in human. And these mice, we breed on the basis of the impaired fasting blood glucose. And by the age of the fourth week, like just after breeding, these mice develop diabetes spontaneously. Not only that, these mice are showing complication of the diabetes, especially retinopathy, nephropathy, neuropathy, and all other complications. Now here is the summary of 30 days triatric clinical trials we do have. So what we did in this type 1, humanized type 1 diabetic mice, we have adaptively transferred normal T-rex cells, irrelevant ipcam-card-controlled T-rex cells, CAR-M, as well as CAR-N T-rex cells. After the 30 days of the adaptively transferred the CAR-T-rex cells in these mice, we recorded the glucose tolerance test. We also scored the islets for the encephalitis. We also profiled the spleen, pancreas, and peripancreatic lymph nodes for the immune cells. So, our top results are from this experiment homing of the homing of the cardiac cells into the islet, then separation of the T effector cells, separation of the T effector cells and reversal of the glucose tolerance. So, you can see in this in this in this presentation in this in this figure you can see here this is the this is the group these mice especially CAR-M treated mice are become tolerant to the glucose challenge and the same in the CAR-M treated mice but whereas in T-reg treated group as well as in the control CART group there wasn't any effect. Not only that we also look for the rescuing of the islet as you can see blue circle is the rescue of the of the of the islet or regeneration of the new islets and some of them have the black circle which have the infiltration but in the untreated as well as in the control mice there hardly you will see any healthy islets and if there are some islets these islets are here they are the infiltrated islets. Now, now most important here we I want to bring the emphasis because this prevent this this the deductively transfer of the CAR-T-reg cells leads to the robust amplification of the CAR-T-reg cells as compared to normal T-reg cells as well as in the APAM CAR-T-reg cells. The amplification is almost the 4 to 5 times as compared to the normal CAR-T-reg cells and that's how they are inducing the they are inducing the separation of the T-effector T-cells. Now, we next coming to the tracking of these CAR-T-reg cells into the pancreatic pancreatic islets. So, GAT-65 CAR-T-reg cells were expressed with the co-transmitted with the inflorescent protein and once it is it will be adaptively transferred to the mice and mice will sacrifice after the 24 hours. We have isolated the peripheral sorry we have harvested peripheral blood, spleen, pancreas and peripancreatic lymph node and results are here you can see in the dot plot you can see the presence of the CAR-T cells in the blood, pancreas and spleen what where it's not present in the peripancreatic lymph node. It's maybe because of the peripancreatic lymph node doesn't have the full proteins always have the peptides for the for the antigen presentation. Now, next is the is the is the cream of the cake. Here we did the we did the homing of the GAT-65 CAR-T-reg cells in autologous human pancreatic islets. So, for so for that so for so for that part what we did we have we have isolated the T-reg cells from the human patients and then we have transduced this this human T-reg cells with the same construct we have used for the mice. As you know mice have the these CAR-T-reg cells are specific to the specific to the human GAT-65 immunodominant regions and so that's how it's recognized and we we we are believing it's recognized the human epitopes of the of the of the pancreatic islet cells and that's how once the transduction of the T-regs will made we develop the CAR-T-reg cells. Now, we culture the CAR-T-reg cells for the amplification and we co-culture the this CAR-T-reg cells with the with the pancreatic islet isolated from the same patients. And that we can say autologous co-culture of the pancreatic islet and CAR-T-reg cells. Now, here is the zero-hour co-culture of the GAT-65 CAR-T-reg cells in an autologous conditions and you can see green fluorescent protein here these are the CAR-T and these are the red strain insulin, the islets. Now here is the one of the best part of the picture you can see the navigation or GPS tracking of the CAR-T-reg cells they are reaching to the islets and and you can see this this this CAR-T-reg cells are finding their path they are they are recognizing the GAT-65 antigen activating themselves and started proliferations and this is the 24-hour post post co-culture. Now in the in the first patient then in the another patient we also did the same this first one is 24 hour another one is the 40 like 46 hour and you can see see the CAR-T-reg cells they get themselves reached to the islets recognize proliferate and that's how in the and in 48 46 to 48 hours they started the the amplifying the signals and you can see their cells are in the in the in the center of the or in the in the in the crown of the islets. Now in the other third patient we have also did the same and you can see this car the same the same car the the autologous CAR-T-reg cells navigate themselves reach to the reach to the islets and that's how amplification of the of the CAR-T-reg cells in an in in the in the islets and this is after the 48 hours the signals been increased and improved. So here in the conclusion this is the first time a therapeutic approach is successfully navigating that is the diabetes phenotype in our spontaneous humanized type 1 diabetic model. The antigen specific proliferative capacity of the GAT-65 CAR-NNM-T-regs was four to five times higher than the antigen specific NAIF-T-reg cells and suppression capacity suppressive capacity of the CAR-T-reg cells was observed at one is weight ratio of the T-responsive cell. This provide evidence for the superiority of the GAT-65 CAR-T-reg cells over normal T-reg cells on their diabetic reversal capacity and that's what we have seen in the glucose tolerance test. Now next we are stepping into the into the human experimentation so homing of the engineered