Well, good morning everyone. Thank you so much for inviting me. I'm honored to have been invited by the Endocrine Society for an oral presentation this morning. My name is Neil Hall. I am a current third year medical student at Mercer University School of Medicine in Savannah, Georgia. And I'm excited to be sharing my research with you this morning. This is a project I've been working on with my colleagues at Mercer since the beginning of medical school in 2019. The title of my talk this morning is Cloning, Sequencing, and Characterization of Aromatase Interacting Partner in Breast, or AIPB. So I'd like to start with a brief background. As many of you are well aware, breast cancer is one of the most common invasive cancers in women in the United States, with statistical data indicating that one in eight women will be affected in their lifetime. There was a recent cancer survey in recent years that showed as of 2012, over 800,000 women were affected. However, as of 2019, that number decreased to 200,000, and most of that can be attributed to advancements in therapies and earlier diagnostics. What we do know is that breast cancer is mostly dependent on the availability of estrogens. And that's why we do have therapies such as aromatase inhibitors, or other treatments such as tamoxifen. However, there are well-documented side effects with these, as well as the development of resistance prior to treatment being completed. Further, there's no therapy that's currently available for triple negative tumors. And that's why we need new therapies and targets for this disease. So in our lab, what we noticed is that in all women, during both menopause and in tumorigenesis, there's a significant increase in estradiol production. And what is known is that estradiol is synthesized from testosterone via the enzyme aromatase. Well, Dave Gosch's group actually saw the 3D configuration and provided significant advancement on the structural side. We felt that this didn't fully explain the mechanism of action, and that's not fully known yet. Now, what we decided to do in our lab to further explore this mechanism of action is we did a short activity experiment where we isolated ER fractions of breast tissue from the same patient, where one side developed tumor and the other side was unaffected. And what we found was quite significant. As you can see here, there was a three-fold increase in estradiol production in the tumor-affected breast as compared to the non-tumor. And further, in the absence of a cofactor, or with just the buffer labeled as the mock here, there was no estradiol production. And so that led us to conclude that there must be a cofactor that's associated with aromatase. So for us to identify the possibility of a cofactor, what we did is we isolated and purified in the plasmid reticulum fractions of breast tissue samples from unaffected tissue, and then separated those 10 kilodalton fractions via SDS-PAGE. We then took those 10 kilodalton fractions to one of our collaborators' labs, and he performed mass spectrometry. And what we found is this 11-mer peptide, which is located here on the screen. Now, subsequently after that, we cloned a full-length cDNA of this from human breast tissues, and then did a variety of biochemical experiments following that. So how did we actually clone this protein? So we used the RACE technique, which is the rapid amplification of cDNA ends, where we cloned the closest three prime ends and the closest five prime ends. And then subsequently, we were able to clone a full-length 927-base pair cDNA that had a poly-A tail at the end. And what we found is that after the 621-base pair, there was a stop codon. And this protein that I had just mentioned on the previous slide was present after the 132-amino acid position. Now, what we saw here is that the molecular weight is calculated to be 21.6 kilodaltons, and this 207-amino acid protein was not present in the database, which confirmed it as a new protein. So in our next step, we wanted to determine what the significance of this protein may be in aromatase activity and estradiol synthesis. And so what we did is we knocked down its activity using an siRNA. And as you can see here, we did this with both MCF12, which is a non-tumorigenic wild-type cell, and a tumorigenic T47D cell line. And when you knock down the expression of this protein, you get a three to five-fold increase in estradiol production in both of these cell lines. And that, for us, suggested that this protein is possibly associated with aromatase. So now, on this slide, I'd like to go over the biochemical experiments that we did and kind of give you a summary. And in subsequent slides, I'll go through the actual experiments and show you the background for that. So first, what we found is it is not present in human ovaries, and it is not a genomic reorganization when we used most five-prime and three-prime primers of both ovarian cDNA and full genomic DNA. We then used conditional expression under the control of a tetracycline promoter. And with the addition of doxycycline, we found that expression of this protein increased, and there was a subsequent decrease in estradiol production. Now, there was ineffective expression with the addition of cyclic AMP and cyclohexamide, and that suggested that it is not transcriptionally regulated or inhibited, and that's also a pattern that's known with aromatase. Now, subsequently after that, we used an estrogen stimulator known as Xeronal, which is a PGE2 analog. And what we saw is that the expression of this new protein increased, but only in the non-tumorigenic cells, and that led to a reduction in estradiol. We saw the opposite effect in the tumorigenic cells. Now, the next thing that we saw was that this protein was co-localized with aromatase at the ER, as well as it was able to be pulled out with an aromatase antibody. And most of this actually came from co-immunoprecipitation reactions that we performed. Now, what is known is that aromatase is a 57 kilodalton protein, but when it interacts with this new protein, it forms a 79 kilodalton protein as shown here on the screen. And that's why we conclude that it is associated with aromatase in breast, and therefore, we have turned this new protein aromatase-interacting partner in breast, or AIPB. Now, as I mentioned before, I wanna go through some of the actual experiments that I just summarized on the screen. And so these are the RT-PCR experiments where we show that it is absent in the ovarian RNA. Now, as you noticed here, this is human tissue samples, and over here on the right is the cell lines. I just wanna draw your attention that it is not present in the ovarian RNA in either of these. Now, next, these are the electron microscopy experiments that were done, and as you notice here on the top left-hand side of the screen, this is an antibody against aromatase, which indicates with the red line that it's at the endoplasmic reticulum. And this slide here on the top right shows aromatase and the antibody that we used against