BMK1/ERK5 Is a Novel Regulator of Angiogenesis by Destabilizing Hypoxia Inducible Factor 1
http://www.100md.com
循环研究杂志 2005年第6期
The Cardiovascular Research Institute and Departments of Medicine (X.P., G.G., Q.Z., H.W., J.A., C.Y., B.C.B.) and Pharmacology (L.X.), University of Rochester, NY
Abstract
Big MAP kinase 1 (BMK1 or ERK5) is a key mediator of endothelial cell (EC) function as shown by impaired embryonic angiogenesis and vascular collapse in BMK1 knockout mice. Hypoxia inducible factor 1 (HIF1), a potent mediator of angiogenesis, is positively regulated by the MAP kinases, ERK1/2. Because BMK1 deficiency is associated with impaired angiogenesis we hypothesized that BMK1 might regulate HIF1. To test this hypothesis, bovine lung microvascular ECs (BLMECs) were transfected with HIF1 and BMK1 cDNAs, and stimulated by hypoxia. HIF1 activity was measured by a reporter gene assay in which luciferase expression was driven by HIF1 activation. Hypoxia (1% O2, 24 hours) stimulated HIF1 activity by 5.1±0.6 fold. In the presence of dominant negative (DN)-BMK1, which inhibited BMK1 activity, hypoxia induced HIF1 activity was enhanced significantly to 6.4±0.4 fold. BMK1 activation by constitutively active (CA)-MEK5 inhibited HIF1 activity by 46±4%, suggesting BMK1 functions as a negative regulator of HIF1 activation. Activation of BMK1 reduced HIF1 protein levels. Ubiquitination inhibitors (30 eol/L ALLN, 2 eol/L lactacystin, or 100 nmol/L MG132) reduced the BMK1-mediated effect on HIF1 expression by >80%, suggesting that BMK1 stimulated HIF1 proteolysis. The negative effect of BMK1 on HIF1 was functionally important because transfection with CA-MEK5 significantly decreased EC migration by 68±10%, and inhibited angiogenesis (in vitro Matrigel assay) by 76±7%. In summary, BMK1 is a novel negative regulator of HIF1 and angiogenesis by increasing HIF1 ubiquitination and inhibiting HIF1 activity in endothelial cells.
Key Words: big MAP kinase 1 hypoxia inducible factor 1 angiogenesis
Introduction
Endothelial cells (EC) play a key role in all aspects of angiogenesis.1,2 Understanding the specific mechanisms that contribute to each step of the angiogenic response may offer the best approach to developing therapies that will augment angiogenesis in diseases such as stroke and myocardial infarction or inhibit angiogenesis in cancer. An important master regulator of angiogenesis (especially induced by hypoxia) is hypoxia inducible factor-1 (HIF1), which controls critical proangiogenic genes such as vascular endothelial growth factor (VEGF).3eC6 Targeted inactivation of HIF1 in mice results in abnormal vascular development and embryonic lethality, demonstrating its key role in the angiogenic processes.7 Multiple mechanisms regulate HIF1 including both transcriptional and posttranslational processes. In EC, posttranslational mechanisms include both protein degradation (mediated by proline hydroxylases and ubiquitination) and protein phosphorylation.8eC11
Previous studies suggest that HIF1 function is regulated by mitogen activated protein (MAP) kinases.6,12,13 Specifically it was shown that ERK1/2 was activated by hypoxia, phosphorylated HIF1 directly, and enhanced HIF1 transcriptional activity as measured by VEGF expression.6,12 The mechanisms by which ERK1/2 stimulates HIF1 transcriptional activity remain unknown because HIF1 phosphorylation by ERK1/2 did not change HIF1 protein stability. However, the ERK1 knockout mouse did not exhibit any defects in angiogenesis.14 In contrast, several recent articles15eC17 suggest a critical role for the closely related MAP kinase, BMK1, in angiogenesis. Sohn et al showed in the BMK1 knockout mouse that initial vascularization occurred normally but subsequent remodeling and maintenance of the vasculature was adversely affected.15 A similar result was reported by Hayashi et al who showed in a conditional BMK1 knockout that vascular integrity was compromised and EC apoptosis increased.16 We hypothesized that the abnormalities in BMK1 knockout vasculature may reflect BMK1-mediated regulation of HIF-1. This hypothesis derived from an article showing that VEGF expression was upregulated in BMK1 knockout embryonic fibroblasts and overexpression of BMK1 cDNA inhibited hypoxia induced VEGF promoter activity.15 The authors suggested that the phenotype of BMK1 null mice was possibly because of disregulation of HIF activity, which disrupted normal vasculogenesis and angiogenesis by overexpression of VEGF. Also, because BMK1 and ERK1/2 have the same substrate recognition motif for phosphorylation, we propose that BMK1 regulates HIF activity in endothelial cells.
In this study, we demonstrate an essential role for BMK1 as a negative feedback regulator of HIF function as measured by VEGF expression, EC migration, and angiogenesis. Furthermore, we show that BMK1 regulation of HIF1 occurs via an ubiquitin dependent degradation mechanism.
Materials and Methods
Reagents
Bovine lung microvascular ECs (BLMEC)s were purchased from VEC Technologies (Renseller, NY). Endothelial basal medium and supplements were purchased from Invitrogen. Mouse anti-GFP antibody and mouse anti-PCNA were from BD Biosciences (San Jose, Calif). Mouse anti-VEGFR2, goat anti-VEGFR1, rabbit anti-pERK1/2, rabbit anti-ERK1/2, rabbit anti-actin, and rabbit anti-HA antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, Calif). Mouse anti-4G10 monoclonal antibody was purchased from Upstate (San Diego, Calif). Rabbit anti-BMK1 and monoclonal anti-FLAG antibodies were bought from Sigma Company. Monoclonal HIF1 antibody was purchased from Novus Biological (Eugene, Ore). A goat anti-mouse, Alexa Fluor 546 (Molecular Probes, Littleton, Co) was used for immunofluorescence. ALLN, Lactacystin, MG132, ALLM were from Calbiochem (San Diego, Calif). Matrigel matrix was purchased from BD Biosciences.
Cell Culture and Transfection
BLMECs at passages 3 to 8 were grown in MCDB-131 media supplemented with 10% FBS, heparin (15300 U/L, Sigma; St. Louis, Mo), hydrocortisone (2.76 eol/L, Sigma), bovine pituitary extract, epidermal growth factor (1.64 nmol/L, Sigma), L-glutamine, and antibiotics (100 U/mL penicillin, 68.6 mol/L streptomycin) in flasks precoated with 2% gelatin. For transient expression experiments, 95% confluent cells in 60 mm dishes were transfected 24 hours after plating with 2 e plasmids using 7 e蘈 Lipofectamine and 10 e蘈 Plus reagent (Invitrogen, Carlsbad, Calif). After 4 hours of incubation, the medium was replaced with complete MCDB-131. One day later, cells were serum starved overnight, then incubated in the hypoxia chamber (1% O2). This protocol yields 60% transfection efficiency based on GFP.17
HIF1-Dependent Reporter Gene Assay
BLMECs were cotransfected with hypoxia response element-firefly-luciferase, TK-renilla-luciferase and CA-MEK5 or other cDNAs indicated in each experiment. After treatment, the luciferase activities in cell lysates were determined using the Dual-Luciferase Reporter Assay system (Promega, Madison, Wis) and Wallac 1420 multilabel counter. Luciferase activity represents the ratio of the luminescence units (RLU) from firefly luciferase and renilla luciferase.8
Western Blot Analysis
The membrane was blocked for 2 hours at room temperature with 5% BSA. The blot was incubated for 2 hours with certain dilution of antibody at room temperature, followed by incubation for 1 hour with horseradish peroxidase-conjugated secondary antibody. Immunoreactive bands were visualized using ECL (Amersham Pharmacia Biotech). Alternatively, the infrared secondary antibody was used and images were analyzed with Odyssey software (LiCor, Inc).
