Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Mechanism of regulation of the hypoxia‐inducible factor‐1α by the von Hippel‐Lindau tumor suppressor protein

Mechanism of regulation of the hypoxia‐inducible factor‐1α by the von Hippel‐Lindau tumor... The EMBO Journal Vol.19 No.16 pp.4298--4309, 2000 Mechanism of regulation of the hypoxia-inducible factor-1a by the von Hippel-Lindau tumor suppressor protein et al., 1996; Iliopoulos et al., 1996) and/or transcriptional Keiji Tanimoto, Vuichi Makino, (Mukhopadhyay et al., 1997) mechanisms. Teresa Pereira and Lorenz Poellinger The VHL protein displays no sequence similarity to Department of Cell and Molecular Biology, Medical Nobel Institute, other known proteins, thus giving no clues about its Karolinska Institutet, S-171 77 Stockholm, Sweden function. Biochemical studies have shown that VHL is Corresponding author associated with elongins B and C, and cullin-2 (Cul-2) e-mail: [email protected] (Kaelin and Maher, 1998), forming the VHL-BC--Cul-2 complex. The crystal structure of the VHL-BC ternary In normoxic cells the hypoxia-inducible factor-la complex shows two interfaces: one between VHL and (IIlF-la.) is rapidly degraded by the ubiquitin-protea­ elongin C and another between elongins B and C (Stebbins some pathway, and activation of HIF-la. to a func­ et al., 1999). The 35 residue-long elongin C binding tional form requires protein stabilization. Here we domain of VHL represents one of the mutational hotspots show that the product of the von Hippel-Lindau in tumors (Kaelin and Maher, 1998), suggesting that (VHL) tumor suppressor gene mediated ubiquityla­ VHL-BC complex formation is critical for tumor sup­ tion and proteasomal degradation of HIF-la under pressor function. In addition, there is a mutational hotspot normoxic conditions via interaction with the core of on a separate domain, the � domain of VHL (Kaelin and the oxygen-dependent degradation domain of IIlF-la. Maher, 1998), which overlaps with a putative macro­ The region of VHL mediating interaction with HIF-la molecular binding site identified in the VHL-BC crystal overlapped with a putative macromolecular binding structure (Stebbins et al., 1999). Elongins B and C and site observed within the crystal structure of VHL. Cul-2 all share homology to components of the SCF This motif of VHL also represents a mutational hot­ (Skpl--Cul-1-F-box protein) multiprotein complex, which spot in tumors, and one of these mutations impaired targets cell cycle regulatory proteins for ubiquitin-medi­ interaction with HIF-la and subsequent degradation. ated proteolysis (Ciechanover, 1998). Importantly, the Interestingly, the VHL binding site within HIF-la structure of the VHL-BC complex extends these similar­ overlapped with one of the minimal transactivation ities to the SCF complex structure (Stebbins et al., 1999), domains. Protection of HIF-la against degradation by indicating that these protein complexes may have similar VHL was a multistep mechanism, including hypoxia­ functions. In excellent agreement with this model, VHL induced nuclear translocation of HIF-la and an intra­ has recently been shown to be associated with an E3 nuclear hypoxia-dependent signal. VHL was not ubiquitin ligase activity in cellular extracts (lwai et al., released from HIF -la during this process. Finally, 1999; Lisztwan et al., 1999). stabilization of HIF-la protein levels per se did not In hypoxic cells, the hypoxia-inducible factor-la. totally bypass the need of the hypoxic signal for gener­ (HIF-la.) mediates transcriptional activation of the ating the transactivation response. VEGF gene (Iyer et al., 1998). HIF-la. mRNA is Keywords: hypoxia-inducible factor-la/transcription/ constitutively expressed in a number of mammalian cells tumor suppression/ubiquitylation/von Rippel-Lindau (Kallio et al., 1997). In contrast, the HIF-la. protein is protein remarkably unstable in cells at normoxia, whereas hypoxia dramatically stabilizes the protein (Kallio et al., 1997). Under normoxic conditions HIF-la. protein degradation is mediated by the ubiquitin-proteasome pathway and a Introduction distinct oxygen-dependent degradation domain of HIF-la. von Rippel-Lindau (VHL) disease is caused by germ line (Huang et al., 1998; Kallio et al., 1999). Stabilization of mutations of the VHL gene. These mutations lead to the HIF-la. initiates a multi-step pathway of activation of development of a variety of tumors including clear cell HIF-la. that includes hypoxia-dependent nuclear trans­ carcinomas of the kidney, pheochromocytomas and vas­ location and dimerization with a partner DNA binding cular tumors of the central nervous system and retina. factor, Arnt, to interact with cognate hypoxia-response VHL-associated neoplasms are typically hypervascular elements of target promoters, followed by recruitment of (reviewed by Kaelin and Maher, 1998), and under transcriptional coactivators (Wenger and Gassmann, 1997; normoxic conditions VHL-deficient cells express vascular Kallio et al., 1998; Ema et al., 1999). Recently, two studies endothelial growth factor (VEGF) mRNA, which is have indicated distinct roles of VHL in regulation of normally expressed in a hypoxia-dependent fashion HIF-la. function: induction of a natural HIF-la. antisense (Gnarra et al., 1996; Iliopoulos et al., 1996). Reintro­ transcript in VHL-deficient cells, resulting in negative duction of VHL into VHL-mutated renal carcinoma cells regulation of HIF-la. function (Thrash-Bingham and indicates that it functions as a negative regulator of Tartof, 1999). On the other hand, VHL has recently been VEGF mRNA levels by post-transcriptional (Gnarra reported to interact physically with HIF-la., possibly 4298 © European Molecular Biology Organization Regulation of HIF-1a. by VHL HIF-1 o:: : + ++ +++ VHL + + Hypoxia: + + Hypoxia: + + + -175 -175 HIF-1a►-- -83 -83 -32 1 2 3 4 5 6 VHL► -25 1 2 3 4 VHL : VHL: + + + -83 MG-132: + + DR► -175 -62 HIF-1o:► -32 VHL► -83 1 2 3 4 2 1 2 Fig. 1. VHL mediates proteasomal degradation of HIP-la under normoxic conditions. (A) Expression of HIP-la under normoxic conditions. Increasing amounts [0.2 µg (+), 0.5 µg (++), 1.0 µg (+++)] of pFLAG CMV2/HIF-la were transiently transfected into COS7 cells, and the cells were incubated for 12 h at normoxia (21 % 0 ) or hypoxia (1 % 0 ), as indicated. Whole-cell extracts were analyzed by immunoblotting using anti-FLAG 2 2 antibodies. (B) Degradation of HIP-la in the presence of VHL. pFLAG CMV2/HIF-la was cotransfected into COS7 cells together with empty vector or wild-type VHL expression vector (pCMX/VHL). Whole-cell extracts were analyzed by immunoblotting using anti-FLAG or anti-VHL antibodies. (C) VHL mediates proteasomal degradation of HIP-la. Cells were transfected with pFLAG CMV2/HIF-la and pCMX/VHL, incubated in the absence or presence of 5 µM MG-132 for 6 h before harvesting, and cellular extracts were analyzed as in (B). (D) VHL does not affect dioxin receptor (DR) protein levels. COS7 cells were transfected with a FLAG-tagged dioxin receptor expression vector (pCMV/DR/FLAG) in the absence or presence of pCMX/VHL, and analyzed as in (B). The mobilities of molecular weight (kDa) markers are shown on the right hand side of the blots. targeting HIF-la for protein degradation (Maxwell et al., genes such as VEGF are constitutively expressed at 1999). Here we show that VHL directly mediated normoxia in VHL-deficient cells (Gnarra et al., 1996; ubiquitylation and ensuing proteasomal degradation of Iliopoulos et al., 1996), indicating dysregulation of HIF-la at normoxia via physical interaction with the core HIF-la function in these cells. We were thus interested of the oxygen-dependent degradation domain. This motif to investigate the potential mechanism of regulation of coincided with one of the minimal transactivation domains HIF-la function by VHL. Due to the pronounced !ability of HIF-1 a. The domain mediating interaction with HIF-1 a of the HIF-la protein under normoxic conditions, HIF-la also represented one of the mutational hotspots of VHL. is normally not detectable by immunoblot analysis of Protection against VHL-mediated degradation required cellular extracts (Kallio et al., 1997, 1998). It was both nuclear translocation of HIF-la and an intranuclear therefore not possible to investigate the effect of expres­ hypoxia-dependent regulatory signal. Finally, stabilization sion of VHL on endogenous HIF-la protein levels. To of HIF-la protein levels per se did not bypass the need of establish experimental conditions to examine the effect of the hypoxic signal for generating the transactivation VHL on HIF-la protein stability, we transiently trans­ response. fected COS7 cells with FLAG epitope-tagged HIF-1 a expression plasmids. As expected, at a low concentration (0.2 µg) of expression vector, HIF-la was not detected at Results normoxia, and we observed potent stabilization ofHIF-la protein levels at hypoxia, as assessed by immunoblot Regulation of HIF-1a protein stability by the VHL analysis (Figure lA). However, at higher concentrations tumor suppressor protein (0.5-1 µg) of expression vector we could detect HIF-la We and others have recently demonstrated that HIF-la is protein expression also under normoxic conditions regulated by the ubiquitin-proteasome pathway under (Figure lA). At the highest concentrations of expression normoxic conditions, resulting in very rapid turnover of vector tested we observed significant HIF-1 a expression the protein, and that one of the early responses to hypoxia levels both at normoxia and hypoxia (Figure lA). Thus, is massive upregulation of HIF-la protein levels (Kallio these experiments suggest that the mechanism of degrad­ et al., 1997; Salceda and Caro, 1997; Huang et al., 1998; Kallio et al., 1999). Interestingly, VHL protein complexes ation of HIF-1 a had become saturated under these have recently been demonstrated to harbor E3 ubiquitin­ conditions and that one or several components of the protein ligase activity, although the target protein for this degradation machinery were limiting. activity has not yet been identified (Lisztwan et al., 1999; We next used the high level HIF-la expression Iwai et al., 1999). Moreover, HIP-la-regulated target conditions for all subsequent experiments to examine the 4299 GAL41VHL ti114-154 _______ o __ _,l K.Tanimoto et al. effect of VHL on HIF-la protein levels. Transient inhibitor MG-132 (Figure IC). Taken together, these coexpression of FLAG/HIF-la and VHL resulted in results strongly suggest that VHL mediates proteasomal degradation of HIF-1 a. This effect of VHL was specific reduction of the HIF-la protein signal under normoxic for HIF-la, as transiently expressed VHL did not produce conditions (Figure IB, upper panel), indicating that VHL this effect on FLAG-tagged dioxin receptor (Figure ID), a may have been limiting under the conditions of expression basic helix-loop-helix(bHLH)/P AS (Per/ Arnt/Sim do­ of HIF-la alone. Interestingly, VHL failed to induce main) protein belonging to the same class of transcription reduction of HIF-la protein levels under hypoxic condi­ factors as HIF-la. tions (Figure IB). In control experiments we detected similar levels of VHL expression in extracts from either Two domains of VHL are required for inducing normoxic or hypoxic cells (Figure IB, lower panel). VHL­ protein degradation of HIF-1a induced reduction of HIF-la protein levels at normoxia Given the potential role of VHL as an E3 ubiquitin ligase was inhibited by treatment of the cells with the proteasome we examined whether VHL physically interacted with HIF-la. S-labeled, in vitro translated HIF-la was incubated with wild-type or mutant GAL4/VHL fusion proteins (schematically represented in Figure 2A) or the elo oglo C minimal GAL4 DNA binding domain alone prior to blndl np GAL41VHLwt II i immunoprecipitation assays. In these experiments, 11' 1H 2U S-labeled HIF-la was co-immunoprecipitated in the __,r---{ ._. 1 110 1H ZU presence of GAL4/VHL by anti-GAL4 specific antibodies, whereas no interaction was observed between HIF-la and GAL4/VHL 114-154 � 114 lk the minimal GAL4 DNA binding domain (Figure 2B, GAL4/VHL 91-154 upper panel). Non-specific pre-immune rabbit antiserum did not precipitate HIF-la protein in the presence of either GAL4/VHL pmt1 213 VHL or GAL4 alone (Figure 2B, lower panel), indicating that wild-type VHL specifically interacted with HIF-1 a B in vitro. GAL4/VHL '5 The VHL M 14-154 deletion mutant showed interaction Cl, ::z ..,. U'I with HIF-la, whereas the VHL 114-154 fragment failed .E ::z ..,. :. '-' ... ":- :. to do so (Figure 2B). A 23 amino acid-long N-terminal Q ... ... i <i ... ;; � extension of this fragment generated VHL 91-154, which ·175 was able to interact with HIF-la, indicating the import­ HIF-1a► o.GAL4 ance of a structure located between residues 91 and 113 of VHL to interact with HIF-la. Interestingly, this region of -89 VHL is not only contained within the putative macro­ molecular binding site observed in the crystal structure of HIF•1o.► Control the VHL-BC complex (Stebbins et al., 1999), but also -89 represents one of the mutational hotspots in tumors 2 9 4 s 6 (Kaelin and Maher, 1998). This fact prompted us to examine whether tumor-derived mutations of VHL would '5 GAL4/VHL affect its ability to interact with HIF-1 a and/or to induce HIF-la degradation. We performed these experiments :s IO ◄ ... .. CJ 0 i >- ·17S o. GAL4 HIF-1a ► Fig. 2. VHL requires two functional domains to induce HIF-lo. degradation. (A) Schematic representation of GAL4-fused wild-type - 8 (wt), deletion or single amino acid point mutant (pmt) forms of VHL . • 175 (B) VHL directly interacts with HIF-lo.. Equal concentrations of Control HIF•1a ► in vitro translated S-labeled full-length HIF-lo. were incubated with in vitro translated wild-type GAL4NHL or VHL deletion mutants or -83 2 3 4 5 GAL4-DBD spanning the GAL4 DNA binding domain alone. Co­ irnmunoprecipitation assays with anti-GAL4 antibody (upper panel) or control preimmune serum (lower panel) were carried out as described GAL41VHL in Materials and methods. The precipitated material was analysed by LI. SOS-PAGE and autoradiography. For loading controls, 10% of UI 35 input S-labeled HIF-lo. is shown in lane 1. (C) A tumor-derived CJ i ! 0 point mutation impairs the interaction between VHL and HIF-la. -175 In vitro translated proteins were incubated and analyzed by co­ immunoprecipitation as in (B). (D) Tumor-derived point-mutated forms of VHL fail to induce HIF-la degradation. pFLAG CMV2/HIF-la was HIF-1a► -83 transiently coexpressed in COS7 cells in the absence or presence of GAL4-fused wild-type or mutant forms of VHL as indicated. Cells were incubated at normoxia for 24 h, and whole cell extracts were -50 prepared and analyzed by immunoblotting using anti-FLAG or anti­ VHL► VHL antibodies. The mobilities of molecular weight (kDa) markers 2 3 4 are shown on the right hand side of the blots. 