Differential roles of prostaglandin E-type receptors in activation of hypoxia-inducible factor 1 by prostaglandin E₁ in vascular-derived cells under non-hypoxic conditions

Cell culture and reagents

Human aortic smooth muscle cells (HASMCs) and human umbilical vein endothelial cells (HUVECs) were purchased from Kurabo (Osaka, Japan) (Oda et al., 2009). Human neuroblastoma cell line SH-SY5Y cells obtained from ATCC (Manassas, VA), were maintained in RPMI 1640 medium containing 10% fetal bovine serum (FBS) (Daijo et al., 2011; Takabuchi et al., 2004). HEK293 cells were maintained in Dulbecco’s Modified Eagle medium (DMEM) containing 10% fetal bovine serum (FBS) (Hirota et al., 2004). PGE₁ was obtained from Sigma-Aldrich (St. Louis, MO). Lipo-PGE₁ was from Mitsubishi Tanabe Pharma (Osaka, Japan). PGE₁-Alfadex (7{-(1 R, 2 R , 3R )-3-Hydroxy-2[-(1 E, 3 S)-3-hydroxyoct-1-en-1-yl]-5-oxocyclopentyl} heptanoic acid-α-Cyclodextrin) was from Ono Pharmaceutical Co. (Osaka, Japan). Cycloheximide (CHX), desferrioxamine (DFX), LY294002, PD98059, and GF109203X were obtained from Calbiochem (San Diego, CA). YC-1 was obtained from Cayman Chemical Company (Ann Arbor, MI).

Anti-HIF-1α antibody was purchased from BD Biosciences (San Jose, CA) and anti-HIF-1β antibody was purchased from Novus Biologicals (Littleton, CO). Anti-β-actin antibody was purchased from Sigma-Aldrich.

EP receptor-specific agonists and antagonists

The specific agonists, ONO-DI-004 for EP1, ONO-AE1-259-01 for EP2, ONO-AE-248 for EP3, ONO-AE1-329 for EP4 and the specific antagonists, ONO-8713 for EP1, ONO-AE3-240 for EP3, and ONO-AE2-227 for EP4 were kindly provided by Ono Pharmaceutical Co. (Suzawa et al., 2000).

Hypoxic treatment

Cells were maintained in a multi-gas incubator (APMW-36, Astec, Japan) and exposed to hypoxia (1% O₂, 5% CO₂, and 94% N₂ or 5% O₂, 5% CO₂, and 90% N₂) at 37°C (Tanaka et al., 2010; Tanaka et al., 2011).

Immunoblot assays

Whole cell lysates were prepared using ice-cold lysis buffer following a protocol described previously (Kasuno et al., 2004; Oda et al., 2006; Tanaka et al., 2010). Aliquots of the cell lysates were fractionated by 7.5% SDS-PAGE and subjected to immunoblot assay using mouse monoclonal antibody against HIF-1α (BD Biosciences, San Jose, CA) or HIF-1β (H1b234; Novus Biologicals, Littleton, CO) at 1:1000 dilution and HRP-conjugated mouse monoclonal antibodies against mouse IgG (Amersham Bioscience, Piscataway, NJ, 1:1000 dilution). Anti-β-actin antibody (Sigma, St. Louis, MO) was used at 1:5000 dilution. Chemiluminescent signals were developed using ECL reagent (Amersham Biosciences). The intensity of each band was quantified using Image J software (Oda et al., 2006).

Metabolic labeling assay

The metabolic-labeling protocol is described elsewhere (Kasuno et al., 2004). Briefly, HASMCs were plated in a 10-cm dish and 24 h later were serum-starved for a further 12 h. The cells were pretreated with 10 µM PGE₁ or 100 µM DFX for 30 min in methionine-free Dulbecco’s modified Eagle’s medium. [35S] Met-Cys was added to a final concentration of 0.3 mCi/ml, and the cells were pulse-labeled for 45 min and then harvested. Whole cell extracts were prepared and 1 mg of extract was precleared with 60 µl of protein A-Sepharose for 1 h. Anti-HIF-1 antibody (20 µl) was added to the supernatant and the mixture was rotated overnight at 4°C. Protein A-Sepharose (40 µl) was added and the mixture rotated for 2 h at 4°C, pelleted, and washed five times with 1 ml of radioimmune precipitation buffer. The samples were analyzed by SDS-polyacrylamide gel electrophoresis. The gel was dried and exposed to X-ray film.