CAR-T-reg cells to the human autologous pancreatic islet and robust proliferation proves that same CAR-NNM-T-reg can be used for a human phase one or phase two clinical trial. Thank you and we are thankful to the Hammersley Foundation for the funding and thank you and open for the questions. Thank you. This is open for questions. That's very exciting. What's your strategy? Are you going to try to partner with pharma to to ramp this up or what what's what's the next step? Our main goal is to bring this technology for the benefits of humanity benefits of the type 1 diabetes whether we are partnering with the government institution or with the with the pharmaceutical institution but this this technology need to come to the market need to come to the public accessible to the public. Yes any any other questions? Any other questions? Okay if not thank you very much that was quite exciting. Thank you. All right so the last talk is from Manisha Taya from UT Southwestern Medical Center works in the cancer center there and her title is ovarian cancer cell glucocorticoid receptor activity is associated with cytokine secretion promoting the infiltration of immunosuppressive cells into the tumor microenvironment. Thank you and good morning everyone and today I'll be talking about how tumor intrinsic glucocorticoid receptor affects the immunosuppressive phenotype observed in ovarian cancer and the potential mechanisms that it may employ. I have nothing to disclose. So in our lab the interest is in studying the glucocorticoid receptor it is a transcription factor however its role is very cell type and cancer type specific. So back in the day before my time in the lab through an unbiased screen the lab unexpectedly found that glucocorticoids can promote cell survival in epithelial tumor cells unlike its well described pro-apoptotic function in other cell types like lymphocytes. Subsequently two genes sgk1 and mkp1 were found to be required for the anti-apoptotic signaling mediated by gr and this is what initiated the lab's interest in studying the role of gr in cancer but it is not till recently that we started studying its impact or its potential impact on tumor immunosuppression. So it is well established that glucocorticoids such as dexamethasone are widely used in conjunction with chemotherapy to prevent chemo associated side effects. However a proof of principle trial in ovarian cancer patients showed that there we go showed that glucocorticoid administration via dexamethasone in patients compared to normal saline treatment is associated with rapid upregulation of these pro-survival genes sgk1 and mkp1 in the ovarian cell tumors. So this really supports that gr activation has this interesting canonical effect of resulting in anti-apoptotic gene expression in patient tumors and may therefore be detrimental to chemo induced apoptosis and reduce chemo effectiveness. Accordingly examining the relationship between gr expression and clinical outcome in ovarian cancer patients an immunohistochemistry study demonstrated high gr expression in primary tumors is significantly associated with decreased progression free survival. Additionally using primary tumor data retrieved from the ovarian cancer TCGA data set transcript counts of FOXP3 which is a marker of immunosuppressive T regulatory cells as well as CD33 a marker of immunosuppressive myeloid cells were compared in gr low and gr high expressing ovarian cancer specimens showing significantly higher expression of these markers is associated with high gr expression suggesting an immunosuppressive phenotype. An important side note however is since gr activity is modulated by the presence of ligand or phosphorylation or ubiquitylation etc. It is in fact more important to understand gr activity in ovarian cancer tumor cells rather than just looking at the gr expression alone. So I mentioned that the anti-apoptotic nature of gr activity in epithelial tumor cells but another area of interest or rather a question is does tumor cell intrinsic gr activity also suppress anti-tumor immunity. And this led us to the hypothesis that tumor intrinsic gr activation in ovarian cancer is associated with increased immune suppressive cell infiltration into the tumor marker environment possibly via modulating tumor cell cytokine secretion. So cytokines are important regulators of all immune responses. What happens is initially during tumor growth there's acute inflammation signifying a tumor killing immune microenvironment with active CD8 T cells and K cells etc. However over time development of chronic inflammation signifies this widespread accumulation of cytokines leading to an immune suppressive microenvironment. So the first question we focus on is whether tumor cell gr activation modulates the tumor cytokine secretome. And to answer this question I performed an in vitro cytokine profiling assay after activating gr in ovarian cancer cell lines via dex treatment. We also use a novel selective gr antagonist, a compound called CORT134 or Relaquarlant. So after activating gr in tumor cells we performed a multiplex bead based immunoassay on tumor condition media allowing for simultaneous measurement of multiple cytokines. And before I go into the data I'd like to mention that our results demonstrate an immunosuppressive trend in the data as tumor intrinsic gr activation upregulated the secretion of specific cytokines that are necessary for the infiltration of a particular immune suppressive cell population called myeloid derived suppressor cells or MDSCs. For instance tumor secreted cytokines such as GCSF which is granulocyte colony stimulating factor was induced by gr activation and this effect was reversed in the presence of the gr antagonist CORT134. Now GCSF along with other cytokines are responsible for the dysregulation of normal myelopoiesis leading to the production or the expansion of MDSCs in the bone marrow. Furthermore CXCL5 is a prominent cytokine or chemokine involved in the migration of MDSCs from the bone marrow through the circulation into the primary tumor site and its protein secretion is significantly upregulated upon tumor