 All right, so good afternoon everyone. Again, my name is Jason Canner and I'd like to welcome you all back for the second of our 2020 Harry's Theme Box Lectures in Biochemistry. It is again my distinct pleasure to introduce Nobel Laureate Dr. Bill Kalen. So again, is the Sydney Farber Professor of Medicine at Harvard Medical School and the Dana Farber Cancer Institute. And is also a senior physician scientist at Bergen and Blumens Hospital. So I'd first like to recognize Professor Kalen for what I think we can all agree was a phenomenal first seminar yesterday. And it really left us with some really critical take-home points. And so for any that may have arrived a bit late yesterday, it does appear that a recording of the lecture can now be found on the Biochemistry website. Now, as I described in part yesterday, Bill has deservedly amassed more accolades that I could possibly tell you about now. But again, include his membership in both the National Academy of Sciences and the National Academy of Medicine, the Canada Gairdner Award, the Lasker Award and the Nobel Prize in Physiology or Medicine that he was awarded just last year, jointly with Peter Ratcliffe and Greg Samenza. And so finally, I'd again really like to thank Professor Kalen for accepting our invitation to deliver this year's Team Bop Lectures and for offering all of us some fantastic advice and things to think about. I know we're really looking forward to this seminar this afternoon. And so now without further ado, again, as best as we can in this context, I'd ask you all to join me in welcoming back Dr. Bill Kalen. Well, thank you very much, Jason. Thank you for this very nice introduction. I wish I could be with you in person, but hopefully this is the next best thing. And I'm honored to be giving these Team Bop Lectures. And deciding what to talk about, looking back, if I'm really being honest, some of the best talks I ever went to, especially when I was younger, were talks that sort of laid out sort of the history of a line of investigation and how it had unfolded and what the thought processes were. And that was very helpful to me. So that's the kind of talk I'm gonna at least try to give. You know, having said that, I suffer from this very common delusion that even though I'm too busy to read everybody else's papers, somehow magically people have read my papers and are just gonna roll their eyes at what I'm about to present because they're gonna view it as all historical and what have you done for me lately. But if you'll forgive me, I think this is probably the better talk, especially for the young people watching today. And so here's my disclosure slide. And here again are the people who've worked on my Hipple-Lindau disease in my laboratory over the years I've been more than lucky in terms of the number of talented people who've worked in my laboratory. And here's just a partial list of our very talented and very generous collaborators. So when I was younger, for a variety of reasons, I thought I was gonna be a clinical doctor. So here I am at Johns Hopkins Hospital as an intern and internal medicine in 1983. My chairman when I was an intern was the great Victor McCusick, who was really one of the fathers of modern human genetics, at least human clinical genetics. And he taught me the power of human genetics. And he was also a bit of a medical historian and he taught me to appreciate the work of people who had come before us. Now, I was so sure I was gonna be a clinical doctor that I spent the next year at Johns Hopkins as a chief resident, which was wonderful for several reasons, not the least of which was I met my late wife, Carolyn, during that year and we had a very happy life together. I came to Boston in 1987 to do an oncology fellowship at the Dana-Farber, again expecting to be a clinical doctor. But the person who changed my life was David Livingston, who was my postdoctoral mentor. I had worked on the RB tumor suppressor gene in David's laboratory and it was really David who taught me how to be a scientist and I think like a scientist. And so I really owe him a great debt. So I started my own laboratory in 1992. I sort of futzed around for a little while trying to figure out what I would work on that would start to distinguish me from my former mentor, David Livingston, whose lab was down the hallway from mine. And fortunately, this paper across my desk, I think in the summer of 1993, which was the cloning of the Von Hippolendau tumor suppressor gene. And for a variety of reasons I'm gonna share with you, I thought this would be a great problem to work on. And as I mentioned, I had worked on another tumor suppressor gene, namely the RB tumor suppressor gene in David's lab. So I thought I was well equipped to study such a gene. So again, I mentioned having a sense of medical history. So this story in hindsight arguably begins with this paper in 1894 by Treacher Collins, who had described a brother and a sister, both of whom had unusual retinal lesions that we would now call either retinal angiomas or retinal hemangioblastomas. These are vascular tumors of the retina that are easiest to see if you inject the patient with a fluorescein dye. So this patient has a large retinal hemangioblastoma. A similar family was described about a decade later by the German ophthalmologist, Eugen von Hippel. But the real hero in the story was Arvin Lindau, who was a Swedish neuropathologist. And being a neuropathologist, he appreciated that these retinal lesions were simply the tip of the iceberg and were often associated with hemangioblastomas of the cerebellum and spinal cord, as well as some tumors of other organs. And so just to show you again what some of these would look like, on the upper left, I'm showing you the fundoscopic view of a patient with a BHL disease. So here, orientation is the optic nerve head and this patient has a hemangioblastoma. Here, I'm showing you a fluorescein angiogram and a BHL patient with multiple small retinal hemangioblastomas. So at this point, they can usually be treated with a laser, but if they're neglected or in a very sensitive place, they can go on to cause retinal's attachment or blindness. But again, these retinal lesions are sometimes associated with hemangioblastomas of the cerebellum or spinal cord. So here's a patient with a BHL disease who has on the sagal MRI, a large cystic hemangioblastoma with the solid component enhancing here indicated with the L triangle. These patients can also develop tumors of some other organs. So for example, they can develop numerous renal cyst. So here, the urologist has opened or unroofed a large renal cyst in a patient with a BHL disease and they've discovered a clear cell, renal cell carcinoma arising from the epithelial cells lining the cyst cavity. And clear cell renal cell carcinoma is the most common form of kidney cancer. So with the apologies to treat your colons, this came to be known as von Hipple-Lindau disease. It affects about one in 35,000 people. Although the first report was about a century ago in hindsight, some families have been affected with this disease for many centuries. It's caused by loss of function germ limutations of the BHL tumor suppressor gene that resides on chromosome 3P25. I've mentioned the hemangioblastomas and the clear cell renal cell carcinomas. These patients can also develop a tumor called paraganglioma, which when it arises in the adrenal gland is referred to as a pheochromocytoma and a few other tumors. And I highlight in red clear cell renal cell carcinoma because hemangioblastomas and pheochromocytomas are interesting but admittedly pretty rare. Whereas