Cell Fractionation
After the different transfection and treatment, BLMEC were washed twice with cold PBS and lysed in hypotonic buffer (10 mmol/L Tris-HCl pH7.5, 10 mmol/L NaCl, 3 mmol/L MgCl2, 1 mmol/L PMSF, 0.5 mmol/L DTT, 0.1 mmol/L EGTA, 2 eol/L leupeptin, 1 e of aprotinin per mL, and 0.05% NP-40).18 The supernatant containing the cytosolic extract was removed and the protein concentration determined by Bradford assay. The nuclei were washed twice with the hypotonic buffer without NP-40 and scraped in Laemmli buffer. The same volume of nuclei extracts was loaded for each sample. PCNA and actin were used as controls of loading for nuclear and cytosolic fractions respectively.
Immunofluorescence
BLMEC were fixed in 3% paraformaldehyde for 15 minutes at room temperature. After 3 washes with PBS, the cells were sequentially treated with 0.1% Triton X-100 for 5 minutes (for permeabilization), with PBS/3% BSA for 1 hour (for blocking), then with an anti-FLAG for 2 hours in PBS/1%BSA. After 3 washes, cells were incubated in dark with antimouse an Alexa Fluor 546 in PBS for 45 minutes at room temperature. After 4 washes in PBS, the cells were counterstained with DAPI and then the fluorescent signal was visualized with an Olympus fluorescence microscope.
Relative Semiquantitative PCR
PCR was performed as previously described.19 RNA of BLMECs was extracted with a Total RNA Isolation Kit (Qiagen). First-strand cDNA was synthesized with the SuperScript Preamplification System (Gibco-BRL). Relative quantitative reverse transcriptioneCpolymerase chain reaction was performed with 18s rRNA as an internal control using Ambion’s competimer technology. VEGF primers (sense: CTTTCTGCTCTCTTGGGTG C; anti-sense: ATCTTCAAGCCATCCTGTGTCCC) were used to generate the 224-bp PCR product. PCR products were separated on a 2% agarose gel, stained with ethidium bromide, and visualized under UV light. The relative intensity of VEGF PCR products was determined by densitometry.
Migration Assay
For detection of cell migration, a wound healing assay was performed.20 BLMECs were grown on 6 cm wells, the monolayer was scratched with a sterile disposable rubber policeman and the edge labeled with a traced line. After injury, the cells were gently washed with normal medium without serum. EC migration from the edge of the injured monolayer was quantified by measuring the area between the wound edges before and the recovered area after injury using light microscopy and the computer program ImageJ.
Angiogenesis Assay
The angiogenic response to HIF1 induction affected by BMK1 activation was assessed using an in vitro Matrigel analysis, using BLMECs transfected with DM-HIF1 and CA-MEK5, or pcDNA3.1 as control. Two days after transfection, the cells were seeded on 6-well cell culture plates precoated with Matrigel at 1x105 cells/well and maintained in serum free medium. The capillary/tube-like structures were visualized by light microscopy at different time points and analyzed using ImageJ.
Statistical Analysis
Data were shown as mean±STDEV (S.E., for data analysis of angiogenesis assay in Figure 5C and 5D) for 3 to 4 separate experiments. Differences were analyzed with one-way analysis of variance (ANOVA). Values of P<0.05 were considered statistically significant.
Results
HIF1 Transcriptional Activity Is Inhibited by BMK1 Activation
To assess the effect of BMK1 on HIF1 function we used CA-MEK5 to stimulate BMK1 activity and measured HIF1-dependent reporter gene expression.6,8,17 CA-MEK5 overexpression in BLMEC increased BMK1 activity by 10-fold.17 Endogenous HIF1 activity was inhibited by 39.0±4.2% after CA-MEK5 overexpression (Figure 1A, second bar). To further assess the effects of BMK1 on HIF1 we also transfected BLMEC with cDNAs of wild-type (WT) or double mutant (DM)- HIF1 (P402A / P564G). DM-HIF1 is stabilized compared with WT-HIF1 protein because mutations of the proline hydroxylase recognition sites Pro 402 and Pro 564 prevent ubiquitination of HIF1.11 Transfection with WT-HIF1 increased luciferase activity by 4.0±0.1-fold compared with endogenous HIF1. CA-MEK5 inhibited HIF1 activity by 68.9±5.8% (Figure 1A, bars 3 and 4). In cells transfected with DM-HIF1, luciferase activity was increased 9.8±0.2 fold. CA-MEK5 dramatically inhibited DM-HIF1 function by 77.9±6.6% (Figure 1A, bars 5 and 6), suggesting that the effect is independent of Pro 402 and Pro 564.
To demonstrate further that BMK1 activation inhibited HIF1 transcriptional activity, we used hypoxia (1% O2) to stimulate HIF1 activity. We first examined the effect of hypoxia on BMK1 activity, and compared the time course and magnitude with activation of ERK1/2.6,12 In response to hypoxia, BMK1 activity increased with peak at 24 hours measured by phosphorylation-dependent bandshift (Figure 1B, 97.5±8.0-fold increase, n=3). We have previously shown an excellent correlation between BMK1 activity measured by bandshift and by in vitro kinase assay.17,21 ERK1/2 activity also increased in response to hypoxia but more rapidly (peak at 6 hours) than BMK1 (Figure 1B). There was no significant change in BMK1 or ERK1/2 protein expression by hypoxia (Figure 1B). HIF1 transcriptional activity was increased 5.0±0.6-fold by hypoxia (for 24 hours) and even further by overexpression of DN-BMK1 (Figure 1C). CA-MEK5 overexpression inhibited HIF1 activity by 44.0±3.7% and this inhibitory effect was prevented by coexpression of DN-BMK1 (Figure 1C). Because ERK1/2 has also been reported to regulate HIF1 function we studied the effect of modulating MEK1 and ERK1/2 activity. Interestingly, HIF1-dependent luciferase activity was significantly increased by CA-MEK1 and inhibited by DN-ERK2 (Figure 1C). These results show that BMK1 and ERK1/2 differentially regulate HIF1 activity. The negative effect of BMK1 on HIF1 activity might serve as a negative feedback regulator of HIF1 to counter-balance the stimulatory effect of ERK1/2.
HIF1 Protein Level Is Decreased by BMK1, but Not by ERK1/2 Activation
The most important mechanism for regulation of HIF1 is protein degradation.8,9,11 To determine whether BMK1 activation decreased HIF1 protein, the protein level of HIF1 was determined by western blotting. First, BLMEC were transfected with either WT-HIF1 or DM-HIF1 in the presence of CA-MEK5 or CA-MEK1 to activate BMK1 or ERK1/2, respectively. Expression of DM-HIF1 is higher than WT-HIF1 because DM-HIF1 has greater protein stability than WT-HIF1 (Figure 2A). HIF1 protein levels were not changed by overexpression of CA-MEK1, which is consistent with previous findings that ERK1/2 activation increased HIF1 transcriptional activity possibly through direct phosphorylation of HIF1 by ERK1/2, without effect on protein degradation.6,12 Interestingly, the level of ectopically expressed HIF1 protein decreased with overexpression of CA-MEK5 (Figure 2A). In addition, the ability of BMK1 to decrease expression of DM-HIF1 lacking the prolines required for ubiquitination and degradation suggested that BMK1 initiated protein degradation via a different mechanism that was independent of Pro 402 and Pro 564.
To confirm the decrease in HIF1 protein by BMK1 activation, endogenous HIF1 was measured after exposure to hypoxia. BLMECs were transfected with CA-MEK5 or DN-BMK1 and exposed to hypoxia for 8 hours. The cell lysates were separated into nuclear and cytoplasmic fractions18 quickly to avoid degradation of HIF1. HIF1 protein induced by hypoxia was exclusively located in the nuclear fraction (Figure 2B). There was no detectable HIF1 in the cytosolic fraction (not shown). Expression of DN-BMK1 significantly increased HIF1 expression in response to hypoxia (Figure 2B, 1% O2). More dramatic was the appearance of HIF1 in the nuclear fraction of cells maintained in 21% O2 indicating a profound effect of DN-BMK1 on HIF1 protein stability. In contrast, CA-MEK5 completely abrogated the increase in nuclear HIF1 induced by 1% O2 (Figure 2B). These results show that BMK1 kinase activity regulates endogenous HIF1 protein level.