AIPB, indicating in blue that it is located together at the endoplasmic reticulum. Now, again, this down here just shows the AIPB antibody only itself in blue, and then on the bottom right is our control, that's GRP78, that's a known resident protein of the endoplasmic reticulum, just confirming that we did have the breast tissue itself. Now, here, these are the Western blotting reactions where we show that it interacts between AIPB and aromatase. And as I mentioned before, aromatase is a 57 kilodalton protein. However, when it interacts with AIPB in both the wild-type MCF12A cells and in the T47D tumorigenic cells, it interacts together and forms this band at 79 kilodaltons in both of these cell lines. And now, as I had mentioned, when we used an estrogen stimulator known as Xeronal, we saw a significant increase in the production, or excuse me, of the expression of AIPB, which led to a reduction in estradiol production, but only in the non-tumorigenic cells. What we found is that tumorigenic cells have reduced levels of this AIPB protein, and that leads to a significant increase in estradiol production in these cell lines, as noted here. So, to summarize all these different findings that I've just discussed with you, it's well-known that aromatase is essential for the production of estradiol from testosterone. However, we propose that AIPB, either the absence of or reduced expression, leads to a significant increase in estradiol production. And so, in conclusion, AIPB regulates aromatase catalytic activity in estradiol synthesis. I'd like to just thank my collaborators, Dr. Brendan Marshall, for providing the transmission electron microscopy experiments, as well as Dr. Elena Rail. She helped us to get the initial breast samples immediately after surgery. I'd also just like to thank Mercer School of Medicine for helping support this research, as well as my other lab members, just for the collegiality and a lot of fun that we had in the lab, as well as financial support from the Landings Women's Golf Association, and Dr. Robert Visali, who was our research chair in Savannah. And I also would be remiss if I didn't mention my wonderful wife and our three children. They just provide support and encouragement for me every single day. Thank you so much for your attention. I would be delighted to answer any questions you may have. Thank you. Thank you, Neil. I'll remind folks, this is being live-streamed, so if you're gonna ask a question, please use the mic and introduce yourself. Zeynep Mollekardan, University of Illinois. The protein is new, novel, but are there any hits when you look at the tumor mutation data or any loss of this, you know, a location of the gene in tumors when you look at the data that's available publicly? I'm sorry, so you're asking why we didn't see the location of the gene in tumor cells? No, so when you look at the publicly available data, tumors, the mutation data, or, you know, so would you see this protein being hit in tumors, actual breast tumors? Yes, so it is expressed in breast tumors as well, as well as normal and unaffected breast tissue. Okay. An interesting discovery, Marina Halls, New York Medical College. So you saw it in the ER, so do you speculate that it might enter the secretory pathway and be present in exosomes and microenvironment of the tumor, or I guess microenvironment in the breast? Correct, so that's what one of our main findings was, and thank you so much for that question. So seeing that it's absent in the ovaries, it's only present in the breast with aromatase at the endoplasmic reticulum, which is where we propose that it actually has its interaction with aromatase, causing an inhibition. Yes. Hi, Shantanu Ghosh, Illinois. Really good talk, I really enjoyed it, thank you. So I had one question, which is, you mentioned that when you knock out this protein, or knock it down, you do see an increased estradiol production, right? So does that affect proliferation? Have you checked proliferation of these breast cancer cells that are dependent on estrogen-regulated growth? We actually haven't checked the proliferation specifically. What we've been doing is radio-immunoassay to look for the activity of estradiol specifically. That is something that we're continuing to work on, and we do actually hope to have more data available in the next year. Thank you. Stéphane Laporte from McGill, so very nice presentation. So can you tell us a little bit more about the motif, or how AAPD is working? So what do you know about the protein itself in terms of structure, or structure probably not, but motifs, or? So we still don't know fully the structure. What we do know is just the full amino acid sequence. Currently, we haven't done any experiments to actually look at the full structure to see what exactly the mechanism may be. We've just looked again at activity experiments and activity data. Do you know, have you identified motifs that can explain its function also in terms of binding partner? No sir, not yet. Thank you. Okay, great, thank you very much. Next speaker is Surjeet Singh Gupta from the Hormel Institute at the University of Minnesota. Thank you. Thank you for the introduction. And I'd like to thank the members of the steering committee for selecting our abstract for this oral presentation. As the title suggests, I will be discussing about sex ed metabolism in endocrine therapy-resistant ER-positive breast cancer cells, which is a developing story in our lab recently. So I don't have any disclosures. So as Neil already mentioned about breast cancer incidences and death, I just want to reiterate that breast cancer is the most diagnosed cancer in the world as well as in the country. And breast cancer is the second leading cause of cancer death in the country with more than 43,000 deaths every year. Most of the breast cancer at the time of diagnosis are ER-positive breast cancer, and this gives just a timeline of different effective drugs that have been approved by FDA over these decade, several decades, first being the tamoxifen. And even before tamoxifen, high-dose estradiol was used, but I'm not going to talk about that. But if you see the tamoxifen and fulvestrin, they fall under the class of anti-estrogen. This was followed by different aromatase inhibitors. And in the last six years, we have seen a new class of drug called CDK4-6 inhibitors, which are being combined with fulvestrin or aromatase inhibitor in the clinic. And I have just one clinical trial just to show that combination of one of the CDK4 inhibitors with fulvestrin gives benefit in terms of survival, whether it is progression-free survival or overall survival, the combination is beneficial. But also one thing is clear looking at these graphs is that 50% of the patients will progress within 21 months, and 50% of the patients will be deceased within 54 months. So which means that we need to understand the mechanisms, underlying mechanisms, which help evade these therapies in the clinic. To understand that, we chose three different cell lines. Actually, they