Nobel Lecture Gregg L Semenza Nobel Prize in Physiology or Medicine 2019

Nobel laureates, excellencies, ladies and gentlemen, it's a great pleasure and indeed a privilege to welcome all of you to Karolinska Institutet's Aula America. We are here for the 2019 Nobel Lectures in Physiology or Medicine. My name is Olli Otterson and I'm the president of Karolinska Institutet. During the Nobel week and particularly during the Nobel lectures, we celebrate science and rejoice at unique contributions of some exceptional individuals. Research is a way of quenching our thirst for knowledge, testing our own limits and having an outlet for our creativity and competitive spirit. I'm convinced we will be witness to all these facets of research as we attend the lectures here today. Within physiology or medicine, the ultimate goal is to achieve scientific breakthroughs that change a review of human health, disease and normal vital processes, the fundamentals of life. Dear laureates, through your seminal discoveries you have indeed made historic contributions to these areas. Today it's an honor for me to extend my warmest welcome on behalf of Karolinska Institutet to William Kellin Jr., Sir Peter Radcliffe and Greg Semenza, who will be receiving the 2019 Nobel Prize in Physiology or Medicine for the discoveries of how cells sense and adapt to oxygen availability. Without further ado, I will now call upon Patrik Arnfors, Professor of Molecular Neurodevelopment and member of the Nobel Committee for Physiology or Medicine. Professor Arnfors will introduce the laureates. Patrik, the floor is yours. Ladies and gentlemen, we are at the most exciting point of the Nobel week. We will hear directly from the laureates about the research that has been awarded this year's Nobel Prize. On behalf of the Nobel Assembly at Karolinska Institutet, it's a great pleasure for me to introduce to you the 2019 Nobel laureates in physiology or medicine, Professor William Kellin, Professor Sir Peter Radcliffe and Professor Greg Semenza. They have received this year's prize in physiology or medicine for the discoveries on how cells sense and adapt to oxygen availability. The need for oxygen to sustain life has been understood since the onset of modern biology. Oxygen is used in oxidative reactions to convert food into energy useful for cells. But animals can be exposed to low oxygen levels. They can also vary between tissues and cells. Sensing and adapting to these changes is critical for survival. Because of this, during evolution, mechanisms have evolved that allow cells to adapt to changes in oxygen levels. But the mechanism by which organisms sense and adapt to oxygen has remained unknown. The story on this discovery has many surprises and starts with two apparently different lines of research. Studies on the regulation of the erythropoietin gene and studies on the function of the hippolytic tumor suppressor gene in cancer. Working in these areas of research, Dr. Kellin, Dr. Radcliffe and Dr. Semenza discovered a master regulator for cellular and systemic oxygen homeostasis. They showed that this machinery exists in cells and tissues throughout our body, constantly monitoring oxygen levels and, when necessary, setting into motion processes which make cells and tissues adapt. The molecular details of oxygen response regulation did not stop unfolding until the discoveries that were awarded with the 2019 Nobel Prize were made. On the contrary, the key discoveries of this year's law rates opened the field and led to an explosion of work that has uncovered further molecular complexity in response to oxygen flux. The oxygen-activated signaling pathway controls at least 300 genes, affecting a wide variety of functional properties ranging from rewiring metabolism, immune and inflammatory responses, respiration, the generation of new blood vessels, altitude adaptation and production of red blood cells. Since the seminal discoveries by this year's law rates, oxygen sensing has also developed into a research field with clinical importance. As expected, based on the fundamental importance of oxygen sensing in normal physiology, it is also central to major diseases, for example, anemia, ischemic vascular diseases and cancer. Intense ongoing efforts now focus on developing drugs that target the oxygen sensing machinery. In addition to this year's Nobel Prize in physiology or medicine, the law rates have been awarded numerous prizes along their distinguished careers, including the Canada Gardner International Award, Scientific Grand Prix, Wiley Prize in Biomedical Sciences, Albert Lasker Basic Medical Research Award and Mastery Prize. It is time to hear the law rates. We'll hear three individual lectures and this is a unique opportunity to gain first-hand insights into the journeys that have led to their seminal discoveries which explain how cells sense and adapt to oxygen availability. Our first speaker is Greg Semenza. He is a professor of pediatrics and genetic medicine and serves as the director for the vascular program at the Institute for Cell Engineering at the Johns Hopkins University School of Medicine in Baltimore. He is a medical doctor and a