To further support these observations, we monitored expression and subcellular localization of transfected HIF1 tagged with GFP (WT HIF1-GFP, Figure 2C through 2E) in presence of DN-BMK1 under hypoxia versus normoxia. In normoxic conditions there was barely detectable WT HIF1-GFP in cytosol (apparent as green fluorescence with long exposures, Figure 2C) or nucleus (DAPI, Figure 2C). Under hypoxia, WT HIF1-GFP was readily detectable in cytosol (green, Figure 2D) and was obviously increased in the nucleus (green, Figure 2D). Merging the DAPI (nuclear) and GFP images showed clear evidence for increased WT HIF1-GFP in the nucleus. Cotransfection of DN-BMK1 and WT HIF1-GFP had a dramatic effect in normoxia. Wild type HIF1-GFP was readily apparent in the nucleus (green, Figure 2E) and colocalized with DAPI (pale blue, Figure 2E merge GFP/DAPI). There was essentially no WT HIF1-GFP in the cytosol (no yellow in Figure 2E merge Flag/GFP). These results show that DN-BMK1 has a powerful effect to enhance HIF1 protein stability and nuclear localization in endothelial cells that should correlate with increased HIF1 transcriptional activity.
HIF1 Protein Degradation Via Ubiquitination Is Enhanced by Activated BMK1
The effect of BMK1 activation on regulation of HIF1 protein level could be caused by alterations in HIF1 mRNA or protein stability. There was no significant change in HIF1 mRNA level with CA-MEK5 overexpression (data not shown). Therefore the effect of BMK1 activation on HIF1 protein degradation was investigated with pharmacological inhibitors of ubiquitination. BLMEC were transfected with DM-HIF1 and CA-MEK5 or pcDNA3.1 as control. The effects of 3 different ubiquitination inhibitors - acetyl-leucinal-leucinal-nonleucinal (ALLN), lactacystin, and MG132 - were then studied. The inhibitors had no effect on HIF1 protein levels in the presence of pcDNA3.1 (Figure 3). CA-MEK5 significantly decreased HIF1 protein level in the absence of inhibitor (0 eol/L in Figure 3). However, all 3 ubiquitination inhibitors prevented CA-MEK5-dependent downregulation of HIF1 at low doses. Although ALLN inhibits both calpain and ubiquitination-dependent proteolysis, the fact that another calpain inhibitor ALLM did not have effect on CA-MEK5-dependent regulation of HIF1 (Figure 3) suggests that BMK1 activation increased HIF1 protein degradation via an ubiquitination pathway, which is independent of proline hydroxylation on Pro 402 and Pro 564.
HIF1 Induced VEGF Gene Expression is Inhibited by BMK1
To determine the functional effects of BMK1 activation on HIF1-mediated gene expression in EC, we studied the expression of VEGF, a known hypoxia-inducible and HIF1-dependent gene.3,4,22 VEGF mRNA level increased by 1.8±0.3-fold after hypoxia for 24 hours (Figure 4A and 4B). Overexpression of CA-MEK5 significantly reduced VEGF mRNA expression induced by hypoxia. The basal level of VEGF in 21% O2 was also decreased significantly (Figure 4A and 4B). Importantly, cotransfection of DN-BMK1 with CA-MEK5 prevented the inhibition of VEGF mRNA expression significantly (Figure 4A and 4B). Tyrosine phosphorylation of the VEGF receptor VEGFR2 was increased dramatically by hypoxia, which was completely blocked by BMK1 activation with CA-MEK5 (Figure 4C), suggesting a paracrine or autocrine effect of VEGF stimulated by hypoxia. The protein levels of VEGFR2 and VEGFR1 were not changed by either hypoxia or BMK1 activation (Figure 4C).
BMK1 Activation Attenuated EC Migration and Tube Formation
EC migration is important in angiogenesis and is regulated by HIF1 target genes such as VEGF.3,4,22 To measure the effect of BMK1 activation on EC migration, a scratch wound healing assay was performed in which the area recovered represents EC migration. Control cells transfected with pcDNA3.1 recovered 25.9±8.4% after 24 hours (Figure 5A, column 1). Transfection of BLMEC with DM-HIF1 increased EC migration by 42.2±6.8% (1.6-fold increase, Figure 5A). Transfection with CA-MEK5 completely inhibited the pro-migratory effects of DM-HIF1 (Figure 5A). In contrast, coexpression of DN-BMK1 relieved the inhibitory effect of CA-MEK5 (43.0±5.1% recovery). These results suggest that BMK1 activation, by inhibiting HIF1 activity, decreases VEGF expression and inhibits EC migration.
To further assess the effect of BMK1 on EC angiogenesis we performed a Matrigel tube-forming assay. When conditioned medium (CM) from control cells was placed on endothelial cells in Matrigel there was infrequent formation of capillary-like tubes (Figure 5B). In response to plating on Matrigel, EC treated with CM from cells transfected with DM-HIF1 rapidly formed capillary-like tube structures (Figure 5B, second image). Cells treated with CM from cells transfected with CA-MEK5 alone did not differ from control. Importantly, in cells treated with CM from cells transfected with CA-MEK5 and DM-HIF1, there was a substantial inhibition of tube formation and branching (Figure, 5B, bottom). The results of these experiments were quantified as described in Methods. The tube length of cells treated with CM collected from the cells expressing DM-HIF1 increased from 64±5 mm/cm2 to 304±12 mm/cm2 (Figure 5C, second white bar). CM harvested from cells cotransfected with CA-MEK5 reduced tube formation from 304±12 mm/cm2 to 75±5 mm/cm2 (Figure 5C, fourth white bar). Similarly, the tube length of cells transfected with DM-HIF1 increased significantly from 92±18 mm/cm2 to 508±39 mm/cm2 (Figure 5C, second black bar). However, coexpression of CA-MEK5 in the cells inhibited tube formation significantly from 508±39 mm/cm2 to 152±11 mm/cm2 (Figure 5C, fourth black bar). The more abundant formation of capillary-like tubes in transfected cells compared with CM-treated cells suggests that autocrine VEGF secretion contributes to angiogenesis, although it is likely not the only mediator.
Cells treated with CM from DM-HIF1 overexpressing cells demonstrated an increased number of formed enclosures (Figure 5D, from 8±5 cmeC2 to 178±21 cmeC2). In contrast, CM from cells in which CA-MEK5 was overexpressed had no enclosures (0±0 enclosures). Importantly, CM from cells expressing both CA-MEK5 and DM-HIF1 had essentially no angiogenesis as measured by enclosure number (Figure 5D). Similar data were obtained when we measured enclosure number in cells transfected with DM-HIF1 or CA-MEK5. The number of enclosure tubes increased in cells that overexpressed DM-HIF1 from 6±4 cmeC2 to 378±49 cmeC2 (Figure 5D). CA-MEK5 coexpression in the DM-HIF1 overexpressing cells inhibited enclosure number dramatically (Figure 5D).
Discussion
The major findings of this study are that BMK1 is a novel regulator of HIF1 and angiogenesis by increasing HIF1 ubiquitination and degradation in EC. BMK1 was recently shown to play a critical role in EC function because the conditional BMK1 knockout mouse rapidly dies because of disruption of vascular integrity.16 We recently showed an essential role for BMK1 in preventing EC apoptosis induced by serum deprivation and tumor necrosis factor-.17 The present study further supports the key role of BMK1 in EC function. In particular, our data show a specific role for BMK1 in regulating HIF1 protein expression and subsequently VEGF expression during hypoxia. An interesting finding is that the effect of BMK1 to promote HIF1 degradation was independent of hydroxylation on Pro 402 and Pro 564, because the HIF1 mutant (P402A/P564G) lacking the consensus prolines was still ubiquitinated and degraded after BMK1 activation.
The mechanisms that regulate HIF1 function have been extensively studied.8,23eC26 In addition to changes in transcription, HIF1 protein level is highly regulated by post-translational mechanisms including hydroxylation and phosphorylation. Indeed, HIF1 is continuously synthesized but degraded via ubiquitination and proteasome activity under normoxia, whereas it accumulates rapidly following exposure to hypoxia. A key aspect of HIF1 degradation is hydroxylation of 2 proline residues (Pro 402 and Pro 564) that promotes HIF1 interaction with an E3 ubiquitin ligase complex targeting HIF1 for proteasomal degradation.24,25 Phosphorylation of HIF1 by ERK1/2 has been shown to enhance HIF1 transcriptional activity,6,12 although phosphorylation did not increase HIF1 stability. The present results indicate that BMK1-mediated HIF1 degradation is independent of proline hydroxylation (because the mutant (P402A/P564G) was degraded) although it requires ubiquitination (because ubiquitination inhibitors blocked the BMK1 effect). Our data also show that BMK1 kinase activity is required because DN-BMK1 blocked the effect of CA-MEK5 to promote HIF1 degradation. Future studies will be required to determine whether the kinase effect of BMK1 is via direct phosphorylation of HIF1 or a regulatory intermediate.
