Induction of Hypoxia-inducible Factor-1α by Transcriptional and Translational Mechanisms*

Hypoxia-inducible factor-1 (HIF-1) regulates the transcription of many genes induced by low oxygen conditions. Recent studies have demonstrated that non-hypoxic stimuli can also activate HIF-1 in a cell-specific manner. Here, we define two key mechanisms that are implicated in increasing the active subunit of the HIF-1 complex, HIF-1α, following the stimulation of vascular smooth muscle cells (VSMC) with angiotensin II (Ang II). We show that, in contrast to hypoxia, the induction of HIF-1α by Ang II in VSMC is dependent on active transcription and ongoing translation. We demonstrate that stimulation of VSMC by Ang II strongly increases HIF-1α gene expression. The activation of diacylglycerol-sensitive protein kinase C (PKC) plays a major role in the increase of HIF-1α gene transcription. We also demonstrate that Ang II relies on ongoing translation to maintain elevated HIF-1α protein levels. Ang II increases HIF-1α translation by a reactive oxygen species (ROS)-dependent activation of the phosphatidylinositol 3-kinase pathway, which acts on the 5′-untranslated region of HIF-1α mRNA. These results establish that the non-hypoxic induction of the HIF-1 transcription factor via vasoactive hormones (Ang II and thrombin) is triggered by a dual mechanism,i.e. a PKC-mediated transcriptional action and a ROS-dependent increase in HIF-1α protein expression. Elucidation of these signaling pathways that up-regulate the vascular endothelial growth factor (VEGF) could have a strong impact on different aspects of vascular biology.

Hypoxia activates a number of genes that are important in the cellular and tissular adaptation to low oxygen conditions (1). These genes include erythropoietin, glucose transporters, glycolytic enzymes, and the vascular endothelial growth factor (VEGF). 1 VEGF is a major angiogenic factor that has been shown to activate new blood vessel formation in a number of experimental models (2)(3)(4)(5). Transcriptional up-regulation was shown to be implicated in the induction of this gene, an action mediated by the specific binding of the hypoxia-inducible factor-1 (HIF-1) to the hypoxic response element (HRE) (6,7). The HIF-1 transcription factor is a heterodimer composed of HIF-1␣ and HIF-1␤. Although the HIF-1␤ protein is readily found in all cells, HIF-1␣ is virtually undetectable in normal oxygen conditions. When cells are subjected to hypoxic conditions (1% oxygen), protein expression levels of the HIF-1␣ subunit are rapidly increased. Instead of acting on HIF-1␣ transcription or translation, hypoxia increases HIF-1␣ protein levels by inhibiting the rapid ubiquitination and degradation of the HIF-1␣ protein by the proteasome. An elegant series of studies has shown that under normal oxygen conditions, HIF-1␣ is modified by prolyl hydroxylation, which permits the binding of pVHL, a recognition component of the E3 ligase complex. This promotes the ubiquitin-degradation of HIF-1␣. Under hypoxic conditions, prolyl hydroxylation of HIF-1␣ is blocked, and the HIF-1 transcription factor is stabilized (8 -13).
We were recently the first to show that VEGF expression in vascular smooth muscle cells (VSMC) induced by non-hypoxic stimuli like angiotensin II (Ang II), thrombin, and plateletderived growth factor is mediated through activation of the HIF-1 complex. The HIF-1␣ subunit was highly elevated at the protein expression level following stimulation with the aforementioned agents, allowing activation of the HIF-1 transcription factor (14). Our studies also demonstrated that reactive oxygen species (ROS) play an important role in the activation of HIF-1. However, the cellular signaling mechanisms that are implicated in increasing HIF-1␣ protein levels in VSMC after stimulation with these agents remain unclear.
In this study, we demonstrate that stimulation of VSMC by Ang II strongly increases HIF-1␣ gene expression. The activation of the diacylglycerol-sensitive protein kinase C plays an important part in the increase of HIF-1␣ transcription. We also demonstrate that Ang II relies on ongoing translation to maintain elevated HIF-1␣ protein levels. Ang II increases HIF-1␣ translation by ROS-dependent activation of phosphatidylinositol-3 kinase (PI3K) pathway and through the 5Ј-untranslated region (UTR) of HIF-1␣ mRNA. These results identify two separate pathways that are responsible for the hormonal induction of the HIF-1 transcription complex.
