William Kaelin, Peter Ratcliffe, and Gregg Semenza Receive the 2016 Albert Lasker Basic Medical Research Award

NEWS The Journal of Clinical Investigation William Kaelin, Peter Ratcliffe, and Gregg Semenza receive the 2016 Albert Lasker Basic Medical Research Award

The importance of adequate oxygen- development of therapies to treat diseas- the 1910 Anglo-American expedition to ation has been recognized for over 200 es such as anemia, cardiovascular dis- Pikes Peak to investigate the effects of years, but how cells and tissues are able ease, pulmonary hypertension, stroke, high altitude on breathing, Mabel Fitz- to monitor and respond to oxygen levels and cancer. gerald found that signs and symptoms of remained elusive until the late twentieth the response of humans to high altitude century. The 2016 Albert Lasker Basic Early evidence of an oxygen were set off even by small reductions in

Medical Research Award honors three sensor the partial pressure of oxygen (pO2) in scientists (Figure 1), William Kaelin, As anyone who has traveled at high alti- the arterial blood (2, 3). Peter Ratcliffe, and Gregg Semenza, for tudes will attest, low levels of atmo- Further support for an oxygen-sens- the discovery of the molecular mecha- spheric oxygen have marked effects on ing mechanism was provided by the nisms by which human and animal cells sense and respond to low or inadequate oxygen levels, referred to as hypoxia. Molecular oxygen is a critical substrate for cellular metabolism and bio- energetics, and cells within each tissue require an adequate oxygen supply to meet the needs of these pathways, as either oxygen deficiency or excess can lead to rapid death of both cells and whole organisms. While organisms with just a few cells can rely on passive oxygen diffusion, large multicellular organisms require multiple complex organ systems, including respiratory, circulatory, and neuroendocrine systems, to ensure Figure 1. The recipients of the 2016 Albert Lasker Basic Medical Research Award. From left to right: William Kaelin, Peter Ratcliffe, and Gregg Semenza discovered the essential pathway by which that all cells and tissues have reliable human and animal cells sense and adapt to changes in oxygen availability. Image credits (left to access to oxygen. In all cells and tissues, right): Sam Ogden/Dana-Farber Cancer Institute, Paul Wilkinson Photography, Jay VanRensselaer/ hypoxia initiates a series of physiological Johns Hopkins Medicine. responses that are geared toward main- taining oxygen homeostasis over a time course of minutes to days. These events physiology, ranging from relatively mild discovery of erythropoietin (EPO), a include upregulation of processes that symptoms, such as breathlessness and glycoprotein hormone that stimulates enhance oxygen delivery, such as eryth- dizziness, to severe symptoms, such as erythrocyte production. In adults, EPO ropoiesis, angiogenesis, and modulation pulmonary and cerebral edema. Conse- is produced in response to low oxygen of vascular tone, and downregulation of quently, some of the earliest evidence levels in the blood by interstitial fibro- oxygen consumption through changes in for an oxygen-sensing mechanism in blasts in the renal cortex. Human EPO cellular metabolism, proliferation, and animal cells came from scientists vis- was purified in 1977 (4), followed by apoptosis. Importantly, these adaptive iting high-altitude locales. In 1890, cloning of the human EPO gene in 1985 responses are dysregulated in a number François-Gilbert Viault noted that the (5, 6). These advances led to the develop- of disease states. Thus, the identification number of erythrocytes in his blood ment of recombinant human EPO for the of the pathways that sense and respond was elevated after a stay in the Peruvian treatment of anemia; however, the mech- to oxygen levels has not only furthered highlands (around 4,500 m above sea anisms underlying the regulation of EPO our understanding of multiple develop- level). He concluded that erythropoiesis by oxygen levels remained enigmatic. mental and physiological processes, but is stimulated when blood oxygen con- By the end of the 1980s, it was well has also opened up new avenues for the tent is reduced (1). Similarly, as part of established that the kidneys controlled the number of erythrocytes and thus the oxygen capacity of the blood through the Reference information: J Clin Invest. 2016;126(10):3628–3638. doi:10.1172/JCI90055. release of EPO (7). Further, it was known

3628 jci.org Volume 126 Number 10 October 2016 The Journal of Clinical Investigation NEWS

that EPO production had three critical During the same time period in which derived from different tissues, Ratcliffe regulatory properties: 1) EPO expression Semenza was developing EPO-transgen- demonstrated that oxygen sensing was is tissue specific; 2) EPO exhibits devel- ic mice, Peter Ratcliffe, a physician and possible in many different cell types, opmental stage specificity, being pro- kidney specialist, was establishing a even in those that did not produce EPO, duced prenatally in the fetal liver, then laboratory in Oxford University’s Nuff- suggesting that similar oxygen-sensing postnatally in the kidney; and 3) EPO ield Department of Medicine to study mechanisms were involved in the regula- expression is inducible, with expression the regulation of EPO, a venture that tion of other genes (21–23). increasing in response to hypoxia, ane- was to be strongly supported by Chris- The identification of a genetical- mia, or cobalt chloride exposure (8–12). topher Pugh, Patrick Maxwell, and other ly encoded hypoxia response element These properties attracted the attention kidney specialists training at Oxford at (HRE) and the demonstration that this of two researchers, Gregg Semenza and the time, who contributed much to the element could be activated in a variety Peter Ratcliffe, who would soon demon- laboratory’s work. As a kidney special- of tissues opened up two new avenues strate that EPO is not the only gene regu- ist, Ratcliffe was initially intrigued by of research: 1) the identification of the lated by hypoxia. the unusual counter-current circulation nuclear factors that regulate hypoxia-in- within the kidney that results in very low duced gene expression and 2) the iden- A genetic hypoxia response oxygen tensions within specific areas of tification of other genes that respond to element the organ. “Somewhat surprisingly, the hypoxia through similar mechanisms. In 1986, Gregg Semenza was a postdoc- kidney is able to distinguish changes in toral fellow in medical genetics at Johns blood oxygen content and the numbers A hypoxia-inducible Hopkins School of Medicine, working of red blood cells in circulation from transcription factor with research teams led by Haig Kaza- changes in renal blood flow. Consider- Semenza next focused on identifying zian and Stylianos Antonarakis, who ation of this remarkable oxygen-sensing the oxygen-regulated nuclear factors were renowned for their ability to dissect capacity was what initially led me into that bind EPO. In 1992, he and his post- disease-associated mutations in order to the field,” said Ratcliffe. By understand- doc, Guang Wang, identified the nuclear understand how these mutations affect- ing how oxygen regulated EPO produc- factor that bound the HRE, which they ed gene expression. “At the time, I was tion, it would be possible to get at the termed hypoxia-inducible factor-1 (HIF- interested in studying developmental oxygen-sensing mechanism in the kid- 1). Moreover, HIF-1 binding of the HRE regulation, and it was known that EPO ney. One of Ratcliffe’s early experiments was required for transcriptional acti- is expressed in the fetal liver and then in demonstrated that Epo mRNA levels in vation and was induced by hypoxia in a the adult kidneys. The original goal was isolated rat kidneys were responsive to variety of mammalian cell lines (24). “No to figure out what sequences were regu- changes in oxygen delivery, confirming matter which mammalian cell line we lating the expression in different organs that all of the components necessary looked at, if we exposed the cells to low during development,” Semenza recent- for Epo regulation were present in the oxygen for four hours, we saw the induc- ly recounted to the JCI. At the time, the kidney (17). At the time, there was evi- tion of HIF activity. So that was a pretty ability to create transgenic animals was dence that diseased kidneys and extra- clear indicator that HIF-1 was not just relatively new, but this technique had the renal tissues could produce EPO, but regulating the expression of EPO, but was advantage of allowing direct manipula- the extent and localization of EPO pro- regulating the expression of other genes tion of various aspects of gene expression, duction were unclear. Using very sen- under hypoxic conditions,” said Semen- making it an interesting model in which sitive assays, Ratcliffe and colleagues za. The DNA-binding activity of HIF-1 to study the regulation of EPO. In a col- showed that Epo mRNA was detectable decayed rapidly when cells were exposed laboration with John Gearhart, Semen- not only in the kidneys and livers of to increased oxygen levels, suggesting za engineered transgenic mice carrying rats, but also that small amounts of Epo that HIF-1 activity itself was directly reg- human EPO. The transgenic mice exhib- mRNA became detectable in the spleen, ulated by oxygen (25). It rapidly became ited increased erythropoiesis compared brain, and testes under hypoxic condi- clear that HIF was not restricted to mam- with their wild-type counterparts. (13). tions, indicating that hypoxia can induce malians cells, as the Ratcliffe group Semenza next began expressing EPO EPO expression outside of the kidney observed a similar activity in Drosophila constructs that contained additional 5′ or and suggesting that the oxygen-sensing cells around the same time (26). 3′ flanking regions to demonstrate that process operated more widely than had Purification of HIF-1 by the Semenza these regions conferred tissue-specific previously been considered (18, 19). In laboratory (27) revealed that it is a het- and hypoxia-inducible expression (14, 15). a parallel line of investigation testing erodimer consisting of an HIF-1α subunit Narrowing down the hypoxia-responsive hypoxia-inducible expression of trun- and an HIF-1β subunit (also known as region of the EPO gene, Semenza iden- cated forms of the mouse Epo gene, Rat- aryl hydrocarbon receptor nuclear trans- tified a region downstream of the EPO cliffe identified a region downstream of locator [ARNT] protein), both of which coding sequence that was bound by mul- the coding sequence that was required are basic helix-loop-helix proteins that tiple nuclear factors. The addition of this for oxygen-regulated expression (20). contain a Per-Arnt-Sim (PAS) domain. region to a different gene construct con- By coupling this region of murine Epo While HIF-1β expression is constitutive ferred hypoxia-inducible expression (16). to a broadly active promoter in cell lines and stable, HIF-1α protein is induced in