BMK1 has been shown to play a critical role in EC homeostasis as shown by the findings that the conditional BMK1 knockout mouse loses vascular integrity and exhibits EC apoptosis.16 Based on the results of the present study and several published studies,15,17 it appears that BMK1 plays a critical role in EC by inhibiting apoptosis and limiting VEGF expression. In particular, the present study suggests that BMK1 mediates many of these effects by altering HIF1 expression (via degradation) and activity. It is somewhat counterintuitive that BMK1 would limit VEGF expression, which is growth promoting for EC. However, BMK1 inhibits EC apoptosis by stimulating Bad phosphorylation and decreasing caspase activation independent of growth factor stimulation. We suggest that BMK1 may play a key role in the initial sprouting stages of angiogenesis or the final differentiation of EC in the maturing vessel. The fact that BMK1 activity also regulates EC migration (Figure 5) suggests that BMK1 may be particularly important in the initial stages of angiogenesis. The negative regulation of VEGF expression may enable the earliest sprouts to maintain vascular integrity and prevent increases in permeability that may be disadvantageous in the early angiogenic response.
The findings that BMK1 activity regulates angiogenesis (via HIF1 and VEGF expression), EC migration, and apoptosis suggests that BMK1 is a potential target for therapies that modulate wound repair, tumor angiogenesis and atherosclerosis. In particular, stimulating BMK1 activity would have beneficial effects to limit edema formation, to promote EC survival, and to decrease cancer angiogenesis. It will be important in the future to discover the BMK1 substrates that mediate these clinically important effects in EC.
Acknowledgments
This work was supported by NIH grant HL64839 to B.C.B.
References
Jain RK. Molecular regulation of vessel maturation. Nat Med. 2003; 9: 685eC693.
Carmeliet P. Angiogenesis in health and disease. Nat Med. 2003; 9: 653eC660.
Plate KH, Breier G, Weich HA, Risau W. Vascular endothelial growth factor is a potential tumour angiogenesis factor in human gliomas in vivo. Nature. 1992; 359: 845eC848.
Shweiki D, Itin A, Soffer D, Keshet E. Vascular endothelial growth factor induced by hypoxia may mediate hypoxia-initiated angiogenesis. Nature. 1992; 359: 843eC845.
Semenza GL. Regulation of mammalian O2 homeostasis by hypoxia-inducible factor 1. Annu Rev Cell Dev Biol. 1999; 15: 551eC578.
Minet E, Arnould T, Michel G, Roland I, Mottet D, Raes M, Remacle J, Michiels C. ERK activation upon hypoxia: involvement in HIF-1 activation. FEBS Lett. 2000; 468: 53eC58.
Iyer NV, Kotch LE, Agani F, Leung SW, Laughner E, Wenger RH, Gassmann M, Gearhart JD, Lawler AM, Yu AY, Semenza GL. Cellular and developmental control of O2 homeostasis by hypoxia- inducible factor 1 alpha. Genes Dev. 1998; 12: 149eC162.
Bruick RK, McKnight SL. A conserved family of prolyl-4-hydroxylases that modify HIF. Science. 2001; 294: 1337eC1340.
Semenza G. Signal transduction to hypoxia-inducible factor 1. Biochem Pharmacol. 2002; 64: 993eC998.
Pugh CW, Ratcliffe PJ. Regulation of angiogenesis by hypoxia: role of the HIF system. Nat Med. 2003; 9: 677eC684.
Schofield CJ, Ratcliffe PJ. Oxygen sensing by HIF hydroxylases. Nat Rev Mol Cell Biol. 2004; 5: 343eC354.
Richard DE, Berra E, Gothie E, Roux D, Pouyssegur J. p42/p44 mitogen-activated protein kinases phosphorylate hypoxia- inducible factor 1alpha (HIF-1alpha) and enhance the transcriptional activity of HIF-1. J Biol Chem. 1999; 274: 32631eC32637.
Berra E, Milanini J, Richard DE, Le Gall M, Vinals F, Gothie E, Roux D, Pages G, Pouyssegur J. Signaling angiogenesis via p42/p44 MAP kinase and hypoxia. Biochem Pharmacol. 2000; 60: 1171eC1178.
Pages G, Guerin S, Grall D, Bonino F, Smith A, Anjuere F, Auberger P, Pouyssegur J. Defective thymocyte maturation in p44 MAP kinase (Erk 1) knockout mice. Science. 1999; 286: 1374eC1377.
Sohn SJ, Sarvis BK, Cado D, Winoto A. ERK5 MAPK regulates embryonicangiogenesis and acts as a hypoxia- sensitive repressor of vascular endothelial growth factor expression. J Biol Chem. 2002; 277: 43344eC43351.
Hayashi M, Kim SW, Imanaka-Yoshida K, Yoshida T, Abel ED, Eliceiri B, Yang Y, Ulevitch RJ, Lee JD. Targeted deletion of BMK1/ERK5 in adult mice perturbs vascular integrity and leads to endothelial failure. J Clin Invest. 2004; 113: 1138eC1148.
Pi X, Yan C, Berk BC. Big Mitogen-Activated Protein Kinase (BMK1)/ERK5 Protects Endothelial Cells From Apoptosis. Circ Res. 2004; 94: 362eC369.
Armesilla AL, Lorenzo E, Gomez del Arco P, Martinez-Martinez S, Alfranca A, Redondo JM. Vascular endothelial growth factor activates nuclear factor of activated T cells in human endothelial cells: a role for tissue factor gene expression. Mol Cell Biol. 1999; 19: 2032eC2043.
Kim D, Rybalkin SD, Pi X, Wang Y, Zhang C, Munzel T, Beavo JA, Berk BC, Yan C. Upregulation of phosphodiesterase 1A1 expression is associated with the development of nitrate tolerance. Circulation. 2001; 104: 2338eC2343.
Poppa V, Miyashiro JK, Corson MA, Berk BC. Endothelial NO synthase is increased in regenerating endothelium after denuding injury of the rat aorta. Arterioscler Thromb Vasc Biol. 1998; 18: 1312eC1321.
Yan C, Takahashi M, Okuda M, Lee JD, Berk BC. Fluid shear stress stimulates big mitogen-activated protein kinase 1 (BMK1) activity in endothelial cells. Dependence on tyrosine kinases and intracellular calcium. J Biol Chem. 1999; 274: 143eC150.
Yamakawa M, Liu LX, Date T, Belanger AJ, Vincent KA, Akita GY, Kuriyama T, Cheng SH, Gregory RJ, Jiang C. Hypoxia-inducible factor-1 mediates activation of cultured vascular endothelial cells by inducing multiple angiogenic factors. Circ Res. 2003; 93: 664eC673.
Ivan M, Kondo K, Yang H, Kim W, Valiando J, Ohh M, Salic A, Asara JM, Lane WS, Kaelin WG Jr. HIFalpha targeted for VHL-mediated destruction by proline hydroxylation: implications for O2 sensing. Science. 2001; 292: 464eC468.
Jaakkola P, Mole DR, Tian YM, Wilson MI, Gielbert J, Gaskell SJ, Kriegsheim A, Hebestreit HF, Mukherji M, Schofield CJ, Maxwell PH, Pugh CW, Ratcliffe PJ. Targeting of HIF-alpha to the von Hippel-Lindau ubiquitylation complex by O2-regulated prolyl hydroxylation. Science. 2001; 292: 468eC472.
Yu F, White SB, Zhao Q, Lee FS. HIF-1alpha binding to VHL is regulated by stimulus-sensitive proline hydroxylation. Proc Natl Acad Sci U S A. 2001; 98: 9630eC9635.
