 Hello everybody. My name is Ethan Abel. I'm from the University of Michigan, and thank you for attending my webinar entitled HNF1A in Human pancreatic cancer stem cells. So begin with I'll talk a little bit about pancreatic ductal adenocarcinoma or pancreatic cancer that I'm going to be discussing. This is not the most common cancer. It's actually the 13th or 11th most common cancer in the United States, which is pretty much where the good news ends for this cancer because it's actually fast on its course to hit the number two cause of cancer deaths in the United States by the year 2020. So coming up very soon. There's a lot of reasons for this. So we believe that a large factor is of course late detection for this disease. This is something that is quite famous. But this is compounded by early and aggressive metastasis as well as a lack of effective therapeutics for treating the disease once we actually do detect it. And a common factor behind both the aggressive metastasis of the disease as well as a lack of effective therapeutics is a cellular compartment that we refer to as pancreatic cancer stem cells. So pancreatic cancer stem cells, these are a subtype of tumor cells. There are multiple types of actual tumor cells within the pancreas as well or pancreatic cancer as well as non-cancer cells. But this subtype of cancer cell is highly plastic. We believe this to be an epigenetic process or transcriptional process that allows them to basically be solely capable of forming new tumors as well as maintaining pre-existing tumors. And through this mechanism, we believe this is the contributes to their ability to promote metastasis as well as resist chemo-therapeutics and allow the regrowth of tumors after treatment. As such pancreatic cancer stem cells are a very desirable therapeutic target in the disease. But as of right now, we don't really understand fully what makes them different from the non-cancer stem cell compartment and what these vulnerabilities might actually be. So a major focus of my research is basically discovering new drivers that actually promote the cancer stem cell state with the hope that these might be something that can be exploited and targeted to get rid of the subtype of cells. So just a little bit of characterization of cancer stem cells that I'll use later in my talk is that these cells are capable of forming tumors. So in this case using data from the simioni lab, which was part of when I did this work, we found that cancer stem cells are marked by the expression of three surface markers CD44, CD24 and ESA or EPCAM. And it's also expressing these markers are able to form tumors where cells lacking these markers are not. Additionally, in vitro, we find that these cells can actually form our cultumor spheres, one grown under non-adherent conditions in silver freak media. And both of these techniques will be used heavily throughout my talk. So keep in mind, these are sort of in vitro and in vivo metrics for the activity of cancer stem cells. So to identify drivers of the pancreatic cancer stem cell state, I started off with patient derived tumors and separated these tumors into cellular compartments of cancer stem cells. So those cells expressing EPCAM, CD44 and CD24, as well as cells lacking those which we refer to as bulk tumor cells. And then performing microarray analysis of these different compartments to look for genes that were upregulated in pancreatic cancer stem cells versus the bulk cells with the idea that these might actually contain drivers for the state. And then through functional validation, figure out which of these genes actually are contributing to the cancer stem cell state. What we found interestingly enough was that using two different primary isolates pancreatic cancer was that these pancreatic cancer stem cells had 50 genes that were conserved between patient isolates that were upregulated in the cancer stem cells versus the non-cancer stem cells. When we look closer at these genes, we found that a large subset of these genes were actually predicted targets of a transcription factor known as HNF1A or HNF1-alpha. And that HNF1A itself was actually one of these genes upregulated in the pancreatic cancer stem cells versus the bulk tumor cells. Which suggested to me that this might actually be essentially the head of the snake driving the pancreatic cancer stem cell state. To look into what actually HNF1A is, this is a transcription factor that is found throughout the gastrointestinal tract in the liver. Its name is hepatocyte nuclear factor 1-alpha. It's also expressed in the pancreas and is important for pancreatic homeostasis. Additionally, there are a number of single nucleotide polymorphisms in HNF1A that are known risk factors for the development of pancreatic cancer. Although this is through unknown mechanisms at this point. Additionally, in the literature, it's been suggested that HNF1A may be a tumor suppressor in some contexts, in the case of the liver, possibly even the pancreas. But also possibly even an oncogene in the context of prostate cancer. So a lot was not really known at the time when I started studying the function of HNF1A. So to learn more about it, I looked at the comparisons of pancreatic cancer tissue versus normal pancreatic tissue. I found that HNF1A levels were elevated in the neoplastic ducts of pancreatic ductal adenocarcinoma versus normal ducts of the pancreas. Additionally, looking across a panel of our primary pancreatic cancer cell lines, compared to immortalized ductal cells, HPNE and HPDE, we found that HNF1A levels tended to be higher in pancreatic cancer versus these normal cell equivalents. To determine what HNF1A actually was doing in pancreatic cancer, I went ahead and knocked down HNF1A with different SI or RNA targeting the gene and found immediately that HNF1A knockdown resulted in a profound effect on proliferation of these cells with a marked reduction when HNF1A was lost. Additionally, loss of HNF1A resulted in an induction of apoptosis in multiple cell lines, suggesting that HNF1A was important for maintaining both the growth and survival of pancreatic cancer cells. To determine whether or not pancreatic cancer stem cells were actually being affected by a loss of HNF1A, I looked at the expression of these different surface markers I previously mentioned, basically demarcate pancreatic cancer stem cells and found that there was a pronounced down-regulation of these markers, in particular CD24, which was down-regulated in multiple cell lines when HNF1A was depleted with different SI or RNA. Additionally, looking at the ability for these cells to form tumor spheres, I found that loss of HNF1A markedly reduced both the size and the number of pancreatic cancer spheres, indicating a loss of in vitro stem cell function. To test what is actually going on in vivo, I went ahead and knocked down these cells with SHRNAs for a stable knockdown