Good morning, and thank you for joining us for the session this morning on Spatial Cellular Interactions and Functions in Liver Health and Disease. Today we're going to have five speakers, followed by a panel discussion at the end. And I wanted to remind the group that immediately following this session we'll be having the business meeting for the Liver Cell Biology Special Interest Group in this room. So if you're interested in participating, please stay after the session has ended. Our first speaker today is Eni Castellari from Mayo Clinic, and she'll be presenting on Spatial Cell-to-Cell Interactions During Liver Fibrosis. Thank you to the organizers, to the moderators, and to all of you for being here. So today I'm going to talk about... Can you hear me now? Yeah? Good. All right. So thanks to everyone for being here, to the organizers and the moderators. I have nothing to disclose. So I will introduce liver fibrosis very briefly, because I'm sure that all of you know what it is. Fibrosis and cirrhosis are the 11th most common cause of death worldwide. It is a scarring process that starts with a good reason, but if the injury persists, this progresses towards cirrhosis. So the main players in liver fibrosis are called hepatic stellate cells. And they get activated by PDGF, where they proliferate and migrate, and by TGF, which induces matrix deposition and the accumulation of collagen and extracellular matrix. In the recent years, we have also shown that activated hepatic stellate cells release fibrogenic extracellular vesicles, which amplify liver fibrosis. Now what are extracellular vesicles? They are nano-sized entities released by any type of cell in the body, presumably by any type of cell. And during liver diseases, many groups, including ours, have shown that the EVs that are released can be pathogenic. So some of the functions or what this pathogenic EVs induces are HSC migration, endothelial cell migration as well, immune cell infiltration, and all of these participate to the progression of liver disease. However, we still do not understand how this EV release is regulated. And most importantly, where these pathogenic EVs are being released in the fibrotic liver. Also, what we don't understand is what is the correlation with the specific cell-to-cell interaction, and what is the role of these EVs in these interactions. So the aim of the study is to understand the cell-to-cell interactions in the fibrogenic niche during liver fibrosis, including the EVs, or extracellular vesicles. So our approach was to utilize the CCL4-mediated model, mouse model, of liver fibrosis. So we injected our mice with olive oil and CCL4. We made sure that we had fibrosis by serious red. And then we utilized the spatial transcriptomics from the 10X genomics technology. And after that, we performed bioinformatics analysis. So the steps that I will be talking about are tissue zonation, spot annotations, differential gene expression. We integrated the spatial with the single-cell RNA sequencing data to perform the neighborhood analysis. So as I mentioned, first of all, we annotated the zones. With the help of the pathologist, we utilized machine learning to annotate the zones in the liver, pericentral versus periportal. And then we annotated each spot. For those who are familiar with the 10X genomics technology, we annotated each spot to the pericentral or periportal zones. And to those who are not familiar with this technology, these spots are about 50 microns and they include several cells in one spot. But it's important to annotate where they are situated. Why? Because the gene expression is different in each of these zones. And here you can see that by the UMAP, here the UMAP graph, you can see that the pericentral is quite distinct from the periportal zone. They have a distinct gene profile. We also saw that the pericentral has increased expression of these hepatic stellate cell markers such as collagen 1 alpha 1 and collagen 3 alpha 1. So hepatic stellate cells activation markers and fibrotic markers. So what we understood from this is that the pericentral zone in this CCL4-mediated liver fibrosis model, the pericentral zone is a fibrotic one where we have HSC activation and extracellular matrix deposition. Next, we wanted to know what pathways are upregulated in this pericentral zone. So we took all of our detected transcripts, around 14,000, and we used a few filters. So we took the pericentral spots with adjusted p-values, p-value less than 10 to the power of minus 6, and a full change more than 2. And then we performed gene ontology analysis of the cellular components. And what we saw is that several of these genes that were upregulated were part of the vesicle trafficking pathways, and many of them were part of the extracellular vesicle pathways, or exosomes as well. And these are a few genes that were, that were, that are part of these pathways where we can see that with CCL4, so with fibrosis, they are upregulated in the pericentral zone. So what we can say here is that on top of HSC activation and matrix deposition, we also have an increase of the EV-related pathways. In addition, we also saw immune response-related pathways that were upregulated in the pericentral fibrotic zone. And the one that caught our attention was the B-cell receptor complex, so B-cell-related pathway. So we were thinking, is there any relationship between hepatic stellate cells and B-cells that we could see in the spatial transcriptomics data? So what we did here, we took our transcripts, we integrated single-cell RNA-seq datasets, so ours and others, with the spatial transcriptomic dataset, and we performed neighborhood analysis. And this is what we got. Here in the olive oil, so in the normal healthy liver, you can see that the HSCs are neighbors with endothelial cells, epithelial cells, stromal cells, some dendritic cells, which seems normal. But when we have fibrotic livers, these HSCs become neighbors with B-cells, T-cells, macrophages. So right now, we are happy to collaborate with Dr. Avello from the University of Minnesota to understand this correlation, this proximity between B-cells and hepatic stellate cells in the context of NASH, MASH, sorry. So as a summary and key takeaways, we can say that the pericentral zone is a fibrogenic zone in the CCL4-mediated liver fibrosis, that there is an increase, an upregulation of the EV-related pathways and immune response pathways in this fibrotic zone, that the HSCs are close neighbors with B-cells during liver fibrosis. And the next question that we can ask is, do EVs mediate the HSC-B-cell communication? And we hope to discover novel cell-to-cell interactions during liver fibrosis, which can be in the long term of therapeutic interest. So with this, I would like to thank all people that work with me, especially Dr. Shalil Kanal and Dr. Leo Liu, who helped with this project, mentors and collaborators, as well as the funding sources. And thank you for your attention. Thank you. Our next speaker today is Dr. Paul Monga from the University of Pittsburgh, and he'll be speaking on endothelial cell hepatocyte interactions in zonation and regeneration. Just give me a few seconds. All right. Is that okay? Oh, great. Absolutely. Well, I'd like to thank the organizers for this kind invitation. And I'm going to change the title a little bit to Metabolic Zonation, Regulation and Implications in Liver Pathophysiology. So these are some of my disclosures. So liver performs more than 500 functions, and I guess the key cell type that performs that function is the hepatocyte within a liver lobule. And in order to sort of perform these very diverse and vital functions, liver is metabolically and functionally zonated. Based on where hepatocytes are located in a liver lobule, if they are closer to portal triad or closer to central vein, they have distinct gene expression, which then results in different function that makes the liver an efficient metabolic synthetic detoxifying machinery. Some of the examples of that zonation include processes like beta-oxidation and lipogenesis, which are compartmentalized, with beta-oxidation-related genes being more periportal versus lipogenesis genes being more pericentral. The same thing is true for gluconeogenesis, glycolysis, and even in ammonia detoxification we have urea production enzymes that are expressed in hepatocytes around the portal triad, and amine production, which is more