geneticist and a member of the National Academy of Sciences. Dr. Semenza received his bachelor's degree from Harvard University in 1978 with the intention of pursuing a career in basic research in the area of genetics. For various reasons he changed his mind and instead applied for an MD-PhD program and joined the University of Pennsylvania where he earned his medical doctorate PhD degree in 1984. His PhD work involved the sequencing of the beta globin gene of a family with beta thalassemia. Dr. Semenza completed his medical training in 1986 with an internship and residency in pediatrics at Duke University in North Carolina. He thereafter moved to Johns Hopkins University in Baltimore where he has been ever since. He started there by attending medical genetics patients and working in the pediatric emergency room and he also obtained a postdoctoral fellowship to study beta thalassemia. His idea was to study human genetic disorders by making transgenic mice carrying the human mutations but by chance he obtained a clone for the erythropoietin gene and this would entirely change the course of his career. Together with John Gerhardt and Stylianos Antonarakis at the Johns Hopkins he made a mouse that expressed human erythropoietin. With the help of his mentors in 1989 when still a postdoc and without a faculty position Dr. Semenza obtained two grants one from the National Institute of Health to study the regulation of the erythropoietin gene. With funding secured he would through tedious biochemical strategies identify the proteins that bind to the regulatory region of the erythropoietin gene and he eventually managed to purify and clone these factors and name them hypoxia inducible factors because of the presence of protein under low oxygen but absence under normoxia. This discovery opened the door to research on oxygen sensing and Dr. Semenza has since continued his research which range from basic science of the molecular mechanism of oxygen sensing to its medical implications. I would like to ask you all to join me in welcoming Professor Greg Semenza to the stage where he will deliver his 2019 Nobel Prize lecture. Thank you for that kind introduction and thank you for this tremendous honor to speak to you all about our research over the last three decades really trying to understand how the human body can adapt to changes in oxygen availability and this is really a critical physiological property of every cell of the human body because we know that cells must receive oxygen on a continuous basis so every one of the hundred trillion cells in your body requires oxygen in order to make energy and so this oxygen demand must be met by the oxygen supply and the delivery of blood which carries oxygen through the blood vessels across the body to ensure that every cell receives adequate oxygen at every moment of the day and the hypoxia inducible factors serve this role to ensure this balance between oxygen supply and demand. So I'd like to tell three stories that illustrate how the hypoxia inducible factors are important in both physiology and medicine so there wasn't any need for the or this year and I'll start where we started. As you heard from Professor Earnforce, we initially tried to understand the control of red blood cell production and at the time that we started this work we knew that red blood cells of course are produced in the bone marrow but the signal that controls the production of red blood cells was made in the kidney and it was known that individuals who had chronic kidney disease stopped making EPO and became anemic and so it was a great breakthrough when the human erythropoietin gene was cloned because it allowed the production of EPO in the laboratory and that became useful as a therapeutic for the treatment of anemia in the context of chronic kidney disease. But what wasn't known was what controlled erythropoietin production and in particular what allowed cells in the kidney to produce EPO in response to changes in oxygen availability and what we were able to do was to first identify a short sequence downstream from the human EPO gene that became known as the hypoxia response element and this is a very short sequence consisting of 33 base pairs of DNA that has the property that it can be taken out of the human EPO gene inserted into a foreign gene and the expression of that gene will be induced by hypoxia by low oxygen and then we use this sequence to identify a protein that bound to it under hypoxic conditions and we call this hypoxia inducible factor one or HIF-1 and we purified the protein based on its binding to the sequence and found that it was a heterodimer composed of two subunits that we called HIF-1 alpha and HIF-1 beta and under hypoxic conditions these two proteins dimerize, they bind to the hypoxia response element and recruit proteins called coactivators that serve as a bridge to the RNA polymerase that's responsible for transcribing the DNA into RNA. And the important question that arose was of course what controlled the levels of HIF-1 protein under hypoxic and normoxic conditions and as you'll hear from Peter Radcliffe and Bill Kalin shortly, this involved a novel post-translational