Rac1 activity is required for the activation of hypoxia-inducible factor 1.

Hypoxia-inducible factor 1 (HIF-1) is a transcription factor that mediates cellular and systemic homeostatic responses (including erythropoiesis, angiogenesis, and glycolysis) to reduced O(2) availability in mammals. Hypoxia induces both the protein expression and transcriptional activity of the HIF-1alpha subunit. However, the molecular mechanisms of sensing and signal transduction by which changes in O(2) concentration result in changes in HIF-1 activity are poorly understood. We report here that the small GTPase Rac1 is activated in response to hypoxia and is required for the induction of HIF-1alpha protein expression and transcriptional activity in hypoxic cells.

Hypoxia-inducible factor 1 (HIF-1) is a transcription factor that mediates cellular and systemic homeostatic responses (including erythropoiesis, angiogenesis, and glycolysis) to reduced O 2 availability in mammals. Hypoxia induces both the protein expression and transcriptional activity of the HIF-1␣ subunit. However, the molecular mechanisms of sensing and signal transduction by which changes in O 2 concentration result in changes in HIF-1 activity are poorly understood. We report here that the small GTPase Rac1 is activated in response to hypoxia and is required for the induction of HIF-1␣ protein expression and transcriptional activity in hypoxic cells.
Mammalian cells exhibit many homeostatic responses to hypoxia, including transcriptional activation of genes encoding proteins that function to increase O 2 delivery and that allow metabolic adaptation under hypoxic or ischemic conditions. Although a variety of transcription factors (including AP-1, Egr-1, and nuclear factor B) mediate hypoxia-inducible gene expression in specific contexts, hypoxia-inducible factor 1 (HIF-1) 1 is an essential global regulator of oxygen homeostasis (1). HIF-1 is a basic helix-loop-helix/PAS (PER-ARNT-SIM homology domain) protein consisting of HIF-1␣ and HIF-1␤ subunits (2). The mechanism by which HIF-1 activity is induced under hypoxic conditions remains to be established. HIF-1␣ and HIF-1␤ mRNAs are constitutively expressed in cultured cells, indicating that HIF-1 activity is regulated by post-transcriptional events. HIF-1␣ protein expression and HIF-1 transcriptional activity are precisely regulated by cellular O 2 concentration, whereas HIF-1␤ protein is constitutively expressed (1). The molecular mechanisms of sensing and signal transduction by which changes in O 2 concentration result in changes in HIF-1 activity are complex and involve regulation at multiple levels, including changes in HIF-1␣ protein stability, nuclear localization, and transactivation function in response to hypoxia (1). HIF-1␣ protein expression is negatively regulated in non-hypoxic cells by the ubiquitin-proteasome system (3). Under hypoxic conditions, HIF-1␣ protein levels increase, and the fraction that is ubiquitinated decreases (4). The carboxyl-terminal half of HIF-1␣ contains a domain that negatively regulates protein stability (5,6) and two transactivation domains that are also negatively regulated under non-hypoxic conditions (7,8).
Although much has been learned about the role of HIF-1 in controlling the expression of hypoxia-inducible genes, the underlying mechanisms by which cells sense a decrease in O 2 concentration and transduce this signal to HIF-1␣ are largely unknown. Presently, four diverse O 2 -sensing mechanisms have been proposed to mediate the hypoxic transcriptional response (9). Two of these models postulate involvement of an ironcontaining unit, in the form of either a heme group or an iron/sulfur cluster, that undergoes a change in activity (10). These models are supported by the observation that exposure of cells to cobaltous ion (CoCl 2 ) or the iron chelator desferrioxamine (DFX) stabilizes HIF-1␣ under non-hypoxic conditions (1). However, no specific proteins with this role have been identified in mammalian cells. Two other models involve the generation of reactive oxygen intermediates by a flavoproteincontaining NAD(P)H oxidase or by mitochondria. In the NAD(P)H model, decreased reactive oxygen intermediate production triggers the transcriptional response to hypoxia (11,12), whereas in the mitochondrial model, increased reactive oxygen intermediate production by the electron transport chain (ETC) is an initial trigger of the response (13)(14)(15). In these latter two models, O 2 signals are converted to redox signals.