in the pericentral hepatocytes. This can be appreciated by many different types of techniques you heard from Menis, the Visium platform. We use molecular cartography or 100-gene single-cell spatial transcriptomics by Resolve Biosciences, and you can see here each dot is one RNA molecule, and each color is a different gene. If you look at 100 genes, they are heavily compartmentalized, some being more expressed around the central vein, others being more expressed in the portal area. We have shown over the last almost 20 years the role of Wnt signaling pathway in the regulation of this process of metabolic zonation, and this can be really attributed to two major Wnts, Wnt2 and Wnt9b, that can be seen to be localized in the endothelial cells of the liver. Wnt9b, which is in green color, can be seen in endothelial cells that are lining the central vein. This is, again, single-molecule fish, which is part of that molecular cartography, and the red is the Wnt2, which is mostly present in the sinusoidal endothelial cells that are located in zone 3 and zone 2. And what we have shown is that the Wnts, Wnt2 and Wnt9b, which need a protein called Wntless to be secreted from a cell, once this Wnt is secreted, it goes and binds to the receptor Frizzled and core receptors LRP5 and 6 on hepatocyte, which then results in activation of beta-catenin. Beta-catenin translocates into the nucleus to turn on target genes. And this particular axis is active in zone 3 of the liver. And how have we shown that? We have generated conditional knockouts of beta-catenin in hepatocytes, conditional knockouts of LRP5, 6 in hepatocytes, or we have knocked out Wntless protein from endothelial cells of the liver, or we have specifically knocked out Wnt2 and Wnt9b from the endothelial cells of the liver, and that has resulted in a significant loss of expression of genes that are normally expressed in zone 3. The top panel, as I said, is the normal liver with 100-gene molecular cartography, and the bottom panel here is endothelial cell Wnt2, Wnt9b double knockout. We can use the data that we have collected from molecular cartography in two ways. We have gene expression of the 100 genes, so we can make U-maps and we can do clustering. But we also have the spatial information of each of the dot, which can be ascribed to a particular cell, so we can do something called reverse tissue mapping of each of the cell in a cluster. And what I mean by that is the following. If you look at 16 genes that are heavily zonated and look at their expression and then try to generate clusters through the single-cell spatial transcriptomic platform, we are able to identify very distinct six clusters in hepatocyte in the liver lobule of a control animal. This is C1, C2, C3, C4, and C5, and C6 clusters. But since we know the spatial localization of each of this dot, which is a cell, back on the tissue, we can reverse map them on a tissue section, and that's what I'm going to show here. This is a tissue section. These are some of the landmark genes, glutamate synthetase in yellow, identifying where a central vein is, and green is SOX9. It can identify where a portal triad is. But we can now reverse map these clusters back on the tissue section. And you can see here the C1 cluster is the one which is pericentral hepatocytes that are right adjacent to your central vein. C2 cluster is more pericentral. And then as you move from C3 to C6, it's more towards the periportal zone. So what happens when Wnt2, Wnt9b are deleted from endothelial cells? You find that these clusters have significantly altered. There's almost absence of C1 and C2 cluster, and there's enrichment of C4, C5, and C6 clusters. And when you map these back onto the liver slide, you can find that although we have identified central vein here, there aren't any typical hepatocytes that were expressing pericentral or zone 3 markers, but instead now the hepatocytes that are adjacent to the central vein are expressing markers of midzone and of periportal hepatocytes, meaning there is pericentralization of the liver lobule. So the whole liver lobule now is expressing periportal and midzonal genes. And this is not just true for Wnt2, Wnt9b endothelial cell knockouts. Any times you perturb Wnt beta-catenin signaling in the liver, you find the same exact pattern. So this is, again, molecular cartography, and I'm just picking up some selective genes to kind of show, to emphasize our observation. CYP2E1 is in green, CYP2F2 is in red, and then CYP2F2 is a periportal hepatocyte marker. CYP2E1 is pericentral. Glutamine, synthetase, Lacto-Regucalcin are, again, decorating your pericentral or zone 3 hepatocytes, and Arginase and ASS1, your periportal hepatocytes. And what happens when beta-catenin is knocked out? You see that there is loss of pericentral marker, but then you have gain of periportal marker in this zone. And the same thing is true on these low-abundance genes. And this is happening all across. This is also happening in LRP56 knockouts in hepatocytes and also in endothelial cell Wnt-less knockout. So we believe that zonation is really dynamic and is a function of both transcriptional activation as well as proactive transcriptional repression of genes that makes this process of metabolic zonation really functional. Very recently, that work was published last year, and I just wanted to share some unpublished work that we are still trying to struggle to make sense of, but I'll be very happy to discuss this. We know that endothelial cells are zonated. Great work from Robert Schwartz and Shaheen Rafi in cell stem cell. They showed that even endothelial cells across the liver lobule are zonated. The endothelial cells that are present in Zone 1 have very different expression profile than Zone 3 endothelial cells. So the question is, endothelial cell Wnts are zonated and control hepatocyte beta-catenin and overall zonation. What controls endothelial cell Wnt zonation? And because we know that there is a very strong and a physical almost relationship between a stellate cell and an endothelial cell, we simply ask the question, if stellate cells are playing any role in regulating the function of endothelial cells? And for that, we used Robert Schwab's L. radcree animals and bred them to our hepatic stellate cell to, sorry, Wnt-less floxed mice to make stellate cell-specific Wnt-less knockout, which means stellate cells are unable to secrete any Wnts. When we did that, what we observed was that in both in males and females, they have lower body weights. And when we did echo MRI, we found there was more lean mass in our knockouts as compared to fat mass. So these animals seem to be more metabolically fit. And then we did single nuc-seq on these Wnt-less stellate cell knockouts and their genotype controls. And what we found was when we did feature plots, we identified different cell populations. And within the hepatocyte fraction, what we find is a unique enrichment of these hepatocytes in the knockout. And when we look specifically at pathway analysis of what is happening in these specific heparocytes, we find they up-regulate. These are heparocytes with up-regulated expression of, or showing FXR-RXR activation, they show fatty acid oxidation, they are also synthesizing more cholesterol, but they're also synthesizing more bile acids. So it feels like when you're knocking out Wnts from the stellate cells and they're not able to secrete any Wnts, the heparocytes are changing in character and now expressing a more metabolically active gene expression. So what is the mechanism behind that? And this is where we are beginning to sort of start to really go deeper into the data. So the first thing was we wanted to confirm some of those observations in gene expression to see if that is indeed happening at a biological and a functional level. So indeed we find increase in bile acids in our stellate cell Wnt-less knockouts and also genes like CYP7A1, which are normally expressed only in the pericentral region because we know bile acid synthesis genes are zonated, the highest expression, which is seen in blue in a control animal is in zone three and then it regresses as you go towards zone one. But what you see is that CYP7A1 is now upregulated almost pan-zonally. It's higher and that is most likely the reason why there are more bile acids in this. The other thing was that we found when we did ORUS-BORUS to look at mitochondrial function that when you are