jci.org Volume 126 Number 10 October 2016 3629 NEWS The Journal of Clinical Investigation cells exposed to 1% oxygen and decays (Glut1), human aldolase A (ALDA), eno- that would be distinct from his post- rapidly when the cells are returned to lase 1 (ENO1), and murine phosphofruc- doctoral work but that could potential- normal oxygen levels (21% pO2). The tokinase (Pfkl), which were bound and ly take advantage of some of the same genes encoding HIF-1α were assigned to regulated by HIF-1 (43–46). Together, experimental approaches he had used murine chromosome 12 and human chro- these studies established HIF-1 as a successfully to study RB. He then read mosome 14 in 1996 (28). The mRNAs critical regulator of glycolysis, allowing a paper describing the identification of encoding HIF-1α and HIF-1β were pres- cells to switch from aerobic to anaerobic the gene that causes von Hippel–Lindau ent in all human, rat, and mouse organs metabolism (47). (VHL) disease (50), a familial cancer tested (29). A closely related gene, Angiogenesis is critical for establish- syndrome that predisposes affected HIF2A, was identified and cloned in ing a blood supply and is therefore also individuals to hemangioblastomas in 1997, followed by HIF3A in 1998 (30– a means by which tissues can increase the CNS and retina, renal cell carcino- 34). HIF-1α homologs were subsequently their oxygenation, making the process a ma, and pheochromocytoma, a benign found in other organisms, including Dro- likely HIF regulatory target. Ratcliffe’s tumor of the adrenal gland. From his sophila melanogaster and Caenorhabditis group showed that the expression of mul- clinical training, Kaelin knew that VHL elegans, indicating an evolutionary con- tiple angiogenic growth factors, includ- mutant–associated tumors are highly servation of the oxygen-sensing mecha- ing PDGFA/B, placental growth factor vascularized (angiogenic) and are nota- nism (26, 35–37). (PLGF), TGF-β1, and VEGF, was regulat- ble for their ability to secrete EPO. By The identification of the HIF pro- ed by hypoxia in a manner similar to that the early 1990s, there was considerable teins and cloning of the genes and seen with EPO (48). This was followed by interest in targeting angiogenesis to treat cDNAs allowed for an in-depth function- a study from the Semenza group demon- tumors, making VHL-associated tumors al analysis of the HIF subunits, providing strating that the VEGF gene was direct- a potentially useful system for the study hints as to how these genes and proteins ly transactivated by HIF-1 (39). Further, of angiogenesis-targeted therapeutic are regulated. Studies in HIF-1β–defi- Semenza and colleagues engineered approaches (51). In addition, he realized cient cells by the Ratcliffe group demon- a mouse lacking both copies of Hif1a, that a defect in oxygen sensing could be strated that this subunit was required for which resulted in developmental arrest the unifying factor underlying the induc- hypoxia-induced gene expression (38). and lethality at mid-gestation (E11). tion of angiogenesis and EPO by the Semenza’s group showed that expression The HIF-1α–deficient embryos exhibit- VHL-mutant tumors, with the tumors of a dominant-negative form of HIF-1α ed multiple malformations of the heart behaving as though they were starved of blocked the transcriptional response to and blood vessels and decreased eryth- oxygen. “These tumors are constantly hypoxia (39). Semenza’s group also iden- ropoiesis (49). “Initially, the embryo is putting out the stress signals that would tified the stretch of amino acids required very small, and it can get all of its oxy- be expected under hypoxic conditions. for dimerization of the HIF-1α and -1β gen just from diffusion from maternal We thought that studying the VHL pro- subunits and DNA binding, as well as blood vessels,” said Semenza. “At some tein could provide some insight into how N-terminal and C-terminal transacti- point, it gets large enough that it has to cells respond to hypoxia,” said Kaelin. vation domains in HIF-1α that interact have its own functioning circulatory sys- Kaelin first demonstrated that VHL with transcriptional coactivators such tem. When we looked at the expression encodes a functional tumor suppressor as CREB-binding protein (CBP), while of HIF-1α in the embryo, we saw that at (pVHL) (52), as reintroduction of wild- both the Ratcliffe and Semenza groups mid-gestation, the levels of HIF-1α went type, but not mutant, VHL in a renal identified a regulatory domain that pre- up in the wild-type embryos, and that’s cell carcinoma line prevented the cells vents transcriptional activity in normox- when the mutant embryos died.” These from forming tumors in mice. Further, ic conditions (40–42). These regulatory studies confirmed the role of oxygen his work, together with findings from domains would soon provide insights homeostasis and HIF-1 in development. Richard Klausner, indicated that pVHL- into the role of oxygen in the regulation Identification of angiogenesis as a tar- mediated tumor suppression required of HIF activity. get of HIF-1 would quickly provide a link that pVHL bind two proteins, elongins between HIF and another highly angio- B and C, originally identified as regula- Identification of hypoxia- genic process, tumorigenesis. tors of transcriptional elongation, but regulated genes which are now known to wear multiple Following the identification of the HRE The cancer connection: hats (53–55). The interactions between and its operation in cell types that do von Hippel–Lindau disease elongins B and C were disrupted by dis- not produce EPO, Ratcliffe and Semen- In 1993, William Kaelin, an oncologist, ease-associated VHL mutations (53, 54, za began hunting for other genes that had just established his own laboratory 56, 57). Importantly, Kaelin showed that exhibited hypoxia-induced expression. after completing a postdoctoral fellow- various hypoxia-inducible mRNAs, such Both groups identified a number of genes ship focused on determining the func- as VEGF, GLUT1, and PDGFB, which had involved in cellular metabolism, includ- tions of the retinoblastoma (RB) tumor recently been shown to be regulated by ing human phosphoglycerate kinase-1 suppressor in the laboratory of David HIF-1 (48, 58, 59), were insensitive to (PGK1), mouse lactate dehydrogenase-A Livingston at the Dana-Farber Cancer oxygen in renal cancer cell lines lacking (Ldha), mouse glucose transporter-1 Institute. He was looking for a project pVHL and, consequently, were over-

3630 jci.org Volume 126 Number 10 October 2016 The Journal of Clinical Investigation NEWS