Lando D, Peet DJ, Whelan DA, Gorman JJ, Whitelaw ML. Asparagine hydroxylation of the HIF transactivation domain a hypoxic switch. Science. 2002; 295: 858eC861.(Xinchun Pi, Gwenaele Gari)
Abstract
Big MAP kinase 1 (BMK1 or ERK5) is a key mediator of endothelial cell (EC) function as shown by impaired embryonic angiogenesis and vascular collapse in BMK1 knockout mice. Hypoxia inducible factor 1 (HIF1), a potent mediator of angiogenesis, is positively regulated by the MAP kinases, ERK1/2. Because BMK1 deficiency is associated with impaired angiogenesis we hypothesized that BMK1 might regulate HIF1. To test this hypothesis, bovine lung microvascular ECs (BLMECs) were transfected with HIF1 and BMK1 cDNAs, and stimulated by hypoxia. HIF1 activity was measured by a reporter gene assay in which luciferase expression was driven by HIF1 activation. Hypoxia (1% O2, 24 hours) stimulated HIF1 activity by 5.1±0.6 fold. In the presence of dominant negative (DN)-BMK1, which inhibited BMK1 activity, hypoxia induced HIF1 activity was enhanced significantly to 6.4±0.4 fold. BMK1 activation by constitutively active (CA)-MEK5 inhibited HIF1 activity by 46±4%, suggesting BMK1 functions as a negative regulator of HIF1 activation. Activation of BMK1 reduced HIF1 protein levels. Ubiquitination inhibitors (30 eol/L ALLN, 2 eol/L lactacystin, or 100 nmol/L MG132) reduced the BMK1-mediated effect on HIF1 expression by >80%, suggesting that BMK1 stimulated HIF1 proteolysis. The negative effect of BMK1 on HIF1 was functionally important because transfection with CA-MEK5 significantly decreased EC migration by 68±10%, and inhibited angiogenesis (in vitro Matrigel assay) by 76±7%. In summary, BMK1 is a novel negative regulator of HIF1 and angiogenesis by increasing HIF1 ubiquitination and inhibiting HIF1 activity in endothelial cells.
Key Words: big MAP kinase 1 hypoxia inducible factor 1 angiogenesis
Introduction
Endothelial cells (EC) play a key role in all aspects of angiogenesis.1,2 Understanding the specific mechanisms that contribute to each step of the angiogenic response may offer the best approach to developing therapies that will augment angiogenesis in diseases such as stroke and myocardial infarction or inhibit angiogenesis in cancer. An important master regulator of angiogenesis (especially induced by hypoxia) is hypoxia inducible factor-1 (HIF1), which controls critical proangiogenic genes such as vascular endothelial growth factor (VEGF).3eC6 Targeted inactivation of HIF1 in mice results in abnormal vascular development and embryonic lethality, demonstrating its key role in the angiogenic processes.7 Multiple mechanisms regulate HIF1 including both transcriptional and posttranslational processes. In EC, posttranslational mechanisms include both protein degradation (mediated by proline hydroxylases and ubiquitination) and protein phosphorylation.8eC11
Previous studies suggest that HIF1 function is regulated by mitogen activated protein (MAP) kinases.6,12,13 Specifically it was shown that ERK1/2 was activated by hypoxia, phosphorylated HIF1 directly, and enhanced HIF1 transcriptional activity as measured by VEGF expression.6,12 The mechanisms by which ERK1/2 stimulates HIF1 transcriptional activity remain unknown because HIF1 phosphorylation by ERK1/2 did not change HIF1 protein stability. However, the ERK1 knockout mouse did not exhibit any defects in angiogenesis.14 In contrast, several recent articles15eC17 suggest a critical role for the closely related MAP kinase, BMK1, in angiogenesis. Sohn et al showed in the BMK1 knockout mouse that initial vascularization occurred normally but subsequent remodeling and maintenance of the vasculature was adversely affected.15 A similar result was reported by Hayashi et al who showed in a conditional BMK1 knockout that vascular integrity was compromised and EC apoptosis increased.16 We hypothesized that the abnormalities in BMK1 knockout vasculature may reflect BMK1-mediated regulation of HIF-1. This hypothesis derived from an article showing that VEGF expression was upregulated in BMK1 knockout embryonic fibroblasts and overexpression of BMK1 cDNA inhibited hypoxia induced VEGF promoter activity.15 The authors suggested that the phenotype of BMK1 null mice was possibly because of disregulation of HIF activity, which disrupted normal vasculogenesis and angiogenesis by overexpression of VEGF. Also, because BMK1 and ERK1/2 have the same substrate recognition motif for phosphorylation, we propose that BMK1 regulates HIF activity in endothelial cells.
In this study, we demonstrate an essential role for BMK1 as a negative feedback regulator of HIF function as measured by VEGF expression, EC migration, and angiogenesis. Furthermore, we show that BMK1 regulation of HIF1 occurs via an ubiquitin dependent degradation mechanism.
Materials and Methods
Reagents
Bovine lung microvascular ECs (BLMEC)s were purchased from VEC Technologies (Renseller, NY). Endothelial basal medium and supplements were purchased from Invitrogen. Mouse anti-GFP antibody and mouse anti-PCNA were from BD Biosciences (San Jose, Calif). Mouse anti-VEGFR2, goat anti-VEGFR1, rabbit anti-pERK1/2, rabbit anti-ERK1/2, rabbit anti-actin, and rabbit anti-HA antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, Calif). Mouse anti-4G10 monoclonal antibody was purchased from Upstate (San Diego, Calif). Rabbit anti-BMK1 and monoclonal anti-FLAG antibodies were bought from Sigma Company. Monoclonal HIF1 antibody was purchased from Novus Biological (Eugene, Ore). A goat anti-mouse, Alexa Fluor 546 (Molecular Probes, Littleton, Co) was used for immunofluorescence. ALLN, Lactacystin, MG132, ALLM were from Calbiochem (San Diego, Calif). Matrigel matrix was purchased from BD Biosciences.
Cell Culture and Transfection
BLMECs at passages 3 to 8 were grown in MCDB-131 media supplemented with 10% FBS, heparin (15300 U/L, Sigma; St. Louis, Mo), hydrocortisone (2.76 eol/L, Sigma), bovine pituitary extract, epidermal growth factor (1.64 nmol/L, Sigma), L-glutamine, and antibiotics (100 U/mL penicillin, 68.6 mol/L streptomycin) in flasks precoated with 2% gelatin. For transient expression experiments, 95% confluent cells in 60 mm dishes were transfected 24 hours after plating with 2 e plasmids using 7 e蘈 Lipofectamine and 10 e蘈 Plus reagent (Invitrogen, Carlsbad, Calif). After 4 hours of incubation, the medium was replaced with complete MCDB-131. One day later, cells were serum starved overnight, then incubated in the hypoxia chamber (1% O2). This protocol yields 60% transfection efficiency based on GFP.17
HIF1-Dependent Reporter Gene Assay
BLMECs were cotransfected with hypoxia response element-firefly-luciferase, TK-renilla-luciferase and CA-MEK5 or other cDNAs indicated in each experiment. After treatment, the luciferase activities in cell lysates were determined using the Dual-Luciferase Reporter Assay system (Promega, Madison, Wis) and Wallac 1420 multilabel counter. Luciferase activity represents the ratio of the luminescence units (RLU) from firefly luciferase and renilla luciferase.8
Western Blot Analysis
The membrane was blocked for 2 hours at room temperature with 5% BSA. The blot was incubated for 2 hours with certain dilution of antibody at room temperature, followed by incubation for 1 hour with horseradish peroxidase-conjugated secondary antibody. Immunoreactive bands were visualized using ECL (Amersham Pharmacia Biotech). Alternatively, the infrared secondary antibody was used and images were analyzed with Odyssey software (LiCor, Inc).
Cell Fractionation
After the different transfection and treatment, BLMEC were washed twice with cold PBS and lysed in hypotonic buffer (10 mmol/L Tris-HCl pH7.5, 10 mmol/L NaCl, 3 mmol/L MgCl2, 1 mmol/L PMSF, 0.5 mmol/L DTT, 0.1 mmol/L EGTA, 2 eol/L leupeptin, 1 e of aprotinin per mL, and 0.05% NP-40).18 The supernatant containing the cytosolic extract was removed and the protein concentration determined by Bradford assay. The nuclei were washed twice with the hypotonic buffer without NP-40 and scraped in Laemmli buffer. The same volume of nuclei extracts was loaded for each sample. PCNA and actin were used as controls of loading for nuclear and cytosolic fractions respectively.