clear cell renal cell carcinoma I mentioned is the most common form of kidney cancer and kidney cancer is one of the 10 most common cancers in the developed world. So I thought if nothing else studying the BHL gene would give us an opportunity to learn something about a common cancer, namely kidney cancer. Now just to make sure everyone's on the same page, I mentioned BHL disease is caused by germline mutations that inactivate the BHL tumor suppressor gene. So in this schematic, it's the maternal allele of the BHL gene that's been altered or has been inactivated by a mutation. But initially these people are okay because they have one remaining wild type of allele from mom or dad and the schematic is from dad. But the problem is they have about a 90% chance that at least one susceptible cell in their eye or brain or kidney, for example, will lose spontaneously the remaining wild type of allele and that's the cell that can go on to form a tumor. Now, as you would predict from the knowledge that germline BHL mutations predisposed to, for example, kidney cancer, if you now look at non-heritotary clearsal renal cell carcinomas, you again see that biolilic inactivation or loss of the BHL gene is a common feature. But here, both the mutational events or hits, if you will, occur somatically in contrast to BHL disease where the first hit has occurred in the germline. Now, I've already mentioned the vascular nature of these retinal hematomas, but another old clinical saw is that kidney cancers are notoriously rich in blood vessels. In fact, in the pre-cat scan are the diagnostic procedure of choice and patients suspected of having a kidney cancer was to do a renal angiogram to look for the characteristic appearance of new blood vessels throughout the kidney. And I should point out that at this time there was tremendous interest in trying to develop angiogenesis inhibitors to treat cancer based on part of the pioneering work of Judah Fogman. And I liked the idea, but I thought if we were gonna develop angiogenesis inhibitors, we really needed to understand the molecular circuits controlling angiogenesis. And so it seemed like studying the BHL gene might also provide insights into the molecular control of angiogenesis. Now, another clinical curiosity about these tumors seen in BHL disease is that they sometimes stimulate excess red blood cell production leading to what you can refer to as secondary polycythemia or perineoplastic erythrocytosis. So normally if you take blood and spin it down into capillary tube, about 40% of the volume should be the red cells themselves. But in polycythemia, now the hematocrit gets increased because you have too many red blood cells. And back when I was a chief resident, I would memorize lists like this. So these are the causes of excess red blood cell production. So clearly some of these are adaptive such as if you live at high altitude or you have certain chronic heart or lung conditions. So here having more red blood cells helps you to deliver oxygen more efficiently. But some tumors are also on this list. Some tumors can also stimulate red blood cell production. And it always struck me as odd that the three tumors seen in BHL disease make this list, especially considering that hematoblastomas and pheochromocytomas are otherwise relatively rare. And so taking this together, it struck me that BHL associated tumors were highly angiogenic. And we now know that's because they overproduce vascular endothelial growth factor. And I just told you they can stimulate red blood cell production. And that's because they connectopically produce erythropoietin. And what these two responses have in common is that they would normally be seen if a tissue wasn't getting enough oxygen. If you weren't getting enough oxygen, you would try to increase oxygen delivery by increasing blood vessel formation and increasing red cell mass. And so if I got one thing right, it was surmising that studying the BHL gene would provide clues into how cells and tissues in our body sense and respond to changes. In oxidants, it seemed like the sensing had gone awry in these various tumors. Now, this is an old fashion experiment. This is for the young people how we used to measure. mRNA abundance and so-called northern blot assays. So here what I'm showing you are northern blots using radio-labeled probes for VEGF, the PDGFB chain, and GLUT1, all of which were well-studied hypoxia-inducible or hypoxia-regulated genes. The trick was to grow the cells under low oxygen or high-oxidant conditions, and then harvest RNA and do these northern blots. So for orientation, let's start with HEP3B cells. So HEP3B cells are VHL proficient. And they're a workhorse cell line in the hypoxia-inducible gene field. And you could see that these three mRNAs accumulated, but only if the cells were made hypoxic, hence hypoxia-inducible mRNA. So that was expected. But now on the left, let's look at a VHL null or defective renal cell cross-nomal line, as was done here by Othoenoleopolis and Andrew Levy. So here you could see that now these hypoxia-inducible mRNAs accumulated, whether oxygen was available or not. And this was specific because when Othoenoleopolis restored the function of the VHL gene by stable transfection, as he did here with three independent subclones, restoring the function of the VHL gene restored the regulation of these hypoxia-inducible mRNAs. And this was specific because if he instead introduced a CDNA for a tumor-derived mutant VHL, or in this case, introduced an empty vector without a VHL CDNA, now he still saw a constitutive high-level expression of these hypoxia-inducible mRNAs. So summarizing this finding, what Othoenoleopolis and the lab had really done was he had isogenic cells where the VHL protein was present or was absent or defective. He grew them under low-oxidant or high-oxidant conditions. And what I just told you is when he measured hypoxia-inducible mRNAs, he found that in the VHL defective cells, they produced high levels of hypoxia-inducible mRNAs even when oxygen was abundant. So in other words, this was the first demonstration that loss of the VHL protein uncoupled oxygen availability from the production of these hypoxia-inducible mRNAs. Now, in parallel, we did biochemical studies, Adam Keivel, Kim Lonergan, and others in my lab. And they showed that the VHL protein binds to a long in C and a long in B and CULTU. And that turned out to be a big break in the story because what I didn't tell you was that the primary sequence of the VHL protein offered no clues as to what it might do. But it was my now colleague, Steve Elage, who pointed out that a long in C looks like a yeast protein called SKIP1 and CULTU, a member of the Cullen family, looks like a yeast protein called CDC53. And Stephen's work and Ray Deshaies' work and others had shown that these two proteins went bound to a cell called F-box protein, generate an SCF-like ubiquitin ligase that can target other proteins for destruction. And so by analogy, we began to think that maybe the VHL protein was likewise the substrate recognition subunit of a ubiquitin ligase complex. I should also point out that there are two hot spots for mutations in the VHL protein and VHL disease, the alpha domain, which we could show was responsible for recruiting the other members of the complex. And the beta domain, which we hypothesized was the substrate docking site. And this idea was strengthened further by X-ray crystallographic studies that we did with Nicola Peblege. So then the $64 question was, well, what is the substrate of this putative ligase? And a good candidate was the HIF transcription factor because