produced (60). In short, loss of pVHL dation under normoxic conditions (66). normoxia or hypoxia, but that these com- uncoupled the accumulation of hypox- Shortly thereafter, Frank Bunn and col- plexes were disrupted by treatment with ia-inducible mRNAs from actual oxygen leagues showed that HIF-1α contains an Co2+ or the iron chelator desferrioxam- availability. oxygen-dependent degradation domain ine. This observation initially led them The connection between HIF and (ODD) consisting of approximately 200 to believe (as it turns out, erroneously — cancer was also emerging. Using hepa- amino acid residues that makes the pro- see below) that an iron-requiring protein toma-derived tumor xenografts, the Rat- tein unstable in the presence of oxy- was essential for complex formation and cliffe group demonstrated that VEGF and gen and allows it to be degraded via the could potentially be involved in oxygen GLUT1 were induced in hypoxic regions ubiquitin/proteasome pathway (67). The sensing through a subsequent action. surrounding areas of tumor necrosis in Semenza group then engineered missense Subsequently, Kaelin’s group showed an HIF-1β–dependent manner. Notably, mutations and deletions within HIF-1α that the pVHL-elongin B/C-CUL2 com- HIF-1β–deficient xenografts exhibited that blocked ubiquitination, resulting in plex binds directly to the HIF-1 ODD slower growth and reduced vasculariza- constitutive expression and transcription- domain and polyubiquitinates HIF-1α tion compared with HIF-1β–expressing al activity in normoxic conditions (68). (75). Similar conclusions were reached xenografts (61). Around the same time, Structural analyses of the pVHL in parallel by the Ratcliffe, Poellinger, the Semenza group demonstrated that complex provided key support for a role and Conaway laboratories and were pub- the level of HIF-1 expression was cor- in HIF-1α degradation. Kaelin’s group lished soon thereafter (76–78). related with tumor growth in murine found that pVHL-mediated regulation xenografts (62). Further, Semenza’s of hypoxia-inducible mRNAs required An oxygen-dependent group found that HIF-1α was overex- binding of pVHL to protein complexes signaling mechanism pressed in multiple human cancers under containing elongins B and C and cul- The studies described above established both normoxic and hypoxic conditions lin 2 (CUL2) (69), a protein that was that pVHL targets HIFα for ubiquitina- (63), while Ratcliffe and his collabora- suspected to be involved in targeting tion and subsequent proteasomal deg- tors found that HIF-1α and HIF-2α were proteins for ubiquitin-dependent pro- radation, but it was still unclear exactly expressed in a number of human can- teolysis (70). Structural studies of the how oxygen regulated this process. The cers and tumor-associated macrophages pVHL-elongin C-elongin B complex by Semenza group had shown that HIF was (64). Together, these studies suggested a Kaelin’s collaborator Nikola Pavletich activated by Co2+ and iron chelators, as broader role for HIFs in oncogenesis, as revealed that it is structurally similar well as by hypoxia; these same factors well as a potential connection to pVHL. to SKP1-CUL1-F-box protein (SCF) E3 induced the transactivation domain ubiquitin ligases (57). This structure also function of HIF (42, 79). These findings Linking VHL and HIF showed that pVHL has two hotspots for suggested the existence of an oxygen- Differences in HIFα regulation at the VHL disease–associated mutations: the and iron-dependent modification gov- protein and mRNA levels pointed to α-domain, which Kaelin had shown is erning the interaction of the HIFα ODD potential regulatory mechanisms. Lorenz required for binding to the elongins and with pVHL. In April 2001, the Kaelin and Poellinger and colleagues found that CUL2, and the β-domain, which was pre- Ratcliffe groups published independent HIF1A and HIF1B mRNAs are constitu- dicted to be a substrate docking site (57). back-to-back studies identifying proline tively expressed under both normoxic and Two additional observations increased hydroxylation as the crucial oxygen- and hypoxic conditions in transformed cell the suspicion that the pVHL complex iron-dependent posttranslational mod- lines; in contrast, HIFα protein levels are was an E3 ubiquitin ligase. First, Kaelin’s ification of HIFα that was required for highly sensitive to changes in oxygen lev- collaborator Joan Conaway showed that recognition by the pVHL complex (80, els, while HIF-1β protein levels are stable the pVHL complex associates with ring- 81); the same mechanism was identified (65). These findings suggested that some box 1 (RBX1), a known ubiquitin-conju- by Frank Lee’s group a few months lat- form of posttranslational modification gating enzyme (71). Second, two groups er (82). Ratcliffe had initially been sur- was responsible for regulating HIF-1α lev- of researchers led by Richard Klausner prised when, as described above, they els. The Ratcliffe laboratory identified dif- and Wilhelm Krek showed that pVHL could isolate the HIFα-pVHL complex ferent domains of HIF-1α that conferred immunoprecipitated from cells displays from hypoxic cells, and postulated that oxygen-regulated activity, distinguishing E3 ubiquitin ligase activity (72, 73). reoxygenation of cell lysates during the domains that altered protein levels from Soon thereafter, the Ratcliffe group experimental procedure might promote others that did not. They also showed that demonstrated that cells lacking pVHL this interaction. “We made a slight mis- oxygen-regulated activity persisted in one cannot target HIF-1α for oxygen-de- take at the time, which is an interest- domain, even when all phospho-accep- pendent proteolysis in vivo (74). HIFα ing one,” said Ratcliffe. “We initially tor amino acids were mutated, directing subunits were constitutively and stably thought that the association between attention away from protein phosphoryla- expressed in VHL-defective cells, but VHL and HIF was regulated by cobalt tion as the signal transduction mechanism reexpression of wild-type pVHL restored and iron chelators, which matched the (41). A study by Jaime Caro demonstrated oxygen-dependent degradation of HIFα. properties of the system, but not by oxy- that HIF-1α undergoes rapid ubiquitina- They also found that HIFα subunits and gen. We were extremely puzzled by that, tion and subsequent proteasomal degra- pVHL formed stable complexes in either as normally those properties were all

jci.org Volume 126 Number 10 October 2016 3631 NEWS The Journal of Clinical Investigation

Figure 2. Regulation of hypoxia-inducible factors. Under normoxic conditions, two proline residues on the HIFα subunit are hydroxylated (OH) by PHD 2+ enzymes (PHD1, -2, and -3), in the presence of O2, Fe , 2-OG, and ascorbate (not shown). Hydroxylated HIFα is recognized by the pVHL E3 ubiquitin ligase complex, which tags HIFα with polyubiquitin, allowing for proteasomal recognition and subsequent degradation. Additionally, the 2-OG dioxygenase FIH-1 hydroxylates an asparagine residue in the C-terminal transactivation domain of HIFα, preventing its interaction with transcriptional coactivators. Under hypoxic conditions, HIFα prolyl hydroxylation is inhibited, preventing recognition of HIFα by pVHL. HIFα can then accumulate and translocate to the nucle- us, where it dimerizes with HIF-1β. The HIF dimer binds to HREs within the promoters of target genes and recruits transcriptional coactivators such as CBP to induce transcription. Asn, asparagine; E2, ubiquitin-conjugating enzyme; Pro, proline; Ub, ubiquitin.

concordant. On reflection, we realized showed that pVHL could bind direct- ylation site, Pro402, which mediates the that oxygen would be in the buffers and ly to HIFα unless the HIFα was derived interaction of HIFα with pVHL, was all the reagents that we used to do the from cells exposed to hypoxia or iron identified by the Ratcliffe group a few work on the bench. So it was possible chelators. Additionally, both groups months later (85). These data revealed that although the cells were made hypox- demonstrated that a cellular factor was that prolyl-4-hydroxylase activity was ic, by the time we’d gotten the extracts, required for the posttranslational mod- necessary to promote the interaction oxygen had gotten in, which turned out ification of HIFα, as pVHL only recog- between HIFα and pVHL. to be correct.” nized recombinant HIF after incubation The well-studied prolyl hydroxy- By repeating these experiments in a with vertebrate cell lysates. A system- lases at that time, which modify colla- hypoxia workstation with deoxygenated atic mutation analysis of the minimal gen, required Fe2+, ascorbate, molecular buffers, two highly skilled postdocs in pVHL-binding domain of HIFα zeroed oxygen, and 2-oxoglutarate (2-OG, also the Ratcliffe laboratory, Panu Jaakkola in on a proline residue, Pro564, which was known as α-ketoglutarate) (86). These and David Mole, showed that the associ- conserved in human, Xenopus, Drosoph- enzymes split molecular oxygen, using ation between HIFα subunits and pVHL ila, and C. elegans HIFα proteins. Mass one atom to add a hydroxyl group to a was governed by both oxygen and iron spectrometric analysis confirmed that target protein and the other to react with availability. Kaelin had reached the same Pro564 was hydroxylated. Importantly, 2-OG, resulting in the generation of suc- 564 conclusion by multiple means, including a hydroxyproline substitution at Pro cinate and CO2. Ratcliffe and colleagues experiments with mouse cells express- promoted the interaction of HIF-1α with determined that the prolyl hydroxy- ing a temperature-sensitive mutant of pVHL. Follow-up studies by the Kaelin lase activity mediating HIFα hydrox- the E1 ubiquitin–activating enzyme that, and Ratcliffe groups showed that the ylation also required these cofactors when grown at nonpermissive tempera- hydroxyproline inserts into a gap within and that HIF was induced by the 2-OG tures, accumulated HIFα under both the pVHL hydrophobic core within a site analog dimethyloxalylglycine (DMOG). normoxic and hypoxic conditions. Using that is a hotspot for tumorigenic muta- Through a combination of structurally so-called “far Western blots,” his group tions (83, 84). A second prolyl hydrox- informed prediction and candidate test-