In addition to changes in cellular redox, hypoxia signal transduction may also require kinase/phosphatase activity because treatment of cells with genistein (a tyrosine kinase inhibitor) or sodium fluoride (a serine/threonine phosphatase inhibitor) blocks hypoxia-induced HIF-1␣ expression (16). In certain cell types, phosphatidylinositol 3-kinase (PI3K) inhibitors such as LY294002 and wortmannin also block hypoxiainduced HIF-1␣ expression (14,17). Reporter assays involving expression of constitutively activated or dominant-negative forms of PI3K or Akt (protein kinase B) demonstrate that the PI3K/Akt pathway modulates hypoxia-induced HIF-1 activation and induces HIF-1 activity in non-hypoxic cells (17)(18)(19). Thus, the signaling pathway from the putative O 2 sensor(s) to HIF-1 may contain several intermediate molecules.
In this study, we have focused on the Rho family small GTPase Rac1 as a potential intermediate in the hypoxia signal transduction pathway. Rac1 plays a pivotal role in multiple cellular processes, including cytoskeletal organization, gene transcription, cell proliferation, and membrane trafficking, through direct or indirect interactions with PI3K, p21-activated kinase (PAK), Ras, and p70 S6 kinase (20 -23). Rac1 also regulates assembly of the active NAD(P)H oxidase complex (24). Rac1 is expressed in most cells and is recognized as a critical determinant of intracellular redox status. We demonstrate here that Rac1 is activated in response to hypoxia and plays an essential role in the induction of HIF-1␣ protein expression and transcriptional activity.
Reporter Gene Assays-All reporter assays were performed in Hep3B cells. Cells were transferred to 24-well plates at a density of 5 ϫ 10 4 cells/well on the day before transfection. Fugene-6 reagent (Roche Molecular Biochemicals) was used for transfection (31). In each transfection, the indicated doses of test plasmids, 200 ng of reporter gene plasmid, and 50 ng of control plasmid pTK-RL (containing a thymidine kinase promoter upstream of Renilla reniformis (sea pansy) luciferase coding sequences; Promega) were premixed with the transfection reagent. In each assay, the total amount of DNA was held constant by addition of empty vector. After treatment, the cells were harvested, and the luciferase activity was determined using the Dual-Luciferase TM reporter assay system (Promega) (17). The ratio of firefly to sea pansy luciferase activity was determined for each reporter experiment; at least two independent transfections were performed in triplicate.
PAK p21-binding Domain (PBD) and Immunoblot Assays-HEK293 cells were transfected with pBOS-HA-Rac1. After 18 h of serum starvation, cells were exposed to 1% O 2 , CoCl 2 , or DFX. Then, cells were lysed in Mg 2ϩ lysis/wash buffer (25 mM HEPES, pH 7.5, 250 mM NaCl, 1% Nonidet P-40, 10 mM MgCl 2 , 1 mM EDTA, and 2% glycerol) supplemented with EDTA-free Complete protease inhibitor mixture (Roche Molecular Biochemicals) in a controlled atmosphere chamber (Plas-Laboratories, Inc.) maintained at 1% O 2 . Lysates (200 g) were incubated with 15 g of GST-PBD (containing amino acids 69 -150 of PAK1), bound to glutathione-agarose beads for 1 h at 4°C, and washed three times with Mg 2ϩ lysis/wash buffer. The bead pellet was finally suspended in 20 l of Laemmli sample buffer (32). Bound proteins were fractionated by SDS-polyacrylamide gel electrophoresis and subjected to immunoblot assay using anti-HA antibody 12CA5 (Roche Molecular Biochemicals). In each experiment, to confirm that equal amounts of HA-tagged Rac1 protein were expressed, immunoblot assay of the starting lysate with the anti-HA antibody was also performed.

FIG. 1. Effect of small GTPases on HIF-1-mediated gene transcription.