it was known from the work of Greg Semenza, Peter Radcliffe, Heine Karo, Frank Fund and others, that the HIF transcription factor was a master regulator of these various hypoxia-inducable mRNAs, such as that Japan Epo. And we knew that it consisted of an unstable alpha subunit that was normally degraded in the presence of oxygen. And a constituent will be stable beta subunit, which is often referred to by the alternative name of ART. And so at this point we were lucky because Greg, excuse me, Peter Radcliffe published a paper in Nature showing that in fact cells lacking the VHL protein are unable to destroy HIF alpha under high oxygen conditions. And so based on that clue, we immediately showed that as we had suspected, this impact is the ubiquitin ligase for the HIF alpha subunits and targets them for proteasome degradation when oxygen is present. Whereas when oxygen levels are low or the VHL protein has been mutated, now HIF alpha can accumulate and dimerize with orange inactivate genes such as VHF and Epo. So that was very gratifying because it explained why tumors lacking the VHL protein would over express things like VHF and Epo. But as you hope happens in science having answered one question and we know we're faced with a bigger question, which is how does the VHL protein know if you will whether oxygen is or is not available and hence whether it should destroy HIF alpha? And since this is a biochemistry talk and since these are some of my favorite experiments of all time, I'm just gonna show you a couple of the experiments that were actually informative in this regard. So this is an experiment that was done by Haifeng Yang when he was in the lab. So he took TS-20 mouse fibroblast, which are VHL proficient, but they have a temperature sensitive mutation in the E1 ubiquitin activating enzyme. So when you shift these cells to the restrictive temperature polyabic relation will stop. So the trick here was to grow them under the permissive or the restrictive temperature. But in the case of the restrictive temperature he grew them under high accident or low accident conditions. So in the bottom block, I'm showing you an anti-HIF alpha immunoblot and you can see that HIF 1 alpha accumulates at the restrictive temperature as you would expect because under the restrictive temperature polyabic relation will cease. But the informative experiment is the top filter. So in the top filter, Haifeng did a so-called far western blot where he incubated the nitrocellulose filter with recombinant VHL protein actually together with O-long and BNC and then he detected the bound VHL after washing with an anti-VHL antibody. And here you could see, first of all, that VHL could bind directly to HIF 1 alpha which frankly wasn't known at that time. So that was an important piece of the puzzle. But secondly you could see that VHL would only bind to the HIF 1 alpha that had accumulated in the presence of oxygen but not the HIF 1 alpha that had accumulated under low accident conditions even though there's comparable amounts of HIF 1 alpha present. And so that said that the signal was going through HIF 1 alpha which was not known at that time. And finally, since this interaction survived boiling in SDS and running through a gel and being slapped on a nitrocellulose filter we could also surmise that the binding region in HIF 1 alpha would likely be peptidic and that turned out to be true. So Mircha Ivan in the lab actually then mapped the minimal region of HIF 1 alpha that was sufficient to bind to VHL. And as we predicted, it was a peptide. In fact, he could get it down to a 25-mer peptide that had at its core the sequence MLIP YIPM. So he made this 25-mer as a biotinylated peptide that he captured on streptabin and agarose and then did pull down studies with S35 labeled VHL protein made by in vitro translation. But he noticed this peptide only acquired the ability to bind to VHL if it was first pre-incubated with a mammalian cell extract I think it was 30 degrees for an hour or so presumably because the mammalian cell extract would then provide the modifying activity that was necessary to modify this peptide such that it could now capture VHL. So now we got lucky here because Jaime Carrow who had been doing linker scanning mutagenesis of HIF 1 alpha had already shown that if you replace these eight residues in HIF 1 alpha with eight consecutive alanines that HIF 1 now became constitutionally stable. So Mercia made the 25-mer peptide with now these eight residues replaced with these eight alanines and now it wouldn't bind to VHL. He then did an alanine scan and it turned out the two other critical residues were this proline residue shown here and not shown here at this leucine. So here I'm showing you if you mutate the proline to an alanine now you don't bind. So then we look for examples of enzymes that might modify proline and might in principle be oxidant sensitive. And also we had shown that iron key ladders could block the interaction between HIF and VHL and it was known that iron key ladders could mimic a hypoxic response. So we added in, are there examples of enzymes that can modify proline, that can potentially be oxidant sensitive and iron sensitive and we learned about certain prolyl hydroxylases. So we made a guess. We guessed that the modification was prolyl hydroxylation. So Mircha made this peptide so that now the proline was already prehydroxylated. And now you can see this peptide captures VHL but now it no longer needed the pre incubation step presumably because we've already introduced the relevant modification. And so this was one of those, you know, once a decade or so Eureka experiments where you're running up and down the hallway and jumping up and down and doing high fives. But I always tell my postdocs before gravitating to the most interesting explanation which in this case would be that we've now discovered the modification. Think about uninteresting explanations. And so if you think about it there would be some uninteresting explanations for this result. So one uninteresting explanation would be that this is just a sticky peptide that would have bound to any S35 label protein. We just happened to look at VHL. So that would not be a very interesting explanation for this result. So to get around this, Billy came in the lab I did an experiment where rather than adding S35 labeled VHL made by in vitro translation he labeled mammalian cells with S35 that on the left were VHL proficient and on the right were VHL defective. On the left these cells are topically expressing an HAVHL. And you can see now when he did the pull down with the hydroxylated peptide that this is exquisitely specific. You get VHL along and B along and C, Col2, a couple other interesting proteins we're not gonna talk about. So this is exquisitely specific. So now we could at least say this was a specific interaction. It wasn't non-specific. The other uninteresting explanation would be that either the modification really is prolohydroxylation or the prolohydroxylation mimics the authentic in vivo oxidant-dependent modification. And so to address that first of all, we did mass spec, which is the high tech way I suppose of showing that it was prolohydroxylation and it showed that hip one alpha is hydroxylated in vivo but the low tech solution was as follows. So Mirchah had already shown that if you produce hip one alpha by in vitro translation using retic lysate it combined to VHL. And we now know that retic lysate contains the necessary prolohydroxylase whereas it be made hip one alpha in a wheat germ extract. It would not bind to VHL and that's because wheat germ lacks the