4300 Regulation of HIF-1a. by VHL VHL: !21212 -175 PA.S N-TAD C-T.AD ,.--, FLAG/HIF-1o. 11111 II a n I I I 1 532 585 82& -8S ► - FLAG/HIF-10.1-652 II n -8S Ell I 1 &52 -47.5 ►- FLAG/HIF-1o. 1-330 11111 II IEI n II 1 330 -32 FLAG/HIF-1cx526-826 WWW -47.5 t I 526 828 B C GAL41HIF-1a a. .E .E 'It ...I 'i/- 'i/- ... < <E 0 0 ; C") ... ... ... CJ C") c:, 3 3 2 2 a-GAL4 a-GAL4 VHL ► VHL ► 5 - ---- ----- ------- 3 · 2 3 Control Control VHL ► VHL ► 2 - 5 2 1 2 3 1 2 3 4 5 6 7 8 Fig. 3. VHL targets the oxygen-dependent degradation domain of HIP-la.. (A) FLAG-tagged wild-type HIP-la. or the indicated HIP-la. deletion mutants were transiently coexpressed in COS? cells at normoxia in the absence or presence of VHL. Whole-cell extracts were prepared and analyzed as described in Figure 1. ODD, oxygen-dependent degradation domain; N- and C-TAD, N- and C-terminal transactivation domains. (B) S-labeled VHL was incubated with equal concentrations of in vitro translated GAL4-fusion proteins spanning full-length HIF-la. or GAL4-DBD alone. Co-immunoprecipitation assays were performed with anti-GAL4 antibodies (upper panel) or control non-specific rabbit antiserum (lower panel). The precipitated material was analyzed by SDS-PAGE and autoradiography. For loading controls, 10% of input S-labeled VHL is shown. (C) VHL, GAL4 fusion proteins spanning the indicated HIP-la. fragments or GAL4 alone were expressed by in vitro translation and analyzed as in (B). using GAL4NHL fusion proteins containing either a mediate degradation of HIF-la, and that regulation of Y98N (the most frequent tumor mutation in this region; HIF-la may be involved in the tumor suppressor fu nction Kaelin and Maher, 1998) or a C162F single amino acid of VHL. mutation. The C162F mutation has been demonstrated to render VHL unable to bind the elongin B-C complex The oxygen-dependent degradation domain of (Lonergan et al., 1998; Lisztwan et al., 1999), and inhibit HIF-1a is targeted for regulation by VHL ubiquitin ligase activity in vitro (Lisztwan et al., 1999). In To identif y the domain of HIF-1 a that is targeted by VHL co-immunoprecipitation experiments, VHL Y98N was to mediate proteasomal degradation at normoxia, we unable to interact with HIF-la, whereas VHL C162F transiently expressed in COS7 cells in the presence or showed wild-type levels of interaction with HIF-la absence ofVHL either wild-type FLAG/HIF-la or a series (Figure 2C, upper panel). In our cellular degradation of FLAG-tagged HIF-la deletion mutants. In analogy to assay we transiently expressed at normoxia FLAG/HIF-la wild-type HIF-la, HIF-la 1-652 lacking the C-terminus in the presence or absence of wild-type or the individual including the C-terminal transactivation domain (sche­ point-mutated forms of VHL. Immunoblot analysis dem­ matically represented in Figure 3A) was degraded in the onstrated that, in contrast to wild-type VHL, both the VHL presence of VHL. However, the protein levels of HIF-la Y98N and VHL C162F mutants failed to induce degrad­ 1-330 lacking structures C-terminal of the PAS domain ation of HIF-la at normoxia (Figure 2D). These results were not affected by VHL. HIF-la 526-826 lacking demonstrate that both the HIF-la interaction domain and N-terminal structures (including the bHLH and PAS the elongin C binding domain of VHL are necessary to domains) was also degraded upon exposure to VHL at 11111 K.Tanimo to et al. A D 1B: a HA 522 541 N-TAO wtN •TAO mt mEPAS -1 ELDLETLAPY IP MDGEDFQL 523 542 VHL: + + + hEPA.S-1 ELDLE TL APYI PMDGEGPQL 555 HA-Ub : + + + + + mHIF-1 a. DLDLEMLAPY IP MD- DDFQI, hH IF-1cr. DLDLBMLAPY IP MD-D DPQL -175 Ubil uitylated [ - I I I -83 N-T. 0 hH IF-1cr. mt DDD IP: o.F LAG -62 lg► B -47.5 a. -3 .E /N CJ '#-- 4:( .... � i e -32 N -TA D► o:-FLAG . VHL WCE -2 1 2 3 4 2 VHL ► -2 1 2 3 4 5 N-TA D wt N-TAD rnt + ++ + ++ VHL: 3 N •TAD N-T AD ► wt K532R K538R K547R - - 2 Hypox ia : + + + + VHL ► N-TAD ► ·2 1 2 3 4 5 6 7 1 2 3 4 5 6 7 8 Fig. 4. The minimal N-T AD of HIF- lu is the target for VHL-mediated ubiquitylation and proteasomal degradation. (A) Alignment of N-T AD sequences of human (h) and mouse (m) HIF- lu and EPAS -1 revealed a conserved core sequence motif. The positions of amino acid substitutions are shown in mutant (ml) N-TAD. (B) Mutation of the N-TAD PYI motif abolishes interaction between HIF- lu and VHL. S-labeled VHL was incubated with equal concentrations of in vitro translated FLAG-tagged wild-type, mutant N-TAD or FLAG epitope alone. Co-immunoprecipitation assays were carried out using anti-FLAG antibodies and analyzed by SOS- PAGE and autoradiography. For loading controls, 10% of input S-la beled VHL is shown. (C) The PYI mutation confers resistance to VHL-mediated degradat ion. FLAG or FLAG-tagged wild-type or mutant N-TAD (I µg/ 6-cm dish) were transiently coexpressed in COS7 cells in the absence or presence of increasing concentrations (1.0 µg, +; 2.0 µg/6-cm dish, ++) of VHL as indicated, and incubated at normoxia for 24 h. Whole-cell extracts were prepared and analyzed by immunoblotting using anti-FLAG or anti­ VHL antibod ies. (D) VHL mediates ubiquitylation of N-TAD. FLAG, or FLAG-tagged wild-type or mutant N-TAD and HA-tagged ubiquitin were transiently coexpressed in COS7 cells in the presence or absence of VHL under normoxic conditions in combination with MG 132 for 6 h. After immunoprecipitation of whole-cell extracts by anti-FLAG, ubiquitylated forms of N-T AD were detected by anti-HA immunoblotting (upper panel). Ten percent of input whole-cell extracts are shown in the two lower panels. (E) Effect on protein stability following substitution of individual lysines to arginines within N-TAD. Wild-type or mutant N-TAD were transiently expressed in 293 cells and incubated at normoxia or hypoxia for 12 h. Whole-cell extracts were prepared and analyzed as in (C). The mobilities of molecular weight (kDa) markers are shown to the right of the blots. (Figure 3C, upper panel). Moreover, GAL4/HIF-la 778- normoxia (Figure 3A). In conclusion, these results indicate that a C-terminal region of HIF-1 a spanning residues 526- 826, spanning the C-terminal transactivation domain of 652 mediated VHL-dependent degradation. HIF-la, did not show any interaction with VHL We proceeded to map the domain of HIF-1 a required to (Figure 3C, upper panel). In contrast, the GAL4/HIF-la interact with VHL. To this end we used fu ll-length HIF-la fragments, which showed VHL-mediated degradation in or a set of HIF-1 a deletion mutants fused to the normoxic cells, HIF-la 1-652 and HIF-la 526-826 GAL4 DNA binding domain, and performed co-immuno­ (Figure 3A), clearly interacted with VHL (Figure 3C). In precipitation assays following incubation with VHL. As control reactions, the GAL4/HIF-la fragments were expected, S-labeled VHL was specifically co-immuno­ expressed at similar levels (data not shown), and non­ precipitated together with full-length HIF-la (Figure 3B). specific pre-immune rabbit antiserum did not precipitate In excellent agreement with the fact that the deletion VHL protein in the presence of any of the used fragments mutant HIF-la 1-330 was not degraded upon over­or GAL4 alone (Figure 3B and C, lower panels). In a more expression of VHL in COS7 cells under normoxic detailed analysis of the interaction of VHL with HIF-la, conditions (Figure 3A), GAL4/HIF-la 1-330 failed VHL was co-immunoprecipitated by anti-GAL4 anti­ to interact physically in vitro with S-labeled VHL bodies in the presence of GAL4/HIF-la 331-641 or F:LA G TAD Reg ulation of HIF-1a. by VHL GAL4/HIF-la 526-641. In fact, when compared with full­ length HIF-la, all these latter fragments of HIF-la interacted with VHL with very similar efficacies (Figure 3C, upper panel). Taken together, these results indicate that a region of HIF-la spanning residues 526- VHL 641 was essential for physical interaction with VHL. Interestingly, this region overlaps with the oxygen/redox­ dependent degradation domain of HIF-la, which has 10µ m previously been demonstrated to mediate proteasomal degradation of HIF-1 a in normoxic cells, and that has broadly been defined to be located between amino acid residues 401 and 603 of hHIF-la (Huang et al., 1998; Kallio et al., 1999). The minimal N-terminal transactivation domain of GFP-HIF-1 a HIF-1a is a target for ubiquitylation and proteasomal degradation by VHL Interestingly, the VHL-interacting fragment GAL4/ HIF-la 526-641 contains not only the core of the oxygen-dependent degradation domain of HIF-1 a but GFP-HIF-1 cx/ also the N-terminal transactivation domain N-TAD (Jiang K71 9T et al., 1997; Pugh et al., 1997; Figure 3A). Within the N-TAD of HIF-la a sequence motif of ~19 amino acid residues (located between amino acids 556 and 574 of hHIF-la) shows the strongest conservation between species and is also conserved in the related hypoxia­ GFP-HIF-1 a/ inducible factor EP AS-1/HLF (Figure 4A), which is �1 78-390 expressed in a tissue-restricted fashion. This motif is also highly conserved in the hypoxia-responsive Droso phila Similar protein (reviewed by Taylor and 10 µm Zhulin, 1999). Interestingly, this sequence motif has recently been reported to be important for the function of the oxygen-dependent degradation domain of HIF-la since alanin substitutions within this domain impair VHL : + + hypoxia-dependent protein stabilization (Srinivas et al., Hypoxia : + + 1999). We initially examined whether VHL could interact 17 5 directly with the minimal N-TAD located between amino HIF 1o.► - acids 532 and 585 of hHIF-la (Jiang et al., 1997; Pugh et al., 1997). In co-immunoprecipitation assays, anti­ 8 -175 FLAG antibodies could precipitate S-labeled VHL in the presence of the FLAG epitope-tagged minimal N-T AD of HIF-1a K71 9T► HIF-la (Figure 4B), demonstrating that this region of -8 HIF-la is sufficient to mediate interaction with VHL. We next generated a point mutation within the context of the HIF-fo 6.1 78-390 ► minimal N-TAD. As assessed in co-immunoprecipitation assays, substitution of the central PYI triplet with aspartic acid totally abolished interaction between VHL and the - VHL► N-TAD (Figure 4B). In conclusion, these experiments -2. 5 demonstrate that the highly conserved core motif of the 1 2 3 4 N-T AD is critical for interaction with VHL. Fi . 5. Protection of HIF-lu against VHL-mediated degradation We next transiently expressed under normoxic con­ requires nuclear translocation of HIF-1 a. and a hypoxia-dependent ditions the wild-type or mutant N-TAD in the absence intranuclear signal. (A) Subcellular localization of VHL. COS7 cells or presence of increasing concentrations of VHL. were transfected with pCMXNHL and after 24 h of expression, the Immunoblot analysis demonstrated that wild-type cells were incubated for 6 h at normoxia or hypoxia. The subcellular localization was determined by indirect immunofluorescence using anti­ N-TAD protein was degraded in a dose-dependent manner VHL antibodies. (B) Subcellular distribution of GFP-HIF-lu chimeric by VHL (Figure 4C). In contrast, the stability of the mutant proteins. GFP fusion proteins spanning wild-type or mutant forms of N-TAD was not affected by identical concentrations of HIF-la. were transiently expressed in COS7 cells and incubated as VHL (Figure 4C). above. Photographs were taken using a Zeiss fluorescent microscope. To investigate the mechanism of VHL-mediated deg­ (C) Effect of VHL on degradation of wild-type and mutant forms of HIF-la. showing constitutively cytoplasmic or nuclear localization. radation of HIF- 1 a, we performed in vivo ubiquitylation FLAG-tagged wild-type or mutant forms of HIF-la. were transiently experiments. We transiently coexpressed the FLAG­ expressed in the absence or presence of VHL and incubated for 12 h at tagged wild-type or mutant N-TAD and HA-tagged normoxia or hypoxia. Whole-cell extracts were prepared and assayed as ubiquitin in the presence or absence of VHL in COS7 in Figure I. 4303 K.Tanimoto et al. GAUlHIF -la immunoblotting. Analogous to full-length HIF-la, the VHL : + + minimal wild-type GAL4/N-TAD fusion protein showed + + + Hypoxia (ha) 1 3 6 significant degradation under normoxic conditions, and -32 was stabilized by hypoxia. Interestingly, mutation of K532 VHL► -- -- IB: o.V HL stabilized the protein at normoxia, whereas expression of -26 ] .. , .. .... the two other lysine mutants was hardly detectable at normoxia (Figure 4E). These results suggest that K532 is critical for degradation of HIF-la. VHL► 1B: a. VHL 2 5 ],�, ..... , -32 Subce/lular localization of VHL at normoxia and ]1 0% Input VHL► - - -- 1B: o. VHL WC E hypoxia -25 We have previously demonstrated that hypoxia induces 1 2 3 4 5 6 nuclear translocation of HIF-la (Kallio et al ., 1998). In the case of VHL, nuclear-cytoplasmic trafficking has been � - __,, GAl41 H IF- 1 a. _ ,.. ::. ..,_ ,.,. .._ .:; ::..... suggested to be required for VHL function (Lee et al., Ar nt : + + + + • + + 1999). To study the intracellular localization of VHL in VHi. : + + + + + + relation to its function in normoxic versus hypoxic cells Hypoxia : • + + .. + - ♦ we transiently expressed VHL in COS7 cells. At -32 normoxia, immunofluorescence by an anti-VHL antibody VHL► ---- IB: o. VHl. was detected throughout the cells with some preference IP: a GAL4 -176 toward localization in the cytoplasmic compartment of the cells (Figure 5A). A very similar distribution of VHL . _ - -83 1B: a Arnt Arnt ► immunoreactivity was observed under hypoxic conditions (Figure 5A). Semiquantitative analysis of the subcellular -32 localization of VHL immunoreactivity (Kallio et al., 1998) VHL► ~- ---- 18: o. VHL under normoxic conditions revealed that 47 % of the ·26 transfected cells had equal distribution of fluorescence in .1 75 lO% input the cytoplasm and the nucleus, whereas in 45% of the -83 1B: a Arnt Arnt► -- -- -- WCE transfected cells cytoplasmic fluorescence predominated over that detected in the nucleus. No transfected cell -1 75 showed exclusive nuclear staining. Hypoxic treatment of GAL41HIF -1 a► • • . ... ,,.. 