are all MCF7 variants. You can see it on the left side. It is the parental MCF7, which are estrogen-dependent, which means estrogen stimulates the growth. And this estrogen stimulation can be blocked with the 4-hydroxytamoxifen and fulvestrin. The LCC9s are on the other side of the spectrum, which are estrogen-independent, which means they are indifferent to estrogen treatment, and they are resistant to endocrine therapy agents, 4-hydroxytamoxifen and fulvestrin. While LCC1 cells, these are somewhere in the middle of the spectrum where they are though estrogen-independent, but they are also sensitive to endocrine therapy agents, like 4-hydroxytamoxifen and fulvestrin. So what we asked the question, and we knew about that LCC9 takes up more glucose than other cell lines. We repeated that experiment, and we saw that LCC9 takes up, glucose uptake was maximum in LCC9, followed by in LCC1 and MCF7. But what we wanted to know is how this glucose is actually metabolized in these cells, and if there is any difference between glucose metabolism in these three cell lines. To look at that, we used C13-labeled glucose, which is a non-radioactive isotope of carbon. And in this case, we have the glucose molecule, we have all the six glucose molecules were labeled with C13. We use this heavy glucose to grow the cells in, and after 16 to 18 hours when there was a steady state, and this is all the basal growth condition, we didn't treat the cells for anything. So after the steady state labeling, we harvested the cells and looked for the C13 metabolites, particularly for glycolysis metabolites and TCA cycle. I'll just go over this very quickly. We all know that when glucose is metabolized, we get acetyl-CoA, and the acetyl-CoA in this C13-labeled glucose, we will get two carbons in the acetyl-CoA, which will be heavy carbon C13. And when in the first cycle of TCA cycle, it integrates with oxaloacetate, we expect to get citrate, which has two carbon-13s. And in the subsequent cycles, we will get M plus four and M plus six citrate. When we looked at the citrate levels in our cell lines, what we found was that M plus four and M plus six citrate was completely missing in LCC9 cell lines, which were resistant for hydroxytamoxifen and fulvestrin. So this indicated that the TCA cycle was somehow truncated or it was impaired in the LCC9 cells, which prevented the regeneration of citrate in these cells. What we also found in these cells was succinate-to-fumarate ratio was very high in these cells, and the succinate-to-fumarate conversion is actually done by an enzyme called succinate dehydrogenase, which is a bidirectional enzyme, basically. It can function both ways, and you will see why I stress on that. But, and SDH is actually a multimeric protein, and one of the subunits of this protein is SDH5, which we saw in LCC9 cells, which was underexpressed in these cells. And one of the readout of succinate accumulation in cells is HIF-1-alpha stabilization. So we also see HIF-1-alpha stabilization in LCC9 cells. So initially we thought probably because SDH5 component was downregulated in LCC9 cells, that's why we were getting the succinate accumulation in this. But when we looked carefully, we found that fumarate, which is the succinate converts into fumarate by this enzyme, fumarate was largely unlabeled, which means that fumarate was not coming from the glucose molecule. And if we think that SDH activity was blocked, then we should have seen at least some accumulation of M plus two succinate, which we didn't find. We actually found the opposite. Succinate was also unlabeled in these cells. So we then looked for the SDH activity per se, and we found that LCC9 also had SDH activity, which was comparable to LCC1 cells. So combining these findings, we think at this moment, and this is also again developing story, that it is actually the fumarate which is getting converted into succinate. So it is a reverse reaction that is taking place in these cells. So when we block this enzyme using a dimethylmalonate, which is an inhibitor of succinate dehydrogenase, what we found was that at a lower concentration of this inhibitor, it didn't do much in the cells, but when we combined it with the full vestrant or 4-hydroxytamoxifen, it completely blocked the growth of the LCC9 cells. And on the right side, you can see that it is a synergistic effect when both compounds are present, they are more effectively inhibiting the growth of these cells. We looked at what is happening in the cells. So they are actually cytotoxic. Again, these red cells represent dead cells in the plates, and when you treat them with full vestrant or dimethylmalonate at the lower concentration, they don't do anything, but the combination is toxic and they kill the cells. We confirmed this in another cell line, in another non-MCF7 cell line, which is T47D, and the T47D HTs are actually resistant to hydroxytamoxifen as well as to full vestrant, and we get similar results in both the cell lines. We further looked at the HIF-1-alpha stabilization. So if SDH is actually blocked, and if the reaction is coming from succinate to fumarate, we should see more succinate accumulation, and as a result of that, more HIF-1-alpha accumulation. But we see exactly opposite, and that's why we think that our hypothesis is probably correct, that when we block the succinate dehydrogenase enzyme, we see that there is a decrease in HIF-1-alpha levels in these cells within four hours. And we tried to look at, to add back succinate to these cells using dimethylsuccinate to see if HIF-1-alpha levels come back. They did partially come back, but this is, again, a four-hour treatment. We are looking at prolonged treatment to see if it comes back to the vehicle-treated cells. So with that, I would summarize our findings. The TCA cycle is impaired in endocrine-therapy-resistant LCC9 cells with high succinate-to-fumarate ratio, and targeting succinate dehydrogenase resensitizes the cells to LCC9, to full Western in LCC9 cells. In the future directions, we want to look at where this fumarate is actually coming from, because if fumarate is not generated by glucose, it is coming from somewhere else, and what is its role in endocrine-therapy resistance? And then we also want to look at and determine the cell death pathway induced by STH inhibition and full Western, in combination with full Western. What kind of cell death pathways are activated? With this, I would like to thank Dr. Robert Clark. Without his support, I wouldn't have completed this work. Lab members, Carla, Lou, Anil, and also I'd like to thank Lena and her lab members. Shared resources from the Hartman Institute, particularly the flow cytometry, and funding from CDMRP, DoD, and the Hartman Foundation. And with this, I will stop, and I thank you for your attention, and I will be happy to take any questions you have. Thank you. coming to the mic, I'll take advantage of asking one as chair. So this is really fascinating, and I'm wondering if you have any ideas or any speculation on what might make