using pyruvate as a substrate versus palmitic acid as a substrate, we found that the L-Ret Wnt-less knockouts are using fatty acids for energy production much more efficiently than glucose. So again, they are using fat as a substrate much better when you're knocking out Wnts from stellate cells. So again, trying to delve deeper into what's going on and knowing specifically that CYP7A1 is very well known to be regulated by Wnt signaling in the liver, we wanted to see if there is any expansion of Wnt activation in endothelial cells, which could be driving beta-catenin-driven gene expression in hepatocytes. And what we did when we looked at molecular cartography of those 100 genes that are known to be zonated and look at the clusters, we find the same six clusters in our knockouts, but what we do find is a gain of number of cells in C2 and C3 cluster, which are more zone 3 hepatocytes. So these are cells that are still under the control of Wnt signaling pathway and you see more expression of Wnt target in these hepatocytes. And you can see here that an example is CYP2E1, which has traditionally been regulated by Wnt signaling, and you can see that in LREP Wntless knockouts, there is expansion of CYP2E1 in liver lobule as compared to CYP2F2, which is a periportal marker. Mid-zonal and periportal expression is not altered much. It's just activation of the Wnt signaling pathway. And obviously the question was, how is it happening? What is the level of Wnt2, which is known to be present in sinusoidal endothelial cells, but only in zone 3? What is happening to it? And what we find is that Wnt2, again, this is using single cell spatial transcriptomics, Wnt2, which is normally only expressed in sinusoidal endothelial cells in zone 3 and then decreases all the way to zone 1, we find almost a pan-zonal increase in Wnt2, meaning that even the periportal sinusoidal endothelial cells are now expressing higher levels of Wnt2. And indeed, when you look at bulk RNA-seq from these animals, we do find several zone 3 genes, which are known targets of beta-catenin to be upregulated in the liver. And then the last question we asked, is it specific only to Wnt2, which is being upregulated, or is the phenotype of sinusoidal endothelial cell changing entirely? So we started looking again at some of the markers that are known to be only expressed in pericentral sinusoidal endothelial cells. And one of those markers is FABP4, fatty acid binding protein 4, and you can see in a normal animal, you see highest expression in zone 3, and it goes down. And then this is an artificial inflation here, because FABP4 is also expressed in portal vasculature. So the portal vein in that area is also high. But you can see FABP4 is almost pan-zonally upregulated. And then when we look at immunohistochemistry, you can see FABP4 in control is only decorating your endothelial cells, sinusoidal endothelial cells, around zone 3. But in a knockout, you see it's pan-zonal. Every endothelial cell in the liver is now expressing FABP4, meaning there is pericentralization of L-seqs throughout the liver lobule, when Wnts cannot be secreted from stellate cells. Who knew? And then the second marker that we used was C-kit. C-kit is, again, another sinusoidal endothelial marker, only in the pericentral zone. But in L-rat Wnt-less Cree, you see pan-zonal expression of C-kit throughout the lobule. So this is where we are. So to summarize, we have endothelial cell Wnts activate beta-catenin, and this axis plays key role in division of labor or metabolic zonation. And it does so by activating pericentral and repressing periportal genes. And Wnts from stellate cells play a key role in patterning of liver sinusoidal endothelial cells and control their zonation and, in turn, metabolic zonation of hepatocytes. And we are still trying to figure out the exact mechanism. So I'll stop here and thank a lot of the members, specifically individuals that are involved in this project were Anya Verma and Leon and Shikai Hu. And this is the individual, so Anya is here, Leon is here, and then I have a couple of other individuals who may be in the audience and are presenting posters tomorrow and day after. So I hope to see you there. Thank you again. Our third speaker for this session is Dr. Xavier Revelo from the University of Minnesota, and he'll be speaking on B-cell interactions with the liver microenvironment and liver disease. All right, thank you. We are ready to start again with Dr. Xavier Revelo. All right, thank you for the invitation, and sorry about that interruption. I'm going to tell you about how B-cells promote that inflammatory process in the liver, specifically in the transition between healthy liver to the more severe MASH condition where you have a very strong inflammatory process. In the healthy liver, you have some immune cells. These are resident immune cells, but notably work from the last 20 years has shown that with disease progression, you have the infiltration of monocytes that come from the bone marrow. These monocytes enter the inflamed tissue where they become macrophages. Some of them become lipid-associated macrophages that recently have been shown to protect the tissue against the excess accumulation of lipids. We also have Kufr cells. These cells, early on, they initiate inflammatory response through the secretion of pro-inflammatory cytokines including CCL2, TNF, IL-1. Some of these cytokines can recruit some of these monocyte-derived cells, and things go wrong. So you have this strong inflammatory response. Other cell types involved include the neutrophils, T-helper cells, cytotoxic T-cells, dendritic cells. Now, notice how there's a question mark next to these B-cells, and that's because the role of these cells has been underappreciated, in my opinion. So a few years ago, Fan Tavaro, who's a PhD student in the lab and is in the job market right now, hypothesized that these B-cells promote a transition from the lipid accumulation to the more severe steatohepatitis, or MASH, and that this process of activation is triggered by the leaky gut and the entry of bacterial products from the intestine. And so the first observation that led to this work was that people, patients with MASH, had an increased accumulation of B-cells in the liver. So these are histology sections of patients with a low NASCOR. And in this representative, you see a patient with a NASCOR of six. And notice the presence of CD19 staining, which is specific for B-cells. You see a cluster of these cells with disease severity. And when we quantify the number of B-cells in the hepatic tissue, we notice a very strong correlation, positive correlation, between disease severity and the presence of B-cells in the portal areas. So we went back to the lab, and we wanted to investigate the mechanisms of B-cell activation as well as the consequences of that infiltration. So in the lab, we use a mouse model of dietary-induced MASH. We feed mice either a normal child diet or a Western diet, 40% fats, 40% carbohydrates, some fructose, some sucrose in the drinking water. And so we've seen that after 15 weeks, these mice present the entire spectrum of disease. You have lipid accumulation. You have fibrosis. You have sustained, strong inflammation. And so we thought it was a good model to study the role of B-cells. And one of the first experiments we did was to harvest the livers of these mice. We isolated the B-cells using immunomagnetic selection. And then we tested the cytokine production by these cells, specifically ex vivo. And as you can see in this mass cytometry plot, this is the frequency of B-cells expressing the cytokine TNF-alpha in a healthy mouse. And then in the Western diet MASH mouse, that frequency of TNF-alpha producing B-cells increases. And we see that in several conditions with different stimulants. We also did a bulk RNA-seq experiment using sorted B-cells from the animals fed the Western diet. And in agreement with the increased cytokine release, we saw a lot of genes associated with an inflammatory response. The top gene was MyD88. So MyD88 is an adapter protein that transduces the signals from the toll-like receptor. So it senses all the inflammation in the environment and induces that signal to NF-kappa-B. And then activation of NF-kappa-B leads to more cytokine production. So at this point, we hypothesized that MyD88 was a key molecule that these B-cells use to become activated. And to formally test