modification in which two proline residues in HIF-1 alpha were hydroxylated under normoxic conditions. So when oxygen was available, an oxygen atom was inserted into HIF-1 alpha subunit and this hydroxylation allowed the protein to be bound by the von Hippel-Lindau or VHL protein and when VHL binds to HIF-1 alpha, it targets the protein for destruction. So when oxygen is available, the protein is hydroxylated and destroyed. However, under hypoxic conditions, when oxygen levels are decreased, the hydroxylation reaction is inhibited. VHL cannot bind and the HIF-1 alpha protein rapidly accumulates within hypoxic cells. In addition to this hydroxylation of proline residues, there's an asparagine residue in the transactivation domain of HIF-1 alpha that's also hydroxylated by another protein that we identified and called factor-inhibiting HIF-1 or FIH-1, because this protein inhibited the ability of HIF-1 alpha to recruit the coactivator proteins. And it was subsequently shown that FIH-1, like the prolyl hydroxylases, was capable of inserting an oxygen atom into a amino acid residue of HIF-1 alpha, in this case an asparagine residue. And in this case, the hydroxylation prevents the interaction with the coactivators. Again, under hypoxic conditions, the hydroxylation does not occur, the coactivators can bind, and HIF-1 is competent to activate gene transcription. So through these oxygen-mediated modifications, the activity of HIF-1 is regulated, and this provides a direct way to transduce changes in oxygen availability to the nucleus as changes in the activity of HIF-1. Early on, after we had purified the protein and isolated the cDNA sequences, we generated antibodies against the HIF-1 alpha and HIF-1 beta proteins, and we used these antibodies to determine the levels of the proteins in cells as a function of the oxygen concentration that they were exposed to. And what we found is that oxygen concentrations below six percent, there was a dramatic increase in the HIF-1 alpha levels. And this increase occurred in a graded fashion, so the more severe the hypoxia, the greater the levels of HIF-1 alpha protein. And six percent oxygen corresponds to venous PO2, and most cells in the body are exposed to between three and six percent oxygen. So any decrease in oxygen availability in the body occurs along the steep portion of this curve, providing a graded transcriptional response to hypoxia. And we started out by identifying the binding of HIF-1 to a hypoxia response element in the EPO gene, but now we know of over 4,000 genes that are regulated in a similar manner by the binding of HIFs to sequences within the gene that result in activation of gene expression when oxygen levels are reduced. And of course hypoxia can result either from an increase in the amount of oxygen that's consumed, so for example if there are more cells consuming more oxygen, as in a tumor, or a decrease in oxygen delivery, so for example in the context of coronary artery disease, where the blood flow to part of the heart muscle is impaired, this will result in the induction of HIF-1 and the expression of downstream target genes. And we know that for example in the case of the EPO gene, this is a response to systemic hypoxia. So if we go to high altitude, every cell in the body will be exposed to hypoxia, and the response is systemic, an increase in red blood cell production to increase delivery of oxygen to all the cells. However in the case of coronary artery disease, there's only a small area of the heart muscle that's deprived of oxygen, and so the response is a local response, for example the production of vascular endothelial growth factor and other proteins that stimulate blood vessel growth in order to increase oxygen supply to the hypoxic tissue. And then regardless of whether the hypoxia is systemic or local, individual cells have to adapt their metabolism to changes in oxygen, and the really classic metabolic adaptation to hypoxia is switching from respiration, which uses oxygen, to glycolysis, which allows energy production in the absence of oxygen. And this of course helps cells to survive oxygen deprivation. Early on we were able to knock out the expression of the HIF1-alpha gene in mice, and we showed that mice that were homozygous for the knockout allele, so these animals did not make any HIF1-alpha protein, the development of the mouse embryos arrested at mid-gestation, and there were defects in the development of the heart, the blood vessels, and the production of red blood cells. So all three components of the circulatory system are dependent on HIF1 for their normal development. Now of course, because these embryos did not survive to be born, we could not study the effects of complete HIF1 deficiency in adult animals, but what we found was that animals that were heterozygous, that contained one of the mutant genes and one of the wild-type genes, developed normally, but as you'll see, had impaired responses to hypoxia and ischemia. Well it turned out that in addition to HIF1-alpha, there were several siblings hiding in the human genome. One of these was named HIF2-alpha, and the other HIF3-alpha, and these three