of HNF1A, and found that knockdown of HNF1A in vivo markedly reduced the ability for these cells to form tumors, and when the tumors were formed, they were substantially smaller than the controlled knockdown cells. This was also observed when we actually implanted the cells into the pancreas of mice, so an orthotopic injection, showing that the microenvironment was not a factor in this loss of tumor growth, and that loss of HNF1A was in fact reducing the ability for these cells to form tumors and to continue to grow. Additionally, if we took tumors out of mice that either had been knocked down for HNF1A or a controlled hairpin, we found that the number of cancer stem cells from tumors of HNF1A knocked down mice were actually markedly reduced, suggesting that the loss of tumor geneticity that we were seeing was a result of a loss of the cancer stem cell compartment. To test whether or not we could actually push the cancer stem cell state in the opposite direction, we went ahead and overexpressed HNF1A, and in this case in an inducible manner, in pancreatic cancer cell lines, and found that consistent with our knockdown data that overexpression of HNF1A promoted a marked increase in the expression of CD24, as well as other cancer stem cell markers, and additionally promoted the formation of tumor spheres. We ended up with larger, as well as more numerous tumor spheres quantitative here. If we go ahead and look at non-cancer cells, in this case HBNE, which is an immortalized pancreatic cell line, lacking both the K-resmutation, which is present in over 90% of pancreatic cancer, as well as lacking HNF1A, we found that if we overexpressed either K-resm or HNF1A alone or in combination, that both of these genes functioned as oncogenes, promoting the ability for cells to form colonies at low density. Additionally, using a separate immortalized cell line, we found that these genes were able to cooperate in the formation of colonies in soft agar, indicating an ability to bypass or actually promote anchorage independent growth, suggesting that HNF1A does in fact function as a novel oncogene in the context of pancreatic cells, and in particular when combined with K-resm, the universal mutation in pancreatic cancer. To determine how HNF1A is actually promoting stemness in these cells, I went about looking at known regulators of stem cell state, including many of the Yamanaka factors known to impart stemness to normal cells such as fibroblastic or tinnocytes, and found that of these genes, Oct4 was down-regulated when HNF1A was knocked down, shown here at the RNA level, as well as at the protein level when we knocked down HNF1A. And importantly, it's worth noting that the form of Oct4 expressed in these cells is Oct4A, which is the subtype of Oct4 capable of imparting stemness to somatic cells. So, indicating that this is actually important for the stemness of these cells. Additionally, when we overexpress HNF1A in these cells, both in cancer cells as well as normal pancreatic cells, we found that overexpression of HNF1A caused an up-regulation of the RNA levels of Oct4. And when we actually look at the promoter regions of Oct4, we found that interestingly, the non-canonical or the canonical promoter of Oct4 was not actually induced by the expression of HNF1A, but instead a non-canonical recently described upstream promoter region, which is essentially a retro-transpose on LTR region that was highly responsive to the present of HNF1A. So, just in that HNF1A is directly regulating Oct4 through a non-canonical promoter, 14 kilobases upstream of the canonical promoter. To determine whether or not Oct4 was actually directly contributing to the stemness of these cells, we knocked down HNF1A alongside Oct4 and found that similar to HNF1A knockdown, loss of Oct4 resulted in a loss of tumor sphere formation. And importantly, when we re-expressed Oct4 in HNF1A knockdown cells, we found that the re-expression of Oct4 was able to rescue sphere formation, indicating that Oct4 is epigenetically located downstream of HNF1A and upstream of stemness in pancreatic cancer stem cells. Now, beyond Oct4, we wanted to know what HNF1A might be regulating in pancreatic cancer cells. So, we went ahead performing RNA-seq as well as chip-seq and identified a number of targets that were both up-regulated and down-regulated by HNF1A expression across multiple pancreatic cancer cell lines. And of these genes, we found an interesting trend where genes that were up-regulated by HNF1A actually were associated with poor survival in patients with pancreatic cancer analyzed by the TCGA, suggesting that beyond Oct4 in cancer stem cells, HNF1A might be regulating a larger malignancy profile or program in pancreatic cancer cells. So, to summarize these findings that we published this summer in Elife, we did find that HNF1A was not only up-regulated in pancreatic cancer stem cells, but actually was important for the function of these cells, both in vitro and in vivo. And this is probably through the positive regulation and direct regulation of Oct4, which links HNF1A directly to stemness. Beyond this, we know that HNF1A now actually is associated with genes that it up-regulates, and that these genes are basically harbingers of bad news for people with pancreatic cancer. They're associated with poor patient survival, suggesting that HNF1A may regulate a larger malignancy profile in pancreatic cancer. So, with that, I'd like to talk a little bit about my future directions. I'm interested in understanding other facets of how HNF1A controls pancreatic cancer stem cell biology, including things like drug resistance and metastasis. These are other hallmarks of pancreatic cancer, as well as understanding ways that we can actually abrogate HNF1A function and expression in cells. And beyond that, look at some of those genes that were associated with poor survival in patients with pancreatic cancer. And find out whether those genes are biomarkers or actually novel targets that could be useful in treating the disease. Beyond HNF1A, I'm interested in other transcription factor networks and understanding how these different networks contribute to pancreatic cancer and the plasticity of this disease. So, with that, I'd like to acknowledge the labs that I've worked with, the Crawford Lab and Simione Lab, that basically funded my research, as well as the many members of these labs, including people with asterisks next to them, who contributed to this paper and are authors on it, as well as my funding from the American Cancer Society, the Pancreatic Cancer Action Network, and AECR. I basically appreciate that and the opportunity to speak to you all the day, so I'd like to thank the audience. And if you have any questions or comments, please do contact me. My email is listed here, and I look forward to hearing from you. With that, have a great World Cancer Day and enjoy the rest of these seminars.