A, Hep3B cells were transfected with pTK-RL (50 ng; encoding Renilla luciferase), p2.1 (100 ng; containing an HRE upstream of an SV40 promoter-firefly luciferase reporter gene), and the indicated amounts of expression vectors encoding either no protein (Empty vector) or a dominant-negative (Rac1-N17) or constitutively activated (Rac1-V12) form of Rac1. The total amount of expression vectors was adjusted to 500 ng with empty vector. Cells were exposed to 20 or 1% O 2 for 16 h. The ratio of firefly to Renilla luciferase activity was determined and normalized to the value obtained from non-hypoxic cells transfected with empty vector to obtain the relative luciferase activity. Results shown represent means Ϯ S.D. of three independent transfections. Reporter p2.4 contains a 3-base pair mutation that eliminates binding of HIF-1 to the HRE (30). B, Hep3B cells were transfected with pTK-RL; p2.1; and expression vector encoding no protein, Rac1-N17, Cdc42-N17, Rho-DN, or Ras-N17. Transfected cells were exposed to 1% O 2 , 100 M CoCl 2 , or 100 M DFX. Relative luciferase activities were determined. Results (means Ϯ S.D. of three independent transfections) are expressed as percent of the control (empty vector). gene transcription in a dose-dependent manner (Fig. 1A). Rac1-V12 expression had a small but reproducible stimulatory effect.
In addition to hypoxia, HIF-1 activity is also induced in cells exposed to CoCl 2 or DFX (7). Rac1-N17 expression significantly attenuated reporter gene transcription in response to 100 M CoCl 2 or DFX, although the degree of inhibition was less than the inhibition of the hypoxic response (Fig. 1B). We next tested two other members of the Rho family of small GTPases, Rho and Cdc42. As shown in Fig. 1B, Cdc42-N17 suppressed hypoxia-induced gene transcription, whereas Rho-DN did not. The dominant-negative form of another small GTPase, Ras-N17, also suppressed hypoxia-induced luciferase expression. Transcription of p2.1 in hypoxic HEK293 and NIH3T3 cells was also inhibited by Rac1-N17 (data not shown). Moreover, transcription of a reporter gene containing the HRE from the human VEGF gene in hypoxic Hep3B, HEK293, and NIH3T3 cells was also inhibited by Rac1-N17 (data not shown).
Kinase Inhibitors Inhibit HIF-1 and Rac1 Activation in Response to Hypoxia-To investigate potential components of the hypoxia signal transduction pathway upstream and downstream of Rac1, we first utilized a PI3K inhibitor, wortmannin. As shown in Fig. 5 (A and B, respectively), treatment with 50 nM wortmannin significantly attenuated HIF-1-and HIF-1␣ TAD-dependent transcriptional activity in response to hypoxia. Hypoxia-induced reporter gene transcription was also inhibited by p85⌬, a dominant-negative form of the PI3K p85 regulatory subunit (Fig. 5C). 10 M PD98059, a MEK1 inhibitor, and 25 M SB203580, a p38 MAPK inhibitor, also reduced HIF-1-dependent gene transcription (Fig. 5, A and B).
LY294002 and genistein also inhibited Rac1 activation in response to hypoxia (Fig. 7). DPI and rotenone also markedly inhibited hypoxia-induced Rac1 activation. In contrast, neither PD98059 nor SB203580 attenuated Rac1 activation in response to hypoxia. These results, which are consistent with the analysis of HIF-1␣ expression and reporter gene transcription above (Figs. 5 and 6), demonstrate that mitochondrial ETC, tyrosine kinase, and PI3K activities are required for the activation of both Rac1 and HIF-1 in response to hypoxia.
Rac1-N17 Suppresses Hypoxia-induced AP-1-dependent Gene Transcription-As the results for the ATF2 TAD indicate, HIF-1 is not the only transcription factor that is activated in response to hypoxia. We therefore explored the possibility that In lanes 1 and 2, lysates were loaded with GDP or GTP␥S, respectively, prior to affinity precipitation. B, after exposure to 1% O 2 for the indicated times (lanes 2-8) or to 1% O 2 for 2 h followed by reoxygenation for 15 min (lane 9), cells were harvested. C, after a 30-min treatment with CoCl 2 or DFX, cells were harvested. Cell lysates were incubated with 15 g of GST-PBD. Activated Rac1 was detected by GST-PBD pull-down and immunoblot (IB) assays with anti-HA antibody (␣HA). Equal amounts of HA-tagged Rac1 proteins were detected by immunoblot assay in all lysates prior to pull-down assay.