proteins are products of different genes, but they have related protein sequences. The proteins are all stabilized under hypoxic conditions through the mechanisms that I described, and all three can heterodimerize with HIF1-beta, and whereas HIF1-alpha is present in all nucleated cells of all animal species, HIF2-alpha and HIF3-alpha occurred only during vertebrate evolution, and these are only expressed in certain cells during the development of vertebrates such as humans. So, we had done these experiments in animals, but what was the proof that this system was important for red blood production in humans? And of course, the answer is genetics, and there were patients identified who made too many red blood cells, and these patients were prone to having strokes and other thrombotic disorders, and we and others attempted to identify the mutations that were responsible for this phenotype called congenital polycythemia, and the first mutations were found in the receptor for EPO that's present on red blood cell progenitors, and this mutation makes the receptor hyperactive so that it stimulates red blood cell production even in the absence of the EPO protein. The second group of mutations that were identified were in the VHL gene, and in the VHL gene, a change of a single amino acid in the protein altered the ability of the VHL protein to interact with the HIF-1-alpha and HIF-2-alpha after they had been hydroxylated, and so as a result, less of the HIF proteins are degraded at any oxygen concentration, so there's more HIF activity, and this results in increased red blood cell production. Other individuals were found not to have mutations in either of those genes and turned out to have mutations in one of the pro-hydroxylase genes, and these mutations, again, a single amino acid change in the protein decreased the ability of the enzyme to insert the oxygen atom into HIF-1-alpha and HIF-2-alpha, and then finally families were identified where there were mutations in the HIF-2-alpha gene itself, again, changing only one amino acid in the protein, and these amino acids made it more difficult for the hydroxylases to bind and modify the protein. So you can see that this pathway is virtually saturated with mutations demonstrating the critical role for this physiological process in human beings, and it's somewhat ironic, of course, that where we started with the EPO gene and HIF-1-alpha, these are the only components in the pathway that have not been shown to have mutations in this disease. So these are the kind of ironic findings that bring us great joy, actually, because they're unpredictable. So now I'd like to turn to disease conditions and first illustrate the role of hypoxia-inducible factors as protection against human disease, and I'll start with cardiovascular disease, which, of course, is the leading cause of death in the industrialized world. And, of course, this comes in two major types. If there's narrowing of a major coronary artery, this, of course, can result in angina or chest pain, and that can eventually lead to myocardial infarction, a heart attack. And then if the arterial obstruction occurs in a large blood vessel in the leg, this results in leg pain and can eventually result in a condition like this where you can see that there's an obstruction to blood flow here, and the only blood—this is about the level of the knee here—so here's the blood vessel in the thigh, and then you can see the knee, there's very little blood flowing to the lower leg, and the only blood flow is coming from these blood vessels that branch off above the obstruction. And this is a condition that's known as critical limb ischemia. It's critical because the amount of blood flow that's reaching the lower leg is not sufficient to maintain the viability of the tissue. And so these patients will have ischemic pain at rest. They'll develop ulcers that will not heal. Eventually gangrene will occur, and in many patients who are not able to be rescued by some sort of revascularization, this will require limb amputation. It's quite hard to believe this, but there are thousands of patients who have their limb amputated every day because of adequate therapy for this condition. So we were interested in modeling this disease in mice, and we're particularly interested in the fact that peripheral arterial disease is really an age-related disease. So the older patients are much more likely to develop critical limb ischemia. And so we looked at the effect of aging in a mouse model where we were able to ligate or cut off the major artery that supplies blood to one of the legs. And we used a technique called laser Doppler perfusion imaging in order to look at the blood flow non-invasively. And you can see that before surgery, the ratio of blood flow in the two limbs is equal. So the ratio is one. And then after the surgery, there's a great decrease in blood flow on this side where the operation has been performed. And so the ratio drops very low. And then over time, in young animals, you can see that there's this dramatic recovery of blood flow. However, if we look at older mice, the recovery of blood flow is not quite as good. And then in very