that reaction work in reverse. So for example, are there cofactors of the enzyme that bias it in a backwards versus forwards direction? What do you think's going on? So there are very few examples of where this reverse reaction has been seen, and particularly it has been seen in ischemic conditions. So in ischemic conditions where there is low blood flow, they have seen that this reaction happens, but we don't know what factors. And one thing which I kind of, it's a wild speculation, but if STH-AF2 downregulation of this factor is actually somehow facilitating the reaction in the reverse direction or not, and I want to look at that thing, yes. Sorry, great talk. My question is about the effect of full western treatment on glucose uptake, because at the beginning of the talk, you showed that when you compared to the parental cells, or MCF7's LCCs have increased glucose uptake. What happens with the treatments, like full western and tamoxifen? Yeah, so I haven't shown that. So we have done that experiment. So with full western and tamoxifen LCC1 cell lines, which are sensitive to products, tamoxifen and full western, the glucose uptake goes down. But not in LCC9 cells. LCC9 cells, they don't care about what we are treating it with. The glucose uptake doesn't go down. Does it go up, actually? It doesn't go up beyond what we see, no. Hi, Shantanu Ghosh, Illinois. So I had actually two questions, and I'll ask the first one first. In cancer cells, from what I know, is that they prefer glycolysis to TCA. So have you seen what percentage of the glucose is actually going into the TCA cycle? We haven't looked at the percentage, but we know that the LCC9 cells utilize more glucose, and they also secrete more lactate. So we think the glycolysis pathway is activated in these cells, yes. Okay, and my second question was, do we know that if estrogen actually regulates any of these TCA cycle enzymes, like just transcriptionally, like SDH, for example? No, I haven't seen any reports where SDH is directly activated or regulated by estrogen as a primary. Yeah, because I was just thinking that since ICI degrades the estrogen receptor, you would essentially be limiting that nuclear pathway at the estrogen receptor ERE pathway. So I was just wondering if that regulates SDH. No, no, no, we haven't. I don't think any of these. I mean, I know some of the glycolytic pathways, there are some enzymes which are regulated by estrogen, but top of my head, I don't remember any of the enzymes in TCA cycle which are regulated by estrogen. Okay, thank you. Sure. Dan Joely, UVA. So one prediction, given the effect on HIF, is that succinate dehydrogenase downregulation would make these cells more sensitive to hypoxia and less likely to induce angiogenesis. Your thoughts? Yes, I think that is, so what we see in LCC9 cells and over the years, we have shown that many of the genes which are overexpressed in LCC9 cells, they are actually HIF-regulated cells. So that is another thing which we want to look at. What we don't know at this moment, that what we are seeing here is actually what is happening or it is a consequence because I think the more important thing is we need to see where the fumarate is coming from. If the fumarate is unlabeled in these cells, there are certain other pathways where the fumarate can come from. And we are right now focusing on those pathways and we don't have a concrete answer at this moment, but we have some encouraging results where these pathways, it looks like they are involved in endocrine therapy resistance. Thank you. Thank you very much. Thank you. Thank you. So we are ready for our next speaker. I will invite Professor Sylvie Madere from University of Montreal. Thank you. Hello, I would like to stress that this presentation should normally have been performed by Dr. Salwa Eddah. She couldn't be here today, so I'm instead going to present her work on the mesenchymal-to-epithelial transition in breast cancer cells, which she performed as part of her PhD thesis in my lab. Is that good? Okay. Okay. So you all know that breast cancer is a heterogeneous disease characterized by the differential expression of therapeutic targets such as estrogen receptor and ERB2 membrane receptor. And that transcriptional profiling has refined the classification of breast cancer subtypes and also revealed interesting parallels between the different breast cancer subtypes and normal mammary epithelial cells at different stages of their differentiation. Most breast cancer subtypes are epithelial in nature, but a particular triple negative subtype called clodin low because of its low expression in this class of tight junction proteins was characterized as having a mesenchymal phenotype and traits reminiscent of normal mammary stem cells. So mesenchymal-to-epithelial transition, or MET, is important during normal mammary development, but it is reversed during tumorigenesis when cells lose cell-cell adhesion properties and acquire cell migration and invasion properties associated with the major remodeling of cell surface proteins. This process is reversible once cancer cells have colonized peripheral tissues. Now, the transcription control of epithelial-to-mesenchymal transition, or EMT, is fairly well-characterized with transcription factors of the Zeb, Snail, and Twist, as well as Foxy and Teed families of transcription factors playing key roles. However, the reverse transition is far less well-characterized. In Drosophila, the graining head transcription factor is known to play important roles in epithelial morphogenesis. And 10 years ago, the lab of Stephen Frisch showed that the homolog of grainy head, GRHL2, in human breast cancer cells can suppress EMT. In normal, in breast cancer subtypes, GRHL2 is, in fact, expressed in all epithelial subtypes, but is depleted in the clodding low subtype associated with the converse overexpression of mesenchymal markers, such as vimentin. This is the same thing in breast cancer cell lines, where cells that overexpress vimentin are also almost completely depleted in GRHL2 protein. When overexpressing GRHL2 in one such cell lines, MDNB231, two levels comparable to those found in epithelial cell lines, this resulted in the acquisition of a more flattened, spread-out phenotype of cells, and RNA profiling revealed both up and down regulation of gene expression, but genes up-regulated were significantly more epithelial in nature, and enriched in a gene signature that is repressed by the mesenchymal factor ZEB1. Conversely, GRHL2 knockdown in epithelial basal-like MDNB468 cells led to EMT, with acquisition of a more fibroblastic phenotype, loss of expression of epithelial genes, and overexpression of mesenchymal markers. We next performed ChIP-seq on mesenchymal cells overexpressing GRHL2, MDNB231, or in MDNB468 cells, which endogenously expressed GRHL2. This revealed a large overlap in GRHL2 ChIP-seq peaks, and a high enrichment in DNA binding motifs, as expected for the GRHL transcription factors, which were present in more than 50% of ChIP-seq