modifying activity. So what Mirchah Ivan did here was the in vitro translated hip one alpha in the presence of tridiated proline so that the prolings would be radioactive. He then hydrolyzed the protein to completion and separated the amino acids by TLC with appropriate standards. And he could see that when he made hip one alpha and retic lysate he could see both proline but he could also see now hydroxylated proline. So the answer turned out to be that in the presence of oxygen hip one alpha gets hydroxylated on one of two prolyl residues and the same conclusion was reached in parallel by Sir Peter Reckliff's group. So this was very exciting and immediately begged the question what enzyme is doing the work. We thought we were well positioned to purify the enzyme biochemically so we teamed up with Joan and Ron Conaway. I already told you that rabbit reticulocyte lysate contain the hydroxylase activity. So we partially purified the hydroxylase using old fashioned, you know, column chromatography and then we monitored which fractions contain the hydroxylase based on their ability to hydroxylate that biotinylated peptide as determined by capture of S35 label VHL. And so we arrived at the proly hydroxylase of Eglenn one as we were putting the paper together Peter Reckliff's group and Steven McKnight's group using genetic approaches in worms and flies came basically to the same answer. So at least we got the right answer. So the answer was that the enzymes here are the Eglenn proly hydroxylases they're sometimes referred to as the PhD proly hydroxylases. This mechanism turned out to be very simple and elegant compared to some of the models that were prevailing at the time. So these enzymes split molecular oxygen and use one of the oxygen atoms to hydroxylate hip alpha. Fortunately, they have very low oxygen affinities and so they're very sensitive to changes in oxygen availability in a physiologically relevant range. This is in contrast to, for example, the collagen proly hydroxylases that have very high oxygen affinities. They require iron which explains the iron chelator result and they also require a co-substrate called two oxyglutrate or alpha ketoglutrate which gets decarboxylated to succinate and hence couples this reaction to metabolism as well. Now there are three members of this family, Eglenn I, Eglenn II and Eglenn III with the alternative names shown here. But by all criteria, the major workhorse regulator of hip is Eglenn I and perhaps for that reason you can't make it through mouse embryogenesis without Eglenn I. And so to circumvent this embryonic lethality Andy Minamishma in the lab made mice where he could conditionally inactivate Eglenn I in adult mice using a fluxed Eglenn I alleol and an inducible creed that can be activated with tamoxifen. And when he did that, the mice develop massive polycythemia. Actually, the first hint of this was the pause here we're getting slightly too red and the vessels were engorged but these animals are developing polycythemia and about the same time work from, frankly, Joseph Burkow and others taught us that certain families around the world that develop excess red blood cell production on a genetic basis have mutations in this pathway. Some have homozygous or less commonly compound heterozygous hypomorphic VHL mutations so-called Chivas polycythemia. Others have hypomorphic Eglenn I mutations and yet others have hypermorphic hip to alpha mutations. And I should also point out that adaptive polymorphisms in this pathway have also been identified in humans that have adapted to life at high altitude such as the Tibetan. So there's pretty good evidence that this is the oxidant sensing pathway at least with respect to red blood cell production. Now you might ask, why don't those families look like VHL disease? Why aren't they conspicuously cancer prone? And I think it's because the genetics here are slightly different. So let's on the left think about a kidney tumor which is going to be effectively VHL null. If it's a sporadic tumor, the host is VHL plus over plus. If this is in the setting of VHL disease, the individual has a defective VHL allele but there's no evidence for haploinsufficiency for the VHL gene. So most of the cells are going to be fine at least with respect to regulation of hip which I'm showing you schematically here. So here we have very high levels of hip but it's confined to a clone that's emerged that's effectively VHL null. In contrast in the familial polycythemia situations the defect is present in the germline and every cell in the body that could be making slightly too much Epo is making slightly too much Epo. And when measured, these mutations lead to a very subtle but measurable increase in hip. So this is a small defect in hip but measured times many cells. This is a bigger defect in hip but confined to a particular clone or clones. Now going back to this model, the polycythemia in these mice not surprisingly is being driven by Epo. And all of the Epo that we can detect in these mice is coming from the kidney and that's not surprising because in the adult it's the kidney which is the major source of Epo. But that's not true during fetal life. During fetal life, it's the liver that makes Epo and then shortly after birth the hepatic Epo locus gets silenced and then the kidney takes over. And this has both economic and medical implications. So for example, there are about 20 million Americans with chronic renal failure and two to four million of these are chronically anemic. So Andy Minamishma wanted to know whether you could somehow reawaken the hepatic Epo locus in an adult mammal. So what he did was he created mice where he knocked out the prolohydroxylases in the mouse liver either singly or in every pairwise combination or he knocked out all three for comparison. He also knocked out the HL. He then measured hepatic Epo mRNA in the black bars and circulating certain people in the white bars. What he saw was when he knocked out all three he got a very robust reawakening or induction of hepatic Epo production. Not shown here is when he knocked out Eglon-1 at least transantly you got a blip of Epo but then it went back down because of compensation by the other paralogs. And so basically depending on what you wanna do you can either cause a pulse of Epo or you can have a high steady state of production of Epo. Now these enzymes turn out potentially to be druggable and in fact we reached out to a company called Fibrogen about 20 years ago. They were making small molecule inhibitors of the collagen prolohydroxylases as potential anti-fibrodics. And so they were making a number of two oxalid rate competitive inhibitors of the collagen prolohydroxylases and we reached out and asked whether we could test whether some of their compounds might fortuitously inhibit the hip prolohydroxylases because if we did we thought that would be a good starting point for making specific inhibitors of the hip prolohydroxylases. So here's a early tool compound FG4497 that was added to cells and culture and you see this nice induction of HIP1-alpha. Macal Saffron in the lab had made a reporter mouse that ubiquitously expresses a HIP1-alpha Luciferous fusion protein that contains at least the business end of HIP1-alpha that's sufficient for oxygen-dependent regulation. And she gave this tool compound to the mice by Orwell-Gravage and you could see that HIP was being stabilized as determined by increased light emission from these mice. Eventually, this gave rise through additional medicinal chemistry efforts at Fibrogen to a compound called Roxadustat, which is the first in class inhibitor of the hip