old mice, because mice only live to the age of two or two and a half, there's very little recovery of blood flow at all. Then at every age, we compared animals that had a mutation that knocked out the expression of HIF1-alpha in one of the two alleles. And these are called heterozygous HIF1-alpha null mice, or HET mice for short. And you can see at every age, the HET mice recover less well than their wild type litter mates. So just by changing two parameters, the age of the mice and the genotype at one locus, we can cause this range of outcomes from very good recovery of blood flow to very little recovery at all. And this has important clinical consequences so we can examine the mice and look for tissue injury. And in the young wild type animals, there's 100% limb salvage, meaning there's no permanent tissue injury. But if we look at the eight-month-old mice, 40% of these animals had either death of tissue of the skin and soft tissue or the loss of several toes below the ankle amputation. And then if we look at the 20-month-old animals, you can see that about 15% of the animals had above the ankle amputation, meaning they lost the whole foot because of the inability to recover sufficient blood flow. And you can see at every age, the heterozygous mice have more severe phenotype compared to the wild type litter mates. Well, we looked at the levels of HIF1-alpha protein in the ischemic compared to the non-ischemic limb three days after femoral artery ligation. And you can see in the young wild type animals, there appears a very strong production of HIF1-alpha protein. In the eight-month-old animals, you can see the production is not quite as brisk. And in 20-month-old animals, there's very little induction of HIF1-alpha at all. And you can see that the heterozygous knockout mice are also impaired in their production of HIF1-alpha protein. So we see a correlation between the effects of aging on the induction of HIF1-alpha protein and the recovery of blood flow, suggesting that one of the major effects of aging is to impair this normal physiological response to hypoxia and ischemia. If that were true, then we hoped that we could provide replacement therapy through gene therapy. And so we engineered an adenovirus that encoded a modified form of HIF1-alpha with a deletion and several point mutations that rendered the protein stable. So it would no longer be degraded under normoxic conditions. And so at the time that we performed the femoral artery ligation, we injected the adenovirus into the leg along the course of the artery that we had removed. And this was either what we call a constitutively active form of HIF1-alpha or an adenovirus that encoded a bacterial enzyme. And you can see that the animals that were treated with the control virus recovered about 55% of their blood flow in three weeks. And the animals that were treated with the HIF1-alpha adenovirus had a much better recovery of blood flow. Now this was done in the young animals. And as you recall, the young animals can recover well on their own. So the question is, what happens when older animals are treated? And here you can see that the control animals, again, have a much less brisk recovery of blood flow. So compared to the recovery here, it's dramatically impaired in the control mice. And when we give the HIF1-alpha adenoviral treatment, you can see that the level of recovery in the treated older animals is now similar to the recovery of the control younger animals. So we've been able to overcome this age-related impairment of vascularization through HIF1-alpha gene therapy. And so we hope that it may be possible to engineer therapies in patients where we can provide the replacement therapy that will allow these patients to avoid limb amputation. And, you know, the reason that we think that this is particularly powerful is that HIF1 does not just control the expression of vascular endothelial growth factor, but it is a master regulator that controls the expression of a whole series of proteins that are made by the ischemic cells, are secreted into the bloodstream, and then bind to receptors on endothelial cells in blood vessels, endothelial progenitor cells, mesenchymal stem cells, and other bone marrow-derived angiogenic cells that together are orchestrating an increase in tissue vascularization to improve blood flow and oxygenation. And again, what we found was that under the conditions of aging, that the expression of all of these enzymes, all these proteins was impaired. And then when we performed the HIF1-alpha gene therapy, we saw an increase in the expression of all of these genes. Well, as I've described, the role of HIF1 is a physiological regulator of red blood cell production and vascularization. And now I'd like to tell you a little bit about the dark side of HIF1, which is cancer. This is a typical picture of an advanced human cancer. These white areas are areas of necrosis, where the cancer cells have died. And they've died because they're too far away from a blood vessel, because all the cells between them and the blood vessel have consumed all of the oxygen. And so these cells have died, but you'll notice that there are