3632 jci.org Volume 126 Number 10 October 2016 The Journal of Clinical Investigation NEWS ing, Ratcliffe and his collaborator Chris- A second oxygen-mediated asparagine residue, thereby inactivating topher Schofield identified the dioxy- HIF regulatory pathway one of its two transactivation domains genase EGL-9 as the enzyme responsible By 2001, it was clear that HIF signaling (101–103). for hydroxylating the HIFα ortholog in can be induced in any cell type under The identification of the prolyl and C. elegans. Further, they identified a set hypoxic conditions and that it plays a asparginyl hydroxylation events revealed of mammalian HIF prolyl hydroxylases critical role in the response to hypox- a dual regulatory system of HIF activity (PHD1, -2, and -3; also known as EglN2, ia through transcriptional activation (Figure 2) involving two oxygen-depen- EglN1, and EglN3, respectively) that of genes encoding proteins that either dent hydroxylases: a PHD (most com- mediate hydroxylation of human HIFα. increase oxygen availability or mediate monly PHD2) and FIH-1 (104). Both Moreover, the activity of these enzymes adaptive responses to intracellular oxy- HIFα and HIFβ are constitutively pro- was modulated by graded hypoxia, iron gen deprivation. Upon the return of nor- duced, but HIFα is only stable and active chelation, and Co2 , as well as the 2-OG mal oxygen levels, oxygen-dependent under hypoxic conditions (around 1%

analog DMOG, mirroring the in vivo binding of pVHL mediates the destruc- pO2). As oxygen levels rise, FIH-1, which characteristics of HIF regulation (87). tion of the HIFα subunit, terminating has activity at lower oxygen levels than Rick Bruick and Steve McKnight iden- the response to hypoxia. Additionally, does PHD2 (105, 106), becomes active tified these same three enzymes, after the C-terminal domains of HIFα sub- and hydroxylates an asparagine residue similar studies that began with the Dro- units were known to contain regulatory in the C-terminal transactivation domain sophila ortholog (88). domains (41, 42), which bind transcrip- of HIFα, preventing interaction with Prolyl hydroxylation has a profound tional coactivators, including CBP, p300, transcriptional coactivators and thereby effect on HIF stability and activity, mak- steroid receptor coactivator-1 (SRC1), partially abrogating HIFα transcription- ing the PHDs potential targets for phar- and transcriptional intermediary factor-1 al activity (as FIH-1 does not inhibit the macological mimicking of the effects (TIF1) (96–98). Ratcliffe, Semenza, and N-terminal transactivation domain). of hypoxia. The Kaelin group had also others had shown that the C-terminal As oxygen levels continue to increase, been working to identify HIFα PHDs transactivation domains were regulated PHD2 becomes active and hydroxylates and, using biochemical approaches, by oxygen, but this regulation influenced one (or both) of two prolines on HIFα. identified PHD2, which has emerged transcriptional activity independently of Hydroxylation of either proline pro- as the workhorse member of the family protein stability (41, 42). motes the interaction of HIFα with the (89). In collaboration with Joan and Ron In order to understand the regu- pVHL E3 ubiquitin ligase complex (107), Conaway, Kaelin’s group purified PHD2 lation of the transactivation domains, resulting in ubiquitination and subse- and demonstrated that some small mol- the Semenza group conducted a yeast quent proteasomal degradation. Thus, ecules designed to inhibit the related two-hybrid screen to identify proteins these hydroxylation events allow cells to collagen prolyl hydroxylases, includ- that interact with HIF-1 to modulate its tightly control responses to alterations ing selected iron chelators and 2-OG biological activity. They identified and in oxygen levels, only allowing for HIF antagonists, also inhibited the activity characterized a protein that they named accumulation and transcriptional activi- of PHD2 (89). Moreover, treatment of factor inhibiting HIF-1 (FIH-1), which ty under the appropriate environmental cultured cells with these inhibitors sta- negatively regulates the function of the conditions (108). bilized HIF-1α, increased expression of HIF C-terminal transactivation domain VEGF (89), and induced EPO in mice, (99). They also found that FIH-1 binds An expanding role for HIF in including mice made anemic by par- to pVHL and that pVHL can function as physiology and disease tial nephrectomy (90). In collaboration a transcriptional corepressor that inhib- The discovery of the HIF pathway not with Christopher Schofield, the Ratcliffe its HIF-1α transactivation function by only unveiled a new signaling mechanism group also developed and tested differ- recruiting histone deacetylases, thereby mediated by oxygen, but also demon- ent analogs of the PHD cofactor 2-OG, closing down the chromatin. strated that every cell in the body is capa- which stabilized HIFα (91). These stud- In 2002, Murray Whitelaw and col- ble of sensing and responding to oxygen ies established PHDs as bona fide phar- leagues showed that an asparagine res- levels. Kaelin, Ratcliffe, and Semenza macological targets for HIF regulation. idue in the C-terminal transactivation have all worked to identify new roles for Given the growing role of HIF signaling domain of HIF was hydroxylated under the pathway and to delineate the mech- in human disease, many such inhibitors normoxia, but that the modification anisms by which HIF signaling is regu- have been developed in the past decade was not present under hypoxia. Aspar- lated in a given context. At the cellular (92). Notably, Josef Prchal, Frank Lee, agine hydroxylation was also prevented level, HIF signaling has a profound effect and others have linked genetic variants by iron chelators or inhibitors of 2-OG– on metabolism, allowing cells to switch of PHD2, HIF2α, and VHL to familial dependent dioxygenases (100). Shortly from the oxygen-consuming TCA cycle polycythemia and high-altitude adap- thereafter, both the collaborating Scho- to glycolysis (109). Additionally, HIF sig- tation. Therefore, there is genetic vali- field and Ratcliffe laboratories and the naling contributes to cell fate decisions, dation in humans as well as in multiple Whitelaw group showed that FIH-1 is including differentiation, senescence, model organisms that ties prolyl hydrox- an iron- and 2-OG–dependent dioxy- and apoptosis (110–115). At the tissue lev- ylation to oxygen sensing (93–95). genase that hydroxylates HIF-1α on an el, HIF signaling is involved in the devel-