peaks. Strikingly, a large number of epithelial genes contain GRHL2-interacting regions, either in their promoters or in their enhancers, indicating direct regulation. We then focused on one of the genes that were most upregulated by GRHL2 in four different breast cancer cell lines, and this gene is VGLL1, which is a transcriptional core factor. Upregulation of GRHL2 in two mesenchymal cell lines led to a major upregulation of VGLL1, while suppression of GRHL2 led to almost a complete loss of VGLL1 protein. The VGLL1 promoter contains a GRHL2-interacting region, and overexpression of GRHL2 in MDNB231 cells resulted in increased histone H3K27 acetylation and loss of histone H9 trimethylation. Interestingly, VGLL1, contrary to GRHL2, has a pattern of expression that is almost restricted to basal subtype of tumors, and the same thing is observed in cell lines. When suppressing VGLL1 expression in mesenchymal, in epithelial MDNB231, MDA468 cells, sorry, which have a basal-like phenotype, this resulted in gene regulation that overlapped by more than 50% with GRHL2 target genes in the same cell line, and strikingly, both factors regulated gene overwhelmingly in the same direction. So what is known about VGLL1? It is known to be a core factor of the TID transcription factors, and as I mentioned before, the TID factors, which are downstream of the heposignaling pathway, are key regulators of EMT, stemness, cell proliferation, migration, and survival. So we verified that VGLL1 overexpression in MDNB231 cells interfered with TID signaling. Indeed, when overexpressing VGLL1, this led to a majority of genes being repressed and enriched by GSEA analysis in YAP target genes, and ingenuity pathway analysis also predicted suppressed signaling by TID transcription factors and YAP cofactors. In MDNB231 cells, characterized TID, YAP, and TAS target genes were also regulated by VGLL1 for a quarter of them, and this overwhelmingly in the negative direction. These genes were repressed by VGLL1. Significantly, we could detect by ChIP-qPCR VGLL1 presence on TID binding regions in TID and VGLL1 target genes, and this in cell lines that overexpressed VGLL1 or that overexpressed GRHL2, resulting in VGLL1 overexpression. However, VGLL1 overexpression by itself was not sufficient to lead to mesenchymal to epithelial transition. So we next investigated whether VGLL1 could cooperate with GRHL2 during MET. First, we overexpressed both factors in HEG293 cells and checked whether they could interact with each other, and indeed, we observed reciprocal co-immunoprecipitation of VGLL1 and GRHL2. In addition, in MDNB231 cells overexpressing GRHL2, immunoprecipitation with VGLL1 led to co-immunoprecipitation of GRHL2, and this band was specific compared to negative controls for parental cells or cells overexpressing VGLL1 alone. ChIP-qPCR also led to the detection of VGLL1 on GRHL2 target genes, and this was observed only in the cell lines that overexpressed GRHL2 and not in the parental cell lines for those that overexpressed VGLL1 only, indicating that GRHL2 is needed for recruitment of VGLL1 on these GRHL2 target genes. This contrasted with the T target genes where VGLL1 was recruited independently from the expression of GRHL2. Here in the absence of GRHL2, VGLL1 overexpressing cells, we could detect VGLL1 at T binding regions. So these results together suggest that VGLL1 can interact with GRHL2 and may act as a cofactor in regulation of a subset of GRHL2 target genes. So in conclusion, GRHL2 acts as a master regulator of the epithelial phenotype of breast cancer cell line. It directly regulates expression of a large number of epithelial target genes, but it also acts in part via the upregulation of transcriptional cofactor VGLL1 and this specifically in basal breast cancer cells. I haven't had time to show you that GRHL2 also acts via other mechanisms as an indirect regulator of gene transcription by upregulating other epithelial transcription factors such as VGLL1 and VGLL2 and via the upregulation of several epithelial mRNAs, including those that target the Z1 mesenchymal factors. So I would like to congratulate Salwa for the superb work that she did and also thank my collaborators at our research institute and at Imperial College London in Sima Kali's lab who performed the ChIP-QPCR on VGLL1. I would like to acknowledge my funding and thank the audience for their attention. Hi, Matt Sikora, University of Colorado. Really nice talk. So I think a lot of us think of GRHL2 as an estrogen receptor co-regulator and I think I noticed that VGLL1 is not expressed in the luminal cells. Absolutely. So if you overexpress it there, does that mess with ER function? Does it reprogram estrogen receptor positive cells to look like basal cells? Well, actually VGLL1 is regulated by estrogen signaling. It's downregulated by estrogen signaling and that's the reason why it's absent in luminal subtype of breast cancer. However, we still could detect some upregulation of VGLL1 by GRHL2 in luminal breast cancer cells. So GRHL2 is in fact amplified in breast cancer. It's on the same large amplicon as C-Myc. And so the question is whether GRHL2 overexpression through VGLL1 upregulation could modulate the transcriptional regulation of luminal breast cancer cells. Thank you. I was wondering, Jennifer Ricker, University of Colorado. I was wondering if you see an effect of the VGLL1 interacting with grainy headlight to when you put them back into the more aggressive triple negative breast cancer cell lines. Do you see an effect on miR-200 family microRNA? Yes. Well, actually the miR-200 are directly upregulated by GRHL2. I cannot tell you whether VGLL1 participates in this regulation. So you don't know if it's helping to repress ZEB1? I know you looked at a couple of other genes. I mean, one of the thing is that VGLL1 is probably one, I mean, GRHL2 cofactors are not very extensively characterized, but VGLL1 is probably one among several. It affected only a subset of GRHL2 target genes, but some of them were common to all the cell line that we studied and are definitely part of the GRHL2 signature. Okay, thank you. Dan Joely, UVA. Do you have any information on the stoichiometry of the interactions between VGLL1 and GRHL2 and evidence for the endogenous proteins interacting? No, we detected this interaction in MDA231 cells when we overexpressed GRHL2, but no, we did not do the co-immunoprecipitation yet in differentiated epithelial cells so far. We, sorry, we actually, however, detected VGLL1 by chip in MDAMB468 cells on some GRHL2 target genes. Have you tried the chip rechip yet? I know those are not impossible. No, not yet. Thank you. Thank you very much. Thank you. Our next speaker is Shruti Bendre from the University of Illinois Urbana-Champaign. Good morning, everyone, and thank you so much for being here and sticking to the very last leg of the conference. My name is Shruti and I'm a second year PhD student in Dr. Eric Nelson's lab. I would like to thank the organizers of the Endocrine Society for giving me a