Cell Culture-VSMC were isolated from the thoracic aortas of 6-week-old male Wistar rats by enzymatic dissociation (16). Cells were cultured in Dulbecco's modified Eagle's medium containing 10% fetal calf serum (FCS), penicillin (50 units/ml), and streptomycin (50 g/ml) (Invitrogen) in a humid atmosphere (5% CO 2 , 95% air). Cells were serially passaged upon reaching confluence, and all experiments were performed on passages 3-10. Quiescent cells were obtained by total deprivation of FCS for 16 -20 h. Pretreatment of cells with different compounds was performed 15 min prior to stimulation. Hypoxic conditions were obtained by placing the cells in a sealed "Bug-Box" anaerobic work station (Ruskinn Technologies, Leeds, United Kingdom). The oxygen levels in this work station were maintained at 1-2% with the residual gas mixture containing 93-94% nitrogen and 5% carbon dioxide.
Western Blot Analysis-Confluent cells were lysed in 2ϫ Laemmli sample buffer. Protein concentration was determined by Lowry assay. 25 g of whole cell extracts were resolved in SDS-polyacrylamide gels (8%) and electrophoretically transferred onto a polyvinylidene difluoride membrane (Immobilon-P, Millipore Corp.). Proteins of interest were revealed with specific antibodies as indicated (1:1000 dilution for primary antibodies and 1:5000 for secondary antibodies). The bands were visualized with the ECL system (Amersham Biosciences).
[ 35 S]Methionine Metabolic Cell Labeling and Immunoprecipitation-Confluent cells were rendered quiescent 18 h prior to a 3 h stimulation in methionine-free Dulbecco's modified Eagle's medium. Easy Tag Express 35 S-protein labeling mix (PerkinElmer Life Sciences) was then added at a final concentration of 0.3 mCi/ml, and cells were labeled for 30 min and harvested in lysis buffer (50 mM Tris, pH 8.0, 0.1% SDS, 1% Nonidet P-40, 5 mM deoxycholic acid, 150 mM NaCl, 2 mM dithiothreitol). Cell extracts (1 mg) were precleared with 50 l of protein A-Sepharose followed by the addition of 3 l of anti-HIF-1␣ antiserum 2087. Extracts were left overnight at 4°C prior to the addition of 50 l of protein A-Sepharose for 2 h at 4°C. Sepharose beads were then washed three times with lysis buffer and resuspended in 2ϫ Laemmli sample buffer. Samples were resolved in SDS-polyacrylamide gels (8%) followed by gel drying and exposure to Kodak BioMax MS film.
Northern Blot Analysis-Confluent cells were lysed, and RNA was isolated with TRIzol reagent (Invitrogen). RNA resolved on 1% agarose/6% formaldehyde gels was transferred to Hybond Nϩ nylon membrane (Amersham Biosciences) and hybridized with a radioactive cDNA probe comprising either the first 900-bp coding sequence of the human HIF-1␣ gene or the total coding sequence of the mouse VEGF gene. A probe against 18 S rRNA was used as a loading control. Results were quantified with the use of a Storm blot imaging system (Amersham Biosciences).
Transient Transfection and Luciferase Assays-Reporter (500 ng/ well) was used along with cytomegalovirus Renilla reniformis luciferase expression vector (250 ng/well; Promega) as a control for transfection efficiency. Transfection of VSMC was performed using the Superfect transfection reagent (Qiagen) at a 1:5 DNA/reagent ratio. At 3 h posttransfection, cells were washed, and fresh medium was added. At 12 h post-transfection, cells were deprived of FCS for 16 h. Stimulation with Ang II and hypoxia was performed for 18 h. Cells were then washed with cold phosphate-buffered saline, and luciferase assays were performed with the Dual-Luciferase reporter assay system (Promega). Results were quantified with a Topcount NXT luminescence counter (Packard Bioscience). Results are expressed as a ratio of beetle luciferase activity over R. reniformis luciferase activity. All data expressed an average of at least three independent experiments performed in triplicate. For the HIF-1A reporters, duplicate wells were lysed in TRIzol reagent, and total RNA was extracted. Reverse transcription was then preformed using the Ready-To-Go T-primed first strand kit (Amersham Biosciences) followed by amplification by PCR of the firefly and Renilla sequences (primer sequences available upon request). Products were then electrophoretically migrated on a 1% agarose gel and stained with ethidium bromide.