Rac1 regulates the activation of AP-1 in response to hypoxia. Hep3B cells were cotransfected with reporter pAP-1-Luc, containing seven copies of an AP-1-binding site, and expression vector encoding either the dominant-negative (Rac1-N17) or constitutively activated (Rac1-V12) form of Rac1. Cells were exposed to 20 or 1% O 2 for 8 h and then subjected to luciferase assays. Rac1-N17 significantly suppressed hypoxia-induced reporter gene transcription (Fig. 9). Furthermore, Rac1-V12 strongly stimulated AP-1-dependent gene transcription in both non-hypoxic and hypoxic cells. DISCUSSION The O 2 -dependent regulation of HIF-1 activity occurs at multiple levels in vivo (1). Among these, the mechanisms regulating HIF-1␣ protein expression and transcriptional activity have been most extensively analyzed. An important recent advance has been the identification of the von Hippel-Lindau tumor suppressor protein (pVHL) as the HIF-1␣-binding component of the ubiquitin-protein ligase that targets HIF-1␣ for proteasomal degradation in non-hypoxic cells (34 -38). Hypoxia may induce changes in the phosphorylation and/or redox status of HIF-1␣, pVHL, or another component of the ubiquitination machinery. Remarkably, exposure of cells to hypoxia, CoCl 2 , or DFX induces both HIF-1␣ protein stabilization and transcriptional activation (7,8), even though these agents are mechanistically distinct. For example, inhibitors of mitochondrial ETC complex I block hypoxia-induced (but not CoCl 2 -or DFXinduced) HIF-1␣ protein expression (1,13). For both protein stabilization and transcriptional activation, hypoxia may induce change(s) in the phosphorylation and/or redox status of HIF-1␣ or HIF-1␣-interacting protein(s).
Rac1 has been shown to modulate both phosphorylation and redox status via its binding to protein kinases (20,39,40) and to the NAD(P)H oxidase complex (24), respectively. Our data indicate that Rac1 is required for the induction of HIF-1␣ protein expression, HIF-1␣ TAD function, and HIF-1-dependent gene transcription in response to hypoxia. Although the dramatic inhibitory effects of the dominant-negative form of Rac1 (Rac1-N17) under hypoxic conditions indicate that Rac1 is necessary for these events, the modest stimulatory effects of its constitutively activated form (Rac1-V12) under non-hypoxic conditions indicate that Rac1-independent signals are also required for HIF-1 activation. Below we consider, first, the relationship of Rac1 to other putative components of the hypoxia signal transduction pathway and, second, the mechanisms by which Rac1 may regulate HIF-1␣ expression and activity.
Rac1 and Hypoxia Signal Transduction-Previous studies have demonstrated that inhibitors of mitochondrial ETC (1,13,15), PI3K (13,14,(17)(18)(19), serine/threonine protein phosphatase (14,16), and protein-tyrosine kinase (16) activities block hypoxia-induced HIF-1␣ expression. The inhibitory effects of DPI, rotenone, LY294002, wortmannin, and genistein on the activation of Rac1 (Fig. 7) indicate that Rac1 is downstream of these putative components of the hypoxia signal transduction pathway (Fig. 10). Hypoxia does not induce PI3K activity (17), and an oxygen-regulated phosphatase or kinase that is required for HIF-1␣ expression has not been identified. Hypoxia-induced hydrogen peroxide generation that is dependent upon mitochondrial ETC activity has been reported (13,14), but how this signal is transduced to HIF-1␣ is unknown. The present data suggest that activation of Rac1 may represent an intermediate step in this process. In contrast, the p38 MAPK activity that is induced by hypoxia is downstream of Rac1 (Fig. 10). HIF-1␣ protein expression and HIF-1 DNA-binding activity increase exponentially as cellular O 2 concentration decreases and rapidly decay upon reoxygenation (2,41,42). In contrast, Rac1 has previously been shown to mediate the effects of hypoxia-reoxygenation on the activity of transcription factors such as nuclear factor B and heat shock factor 1 via generation of reactive oxygen intermediates (43,44). In a recent study, hypoxiareoxygenation, but not hypoxia, was shown to induce heat shock factor 1 activation as a result of Rac1-mediated H 2 O 2 generation (45). Thus, the involvement of Rac1 in hypoxiainduced HIF-1 activation represents a novel pathway, and delineation of both the upstream signal for Rac1 activation in response to hypoxia as well as the downstream signal leading to HIF-1 activation will require further studies.