that hypothesis, we created a mouse. We created a cross-mouse to delete MyD88 specifically in the B-cells. That was a CD19cree with a MyD88 flux mouse. We put these mice on the Western diet for 15 weeks, and then we assessed disease progression. We didn't see a lot of effects of MyD88 depletion on lipid accumulation. The most striking phenotype was fibrosis. So these mice lacking MyD88 specifically on the B-cells had about 50% decrease in fibrosis. So that was very interesting to us. We wanted to understand some of the potential mechanisms. First, we analyzed the B-cells in these mice. And indeed, we noticed that these cells have a decreased level of activation as evidenced by lower expression of class 1 and class 2 antigen presentation molecules and CD86. And then, of course, we are wondering if the B-cells can then crosstalk with other immune cells that are well known to promote inflammation and fibrosis, such as the T-cells. And what we found is that the CD8 and CD4 T-cells, they have a phenotype suggesting less effector memory responses. So the B-cell become activated through MyD88 and then crosstalks to the T-cell. The T-cell gets a while and makes more inflammation, more cytokines. We also did some single-cell RNA-seq experiments, because that's what everybody's doing now. In this case, we sorted all immune cells, and we focused on the B-cells, but also noticed that we have enough resolution to see some of the changes in monocyte macrophages, T-cells, NKT cells, and neutrophils. But we focused on the B-cells. We were able to identify four different clusters with a unique gene expression pattern. Looking further into these genes, we concluded that the major cluster, cluster number one, was a mature B-cell cluster. These are the classical B-cells that make lots of antibodies that are making the pro-inflammatory cytokines. One interesting finding here was that there was a second cluster, cluster 11, that expressed a gene expression pattern suggesting that these cells are immature. They're not fully mature, not yet antibody-producing cells, but in the process of differentiation into a full, mature B-cell. Also, we detected a small cluster of B-cells with a high mitochondrial gene content, suggesting that they're metabolically active. We're working on this right now. We're trying to investigate what is the significance of that population of immature B-cells in the MASH liver. We confirmed some of these findings by flow cytometry. We see that there's a decrease in immature, increase in mature B-cells in the MASH liver. And also, we're studying the metabolism of these B-cells in healthy and MASH conditions using seahorse assays. One of the main findings we have is that with the inflammatory MASH condition, these B-cells need more ATP to be pro-inflammatory, to make more cytokines, to produce more antibodies. What is the source? What are the mechanisms underlying that increased ATP production? We see that they abrogate Oxford's capacity to be able to come up with that ATP that they need, not so much glycolysis. If you want to know more about the metabolic flexibility of these cells during MASH, Fanta is presenting some of his work on Monday afternoon. To quickly summarize these findings that we published a couple years ago, what we've seen is that during the transition from simple fatty liver to the more severe MASH, these B-cells become activated through toll-like receptor signaling. They need MyD88. Also, we have some evidence that the B-cell receptor engagement is activated. Once the cell becomes activated, they can crosstalk with CD4 T-cells, CD8 T-cells, and these cells make interferon gamma. Along with the other B-cell cytokines, the most striking phenotype we see when we deplete these cells and MyD88 in these cells is the lack of fibrosis. Now, I didn't have time to show you today, but we've also done fecal microbiota transplants using human material into antibiotic-treated mice to show that the leaky gut and the entry of endotoxin and antigen through the endothelial barrier is one of the stimulants leading to B-cell activation in the MASH liver. So in the next couple minutes, I just want to show you some data, the recent work that we're working in the lab right now, trying to understand what are the precise organs where B-cells become activated during disease. We see local activation in the liver tissue, but also we're interested in the activation of B-cells in the small intestine. And this work is led by a postdoc in the lab, Dr. Wan, who hypothesized that Peyer's patches in the gut are probably involved in the activation of B-cells. And that's because here in these tertiary lymphoid organs, there's a strong communication between the B-cells and the T-cells. Mainly in this germinal center response, the T-cells, specifically TFH cells, follicular helper cells, talk to the B-cells and send signals in the form of cytokines or direct contact to tell the B-cell, hey, you can go ahead and class switch to a specific immunoglobulin. Now in the Peyer's patches, there's an important immunoglobulin being made by the B-cells. It's IgA. And IgA is very important because once secreted into the lumen, IgA is very good at binding pathogenic bacteria, preventing gut damage, preventing those endotoxins from entering the system. So we wanted to study those Peyer's patches in more depth, and we did simple experiments like histology. And the first observation we had was the size of those Peyer's patches. So in a lean, healthy mouse, this is a Peyer's patch with a reasonable germinal center. However, in a mashed mouse, that Peyer's patch becomes enlarged, but the germinal center is much smaller. We've done a lot of characterization looking at the cell types. Here you see an increase in B-cells in the Peyer's patches of these mice. But what is the functionality of those cells? We've done some histology, IF, as well as flow cytometry to learn about the germinal center reactions in the Peyer's patches of these mice. And what we've seen is that the germinal centers are smaller in the MASH mouse, and that the germinal center B-cells are dysfunctional. They lack the ability to move to the germinal center where they will interact with the T-cells. As a consequence, we see a loss of IgA in the Peyer's patches, as well as in the gut lumen. So here, we have a system where the B-cells are unable to help the T-cells, and then the T-cells are also unable to help the B-cells make antibodies that, in the gut, these antibodies are protective because they help maintain barrier integrity. So we have a model. We hypothesize, we believe that the dysfunctional intestinal germinal centers in MASH mice lead to a loss of secretory IgA and leaky gut, and all this fuels the inflammatory process in the MASH liver. With that, I'll stop here, and I thank you for your attention. Good morning, everyone. Our fourth speaker is Dr. Matthew Bergio from University of Colorado. His talk entitles T-cell and dendritic cell interactions in the progression of liver disease. Go ahead. Perfect. Thank you, everybody, for showing up to hear my talk on T-cell-dendritic cell interactions in chronic liver disease, and I'd be remiss if I didn't take this opportunity to highlight some of the cool work going on in my lab that's being presented here. So if you're interested in the lymphatics in the liver, Dustin DiNicola is giving a poster on Monday. And if you want to hear more detail about the T-cells, Abby Burtis is presenting today. Thank you. So we really became interested in T-cells in non-viral liver diseases because of a paper we published several years ago in which we wanted to phenotype the immune response in the liver of individuals with cirrhosis. And what we found, which wasn't particularly surprising, was that the T-cells here in the white, and I'm using MASH as an example, are dramatically increased compared to a normal healthy liver. What was surprising to me was that when we quantitated the number of T-cells in MASH, and we see similar things with ALD, that the frequency of T-cells was roughly equivalent to the number of T-cells we see in an individual with HCV-induced cirrhosis. So HCV we know is a T-cell-mediated disease, whereas I thought that MASH was more of a monocyte macrophage-driven disease. So that really reframed our thinking about the progression of MASH, and we wanted to ask the question, are environmental chronic liver disease pseudo-autoimmune disorders? So during