cells immediately around the areas of necrosis that stain very intensely with this brown stain. And that's because we've applied an antibody that binds to HIF1-alpha protein. So these cells are making incredibly high levels of HIF1-alpha protein in the nucleus. And we believe that these cells play a particularly important role in the invasion and metastasis of cancer and the ability of cancers to avoid killing by our therapies and also by the immune system. And whereas most of the cancer chemotherapies target rapidly dividing cancer cells that are present around blood vessels, there are no approved therapies that target these hypoxic cancer cells that express very high levels of HIF1 and HIF2. And so this immunohistochemistry using antibodies against HIF1-alpha on primary patient biopsy samples have shown in a large variety of human cancers that high HIF1-alpha levels are associated with increased patient mortality. So looking at the biopsy at the beginning of the patient's treatment, one can predict patients who are at high risk of failing therapy and dying of the disease by their high expression of HIF1-alpha. And I've highlighted five out of actually 10 studies or so in breast cancer associating high HIF1-alpha with increased patient mortality. So we were interested in identifying drugs that inhibit HIF1 that might be used first of all in our cancer models in animals as a way of evaluating whether HIF1 inhibition might be useful for cancer therapy. And so we identified several drugs that inhibit HIF1 by different mechanisms. And the first one I'd like to describe is acroflavin. And this was one of Ehrlich's original antimicrobials. And it turns out that acroflavin can bind to HIF1-alpha and HIF2-alpha and prevent their dimerization with HIF1-beta. And without dimerization, the protein is inactive. And so we showed that acroflavin would inhibit the production of different HIF target genes. And when we treated tumor-bearing mice, what we saw was, so in this case, we've taken prostate cancer cells, injected them into an immunodeficient mouse, and waited for the tumor to grow, and then started treatment either with saline or with acroflavin at different concentrations. And you can see that the animals treated with saline injections, the tumor grows very rapidly, whereas the animals treated with the HIF inhibitor, there's very little tumor growth after we start the treatment. After two weeks of treatment, we removed the tumors and analyzed the expression of different genes that are regulated by the HIFs. And you can see that compared to the controls, there's a marked decrease in the expression of every HIF target gene that we looked at. And then if we take sections of the tumor and investigate the presence of blood vessels, this brown staining indicates blood vessels, and it's very easy for you to see them in the tumors from the control-treated animals, whereas in the animals treated with the HIF inhibitor, it's very hard to see the blood vessels because they're much smaller. And therefore, of course, there's much less blood flow, and as a result, the growth of the tumor is impaired. We've been interested in particular in breast cancer, and this is a model of so-called triple negative breast cancer. These are breast cancers that do not respond to treatments targeting the estrogen receptor, the progesterone receptor, or HER2. And as a result, they're treated with cytotoxic chemotherapy, and the women often have a relapse after treatment and account for much of the mortality of breast cancer. And what we found was that the expression of the HIFs was particularly high in these tumors, so we injected one of these tumors into mice, into the mammary gland of the mouse, and these tumors will grow rapidly in the mice if we treat only with a saline control. If we treat with a HIF inhibitor, in this case, a drug called digoxin that blocks the production of HIF1-alpha protein, we see that the primary tumor growth is impaired. Compared to the control, the metastasis of cells from the primary tumor to lymph nodes is greatly decreased. And then finally, the metastasis through the blood to the lungs, these purple areas, are all areas of breast cancer cells that have metastasized to the lung in the saline-treated animals. And you can see in the animals treated with the HIF inhibitor digoxin, there's no metastatic cells in the lungs. So we also looked to see what would be the effect of combining a HIF inhibitor with cytotoxic chemotherapy, in this case, a drug called gemcitabine. And you can see that when gemcitabine is administered to the animals, the tumor growth is greatly impaired, but as soon as we stop treatment, tumor growth resumes at the same rate as in the control animals. Whereas if we treat with gemcitabine and the HIF inhibitor, we can eradicate the tumor in this animal model. We've gone on to look at different steps in the metastatic process so we can break down this into individual steps and then identify specific genes that are regulated by the HIFs at each of these steps and show that the expression of all of these genes is inhibited when we administer one of these drugs. We were particularly interested in