Implication of Transcriptional Activity in HIF-1␣
Induction by Ang II in VSMC-As mentioned previously, we have shown that Ang II strongly induced an increase in HIF-1␣ protein expression levels in VSMC (14). In the same study, we also showed that HIF-1␣ induction by Ang II proceeded in a manner distinct from that mediated by hypoxia. Interestingly, when time course studies of HIF-1␣ protein induction were performed in VSMC following stimulation with Ang II, we noticed a striking difference between the kinetics of HIF-1␣ protein induction mediated by hypoxia and that mediated by Ang II. As seen in Fig. 1A, the hypoxic induction of HIF-1␣ was rapid, and an increase could be seen at as early as 5 min following hypoxic incubation. Maximal stimulation was attained between 1-2 h of incubation in oxygen-deprived conditions. In contrast, stimulation with Ang II caused a more latent time course. No increase in HIF-1␣ protein expression levels could be seen for times up to 1 h following Ang II addition (Fig. 1A). A small but detectable increase in HIF-1␣ protein expression levels could be observed at 1 h following Ang II addition, with maximal stimulation obtained between 3 and 4 h of incubation in the presence of Ang II. This was not due to slow receptor activation, because p42/p44 MAPK phosphorylation, a control for AT1 receptor activity, was strongly increased at the first time point (5 min). Given the slow induction of HIF-1␣, these results suggest that transcriptional activity is implicated in the maximal increase of HIF-1␣ in VSMC following Ang II stimulation. To confirm this, we pretreated cells for 15 min with the commonly used inhibitor of transcription, actinomycin D, prior to stimulation with Ang II for 4 h. Following this treatment, the activation of HIF-1␣ by Ang II in VSMC was almost completely inhibited (Fig. 1B). Actinomycin D did not block normal receptor signaling, because the ability of Ang II to activate p42/p44 MAPK was not decreased. It is also noteworthy to mention that HIF-1␣ protein levels induced during hypoxia were not decreased by treatment with actinomycin D. Interestingly, actinomycin D treatment in hypoxia induced a marked increase in the molecular weight of the HIF-1␣ protein (Fig. 1B). We believe that this is due to increased HIF-1␣ phosphorylation in these conditions. We have previously shown that p42/p44 MAPK activity can induce a shift in HIF-1␣ molecular weight (15). p42/p44 MAPK activity was also increased when VSMC were treated with actinomycin D (Fig. 1B, lower panel). It is therefore possible that this increased p42/p44 MAPK activity causes the shift in the molecular weight of the HIF-1␣ protein.
We are currently investigating this possibility. Similar results were obtained when the experiment was performed in the presence of another transcriptional inhibitor, 5,6-dichlorobenzimidazole riboside (data not shown). These results demonstrate that, in contrast to hypoxia, transcriptional activity is essential for Ang II-increased HIF-1␣ protein levels.
Regulation of HIF-1␣ mRNA Expression by Ang II-Given the ability of actinomycin D to attenuate Ang II-mediated induction of HIF-1␣ protein expression levels, we wondered whether Ang II stimulation could increase HIF-1␣ mRNA levels. If this was the case, this increase would directly influence the turnover of HIF-1␣ protein. As mentioned previously, traditional HIF-1␣ inducers such as hypoxia or cobalt have very little effect on HIF-1␣ mRNA levels in a number of cell lines (17)(18)(19). This is also the case in VSMC, because a 4-h incubation period in hypoxic conditions did not affect the basal level of HIF-1␣ mRNA in these cells ( Fig. 2A). However, the expression level of HIF-1␣ mRNA was strongly increased in VSMC stimulated with Ang II. This increase was maximal after a 4-h stimulation with 100 nM Ang II (Fig. 2B). As with the increase in HIF-1␣ protein expression levels, the increase in HIF-1␣ mRNA was mediated through the AT1 receptor subtype because the AT1-specific antagonist, losartan, completely blocked the increase in HIF-1␣ mRNA ( Fig. 2A). PD123319, an AT2specific inhibitor, had no effect on the ability of Ang II to up-regulate HIF-1␣ mRNA expression levels (results not shown).