prolohydroxylases. And I should mention I have a financial conflict of interest again in Roxadustat. But here I'm showing you data from phase two trial and patient with chronic kidney failure who were treated with this orally available a hip stabilizer. So in blue, I'm showing you plasma Epo levels in the individuals got Roxadustat compared to the placebo in red. There's no evidence of pharmacological tolerance. You can bring the patients back in a month and dose them again and turn Epo on or off at will. As you would hope this translates into an improvement in hemoglobin production or measure of red blood cell mass, even in patients frankly who are Epo refractory. For a variety of reasons, the phase three trials in China were finished before some of the phase three trials in the West although those trials have now been completed as well. And Roxadustat has now been approved for the treatment of anemia and chronic kidney failure in China and Japan. And the US so-called Padoofidate is December 20th of this year. So hopefully we'll learn by the end of the year. But there are a number of these drugs that have advanced to phase three testing. And so hopefully some of them will actually turn out to be useful for patients with anemia. Now, one thing that's very gratifying is having discovered this system. It turns out that every multicellular animal on the planet is using this system. So here I slide, I borrowed from Peter Ratcliffe where I'm showing you the conservation of the HIF family. He prefers the PhD nomenclature for the prolydoxases and BHL. We're not gonna talk about FIH today. But you can see all multicellular animals have this system as often happened with genes during evolution, there were some gene duplication events which gave this system some added richness. But this is the system that multicellular animals on the planet are using to sense and respond to changes of oxidin, at least in terms of large scale changes in gene expression. Now, presumably the selection pressure here is not to cause BHL disease. The selection pressure is to have a system that would allow you to survive in a low oxidin environment if you were faced with that insult or situation. And so just based on teleological arguments, we wondered whether these drugs that inhibit Eglene or PhD prolydoxases would also be useful in various diseases that are characterized by impaired oxidin delivery, such as heart attack or stroke. And it's certainly true in preclinical models. So we and others have modeled this in, for example, rodents. Here I'm showing you data from Ben Olenchuk and Java Moslehi in my lab where here using that inducible Cree system and the Flock's Eglene-1, they've acutely inactivated Eglene-1 at the time of experimental myocardial infarction and mice. And here on the y-axis, I'm showing you the amount of heart muscle damage after including a major coronary artery and releasing it. So using either here a genetic tool or not shown here, pharmacological tools, they see substantial protection. So we are hopeful that maybe someday these drugs will also be useful for the treatment of ischemic diseases. So I'm now gonna put my cancer biology hat back on, which on one level is where I started. So let's now return to the role of VHL in cancer. So let's look at, for example, kidney cancer. So again, patients with VHL disease are effectively VHL-Hutters-Igoes, but I already told you there's no evidence for haploinsufficiency of the VHL gene. So then what happens over time is they can develop numerous renal cysts. And when examined, these cysts are lined by epithelial cells that are VHL and null. So apparently at least in the human kidney, VHL loss causes perineoplastic renal cyst. And then over years to decades, they can develop renal cell carcinomas of the clear cell type. And when examined, these tumors have stereotypical mutations involving other genes, presumably these reflect cooperating events. So one conclusion from this experiment of nature is that VHL loss, even if it's a critical step in renal carcinogenesis, is not sufficient for renal carcinogenesis. And it also says that VHL loss can be the initiating event or some would call it the truncal event. But of course here, the deck was rigged because the person of the germline VHL mutation. So what about its sporadic clear cell renal cell carcinomas? Well, here are very similar picture emerges. And this is based on the work of Charlie Swanton and co-workers who've taken kidney tumors and done multiple spatially distinct biopsies of those tumors, including when possible of metastatic deposits, and then have done deep sequencing and abused mutant allele frequencies to infer the evolutionary histories of those tumors. And almost invariably, they see that inactivation of the VHL gene is the initiating or truncal event and that there are then these later cooperating events that often occur in a subclonal or branching pattern. So it appears whether you're a hereditary kidney cancer or a sporadic kidney cancer, loss of VHL is sort of the gatekeeper. So by the time you're a clear cell renal cell carcinoma, it is VHL function even matter anymore. And so to address that, Osana Eliopoulos in the lab took VHL null renal carcinoma cell lines, he restored the function of the VHL protein and he saw that these cells could still grow on a plastic dish, but they lost the ability to form tumors. So then to address the role of HIF, KG Condor took these cells and he introduced into these cells a version of HIF alpha and in particular HIF2 alpha that couldn't be recognized by the VHL protein because the proline hydroxylation sites have been converted to alanine and this restored the ability of these cells to form tumors. Conversely, when he took these cells and eliminated HIF2 alpha, initially using short hairpin RNA technology and later we did this with CRISPR. Once again, these cells can grow pretty well on plastic but they lose the ability to form tumors. And this is really a specific property of HIF2 alpha. In fact, when we did analog experiments with HIF1 alpha, if anything, HIF1 alpha had the opposite effects, which led us kicking and screaming to the idea that at least in this context, a HIF2 alpha is acting like an uncle protein and HIF1 alpha is acting as a tumor suppressant. In fact, in some kidney cancers, you can't even detect HIF1 alpha expression. So what can we do about this? Well, I already told you that it was known that HIF was a regulator of VEGF and a number of companies fortunately were making VEGF inhibitors by the 90s and we argued if they were gonna work in any solid tumor, they would work in kidney cancer. It was very gratifying to see these data from Genentech were on the Y axis, I'm showing you a measure of VEGF production and along the X axis, I'm showing you different tissues and in green are the VEGF levels in the normal tissues and in red are the VEGF levels in the tumors derived from those tissues. So it is true that VEGF is modestly elevated in a variety of solid tumors. That's because most solid tumors contain regions that are hypoxic, but the 800 pound gorilla is kidney cancer where VEGF goes through the ceiling. Presumably this reflects the fact that from the earliest days of carcinogenesis, VHL was lost, HIF was upregulated and hence VEGF was upregulated and presumably that led to decreased selection pressure to turn on alternative or collateral angiogenic pathways. So in fact, we now have seven VEGF inhibitors approved for the treatment of kidney cancer and that's been very gratifying, but some patients don't respond and even those patients who do respond will eventually progress or relapse. So how can