Quantitative real-time reverse transcriptase (RT)-PCR analysis (qRT-PCR)

RNA was purified using RNeasy™ (Qiagen, Valencia, CA) and treated with DNase. First-strand synthesis and real-time PCR were performed using the QuantiTect SYBR Green PCR Kit (Qiagen) following the manufacturer’s protocol. Amplification and detection were performed using the Applied Biosystems 7300 Real-time PCR System (Applied Biosystems, Foster City, CA) (Oda et al., 2008). PCR primers were purchased from Qiagen (Oda et al., 2009; Tanaka et al., 2010). The change in expression of each target mRNA relative to 18S rRNA was calculated.

VEGF ELISA

Secreted VEGF protein levels in culture supernatant were determined with an ELISA kit (Quantikine Human VEGF ELISA kit; R & D Systems) according to the manufacturer’s instruction. Values of optical density were measured spectrophotometrically (Microplate Reader 680 XR; Bio-Rad) at 450 nm (correction wavelength set at 570 nm). The standard curve was made with Prism™ software. In each experiment, all samples and standards were measured in duplicate. The amount of VEGF released in the supernatant was expressed in pg/ml (Tanaka et al., 2011).

In vitro vascular permeability assay

The assay kit was purchased from Milliore (Billerica, MA). HUVEC cells were seeded at 200,000 cells per insert and cultured for 72 h. HUVEC monolayer permeability was tested after a 24 h starvation period. Test inserts were then treated for 18 h with 1 µM PGE₁ in basal medium. Monolayers were also treated with basal medium and growth medium only. Inserts were also tested without the cell monolayer. The fluorescence of the plate well solution was determined using a standard fluorescent plate reader.

Transepithelial electrical resistance measurements

Measurement of transepithelial electrical resistance in HUVEC monolayer was performing using a Millicell electric resistance system (ERS) ohmmeter (Millipore Corporation, Bedford, MA) as a barrier function assay (Kimura et al., 2008). This device can measure electric resistance of epithelial cells in culture using a separate pair of Ag-AgCl electrolodes and a resistance meter. Fluid resistance was subtracted and net resistance was calculated as ohm-square centimeter. Results are presented as the mean ± S.D. of three independent wells.

Statistical analysis

Data are presented as means ± S.D. Data were analyzed using one-way analysis of variance followed by Dunnett’s multiple comparison test using Prism version 4c. P < 0.05 was considered statistically significant (Tanaka et al., 2010; Tanaka et al., 2011).

PGE₁ induces HIF-1α protein accumulation in vascular-derived cells

With the aim of evaluating the effect of PGE₁ (alprostadil) on HIF-1 activity, HASMCs and HUVECs were treated with or without 1 or 10 µM PGE₁ under 20% O₂ conditions for 4 h and HIF-1 protein expression was investigated (Figs. 1A–1D). HIF-1α protein expression was low under 20% O₂ conditions (lane 1) and increased in response to PGE₁ treatment in both HASMCs and HUVECs (lanes 2 and 3) as well as under 1% O₂ conditions (lane 4) (Figs. 1A and 1C). HIF-1β protein expression was not affected by PGE₁ treatment in either HASMCs or HUVECs. β-actin expression remained constant during treatment. Densitometric analysis showed that both 1 µM and 10 µM PGE₁ statistically significantly induced HIF-1α protein expression as early as 4 h (Figs. 1B and 1D). With the aim of constructing a time course of the effect of PGE₁ on HIF-1α protein accumulation, cells were exposed to 1 µM PGE₁ for 2, 4, and 8 h. The time-course study demonstrated that 1 µM PGE₁ induced HIF-1α protein expression as early as 1 h with peak activity at 4 h (lanes 4) in both HASMCs (Fig. 1E) and HUVECs (Fig. 1F). No toxic effect was detected as judged from an MTT assay (data not shown). PGE₁ did not induce HIF-1α protein accumulation in HEK293 cells (Fig. 1G). In contrast, 1 µM PGE₁ did not affect HIF-2α protein expression under 20% O₂ conditions in either HASMCs or HUVECs at 4 h (Fig. 1H). There are several preparations of PGE₁ available in clinical settings. We tested the two PGE₁ analogues; lipo-PGE₁ and PGE₁-Alfadex (Fig. 1I). Both lipo-PGE₁ and PGE₁-Alfadex induced HIF-1α protein accumulation in HASMCs under normoxic conditions.