chance to present my research. This is actually my very first in-person presentation, so I'm really excited to be here today. So we're hearing this probably for the fourth time today, but I'm gonna say it anyway. Breast cancer is the most commonly diagnosed and second leading cause of cancer-related death among women in the United States. Approximately 13% of women in the United States will be diagnosed with breast cancer in their lifetime. And in contrast to the overall decreasing mortality rate for breast cancer, the mortality rate for stage four metastatic breast cancer still remains high. The focus of my study is triple negative breast cancer. As we know, TNBC lacks the expression of estrogen progesterone receptor, as well as overexpression of HER2 receptors, and so definitely lacks targeted therapeutic strategies. Additionally, TNBC is also characterized by a more aggressive histology, a poor prognosis, shorter survival rate, and unresponsiveness to hormone therapy. TNBC is characterized by a higher number of tumor-infiltrating lymphocytes, and therefore, in theory, is supposed to work better with immune therapy. However, most breast cancers have been refractory to immune therapy, and therefore, novel therapeutic targets against triple negative breast cancer, as well as stage four metastatic cancer are definitely required. Cholesterol is an established risk, poor prognostic factor for breast cancer. As you can see here, high plasma cholesterol levels are associated with a higher risk of breast cancer occurrence. Statins, which are drugs given to lower the blood cholesterol levels, you can see that a prolonged use of statins is associated with a lower risk of breast cancer occurrence, too. Our lab has previously shown that a specific metabolite of cholesterol, 27-hydroxycholesterol, or 27HC, has pro-metastatic effects on breast cancer. So when mice were treated with 27HC, and these mice were previously grafted with breast cancer cells, 27HC increased metastasis in them. However, when we treated mice with clotonate liposomes, so these clotonate liposomes are essentially taken up by phagocytic cells and results in their depletion. So when we depleted myeloid cells, you can see that the pro-metastatic effects of 27HC were attenuated, indicating that 27HC impacts breast cancer progression, but in a myeloid cell-dependent manner. Further work has shown that 27HC within myeloid cells, such as macrophages, actually impaired their ability to activate T cells, and those T cells that were activated showed a decreased cytotoxic response. T cells being key mediators of tumor destruction, such a decreased cytotoxic response allowed the breast cancer lesions to grow unchallenged and further increasing metastasis. So initially, my project was focused on finding out the mechanism by which 27HC actually mediates this pro-metastatic effects. 27HC is a known ligand which binds to the liver X receptor. LXR is a nuclear receptor, and it's actually a sensor of cholesterol homeostasis. So when the intracellular levels of cholesterol rise, LXR is known to activate ABCA1, which then mediates the efflux of cholesterol. And in addition to cholesterol efflux, ABCA1 can also translocate cholesterol from the inner leaflet of the plasma membrane to the outer leaflet. Cholesterol in the inner leaflet of the plasma membrane is shown to initiate a bunch of signaling cascades. So we believe that by the translocation activity of ABCA1, this could potentially influence the membrane-initiated signaling of cholesterol too. So a very simple hypothesis initially was that 27HC acts on LXR, which activates ABCA1 and that's how the prometastatic effects of 27HC are initiated. However, when I actually mined the clinical data for ABCA1, to our surprise, we saw that ABCA1 is actually associated with good prognosis. So a higher expression of ABCA1 was associated with a higher survival rate, a higher distant metastasis-free survival, as well as a recurrence-free survival. So now, because we had this very surprising but very significant clinical data, we further set out and laid a bunch of experiments to actually probe ABCA1 and see how it functions in these myeloid cells. So we started by either knocking down ABCA1 in primary macrophages or overexpressing them. And I checked for the gene expression of genes that are associated with antigen presentation and processing. And we saw that when ABCA1 was knocked down, it down-regulated the genes associated with antigen presentation and processing, versus when ABCA1 was overexpressed, all of these genes were up-regulated. And this was particularly evident in ABCA1 mutant. ABCA1 mutant is essentially a mutant form of the protein, which is constitutively active for cholesterol translocation and therefore has decreased cholesterol in the inner leaflet of the plasma membrane. So after this, we wanted to study how these macrophages, which have altered ABCA1, actually affect T cells. So to do this, we first isolated and differentiated macrophages from the bone marrow of wild-type mice. These macrophages were then transfected, in this case, to overexpress ABCA1, after which we isolated pan-T cells from the spleen and lymph node, which were activated using anti-CD3, anti-CD28, and co-cultured with these macrophages, and their expansion was assessed by flow cytometry. And we found that overexpression of ABCA1 in macrophages actually increased T cell expansion, and again, this was particularly evident in ABCA1 mutant. So after this, to confirm our results, we wanted to do like sort of a complementary experiment, where in this case, I knocked down ABCA1, as well as we wanted to see how this particularly affects CD8-positive cytotoxic cells. And so we saw that when ABCA1 was knocked down, so like loss of ABCA1 decreased cytotoxic T cell expansion, and this was particularly dramatic towards the further generations of T cells, which is a characteristic of T cell exhaustion. Furthermore, when we assessed these T cells for their expression of effector enzymes, we saw that loss of ABCA1 resulted in a down-regulation of granzyme B interferon gamma, as well as porphyrin 1, which are all effector enzymes that are secreted by CD8-cytotoxic cells. So in all, we conclude that ABCA1 is in fact positively correlated with better prognosis within patients. Its expression in myeloid cells results in altered T cell expansion and expression of genes. And we feel that the immune suppressive effects of 27HC are not likely mediated via ABCA1, but in fact, it's probably a compensatory mechanism for 27HC, and my project actually is now driven towards developing ABCA1 as a novel therapeutic target to treat metastatic breast cancer. And with that, I would like to thank all my funding agencies, my lab, my mentor, Dr. Eric Nelson, as well as the Endocrine Society. I was a part of the Regim's Fellowship, which gave me the opportunity to fly out and attend this amazing conference. Thank you. Dan Jewel, UVA. So I've already forgotten the name. So what's the effect of ABCA1 expression on macrophage