we do better? So just based on first principles, you might argue, well, why target any one HIF target gene? Why not just target HIF itself? And based on what I told you, why not target HIF2? But unfortunately the conventional dogma at the time was that HIF2 was not drugable, but fortunately Rick Rueck and Kevin Gardner ignored that and they identified a drug-able pocket in HIF2 alpha. They also developed or identified chemical scaffolds that could bind to this pocket and in so doing induce an allosteric change such as HIF2 alpha would no longer bind to Arnton, hence could no longer bind to DNA. They then out licensed these scaffolds to Peloton and they did medicinal chemistry to make PT2399 which is a more potent, more specific, more bioavailable HIF2 alpha inhibitor. They were kind enough to share it with us and we used it in pre-clinical models and showed that this compound could decrease HIF dependent mRNAs, could decrease proliferation X vivo and could decrease orthotopic tumor assays. And I think I showed this yesterday about, here's an example of such an experiment where Chin Cho in the lab took arino-carcinoma cell line that was VHL defective and treated them with 0.2 or two micromolar PT2399 and she saw a decrease in soft agar growth. But again, how would you know this is on target? Maybe this is just some noxious poison that inhibits HIF2 but also does many other things and this is really just a toxicity effect. While the way as we discussed yesterday to deal with that is to do a rescue experiment and here we were aided by Bruick and Gardner who had identified a point mutation in HIF2 alpha that would prevent these drugs from binding to the pocket, such drugs from binding to the pocket but would otherwise leave HIF2 alpha intact and so that allowed Chin in my lab to really do the experiment this way where using CRISPR, she generated isogenic VHL defective arino-carcinoma cells that were expressing wild type HIF2 alpha or this drug resistant variant and with the drug resistant variants you could see that now these cells continued to grow in soft agar assays despite the presence of the compound. So this said that this and the other effects we were measuring were on target. Eventually a clinical trial occurred with what is now the most current version of this HIF2 inhibitor. These were patients with advanced kidney cancer who had failed multiple lines of therapy. 90% of them had failed a VEGF inhibitor and 70% had failed a immune checkpoint inhibitor. These are so-called swimmers plots where each horizontal bar is a patient and how long they were on study at the time of this analysis. So orientation here's one year on therapy. The patients with the black arrows were doing well at the time of this analysis and the patients with the yellow stars had achieved the partial response by resist criteria. So at least some of the patients seem to be benefiting and based on this the drug has gone into phase three testing but some of the patients were not benefiting and were trying to understand this. Now I mentioned these were heavily pretreated patients and you might have wondered how might this drug have fared if it had been used more in a frontline setting or had been used in patients who didn't have such advanced disease? And we were fortunate that we were able to convince Peloton to also test this HIF2 alpha inhibitor which is now called MK6482 by the way now that it's been acquired by Merck in patients with von Hippo-Lindau disease who had not been treated medically before. To be on this study these patients had to have measurable renal tumors that were currently being monitored in careful surveillance programs in an attempt to delay or prevent the need for repeated partial nephrectomies. So you had to have a measurable kidney tumor but you could have incidentally a mangyoblastomas of the eye and brain. You could have pancreatic lesions, other lesions. And so these data were presented by Eric Yonash and colleagues at ASCO this year but I'm pleased to report that 87% of the patients had some measurable tumor shrinkage of their kidney tumors. About 40% had a confirmed partial response or a partial response awaiting independent confirmation that the median progression free survival has not been reached and the 12 month progress free survival was 98.3%. And I should also point out that response is worst seen in some of the mangyoblastomas and some of the other incidental tumors. And so now this is what the swimmers looks like for the BHL population. You can see now that it looks like at this analysis most patients are doing pretty well on therapy and most have passed the one year mark. Frankly, even before the data were presented I knew that things were going in the right direction because some of the patients were posting their findings on social media sites. So here for example, as a patient saying I never thought I'd see this day describing that some of their tumors were getting smaller or in some cases had disappeared. And frankly, these patients have been living with this sort of democles over their neck for years having watched other members of their family over time suffer from this disease. So this has been really very gratifying. Everybody loves a video. So I'm gonna share a video that was sent to me. Hey everybody, it's Justin. I just wanted to give you a quick update. I am in a gondola right now and Taiwan over there is Taipei 101. I'm actually right by the Taipei Zoo. But I just wanted to give you a quick update and say I'm doing well and enjoying my trip. If it wasn't for the PT2977 drug trial I would have never been able to come out here and do what I'm doing right now. So I just wanted to thank Peloton and I hope Merck will fast track this drug for a BHL treatment. So if you guys are listening hopefully you guys will put it on market to help BHL. But yeah, keep watching these videos. I'll be making more and I'll get better at it and I have to get the angles right because it kind of looks fat. I love that because at that age that's the kind of thing you should be worried about. You should be worried about whether you look fat in your vlog and not what you're not being worried about what your next cat scan looks like. So hopefully today I've shared a couple examples of things that actually got translated. But one of the things that I'm very concerned about is I think there's just too much emphasis frankly on translation. I think we should be in the knowledge generation business. So I think real translation happens not when you put a gun to someone's head and say, go translate. Real translation happens when you actually understand something well enough that you've created an opportunity to do something good. So I call this the translational moment. So for example, think about all the things that had to be learned before we got to Gleevec for the treatment of CML. We had to learn about the Philadelphia chromosome. We had to clone the BCR-ABL fusion. We had to understand that ABL was a kinase. We had to show that this fusion protein caused the disease. The kinases could be inhibited with ATP analogs, et cetera, et cetera, et cetera. And so I think the stuff on the left which is the knowledge generation stuff that to me is science. And then as you start to learn more and more of the rules and you're gonna apply them that gets to be more like engineering at least to me. And maybe for that reason, the stuff on the left has usually been dominated by academia and the stuff on the right has usually been dominated by biotech and pharma. And my friends in biotech and pharma tell me, it's great when you academics wanna play over here but please don't