To evaluate the effect of Ang II on HIF-1␣ mRNA transcription, we used a luciferase reporter, pHIF-1A-572/ϩ32Luc, driven by the human HIF-1A gene promoter region (20). When cells were transfected with this reporter and stimulated with Ang II, a strong increase (6.4-fold over basal levels) in reporter activity was observed (Fig. 3). In contrast, hypoxia did not significantly increase HIF-1 promoter activity. These results indicate that Ang II increases the rate of transcription of the HIF-1A gene in VSMC. Studies were also performed to evalu-ate HIF-1␣ mRNA stability following Ang II stimulation. To perform these studies, we evaluated the HIF-1␣ mRNA degradation rate after 4 h of Ang II treatment by blocking transcription with actinomycin D in either the presence or absence of losartan, which inhibits Ang II signaling. However, no differences could be observed between the HIF-1␣ mRNA half-life in these two conditions, indicating that HIF-1␣ mRNA stability is not increased during Ang II stimulation (results not shown).
Protein kinase C (PKC) Is Implicated in the Induction of HIF-1␣ mRNA Expression by Ang II-Activation of PKC is one pathway by which Ang II regulates gene expression. In VSMC, Ang II potently activates a number of PKC isoforms, including the diacylglycerol (DAG)-sensitive forms of PKC. To investigate whether these PKCs are implicated in the regulation of HIF-1␣ mRNA, we treated VSMC with PMA. As seen in Fig. 4A, PMA strongly increased HIF-1␣ protein expression levels in VSMC. This increase was similar to that induced by Ang II and also followed a similar time course (results not shown), again suggesting the implication of transcriptional activation. Indeed, in the presence of 100 nM PMA, HIF-1␣ mRNA levels are also strikingly increased (Fig. 4B). Given the common pathways activated by Ang II and PMA and the similarity between the effects of Ang II and PMA on HIF-1␣, our results suggested that PKC activation was implicated in Ang II-mediated increases in HIF-1␣. To confirm this, we treated cells with GF109203X, a compound shown to inhibit PKC activation by Ang II in VSMC (21,22). Preincubation of VSMC with GF109203X strongly inhibited HIF-1␣ protein levels following stimulation with Ang II and PMA (Fig. 4A). It is important to note that hypoxic induction of HIF-1␣ protein levels was not modified following incubation with GF109203X (Fig. 4A). In addition, GF109203X did not block normal receptor signaling, because p42/p44 MAPK activation by Ang II was not decreased. It has been shown that DAG-sensitive forms of PKC are only minimally responsible for p42/p44 MAPK activation by Ang II (23). However, PKC activation was indeed inhibited by GF109203X, because the activation of p42/p44 MAPK by PMA was completely blocked in the presence of GF109203X. Taken together, these results suggest that PKC activity is implicated in the non-hypoxic up-regulation of HIF-1␣ by Ang II. More importantly, GF109203X inhibited Ang II-and PMA-induced HIF-1␣ and VEGF mRNA levels (Fig. 4B). GF109203X had no effect on HIF-1␣ mRNA levels in unstimulated cells (results not shown). These results indicate that activation of PKC is involved in the induction of HIF-1 by increasing HIF-1␣ mRNA levels.