polarization in the M1 to M2-like continuum? So we haven't done like an exclusive area of M1, M2, but the few genes that I actually looked at, it was towards more like an M1 pro-inflammatory. Thank you. Have you actually, again, Jen Ricker from Colorado. Have you actually measured the ability of the macrophages to present an antigen or their function, like in atherosclerosis or not yet? No. And also, does the 27HC affect ABCA1 on the tumor cells as well as the macrophages? So we are specifically focused on macrophages or like myeloid cells as such. We haven't looked at 27HC in tumor cells. Okay. And have you happened to look at what it does in the macrophages to, there's a thing that moves cholesterol from lysosomes called MPC1. I wonder if you've looked at that at all. We haven't, but I'll keep it in mind. How did you, I may have missed it at the beginning. How did you discover its effects on the ABC1 in the macrophages? Effects of ABC1? How did you see that it was the ABC1 that was changing initially? So initially, so we worked with 27HC and ABCA1 is a known target, downstream target gene. Oh, so that part was already known. Yeah. So ABCA1 was in our mind. Oh, okay. The effect was surprising to what we thought it would be. Oh, yeah. It was kind of opposite to what, okay. Great. Thank you. Thank you. Thank you so much. So we have one final presentation in this session and it will be a virtual one from Lindsey Crump at the University of Colorado Anschutz Medical Center, who's not able to be with us in person today, but she has a virtual presentation. Her PI, Dr. Jen Ricker is here. So for Q&A period, I don't believe they were able to phone her in as it were. So if there's questions, we can still have a robust discussion. I'd like to begin by thanking the organizers of the meeting for the invitation to speak today. I regret not being there in person, but I am very excited to tell you about our work looking at some androgen regulated secreted factors and their role in breast cancer and affecting the immune system. I have no disclosures. So breast cancer is the most commonly diagnosed cancer in women worldwide. And it's the second leading cause of cancer related death in women in the United States. And the majority of breast cancers, about 70%, express the estrogen receptor alpha or ER. And we've become pretty effective at targeting ER signaling in breast cancer. But despite this, there are multiple mechanisms of resistance that can arise to tumors on these treatment regimens. And so one of these that has been identified are mutations in ESR1, which is the gene encoding ER alpha. These are two of the most common ones shown here. And these can arise in about 40% of metastatic ER positive breast cancer that's been treated with the endocrine therapy of aromatase inhibition or AI. And so in the lab, we use long term estrogen deprivation or LTED to mimic aromatase inhibitor therapy, which blocks the conversions of androgens into estrogens that then signal through ER to promote breast cancer cell survival. Importantly, nearly all ER positive breast cancers also express the androgen receptor or AR. And we recently published that AR is actually pretty highly expressed in ER positive breast cancer. Breast cancer cells that have these ESR1 mutations, particularly in long term estrogen deprived conditions as shown here. And so this really leads to the main premise of this work, which is that a long term estrogen deprivation that's caused by aromatase inhibitor therapies could shift tumor cell reliance from estrogens and ER to androgens and AR to promote breast cancer cell survival. And that leads to our overall hypothesis, which is that AR regulates secreted pro survival immunomodulatory factors in AI resistant ER positive breast cancer that could potentially serve as biomarkers or therapeutic targets for these patients. So we first identified these actually in a triple negative breast cancer model or a model that blocks the estrogen receptor. But as you can see, this is a PDX model that has very high amounts of the androgen receptor and these tumors are highly sensitive to treatment with the androgen DHT as shown here. So using this model, we compared DHT treated factors that increase with DHT treatment to a known secretome signature to find secreted factors that were regulated by androgens. And we got multiple hits and I'm just going to point two of them out to you here, which are PIP and AZGP1. We next verified at the protein level that PIP was increased with DHT treatment. And we also looked at another known AR regulated secreted factor, which is kinase three like one that was also up at the protein level with DHT treatment. So I'll tell you a little bit about what these proteins actually are. PIP is prolactin inducible protein. It's a known AR regulated gene that makes a protein that is secreted. It has multiple known receptors. And the one that I mentioned earlier is AZGP1 or ZAG. So that's a PIP receptor that was up in our screen as well. And PIP can actually initiate numerous pro survival signaling pathways like map kinase and AKD signaling. It's also been shown to affect tumor promotional macrophages as well. And interestingly, PIP is actually used in the clinic to identify metastatic disease of unknown origin as arising from the breast tissue because this protein is so highly expressed in the breast. The other factor is kinase three like one. And so this is another known AR regulated factor that can be secreted and initiate programs of cell survival and invasion and can interact with macrophages as well. And so it has multiple known receptors. And recently it's been published that kinase can promote PD-L1 expression in melanoma models. And our group actually recently published that those ESR1 mutant cells that I showed you have high amounts of kinase. And it actually promotes invasive programs in those cells. So we looked at ER positive cells next that were treated with DHT or the anti-AR drug enzalutamide. And we found that both PIP and its receptor ZAG were increased with DHT treatment. And that was blocked with ENZA. We also looked in T470 cells that were long term estrogen deprived and saw again that they had very high amounts of AR and a corresponding increase in PIP. I'm sorry, in kinase as well. Using an in vivo model, this was actually recently published too. We saw that this PDX model that has an ESR1 mutation that was grown with or without estrogen supplementation. We observed that there was an increase in kinase with the loss or with the lack of estrogen supplementation showing that estrogen depletion is increasing these proteins. We next looked at long term estrogen deprived cells that had these ESR1 mutations. So these are cell pellets and they were treated with DHT. So you can see here in both the D mutant and the Y mutant treated with DHT increases PIP in both mutants as well as kinase. And that's those images are shown at the bottom here. So this suggests that in this LTED ESR1 mutant cells that these proteins are remain estrogen depleted