Immunofluorescence
BLMEC were fixed in 3% paraformaldehyde for 15 minutes at room temperature. After 3 washes with PBS, the cells were sequentially treated with 0.1% Triton X-100 for 5 minutes (for permeabilization), with PBS/3% BSA for 1 hour (for blocking), then with an anti-FLAG for 2 hours in PBS/1%BSA. After 3 washes, cells were incubated in dark with antimouse an Alexa Fluor 546 in PBS for 45 minutes at room temperature. After 4 washes in PBS, the cells were counterstained with DAPI and then the fluorescent signal was visualized with an Olympus fluorescence microscope.
Relative Semiquantitative PCR
PCR was performed as previously described.19 RNA of BLMECs was extracted with a Total RNA Isolation Kit (Qiagen). First-strand cDNA was synthesized with the SuperScript Preamplification System (Gibco-BRL). Relative quantitative reverse transcriptioneCpolymerase chain reaction was performed with 18s rRNA as an internal control using Ambion’s competimer technology. VEGF primers (sense: CTTTCTGCTCTCTTGGGTG C; anti-sense: ATCTTCAAGCCATCCTGTGTCCC) were used to generate the 224-bp PCR product. PCR products were separated on a 2% agarose gel, stained with ethidium bromide, and visualized under UV light. The relative intensity of VEGF PCR products was determined by densitometry.
Migration Assay
For detection of cell migration, a wound healing assay was performed.20 BLMECs were grown on 6 cm wells, the monolayer was scratched with a sterile disposable rubber policeman and the edge labeled with a traced line. After injury, the cells were gently washed with normal medium without serum. EC migration from the edge of the injured monolayer was quantified by measuring the area between the wound edges before and the recovered area after injury using light microscopy and the computer program ImageJ.
Angiogenesis Assay
The angiogenic response to HIF1 induction affected by BMK1 activation was assessed using an in vitro Matrigel analysis, using BLMECs transfected with DM-HIF1 and CA-MEK5, or pcDNA3.1 as control. Two days after transfection, the cells were seeded on 6-well cell culture plates precoated with Matrigel at 1x105 cells/well and maintained in serum free medium. The capillary/tube-like structures were visualized by light microscopy at different time points and analyzed using ImageJ.
Statistical Analysis
Data were shown as mean±STDEV (S.E., for data analysis of angiogenesis assay in Figure 5C and 5D) for 3 to 4 separate experiments. Differences were analyzed with one-way analysis of variance (ANOVA). Values of P<0.05 were considered statistically significant.
Results
HIF1 Transcriptional Activity Is Inhibited by BMK1 Activation
To assess the effect of BMK1 on HIF1 function we used CA-MEK5 to stimulate BMK1 activity and measured HIF1-dependent reporter gene expression.6,8,17 CA-MEK5 overexpression in BLMEC increased BMK1 activity by 10-fold.17 Endogenous HIF1 activity was inhibited by 39.0±4.2% after CA-MEK5 overexpression (Figure 1A, second bar). To further assess the effects of BMK1 on HIF1 we also transfected BLMEC with cDNAs of wild-type (WT) or double mutant (DM)- HIF1 (P402A / P564G). DM-HIF1 is stabilized compared with WT-HIF1 protein because mutations of the proline hydroxylase recognition sites Pro 402 and Pro 564 prevent ubiquitination of HIF1.11 Transfection with WT-HIF1 increased luciferase activity by 4.0±0.1-fold compared with endogenous HIF1. CA-MEK5 inhibited HIF1 activity by 68.9±5.8% (Figure 1A, bars 3 and 4). In cells transfected with DM-HIF1, luciferase activity was increased 9.8±0.2 fold. CA-MEK5 dramatically inhibited DM-HIF1 function by 77.9±6.6% (Figure 1A, bars 5 and 6), suggesting that the effect is independent of Pro 402 and Pro 564.
To demonstrate further that BMK1 activation inhibited HIF1 transcriptional activity, we used hypoxia (1% O2) to stimulate HIF1 activity. We first examined the effect of hypoxia on BMK1 activity, and compared the time course and magnitude with activation of ERK1/2.6,12 In response to hypoxia, BMK1 activity increased with peak at 24 hours measured by phosphorylation-dependent bandshift (Figure 1B, 97.5±8.0-fold increase, n=3). We have previously shown an excellent correlation between BMK1 activity measured by bandshift and by in vitro kinase assay.17,21 ERK1/2 activity also increased in response to hypoxia but more rapidly (peak at 6 hours) than BMK1 (Figure 1B). There was no significant change in BMK1 or ERK1/2 protein expression by hypoxia (Figure 1B). HIF1 transcriptional activity was increased 5.0±0.6-fold by hypoxia (for 24 hours) and even further by overexpression of DN-BMK1 (Figure 1C). CA-MEK5 overexpression inhibited HIF1 activity by 44.0±3.7% and this inhibitory effect was prevented by coexpression of DN-BMK1 (Figure 1C). Because ERK1/2 has also been reported to regulate HIF1 function we studied the effect of modulating MEK1 and ERK1/2 activity. Interestingly, HIF1-dependent luciferase activity was significantly increased by CA-MEK1 and inhibited by DN-ERK2 (Figure 1C). These results show that BMK1 and ERK1/2 differentially regulate HIF1 activity. The negative effect of BMK1 on HIF1 activity might serve as a negative feedback regulator of HIF1 to counter-balance the stimulatory effect of ERK1/2.
HIF1 Protein Level Is Decreased by BMK1, but Not by ERK1/2 Activation
The most important mechanism for regulation of HIF1 is protein degradation.8,9,11 To determine whether BMK1 activation decreased HIF1 protein, the protein level of HIF1 was determined by western blotting. First, BLMEC were transfected with either WT-HIF1 or DM-HIF1 in the presence of CA-MEK5 or CA-MEK1 to activate BMK1 or ERK1/2, respectively. Expression of DM-HIF1 is higher than WT-HIF1 because DM-HIF1 has greater protein stability than WT-HIF1 (Figure 2A). HIF1 protein levels were not changed by overexpression of CA-MEK1, which is consistent with previous findings that ERK1/2 activation increased HIF1 transcriptional activity possibly through direct phosphorylation of HIF1 by ERK1/2, without effect on protein degradation.6,12 Interestingly, the level of ectopically expressed HIF1 protein decreased with overexpression of CA-MEK5 (Figure 2A). In addition, the ability of BMK1 to decrease expression of DM-HIF1 lacking the prolines required for ubiquitination and degradation suggested that BMK1 initiated protein degradation via a different mechanism that was independent of Pro 402 and Pro 564.
To confirm the decrease in HIF1 protein by BMK1 activation, endogenous HIF1 was measured after exposure to hypoxia. BLMECs were transfected with CA-MEK5 or DN-BMK1 and exposed to hypoxia for 8 hours. The cell lysates were separated into nuclear and cytoplasmic fractions18 quickly to avoid degradation of HIF1. HIF1 protein induced by hypoxia was exclusively located in the nuclear fraction (Figure 2B). There was no detectable HIF1 in the cytosolic fraction (not shown). Expression of DN-BMK1 significantly increased HIF1 expression in response to hypoxia (Figure 2B, 1% O2). More dramatic was the appearance of HIF1 in the nuclear fraction of cells maintained in 21% O2 indicating a profound effect of DN-BMK1 on HIF1 protein stability. In contrast, CA-MEK5 completely abrogated the increase in nuclear HIF1 induced by 1% O2 (Figure 2B). These results show that BMK1 kinase activity regulates endogenous HIF1 protein level.