Implication of Translational Activity in HIF-1␣ Induction by Ang II in VSMC-We have shown that ROS are implicated in the increase in HIF-1␣ protein levels following Ang II stimulation (14). We therefore wanted to evaluate whether ROS activity was implicated in the Ang II-stimulated increase of HIF-1␣ mRNA. Cells were pretreated with inhibitors of Ang II-stimulated ROS production in VSMC, diphenyleneiodonium (DPI) and catalase. We have previously shown that these two agents can potently inhibit HIF-1␣ protein induction by Ang II (14). When cells were pretreated with DPI or catalase prior to stimulation with Ang II, no inhibition of HIF-1␣ mRNA expression could be observed (Fig. 5A). As a control, Fig. 5B demonstrates that these compounds effectively inhibited HIF-1␣ protein expression. These results demonstrate that ROS are not implicated in increasing HIF-1␣ mRNA and suggest that Ang II activates a secondary ROS-dependent mechanism that is responsible for controlling HIF-1␣ protein induction.
Recent studies brought us to consider that ROS activity could be implicated in HIF-1␣ protein translation. These studies have shown the following. 1) Ang II activates the PI3K/AKT pathway in a ROS-dependent manner in VSMC (24); and 2) the PI3K pathway is implicated in heregulin receptor-mediated increases in HIF-1␣ protein translation in breast cancer cells (20). We therefore set out to investigate the possibility that, through the activation of ROS production and hence the PI3K pathway, Ang II could increase HIF-1␣ protein translation in VSMC.
Our previous study (14) demonstrated that the PI3K inhibitor Ly294002 could partially inhibit HIF-1␣ protein levels in VSMC stimulated by Ang II. When this experiment was repeated in new cell lines with different batches of Ly294002, we noted that this inhibition could be considerably stronger (Fig.  6A). Additionally, wortmannin, another PI3K inhibitor, and rapamycin, an inhibitor of a downstream target of the PI3K pathway, mTOR/FRAP, also impaired the induction of HIF-1␣ protein expression levels in response to Ang II. These results point to the implication of the PI3K pathway in HIF-1␣ induction by Ang II in VSMC. As a control, all these compounds effectively inhibited the phosphorylation of a downstream target of this pathway, p70S6 kinase (p70S6K) (Fig. 6A). However, these compounds had no effect on the Ang II-mediated increases in HIF-1␣ mRNA, indicating that PI3K is not implicated in the transcriptional regulation of HIF-1␣ (Fig. 6B). In agreement with previous studies by Ushio-Fukai et al. (24), Fig. 5B shows that ROS inhibitors DPI and catalase also inhibited PI3K pathway activation (phosphorylation of p70S6K) by Ang II in VSMC.
To demonstrate that Ang II increases HIF-1␣ protein synthesis, VSMC were or were not stimulated for 3 h followed by a pulse-labeling period with [ 35 S]methionine/cysteine for 30 min. Very little labeled HIF-1␣ was immunoprecipitated from VSMC under control conditions (Fig. 7). Only minor changes in labeled HIF-1␣ could be seen in hypoxic conditions. These results indicate that, in VSMC, hypoxia relies on protein stabilization rather than increases in translation. This has been seen in other cell models, and it is possible that during the 30 min labeling following 3 h of hypoxia, the rate of new protein turnover was not elevated enough to be detected by this technique (20). However, a strong labeling of HIF-1␣ protein could be seen in cells treated with Ang II. This indicates that, in opposition to hypoxic induction, Ang II utilizes continued novel protein synthesis to maintain elevated levels of HIF-1␣ in VSMC. When cells were treated with rapamycin 10 min before the pulse-labeling period, HIF-1␣ labeling was blocked (Fig. 7). This implicates mTOR and the PI3K pathway in increasing HIF-1␣ protein synthesis.
Laughner et al. (20) have shown that the 5Ј-UTR of HIF-1␣ mRNA is involved in the heregulin-mediated induction of HIF-1␣ protein synthesis. We therefore wanted to investigate whether the 5Ј-UTR of HIF-1␣ mRNA could be implicated in increasing HIF-1␣ translation in VSMC following Ang II stim- A, total RNA was extracted from cells and resolved on formaldehyde/agarose gels. Northern blotting was performed using a specific radiolabeled HIF-1␣ or VEGF probe. An 18 S rRNA probe was used as a control for gel loading. B, total cell extracts (25 g) were resolved by SDS-PAGE (8% gel) and immunoblotted using an anti-HIF-1␣ antiserum, an anti-phospho-p70S6 kinase polyclonal antibody, or an anti-phospho-p44/p42 MAPK monoclonal antibody.