Rac1 and HIF-1␣ Protein Expression-As in the case of ETC activity, Rac1 activity is specifically required for hypoxia-induced (but not CoCl 2 -or DFX-induced) HIF-1␣ expression. These results are consistent with data indicating that CoCl 2 FIG. 8. Effect of Rac1 on p38 MAPK activation. A, constructs encoding the GAL4 DNA-binding domain (amino acids 1-147) fused to the transactivation domain (amino acids 1-96) of ATF2 were analyzed for their ability to transactivate reporter gene G5E1bLuc, containing five GAL4-binding sites. Hep3B cells were cotransfected with pTK-RL (50 ng), G5E1bLuc (100 ng), expression vectors encoding Rac1-V12 or Rac1-N17 (250 ng), and GAL4-ATF2 fusion protein (100 ng). Cells were exposed to 20 or 1% O 2 for 8 h and then harvested. The ratio of firefly to Renilla luciferase activity was determined and normalized to the value obtained from cells transfected with GAL4-ATF2 at 20% O 2 to obtain the relative luciferase activity (means Ϯ S.D. of three independent transfections). B, HEK293 cells were transfected with pSR␣-HA-p38 MAPK and either Rac1-N17 or empty vector (pEF-BOS). Transfected cells were exposed to 20 or 1% O 2 for 6 h or to 100 M H 2 O 2 for 15 min, and cell lysates were subjected to immunoprecipitation using anti-HA antibody matrix. Precipitates were analyzed by Western immunoblotting (IB) using anti-phospho-p38 MAPK (␣Pp38) and anti-HA (␣HA) antibodies.
FIG. 9. Effect of Rac1 on AP-1-dependent reporter gene expression. Hep3B cells were cotransfected with pTK-RL (50 ng); pAP-1-Luc (100 ng); and expression vectors (250 ng) encoding no protein (Empty vector), Rac1-N17, or Rac1-V12. Cells were exposed to 20 or 1% O 2 for 8 h and then harvested. The ratio of firefly to Renilla luciferase activity was determined and normalized to the value obtained from cells transfected with empty vector at 20% O 2 to obtain the relative luciferase activity (means Ϯ S.D. of three independent transfections). and DFX directly disrupt the interaction of HIF-1␣ with pVHL (3,4), i.e. at a step downstream of Rac1.
Rac1 and HIF-1␣ TAD Function-The carboxyl-terminal half of HIF-1␣ consists of two TADs separated by an inhibitory domain that represses TAD function especially under nonhypoxic conditions (7,8). TAD-N function (either in the presence or absence of the inhibitory domain) is induced by hypoxia, an effect that is dependent upon Rac1 activity (Fig. 3). In contrast, TAD-C function is independent of both O 2 concentration and Rac1, again demonstrating that Rac1 is specifically required to transduce hypoxic signals to HIF-1␣. Hypoxia also induced p38 MAPK activity in a Rac1-dependent manner, and the p38 inhibitor SB203580 attenuated hypoxia-induced TAD function (Fig. 5). Rac1 is known to interact with the MAPK kinase kinase PAK1 (20), and p38 MAPK has been shown to phosphorylate the HIF-1␣ inhibitory domain in vitro (46). Taken together, these data suggest that in response to hypoxia, activated Rac1 induces p38 MAPK activity, leading to HIF-1␣ phosphorylation and increased TAD function. Rac1-N17 completely blocked hypoxia-induced transactivation, whereas SB203580 had only a partial inhibitory effect, suggesting that in addition to p38 MAPK activation, there may be other pathways by which Rac1 induces HIF-1␣ TAD function in response to hypoxia. The p42/p44 ERK MAPKs phosphorylate HIF-1␣ and stimulate HIF-1 transcriptional activity (46,47), but this process in not regulated by O 2 concentration.
Broader Role for Rac1 in Hypoxia-induced Gene Transcription-Hypoxia induces the activity of multiple transcription factors in addition to HIF-1. Transcription of an AP-1-dependent reporter gene was induced by hypoxia, and the induction was specifically blocked by Rac1-N17 (Fig. 9), as in the case of the HIF-1-dependent reporter gene (Fig. 1). However, Rac1-V12 markedly induced AP-1-dependent transcription under both hypoxic and non-hypoxic conditions, whereas it had only a minor effect on HIF-1-dependent transcription. These data indicate that Rac1 plays an important role in hypoxia signal transduction in other systems, although the specific mechanisms of transcriptional regulation involved may differ. With these results as a foundation, future studies will be necessary to further delineate the mechanisms and consequences of Rac1 activation in response to hypoxia.