1B: o.H IF lo: the cells had no effect on the intracellular distribution of -83 1234 56789 VHL since under these conditions 47, 51 and 0% of the transfected cells fell into the three different categories, Fig. 6. VHL, Arnt and HIP-la. form a ternary nuclear complex. respectively. (A) VHL is not released from HIP-la. in hypoxic cells. GAIA, GAIA! HIP-la. and/or VHL were transiently coexpressed in COS? cells in the presence of MG 132 at normoxia or hypoxia for the indicated periods of Protection of HIF-1a from VHL- dependent time. Following immunoprecipitation of whole cell extracts by anti­ degradation is a multi-step pathway requiring GAIA (upper panel) or control (middle panel) antibodies, VHL was hypoxia-induced nuclear translocation of HIF-1a detected by immunoblotting using anti-VHL antibodies. Ten percent of and a hypoxia -dependent regulatory signal input whole-cell extracts is shown in the lower panel. (B) Association with Arnt. GAIA or GAIA/HIF-la. were transiently expressed in COS? We next compared the effect of VHL on the stability of cells in the absence or presence of Arnt or VHL at normoxia or wild-type HIF-la versus the stability of a HIF-la single hypoxia for 6 h, and analyzed as above. amino acid point mutant, HIF-la K719T, which fails to enter the nucleus at hypoxia and is constitutively localized in the cytoplasm (Kallio et al., 1998; Figure 5B). Under cells under normoxic conditions in the presence of normoxic conditions, both wild-type HIF-la and HIF-la the proteasome inhibitor MG 132. After anti-FLAG K719T were degraded in the presence of VHL (Figure 5C). immunoprecipitation of whole-cell extracts, ubiquitylated Interestingly, whereas wild-type HIF-la showed signifi­ N-T AD forms were detected by anti-HA immunoblotting cant (albeit not total) resistance to VHL under hypoxic analysis. Ubiquitylation of wild-type N-T AD was strongly conditions (Figures 1B and 5C), the HIF-la K719T enhanced by coexpression of VHL (Figure 4D). On the mutant was potently degraded upon exposure to VHL in other hand, the mutant N-TAD showed very low levels of hypoxic cells (Figure 5C). In conclusion, these results ubiquitylation even in presence of VHL. Importantly, suggest that hypoxia-induced nuclear translocation of these results clearly demonstrate that VHL mediates HIF-la protects HIF-la from VHL-mediated proteasomal ubiquitylation of HIF-la via physical interaction with degradation. the minimal N-TAD motif. To test whether nuclear localization is sufficient to Given the ubiquitylation of the minimal N-TAD motif, protect HIF-la against VHL-induced proteolysis we we next examined which of the three lysines of N-TAD examined the effect of VHL on the stability of a HIF-la were targeted for degradation at normoxia. FLAG-tagged mutant, HIF-la Li178-390, which shows constitutive wild type or single lysine mutants (K532R, K538R and nuclear localization (Kallio et al ., 1998; Figure 5B). K547R, respectively) of N-TAD were transiently expres­ Strikingly, this protein was degraded at normoxia upon sed in 293 cells. The cells were incubated at normoxia or exposure to VHL and showed hypoxia-dependent protein hypoxia for 12 h, and whole-cell extracts were analyzed by stabilization (Figure 5C). In biochemical experiments, 4304 Reg ulation of HIF-1a. by VHL A FLAG /HIF -11X wt mt GAL4/N-TAD wt mt VHL : - + + Hypox ia : + Hypoxia : + + -32 N-TAO ► HIF-1a ► 2 3 4 VHL ► -- -- -25 12 3 4 5 67 89 30T"" -------=- , -- ----� 10 .---------- --, ■ Hormoxla • Normoxill Hy po xl ■ D Hypoxia 7.5 ,, 5 2. 5 mt wt wt mt GAL4 GAL 41M-TAD GAL4 GAU/Hlf-1 ex Fig. 7. Mutant N-TAD maintains an hypoxia-inducible transactivation response. (A) Constitutive stabilization of PYI mutant N-TAD. GAL4-fused wild-type and mutant N-TAD were transiently expressed in C0S7 cells at normoxia or hypoxia for 12 h. Whole-cell extracts were prepared and assayed as in Figure 1. (B) Hypoxia-inducible transactivation function of mutant N-T AD. The transcriptional activation function by GAL4-fused wild­ type or mutant N-TAD proteins was monitored using a GAL4-responsive luciferase reporter gene as described in Materials and methods. Six hours after transfection, cells were incubated for 36 h at either normoxia or hypoxia. Reporter gene activities are expressed relative to the activity of the GAL4 DNA binding domain alone at normoxia, and normalized against internal control �-galactosidase activit ies. Values represent the mean ± SD of three independent experiments. (C) PYI mutant stabilizes full-length HIF- la. in normoxic cells. FLAG-tagged wild-type and mutant HIF- la were transiently expressed in absence or presence of VHL and analyzed as above. (D) Transcription activation functions of wild-type and mutant HIF- la proteins were assayed as in (B ). cytosolic or nuclear extracts were prepared from COS7 protect HIF-la, from VHL-mediated degradation, we next cells transiently expressing wild-type or mutant HIF-la, in examined the mechanistically important question of the absence or presence of VHL . HIF-la, and VHL co­ whether VHL was released from HIF- la, under hypoxic immunoprecipitation was detected only in nuclear extracts conditions. We transiently expressed GAL4/HIF-1 a, or the of hypoxic cells; whereas VHL was constitutively asso­ minimal GAL4 DNA binding domain in COS7 cells in the ciated with the nuclear HIF-la, A178-390 mutant, and no absence or presence of VHL under normoxic or hypoxic interaction was observed between HIF-la, K719T and conditions for increasing periods of time. In co-immuno­ VHL in nuclear extracts from either hypoxic or normoxic precipitation experiments using whole-cell extracts, VHL cells (data not shown). In summary, nuclear localization was detected by immunoblotting using anti-VHL anti­ per se cannot explain protection of HIF-la. against bodies. As expected, VHL was specifically co-immuno­ degradation by VHL. Our results suggest that, in addition precipitated together with GAL4/HIF-la, under normoxic to nuclear translocation, a distinct hypoxia-dependent conditions. However, VHL was also co-immunoprecipi­ intranuclear event or signal is required for stabilization of tated under hypoxic conditions at levels similar to those HIF-la.. Given the striking overlap within the N-TAD observed at normoxia (Figure 6A). These results suggest between the structures that mediate degradation, physical that dissociation of the HIF-la.-VHL complex is not interaction with VHL, and hypoxia-inducible transactiv­ necessary for protection of HIF-la. from VHL-mediated ation, it is possible that this putative intranuclear degradation, and that there exists a mechanism for stabilizing signal may be linked to the transactivation hypoxia-dependent inactivation of VHL function when function of the protein. remaining associated with HIF-la.. Next, we examined whether Arnt was associated with the HIF-la.-VHL complex. GAL4/HIF-la, or the minimal VHL is not released from HIF-1a in hypoxic cells GAL4 DNA binding domain transiently expressed in Since we found that both nuclear translocation of HIF-la, COS7 cells in the absence or presence of VHL and/or Arnt and a hypoxia-induced activation signal were necessary to 4305 K.Tanimoto et al. HYPOXIA NO RMOXIA jld omain CBD � domain CBD r--7 .----, .------, ,--------, Nuclear Pol y-Ubiqu itylation Transl ocation Arn t Cofactors Ref-1 !f 0 0. Proteas ome­ Target Gene Oo Mediated 0 oo O Acti vati on Degrada tion Fig. 8. Model of conditional regulation of HIF- la function. Under normoxic conditions (left panel) HIF- la is targeted for ubiquitin-proteasomal degradation by VHL. The shaded areas in VHL indicate mutational hotspots in tumors that coincide with the HIF- 1 a or the elongin C binding domain (CBD) of VHL, respectively. Mutations in either of these two domains stabilize HIF- la protein. Hypoxia (right panel) leads to inhibition of degradation of HIF- la by induction of nuclear translocation and an as yet unidentified nuclear regulatory signal that may be linked to recruitment of a partner DNA binding factor, Arnt, transcriptional coactivators and/or the redox regulator Ref -I. See text for details. at normoxia or hypoxia. As expected, Arnt was specifically those generated by wild-type GAL4/N-TAD under co-immunoprecipitated together with GAL4/HIF-la. under hypoxic conditions. In transactivation assays using a hypoxic conditions in the absence of VHL (Figure 6B). GAL4-driven luciferase reporter gene, we observed, as Moreover, in the presence of VHL, Arnt and GAL4/HIF-1 a. expected (Carrero et al ., 2000), rather modest (~3-fold) were also co-immunoprecipitated in a hypoxia-dependent activation of the fu nction of the minimal wild-type GAL4/ fashion (Figure 6B), demonstrating that these proteins N-T AD by hypoxia (Figure 7B). In the context of full­ formed a ternary complex in hypoxic cells. length HIF-la., mutation of the PYI motif rendered the protein stable under normoxic conditions and resistant to VHL-mediated degradation (Figure 7C), demonstrating Role of protein stabilization in regulation of the that this motif is the critical determinant for VHL­ hypoxia-dependent transactivation function of the dependent degradation of HIF-1 a.. Although the mutant HIF-1a N- TAD domain protein showed an increased constitutive transcriptional We and others have observed that transcriptional acti­ activity in comparison with wild-type HIF-1 a., it was still vation by HIF-la. in hypoxic cells is mediated by two hypoxia-inducible (Figure 7D). Within the minimal distinct transactivation domains, N- and C-TAD (Jiang N-TAD domain, mutation of the PYI motif reduced the et al ., 1997; Pugh et al ., 1997; Ema et al ., 1999; Carrero transactivation function with regard to both the activity et al., 2000). Given the fact that VHL interacted with the observed at normoxia and hypoxia (Figure 7B). These minimal N-TAD structure, we were interested to compare results indicate that the mutation may generally have under normoxic and hypoxic conditions both the stability altered the structure important for transactivation, possibly and transactivation functions of GAL4 fusion proteins impairing some of the protein contacts that may be harboring either the wild-type and/or the PYI mutant required for the full activation response. Nevertheless, this N-T AD motifs (Figure 4A). Whereas wild-type GAL4/ mutated construct was inducible by hypoxia, yielding an N-TAD showed hypoxia-dependent protein stabilization ~2-fold increase in transcriptional activity (Figure 7B). (Figures 4E and 7A), PYI mutant GAL4/N-TAD protein These data suggest that the stabilized protein resistant to levels were readily detectable in extracts from normoxic degradation by VHL is still capable of mediating a cells, and not significantly increased following exposure of hypoxia-dependent activation response, and that protein the cells to hypoxia (Figure 7 A). Thus, the failure of this stabilization per se does not bypass the need of the hypoxic mutant to interact with VHL (Figure 4B) correlated with signal for transactivation. constitutively stable protein expression levels similar to 4306 Regulation of HIF-1a by VHL This effect of hypoxia, in turn, correlates with the hypoxia­ Di scussion induced stabilization of HIF-la. protein levels (Kallio We have demonstrated that VHL interacts with HIF-la. to et al., 1999). Thus, HIF-la. is differentially regulated by mediate ubiquitin-proteasomal degradation of HIF-la. ubiquitylation, and inhibition of ubiquitylation constitutes under normoxic conditions. The HIF-la.-binding domain an early and critical step in hypoxia-dependent activation of VHL coincides with one of the hotspots of cancer­ of HIF-1 a, function. It is likely that these effects are causing mutations in the VHL gene (Kaelin and Maher, correlated with protection of HIF-la. against regulation by 1998). In fact, a frequent tumor mutation occurring within VHL. If this is the case, what is the mechanism that this region of VHL abrogated VHL-HIF-la, complex renders VHL-mediated degradation of HIF- la. inactive in formation, suggesting the importance of VHL-HIF-la. hypoxic cells? Obviously, this is a key question to interaction for the tumor suppressor function of VHL. understanding the mechanism of conditional regulation HIF-la, is among the most short-lived proteins currently of HIF-1 a, function. As schematically illustrated in known. The half-life of HIF-la, after exposure of cells to Figure 8, we show here that hypoxia-induced protection hypoxia and subsequent return to normoxia is in the range of HIF-1 a, against regulation by VHL involves two distinct of a few minutes (Wang et al., 1995). The present data and successive steps: nuclear translocation of HIF-1 a, and indicate that VHL functions by recruiting HIF-la, to the an intranuclear event or signal required for protection of VHL-BC--Cul-2 complex. Importantly, the naturally HIF-la, against VHL-induced proteolysis. Given the occurring C162F mutation of VHL fails to bind the striking overlap within the N-TAD of HIF-la. between elongin B and C complex, and has recently been demon­ the structures that mediate oxygen-dependent degradation, strated to abrogate VHL-dependent ubiquitylation activity physical interaction with VHL and hypoxia-inducible in vitro (Lisztwan et al., 1999). As schematically transactivation, it is possible that the putative intranuclear summarized in Figure 8, this mutant also failed to mediate stabilizing signal is linked to the transactivation fu nction degradation of HIF- la., strongly supporting the notion that of the protein. We and others have previously observed VHL regulates HIF-la, fu nction via targeting it for that the hypoxia-inducible fu nction of both the N-TAD ubiquitylation. and C-T AD is critically dependent on the recruitment of Degradation of a protein by the ubiquitin system the transcriptional coactivator CBP/p300 and SRC-1/p 160 involves two critical steps: covalent attachment of mul­ (Arany et al., 1996; Kallio et al., 1998; Ema et al., 1999; tiple ubiquitin molecules to the target protein, and Carrero et al., 2000). This recruitment appears to be degradation of the ubiquitin-tagged substrate by the 26S facilitated by the redox regulator Ref-1 (Ema et al., 1999; proteasome (reviewed by Ciechanover, 1998). Although Carrero et al., 2000). Thus, the functional architecture of the cascade of enzymatic pathways that mediate con juga­ the N-TAD encompasses overlapping structures that have tion of ubiquitin to its substrates has been rather well two different, in fact opposing functions, i.e. protein characterized (for recent reviews see Ciechanover, 1998; degradation versus activation of gene transcription, creat­ Hershko and Ciechanover, 1998), one of the central ing an important 'switch' in regulation of HIF-la. protein questions remains how proteins are selected for degrad­ function. The hypoxia-dependent intranuclear mechanism ation (reviewed by Laney and Hochstrasser, 1999). of protection may involve dimerization with