ulation. For this experiment, we used a luciferase reporter construct under the control of the HIF-1A gene promoter that included the 5Ј-UTR pHIF-1A-572/ϩ284Luc (20). As was the case for the activation of the HIF-1␣ promoter construct (Fig.  3), this reporter was not activated by hypoxia (Fig. 8A). However, when cells were stimulated with Ang II, a strong activation of pHIF-1A-572/ϩ284Luc could be observed. This increase was superior (by nearly 4-fold) to assays performed with the HIF-1␣ promoter only (compare Figs. 3 and 8A). To ensure that the 5Ј-UTR was not increasing transcription of the reporter, we performed a reverse transcription PCR analysis of firefly and Renilla luciferase mRNA. As seen in Fig. 8B, when cells transfected with the pHIF-1A-572/ϩ284Luc or pHIF-1A-572/ ϩ32Luc were stimulated with Ang II, expression of firefly luciferase mRNA for both constructs was significantly increased to similar levels as compared with the control Renilla luciferase mRNA. These results demonstrate that the HIF-1␣ 5Ј-UTR does not affect the transcription of the luciferase reporter. Taken together, these results indicate that the 5Ј-UTR of HIF-1␣ mRNA controls the increase of HIF-1␣ translation following Ang II stimulation.
To investigate the implication of the ROS-activated PI3K pathway on this increase in HIF-1␣ translation, we evaluated the effect of a p22 phox antisense plasmid and a dominant negative form of PI3K, p85⌬PI3K. p22 phox is a critical component of the VSMC NADH/NADPH oxidase and is essential for Ang II-mediated ROS production in VSMC (24). p85⌬PI3K is a mutant form of the PI3K subunit p85, which cannot interact with the P110 subunit and acts as a dominant negative (25). When VSMC were transfected with either p85⌬PI3K or antisense p22 phox , a significant decrease was seen in the activation of the pHIF-1A-572/ϩ284Luc reporter (Fig. 9, Ϫ572/ϩ284). The constructs also decreased the activation of the HIF-1-targeted reporter construct PRE-tk-LUC (Fig. 9, HRE). However, p85⌬PI3K and antisense p22 phox had no significant effect on the HIF-1A gene promoter construct pHIF-1A-572/ϩ32Luc, which contains no 5Ј-UTR (Fig. 9, Ϫ572/ϩ32). These results indicate that, in addition to increases in HIF-1␣ transcription, HIF-1 activation by Ang II is mediated through the ROSactivated PI3K pathway, which increases HIF-1␣ translation. DISCUSSION Ubiquitously, hypoxia strongly increases HIF-1␣ protein expression levels by blocking the ubiquitin/proteasome-mediated degradation of HIF-1␣ through the inhibition of prolyl hydroxylation (8 -13). In VSMC, we have previously shown that Ang II can potently increase and activate the HIF-1 transcription complex (14) to levels that can even surpass those achieved by hypoxic activation. In this study, we set out to identify the mechanisms that control this increase. We have found that, in FIG. 6. Implication of the PI3K pathway in HIF-1␣ induction. Quiescent VSMC were pretreated or not for 15 min with either wortmannin (100 nM), Ly294002 (20 M), or rapamycin (10 nM) prior to stimulation with Ang II (100 nM) for 4 h. A, total cell extracts (25 g) were resolved by SDS-PAGE (8% gel) and immunoblotted using an anti-HIF-1␣ antiserum, an anti-phospho-p70S6 kinase polyclonal antibody, or an anti-phospho-p44/p42 MAPK monoclonal antibody. B, total RNA was extracted from cells and resolved on formaldehyde/agarose gels. Northern blotting was performed using a specific radiolabeled HIF-1␣ or VEGF probe. An 18 S rRNA probe was used as a control for gel loading. contrast to hypoxia, Ang II modulates two different pathways to increase HIF-1␣ protein expression levels. The first involves an increase in the rate of HIF-1␣ mRNA transcription through a mechanism requiring the activation of the DAG-sensitive PKC. The second involves increasing the rate of translation of this newly produced HIF-1␣ mRNA by activating ROS production and the subsequent activation of the PI3K pathway. Together, these two pathways boost VSMC HIF-1␣ protein expression to levels that surpass hypoxic induction and lead to the activation of HIF-1-responsive genes such as VEGF.