and androgen sensitive for regulation of PIP and kinase. With that, I'm going to switch gears a little bit and talk about the more focused hypothesis that I've been working on recently, which is that these AR regulated factors promote breast cancer cell survival and can actually impact the treatment of breast cancer. Cell survival and can actually impact the tumor micro environment, namely macrophages. And for the sake of time today, I'm not going to show the cell survival data, but I will share some of the new macrophage data that we have. So I think it's pretty well appreciated now that macrophages can be polarized into tumor promotional M1 like or tumor suppressive M2 like states. But macrophages can also acquire other markers that help them become more pro metastatic. And these include lymphatic markers. And so this is adapted from a nice review showing that different secreted factors have been identified that cause macrophage polarization into what's called a poem like state. And this stands for protoplanin expressing macrophage. So these are macrophages that have acquired lymphatic markers that allow them to incorporate into the lymphatic vasculature and become more permissive to tumor cell escape from the primary site and lymphatic mediated metastasis. It's not been shown, however, if other factors such as kinase and PIP can affect this phenotype. So we took PMA treated THP1 macrophages and treated them with exogenous kinase and PIP and looked at numerous gene expression profiles. There weren't really notable changes in the traditional anti-tumor and pro-tumor macrophage markers with these recombinant proteins. However, we did see that there was a pretty universal increase in lymphatic markers with the addition of kinase and PIP, suggesting they may be shifting to that more poem-like state. We also saw that there was a reduction in an efferocytosis gene. So efferocytosis is a specialized phagocytic process whereby macrophages eat dead cells. And so this suggested that at least with PIP treatment that they might have reduced efferocytotic ability. So to test that, we used this photo efferocytosis assay. To do this, you dye tumor cells and then feed them to macrophages. And as the macrophages eat the cancer cells, they emit a red signature, which you can see here that's quantifiable. So we performed this experiment by treating THP1 cells with these proteins and feeding them cancer cells. And you can see that compared to the vehicle treatment control, the addition of kinase and PIP reduced the efferocytotic ability of these macrophage cells. So with that, I'll just conclude. I've shown you that AR upregulates kinase and PIP in multiple models, particularly in those ESR1 mutant estrogen-deprived cells that have high AR. We also see that kinase and PIP can alter macrophage marker expression and reduce their efferocytotic function. We think that these proteins could be shifting them to a more poem-like state that could impact downstream metastasis. So moving forward, we want to determine if tumor cell-derived PIP and kinase can affect macrophage infiltration function or incorporation into the lymphatics that could mediate metastasis. And then importantly, we also want to block these proteins to see how that affects macrophage and metastasis function in syngeneic mouse models, particularly estrogen-independent models. So with that, I'd like to thank my lab, especially the people in BOLD who contributed to this work, as well as the Cancer Center and our other collaborators. And thank you very much. Folks? All right, actually, if you wanna come up to the mic and then, just because we do have a few people online, if you wanna come to the mic and if, Jen, you wanna sort of sneak around to one of these mics, maybe you can borrow this one. Yeah, go ahead. Okay, so I was actually wondering, the ER mutations that were mentioned, the Y537S and the D538G mutations, from what I know, they actually make the ER constitutively active. So when you have constitutively active estrogen receptor, I was wondering, estrogen deprivation, why would that increase androgen levels? Androgen receptor levels, I'm sorry. I mean, they're not dependent on ER estrogens for growth. So is it a non-canonical estrogen pathway that is causing androgen receptor to be upregulated? Yeah, actually, in the normal mammary gland and the breast cancer cells, we do see when you long-term estrogen deprive them, which is the condition that the ESR1 mutations arise, that the androgen receptor does increase. And so you also, when you're blocking conversion of androgens to estrogens in women on aromatase inhibitors, you do get androgenic side effects. And so you get an increase of androgens when you're blocking the conversion of androgens to estrogens. And AR is stabilized by its ligand. So you actually see in patients' tumors, especially after resistance to aromatase inhibitors, that the androgen receptor is increased. And that's actually how we first started studying the androgen receptor. Okay, thank you. Does that answer the question? Yeah, I'll think about it a little bit more. Thank you. Okay, or we can talk later, too. Anybody else? Were there any online? So one of the purposes was to see if there were secreted factors that we could actually measure in the blood that would be indicators of tumors that were switching over from dependence on estrogens to androgens. Yeah, there's PIP, and it's actually used in the clinic to look for metastases of unknown origin. They stain it with this antibody to PIP to see if it came from the breast. But one of its receptors is ZAG, or AZGP1, and then chitinase is sort of a separate story. But since they're both regulated by AR, do you seek cooperativity if you co-over-express or co-knock down the two on macrophage function, et cetera? Yeah, we're just now starting to manipulate them. And some of the stuff that Lindsey showed was just adding exogenous PIP or chitinase. But now we're starting to knock them down or over-express them in tumor cells and then see if the PIP affects the receptor, the AZGP1, or vice versa, and how that affects AR activity, because they are known to be AR-regulated in prostate cancer. So yeah, it's really early days. But there was stuff in the literature about all of these affecting macrophages and other systems, like in allergies and stuff, sort of recruiting into macrophages. So we wanted to look at their effect when they were made by tumor cells, their effect on macrophages and potential immune suppression. Okay, thanks. Thank you. All right. Yeah, thanks for that round of applause. It's tough to give a talk on the last morning of a meeting when everybody's tired. But thanks to the audience too for good questions and interaction, and my co-chair. Thank you very much again. And maybe a round of applause for all the speakers this morning. Thank you.

breast cancer

aromatase inhibitors

hormone receptor expression

immune responses

cholesterol

ABCA1

prognosis

androgen receptor-regulated factors

macrophage function

therapeutic targets