jci.org Volume 126 Number 10 October 2016 3633 NEWS The Journal of Clinical Investigation opment and maintenance of numerous targeting the HIF-signaling pathway are to ask big questions. When you say the organs and tissues, including those in under development, with several ther- question out loud, it should sound a little the cardiovascular, skeletal, and immune apeutic modalities advancing to late- audacious.” These three physician-sci- systems (116–118). Further, HIF has been stage clinical trials. Since PHD inhibitors entists became interested in pursuing shown to play a critical role in mucosal stabilize HIFα, they cause increased HIF one of the most audacious of questions: barrier functions and inflammation (119, signaling. Such drugs could potentially what are the mechanisms that underlie 120). “It really has become a situation be used in disease states such as anemia, oxygen sensing in animals? The answer where I assume that HIF is involved in in which HIF signaling is beneficial, or to to that question has expanded our under- a given process until proven otherwise, protect against ischemic injury in CAD standing of many different aspects of because in so many contexts this pathway and PAD (158). Conversely, in diseases in biology, ranging from metazoan evolu- is important,” said Semenza. which HIF signaling is detrimental, HIFα tion to cancer biology, and underscores HIF-signaling pathways have been inhibitors may be beneficial. For exam- the importance of research focused on implicated in a variety of diseases states, ple, a direct HIF inhibitor is currently in the most basic of scientific questions. with HIF playing a beneficial or a detri- phase II clinical trials for kidney cancer mental role, depending on the context. (159), while drugs that indirectly down- Jillian H. Hurst HIF signaling has been shown to medi- regulate HIF-1α, including the cardiac ate protective responses in diseases glycoside digoxin (160), are currently in 1. Viault FG. Sur l’augmentation considérable du characterized by impaired tissue oxy- clinical trials for various forms of cancer. nombre des globules rouges dans le sang chez genation and inflammation, such as cor- les habitants des hauts plateaux de l’Amérique du Sud. C R Acad Sci Paris. 1890;111:917–918. onary artery disease (CAD) (121–124), Coda 2. FitzGerald MP. The Changes in the Breathing peripheral arterial disease (PAD) (125– Each of these researchers approached and the Blood at Various High Altitudes. 127), wound healing (128–130), organ this essential biological question in a Philos Trans R Soc Lond Ser B Contain Pap Biol. transplant rejection (131, 132), and coli- slightly different way, based on their 1913;203:351–371. http://www.jstor.org/stable/92005. Accessed August, 23, 2016. tis (133–135). In contrast, HIF signaling area of specialty: Kaelin from oncology, 3. FitzGerald MP. Further Observations on the might be maladaptive in other disease Ratcliffe from nephrology, and Semenza Changes in the Breathing and the Blood at Vari- states, including hereditary erythrocy- from medical genetics. Since their ini- ous High Altitudes. Proc R Soc Lond Ser B Contain tosis (136), pulmonary arterial hyper- tial discoveries, they have all continued Pap Biol. 1914;88(602):248–258. http://www.jstor. tension (137–139), chronic ischemic car- to examine the mechanisms by which org/stable/80707. Accessed August 23, 2016. diomyopathy (140, 141), and obstructive oxygen sensing impacts human physiol- 4. Miyake T, Kung CK, Goldwasser E. Purifica- tion of human erythropoietin. J Biol Chem. sleep apnea (142, 143). ogy and disease. William Kaelin is cur- 1977;252(15):5558–5564. HIF signaling plays complex role in rently a professor in the Department of 5. Jacobs K, et al. Isolation and characterization cancer (144–146). Hypoxia and expres- Medicine and associate director of Basic of genomic and cDNA clones of human eryth- sion of HIF in tumors are associated with Science at the Dana-Farber Institute, ropoietin. Nature. 1985;313(6005):806–810. poor prognosis and have been shown to Harvard Medical School, where he and 6. Lin FK, et al. Cloning and expression of the human erythropoietin gene. Proc Natl Acad Sci promote tumor angiogenesis, epithe- his team study the function of specific U S A. 1985;82(22):7580–7584. lial-to-mesenchymal transition, stem tumor-suppressor genes, including VHL. 7. Bauer C, Kurtz A. Oxygen sensing in the kidney cell maintenance, invasion and metas- Peter Ratcliffe is the clinical research and its relation to erythropoietin production. tasis, therapy resistance, and induction director at the Francis Crick Institute, Annu Rev Physiol. 1989;51:845–856. of metabolic alterations (147–149). The director of the Target Discovery Institute 8. Beru N, McDonald J, Lacombe C, Goldwasser E. Expression of the erythropoietin gene. role of HIF in cancer has best been illus- and the Oxford Hypoxia Biology Labo- Mol Cell Biol. 1986;6(7):2571–2575. trated in the context of pVHL-defective ratory at the University of Oxford, and 9. Bondurant MC, Koury MJ. Anemia induc- kidney cancers. In this setting, HIF-2, a member of the Ludwig Institute for es accumulation of erythropoietin mRNA rather than its better-studied cousin Cancer Research. His current research in the kidney and liver. Mol Cell Biol. HIF-1, appears to the be the main culprit explores mechanisms of oxygen sensing 1986;6(7):2731–2733. 10. Koury ST, Bondurant MC, Koury MJ, Semenza (150–157). The identification of HIF-2 as mediated by 2-OG oxygenases, as well as GL. Localization of cells producing erythro- a driving force in kidney cancer helped the oncogenic mechanisms mediated by poietin in murine liver by in situ hybridization. to motivate and accelerate the success- HIF signaling. Gregg Semenza is the C. Blood. 1991;77(11):2497–2503. ful development of drugs that inhibit Michael Armstrong Professor of Genet- 11. Koury MJ, Bondurant MC, Graber SE, Sawyer the HIF-responsive growth factor VEGF ic Medicine and director of the vascular ST. Erythropoietin messenger RNA levels in developing mice and transfer of 125I-eryth- for the treatment of this disease (145). program at the Institute for Cell Engi- ropoietin by the placenta. J Clin Invest. Delineation of the pathways and factors neering at the Johns Hopkins University 1988;82(1):154–159. that interact with HIF in a specific dis- School of Medicine, where his research 12. Schuster SJ, Wilson JH, Erslev AJ, Caro J. Phys- ease context has and will continue to group investigates oxygen homeostasis iologic regulation and tissue localization of help identify therapeutic strategies cen- and HIF signaling in ischemic cardiovas- renal erythropoietin messenger RNA. Blood. 1987;70(1):316–318. tered on HIF signaling. cular disease and cancer. 13. Semenza GL, Traystman MD, Gearhart JD, Given its role in so many different As Kaelin told the JCI, “in order to Antonarakis SE. Polycythemia in trans- disease states, a number of therapies make big advances in science, you need genic mice expressing the human eryth-