stop doing this stuff because we can't do this stuff. The timelines and the deliverables are too uncertain. We can't do this stuff to the far left. And in fact, it's even worse because you can think of this stuff on the right being disease-oriented applied research and maybe this stuff over here as disease-oriented basic research. But of course, even this stuff was made possible if I can advance the slide by decades of investment over here which I call not disease-oriented basic research. This is what really enabled this which eventually led to translation. And parenthetically, there's a lot of talk about how quickly we're making progress in COVID-19 but that's because of all the investments we made over here such that when the COVID-19 virus came along we kind of knew most of the rules and we knew some of the principles in terms of what genes would you target and how would you design a vaccine, et cetera, et cetera. So I think there's much too much pressure to try to justify one's work in terms of translatability. And so that's why I wrote this little editorial about publishing houses of brick, not mansions of straw. I think we should get back to just judging things based on whether they're likely to be true and robust and not have every last figure of every paper be some attempt, some gratuitous attempt to link it to clinical translation. And so I think we should get back to forming sort of a symbiotic relationship with pharma where we create the knowledge and learn the rules and then they apply that knowledge for the betterment of patients. And with two other philosophical musings. So one is this is the first slide I showed you, the paper by Treacher Collins. And you might say, well, why did it come to be known as von Hippel-Lindau disease? Well, here was the paper by von Hippel. Now he was a German ophthalmologist. So it's perhaps not surprising that he published in a German language journal. But Arvin Lindau was a Swedish and he published in a German language journal as well. And that's because this was the great descriptive error in medicine. And if you could only publish in one literature and only read one literature, it was the German literature. In fact, I studied German in high school in the 70s because I was told if you wanted to go into science and medicine, you should learn German. And when I was a medical student, I once had a professor tell me, if you thought you had a new clinical observation, you should take a sabbatical and read the German literature from the late 19th century and early 20th century because what you had seen was almost certainly already described in the German literature. So the irony is Treacher Collins got penalized for publishing in an English speaking journal. Germany was dominant in biomedicine, but of course due to the horrific events in the middle of the last century, German science never really recovered. And so I tell you this because there's no rule that societies always continue to go in a good direction. They could also go backwards. And so I've been trying to lend my voice to those who are concerned about the attack on science that we've witnessed especially over the past four years and that the road we might have found ourselves on and hopefully we can start to change course here. And then finally to end on a happy note for the students, my very first laboratory experience was when I was a junior in college at Duke University. I worked in a chemistry lab on a project with sort of an absentee mentor in a dysfunctional laboratory. My project was uninteresting, unimportant and undoable. And then to make matters worse, I had the audacity during my last lab meeting to correctly say that this project, which had started seven years before I entered the lab was really based on an artifact and would never be brought to completion. So to reward me, my professor gave me a C minus, which for pre-med is like having a wooden stake driven through your heart. And then as an added punitive flourish, she wrote in the transcript of my college transcript that Mr. Kalin appears to be a bright young man who's future lies outside the laboratory. And I share this with the young people because it's almost certain that at some point in your career, you're gonna get knocked down and maybe someone won't believe in you. And you have two responses. One is you can wellow in self pity or you can use this as motivation and hopefully show that they were wrong. And I can tell you at the time, I was tempted to wallow in self pity because for example, the C minus had the expected chilling effect on some of my medical school applications. Thank goodness, Duke still took a chance on me. But in the end, it turned out okay. So thank you very much for your attention and I'll be happy to answer questions. Thank you. So I think what worked yesterday that we can try again is if people would like to input their questions into the chat window, like that's probably easiest approach. I guess while we wait, I have a quick question. In terms of the knowledge base that you were describing therapeutically, what is your take on drug repurposing and some of the serendipity behind that? Yeah, no, I think if you mean by that, for example, doing phenotypic screens with chemicals with at least thought of as having known functions and getting interesting phenotypes and then doing the work to show that either that phenotype really was due to the known activity of the drug or maybe in some cases will reflect some new activity, maybe some other target we didn't know about and maybe some fortuitous off-target effect, for example. You know, I think that's great. But I think more broadly, whatever the approach, whether it's repurposing chemicals, whether it's doing CRISPR screens, et cetera, et cetera, as we talked about yesterday, I actually think the pharmaceutical and biotech company is really good if you provide them high-quality actionable information. But as I discussed yesterday, they are wallowing in things that are either not completely true or true only under a very narrow set of conditions. That is to say, you're not very robust. So I think we have to just do a much better job. All right. One brave soul has decided to submit a question so far. So do you have any idea of what drives the acquisition of the loss of the second allele in germline mutation carriers with VHL? You mentioned no difference in his access. Could you prevent renal and other cancers in VHL or those at increased risk of RCC by targeting whatever the heterozygous effect is? Yeah, several great questions. So my thought here is colored a lot by my work on the RB gene. So I think the prevailing wisdom is that the loss of the remaining wild-type allele is simply a stochastic event, which frankly, in a lot of cells and tissues is probably not tolerated. So for example, one of the questions Al Coneuchin used to always ask me was why is retinoblastoma retinoblastoma? For example, RB is thought to be this master cell cycle regulator. Why the itumors and not other types of tumors? And I think part of the answer is almost certainly that most cells actually don't like to lose RB and don't tolerate it. And I'm sure the same is true for VHL. In fact, we know that's true. Most cells do not like biolilic inactivation of VHL. So it's only rare cell types, including specific cells in the kidney that will tolerate VHL loss. So I think it's stochastic. I might be wrong on that, but that's what I think. In terms, I think it's a very intriguing idea to target cells that are heterozygous for VHL as a chemo-preventative strategy. Now, of course, you couldn't do that in the VHL setting because every cell