a normal autoimmune disease, or I shouldn't say normal autoimmune disease, during autoimmune disease, you have a loss of tolerance. This allows a pathogenic antigen to be presented by antigen-presenting cells. These antigen-presenting cells then present to T-cells, which undergo clonal expansion, get to the target tissue, and kill its target, like beta cells in type 1 diabetes. So today, I will be talking about the interactions between dendritic cells and T-cells, and how that influences progression of chronic liver disease. And for the sake of time, I'm just going to focus on MASH here, and what do we know about dendritic cells and T-cells in MASH? So this is a rapidly emerging field, and a recent paper from Ito Mit's lab nicely identified a population of XCR1-positive CDC1s, or conventional dendritic cell subtype 1, that drive liver fibrosis in MASH. In contrast, Veronica Lucas-Kornick's lab showed that CD103-positive dendritic cells actually protected against steatosis and steatohepatitis. Now this is in contrast, because in the liver, the XCR1-positive dendritic cells are all CD103-positive. So really, there's a little bit of, I wouldn't say controversy, but unresolved nature of dendritic cells in the progression of chronic liver disease. We see similar things on the T-cell side of things, that there was a nice paper that demonstrated and identified a population of cytotoxic T-cells that were CXCR6-positive that were autoaggressive and drove liver fibrosis from Percy Noel's lab. Furthermore, there was several papers that demonstrated that CD8 T-cells were critical for resolution of disease and surveillance against hepatocellular carcinoma. So like the dendritic cells, really the role of T-cells in the progression of, in this case MASH, is yet unresolved. So we wanted to utilize our expertise to deeply phenotype the T-cells in the liver during the progression of MASH, and we went into our high-fat, high-cholesterol diet, which is 40% fat, 2% cholesterol, and we used single-cell sequencing and flow cytometry to really interrogate the T-cells within the liver during disease progression. And like other people, we see mice fed a high-fat, high-cholesterol diet have increased T-cell accumulation in the liver, but what was kind of interesting to us was that not all T-cells accumulate equally. As an example of this, if you look in a chow-fed healthy mouse, the CD4 helper T-cell ratio to the CD8 T-cell, cytotoxic T-cell ratio, is roughly 2 to 1. When we feed mice a high-fat, high-cholesterol diet, this ratio is now skewed more towards the CD8 T-cell compartment. There's more CD8s compared to the CD4 helper cells seen here. So there's not a general accumulation of all T-cells, it's really specifically CD8 T-cells. So we wanted to ask why these CD8 T-cells are accumulated in the liver, and we boiled it down to two possible hypotheses. One is there's general inflammation that's causing the recruitment of cells through the production of chemokines. So we and others have demonstrated during the progression of MASH, there's upregulation of chemokines such as CXCL10 and CXCL16, which can recruit immune cells, whether it be B-cells, monocytes, T-cells, to the inflamed liver. The alternative hypothesis that we're proposing is that there's specific antigens that drive the activation and clonal expansion of T-cells, causing them to migrate to the liver. So what clonal expansion is, is when a T-cell sees its cognate antigen, like this orange T-cell here, it undergoes massive division, and then it will out-populate the other T-cells in the affected organ. So how do we tease apart these two hypotheses? So we decided to use single-cell sequencing of the T-cell receptor. So if the T-cell receptor, because we have many different combinations, potential combinations of T-cells, I think there's 10 to the 15 different combinations of T-cell receptors that's utilized in the human, what we would predict if there was clonal expansion occurring in the liver is that there would be a decreased diversity or decreased usage of T-cell receptors among liver-resident T-cells. And using single-cell sequencing, we find each dot here is an individual mouse. The high-fat, high-cholesterol, diet-fed mice, the T-cells have much less diversity by multiple diversity indices compared to the chow in the blue. As a representative example of that, you can see we gated on the top 10 largest clonal-expanded groups. You can see there's very few clonal expansions in the chow, none really above 1% of total T-cells, whereas in this high-fat, high-cholesterol mouse, you can see there's a T-cell clone that represents between 12% and 15% of the total T-cell pool within the liver. In support of antigen activation, we looked at the transcriptional profile of these T-cells, and on the y-axis here is clonal group size, so non-clonally expanded up to hyper-expanded, greater than 20 T-cells with the exact same T-cell receptor. You can see as you increase in clonal group size, you get skewed towards CD8 T-cells, and they express markers of T-cell activation, including TIGIT, TOX, PD-1, and the autoaggressive chemokine receptor CXCR6. Interestingly, this correlates to a similar phenotype in hyper-expanded cells from humans with MASH-induced cirrhosis. They express TIGIT, TOX, and PD-1, and what was kind of interesting here was that despite the fact that there was equivalent accumulation of T-cells in the liver of individuals with ALD and MASH, the phenotype of these clonally expanded T-cells was slightly different. So this suggests that the inflammatory environment in ALD and MASH program the T-cells differently. So this supports the hypothesis that we have that antigen-driven clonal expansion occurs during the progression of MASH. So the next question we wanted to ask is, what cell type is causing clonal expansion in the liver? And so we wanted to block antigen presentation to prevent clonal expansion. Now, we know from decades of work that many different cell types in the liver can present antigen to T-cells, whether it be B-cells like we just heard about, macrophages, KUFR cells, sinusoidal endothelial cells, hepatocytes, the list goes on and on. But we wanted to focus on the CDC-1s, mainly because of work we did and, you know, Mitt's paper in which we were able to isolate the liver draining lymph nodes from mice fed either a high-fat, high-cholesterol, or chow diet. And we could use flow cytometry to interrogate the migratory DCs specifically. So these are the dendritic cells that migrate from the liver to the liver-draining lymph node. And what we found is that the CDC-1s, but not the CDC-2s, upregulated the expression of a costimulatory marker CD80 when fed a high-fat, high-cholesterol diet compared to the chow. So with Terry Fry's lab and Prashant Francis, we developed a strategy, a chimeric antigen receptor T-cell strategy, to specifically target CDC-1s via the XCR-1 receptor. So these CAR T-cells have a CAR construct that specifically recognizes XCR-1, which is almost exclusively expressed by CDC-1s, and then they degranulate and kill these CDC-1s. And so we fed mice a high-fat, high-cholesterol diet to induce early steatosis, injected either anti-XCR-1 CARs or mock CAR T-cells, waited several weeks, and looked at T-cell activation in liver pathology. And what we found was that, yes, these anti-XCR-1 CAR T-cells could effectively deplete CDC-1s from the liver, as well as the secondary lymphoid organs. As you can see here, there's only 7% of XCR-1-positive DCs compared to 40% in the mock. But when we looked at the liver pathology, what was interesting to us was that if you compare the mock to the anti-XCR-1, there's no difference in fat accumulation in the liver or steatosis. So we're not influencing diet uptake or fat deposition in the liver, but what we are doing is we are reducing ALT levels back to almost baseline. Here's the XCR-1 compared to the mock. The activation score is measured by a blinded pathologist, as well as we essentially normalize picroserous red staining and fibrosis. So this suggests, like the paper from Ito Amit, that XCR-1-positive CDC-1s are really driving disease progression in a mouse model of MASH. So the next question we wanted to ask is, what is the antigen? What is the antigen presented by XCR-1-positive dendritic cells that drives T-cell activation? And so we focused in on this T-cell receptor MHC peptide complex interaction, and we wanted to do this