3634 jci.org Volume 126 Number 10 October 2016 The Journal of Clinical Investigation NEWS

ropoietin gene. Proc Natl Acad Sci U S A. 28. Semenza GL, Rue EA, Iyer NV, Pang MG, domains within the alpha subunit. J Biol Chem. 1989;86(7):2301–2305. Kearns WG. Assignment of the hypoxia- 1997;272(17):11205–11214. 14. Semenza GL, Dureza RC, Traystman MD, inducible factor 1alpha gene to a region of 42. Jiang BH, Zheng JZ, Leung SW, Roe R, Semen- Gearhart JD, Antonarakis SE. Human eryth- conserved synteny on mouse chromosome za GL. Transactivation and inhibitory domains ropoietin gene expression in transgenic mice: 12 and human chromosome 14q. Genomics. of hypoxia-inducible factor 1alpha. Modula- multiple transcription initiation sites and 1996;34(3):437–439. tion of transcriptional activity by oxygen ten- cis-acting regulatory elements. Mol Cell Biol. 29. Wiener CM, Booth G, Semenza GL. In vivo sion. J Biol Chem. 1997;272(31):19253–19260. 1990;10(3):930–938. expression of mRNAs encoding hypoxia- 43. Firth JD, Ebert BL, Pugh CW, Ratcliffe PJ. 15. Semenza GL, Koury ST, Nejfelt MK, Gearhart inducible factor 1. Biochem Biophys Res Oxygen-regulated control elements in the JD, Antonarakis SE. Cell-type-specific and Commun. 1996;225(2):485–488. phosphoglycerate kinase 1 and lactate dehy- hypoxia-inducible expression of the human 30. Ema M, Taya S, Yokotani N, Sogawa K, Mat- drogenase A genes: similarities with the eryth- erythropoietin gene in transgenic mice. Proc suda Y, Fujii-Kuriyama Y. A novel bHLH-PAS ropoietin 3’ enhancer. Proc Natl Acad Sci U S A. Natl Acad Sci U S A. 1991;88(19):8725–8729. factor with close sequence similarity to 1994;91(14):6496–6500. 16. Semenza GL, Nejfelt MK, Chi SM, Antonarakis hypoxia-inducible factor 1alpha regulates the 44. Ebert BL, Firth JD, Ratcliffe PJ. Hypoxia SE. Hypoxia-inducible nuclear factors bind to VEGF expression and is potentially involved and mitochondrial inhibitors regulate an enhancer element located 3’ to the human in lung and vascular development. Proc Natl expression of glucose transporter-1 via erythropoietin gene. Proc Natl Acad Sci U S A. Acad Sci U S A. 1997;94(9):4273–4278. distinct Cis-acting sequences. J Biol Chem. 1991;88(13):5680–5684. 31. Flamme I, Fröhlich T, von Reutern M, Kappel 1995;270(49):29083–29089. 17. Ratcliffe PJ, Jones RW, Phillips RE, Nicholls A, Damert A, Risau W. HRF, a putative basic 45. Semenza GL, Roth PH, Fang HM, Wang GL. LG, Bell JI. Oxygen-dependent modulation of helix-loop-helix-PAS-domain transcription Transcriptional regulation of genes encoding erythropoietin mRNA levels in isolated rat kid- factor is closely related to hypoxia-inducible glycolytic enzymes by hypoxia-inducible fac- neys studied by RNase protection. J Exp Med. factor-1 alpha and developmentally expressed tor 1. J Biol Chem. 1994;269(38):23757–23763. 1990;172(2):657–660. in blood vessels. Mech Dev. 1997;63(1):51–60. 46. Semenza GL, et al. Hypoxia response elements 18. Tan CC, Eckardt KU, Ratcliffe PJ. Organ 32. Hogenesch JB, et al. Characterization of a in the aldolase A, enolase 1, and lactate dehy- distribution of erythropoietin messenger subset of the basic-helix-loop-helix-PAS drogenase A gene promoters contain essential RNA in normal and uremic rats. Kidney Int. superfamily that interacts with components binding sites for hypoxia-inducible factor 1. 1991;40(1):69–76. of the dioxin signaling pathway. J Biol Chem. J Biol Chem. 1996;271(51):32529–32537. 19. Tan CC, Eckardt KU, Firth JD, Ratcliffe PJ. 1997;272(13):8581–8593. 47. Ebert BL, Gleadle JM, O’Rourke JF, Bartlett Feedback modulation of renal and hepatic 33. Tian H, McKnight SL, Russell DW. Endothelial SM, Poulton J, Ratcliffe PJ. Isoenzyme-specific erythropoietin mRNA in response to graded PAS domain protein 1 (EPAS1), a transcription regulation of genes involved in energy metab- anemia and hypoxia. Am J Physiol. 1992; factor selectively expressed in endothelial olism by hypoxia: similarities with the regula- 263(3 Pt 2):F474–F481. cells. Genes Dev. 1997;11(1):72–82. tion of erythropoietin. Biochem J. 1996; 20. Pugh CW, Tan CC, Jones RW, Ratcliffe PJ. 34. Gu YZ, Moran SM, Hogenesch JB, Wartman L, 313(Pt 3):809–814. Functional analysis of an oxygen-regulated Bradfield CA. Molecular characterization and 48. Gleadle JM, Ebert BL, Firth JD, Ratcliffe transcriptional enhancer lying 3’ to the mouse chromosomal localization of a third alpha- PJ. Regulation of angiogenic growth factor erythropoietin gene. Proc Natl Acad Sci U S A. class hypoxia inducible factor subunit, HIF3 expression by hypoxia, transition metals, and 1991;88(23):10553–10557. alpha. Gene Expr. 1998;7(3):205–213. chelating agents. Am J Physiol. 1995; 21. Maxwell PH, Pugh CW, Ratcliffe PJ. Inducible 35. Iyer NV, Leung SW, Semenza GL. The human 268(6 Pt 1):C1362–C1368. operation of the erythropoietin 3’ enhancer in hypoxia-inducible factor 1alpha gene: HIF1A 49. Iyer NV, et al. Cellular and developmental con- multiple cell lines: evidence for a widespread structure and evolutionary conservation. trol of O2 homeostasis by hypoxia-inducible oxygen-sensing mechanism. Proc Natl Acad Genomics. 1998;52(2):159–165. factor 1 alpha. Genes Dev. 1998;12(2):149–162. Sci U S A. 1993;90(6):2423–2427. 36. Bacon NC, et al. Regulation of the Drosophila 50. Latif F, et al. Identification of the von Hip- 22. Pugh CW, Ebert BL, Ebrahim O, Ratcliffe bHLH-PAS protein Sima by hypoxia: function- pel-Lindau disease tumor suppressor gene. PJ. Characterisation of functional domains al evidence for homology with mammalian Science. 1993;260(5112):1317–1320. within the mouse erythropoietin 3’ enhancer HIF-1 alpha. Biochem Biophys Res Commun. 51. Folkman J. The role of angiogenesis in tumor conveying oxygen-regulated responses in 1998;249(3):811–816. growth. Semin Cancer Biol. 1992;3(2):65–71. different cell lines. Biochim Biophys Acta. 37. Jiang H, Guo R, Powell-Coffman JA. The 52. Iliopoulos O, Kibel A, Gray S, Kaelin WG. 1994;1217(3):297–306. Caenorhabditis elegans hif-1 gene encodes a Tumour suppression by the human von 23. Bauer C, Koch KM, Scigalla P, Wieczorek L, bHLH-PAS protein that is required for adap- Hippel-Lindau gene product. Nat Med. eds. Erythropoietin: molecular physiology and tation to hypoxia. Proc Natl Acad Sci U S A. 1995;1(8):822–826. clinical applications. New York, New York, 2001;98(14):7916–7921. 53. Kibel A, Iliopoulos O, DeCaprio JA, Kaelin USA: Marcel Dekker; 1993. 38. Wood SM, Gleadle JM, Pugh CW, Hankinson WG. Binding of the von Hippel-Lindau tumor 24. Wang GL, Semenza GL. General involvement O, Ratcliffe PJ. The role of the aryl hydrocar- suppressor protein to Elongin B and C. Science. of hypoxia-inducible factor 1 in transcriptional bon receptor nuclear translocator (ARNT) 1995;269(5229):1444–1446. response to hypoxia. Proc Natl Acad Sci U S A. in hypoxic induction of gene expression. 54. Duan DR, et al. Characterization of the VHL 1993;90(9):4304–4308. Studies in ARNT-deficient cells. J Biol Chem. tumor suppressor gene product: localization, 25. Wang GL, Semenza GL. Characterization of 1996;271(25):15117–15123. complex formation, and the effect of natural hypoxia-inducible factor 1 and regulation of 39. Forsythe JA, et al. Activation of vascular inactivating mutations. Proc Natl Acad Sci U S A. DNA binding activity by hypoxia. J Biol Chem. endothelial growth factor gene transcription 1995;92(14):6459–6463. 1993;268(29):21513–21518. by hypoxia-inducible factor 1. Mol Cell Biol. 55. Duan DR, et al. Inhibition of transcription 26. Nagao M, Ebert BL, Ratcliffe PJ, Pugh CW. 1996;16(9):4604–4613. elongation by the VHL tumor suppressor pro- Drosophila melanogaster SL2 cells contain a 40. Jiang BH, Rue E, Wang GL, Roe R, Semenza tein. Science. 1995;269(5229):1402–1406. hypoxically inducible DNA binding complex GL. Dimerization, DNA binding, and transacti- 56. Ohh M, et al. Synthetic peptides define critical which recognises mammalian HIF-binding vation properties of hypoxia-inducible factor 1. contacts between elongin C, elongin B, and sites. FEBS Lett. 1996;387(2-3):161–166. J Biol Chem. 1996;271(30):17771–17778. the von Hippel-Lindau protein. J Clin Invest. 27. Wang GL, Semenza GL. Purification and char- 41. Pugh CW, O’Rourke JF, Nagao M, Gleadle 1999;104(11):1583–1591. acterization of hypoxia-inducible factor 1. JM, Ratcliffe PJ. Activation of hypoxia- 57. Stebbins CE, Kaelin WG, Pavletich NP. Struc- J Biol Chem. 1995;270(3):1230–1237. inducible factor-1; definition of regulatory ture of the VHL-ElonginC-ElonginB complex:

jci.org Volume 126 Number 10 October 2016 3635 NEWS The Journal of Clinical Investigation

implications for VHL tumor suppressor func- 72. Iwai K, et al. Identification of the von Hip- Cell. 2001;107(1):43–54. tion. Science. 1999;284(5413):455–461. pel-lindau tumor-suppressor protein as part 88. Bruick RK, McKnight SL. A conserved family 58. Levy AP, Levy NS, Wegner S, Goldberg MA. of an active E3 ubiquitin ligase complex. Proc of prolyl-4-hydroxylases that modify HIF. Transcriptional regulation of the rat vascular Natl Acad Sci U S A. 1999;96(22):12436–12441. Science. 2001;294(5545):1337–1340. endothelial growth factor gene by hypoxia. 73. Lisztwan J, Imbert G, Wirbelauer C, Gstaiger 89. Ivan M, et al. Biochemical purification and J Biol Chem. 1995;270(22):13333–13340. M, Krek W. The von Hippel-Lindau tumor pharmacological inhibition of a mamma- 59. Liu Y, Cox SR, Morita T, Kourembanas S. suppressor protein is a component of an E3 lian prolyl hydroxylase acting on hypoxia- Hypoxia regulates vascular endothelial ubiquitin-protein ligase activity. Genes Dev. inducible factor. Proc Natl Acad Sci U S A. growth factor gene expression in endothelial 1999;13(14):1822–1833. 2002;99(21):13459–13464. cells. Identification of a 5’ enhancer. Circ Res. 74. Maxwell PH, et al. The tumour suppressor 90. Safran M, et al. Mouse model for noninvasive 1995;77(3):638–643. protein VHL targets hypoxia-inducible factors imaging of HIF prolyl hydroxylase activity: 60. Iliopoulos O, Levy AP, Jiang C, Kaelin WG, for oxygen-dependent proteolysis. Nature. assessment of an oral agent that stimulates Goldberg MA. Negative regulation of 1999;399(6733):271–275. erythropoietin production. Proc Natl Acad Sci hypoxia-inducible genes by the von Hip- 75. Ohh M, et al. Ubiquitination of hypoxia-induc- U S A. 2006;103(1):105–110. pel-Lindau protein. Proc Natl Acad Sci U S A. ible factor requires direct binding to the beta- 91. Mole DR, et al. 2-oxoglutarate analogue inhib- 1996;93(20):10595–10599. domain of the von Hippel-Lindau protein. Nat itors of HIF prolyl hydroxylase. Bioorg Med 61. Maxwell PH, et al. Hypoxia-inducible fac- Cell Biol. 2000;2(7):423–427. Chem Lett. 2003;13(16):2677–2680. tor-1 modulates gene expression in solid 76. Kallio PJ, Wilson WJ, O’Brien S, Makino Y, 92. Chan MC, Holt-Martyn JP, Schofield CJ, Rat- tumors and influences both angiogenesis Poellinger L. Regulation of the hypoxia- cliffe PJ. Pharmacological targeting of the HIF and tumor growth. Proc Natl Acad Sci U S A. inducible transcription factor 1alpha by the hydroxylases--A new field in medicine devel- 1997;94(15):8104–8109. ubiquitin-proteasome pathway. J Biol Chem. opment. Mol Aspects Med. 2016;47-48:54–75. 62. Jiang BH, Agani F, Passaniti A, Semenza GL. 1999;274(10):6519–6525. 93. Gordeuk VR, et al. Congenital disorder of oxy- V-SRC induces expression of hypoxia-induc- 77. Cockman ME, et al. Hypoxia inducible fac- gen sensing: association of the homozygous ible factor 1 (HIF-1) and transcription of genes tor-alpha binding and ubiquitylation by the Chuvash polycythemia VHL mutation with encoding vascular endothelial growth factor and von Hippel-Lindau tumor suppressor protein. thrombosis and vascular abnormalities but not enolase 1: involvement of HIF-1 in tumor pro- J Biol Chem. 2000;275(33):25733–25741. tumors. Blood. 2004;103(10):3924–3932. gression. Cancer Res. 1997;57(23):5328–5335. 78. Kamura T, Sato S, Iwai K, Czyzyk-Krzeska 94. Song D, et al. Defective Tibetan PHD2 63. Zhong H, et al. Increased expression of hypoxia M, Conaway RC, Conaway JW. Activation binding to p23 links high altitude adaption inducible factor-1alpha in rat and human pros- of HIF1alpha ubiquitination by a reconsti- to altered oxygen sensing. J Biol Chem. tate cancer. Cancer Res. 1998;58(23):5280–5284. tuted von Hippel-Lindau (VHL) tumor sup- 2014;289(21):14656–14665. 64. Talks KL, et al. The expression and distribution pressor complex. Proc Natl Acad Sci U S A. 95. Beall CM, et al. Natural selection on EPAS1 of the hypoxia-inducible factors HIF-1alpha 2000;97(19):10430–10435. (HIF2alpha) associated with low hemoglobin and HIF-2alpha in normal human tissues, can- 79. Wang GL, Semenza GL. Desferrioxamine concentration in Tibetan highlanders. Proc Natl cers, and tumor-associated macrophages. Am induces erythropoietin gene expression and Acad Sci U S A. 2010;107(25):11459–11464. J Pathol. 2000;157(2):411–421. hypoxia-inducible factor 1 DNA-binding activ- 96. Arany Z, et al. An essential role for p300/CBP 65. Kallio PJ, Pongratz I, Gradin K, McGuire J, ity: implications for models of hypoxia signal in the cellular response to hypoxia. Proc Natl Poellinger L. Activation of hypoxia-inducible transduction. Blood. 1993;82(12):3610–3615. Acad Sci U S A. 1996;93(23):12969–12973. factor 1alpha: posttranscriptional regulation 80. Jaakkola P, et al. Targeting of HIF-alpha to the 97. Carrero P, Okamoto K, Coumailleau P, O’Brien and conformational change by recruitment of von Hippel-Lindau ubiquitylation complex by S, Tanaka H, Poellinger L. Redox-regulated the Arnt transcription factor. Proc Natl Acad O2-regulated prolyl hydroxylation. Science. recruitment of the transcriptional coacti- Sci U S A. 1997;94(11):5667–5672. 2001;292(5516):468–472. vators CREB-binding protein and SRC-1 to 66. Salceda S, Caro J. Hypoxia-inducible factor 81. Ivan M, et al. HIFalpha targeted for VHL- hypoxia-inducible factor 1alpha. Mol Cell Biol. 1alpha (HIF-1alpha) protein is rapidly degrad- mediated destruction by proline hydroxyl- 2000;20(1):402–415. ed by the ubiquitin-proteasome system under ation: implications for O2 sensing. Science. 98. Ema M, et al. Molecular mechanisms of tran- normoxic conditions. Its stabilization by 2001;292(5516):464–468. scription activation by HLF and HIF1alpha in hypoxia depends on redox-induced changes. 82. Yu F, White SB, Zhao Q, Lee FS. HIF-1alpha response to hypoxia: their stabilization and J Biol Chem. 1997;272(36):22642–22647. binding to VHL is regulated by stimulus-sen- redox signal-induced interaction with CBP/ 67. Huang LE, Gu J, Schau M, Bunn HF. Regulation sitive proline hydroxylation. Proc Natl Acad Sci p300. EMBO J. 1999;18(7):1905–1914. of hypoxia-inducible factor 1alpha is mediated U S A. 2001;98(17):9630–9635. 99. Mahon PC, Hirota K, Semenza GL. FIH-1: by an O2-dependent degradation domain via 83. Min JH, Yang H, Ivan M, Gertler F, Kaelin WG, a novel protein that interacts with HIF-1 the ubiquitin-proteasome pathway. Proc Natl Pavletich NP. Structure of an HIF-1alpha -pVHL alpha and VHL to mediate repression of Acad Sci U S A. 1998;95(14):7987–7992. complex: hydroxyproline recognition in signal- HIF-1 transcriptional activity. Genes Dev. 68. Sutter CH, Laughner E, Semenza GL. Hypoxia- ing. Science. 2002;296(5574):1886–1889. 2001;15(20):2675–2686. inducible factor 1alpha protein expression 84. Hon WC, et al. Structural basis for the recog- 100. Lando D, Peet DJ, Whelan DA, Gorman JJ, is controlled by oxygen-regulated ubiquiti- nition of hydroxyproline in HIF-1 alpha by Whitelaw ML. Asparagine hydroxylation of the nation that is disrupted by deletions and pVHL. Nature. 2002;417(6892):975–978. HIF transactivation domain a hypoxic switch. missense mutations. Proc Natl Acad Sci U S A. 85. Masson N, Willam C, Maxwell PH, Pugh CW, Science. 2002;295(5556):858–861. 2000;97(9):4748–4753. Ratcliffe PJ. Independent function of two 101. Lando D, Peet DJ, Gorman JJ, Whelan DA, 69. Lonergan KM, et al. Regulation of hypoxia- destruction domains in hypoxia-inducible Whitelaw ML, Bruick RK. FIH-1 is an aspar- inducible mRNAs by the von Hippel-Lindau factor-alpha chains activated by prolyl hydrox- aginyl hydroxylase enzyme that regulates the tumor suppressor protein requires binding to ylation. EMBO J. 2001;20(18):5197–5206. transcriptional activity of hypoxia-inducible complexes containing elongins B/C and Cul2. 86. Myllyharju J, Kivirikko KI. Characterization factor. Genes Dev. 2002;16(12):1466–1471. Mol Cell Biol. 1998;18(2):732–741. of the iron- and 2-oxoglutarate-binding sites 102. Hewitson KS, et al. Hypoxia-inducible factor 70. Jackson PK. Cell cycle: cull and destroy. of human prolyl 4-hydroxylase. EMBO J. (HIF) asparagine hydroxylase is identical Curr Biol. 1996;6(10):1209–1212. 1997;16(6):1173–1180. to factor inhibiting HIF (FIH) and is related 71. Kamura T, et al. Rbx1, a component of the VHL 87. Epstein AC, et al. C. elegans EGL-9 and mam- to the cupin structural family. J Biol Chem. tumor suppressor complex and SCF ubiquitin malian homologs define a family of dioxygen- 2002;277(29):26351–26355. ligase. Science. 1999;284(5414):657–661. ases that regulate HIF by prolyl hydroxylation. 103. Elkins JM, et al. Structure of factor-inhibiting

3636 jci.org Volume 126 Number 10 October 2016 The Journal of Clinical Investigation NEWS