In VSMC, Ang II activates a large number of signaling pathways that are responsible for the modulation of vasomotor tone, cell growth, migration, and apoptosis (26). In this study, our task was to identify the signaling pathways that were responsible for increasing the expression levels of the HIF-1 complex. One signaling mechanism of interest was the protein kinase C pathway. In these cells, Ang II is a major activator of various forms of PKC that have been shown to activate a number of pathways and genes. PMA is an agonist that directly and potently activates DAG-sensitive isoforms of PKC. Our results show that PMA is one of the strongest activators of HIF-1␣ activity in VSMC. 2 We noticed that the time course of HIF-1␣ induction and activation by PMA closely resembles that of Ang II. We show that the common mediator involved in the increase of HIF-1␣ mRNA expression by Ang II and PMA is indeed PKC. The use of specific inhibitors of DAG-sensitive PKCs also potently blocks the induction of HIF-1␣ protein levels. Previous studies have also shown that PKC activity is required for Ang II-induced expression of nox1, a main component of the VSMC ROS-producing NADPH oxidase (22). The expression of nox1 or other elements during Ang II stimulation would also be blocked by actinomycin D, thereby explaining the strong potency of this drug on HIF-1␣ protein expression. Taken together, these results define a central role for PKC in Ang II-induced HIF-1 activation. PKC may control HIF at two levels: 1) by increasing HIF-1␣ mRNA levels; and 2) by increasing cellular ROS levels, which are essential in increasing the rate of HIF-1␣ protein translation.
The HIF-1␣ gene promoter has been cloned from the human and mouse genome and shows an 80% conservation of the nucleotide sequence (17,27). Although the promoter of rat HIF-1␣ gene has not yet been sequenced, we suspect that the sequence is similar to those of the mouse and human genes given the strong homology of the HIF-1␣ gene product. At this time, we have not identified the sequences that are responsible for the activation of the HIF-1␣ gene promoter by Ang II. Interestingly, a number of Sp1 sites exist in the HIF-1 promoter that are conserved between the mouse and the human sequence (27). Studies have shown that PKC can increase the rate of gene transcription via the Sp1 transcription factor (28,29). We are currently undertaking studies to investigate this possibility.
Free radicals have been shown to play a crucial role in the activity of a number of signaling pathways in VSMC (24, 30 -33). We and others have shown that ROS production is important for the induction of HIF-1␣ protein levels by vasoactive hormones (14,34). However, ROS inhibitors do not affect the ability of Ang II to up-regulate the rate of HIF-1␣ gene transcription, indicating that a second pathway was involved in the induction of HIF-1␣ by Ang II. Ushio-Fukai et al. (24) has clearly shown that Ang II activates the PI3K pathway in VSMC. Recently, a number of studies have shown that this pathway plays an important role in the induction of HIF-1␣ by non-hypoxic factors such as heregulin and insulin (20,35). These factors have no effect on HIF-1␣ mRNA transcription but instead increase HIF-1␣ protein translation. We therefore wanted to investigate whether ROS-dependent activation of the PI3K pathway by Ang II in VSMC could increase HIF-1␣ translation of the newly increased HIF-1␣ mRNA. We demonstrate that this is indeed the case and that the 5Ј-UTR is critically involved in this increase. A possible mechanism by which Ang II could increase the rate of HIF-1␣ translation involves the activation of eukaryotic translation initiation factor 4F (eIF-4F) and/or the ribosomal S6 protein by the PI3K/ p70S6K/mTOR pathway. p70S6K regulates the translation of a group of mRNAs possessing a 5Ј-terminal oligopyrimidine tract (5ЈTOP), a stretch of 4 -14 pyrimidines found at the extreme 5Ј-terminus of certain mRNAs (36). The HIF-1␣ 5Ј-UTR contains these tracts, including a long conserved sequence in the extreme 5Ј terminus (27). Phosphorylation of the