in an unbiased manner. So what we did is we took the sequences from the clonally expanded T-cells from a mouse with a high-fat, high-cholesterol diet, cloned it into a reporter cell line, and this reporter cell line actually fluoresces green when stimulated through the T-cell receptor. And this is anti-CD3 as a mimic of T-cell receptor signaling. You can see it's green. Without anti-CD3, there's nothing. But how do we do this unbiasedly? And on our campus and other places, one of the ways that you can do this is called a combinatorial peptide library. And this has been used many times to identify antigens in autoimmune disease as well as infectious challenge. And what combinatorial peptide libraries are is it allows you to unbiasedly interrogate the preferred amino acid at each position along a peptide, scanning either eight through 12 amino acids long. So you systematically go through and test every single position and see which amino acid is preferred at that position. And as a representative example of this, I'm showing you a TCR from a mouse fed a high-fat, high-cholesterol diet, and you can see arginine here is highly stimulates this T-cell receptor, whereas n-proline doesn't. So once we went through all the different positions of the peptide and discovered the preferred amino acid, we can use traditional deconvolution to really narrow down and use individual peptides to stimulate this TCR. And what I'm showing you as an example here is a 9-mer. Eight of the nine amino acids are exactly the same, with the exception of amino acid six. When you have a phenylalanine here, it does not stimulate the T-cell receptor at all. If you have a proline here, now you get robust T-cell receptor activation. So from that, we can go from a pool of millions of peptides down to a pool of a couple of hundred and use the Swiss database to predict where these antigens come from. And so as an example of this, we see roughly half the predicted antigens come from self, half come from formin. So a kind of a flip of a coin here. And so how do we figure out exactly where the antigen's coming from without testing 250 peptides? And so what we decided to do in the short term was go back to our days of hardcore T-cell immunology, and we have this vaccination strategy that we've shown when you combine a TLR agonist, in this case LPS, along with a CD40 agonistic antibody, this induced robust T-cell responses to a foreign antigen. So this is a foreign peptide combined with LPS and anti-CD40. So we immunize with a foreign antigen or the peptides derived from that stimulate in vitro a high-fat, high-cholesterol TCR, and seven days later, we can look for T-cell responses in the liver and the spleen. And what we see, as we've published many times, a foreign peptide induces a very nice, robust T-cell response, as measured by interferon gamma, ICCS. What was surprising to us is the high-fat, high-cholesterol peptides induced nothing, no T-cell response in either the CD8 or the CD4 T-cell compartment. So this suggests to us that it's potentially a self-peptide that's driving T-cell activation during the progression of MASH. So this leads us to a model of CD8 T-cell activity in MASH. We think that the chronic inflammatory environment breaks tolerance. This allows for productive presentation by XCR1-positive dendritic cells to CD8 T-cells. They undergo clonal expansion, and then they target the hepatic parenchyma for progression of liver disease. So with that, I have to thank the people involved in these studies, my lab, Destiny and Abby are here, so please go see their posters. We have a rotation student, Natalie, Prashanth Francis, and Beth Tamarini on campus, as well as collaborators at Minnesota and NOVA. And lastly, I would have to say I would not be here if it weren't for the mentorship of Hugo Rosen, and we tragically lost him a couple years ago. And the AASLD Foundation was kind enough to set up a fund to help the career development of young investigators. So if you have the inclination and the ability, please donate. Thank you. Thank you, Matt. It's my pleasure to introduce the last speaker, Dr. Li-Chen Ma, from NCI. Her talk entitles Neighborhood Differentiation in HCC with Multi-Regional Single-Cell RNA Sequencing. Go ahead. Yeah, thank you so much for the introduction, and thanks so much to Dr. Monga for the very kind invitation. So I'm Li-Chen, and I'm a tenure-track investigator from the Cancer Data Science Lab of the National Cancer Institute with a co-affiliation of the Liver Cancer Program. And my lab mainly focuses on understanding tumor heterogeneity by using single-cell and spatial approaches. And specifically, we develop computational strategies to understand tumor heterogeneity in the context of tumor initiation and evolution. Today, I'm going to talk about our understanding of tumor heterogeneity in hepatocellular carcinoma by using single-cell analysis of the communications as well as visual analysis of the cellular neighborhood. So I have no financial relationships to disclose. As you may know, liver cancer comprises mainly two clinical subtypes. Intrahepatic cholangiocarcinoma, HCC, and intrahepatic cholangiocarcinoma, ICCA, accounting for about 90 percent and 10 percent of liver cancer, respectively. Each clinical type further consists of different molecular subtypes determined based on molecular features. This complicated molecular landscape of liver cancer is related to the complex etiological factors such as age, gender, ethnicity, HPV, HCV, HDV infections. In addition, liver cancer is a dynamic disease with constantly evolution of tumor cells in maintaining their continuous survival. Selective pressure from tumor microenvironment such as nutrients and space allow sub-subclones to expand while others become extinct or remain dormant. This evolutionary feature of cancer is driven by selecting favorable phenotypes for survival fitness. As such, the molecular landscape of liver cancer is very complicated with vast molecular heterogeneity both between patients and the single patient, along with immune invasion, tumor progression, metastasis, and treatment resistance. Therefore, our ability to understand tumor heterogeneity is very critical to improve liver cancer treatment. Single-cell analysis has been very successfully applied to characterize both tumor cells and the cells in the tumor microenvironment as demonstrated in both my work as well as the work from other people. But for single-cell approaches, the limitation is the missing of a spatial context, which is very critical to understand the cellular functions of the cells. In addition, the growth and progression of a solid tumor are influenced by a community of malignant cells and non-malignant cells. So we've seen the community and the tumor cells and non-tumor cells, they continuously communicate with each other. So based on these readings, we determine the cellular communications between tumor cells and non-tumor cells to try to understand tumor heterogeneity using single-cell and spatial approaches. We first performed a single-cell profiling of tumor samples collected from multiple locations of a single-tumor region. And specifically, we collected three regions from tumor core, one from tumor border, and one from adjacent normal tissue. With the collected tumor samples, we performed a single-cell RNA sequencing to build a multi-regional single-cell landscape in liver cancer. And this study was in collaboration with Dr. Sophia and Dr. Jens Markler from Germany. For the malignant cells, we found patient-specific clusters, which is consistent with previous studies of the inter-tumor heterogeneity between patients. But interestingly, with single patient, the tumor cells are very well mixed from different tumor regions, including three tumor core and one tumor border, indicating smaller inter-regional heterogeneity than inter-tumor heterogeneity. We also determined the landscape of non-malignant cells, including T cells, B cells, macrophages, et cetera. Based on the landscape of the malignant cells and non-malignant cells, we determined the communications between tumor cells and different type of non-tumor cells in tumor microenvironment by using a method called CellphoneDB, developed by Dr. Sarah Taksman's lab. So in this dot plot, each dot represents a ligand receptor pair. The two colors, purple and gray, represent the directions of interactions, either from tumor cells to tumor