would presumably have the same defect, but it's a very intriguing idea with respect to sporadic kidney cancers. But there, I think the challenge is, again, to a first approximation, there is no phenotype of having only a single VHL allele. So I'm not even sure what the basis would be for selectivity. So in the meantime, what I'm very hopeful about are the HIP2 inhibitors, which are reasonably well tolerated. And so there you could start to imagine certain high-risk settings such as in VHL disease using the HIP2 inhibitor as a chemo-preventative strategy. And so that's something I have my eyes open to. Great. So the next question. I wonder if germline VHL null have any impact on the transcription factor regulating hematopoiesis during embryogenesis? And do VHL null red blood cells, macrophages, or neutrophils function normally? Yeah, that sort of relates a little bit to the previous question. So to a first approximation, there is no phenotype that I'm aware of for being VHL herozygous, meaning plus over minus. Most cells do not like being VHL null. So yes, there would be a phenotype. In most cells, for example, in most cells, paradoxically, you see an antiproliferative effect. If you knock out VHL, actually in most cells, what HIF does is it kind of puts you into a state of hibernation where you're trying to conserve energy, because you don't have enough oxygen and the cells don't proliferate very well. And there could also be indirect effects on differentiation. So I don't know about these specific examples listed here, such as macrophages and neutrophils, but I can almost assure you there would be a phenotype of being VHL null in those cells. All right. Next question. Do individuals slash populations with more quote unquote robust or responsive HIF transcriptional programs, e.g. communities that live at higher altitudes, have a higher susceptibility for developing diseases such as kidney cancer? Yeah, that's a great question. I mean, I guess if I'm really being honest, the short answer is I don't know. So let me answer some slightly related questions, because as soon as we started talking about stabilizing HIF with small molecules, the naysayers lined up and said, well, you're just going to cause cancer. You're just going to cause VHL disease. You're just going to cause kidney cancer. Now, this is a long and lengthy discussion why that probably won't be the case, but certainly one experiment of nature are people who live at high altitude. If anything, I'm told that epidemiologically, they have longer life spans than people who live at sea level. They certainly don't spontaneously develop VHL disease by virtue of living high altitude. At worst, there have been some case reports of carotid body tumors, which integral to how you sense oxygen and ultra ventilation and there have been a few other rare things, but they're not really a conspicuously increased risk for cancer. So I think HIF alone doesn't do it. I think it gets back to part of what I told you earlier. I don't think HIF alone can promote cancer, even in the kidney cancer setting, you need multiple other mutations and many other genes. And we didn't talk about, in fact, the VHL also does do some things other than regulate HIF. So, but I think to be fair to the questioner, I think it could be true that in time, we'll understand that polymorphisms in the HIF response do modulate your risk positively or negatively for cancer and other diseases. So I think it's a good question. So I will take a leap of faith in translating the next question, but is hydroxylated protein or protein hydroxylation involved in other diseases? Well, prior to our work, it was thought that hydroxylation of protein only took place in the endoplasmic reticulum. In fact, one of the reviewers who tried to reject our paper only wrote about a two or three line review, which was just has to be wrong because it's already known that hydroxylation of protein only takes place in the endoplasmic reticulum. But fortunately, we had a strong editor who overruled that reviewer. But the statement is still true. Most hydroxylation of protein occurs in the endoplasmic reticulum and the bulk of that is collagen. And so, for example, the classical answer to the question would be scurvy. So scurvy, which is caused by vitamin C deficiency is because you need also vitamin C for the collagen prolohydroxylases to function properly. So scurvy is caused by failure to hydroxylate proline on collagen. So that would be the best example. Great. Next question, is rocks and dust that specific for PhD2 or does it also inhibit PHP1, PHP3 and collagen prolohydroxylase? Okay, so we have a ringer in that. We have a true way. It actually is a ringer. There's always a couple in the crowd. So no, so for better or worse, and we could argue both sides of that, rocks and dust that is not specific for PhD2. It does inhibit PhD1, 2, and 3. And I don't know yet whether that's really a good thing or a bad thing. Certainly for sustained production, you'd want to inactivate all three. But frankly, the way it's being dosed currently, which is orally three times a week, the drug is still being given a chance to wash out and you're still resetting, which again, I think also maybe speaks to the safety issue. But since Ron's an expert, I will point out that fortuitously however, rocks and dust that does not inhibit that highly related or I should say that other regulator of FIH that I've only mentioned in passing. And that turns out to be probably important in terms of the ability to turn on EFO, but not to turn on things like that, Jeff. I mean, one final question. Do the HIP stabilizing inhibitors of prolohydroxylases inhibit the collagen processing enzymes with effect on extracellular matrices? Yeah, so there has been, I believe, a successful attempt to dial out the collagen prolohydroxylases. So to my knowledge, these are specific for the HIP prolohydroxylases relative to the collagen prolohydroxylases. I suspect for the reason hinted at here that you wouldn't start to, you wouldn't want to start affecting the extracellular matrix as an off-target effective drug. So that was dialed out, but I should point out that there are about 70 enzymes in this super family of so-called two-oxycutoride-dependent deoxynases. So it could be the case that loxobestate does inhibit some other members of the family, but I don't believe that the collagen prolohydroxylases are inhibited. Great. So, sorry, one more. So we focused on HIP 1-alpha degradation as a regulatory step. Is there variation or regulation at the production rate of the protein? Great question. The answer is yes. So because HIP 1-alpha is turning over so rapidly, it's exquisitely sensitive to both changes in degradation, but also changes in synthesis. And so, for example, if you want to down-regulate HIP 1-alpha, one thing you can do in many cell types is block M-tor, and you'll decrease the rate of translation of HIP 1-alpha. HIP 1-alpha is very sensitive to M-tor inhibitors. But more generally, a cautionary note for the students is precisely because HIP 1-alpha is turning over so rapidly, lots of other noxious things will make HIP 1-alpha disappear before typical loading controls, such as actin and tubulin. So if you read a paper about a HIP 1-degrader or a HIP 1 antagonist, keep that in mind as a potentially uninteresting explanation for what the rest of the body is showing you. Great, well, if there's no other questions, again, if we can, as best as possible, in this setting, thanks to Mr. Kaelin again for another outstanding sound arm. Thank you so much for joining. Thank you for the invite. Thank you, my pleasure.