cell gr activation. Finally once MDSCs are in the tumor microenvironment they undergo an activation process via key cytokines and growth factors such as TGF-beta and its secretion is also upregulated after gr is activated in ovarian cancer cells. Once activated MDSCs of course strongly inhibit T cells particularly CD8 positive cytotoxic T cells via various mechanisms. So it seems from our preliminary data that there is definitely an evident glucocorticoid receptor and cytokine circuit in tumor tissues. So what are MDSCs? They're a heterogeneous population of immune suppressive cells. Normally bone marrow derived myeloid cells develop into cells of the innate immune system such as dendritic cells, macrophages, neutrophils, etc. But during chronic inflammation such as in cancer there are signals particularly from the tumors in the form of cytokines that cause the differentiation of myeloid cell precursors into MDSCs. There are two subsets of MDSCs that are both functionally and phenotypically different from each other. Importantly MDSCs through recent studies there's some evidence showing that MDSCs are significantly increased both systemically in the peripheral blood as well as locally in the tumor or ascites in ovarian cancer patients. However biomarkers used to identify patients that might benefit from MDSC targeting therapy have not been developed and the cause or the mechanism for this observed MDSC infiltration in patients is unknown and that's sort of the idea behind our project is to understand whether tumor cell glucocorticoid receptor activity might play a role in the infiltration of these immune suppressive cells. So to investigate whether ovarian tumor cell GR activity regulates the production of MDSCs we perform a differentiation assay culturing healthy PBMCs in tumor conditioned media that consists the cytokine secretome released from tumor cells after GR was activated using dexamethasone. We identify the differentiated MDSCs via flow cytometry and further characterize both their activation state as well as functional state through T cell suppression assays. So our results demonstrate that the tumor conditioned media collected after GR activation which is in the middle panel has a larger potential to differentiate MDSCs from precursor myeloid cells compared to vehicle treated tumor conditioned media as identified by the CD11B, CD33 double positive population and this data is quantified on the right. So then we sort for these MDSCs and upon sorting we characterize them by measuring their activation markers like arginase 1, nitric oxide synthase 2 that are required for their suppressive function on T cells and our data illustrates that these MDSCs are indeed in their activated state and most importantly we have ongoing experiments to determine whether these MDSCs are functional and have the potential to suppress T cell proliferation or activation. Finally to further explore the effect of tumor GR activity on MDSC recruitment or activation in vivo I'm complementing the in vitro studies with a mouse IDH tumor model. These cells grow into multiple tumors throughout the abdominal cavity much like those observed in women with stage 3 and stage 4 ovarian cancer and using this model we can study key mechanisms detailed here like whether tumor cell GR activity is involved in cytokine release, whether it's involved in MDSC infiltration and production and finally whether these MDSCs as a result can suppress CD8 positive T cell responses. So really the idea is to understand whether GR antagonism can improve anti-tumor immune responses in patients. Lastly I'd like to share that a recent phase 2 clinical trial presented at ESMO in September and at ASCO last week demonstrated that intermittent GR antagonism by using CORT134 arelopurulent with chemo significantly improved progression-free survival. And with that I'd like to conclude that ovarian tumor intrinsic GR activation regulates the cytokine secretome. This tumor's cytokine secretome allows for MDSC generation and ovarian tumor cell GR activation and the resulting MDSC generation or the infiltration could possibly contribute to the tumor regrowth we see in ovarian cancer patients and the recurrence. And hopefully the effects of tumor cell GR activation on increased MDSCs may be reversed by selective GR modulation. And I'd like to give my acknowledgments and happy to take questions. Thank you. Thank you for that excellent talk. It's open for questions. Any questions? So how widely applicable is this across cancers? This surprises me. It's kind of counterintuitive but do you think it's mainly in the ovarian cancer realm or breast cancers and other cancers as well? That's a great and complicated question. So as I mentioned GR is a very cell type and cancer type specific. So in ovarian cancer we do see that it could have immunosuppressive effects as well as in prostate cancer. But when it comes to breast cancers that's very complicated because for instance ER positive breast cancer GR is actually good. It has better prognosis but for ER negative like triple negative breast cancer it is it's also immunosuppressive. So it's very cancerous. And I guess the reason behind that would be GR is a transcription factor and it all depends on the upstream co-regulators and the downstream epigenetics in the tissue. Any other questions? So one more I have. Obviously this is pathophysiology. What about the physiologic role in normal tissue for this suppression? So suppressing glucocorticoid receptor in regular physiology. I don't see how that would be beneficial in any way. I guess it's mostly GR modulation would be most beneficial pathophysiologically and physiologically in cancers or in like Cushing syndrome or hyper cortisolism. Any other questions? If not let's thank the speakers for a great session and everyone staying on time. Thank you.

TGF-beta signaling

endometrial cell homeostasis

endometrial regeneration

organoid morphology

gene expression

melatonin receptor 1

glucose metabolism

lipid levels

engineered CAR-T-reg cells

type 1 diabetes

pancreatic islets

tumor intrinsic glucocorticoid receptor activity