Arnt, recruit­ Obviously, this process must be highly specific since ment of coactivators and/or recruitment of Ref -1 . short-lived proteins need to be identified and differentiated Importantly, the present data indicate that protein from more stable proteins within the cells. stabilization per se does not provide the sole basis for Earlier experiments have indicated that regulation of rendering HIF-la. transcriptionally active. It is possible HIF-la. protein levels by the proteasome pathway is that covalent modification of the C-terminus of HIF-la. mediated by a C-terminal structure of HIF-la. spanning may play a role in determining this regulatory effect. For PEST sequence motifs (Huang et al., 1998; Kallio et al., instance, it is also an attractive scenario that the hypoxic 1999). This structure has been termed the oxygen­ signal may determine a conformational change in HIF-la., dependent degradation domain (Huang et al., 1998), and which facilitates recruitment of the coactivators and notably, this region harbors the N-TAD domain of inactivates VHL function. HIF-la.. What is the degradation motif within the C-terminal region of HIF-la. that mediates regulation by VHL? Transplantable sequence elements, destruction Materials and methods boxes recognized and targeted by a proteolytic apparatus, Plasmid constructs have been identified in many short-lived proteins (Laney pFLAG CMV2/HIF-lal- 826, pCMX-GAL4/HIF-la and deletion and Hochstrasser, 1999) but there is still limited informa­ mutants thereof, and pCMV/Arnt have been described elsewhere tion with regard to the structural characteristics of (Kallio et al., 1997, 1998; Carrero et al., 2000), or were constructed by these elements. Here we demonstrate that VHL directly inserting relevant fragments of hHIF-la into pFLAG-CMV2 (Kodak). Amino acid substitutions in pFLAG or GAL4/HIF-la or NTAD were interacts with the N-TAD of HIF-la., and specifically generated using a QuikCbange site-directed mutagenesis kit (Stratagene ). ubiquitylates this structure. More specifically, point Expression plasmids for green fluorescent protein (GFP) fused to wild­ mutagenesis indicates that the highly conserved central type or mutant forms of HIP-la have been described previously (Kallio PYI motif of the N-TAD was critical for interaction et al., 1998). Wild-type or mutant forms of VHL and VHL GAL4 fusion proteins were constructed by inserting a blunted Nhel-Ecoru. fragment of with VHL and for VHL-induced ubiquitylation and pCI/VHL wild-type or mutant/FLAG (generously provided by Dr Joan proteasomal degradation. W.Conaway, Howard Hughes Medical Institute, USA) into EcoRV­ We have recently observed that exposure of cells to digested pCMX or pCMX-GAL4. The hemagglutinin (HA)-tagged hypoxia or a hypoxia-mimicking agent results in marked ubiquitin expression plasmid, pMT123, was kindly provided by reduction in recovery of ubiquitylated HIF-la, complexes. Dr D.Bohmann (European Molecular Biology Laboratory, Germany). 4307 K.Tanimoto et al. The FLAG-tagged dioxin receptor expression plasmid, pCMV/DR/ Acknowledgements FLAG, was obtained from Dr I.McGuire (Karolinska Institutet, Sweden). We thank Dr Joan W.Conaway for kindly providing wild-type and mutant VHL constructs, and Dr Dirk Bohmann for supplying the HA-tagged Cell culture and transient transfection experiments ubiquitin expression vector. This study was supported by grants from the COS? and 293 cells (obtained from ATCC) were routinely maintained in Swedish Medical Research Council, the JSPS Research Fellowships for Dulbec co 's minimal essential medium supplemented with 10% fetal calf Young Scientists (to K.T.), the Swedish Cancer Society (to Y.M.) and the serum plus penicillin (50 IU/ml) and streptomycin (50 µg/ml). For Fundaciio para a Ciencia ea Technologia (to T.P.). analysis of protein expression, we transiently transfected using the FuGene6 (Boehringer Mannheim) reagent wild-type or mutant forms of FLAG-tagged HIF-la (1 µg) and VHL (2 µg) into COS? cells in 6-cm References diameter plastic dishes. After 12 h of incubation, cells were treated for Arany,Z., Huang,L.E., Eckner,R., Bhattacharya,S., Jiang,C., 12 h at hypoxia (1 % 0 ) or normoxia (21 % 0 ). GAU reporter gene 2 2 Goldberg,M.A., Bunn,H.F. and Livingston,D.M. (1996) An essential assays were performed as described using a �-galactosidase expression role for p300/CBP in the cellular response to hypoxia. Proc. Natl plasmid as an internal control (Carrero et al., 2000). After 6 h of Acad. Sci. USA, 93, 12969-12973. transfection, cells were incubated for 36 h under hypoxic or normoxic Carrero,P., Okamoto,K., Coumailleau,P., O'Brien,S., Tanaka,H. and conditions prior to analysis of reporter gene activity. Poellinger,L. (2000) Redox-regulated recruitment of the transcriptional coactivators CBP and SRC-1 to the hypoxia­ lmmunoblotting and immunoprecipitation assays inducible factor-la. Mol. Cell. Biol., 20, 402-415. For the detection of HIF-la or VHL fusion protein expression whole-cell Ciechanover,A. (1998) The ubiquitin-proteasome pathway: on protein extracts were prepared as previously described (Kallio et al., 1997). death and cell life. EMBO J., 17, 7151-7160. Nuclear and cytosolic extracts were prepared as previously described Ema,M., Hirota,K., Mimura,J., Abe,H., Yodoi,J., Sogawa,K., (Gradin et al., 1996). Fifty micrograms of protein were blotted onto Poellinger,L. and Fu jii-Kuriyama,Y. (1999) Molecular mechanisms nitrocellulose filters following SOS-PAGE. Anti-FLAG M2 (Kodak), of transcription activation by HLF and HIFla in response to hypoxia: anti-VHL (PharMingen) or anti-Arnt (Kallio et al., 1997) antibodies were their stabilization and redox signal-induced interaction with CBP/ used as primary antibodies, diluted l :500 in TBS containing 0.1 % Tween- p300. EMB O J., 18, 1905-1914. 20 (TBS-T) and 1 % non-fat milk for 1 h. After several washes, a 1: 1000 Gnarra,J.R., Zhou,S., Merrill,M.J., Wagner,J.R., Krumm,A., dilution of anti-mouse lg-horseradish peroxidase con jugate (Amersham Papavassiliou,E., Oldfield,E.H., Klausner,R.D. and Linehan,W.M. Life Science) in TBS-T buffer containing 1% non-fat milk was used as a (1996) Post-transcriptional regulation of vascular endothelial growth secondary antibody and incubated with the sample for 1 h at room factor mRNA by the product of the VHL tumor suppressor gene. Proc. temperature. After extensive washing with TBS-T buffer, immuno­ Natl Acad. Sci. USA, 93, 10589-10594. complexes were visualized using enhanced chemiluminescence Gradin,K., McGuire,J., Wenger,R.H., Kvietik:ova,I., Whitelaw,M.L., (Amersham Pharmacia Biotech). Toftgard,R., Tora,L., Gassmann,M. and Poellinger,L. (1996) For in vivo immunoprecipitation experiments, we transiently expressed Functional interference between hypoxia and dioxin signal wild-type or mutant forms of FLAG-tagged or GAU-fused HIF-la, VHL transduction pathways: competition for recruitment of the Arnt and HA-tagged ubiquitin or Arnt (4 µg each) in COS? cells in 10-cm transcription factor. Mol. Cell. Biol., 16, 5221-5231. diameter plastic dishes. After 12 h of transfection, cells were treated for 1, Hershko,A. and Ciechanover,A. (1998) The ubiquitin system. Annu. Rev. 3 or 6 h under hypoxic or normoxic conditions before harvesting the cells. Biochem., 67, 425-479. The cell pellet was resuspended in 200 µl of whole-cell extract buffer Huang,L.E., Gu,J., Schau,M. and Bunn,H.F. (1998) Regulation of (Kallio et al., 1997) supplemented with 20 µM N-ethylmaleimide, -dependent hypoxia-inducible factor la is mediated by an O followed by centrifugation for 30 min at 14 000 r.p.m. For detection of degradation domain via the ubiquitin-proteasome pathway. Proc. co-immunoprecipitated proteins, cells were treated with 5 µM MG-132 Natl Acad. Sci. USA, 95, 7987-7992. (Calbiochem) for 6 h prior to harvest. Subsequent to protein extraction, Iliopoulos,O., Levy,A.P., Jiang,C., Kaelin,W.G.,Jr and Goldberg,M.A. 900 µg of total cell proteins were incubated with anti-FLAG antibodies at (1996) Negative regulation of hypoxia-inducible genes by the von room temperature for 1 h. Thirty microliters of a 50% slurry of protein G­ Rippel-Lindau protein. Proc. Natl Acad. Sci. USA , 93, 10595-10599. Sepharose in TEG buffer (20 mM Tris-HCl pH 7.4, 1 mM EDTA, 10% Iwai,K., Yamanaka,K., Kamura,T., Minato,N., Conaway,R.C., glycerol, 1 mM dithiothreitol), containing 150 mM NaCl and 0.1 % Triton Conaway,J.W., Klausner,R.D. and Pause,A. (1999) Identification of X-100, were then added to the reaction mixtures and incubated for 12 ha t the von Rippel-Lindau tumor-suppressor protein as part of an active 4 °C under rotation. After rapid centrifugation, the resulting Sepharose E3 ubiquitin ligase complex. Proc. Natl Acad. Sci. USA, 96, 12436-- pellets were washed three times with supplemented TEG buffer, and co­ immunoprecipitated proteins were analyzed by SOS-PAGE followed by Iyer,N.V. et al. (1998) Cellular and developmental control of 0 immunoblotting using anti-HA (Santa Cruz Biotechnology), anti-VHL or homeostasis by hypoxia-inducible factor la. Genes Dev., 12, 149- anti-Arnt antibodies diluted 1:500 in TBS-T and 1 % non-fat milk. In in vitro immunoprecipitation assays, GAU and FLAG fusion Jiang,B.H., Zheng,J.Z., Leung,S.W., Roe,R. and Semenza,G.L. (1997) proteins containing full-length or various deletion mutants of HIF-la, Transactivation and inhibitory domains of hypoxia-inducible factor and/or wild-type or mutant VHL proteins were translated either in the la. Modulation of transcriptional activity by oxygen tension. J. Biol. presence or absence of [ S]methionine in rabbit reticulocyte lysate Chem., 272, 19253-19260. (Promega). Labeled translation products were separated by SOS-PAGE Kaelin,W.G.,Jr and Maher,E.R. (1998) The VHL tumour-suppressor and analysed by phosphoimager (Fuji) for calculation of the protein gene paradigm. Trends Genet., 14, 423-426. concentration on the basis of incorporated [ S]methionine. Kallio,P.J., Pongratz,I., Gradin,K., McGuire,J. and Poellinger,L. (1997) Immunoprecipitation experiments were performed as previously Activation of hypoxia-inducible factor la: posttranscriptional described (Gradin et al., 1996) using GAU (Upstate Biotech) or FLAG regulation and conformational change by recruitment of the Arnt antibodies or pre-immune rabbit serum, and precipitated proteins were transcription factor. Proc. Natl Acad. Sci. USA, 94, 5667-5672. analyzed by SOS-PAGE followed by autoradiography. Kallio,P.J., Okamoto,K., O'Brien,S., Carrero,P., Makino,Y., Tanaka,H. and Poellinger,L. (1998) Signal transduction in hypoxic cells: lmmunostaining of VHL and visualization of intracellular inducible nuclear translocation and recruitment of the CBP/p300 coactivator by the hypoxia-inducible factor-la. EMBO J., 17, 6573- trafficking of GFP-tagged proteins in living cells COS? cells grown on cover slips were transiently transfected with VHL expression plasmid and immunostained, and analyzed as described Kallio,P.J., Wilson,W.J., O'Brien,S., Makino,Y. and Poellinger,L. (Kallio et al., 1998). COS? cells were transiently transfected with GFP­ (1999) Regulation of the hypoxia-inducible transcription factor la tagged protein expression plasmids, and after 24 h of expression the cells by the ubiquitin-proteasome pathway. J. Biol. Chem., 274, 6519--6525. were incubated under either normoxic or hypoxic conditions for 6 h. Laney,J.D. and Hochstrasser,M. (1999) Substrate targeting in the Subcellular distribution of fluorescence activity was examined using a ubiquitin system. Cell, 97, 427-430. Zeiss Axiovert 135 microscope with an FITC-filter set, and epifluores­Lee,S., Neumann,M., Stearman,R., Stauber,R., Pause,A., Pavlakis,G.N. cence with illumination from a Gixenon burner (Carl Zeiss Jena GmbH, and Klausner,R.D. (1999) Transcription-dependent nuclear­ Jena, Germany). cytoplasmic trafficking is required for the function of the von 4308 Regulation of HIF-1a by VHL Rippel-Lindau tumor suppressor protein. Mol. Cell. Biol. , 19, 1486- Lisztwan,J., Imbert,G., Wirbelauer,C., Gstaiger,M. and Krek,W. (1999) The von Rippel-Lindau tumor suppressor protein is a component of an E3 ubiquitin-protein ligase activity. Genes Dev. , 13, 1822-1833. Lonergan,K.M., Iliopoulos,O., Ohh,M., Kamura,T., Conaway,R.C., Conaway,J.W. and Kaelin,W.G.,Jr (1998) Regulation of hypoxia­ inducible mRNAs by the von Rippel-Lindau tumor suppressor protein requires binding to complexes containing elongins B/C and Cul2. Mol. Cell. Biol. , 18, 732-741. Maxwell,P.H. et al. (1999) The tumour suppressor protein VHL targets hypoxia-inducible factors for oxygen-dependent proteolysis. Natu re, 399, 271-275. Mukhopadhyay,D., Knebelmann,B., Cohen,H.T., Ananth,S. and Sukhatme,V.P. (1997) The von Rippel-Lindau tumor suppressor gene product interacts with Sp 1 to repress vascular endothelial growth factor promoter activity. Mol. Cell. Biol. , 17, 5629-5639. Pugh,C.W., O'Rourke,J.F., Nagao,M., Gleadle,J.M. and Ratcliffe,P.J. (1997) Activation of hypoxia-inducible factor- I; definition of regulatory domains within the a subunit. J. Biol. Chem. , 272, 11205-11214. Salceda,S. and Caro,J. (1997) Hypoxia-inducible factor la (HIF-la) protein is rapidly degraded by the ubiquitin-proteasome system under normoxic conditions. Its stabilization by hypoxia depends on redox­ induced changes. J. Biol. Chem. , 272, 22642-22647. Srinivas,V., Zhang,L.P., Zhu,X.H. and Caro,J. (1999) Characterization of an oxygen/redox-dependent degradation domain of hypoxia­ inducible factor la (HIF-la) proteins. Biochem. Biophys. Res. Com mun. , 260, 557-561. Stebbins,C.E., Kaelin,W.G.,Jr and Pavletich,N.P. (1999) Structure of the VHL-ElonginC-ElonginB complex: implications for VHL tumor suppressor function. Science, 284, 455-461. Taylor,B.L. and Zhulin,1.B. (1999) PAS domains: internal sensors of oxygen, redox potential, and light. Microbiol. Mol. Biol. Rev. , 63, 479-506. Thrash-Bingham,C.A. and Tartof,K.D. (1999) aHIF: a natural antisense transcript overexpressed in human renal cancer and during hypoxia. J. Natl Cancer Inst. , 91, 143-151. Wang,G.L., Jiang,B.H., Rue,E.A. and Semenza,G.L. (1995) Hypoxia­ inducible factor 1 is a basic-helix-loop-helix-PAS heterodimer regulated by cellular 0 tension. Proc. Natl Acad. Sci. USA , 92, 5510-5514. Wenger,R.H. and Gassmann,M. (1997) Oxygen(es) and the hypoxia­ inducible factor- I. Biol. Chem. , 378, 609---616. Received April 5, 2000; accepted June 30, 2000 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The EMBO Journal Springer Journals

Mechanism of regulation of the hypoxia‐inducible factor‐1α by the von Hippel‐Lindau tumor suppressor protein

Download PDF
12 pages

Loading next page...