looking at the regulation of the breast cancer stem cells. So most of the cells in the tumor divide very rapidly but can only divide a certain number of times and then stop dividing. But the cancer stem cells have the property of unlimited division. So eradicating these cells is critical for eradicating the tumor. And what we found is that when cancer cells were exposed either to hypoxia or to chemotherapy, there was a great increase in the expression of proteins that control the breast cancer stem cell phenotype. And these are proteins, Nanog being one of them, that were first identified in embryonic stem cells and are responsible for the stem cell phenotype in ES cells. And what we found was that all of these boxed symbols represent genes that are induced by either hypoxia or treatment with chemotherapy and are controlled by the HIFs. And they all lead through different pathways to the production of this protein Nanog that determines the cancer stem cell fate. If we treat with chemotherapy and a HIF inhibitor, we don't see any increase in Nanog production. Finally, we've been interested in the role of HIFs in controlling the production of proteins that allow cancer cells to avoid killing by the immune system. And these come in two types. The first is what's called the adaptive immune system. And these are particularly T lymphocytes that can recognize cancer cells as foreign and kill them. But cancer cells express several proteins. One of them is a protein called PD-L1 that binds to a protein called PD-1 on T-cells and causes the T-cells to become anergic or can even kill the T-cells. A second protein called CD73 makes a small molecule called adenosine that binds to a receptor on T-cells and has the same effect. It causes the T-cells to become inactive or to actually kill them. Then a second branch of the immune system that can kill cancer cells is called the innate immune system. And this consists principally of macrophages. And macrophages are very good at eating cancer cells. So, the macrophage will wrap itself around the cancer cell and eat it. However, if the cancer cell expresses a protein called CD47, it signals to the cancer cell and it blocks the ability of the macrophage to kill the cancer cell. So, you can see that all of these proteins are induced when cancer cells are exposed to either hypoxia or chemotherapy. And we think that this is an important mechanism by which these cells can evade killing by the immune system. And again, when we use HIF inhibitors, we find that the expression of these genes is blocked and we can prevent the loss of active T-cells within the tumor microenvironment. So, we're quite interested in developing inhibitors that could be used in patients. The drugs that I've described to you are good for our animal models, but for a variety of reasons are not appropriate for use in patients. And so, a great effort in my lab now is identifying new inhibitors that will block the HIFs and will be appropriate for use in patients. I'd just like to summarize then what I've tried to illustrate to you, the role of the HIF pathway in biology and medicine through development, physiology, disease, and even in evolution where people who live at high altitude in Tibet and the Andes have genetic alterations in the HIF pathway that allow them to adapt to life at high altitude. Finally, I'd like to dedicate the lecture to Dr. Rose Nelson, the late Dr. Rose Nelson. She was my high school biology teacher and my inspiration in science, and I'm here because of her. The other people I would like to thank are my mentors, Haig Kazazian and Stylianos Antonarakis, who are here today. They were very generous in helping me and allowing me to have independence to pursue this project, and I can't thank them enough. And the late Victor McKusick, who also was extremely generous in helping me in my early career. I'd also like to acknowledge two faculty members at Hopkins, John Gearhart and Tom Kelly, who played critical roles in our studies of the erythropoietin gene and the purification of HIF1, and Guang Wang, who was my first postdoc and who purified HIF1. Finally, colleagues that I've collaborated over the years, Larissa Shimoda, Chi Dong, Akrit Sodhi at Hopkins, Joe Prakal at the University of Utah, Nanduri Prabhakar at the University of Chicago, and finally, Mike Armstrong, who had endowed a professorship that I hold and has provided us with great support in our cancer research. And then finally, I'd just like to illustrate that this work was done by my lab, not by me, and that I've been very fortunate to have a large group of undergraduate students, graduate students, and postdocs over the years who have performed these studies. I've been fortunate to have many colleagues on the Hopkins faculty with whom I've collaborated, and then faculty elsewhere around the world that we've worked with to try to understand the role of HIFs in physiology and medicine. And last but not least, I'd like to thank—thank you. Okay, I've got it together. I'd like to thank my wife, Laura, my sons, Evan and Gabe, my daughter, Ali, for always being there for me, giving me unconditional love and support. Thank you so much.