To further support these observations, we monitored expression and subcellular localization of transfected HIF1 tagged with GFP (WT HIF1-GFP, Figure 2C through 2E) in presence of DN-BMK1 under hypoxia versus normoxia. In normoxic conditions there was barely detectable WT HIF1-GFP in cytosol (apparent as green fluorescence with long exposures, Figure 2C) or nucleus (DAPI, Figure 2C). Under hypoxia, WT HIF1-GFP was readily detectable in cytosol (green, Figure 2D) and was obviously increased in the nucleus (green, Figure 2D). Merging the DAPI (nuclear) and GFP images showed clear evidence for increased WT HIF1-GFP in the nucleus. Cotransfection of DN-BMK1 and WT HIF1-GFP had a dramatic effect in normoxia. Wild type HIF1-GFP was readily apparent in the nucleus (green, Figure 2E) and colocalized with DAPI (pale blue, Figure 2E merge GFP/DAPI). There was essentially no WT HIF1-GFP in the cytosol (no yellow in Figure 2E merge Flag/GFP). These results show that DN-BMK1 has a powerful effect to enhance HIF1 protein stability and nuclear localization in endothelial cells that should correlate with increased HIF1 transcriptional activity.
HIF1 Protein Degradation Via Ubiquitination Is Enhanced by Activated BMK1
The effect of BMK1 activation on regulation of HIF1 protein level could be caused by alterations in HIF1 mRNA or protein stability. There was no significant change in HIF1 mRNA level with CA-MEK5 overexpression (data not shown). Therefore the effect of BMK1 activation on HIF1 protein degradation was investigated with pharmacological inhibitors of ubiquitination. BLMEC were transfected with DM-HIF1 and CA-MEK5 or pcDNA3.1 as control. The effects of 3 different ubiquitination inhibitors - acetyl-leucinal-leucinal-nonleucinal (ALLN), lactacystin, and MG132 - were then studied. The inhibitors had no effect on HIF1 protein levels in the presence of pcDNA3.1 (Figure 3). CA-MEK5 significantly decreased HIF1 protein level in the absence of inhibitor (0 eol/L in Figure 3). However, all 3 ubiquitination inhibitors prevented CA-MEK5-dependent downregulation of HIF1 at low doses. Although ALLN inhibits both calpain and ubiquitination-dependent proteolysis, the fact that another calpain inhibitor ALLM did not have effect on CA-MEK5-dependent regulation of HIF1 (Figure 3) suggests that BMK1 activation increased HIF1 protein degradation via an ubiquitination pathway, which is independent of proline hydroxylation on Pro 402 and Pro 564.
HIF1 Induced VEGF Gene Expression is Inhibited by BMK1
To determine the functional effects of BMK1 activation on HIF1-mediated gene expression in EC, we studied the expression of VEGF, a known hypoxia-inducible and HIF1-dependent gene.3,4,22 VEGF mRNA level increased by 1.8±0.3-fold after hypoxia for 24 hours (Figure 4A and 4B). Overexpression of CA-MEK5 significantly reduced VEGF mRNA expression induced by hypoxia. The basal level of VEGF in 21% O2 was also decreased significantly (Figure 4A and 4B). Importantly, cotransfection of DN-BMK1 with CA-MEK5 prevented the inhibition of VEGF mRNA expression significantly (Figure 4A and 4B). Tyrosine phosphorylation of the VEGF receptor VEGFR2 was increased dramatically by hypoxia, which was completely blocked by BMK1 activation with CA-MEK5 (Figure 4C), suggesting a paracrine or autocrine effect of VEGF stimulated by hypoxia. The protein levels of VEGFR2 and VEGFR1 were not changed by either hypoxia or BMK1 activation (Figure 4C).
BMK1 Activation Attenuated EC Migration and Tube Formation
EC migration is important in angiogenesis and is regulated by HIF1 target genes such as VEGF.3,4,22 To measure the effect of BMK1 activation on EC migration, a scratch wound healing assay was performed in which the area recovered represents EC migration. Control cells transfected with pcDNA3.1 recovered 25.9±8.4% after 24 hours (Figure 5A, column 1). Transfection of BLMEC with DM-HIF1 increased EC migration by 42.2±6.8% (1.6-fold increase, Figure 5A). Transfection with CA-MEK5 completely inhibited the pro-migratory effects of DM-HIF1 (Figure 5A). In contrast, coexpression of DN-BMK1 relieved the inhibitory effect of CA-MEK5 (43.0±5.1% recovery). These results suggest that BMK1 activation, by inhibiting HIF1 activity, decreases VEGF expression and inhibits EC migration.
To further assess the effect of BMK1 on EC angiogenesis we performed a Matrigel tube-forming assay. When conditioned medium (CM) from control cells was placed on endothelial cells in Matrigel there was infrequent formation of capillary-like tubes (Figure 5B). In response to plating on Matrigel, EC treated with CM from cells transfected with DM-HIF1 rapidly formed capillary-like tube structures (Figure 5B, second image). Cells treated with CM from cells transfected with CA-MEK5 alone did not differ from control. Importantly, in cells treated with CM from cells transfected with CA-MEK5 and DM-HIF1, there was a substantial inhibition of tube formation and branching (Figure, 5B, bottom). The results of these experiments were quantified as described in Methods. The tube length of cells treated with CM collected from the cells expressing DM-HIF1 increased from 64±5 mm/cm2 to 304±12 mm/cm2 (Figure 5C, second white bar). CM harvested from cells cotransfected with CA-MEK5 reduced tube formation from 304±12 mm/cm2 to 75±5 mm/cm2 (Figure 5C, fourth white bar). Similarly, the tube length of cells transfected with DM-HIF1 increased significantly from 92±18 mm/cm2 to 508±39 mm/cm2 (Figure 5C, second black bar). However, coexpression of CA-MEK5 in the cells inhibited tube formation significantly from 508±39 mm/cm2 to 152±11 mm/cm2 (Figure 5C, fourth black bar). The more abundant formation of capillary-like tubes in transfected cells compared with CM-treated cells suggests that autocrine VEGF secretion contributes to angiogenesis, although it is likely not the only mediator.
Cells treated with CM from DM-HIF1 overexpressing cells demonstrated an increased number of formed enclosures (Figure 5D, from 8±5 cmeC2 to 178±21 cmeC2). In contrast, CM from cells in which CA-MEK5 was overexpressed had no enclosures (0±0 enclosures). Importantly, CM from cells expressing both CA-MEK5 and DM-HIF1 had essentially no angiogenesis as measured by enclosure number (Figure 5D). Similar data were obtained when we measured enclosure number in cells transfected with DM-HIF1 or CA-MEK5. The number of enclosure tubes increased in cells that overexpressed DM-HIF1 from 6±4 cmeC2 to 378±49 cmeC2 (Figure 5D). CA-MEK5 coexpression in the DM-HIF1 overexpressing cells inhibited enclosure number dramatically (Figure 5D).
Discussion
The major findings of this study are that BMK1 is a novel regulator of HIF1 and angiogenesis by increasing HIF1 ubiquitination and degradation in EC. BMK1 was recently shown to play a critical role in EC function because the conditional BMK1 knockout mouse rapidly dies because of disruption of vascular integrity.16 We recently showed an essential role for BMK1 in preventing EC apoptosis induced by serum deprivation and tumor necrosis factor-.17 The present study further supports the key role of BMK1 in EC function. In particular, our data show a specific role for BMK1 in regulating HIF1 protein expression and subsequently VEGF expression during hypoxia. An interesting finding is that the effect of BMK1 to promote HIF1 degradation was independent of hydroxylation on Pro 402 and Pro 564, because the HIF1 mutant (P402A/P564G) lacking the consensus prolines was still ubiquitinated and degraded after BMK1 activation.
The mechanisms that regulate HIF1 function have been extensively studied.8,23eC26 In addition to changes in transcription, HIF1 protein level is highly regulated by post-translational mechanisms including hydroxylation and phosphorylation. Indeed, HIF1 is continuously synthesized but degraded via ubiquitination and proteasome activity under normoxia, whereas it accumulates rapidly following exposure to hypoxia. A key aspect of HIF1 degradation is hydroxylation of 2 proline residues (Pro 402 and Pro 564) that promotes HIF1 interaction with an E3 ubiquitin ligase complex targeting HIF1 for proteasomal degradation.24,25 Phosphorylation of HIF1 by ERK1/2 has been shown to enhance HIF1 transcriptional activity,6,12 although phosphorylation did not increase HIF1 stability. The present results indicate that BMK1-mediated HIF1 degradation is independent of proline hydroxylation (because the mutant (P402A/P564G) was degraded) although it requires ubiquitination (because ubiquitination inhibitors blocked the BMK1 effect). Our data also show that BMK1 kinase activity is required because DN-BMK1 blocked the effect of CA-MEK5 to promote HIF1 degradation. Future studies will be required to determine whether the kinase effect of BMK1 is via direct phosphorylation of HIF1 or a regulatory intermediate.