 
/lp/springer-journals/mechanism-of-regulation-of-the-hypoxia-inducible-factor-1-by-the-von-2sdGDAkwgs

References (34)

Publisher
Springer Journals
Copyright
Copyright © European Molecular Biology Organization 2000
ISSN
0261-4189
eISSN
1460-2075
DOI
10.1093/emboj/19.16.4298
Publisher site
See Article on Publisher Site

Abstract

The EMBO Journal Vol.19 No.16 pp.4298--4309, 2000 Mechanism of regulation of the hypoxia-inducible factor-1a by the von Hippel-Lindau tumor suppressor protein et al., 1996; Iliopoulos et al., 1996) and/or transcriptional Keiji Tanimoto, Vuichi Makino, (Mukhopadhyay et al., 1997) mechanisms. Teresa Pereira and Lorenz Poellinger The VHL protein displays no sequence similarity to Department of Cell and Molecular Biology, Medical Nobel Institute, other known proteins, thus giving no clues about its Karolinska Institutet, S-171 77 Stockholm, Sweden function. Biochemical studies have shown that VHL is Corresponding author associated with elongins B and C, and cullin-2 (Cul-2) e-mail: [email protected] (Kaelin and Maher, 1998), forming the VHL-BC--Cul-2 complex. The crystal structure of the VHL-BC ternary In normoxic cells the hypoxia-inducible factor-la complex shows two interfaces: one between VHL and (IIlF-la.) is rapidly degraded by the ubiquitin-protea­ elongin C and another between elongins B and C (Stebbins some pathway, and activation of HIF-la. to a func­ et al., 1999). The 35 residue-long elongin C binding tional form requires protein stabilization. Here we domain of VHL represents one of the mutational hotspots show that the product of the von Hippel-Lindau in tumors (Kaelin and Maher, 1998), suggesting that (VHL) tumor suppressor gene mediated ubiquityla­ VHL-BC complex formation is critical for tumor sup­ tion and proteasomal degradation of HIF-la under pressor function. In addition, there is a mutational hotspot normoxic conditions via interaction with the core of on a separate domain, the � domain of VHL (Kaelin and the oxygen-dependent degradation domain of IIlF-la. Maher, 1998), which overlaps with a putative macro­ The region of VHL mediating interaction with HIF-la molecular binding site identified in the VHL-BC crystal overlapped with a putative macromolecular binding structure (Stebbins et al., 1999). Elongins B and C and site observed within the crystal structure of VHL. Cul-2 all share homology to components of the SCF This motif of VHL also represents a mutational hot­ (Skpl--Cul-1-F-box protein) multiprotein complex, which spot in tumors, and one of these mutations impaired targets cell cycle regulatory proteins for ubiquitin-medi­ interaction with HIF-la and subsequent degradation. ated proteolysis (Ciechanover, 1998). Importantly, the Interestingly, the VHL binding site within HIF-la structure of the VHL-BC complex extends these similar­ overlapped with one of the minimal transactivation ities to the SCF complex structure (Stebbins et al., 1999), domains. Protection of HIF-la against degradation by indicating that these protein complexes may have similar VHL was a multistep mechanism, including hypoxia­ functions. In excellent agreement with this model, VHL induced nuclear translocation of HIF-la and an intra­ has recently been shown to be associated with an E3 nuclear hypoxia-dependent signal. VHL was not ubiquitin ligase activity in cellular extracts (lwai et al., released from HIF -la during this process. Finally, 1999; Lisztwan et al., 1999). stabilization of HIF-la protein levels per se did not In hypoxic cells, the hypoxia-inducible factor-la. totally bypass the need of the hypoxic signal for gener­ (HIF-la.) mediates transcriptional activation of the ating the transactivation response. VEGF gene (Iyer et al., 1998). HIF-la. mRNA is Keywords: hypoxia-inducible factor-la/transcription/ constitutively expressed in a number of mammalian cells tumor suppression/ubiquitylation/von Rippel-Lindau (Kallio et al., 1997). In contrast, the HIF-la. protein is protein remarkably unstable in cells at normoxia, whereas hypoxia dramatically stabilizes the protein (Kallio et al., 1997). Under normoxic conditions HIF-la. protein degradation is mediated by the ubiquitin-proteasome pathway and a Introduction distinct oxygen-dependent degradation domain of HIF-la. von Rippel-Lindau (VHL) disease is caused by germ line (Huang et al., 1998; Kallio et al., 1999). Stabilization of mutations of the VHL gene. These mutations lead to the HIF-la. initiates a multi-step pathway of activation of development of a variety of tumors including clear cell HIF-la. that includes hypoxia-dependent nuclear trans­ carcinomas of the kidney, pheochromocytomas and vas­ location and dimerization with a partner DNA binding cular tumors of the central nervous system and retina. factor, Arnt, to interact with cognate hypoxia-response VHL-associated neoplasms are typically hypervascular elements of target promoters, followed by recruitment of (reviewed by Kaelin and Maher, 1998), and under transcriptional coactivators (Wenger and Gassmann, 1997; normoxic conditions VHL-deficient cells express vascular Kallio et al., 1998; Ema et al., 1999). Recently, two studies endothelial growth factor (VEGF) mRNA, which is have indicated distinct roles of VHL in regulation of normally expressed in a hypoxia-dependent fashion HIF-la. function: induction of a natural HIF-la. antisense (Gnarra et al., 1996; Iliopoulos et al., 1996). Reintro­ transcript in VHL-deficient cells, resulting in negative duction of VHL into VHL-mutated renal carcinoma cells regulation of HIF-la. function (Thrash-Bingham and indicates that it functions as a negative regulator of Tartof, 1999). On the other hand, VHL has recently been VEGF mRNA levels by post-transcriptional (Gnarra reported to interact physically with HIF-la., possibly 4298 © European Molecular Biology Organization Regulation of HIF-1a. by VHL HIF-1 o:: : + ++ +++ VHL + + Hypoxia: + + Hypoxia: + + + -175 -175 HIF-1a►-- -83 -83 -32 1 2 3 4 5 6 VHL► -25 1 2 3 4 VHL : VHL: + + + -83 MG-132: + + DR► -175 -62 HIF-1o:► -32 VHL► -83 1 2 3 4 2 1 2 Fig. 1. VHL mediates proteasomal degradation of HIP-la under normoxic conditions. (A) Expression of HIP-la under normoxic conditions. Increasing amounts [0.2 µg (+), 0.5 µg (++), 1.0 µg (+++)] of pFLAG CMV2/HIF-la were transiently transfected into COS7 cells, and the cells were incubated for 12 h at normoxia (21 % 0 ) or hypoxia (1 % 0 ), as indicated. Whole-cell extracts were analyzed by immunoblotting using anti-FLAG 2 2 antibodies. (B) Degradation of HIP-la in the presence of VHL. pFLAG CMV2/HIF-la was cotransfected into COS7 cells together with empty vector or wild-type VHL expression vector (pCMX/VHL). Whole-cell extracts were analyzed by immunoblotting using anti-FLAG or anti-VHL antibodies. (C) VHL mediates proteasomal degradation of HIP-la. Cells were transfected with pFLAG CMV2/HIF-la and pCMX/VHL, incubated in the absence or presence of 5 µM MG-132 for 6 h before harvesting, and cellular extracts were analyzed as in (B). (D) VHL does not affect dioxin receptor (DR) protein levels. COS7 cells were transfected with a FLAG-tagged dioxin receptor expression vector (pCMV/DR/FLAG) in the absence or presence of pCMX/VHL, and analyzed as in (B). The mobilities of molecular weight (kDa) markers are shown on the right hand side of the blots. targeting HIF-la for protein degradation (Maxwell et al., genes such as VEGF are constitutively expressed at 1999). Here we show that VHL directly mediated normoxia in VHL-deficient cells (Gnarra et al., 1996; ubiquitylation and ensuing proteasomal degradation of Iliopoulos et al., 1996), indicating dysregulation of HIF-la at normoxia via physical interaction with the core HIF-la function in these cells. We were thus interested of the oxygen-dependent degradation domain. This motif to investigate the potential mechanism of regulation of coincided with one of the minimal transactivation domains HIF-la function by VHL. Due to the pronounced !ability of HIF-1 a. The domain mediating interaction with HIF-1 a of the HIF-la protein under normoxic conditions, HIF-la also represented one of the mutational hotspots of VHL. is normally not detectable by immunoblot analysis of Protection against VHL-mediated degradation required cellular extracts (Kallio et al., 1997, 1998). It was both nuclear translocation of HIF-la and an intranuclear therefore not possible to investigate the effect of expres­ hypoxia-dependent regulatory signal. Finally, stabilization sion of VHL on endogenous HIF-la protein levels. To of HIF-la protein levels per se did not bypass the need of establish experimental conditions to examine the effect of the hypoxic signal for generating the transactivation VHL on HIF-la protein stability, we transiently trans­ response. fected COS7 cells with FLAG epitope-tagged HIF-1 a expression plasmids. As expected, at a low concentration (0.2 µg) of expression vector, HIF-la was not detected at Results normoxia, and we observed potent stabilization ofHIF-la protein levels at hypoxia, as assessed by immunoblot Regulation of HIF-1a protein stability by the VHL analysis (Figure lA). However, at higher concentrations tumor suppressor protein (0.5-1 µg) of expression vector we could detect HIF-la We and others have recently demonstrated that HIF-la is protein expression also under normoxic conditions regulated by the ubiquitin-proteasome pathway under (Figure lA). At the highest concentrations of expression normoxic conditions, resulting in very rapid turnover of vector tested we observed significant HIF-1 a expression the protein, and that one of the early responses to hypoxia levels both at normoxia and hypoxia (Figure lA). Thus, is massive upregulation of HIF-la protein levels (Kallio these experiments suggest that the mechanism of degrad­ et al., 1997; Salceda and Caro, 1997; Huang et al., 1998; Kallio et al., 1999). Interestingly, VHL protein complexes ation of HIF-1 a had become saturated under these have recently been demonstrated to harbor E3 ubiquitin­ conditions and that one or several components of the protein ligase activity, although the target protein for this degradation machinery were limiting. activity has not yet been identified (Lisztwan et al., 1999; We next used the high level HIF-la expression Iwai et al., 1999). Moreover, HIP-la-regulated target conditions for all subsequent experiments to examine the 4299 GAL41VHL ti114-154 _______ o __ _,l K.Tanimoto et al. effect of VHL on HIF-la protein levels. Transient inhibitor MG-132 (Figure IC). Taken together, these coexpression of FLAG/HIF-la and VHL resulted in results strongly suggest that VHL mediates proteasomal degradation of HIF-1 a. This effect of VHL was specific reduction of the HIF-la protein signal under normoxic for HIF-la, as transiently expressed VHL did not produce conditions (Figure IB, upper panel), indicating that VHL this effect on FLAG-tagged dioxin receptor (Figure ID), a may have been limiting under the conditions of expression basic helix-loop-helix(bHLH)/P AS (Per/ Arnt/Sim do­ of HIF-la alone. Interestingly, VHL failed to induce main) protein belonging to the same class of transcription reduction of HIF-la protein levels under hypoxic condi­ factors as HIF-la. tions (Figure IB). In control experiments we detected similar levels of VHL expression in extracts from either Two domains of VHL are required for inducing normoxic or hypoxic cells (Figure IB, lower panel). VHL­ protein degradation of HIF-1a induced reduction of HIF-la protein levels at normoxia Given the potential role of VHL as an E3 ubiquitin ligase was inhibited by treatment of the cells with the proteasome we examined whether VHL physically interacted with HIF-la. S-labeled, in vitro translated HIF-la was incubated with wild-type or mutant GAL4/VHL fusion proteins (schematically represented in Figure 2A) or the elo oglo C minimal GAL4 DNA binding domain alone prior to blndl np GAL41VHLwt II i immunoprecipitation assays. In these experiments, 11' 1H 2U S-labeled HIF-la was co-immunoprecipitated in the __,r---{ ._. 1 110 1H ZU presence of GAL4/VHL by anti-GAL4 specific antibodies, whereas no interaction was observed between HIF-la and GAL4/VHL 114-154 � 114 lk the minimal GAL4 DNA binding domain (Figure 2B, GAL4/VHL 91-154 upper panel). Non-specific pre-immune rabbit antiserum did not precipitate HIF-la protein in the presence of either GAL4/VHL pmt1 213 VHL or GAL4 alone (Figure 2B, lower panel), indicating that wild-type VHL specifically interacted with HIF-1 a B in vitro. GAL4/VHL '5 The VHL M 14-154 deletion mutant showed interaction Cl, ::z ..,. U'I with HIF-la, whereas the VHL 114-154 fragment failed .E ::z ..,. :. '-' ... ":- :. to do so (Figure 2B). A 23 amino acid-long N-terminal Q ... ... i <i ... ;; � extension of this fragment generated VHL 91-154, which ·175 was able to interact with HIF-la, indicating the import­ HIF-1a► o.GAL4 ance of a structure located between residues 91 and 113 of VHL to interact with HIF-la. Interestingly, this region of -89 VHL is not only contained within the putative macro­ molecular binding site observed in the crystal structure of HIF•1o.► Control the VHL-BC complex (Stebbins et al., 1999), but also -89 represents one of the mutational hotspots in tumors 2 9 4 s 6 (Kaelin and Maher, 1998). This fact prompted us to examine whether tumor-derived mutations of VHL would '5 GAL4/VHL affect its ability to interact with HIF-1 a and/or to induce HIF-la degradation. We performed these experiments :s IO ◄ ... .. CJ 0 i >- ·17S o. GAL4 HIF-1a ► Fig. 2. VHL requires two functional domains to induce HIF-lo. degradation. (A) Schematic representation of GAL4-fused wild-type - 8 (wt), deletion or single amino acid point mutant (pmt) forms of VHL . • 175 (B) VHL directly interacts with HIF-lo.. Equal concentrations of Control HIF•1a ► in vitro translated S-labeled full-length HIF-lo. were incubated with in vitro translated wild-type GAL4NHL or VHL deletion mutants or -83 2 3 4 5 GAL4-DBD spanning the GAL4 DNA binding domain alone. Co­ irnmunoprecipitation assays with anti-GAL4 antibody (upper panel) or control preimmune serum (lower panel) were carried out as described GAL41VHL in Materials and methods. The precipitated material was analysed by LI. SOS-PAGE and autoradiography. For loading controls, 10% of UI 35 input S-labeled HIF-lo. is shown in lane 1. (C) A tumor-derived CJ i ! 