microenvironment, or from tumor microenvironment to tumor cells. And the size of each dot represents the proportion of tumor regions in identifying a specific ligand receptor interaction, with the largest size indicating the occurrence of a specific pair in all the tumor regions, including three-tumor CAR and one-tumor BADR. As you may notice from this dot plot, the majority of the dots are this very large size, indicating that the communications between tumor cells and tumor microenvironment are very stable within a specific case. In addition, patient-specific communication networks are evident. For example, this group of communications only occur in this hepatocellular carcinoma patient. This is also revealed from this heat map. It's very stable, but unique communications within each individual case. To demonstrate whether those patient-specific communication networks are biologically important, we extended the search for ligand receptor pairs in additional 24 samples from the NIH Clinical Center. And we identified two major clusters based on the ligand receptor interaction activities. And interestingly, we found that the two clusters of patients they have significantly different overall survival. Because this cohort contains both HCC and ICC patients, and we also analyzed HCC patients separately, there's a significant, there's a kind of a trend between the two clusters of patients. We didn't perform survival analysis for ICC because they were all finding in one cluster. So this analysis, this really demonstrated that the communications between tumor cells and tumor microenvironment can reflect the intrinsic tumor biology. To validate, kind of demonstrate robustness of those ligand receptor interaction pairs in HCC prognosis, we validated the top two pairs, the LGALS9, SLC1A5, SPP1PTGR4, representing the communications between tumor cells and tumor-associated macrophages by using RNA-scope multiplex in-situ hybridization. And consistent with our single cell analysis, and we found that patient with high impression of both the two pairs had a much shorter survival than patient from the low impression group. So this multi-regional single cell analysis demonstrate that there's very stable, but unique communication networks otherwise in tumor biodiversity within each patient, and we call it log and key features. And those features are very stable, and it can reflect inter-tumor heterogeneity. So this multi-regional single cell study can allow us to understand the tumor biology at different locations within a tumor region, but we still don't know the spatial locations of the individual cells. To define the spatial organization of the cells, including immune cells, stroma cells, tumor cells, in hepatocellular carcinoma, we used a platform called CODACS, and the CODACS use an antibody panel and can juggle with dilabeled kind of reporter for highly specific detection. And we performed the CODACS staining for tissue microarrays of more than 200 HCC cases, and the marker list, including markers specific to T-cells, B-cells, macrophages, et cetera. The computational analysis of this project is leading by my former post-bac fellow, Moller, Evan, and this study was in collaboration with Dr. Noemi K.D. and Dr. Xinwei Wang from NCI. So here is a representative HCC curve. It's very specific staining for the nucleus, the epithelial cells using E-cadherin beta-containing HNF alpha. The CD4 and CD80 cells, CD31 part of endothelial cells, CD11C dendritic cells, CD20 B-cells, and as well as the CD68 and CD163 macrophages. So with the result, this visual single cell landscape of the cell types, a central question is what and how do we learn from the spatial omics data? A potential analysis is to understand the cellular neighborhood to determine the communications between the cells and whether a cellular neighborhood are spatially enriched, very similar to this neighborhood analysis to define the locations of retail business in California. So to understand the cellular neighborhood or understand the cellular communications, we apply the Varanian networks based on Delaunay triangulation of individual cells. And so, for example, in this Varanian network, this adjacent cells connected within 100 pixels of 40 microns were used to identify specific cell-to-cell interaction. And here is an example of the interactions between tumor cells and CD80 cells. So based on the result of the cell-to-cell interactions, we performed hierarchical clustering analysis of the interactions between tumor cells and different type of immune cells. And we identified four major immune classes, IC1, IC2, IC3, and IC4. And interestingly, we found that patient from IC2 and IC4 had much better survival than patient from IC1 and IC3. And we found that in the best survival group, IC2, there's a significant increase or enrichment of CD80 cells to tumor cell interactions. And further analysis of the interactions among different immune cells demonstrate that there's an enrichment of CD40 cells and CD80 cell interactions in the best survival group. This analysis demonstrate that the spatial interactions between tumor cells and immune cells can dictate intertumor heterogeneity. The limitation of the CODEX platform is the antibody panel and the list of markers we can stain. So to build a comprehensive spatial single-cell landscape in liver cancer, we're applying the COSMICS platform. And the COSMICS allows the quantification of 1,000 RA targets and 64 protein targets for FFP tissue or French foreign tissue. So this platform was just released this year. So by applying this cutting-edge platform, we perform profiling of a list of samples and resolve the spatial single-cell landscape of more than two million cells. And this study was in collaboration with Maria from the NCI and Dr. Jens Marker from Germany. And the computational analysis was leading by my postdoc fellow, Meng. So we still determined the spatial single-cell profiles. We annotated cells as tumor cells, macrophages, T-cells, B-cells, et cetera. And then Meng mapped the cells back to their spatial locations, where we found a very nice separation of the cell types from tumor tissues and non-tumor tissues corresponding to the HNE image. And we also determined the cellular neighborhood by using niche analysis. So this is still an ongoing study and hope I will share with you a complete story very soon. So to summarize, liver cancer is very heterogeneous and there's a vast molecular heterogeneity between patients called inter-tumor heterogeneity and the single patient called inter-tumor heterogeneity. And we found that the tumor immune communications determined by using spatial single-cell profiling can drive inter-tumor heterogeneity. And the single-cell approaches is superior in determining those cellular communications, but spatial approaches is very promising. We also have an online interface called Single-Cell Atlas in Liver Cancer for researchers to explore our dataset. And most recently, we are updating this website to include a more in-house single-cell dataset. Lastly, I would like to thank my lab members as well as my collaborators. Thank you so much and I'm happy to take questions. Thank you. All right. Thank you to all the speakers who presented today. We're now having a Q&A session, so please, if you have questions for any of our speakers, line up at the microphone. And as a reminder, we will be having a business meeting for the liver cell biology special interest group immediately following this session. Please. Oh. Thank you for that session. That was awesome. I just, I have a few questions, but I just wanted to pick Enos' brain briefly, Enos, if it's okay. So you had mentioned that you used the bead-based spatial evaluation, and each one of those beads are about 50 microns. And then from what I understand, you were able to use that with single cell to then evaluate cellular interaction. But in my mind, with the beads capturing more than one cell type, it's almost like you're getting multiple sort of bulk sequencing data. So how did you sort of tease that out with single cell data? Thank you for your question. Yeah. I have a question that bioinformatician working with us would be better to answer, but the short answer for a biologist as me is that, so we use this 10X genomics, 10X genomics. So these are spots, 15 microns, and from one spot, from the center of one spot to the other is 100 microns, so yes. We do have five to 10 