hypoxia-inducible factor (HIF) reveals mecha- sentation of coronary disease. Am Heart J. 137. Shimoda LA, Semenza GL. HIF and the lung: nism of oxidative modification of HIF-1 alpha. 2007;154(6):1035–1042. role of hypoxia-inducible factors in pulmonary J Biol Chem. 2003;278(3):1802–1806. 122. Resar JR, et al. Hypoxia-inducible factor development and disease. Am J Respir Crit 104. Berra E, Benizri E, Ginouvès A, Volmat V, Roux 1alpha polymorphism and coronary collaterals Care Med. 2011;183(2):152–156. D, Pouysségur J. HIF prolyl-hydroxylase 2 is in patients with ischemic heart disease. Chest. 138. Brusselmans K, et al. Heterozygous deficiency the key oxygen sensor setting low steady-state 2005;128(2):787–791. of hypoxia-inducible factor-2alpha protects levels of HIF-1alpha in normoxia. EMBO J. 123. Cai Z, Luo W, Zhan H, Semenza GL. Hypoxia- mice against pulmonary hypertension and 2003;22(16):4082–4090. inducible factor 1 is required for remote isch- right ventricular dysfunction during prolonged 105. Flashman E, Hoffart LM, Hamed RB, emic preconditioning of the heart. Proc Natl hypoxia. J Clin Invest. 2003;111(10):1519–1527. Bollinger JM, Krebs C, Schofield CJ. Evidence Acad Sci U S A. 2013;110(43):17462–17467. 139. Bonnet S, et al. An abnormal mitochondrial- for the slow reaction of hypoxia-inducible fac- 124. Olenchock BA, et al. EGLN1 Inhibition hypoxia inducible factor-1alpha-Kv chan- tor prolyl hydroxylase 2 with oxygen. FEBS J. and Rerouting of α-Ketoglutarate Suffice nel pathway disrupts oxygen sensing and 2010;277(19):4089–4099. for Remote Ischemic Protection. Cell. triggers pulmonary arterial hypertension 106. Tarhonskaya H, et al. Kinetic investigations of 2016;164(5):884–895. in fawn hooded rats: similarities to human the role of factor inhibiting hypoxia-inducible 125. Bosch-Marce M, et al. Effects of aging pulmonary arterial hypertension. Circulation. factor (FIH) as an oxygen sensor. J Biol Chem. and hypoxia-inducible factor-1 activity on 2006;113(22):2630–2641. 2015;290(32):19726–19742. angiogenic cell mobilization and recovery 140. Moslehi J, et al. Loss of hypoxia-inducible 107. Chan DA, Sutphin PD, Yen SE, Giaccia of perfusion after limb ischemia. Circ Res. factor prolyl hydroxylase activity in cardiomyo- AJ. Coordinate regulation of the oxygen- 2007;101(12):1310–1318. cytes phenocopies ischemic cardiomyopathy. dependent degradation domains of hypoxia- 126. Loinard C, et al. Inhibition of prolyl Circulation. 2010;122(10):1004–1016. inducible factor 1 alpha. Mol Cell Biol. hydroxylase domain proteins promotes 141. Bekeredjian R, et al. Conditional HIF-1alpha 2005;25(15):6415–6426. therapeutic revascularization. Circulation. expression produces a reversible cardiomyo 108. Masson N, Ratcliffe PJ. HIF prolyl and asparag- 2009;120(1):50–59. pathy. PLoS One. 2010;5(7):e11693. inyl hydroxylases in the biological response to 127. Rey S, et al. Synergistic effect of HIF-1alpha 142. Belaidi E, Morand J, Gras E, Pépin JL, intracellular O(2) levels. J Cell Sci. 2003; gene therapy and HIF-1-activated bone mar- Godin-Ribuot D. Targeting the ROS-HIF-1- 116(Pt 15):3041–3049. row-derived angiogenic cells in a mouse model endothelin axis as a therapeutic approach 109. Semenza GL. Regulation of metabolism by of limb ischemia. Proc Natl Acad Sci U S A. for the treatment of obstructive sleep hypoxia-inducible factor 1. Cold Spring Harb 2009;106(48):20399–20404. apnea-related cardiovascular complications Symp Quant Biol. 2011;76:347–353. 128. Botusan IR, et al. Stabilization of HIF-1 [published online ahead of print August 110. Semenza GL. Hypoxia. Cross talk between oxy- alpha is critical to improve wound healing 3, 2016]. Pharmacol Ther. doi:10.1016/j. gen sensing and the cell cycle machinery. Am J in diabetic mice. Proc Natl Acad Sci U S A. pharmthera.2016.07.010. Physiol Cell Physiol. 2011;301(3):C550–C552. 2008;105(49):19426–19431. 143. Semenza GL, Prabhakar NR. Neural regulation 111. Semenza GL. Dynamic regulation of stem cell 129. Mace KA, Yu DH, Paydar KZ, Boudreau N, of hypoxia-inducible factors and redox state specification and maintenance by hypoxia- Young DM. Sustained expression of Hif-1alpha drives the pathogenesis of hypertension in a inducible factors. Mol Aspects Med. 2016; in the diabetic environment promotes angio- rodent model of sleep apnea. J Appl Physiol. 47–48:15–23. genesis and cutaneous wound repair. Wound 2015;119(10):1152–1156. 112. Hubbi ME, Semenza GL. Regulation of cell Repair Regen. 2007;15(5):636–645. 144. Ratcliffe PJ. Oxygen sensing and hypoxia sig- proliferation by hypoxia-inducible factors. Am J 130. Zhang X, et al. Impaired angiogenesis and nalling pathways in animals: the implications Physiol Cell Physiol. 2015;309(12):C775–C782. mobilization of circulating angiogenic of physiology for cancer. J Physiol (Lond). 113. Ratcliffe PJ. Understanding hypoxia signalling cells in HIF-1alpha heterozygous-null mice 2013;591(8):2027–2042. in cells--a new therapeutic opportunity? Clin after burn wounding. Wound Repair Regen. 145. Shen C, Kaelin WG. The VHL/HIF axis in Med (Lond). 2006;6(6):573–578. 2010;18(2):193–201. clear cell renal carcinoma. Semin Cancer Biol. 114. Hubbi ME, et al. A nontranscriptional role for 131. Jiang X, et al. Adenovirus-mediated HIF- 2013;23(1):18–25. HIF-1α as a direct inhibitor of DNA replica- 1α gene transfer promotes repair of mouse 146. Kim W, Kaelin WG. The von Hippel-Lindau tion. Sci Signal. 2013;6(262):ra10. airway allograft microvasculature and tumor suppressor protein: new insights into 115. Hubbi ME, et al. Cyclin-dependent kinases attenuates chronic rejection. J Clin Invest. oxygen sensing and cancer. Curr Opin Genet regulate lysosomal degradation of hypox- 2011;121(6):2336–2349. Dev. 2003;13(1):55–60. ia-inducible factor 1α to promote cell-cycle 132. Bernhardt WM, et al. Donor treatment with a 147. Kaelin WG. Cancer and altered metabolism: progression. Proc Natl Acad Sci U S A. PHD-inhibitor activating HIFs prevents graft potential importance of hypoxia-inducible 2014;111(32):E3325–3334. injury and prolongs survival in an allogenic factor and 2-oxoglutarate-dependent dioxy- 116. Bishop T, Ratcliffe PJ. HIF hydroxylase path- kidney transplant model. Proc Natl Acad Sci genases. Cold Spring Harb Symp Quant Biol. ways in cardiovascular physiology and medi- U S A. 2009;106(50):21276–21281. 2011;76:335–345. cine. Circ Res. 2015;117(1):65–79. 133. Karhausen J, Furuta GT, Tomaszewski JE, 148. Semenza GL. The hypoxic tumor micro- 117. Pugh CW, Ratcliffe PJ. Regulation of angiogen- Johnson RS, Colgan SP, Haase VH. Epithelial environment: A driving force for breast esis by hypoxia: role of the HIF system. Nat hypoxia-inducible factor-1 is protective in cancer progression. Biochim Biophys Acta. Med. 2003;9(6):677–684. murine experimental colitis. J Clin Invest. 2016;1863(3):382–391. 118. Kaelin WG, Ratcliffe PJ. Oxygen sensing by 2004;114(8):1098–1106. 149. Semenza GL. Cancer-stromal cell interactions metazoans: the central role of the HIF hydrox- 134. Cummins EP, et al. The hydroxylase inhib- mediated by hypoxia-inducible factors pro- ylase pathway. Mol Cell. 2008;30(4):393–402. itor dimethyloxalylglycine is protective in mote angiogenesis, lymphangiogenesis, and 119. Taylor CT, Doherty G, Fallon PG, Cummins a murine model of colitis. Gastroenterology. metastasis. Oncogene. 2013;32(35):4057–4063. EP. Hypoxia-dependent regulation of inflam- 2008;134(1):156–165. 150. Kondo K, Klco J, Nakamura E, Lechpam- matory pathways in immune cells. J Clin Invest. 135. Hirota SA, et al. Hypoxia-inducible factor mer M, Kaelin WG. Inhibition of HIF is 2016;126(10):3716–3724. signaling provides protection in Clostridium necessary for tumor suppression by the 120. Glover LE, Lee JS, Colgan SP. Oxygen metab- difficile-induced intestinal injury. Gastroenter- von Hippel-Lindau protein. Cancer Cell. olism and barrier regulation in the intestinal ology. 2010;139(1):259–269.e3. 2002;1(3):237–246. mucosa. J Clin Invest. 2016;126(10):3680–3688. 136. Yoon D, Ponka P, Prchal JT. Hypoxia. 5. 151. Kondo K, Kim WY, Lechpammer M, Kaelin 121. Hlatky MA, et al. Polymorphisms in hypoxia Hypoxia and hematopoiesis. Am J Physiol Cell WG. Inhibition of HIF2alpha is sufficient to inducible factor 1 and the initial clinical pre- Physiol. 2011;300(6):C1215–C1222. suppress pVHL-defective tumor growth. PLoS

jci.org Volume 126 Number 10 October 2016 3637 NEWS The Journal of Clinical Investigation

Biol. 2003;1(3):E83. cinoma. Cancer Cell. 2002;1(3):247–255. Cancer Cell. 2008;14(6):435–446. 152. Shen C, et al. Genetic and functional studies 155. Raval RR, et al. Contrasting properties of 158. Maxwell PH, Eckardt KU. HIF prolyl hydrox- implicate HIF1α as a 14q kidney cancer suppres- hypoxia-inducible factor 1 (HIF-1) and HIF-2 in ylase inhibitors for the treatment of renal sor gene. Cancer Discov. 2011;1(3):222–235. von Hippel-Lindau-associated renal cell carci- anaemia and beyond. Nat Rev Nephrol. 153. Zimmer M, Doucette D, Siddiqui N, Iliopoulos noma. Mol Cell Biol. 2005;25(13):5675–5686. 2016;12(3):157–168. O. Inhibition of hypoxia-inducible factor is 156. Kim WY, et al. Failure to prolyl hydroxylate 159. Rogers JL, et al. Development of inhibitors of sufficient for growth suppression of VHL-/- hypoxia-inducible factor alpha pheno- the PAS-B domain of the HIF-2α transcription tumors. Mol Cancer Res. 2004;2(2):89–95. copies VHL inactivation in vivo. EMBO J. factor. J Med Chem. 2013;56(4):1739–1747. 154. Maranchie JK, Vasselli JR, Riss J, Bonifacino 2006;25(19):4650–4662. 160. Zhang H, et al. Digoxin and other cardiac JS, Linehan WM, Klausner RD. The contribu- 157. Gordan JD, et al. HIF-alpha effects on c-Myc glycosides inhibit HIF-1alpha synthesis and tion of VHL substrate binding and HIF1-alpha distinguish two subtypes of sporadic VHL- block tumor growth. Proc Natl Acad Sci U S A. to the phenotype of VHL loss in renal cell car- deficient clear cell renal carcinoma. 2008;105(50):19579–19586.

3638 jci.org Volume 126 Number 10 October 2016