0 point mutation impairs the interaction between VHL and HIF-la. -175 In vitro translated proteins were incubated and analyzed by co­ immunoprecipitation as in (B). (D) Tumor-derived point-mutated forms of VHL fail to induce HIF-la degradation. pFLAG CMV2/HIF-la was HIF-1a► -83 transiently coexpressed in COS7 cells in the absence or presence of GAL4-fused wild-type or mutant forms of VHL as indicated. Cells were incubated at normoxia for 24 h, and whole cell extracts were -50 prepared and analyzed by immunoblotting using anti-FLAG or anti­ VHL► VHL antibodies. The mobilities of molecular weight (kDa) markers 2 3 4 are shown on the right hand side of the blots. 4300 Regulation of HIF-1a. by VHL VHL: !21212 -175 PA.S N-TAD C-T.AD ,.--, FLAG/HIF-1o. 11111 II a n I I I 1 532 585 82& -8S ► - FLAG/HIF-10.1-652 II n -8S Ell I 1 &52 -47.5 ►- FLAG/HIF-1o. 1-330 11111 II IEI n II 1 330 -32 FLAG/HIF-1cx526-826 WWW -47.5 t I 526 828 B C GAL41HIF-1a a. .E .E 'It ...I 'i/- 'i/- ... < <E 0 0 ; C") ... ... ... CJ C") c:, 3 3 2 2 a-GAL4 a-GAL4 VHL ► VHL ► 5 - ---- ----- ------- 3 · 2 3 Control Control VHL ► VHL ► 2 - 5 2 1 2 3 1 2 3 4 5 6 7 8 Fig. 3. VHL targets the oxygen-dependent degradation domain of HIP-la.. (A) FLAG-tagged wild-type HIP-la. or the indicated HIP-la. deletion mutants were transiently coexpressed in COS? cells at normoxia in the absence or presence of VHL. Whole-cell extracts were prepared and analyzed as described in Figure 1. ODD, oxygen-dependent degradation domain; N- and C-TAD, N- and C-terminal transactivation domains. (B) S-labeled VHL was incubated with equal concentrations of in vitro translated GAL4-fusion proteins spanning full-length HIF-la. or GAL4-DBD alone. Co-immunoprecipitation assays were performed with anti-GAL4 antibodies (upper panel) or control non-specific rabbit antiserum (lower panel). The precipitated material was analyzed by SDS-PAGE and autoradiography. For loading controls, 10% of input S-labeled VHL is shown. (C) VHL, GAL4 fusion proteins spanning the indicated HIP-la. fragments or GAL4 alone were expressed by in vitro translation and analyzed as in (B). using GAL4NHL fusion proteins containing either a mediate degradation of HIF-la, and that regulation of Y98N (the most frequent tumor mutation in this region; HIF-la may be involved in the tumor suppressor fu nction Kaelin and Maher, 1998) or a C162F single amino acid of VHL. mutation. The C162F mutation has been demonstrated to render VHL unable to bind the elongin B-C complex The oxygen-dependent degradation domain of (Lonergan et al., 1998; Lisztwan et al., 1999), and inhibit HIF-1a is targeted for regulation by VHL ubiquitin ligase activity in vitro (Lisztwan et al., 1999). In To identif y the domain of HIF-1 a that is targeted by VHL co-immunoprecipitation experiments, VHL Y98N was to mediate proteasomal degradation at normoxia, we unable to interact with HIF-la, whereas VHL C162F transiently expressed in COS7 cells in the presence or showed wild-type levels of interaction with HIF-la absence ofVHL either wild-type FLAG/HIF-la or a series (Figure 2C, upper panel). In our cellular degradation of FLAG-tagged HIF-la deletion mutants. In analogy to assay we transiently expressed at normoxia FLAG/HIF-la wild-type HIF-la, HIF-la 1-652 lacking the C-terminus in the presence or absence of wild-type or the individual including the C-terminal transactivation domain (sche­ point-mutated forms of VHL. Immunoblot analysis dem­ matically represented in Figure 3A) was degraded in the onstrated that, in contrast to wild-type VHL, both the VHL presence of VHL. However, the protein levels of HIF-la Y98N and VHL C162F mutants failed to induce degrad­ 1-330 lacking structures C-terminal of the PAS domain ation of HIF-la at normoxia (Figure 2D). These results were not affected by VHL. HIF-la 526-826 lacking demonstrate that both the HIF-la interaction domain and N-terminal structures (including the bHLH and PAS the elongin C binding domain of VHL are necessary to domains) was also degraded upon exposure to VHL at 11111 K.Tanimo to et al. A D 1B: a HA 522 541 N-TAO wtN •TAO mt mEPAS -1 ELDLETLAPY IP MDGEDFQL 523 542 VHL: + + + hEPA.S-1 ELDLE TL APYI PMDGEGPQL 555 HA-Ub : + + + + + mHIF-1 a. DLDLEMLAPY IP MD- DDFQI, hH IF-1cr. DLDLBMLAPY IP MD-D DPQL -175 Ubil uitylated [ - I I I -83 N-T. 0 hH IF-1cr. mt DDD IP: o.F LAG -62 lg► B -47.5 a. -3 .E /N CJ '#-- 4:( .... � i e -32 N -TA D► o:-FLAG . VHL WCE -2 1 2 3 4 2 VHL ► -2 1 2 3 4 5 N-TA D wt N-TAD rnt + ++ + ++ VHL: 3 N •TAD N-T AD ► wt K532R K538R K547R - - 2 Hypox ia : + + + + VHL ► N-TAD ► ·2 1 2 3 4 5 6 7 1 2 3 4 5 6 7 8 Fig. 4. The minimal N-T AD of HIF- lu is the target for VHL-mediated ubiquitylation and proteasomal degradation. (A) Alignment of N-T AD sequences of human (h) and mouse (m) HIF- lu and EPAS -1 revealed a conserved core sequence motif. The positions of amino acid substitutions are shown in mutant (ml) N-TAD. (B) Mutation of the N-TAD PYI motif abolishes interaction between HIF- lu and VHL. S-labeled VHL was incubated with equal concentrations of in vitro translated FLAG-tagged wild-type, mutant N-TAD or FLAG epitope alone. Co-immunoprecipitation assays were carried out using anti-FLAG antibodies and analyzed by SOS- PAGE and autoradiography. For loading controls, 10% of input S-la beled VHL is shown. (C) The PYI mutation confers resistance to VHL-mediated degradat ion. FLAG or FLAG-tagged wild-type or mutant N-TAD (I µg/ 6-cm dish) were transiently coexpressed in COS7 cells in the absence or presence of increasing concentrations (1.0 µg, +; 2.0 µg/6-cm dish, ++) of VHL as indicated, and incubated at normoxia for 24 h. Whole-cell extracts were prepared and analyzed by immunoblotting using anti-FLAG or anti­ VHL antibod ies. (D) VHL mediates ubiquitylation of N-TAD. FLAG, or FLAG-tagged wild-type or mutant N-TAD and HA-tagged ubiquitin were transiently coexpressed in COS7 cells in the presence or absence of VHL under normoxic conditions in combination with MG 132 for 6 h. After immunoprecipitation of whole-cell extracts by anti-FLAG, ubiquitylated forms of N-T AD were detected by anti-HA immunoblotting (upper panel). Ten percent of input whole-cell extracts are shown in the two lower panels. (E) Effect on protein stability following substitution of individual lysines to arginines within N-TAD. Wild-type or mutant N-TAD were transiently expressed in 293 cells and incubated at normoxia or hypoxia for 12 h. Whole-cell extracts were prepared and analyzed as in (C). The mobilities of molecular weight (kDa) markers are shown to the right of the blots. (Figure 3C, upper panel). Moreover, GAL4/HIF-la 778- normoxia (Figure 3A). In conclusion, these results indicate that a C-terminal region of HIF-1 a spanning residues 526- 826, spanning the C-terminal transactivation domain of 652 mediated VHL-dependent degradation. HIF-la, did not show any interaction with VHL We proceeded to map the domain of HIF-1 a required to (Figure 3C, upper panel). In contrast, the GAL4/HIF-la interact with VHL. To this end we used fu ll-length HIF-la fragments, which showed VHL-mediated degradation in or a set of HIF-1 a deletion mutants fused to the normoxic cells, HIF-la 1-652 and HIF-la 526-826 GAL4 DNA binding domain, and performed co-immuno­ (Figure 3A), clearly interacted with VHL (Figure 3C). In precipitation assays following incubation with VHL. As control reactions, the GAL4/HIF-la fragments were expected, S-labeled VHL was specifically co-immuno­ expressed at similar levels (data not shown), and non­ precipitated together with full-length HIF-la (Figure 3B). specific pre-immune rabbit antiserum did not precipitate In excellent agreement with the fact that the deletion VHL protein in the presence of any of the used fragments mutant HIF-la 1-330 was not degraded upon over­or GAL4 alone (Figure 3B and C, lower panels). In a more expression of VHL in COS7 cells under normoxic detailed analysis of the interaction of VHL with HIF-la, conditions (Figure 3A), GAL4/HIF-la 1-330 failed VHL was co-immunoprecipitated by anti-GAL4 anti­ to interact physically in vitro with S-labeled VHL bodies in the presence of GAL4/HIF-la 331-641 or F:LA G TAD Reg ulation of HIF-1a. by VHL GAL4/HIF-la 526-641. In fact, when compared with full­ length HIF-la, all these latter fragments of HIF-la interacted with VHL with very similar efficacies (Figure 3C, upper panel). Taken together, these results indicate that a region of HIF-la spanning residues 526- VHL 641 was essential for physical interaction with VHL. Interestingly, this region overlaps with the oxygen/redox­ dependent degradation domain of HIF-la, which has 10µ m previously been demonstrated to mediate proteasomal degradation of HIF-1 a in normoxic cells, and that has broadly been defined to be located between amino acid residues 401 and 603 of hHIF-la (Huang et al., 1998; Kallio et al., 1999). The minimal N-terminal transactivation domain of GFP-HIF-1 a HIF-1a is a target for ubiquitylation and proteasomal degradation by VHL Interestingly, the VHL-interacting fragment GAL4/ HIF-la 526-641 contains not only the core of the oxygen-dependent degradation domain of HIF-1 a but GFP-HIF-1 cx/ also the N-terminal transactivation domain N-TAD (Jiang K71 9T et al., 1997; Pugh et al., 1997; Figure 3A). Within the N-TAD of HIF-la a sequence motif of ~19 amino acid residues (located between amino acids 556 and 574 of hHIF-la) shows the strongest conservation between species and is also conserved in the related hypoxia­ GFP-HIF-1 a/ inducible factor EP AS-1/HLF (Figure 4A), which is �1 78-390 expressed in a tissue-restricted fashion. This motif is also highly conserved in the hypoxia-responsive Droso phila Similar protein (reviewed by Taylor and 10 µm Zhulin, 1999). Interestingly, this sequence motif has recently been reported to be important for the function of the oxygen-dependent degradation domain of HIF-la since alanin substitutions within this domain impair VHL : + + hypoxia-dependent protein stabilization (Srinivas et al., Hypoxia : + + 1999). We initially examined whether VHL could interact 17 5 directly with the minimal N-TAD located between amino HIF 1o.► - acids 532 and 585 of hHIF-la (Jiang et al., 1997; Pugh et al., 1997). In co-immunoprecipitation assays, anti­ 8 -175 FLAG antibodies could precipitate S-labeled VHL in the presence of the FLAG epitope-tagged minimal N-T AD of HIF-1a K71 9T► HIF-la (Figure 4B), demonstrating that this region of -8 HIF-la is sufficient to mediate interaction with VHL. We next generated a point mutation within the context of the HIF-fo 6.1 78-390 ► minimal N-TAD. As assessed in co-immunoprecipitation assays, substitution of the central PYI triplet with aspartic acid totally abolished interaction between VHL and the - VHL► N-TAD (Figure 4B). In conclusion, these experiments -2. 5 demonstrate that the highly conserved core motif of the 1 2 3 4 N-T AD is critical for interaction with VHL. Fi . 5. Protection of HIF-lu against VHL-mediated degradation We next transiently expressed under normoxic con­ requires nuclear translocation of HIF-1 a. and a hypoxia-dependent ditions the wild-type or mutant N-TAD in the absence intranuclear signal. (A) Subcellular localization of VHL. COS7 cells or presence of increasing concentrations of VHL. were transfected with pCMXNHL and after 24 h of expression, the Immunoblot analysis demonstrated that wild-type cells were incubated for 6 h at normoxia or hypoxia. The subcellular localization was determined by indirect immunofluorescence using anti­ N-TAD protein was degraded in a dose-dependent manner VHL antibodies. (B) Subcellular distribution of GFP-HIF-lu chimeric by VHL (Figure 4C). In contrast, the stability of the mutant proteins. GFP fusion proteins spanning wild-type or mutant forms of N-TAD was not affected by identical concentrations of HIF-la. were transiently expressed in COS7 cells and incubated as VHL (Figure 4C). above. Photographs were taken using a Zeiss fluorescent microscope. To investigate the mechanism of VHL-mediated deg­ (C) Effect of VHL on degradation of wild-type and mutant forms of HIF-la. showing constitutively cytoplasmic or nuclear localization. radation of HIF- 1 a, we performed in vivo ubiquitylation FLAG-tagged wild-type or mutant forms of HIF-la. were transiently experiments. We transiently coexpressed the FLAG­ expressed in the absence or presence of VHL and incubated for 12 h at tagged wild-type or mutant N-TAD and HA-tagged normoxia or hypoxia. Whole-cell extracts were prepared and assayed as ubiquitin in the presence or absence of VHL in COS7 in Figure I. 4303 K.Tanimoto et al. GAUlHIF -la immunoblotting. Analogous to full-length HIF-la, the VHL : + + minimal wild-type GAL4/N-TAD fusion protein showed + + + Hypoxia (ha) 1 3 6 significant degradation under normoxic conditions, and -32 was stabilized by hypoxia. Interestingly, mutation of K532 VHL► -- -- IB: o.V HL stabilized the protein at normoxia, whereas expression of -26 ] .. , .. .... the two other lysine mutants was hardly detectable at normoxia (Figure 4E). These results suggest that K532 is critical for degradation of HIF-la. VHL► 1B: a. VHL 2 5 ],�, ..... , -32 Subce/lular localization of VHL at normoxia and ]1 0% Input VHL► - - -- 1B: o. VHL WC E hypoxia -25 We have previously demonstrated that hypoxia induces 1 2 3 4 5 6 nuclear translocation of HIF-la (Kallio et al ., 1998). In the case of VHL, nuclear-cytoplasmic trafficking has been � - __,, GAl41 H IF- 1 a. _ ,.. ::. ..,_ ,.,. .._ .:; ::..... suggested to be required for VHL function (Lee et al., Ar nt : + + + + • + + 1999). To study the intracellular localization of VHL in VHi. : + + + + + + relation to its function in normoxic versus hypoxic cells Hypoxia : • + + .. + - ♦ we transiently expressed VHL in COS7 cells. At -32 normoxia, immunofluorescence by an anti-VHL antibody VHL► ---- IB: o. VHl. was detected throughout the cells with some preference IP: a GAL4 -176 toward localization in the cytoplasmic compartment of the cells (Figure 5A). A very similar distribution of VHL . _ - -83 1B: a Arnt Arnt ► immunoreactivity was observed under hypoxic conditions (Figure 5A). Semiquantitative analysis of the subcellular -32 localization of VHL immunoreactivity (Kallio et al., 1998) VHL► ~- ---- 18: o. VHL under normoxic conditions revealed that 47 % of the ·26 transfected cells had equal distribution of fluorescence in .1 75 lO% input the cytoplasm and the nucleus, whereas in 45% of the -83 1B: a Arnt Arnt► -- -- -- WCE transfected cells cytoplasmic fluorescence predominated over that detected in the nucleus. No transfected cell -1 75 showed exclusive nuclear staining. Hypoxic treatment of GAL41HIF -1 a► • • . ... ,,.. 