BMK1 has been shown to play a critical role in EC homeostasis as shown by the findings that the conditional BMK1 knockout mouse loses vascular integrity and exhibits EC apoptosis.16 Based on the results of the present study and several published studies,15,17 it appears that BMK1 plays a critical role in EC by inhibiting apoptosis and limiting VEGF expression. In particular, the present study suggests that BMK1 mediates many of these effects by altering HIF1 expression (via degradation) and activity. It is somewhat counterintuitive that BMK1 would limit VEGF expression, which is growth promoting for EC. However, BMK1 inhibits EC apoptosis by stimulating Bad phosphorylation and decreasing caspase activation independent of growth factor stimulation. We suggest that BMK1 may play a key role in the initial sprouting stages of angiogenesis or the final differentiation of EC in the maturing vessel. The fact that BMK1 activity also regulates EC migration (Figure 5) suggests that BMK1 may be particularly important in the initial stages of angiogenesis. The negative regulation of VEGF expression may enable the earliest sprouts to maintain vascular integrity and prevent increases in permeability that may be disadvantageous in the early angiogenic response.
The findings that BMK1 activity regulates angiogenesis (via HIF1 and VEGF expression), EC migration, and apoptosis suggests that BMK1 is a potential target for therapies that modulate wound repair, tumor angiogenesis and atherosclerosis. In particular, stimulating BMK1 activity would have beneficial effects to limit edema formation, to promote EC survival, and to decrease cancer angiogenesis. It will be important in the future to discover the BMK1 substrates that mediate these clinically important effects in EC.
Acknowledgments
This work was supported by NIH grant HL64839 to B.C.B.
References
Jain RK. Molecular regulation of vessel maturation. Nat Med. 2003; 9: 685eC693.
Carmeliet P. Angiogenesis in health and disease. Nat Med. 2003; 9: 653eC660.
Plate KH, Breier G, Weich HA, Risau W. Vascular endothelial growth factor is a potential tumour angiogenesis factor in human gliomas in vivo. Nature. 1992; 359: 845eC848.
Shweiki D, Itin A, Soffer D, Keshet E. Vascular endothelial growth factor induced by hypoxia may mediate hypoxia-initiated angiogenesis. Nature. 1992; 359: 843eC845.
Semenza GL. Regulation of mammalian O2 homeostasis by hypoxia-inducible factor 1. Annu Rev Cell Dev Biol. 1999; 15: 551eC578.
Minet E, Arnould T, Michel G, Roland I, Mottet D, Raes M, Remacle J, Michiels C. ERK activation upon hypoxia: involvement in HIF-1 activation. FEBS Lett. 2000; 468: 53eC58.
Iyer NV, Kotch LE, Agani F, Leung SW, Laughner E, Wenger RH, Gassmann M, Gearhart JD, Lawler AM, Yu AY, Semenza GL. Cellular and developmental control of O2 homeostasis by hypoxia- inducible factor 1 alpha. Genes Dev. 1998; 12: 149eC162.
Bruick RK, McKnight SL. A conserved family of prolyl-4-hydroxylases that modify HIF. Science. 2001; 294: 1337eC1340.
Semenza G. Signal transduction to hypoxia-inducible factor 1. Biochem Pharmacol. 2002; 64: 993eC998.
Pugh CW, Ratcliffe PJ. Regulation of angiogenesis by hypoxia: role of the HIF system. Nat Med. 2003; 9: 677eC684.
Schofield CJ, Ratcliffe PJ. Oxygen sensing by HIF hydroxylases. Nat Rev Mol Cell Biol. 2004; 5: 343eC354.
Richard DE, Berra E, Gothie E, Roux D, Pouyssegur J. p42/p44 mitogen-activated protein kinases phosphorylate hypoxia- inducible factor 1alpha (HIF-1alpha) and enhance the transcriptional activity of HIF-1. J Biol Chem. 1999; 274: 32631eC32637.
Berra E, Milanini J, Richard DE, Le Gall M, Vinals F, Gothie E, Roux D, Pages G, Pouyssegur J. Signaling angiogenesis via p42/p44 MAP kinase and hypoxia. Biochem Pharmacol. 2000; 60: 1171eC1178.
Pages G, Guerin S, Grall D, Bonino F, Smith A, Anjuere F, Auberger P, Pouyssegur J. Defective thymocyte maturation in p44 MAP kinase (Erk 1) knockout mice. Science. 1999; 286: 1374eC1377.
Sohn SJ, Sarvis BK, Cado D, Winoto A. ERK5 MAPK regulates embryonicangiogenesis and acts as a hypoxia- sensitive repressor of vascular endothelial growth factor expression. J Biol Chem. 2002; 277: 43344eC43351.
Hayashi M, Kim SW, Imanaka-Yoshida K, Yoshida T, Abel ED, Eliceiri B, Yang Y, Ulevitch RJ, Lee JD. Targeted deletion of BMK1/ERK5 in adult mice perturbs vascular integrity and leads to endothelial failure. J Clin Invest. 2004; 113: 1138eC1148.
Pi X, Yan C, Berk BC. Big Mitogen-Activated Protein Kinase (BMK1)/ERK5 Protects Endothelial Cells From Apoptosis. Circ Res. 2004; 94: 362eC369.
Armesilla AL, Lorenzo E, Gomez del Arco P, Martinez-Martinez S, Alfranca A, Redondo JM. Vascular endothelial growth factor activates nuclear factor of activated T cells in human endothelial cells: a role for tissue factor gene expression. Mol Cell Biol. 1999; 19: 2032eC2043.
Kim D, Rybalkin SD, Pi X, Wang Y, Zhang C, Munzel T, Beavo JA, Berk BC, Yan C. Upregulation of phosphodiesterase 1A1 expression is associated with the development of nitrate tolerance. Circulation. 2001; 104: 2338eC2343.
Poppa V, Miyashiro JK, Corson MA, Berk BC. Endothelial NO synthase is increased in regenerating endothelium after denuding injury of the rat aorta. Arterioscler Thromb Vasc Biol. 1998; 18: 1312eC1321.
Yan C, Takahashi M, Okuda M, Lee JD, Berk BC. Fluid shear stress stimulates big mitogen-activated protein kinase 1 (BMK1) activity in endothelial cells. Dependence on tyrosine kinases and intracellular calcium. J Biol Chem. 1999; 274: 143eC150.
Yamakawa M, Liu LX, Date T, Belanger AJ, Vincent KA, Akita GY, Kuriyama T, Cheng SH, Gregory RJ, Jiang C. Hypoxia-inducible factor-1 mediates activation of cultured vascular endothelial cells by inducing multiple angiogenic factors. Circ Res. 2003; 93: 664eC673.
Ivan M, Kondo K, Yang H, Kim W, Valiando J, Ohh M, Salic A, Asara JM, Lane WS, Kaelin WG Jr. HIFalpha targeted for VHL-mediated destruction by proline hydroxylation: implications for O2 sensing. Science. 2001; 292: 464eC468.
Jaakkola P, Mole DR, Tian YM, Wilson MI, Gielbert J, Gaskell SJ, Kriegsheim A, Hebestreit HF, Mukherji M, Schofield CJ, Maxwell PH, Pugh CW, Ratcliffe PJ. Targeting of HIF-alpha to the von Hippel-Lindau ubiquitylation complex by O2-regulated prolyl hydroxylation. Science. 2001; 292: 468eC472.
Yu F, White SB, Zhao Q, Lee FS. HIF-1alpha binding to VHL is regulated by stimulus-sensitive proline hydroxylation. Proc Natl Acad Sci U S A. 2001; 98: 9630eC9635.
Lando D, Peet DJ, Whelan DA, Gorman JJ, Whitelaw ML. Asparagine hydroxylation of the HIF transactivation domain a hypoxic switch. Science. 2002; 295: 858eC861.(Xinchun Pi, Gwenaele Gari)