cells in one spot maybe, depends on the size of the cells. So what we do is that we use, I believe, the fissures that test to integrate single cell. So how do we integrate them? We know the annotations, we know the markers from the single cell, we know the markers of each cell type, and we use those markers to see in each spot the percentage of the expression of each marker and to see what cell markers are next to each other. So it's mRNA-based. We need to confirm this with IF, et cetera, but yeah, it's basically this. Robert Schwabe, Columbia University. First of all, great talks by everybody and great organization. I mean, this is a full room, amazing. So I have a question for Matt, or actually two, one very obvious question. Do you see the clonal T cell expansion in patients? And maybe it's a little bit more heterogeneous. And my other question is, where does the education happen? Is that in the liver, the T cell education in the liver, or is it in the lymph nodes? So it does happen in patients. It's highly variable. I mean, the patient samples that we've used have been cirrhosis. So whether that's a cirrhosis-dependent effect or a progression of disease-dependent effect, we don't know. Education, again, we don't know, but traditional immunology dogma would say that it's the regional lymph node, the portal and the celiac. And that's where we see during disease progression in the mouse, you get an accumulation of these migratory DCs as disease progresses. So initial priming, I think, happens in the portal and the celiac. And David DeBattista from Bethesda, congratulations for the presentation, very nice work. I have just one question for Dr. Monga. So the knockdown of the several genes are in the embryonic stage or during the adult? Because I'm wondering if it's an impairment due to the development of the organ, or can be a mature tissue with the knockdown of that genes have the same impairment in the localization of the gene expression? Yeah, that's a very good question. So the analysis that I presented here was mixed, meaning that some of the models were inducible models and others were developmental. So the live one, CRE, for example, which we use to delete from endothelial cells, Wnt2, Wnt9b, or LRAT, CRE for Wntless and stellate cells, they are all developmental. So as animal develops, as these expression of these genes is induced, the CRE decombination is functional and it deletes. But for beta-catenin, LRP56, we have done adeno-associated virus. And that is, you know, within 10 days of single IP injection, it's enough. And the phenotypes are very similar. So I do think there may be some developmental play in there, some adaptation that may happen. But I think overall, the phenotypes are preserved and conserved across the board. Thank you. Thank you. Dave Kleiner from the NCI. One question I get asked a lot as a pathologist is, especially when people are worried about autoimmune disease in the liver, is plasma cells. And so this question is for Dr. Rivello. So you do see plasma cells in ash, typically with the more aggressive inflammatory infiltrates. This has always been a shock to my colleagues in hepatology, but I have seen them for many years. So what do you know about the plasma cells? What are they doing? Where do they come from? I love that question. I think in mice, we see some plasma cells in the liver. The frequency is very low. The better question is, what are they doing? And I think what's happening is that they're undergoing antibody production. They do that. And what we see is that the type of antibody is very important. And so if it's an IgG, that would be an antibody that can actually activate the stellar cells. And we're working with any on some of that aspect. But then when the plasma cell is an IgA positive cell, and if it's coming from the gut, and some of them do, it seems that those cells, those antibodies are protective in the intestinal environment. And so we do see evidence of increased presence of plasma cells. I think we're trying to understand a little more, what are they doing? And not all of them are equal in function, I would say is my takeaway. Okay. It's a very interesting area. Thank you. Wonderful talks. Thanks. Hi. I'm Chico Zeri from Yale. I have a question for Dr. Manga. So obviously, there's been a lot of work showing the role of Wnt-beta-catenin signaling in liver regeneration. And I'm curious to know your thoughts on the contribution of Wnts from these different cell sources. And also, there's also been some work, I guess, recently showing the contribution of Wnts from comfort cells as well. And again, I'm curious to know your thoughts about how modulating with these Wnts from different cells also contribute. Yes, that's a very long answer. I'm going to just keep it short. So again, I think the physiological Wnts that are expressed in endothelial cells, Wnt2 and Wnt9b, are the Wnts that are activated in the same endothelial cells all across the lobule after partial hepatectomy. In toxicant-induced injury, like in, for example, acetaminophen overdose, macrophages become the source of Wnts. And we haven't, but Udayanapte and his group from Kansas University has shown the role of macrophage Wnts, and Bharath Bhushan also from Pittsburgh. And from stellate cell Wnts, you know, is interesting. We find, and we have to discuss more with Robert at some point in time, is that when we knock out Wnt-less, meaning that we completely knock out Wnt secretion from stellate cell, the phenotype is very, it's opposite. So meaning that these animals do not do well. They don't regenerate their livers well. So and why is that? Is that a direct effect that a Wnt from stellate cell is required for regeneration, or is it because what we think that hepatocytes in a LRAT Wnt-less animal are metabolically hyperactive, and they just do not go into proliferative stage, and hence are very poor at proliferating because they're already engaged in a lot of metabolic function. So again, this is, you know, mostly a conjecture, and we need to prove that, but that's where we are with this entire thing. It's context-dependent after injury where the Wnts are going to come from, but it's absolutely critical for normal regeneration. Yeah. Thank you. Thank you. Well, thank you. Thank you. It was really outstanding. This is Samar Ibrahim from Mayo Clinic. I would have a question for all of you. I'm going to limit to one, to Matt. And it was really inspiring at the end of your talk, like, speaking about this break of self-tolerance as a driver of the inflammatory response in MASH, and if we can put it back in context to the clinically, like, there is a subset of patients with steatosis who would never proceed and develop MASH, and I'm wondering if this is something you thought about, what are the drivers of this break of self-tolerance? I know it's kind of a loaded one. That is also a long answer. I think there's probably some interplay. I focused everything on the CD8 T-cell side of it. There's probably a progressive decline in regulatory T-cell function. So I think, at least in the mouse, these auto-aggressive CD8 T-cells are there, but they're, you know, in a normal environment, healthy environment, they're being suppressed by the regulatory and the tolerance mechanisms within the liver. And I think, as Dr. Vovello said, the TLR stimulation due to leaky gut, I guess, is, I would propose would be the driver of that initial break in tolerance. And it's the chronic stimulation through TLRs. Thank you. Hello, thank you all for great talks. I had a question for Dr. Munga. I'm Elizabeth from Baton Rouge Pennington Biomedical. With your respiratory phenotyping and your knockout, I was really interested in the succinate story because it actually looks like your phenotype is in the succinate and not in the fatty acid. That is correct. So. Is in the? Not in the fatty acid, the one I mentioned last. It is in the fatty acid. In the ADP response after the fatty acids? Okay, I wasn't sure if there was a difference there. But. We can talk more. Yeah, okay. I'll show you the data. Thanks. Sorry. All right, let's give one more hand to our speakers for their fantastic talks today. Thank you.

liver health

liver disease

spatial cellular interactions

liver fibrosis

metabolic zonation

B-cell interactions

T-cell interactions

dendritic cell interactions

hepatic stellate cells

extracellular vesicles

Wnt signaling

autoimmune responses