S6 protein of the 40 S ribosomal unit by p70S6K increases the translation of the 5Ј-TOP mRNAs. In VSMC, the p70S6K pathway has been shown to be implicated in the increases in protein translation induced by Ang II (37). The eIF-4F complex is formed by interactions between eIF-4E, eIF-4G, and other factors. This complex binds to the 5Ј-mRNA cap structure and permits the recruitment of the eukaryotic ribosome (38). eIF-4E activity is blocked by interaction with 4E-binding proteins (4E-BPs). This interaction is regulated by phosphorylation. Hypo-phosphorylated 4E-BPs strongly bind eIF-4E, whereas hyper-phosphorylated 4E-BPs release eIF-4E. The phosphorylation of 4E-BP is regulated by PI3K/mTOR. In VSMC, Ang II increases 4E-BP1 phosphorylation (39) through the PI3K pathway. 3 eIF-4E is also phosphorylated in VSMC following Ang II stimulation, leading to increased protein synthesis (40 9. PI3K and reactive oxygen species are involved in increasing HIF-1␣ translation and activity. VSMC (5 ϫ 10 5 cells/well, six-well plate) were transfected with 500 ng of either the PRE-tk-LUC (HRE), pHIF-1A-572/ϩ32Luc (Ϫ572/ϩ32) or pHIF-1A-572/ϩ284Luc (Ϫ572/ϩ284) reporter plasmid and 250 ng of an expression vector coding for R. reniformis luciferase. Also, cells were co-transfected with an empty pcDNA3.1 expression vector (white bars), a p85⌬PI3K expression vector (black bars), or a p22 antisense expression vector (mixed bars). 12 h after transfection, cells were deprived of FCS for 16 h. Cells were then maintained under control conditions or in the presence of Ang II (100 nM) for 18 h. At this point, VSMC were lysed, and luciferase activity was measured using the Dual-Luciferase reporter assay system. Results are expressed as a ratio of beetle luciferase activity over R. reniformis luciferase activity. Data expressed are an average of at least three independent experiments performed in triplicate.
shown to be implicated in this phosphorylation, suggesting again that PKC plays a central role in the induction of HIF-1␣ by Ang II. A second study has also shown that ROS are implicated in eIF-4E phosphorylation, further supporting our findings (33).
During our previous study, two observations indicated that, in VSMC, hypoxia and Ang II acted through different pathways. First, the hypoxic and Ang II effects were additive, and, second, DPI completely blocked Ang II-mediated HIF-1␣ protein increases while having no effect on the HIF-1␣ induction by hypoxia. In this study we clearly demonstrate that the actions of Ang II in VSMC are distinct from those of hypoxia. It is essential to understand that, in VSMC, hypoxia had no effect on the rate of HIF-1␣ transcription or translation. We believe that, in this way, Ang II can increase the low basal expression levels of HIF-1␣ mRNA in quiescent VSMC and effectively translate the messenger, leading to high HIF-1␣ protein expression levels during Ang II stimulation, whereas hypoxia relies on increased protein stability to maintain elevated HIF-1␣ protein expression levels. 4 Given the low expression levels of HIF-1␣ mRNA in quiescent VSMC, this would explain the mediocre induction of HIF-1␣ in response to hypoxia. However, when cells are stimulated with both hypoxia and hormones or growth factors, the induction of HIF-1␣ is additive because of the different pathways activated (see Ref. 14).
In conclusion, our study provides data that identifies two interesting mechanisms by which Ang II (and other hormones like thrombin) activates the HIF-1 complex. By binding to the AT1 receptor, Ang II activates at least two distinct pathways involved in the up-regulation of HIF-1. The first, by increasing HIF-1␣ mRNA expression levels, leads to increased HIF-1␣ protein expression activity. The second, which involves ROS production, results in an increased rate of translation of the HIF-1␣ protein. An active HIF-1 complex is thus formed that increases the expression of its target genes, including VEGF. Given the importance of these genes in cellular physiology, we strongly believe that these studies will have a strong impact on vascular biology.