1B: o.H IF lo: the cells had no effect on the intracellular distribution of -83 1234 56789 VHL since under these conditions 47, 51 and 0% of the transfected cells fell into the three different categories, Fig. 6. VHL, Arnt and HIP-la. form a ternary nuclear complex. respectively. (A) VHL is not released from HIP-la. in hypoxic cells. GAIA, GAIA! HIP-la. and/or VHL were transiently coexpressed in COS? cells in the presence of MG 132 at normoxia or hypoxia for the indicated periods of Protection of HIF-1a from VHL- dependent time. Following immunoprecipitation of whole cell extracts by anti­ degradation is a multi-step pathway requiring GAIA (upper panel) or control (middle panel) antibodies, VHL was hypoxia-induced nuclear translocation of HIF-1a detected by immunoblotting using anti-VHL antibodies. Ten percent of and a hypoxia -dependent regulatory signal input whole-cell extracts is shown in the lower panel. (B) Association with Arnt. GAIA or GAIA/HIF-la. were transiently expressed in COS? We next compared the effect of VHL on the stability of cells in the absence or presence of Arnt or VHL at normoxia or wild-type HIF-la versus the stability of a HIF-la single hypoxia for 6 h, and analyzed as above. amino acid point mutant, HIF-la K719T, which fails to enter the nucleus at hypoxia and is constitutively localized in the cytoplasm (Kallio et al., 1998; Figure 5B). Under cells under normoxic conditions in the presence of normoxic conditions, both wild-type HIF-la and HIF-la the proteasome inhibitor MG 132. After anti-FLAG K719T were degraded in the presence of VHL (Figure 5C). immunoprecipitation of whole-cell extracts, ubiquitylated Interestingly, whereas wild-type HIF-la showed signifi­ N-T AD forms were detected by anti-HA immunoblotting cant (albeit not total) resistance to VHL under hypoxic analysis. Ubiquitylation of wild-type N-T AD was strongly conditions (Figures 1B and 5C), the HIF-la K719T enhanced by coexpression of VHL (Figure 4D). On the mutant was potently degraded upon exposure to VHL in other hand, the mutant N-TAD showed very low levels of hypoxic cells (Figure 5C). In conclusion, these results ubiquitylation even in presence of VHL. Importantly, suggest that hypoxia-induced nuclear translocation of these results clearly demonstrate that VHL mediates HIF-la protects HIF-la from VHL-mediated proteasomal ubiquitylation of HIF-la via physical interaction with degradation. the minimal N-TAD motif. To test whether nuclear localization is sufficient to Given the ubiquitylation of the minimal N-TAD motif, protect HIF-la against VHL-induced proteolysis we we next examined which of the three lysines of N-TAD examined the effect of VHL on the stability of a HIF-la were targeted for degradation at normoxia. FLAG-tagged mutant, HIF-la Li178-390, which shows constitutive wild type or single lysine mutants (K532R, K538R and nuclear localization (Kallio et al ., 1998; Figure 5B). K547R, respectively) of N-TAD were transiently expres­ Strikingly, this protein was degraded at normoxia upon sed in 293 cells. The cells were incubated at normoxia or exposure to VHL and showed hypoxia-dependent protein hypoxia for 12 h, and whole-cell extracts were analyzed by stabilization (Figure 5C). In biochemical experiments, 4304 Reg ulation of HIF-1a. by VHL A FLAG /HIF -11X wt mt GAL4/N-TAD wt mt VHL : - + + Hypox ia : + Hypoxia : + + -32 N-TAO ► HIF-1a ► 2 3 4 VHL ► -- -- -25 12 3 4 5 67 89 30T"" -------=- , -- ----� 10 .---------- --, ■ Hormoxla • Normoxill Hy po xl ■ D Hypoxia 7.5 ,, 5 2. 5 mt wt wt mt GAL4 GAL 41M-TAD GAL4 GAU/Hlf-1 ex Fig. 7. Mutant N-TAD maintains an hypoxia-inducible transactivation response. (A) Constitutive stabilization of PYI mutant N-TAD. GAL4-fused wild-type and mutant N-TAD were transiently expressed in C0S7 cells at normoxia or hypoxia for 12 h. Whole-cell extracts were prepared and assayed as in Figure 1. (B) Hypoxia-inducible transactivation function of mutant N-T AD. The transcriptional activation function by GAL4-fused wild­ type or mutant N-TAD proteins was monitored using a GAL4-responsive luciferase reporter gene as described in Materials and methods. Six hours after transfection, cells were incubated for 36 h at either normoxia or hypoxia. Reporter gene activities are expressed relative to the activity of the GAL4 DNA binding domain alone at normoxia, and normalized against internal control �-galactosidase activit ies. Values represent the mean ± SD of three independent experiments. (C) PYI mutant stabilizes full-length HIF- la. in normoxic cells. FLAG-tagged wild-type and mutant HIF- la were transiently expressed in absence or presence of VHL and analyzed as above. (D) Transcription activation functions of wild-type and mutant HIF- la proteins were assayed as in (B ). cytosolic or nuclear extracts were prepared from COS7 protect HIF-la, from VHL-mediated degradation, we next cells transiently expressing wild-type or mutant HIF-la, in examined the mechanistically important question of the absence or presence of VHL . HIF-la, and VHL co­ whether VHL was released from HIF- la, under hypoxic immunoprecipitation was detected only in nuclear extracts conditions. We transiently expressed GAL4/HIF-1 a, or the of hypoxic cells; whereas VHL was constitutively asso­ minimal GAL4 DNA binding domain in COS7 cells in the ciated with the nuclear HIF-la, A178-390 mutant, and no absence or presence of VHL under normoxic or hypoxic interaction was observed between HIF-la, K719T and conditions for increasing periods of time. In co-immuno­ VHL in nuclear extracts from either hypoxic or normoxic precipitation experiments using whole-cell extracts, VHL cells (data not shown). In summary, nuclear localization was detected by immunoblotting using anti-VHL anti­ per se cannot explain protection of HIF-la. against bodies. As expected, VHL was specifically co-immuno­ degradation by VHL. Our results suggest that, in addition precipitated together with GAL4/HIF-la, under normoxic to nuclear translocation, a distinct hypoxia-dependent conditions. However, VHL was also co-immunoprecipi­ intranuclear event or signal is required for stabilization of tated under hypoxic conditions at levels similar to those HIF-la.. Given the striking overlap within the N-TAD observed at normoxia (Figure 6A). These results suggest between the structures that mediate degradation, physical that dissociation of the HIF-la.-VHL complex is not interaction with VHL, and hypoxia-inducible transactiv­ necessary for protection of HIF-la. from VHL-mediated ation, it is possible that this putative intranuclear degradation, and that there exists a mechanism for stabilizing signal may be linked to the transactivation hypoxia-dependent inactivation of VHL function when function of the protein. remaining associated with HIF-la.. Next, we examined whether Arnt was associated with the HIF-la.-VHL complex. GAL4/HIF-la, or the minimal VHL is not released from HIF-1a in hypoxic cells GAL4 DNA binding domain transiently expressed in Since we found that both nuclear translocation of HIF-la, COS7 cells in the absence or presence of VHL and/or Arnt and a hypoxia-induced activation signal were necessary to 4305 K.Tanimoto et al. HYPOXIA NO RMOXIA jld omain CBD � domain CBD r--7 .----, .------, ,--------, Nuclear Pol y-Ubiqu itylation Transl ocation Arn t Cofactors Ref-1 !f 0 0. Proteas ome­ Target Gene Oo Mediated 0 oo O Acti vati on Degrada tion Fig. 8. Model of conditional regulation of HIF- la function. Under normoxic conditions (left panel) HIF- la is targeted for ubiquitin-proteasomal degradation by VHL. The shaded areas in VHL indicate mutational hotspots in tumors that coincide with the HIF- 1 a or the elongin C binding domain (CBD) of VHL, respectively. Mutations in either of these two domains stabilize HIF- la protein. Hypoxia (right panel) leads to inhibition of degradation of HIF- la by induction of nuclear translocation and an as yet unidentified nuclear regulatory signal that may be linked to recruitment of a partner DNA binding factor, Arnt, transcriptional coactivators and/or the redox regulator Ref -I. See text for details. at normoxia or hypoxia. As expected, Arnt was specifically those generated by wild-type GAL4/N-TAD under co-immunoprecipitated together with GAL4/HIF-la. under hypoxic conditions. In transactivation assays using a hypoxic conditions in the absence of VHL (Figure 6B). GAL4-driven luciferase reporter gene, we observed, as Moreover, in the presence of VHL, Arnt and GAL4/HIF-1 a. expected (Carrero et al ., 2000), rather modest (~3-fold) were also co-immunoprecipitated in a hypoxia-dependent activation of the fu nction of the minimal wild-type GAL4/ fashion (Figure 6B), demonstrating that these proteins N-T AD by hypoxia (Figure 7B). In the context of full­ formed a ternary complex in hypoxic cells. length HIF-la., mutation of the PYI motif rendered the protein stable under normoxic conditions and resistant to VHL-mediated degradation (Figure 7C), demonstrating Role of protein stabilization in regulation of the that this motif is the critical determinant for VHL­ hypoxia-dependent transactivation function of the dependent degradation of HIF-1 a.. Although the mutant HIF-1a N- TAD domain protein showed an increased constitutive transcriptional We and others have observed that transcriptional acti­ activity in comparison with wild-type HIF-1 a., it was still vation by HIF-la. in hypoxic cells is mediated by two hypoxia-inducible (Figure 7D). Within the minimal distinct transactivation domains, N- and C-TAD (Jiang N-TAD domain, mutation of the PYI motif reduced the et al ., 1997; Pugh et al ., 1997; Ema et al ., 1999; Carrero transactivation function with regard to both the activity et al., 2000). Given the fact that VHL interacted with the observed at normoxia and hypoxia (Figure 7B). These minimal N-TAD structure, we were interested to compare results indicate that the mutation may generally have under normoxic and hypoxic conditions both the stability altered the structure important for transactivation, possibly and transactivation functions of GAL4 fusion proteins impairing some of the protein contacts that may be harboring either the wild-type and/or the PYI mutant required for the full activation response. Nevertheless, this N-T AD motifs (Figure 4A). Whereas wild-type GAL4/ mutated construct was inducible by hypoxia, yielding an N-TAD showed hypoxia-dependent protein stabilization ~2-fold increase in transcriptional activity (Figure 7B). (Figures 4E and 7A), PYI mutant GAL4/N-TAD protein These data suggest that the stabilized protein resistant to levels were readily detectable in extracts from normoxic degradation by VHL is still capable of mediating a cells, and not significantly increased following exposure of hypoxia-dependent activation response, and that protein the cells to hypoxia (Figure 7 A). Thus, the failure of this stabilization per se does not bypass the need of the hypoxic mutant to interact with VHL (Figure 4B) correlated with signal for transactivation. constitutively stable protein expression levels similar to 4306 Regulation of HIF-1a by VHL This effect of hypoxia, in turn, correlates with the hypoxia­ Di scussion induced stabilization of HIF-la. protein levels (Kallio We have demonstrated that VHL interacts with HIF-la. to et al., 1999). Thus, HIF-la. is differentially regulated by mediate ubiquitin-proteasomal degradation of HIF-la. ubiquitylation, and inhibition of ubiquitylation constitutes under normoxic conditions. The HIF-la.-binding domain an early and critical step in hypoxia-dependent activation of VHL coincides with one of the hotspots of cancer­ of HIF-1 a, function. It is likely that these effects are causing mutations in the VHL gene (Kaelin and Maher, correlated with protection of HIF-la. against regulation by 1998). In fact, a frequent tumor mutation occurring within VHL. If this is the case, what is the mechanism that this region of VHL abrogated VHL-HIF-la, complex renders VHL-mediated degradation of HIF- la. inactive in formation, suggesting the importance of VHL-HIF-la. hypoxic cells? Obviously, this is a key question to interaction for the tumor suppressor function of VHL. understanding the mechanism of conditional regulation HIF-la, is among the most short-lived proteins currently of HIF-1 a, function. As schematically illustrated in known. The half-life of HIF-la, after exposure of cells to Figure 8, we show here that hypoxia-induced protection hypoxia and subsequent return to normoxia is in the range of HIF-1 a, against regulation by VHL involves two distinct of a few minutes (Wang et al., 1995). The present data and successive steps: nuclear translocation of HIF-1 a, and indicate that VHL functions by recruiting HIF-la, to the an intranuclear event or signal required for protection of VHL-BC--Cul-2 complex. Importantly, the naturally HIF-la, against VHL-induced proteolysis. Given the occurring C162F mutation of VHL fails to bind the striking overlap within the N-TAD of HIF-la. between elongin B and C complex, and has recently been demon­ the structures that mediate oxygen-dependent degradation, strated to abrogate VHL-dependent ubiquitylation activity physical interaction with VHL and hypoxia-inducible in vitro (Lisztwan et al., 1999). As schematically transactivation, it is possible that the putative intranuclear summarized in Figure 8, this mutant also failed to mediate stabilizing signal is linked to the transactivation fu nction degradation of HIF- la., strongly supporting the notion that of the protein. We and others have previously observed VHL regulates HIF-la, fu nction via targeting it for that the hypoxia-inducible fu nction of both the N-TAD ubiquitylation. and C-T AD is critically dependent on the recruitment of Degradation of a protein by the ubiquitin system the transcriptional coactivator CBP/p300 and SRC-1/p 160 involves two critical steps: covalent attachment of mul­ (Arany et al., 1996; Kallio et al., 1998; Ema et al., 1999; tiple ubiquitin molecules to the target protein, and Carrero et al., 2000). This recruitment appears to be degradation of the ubiquitin-tagged substrate by the 26S facilitated by the redox regulator Ref-1 (Ema et al., 1999; proteasome (reviewed by Ciechanover, 1998). Although Carrero et al., 2000). Thus, the functional architecture of the cascade of enzymatic pathways that mediate con juga­ the N-TAD encompasses overlapping structures that have tion of ubiquitin to its substrates has been rather well two different, in fact opposing functions, i.e. protein characterized (for recent reviews see Ciechanover, 1998; degradation versus activation of gene transcription, creat­ Hershko and Ciechanover, 1998), one of the central ing an important 'switch' in regulation of HIF-la. protein questions remains how proteins are selected for degrad­ function. The hypoxia-dependent intranuclear mechanism ation (reviewed by Laney and Hochstrasser, 1999). of protection may involve dimerization with Arnt, recruit­ Obviously, this process must be highly specific since ment of coactivators and/or recruitment of Ref -1 . short-lived proteins need to be identified and differentiated Importantly, the present data indicate that protein from more stable proteins within the cells. stabilization per se does not provide the sole basis for Earlier experiments have indicated that regulation of rendering HIF-la. transcriptionally active. It is possible HIF-la. protein levels by the proteasome pathway is that covalent modification of the C-terminus of HIF-la. mediated by a C-terminal structure of HIF-la. spanning may play a role in determining this regulatory effect. For PEST sequence motifs (Huang et al., 1998; Kallio et al., instance, it is also an attractive scenario that the hypoxic 1999). This structure has been termed the oxygen­ signal may determine a conformational change in HIF-la., dependent degradation domain (Huang et al., 1998), and which facilitates recruitment of the coactivators and notably, this region harbors the N-TAD domain of inactivates VHL function. HIF-la.. What is the degradation motif within the C-terminal region of HIF-la. that mediates regulation by VHL? Transplantable sequence elements, destruction Materials and methods boxes recognized and targeted by a proteolytic apparatus, Plasmid constructs have been identified in many short-lived proteins (Laney pFLAG CMV2/HIF-lal- 826, pCMX-GAL4/HIF-la and deletion and Hochstrasser, 1999) but there is still limited informa­ mutants thereof, and pCMV/Arnt have been described elsewhere tion with regard to the structural characteristics of (Kallio et al., 1997, 1998; Carrero et al., 2000), or were constructed by these elements. Here we demonstrate that VHL directly inserting relevant fragments of hHIF-la into pFLAG-CMV2 (Kodak). Amino acid substitutions in pFLAG or GAL4/HIF-la or NTAD were interacts with the N-TAD of HIF-la., and specifically generated using a QuikCbange site-directed mutagenesis kit (Stratagene ). ubiquitylates this structure. More specifically, point Expression plasmids for green fluorescent protein (GFP) fused to wild­ mutagenesis indicates that the highly conserved central type or mutant forms of HIP-la have been described previously (Kallio PYI motif of the N-TAD was critical for interaction et al., 1998). Wild-type or mutant forms of VHL and VHL GAL4 fusion proteins were constructed by inserting a blunted Nhel-Ecoru. fragment of with VHL and for VHL-induced ubiquitylation and pCI/VHL wild-type or mutant/FLAG (generously provided by Dr Joan proteasomal degradation. W.Conaway, Howard Hughes Medical Institute, USA) into EcoRV­ We have recently observed that exposure of cells to digested pCMX or pCMX-GAL4. The hemagglutinin (HA)-tagged hypoxia or a hypoxia-mimicking agent results in marked ubiquitin expression plasmid, pMT123, was kindly provided by reduction in recovery of ubiquitylated HIF-la, complexes. Dr D.Bohmann (European Molecular Biology Laboratory, Germany). 4307 K.Tanimoto et al. The FLAG-tagged dioxin receptor expression plasmid, pCMV/DR/ Acknowledgements FLAG, was obtained from Dr I.McGuire (Karolinska Institutet, Sweden). We thank Dr Joan W.Conaway for kindly providing wild-type and mutant VHL constructs, and Dr Dirk Bohmann for supplying the HA-tagged Cell culture and transient transfection experiments ubiquitin expression vector. This study was supported by grants from the COS? and 293 cells (obtained from ATCC) were routinely maintained in Swedish Medical Research Council, the JSPS Research Fellowships for Dulbec co 's minimal essential medium supplemented with 10% fetal calf Young Scientists (to K.T.), the Swedish Cancer Society (to Y.M.) and the serum plus penicillin (50 IU/ml) and streptomycin (50 µg/ml). For Fundaciio para a Ciencia ea Technologia (to T.P.). analysis of protein expression, we transiently transfected using the FuGene6 (Boehringer Mannheim) reagent wild-type or mutant forms of FLAG-tagged HIF-la (1 µg) and VHL (2 µg) into COS? cells in 6-cm References diameter plastic dishes. After 12 h of incubation, cells were treated for Arany,Z., Huang,L.E., Eckner,R., Bhattacharya,S., Jiang,C., 12 h at hypoxia (1 % 0 ) or normoxia (21 % 0 ). GAU reporter gene 2 2 Goldberg,M.A., Bunn,H.F. and Livingston,D.M. (1996) An essential assays were performed as described using a �-galactosidase expression role for p300/CBP in the cellular response to hypoxia. Proc. Natl plasmid as an internal control (Carrero et al., 2000). After 6 h of Acad. Sci. USA, 93, 12969-12973. transfection, cells were incubated for 36 h under hypoxic or normoxic Carrero,P., Okamoto,K., Coumailleau,P., O'Brien,S., Tanaka,H. and conditions prior to analysis of reporter gene activity. Poellinger,L. (2000) Redox-regulated recruitment of the transcriptional coactivators CBP and SRC-1 to the hypoxia­ lmmunoblotting and immunoprecipitation assays inducible factor-la. Mol. Cell. Biol., 20, 402-415. For the detection of HIF-la or VHL fusion protein expression whole-cell Ciechanover,A. (1998) The ubiquitin-proteasome pathway: on protein extracts were prepared as previously described (Kallio et al., 1997). death and cell life. EMBO J., 17, 7151-7160. Nuclear and cytosolic extracts were prepared as previously described Ema,M., Hirota,K., Mimura,J., Abe,H., Yodoi,J., Sogawa,K., (Gradin et al., 1996). Fifty micrograms of protein were blotted onto Poellinger,L. and Fu jii-Kuriyama,Y. (1999) Molecular mechanisms nitrocellulose filters following SOS-PAGE. Anti-FLAG M2 (Kodak), of transcription activation by HLF and HIFla in response to hypoxia: anti-VHL (PharMingen) or anti-Arnt (Kallio et al., 1997) antibodies were their stabilization and redox signal-induced interaction with CBP/ used as primary antibodies, diluted l :500 in TBS containing 0.1 % Tween- p300. EMB O J., 18, 1905-1914. 20 (TBS-T) and 1 % non-fat milk for 1 h. After several washes, a 1: 1000 Gnarra,J.R., Zhou,S., Merrill,M.J., Wagner,J.R., Krumm,A., dilution of anti-mouse lg-horseradish peroxidase con jugate (Amersham Papavassiliou,E., Oldfield,E.H., Klausner,R.D. and Linehan,W.M. Life Science) in TBS-T buffer containing 1% non-fat milk was used as a (1996) Post-transcriptional regulation of vascular endothelial growth secondary antibody and incubated with the sample for 1 h at room factor mRNA by the product of the VHL tumor suppressor gene. Proc. temperature. After extensive washing with TBS-T buffer, immuno­ Natl Acad. Sci. USA, 93, 10589-10594. complexes were visualized using enhanced chemiluminescence Gradin,K., McGuire,J., Wenger,R.H., Kvietik:ova,I., Whitelaw,M.L., (Amersham Pharmacia Biotech). Toftgard,R., Tora,L., Gassmann,M. and Poellinger,L. (1996) For in vivo immunoprecipitation experiments, we transiently expressed Functional interference between hypoxia and dioxin signal wild-type or mutant forms of FLAG-tagged or GAU-fused HIF-la, VHL transduction pathways: competition for recruitment of the Arnt and HA-tagged ubiquitin or Arnt (4 µg each) in COS? cells in 10-cm transcription factor. Mol. Cell. Biol., 16, 5221-5231. diameter plastic dishes. After 12 h of transfection, cells were treated for 1, Hershko,A. and Ciechanover,A. (1998) The ubiquitin system. Annu. Rev. 3 or 6 h under hypoxic or normoxic conditions before harvesting the cells. Biochem., 67, 425-479. The cell pellet was resuspended in 200 µl of whole-cell extract buffer Huang,L.E., Gu,J., Schau,M. and Bunn,H.F. (1998) Regulation of (Kallio et al., 1997) supplemented with 20 µM N-ethylmaleimide, -dependent hypoxia-inducible factor la is mediated by an O followed by centrifugation for 30 min at 14 000 r.p.m. For detection of degradation domain via the ubiquitin-proteasome pathway. Proc. co-immunoprecipitated proteins, cells were treated with 5 µM MG-132 Natl Acad. Sci. USA, 95, 7987-7992. (Calbiochem) for 6 h prior to harvest. Subsequent to protein extraction, Iliopoulos,O., Levy,A.P., Jiang,C., Kaelin,W.G.,Jr and Goldberg,M.A. 900 µg of total cell proteins were incubated with anti-FLAG antibodies at (1996) Negative regulation of hypoxia-inducible genes by the von room temperature for 1 h. Thirty microliters of a 50% slurry of protein G­ Rippel-Lindau protein. Proc. Natl Acad. Sci. USA , 93, 10595-10599. Sepharose in TEG buffer (20 mM Tris-HCl pH 7.4, 1 mM EDTA, 10% Iwai,K., Yamanaka,K., Kamura,T., Minato,N., Conaway,R.C., glycerol, 1 mM dithiothreitol), containing 150 mM NaCl and 0.1 % Triton Conaway,J.W., Klausner,R.D. and Pause,A. (1999) Identification of X-100, were then added to the reaction mixtures and incubated for 12 ha t the von Rippel-Lindau tumor-suppressor protein as part of an active 4 °C under rotation. After rapid centrifugation, the resulting Sepharose E3 ubiquitin ligase complex. Proc. Natl Acad. Sci. USA, 96, 12436-- pellets were washed three times with supplemented TEG buffer, and co­ immunoprecipitated proteins were analyzed by SOS-PAGE followed by Iyer,N.V. et al. (1998) Cellular and developmental control of 0 immunoblotting using anti-HA (Santa Cruz Biotechnology), anti-VHL or homeostasis by hypoxia-inducible factor la. Genes Dev., 12, 149- anti-Arnt antibodies diluted 1:500 in TBS-T and 1 % non-fat milk. In in vitro immunoprecipitation assays, GAU and FLAG fusion Jiang,B.H., Zheng,J.Z., Leung,S.W., Roe,R. and Semenza,G.L. (1997) proteins containing full-length or various deletion mutants of HIF-la, Transactivation and inhibitory domains of hypoxia-inducible factor and/or wild-type or mutant VHL proteins were translated either in the la. Modulation of transcriptional activity by oxygen tension. J. Biol. presence or absence of [ S]methionine in rabbit reticulocyte lysate Chem., 272, 19253-19260. (Promega). Labeled translation products were separated by SOS-PAGE Kaelin,W.G.,Jr and Maher,E.R. (1998) The VHL tumour-suppressor and analysed by phosphoimager (Fuji) for calculation of the protein gene paradigm. Trends Genet., 14, 423-426. concentration on the basis of incorporated [ S]methionine. Kallio,P.J., Pongratz,I., Gradin,K., McGuire,J. and Poellinger,L. (1997) Immunoprecipitation experiments were performed as previously Activation of hypoxia-inducible factor la: posttranscriptional described (Gradin et al., 1996) using GAU (Upstate Biotech) or FLAG regulation and conformational change by recruitment of the Arnt antibodies or pre-immune rabbit serum, and precipitated proteins were transcription factor. Proc. Natl Acad. Sci. USA, 94, 5667-5672. analyzed by SOS-PAGE followed by autoradiography. Kallio,P.J., Okamoto,K., O'Brien,S., Carrero,P., Makino,Y., Tanaka,H. and Poellinger,L. (1998) Signal transduction in hypoxic cells: lmmunostaining of VHL and visualization of intracellular inducible nuclear translocation and recruitment of the CBP/p300 coactivator by the hypoxia-inducible factor-la. EMBO J., 17, 6573- trafficking of GFP-tagged proteins in living cells COS? cells grown on cover slips were transiently transfected with VHL expression plasmid and immunostained, and analyzed as described Kallio,P.J., Wilson,W.J., O'Brien,S., Makino,Y. and Poellinger,L. (Kallio et al., 1998). COS? cells were transiently transfected with GFP­ (1999) Regulation of the hypoxia-inducible transcription factor la tagged protein expression plasmids, and after 24 h of expression the cells by the ubiquitin-proteasome pathway. J. Biol. Chem., 274, 6519--6525. were incubated under either normoxic or hypoxic conditions for 6 h. Laney,J.D. and Hochstrasser,M. (1999) Substrate targeting in the Subcellular distribution of fluorescence activity was examined using a ubiquitin system. Cell, 97, 427-430. Zeiss Axiovert 135 microscope with an FITC-filter set, and epifluores­Lee,S., Neumann,M., Stearman,R., Stauber,R., Pause,A., Pavlakis,G.N. cence with illumination from a Gixenon burner (Carl Zeiss Jena GmbH, and Klausner,R.D. (1999) Transcription-dependent nuclear­ Jena, Germany). cytoplasmic trafficking is required for the function of the von 4308 Regulation of HIF-1a by VHL Rippel-Lindau tumor suppressor protein. Mol. Cell. Biol. , 19, 1486- Lisztwan,J., Imbert,G., Wirbelauer,C., Gstaiger,M. and Krek,W. (1999) The von Rippel-Lindau tumor suppressor protein is a component of an E3 ubiquitin-protein ligase activity. Genes Dev. , 13, 1822-1833. Lonergan,K.M., Iliopoulos,O., Ohh,M., Kamura,T., Conaway,R.C., Conaway,J.W. and Kaelin,W.G.,Jr (1998) Regulation of hypoxia­ inducible mRNAs by the von Rippel-Lindau tumor suppressor protein requires binding to complexes containing elongins B/C and Cul2. Mol. Cell. Biol. , 18, 732-741. Maxwell,P.H. et al. (1999) The tumour suppressor protein VHL targets hypoxia-inducible factors for oxygen-dependent proteolysis. Natu re, 399, 271-275. Mukhopadhyay,D., Knebelmann,B., Cohen,H.T., Ananth,S. and Sukhatme,V.P. (1997) The von Rippel-Lindau tumor suppressor gene product interacts with Sp 1 to repress vascular endothelial growth factor promoter activity. Mol. Cell. Biol. , 17, 5629-5639. Pugh,C.W., O'Rourke,J.F., Nagao,M., Gleadle,J.M. and Ratcliffe,P.J. (1997) Activation of hypoxia-inducible factor- I; definition of regulatory domains within the a subunit. J. Biol. Chem. , 272, 11205-11214. Salceda,S. and Caro,J. (1997) Hypoxia-inducible factor la (HIF-la) protein is rapidly degraded by the ubiquitin-proteasome system under normoxic conditions. Its stabilization by hypoxia depends on redox­ induced changes. J. Biol. Chem. , 272, 22642-22647. Srinivas,V., Zhang,L.P., Zhu,X.H. and Caro,J. (1999) Characterization of an oxygen/redox-dependent degradation domain of hypoxia­ inducible factor la (HIF-la) proteins. Biochem. Biophys. Res. Com mun. , 260, 557-561. Stebbins,C.E., Kaelin,W.G.,Jr and Pavletich,N.P. (1999) Structure of the VHL-ElonginC-ElonginB complex: implications for VHL tumor suppressor function. Science, 284, 455-461. Taylor,B.L. and Zhulin,1.B. (1999) PAS domains: internal sensors of oxygen, redox potential, and light. Microbiol. Mol. Biol. Rev. , 63, 479-506. Thrash-Bingham,C.A. and Tartof,K.D. (1999) aHIF: a natural antisense transcript overexpressed in human renal cancer and during hypoxia. J. Natl Cancer Inst. , 91, 143-151. Wang,G.L., Jiang,B.H., Rue,E.A. and Semenza,G.L. (1995) Hypoxia­ inducible factor 1 is a basic-helix-loop-helix-PAS heterodimer regulated by cellular 0 tension. Proc. Natl Acad. Sci. USA , 92, 5510-5514. Wenger,R.H. and Gassmann,M. (1997) Oxygen(es) and the hypoxia­ inducible factor- I. Biol. Chem. , 378, 609---616. Received April 5, 2000; accepted June 30, 2000

Journal

The EMBO Journal – Springer Journals

Published: Aug 15, 2000

Keywords: hypoxia‐inducible factor‐1α; transcription; tumor suppression; ubiquitylation; von Hippel‐Lindau protein

There are no references for this article.