PGE₁ induces HIF-1 activation and expression of its downstream genes

We next investigated the effect of PGE₁ on HIF-1-mediated gene expression. The mRNA expression of genes was assayed by qRT-PCR. PGE₁ treatment induced VEGF and GLUT1 mRNA expression in a PGE1-dose dependent manner under 20% O₂ conditions in both HASMCs (Fig. 2A) or HUVECs (Fig. 2B) at 4 h. The fold induction of VEGF mRNA by PGE1 was less than that of hypoxic exposure (1% O₂ conditions). In contrast, the induction of GLUT1 mRNA was comparable to that of hypoxic exposure. This is consistent with the results of HIF-1α protein expression (Figs. 1A and 1C). The expression of either HIF-1α or HIF-1β mRNA was not affected by PGE₁. Moreover, PGE₁ induced VEGF protein secretion in both HASMCs (Fig. 3A) or HUVECs (Fig. 3B). The induction of VEGF mRNA in HASMCs was inhibited by the inhibitor of HIF-1 activation YC-1 or downregulation of HIF-1α expression by siRNA, indicating that the effect of PGE₁ is dependent on HIF-1 activation (Fig. 3C). Glut1 mRNA expression was also affected by YC-1 and downregulation of HIF-1α expression by siRNA similarly to VEGF mRNA (Fig. 3D).

Differential involvement of EP receptors in PGE₁-induced HIF-1 activation

Four types of EP receptors have been identified for PGE₁ (Sugimoto & Narumiya, 2007). We sought to identify those responsible for PGE₁-induced HIF-1 activation. Expression of EP receptors in HUVECs and HSMACs was examined by qRT-PCR. It is reported that SH-SY5Y cells express EP1, EP2, EP3, and EP4 receptors (Ji et al., 2010). As shown in Fig. 4A, HASMCs express all types of EP receptor. HUVECs express the EP1, EP3, and EP4 receptors. The EP2 receptor was not detected in HUVECs.

Among EP receptor-subtype-specific ligands, the EP1-specific ligand ONO-DI-004 (Fig. 4B, lanes 2) and the EP3-specific ligand ONO-AE-248 (Fig. 4B; lanes 4) induced HIF-1α protein accumulation in both HASMCs and HUVECs. In contrast, neither the EP2-specific ligand ONO-AE1-259-01 (Fig. 4B; lanes 3) nor the EP4-specific ligand ONO-AE1-329 (Fig. 4B; lanes 5) induced HIF-1α protein expression at 1 µM concentration. Consistent with these observations, the EP1-specific antagonist ONO-8713 or the EP3-specific antagonist ONO-AE3-240 inhibited PGE₁-induced HIF-1α protein accumulation in both HASMCs and HUVECs (Fig. 4C). HIF-1β expression was not affected by any agonists of the EP receptors (Fig. 4B). Moreover, qRT-PCR analysis showed that ONO-DI-004 and ONO-AE-248 induced expression of VEGF mRNA in both HASMCs (Fig. 5A) and HUVECs (Fig. 5B) at 4 h. qRT-PCR analysis showed that ONO-8713 and ONO-AE3-240 did suppress but the EP4-specific antagonist ONO-AE2-227 did not suppress VEGF or GLUT1 mRNA induction by PGE₁ expression statistically significantly in both HASMCs (Fig. 5C) or HUVECs (Fig. 5D).

Effect of PGE₁ on stability and synthesis of HIF-1α

The balance between synthesis and degradation determines HIF-1α protein levels. With the aim of exploring the mechanism for HIF-1α protein induction by PGE₁, HIF-1α protein half-life in the cells was examined. HASMCs and HUVECs were treated with PGE₁ or DFX for 4 h to induce HIF-1α protein accumulation. Then, CHX was added to block ongoing protein synthesis. In the presence of CHX, HIF-1α protein expression was not detected at 30 min in PGE₁-treated cells but persisted in DFX-treated cells in both HASMCs and HUVECs (Fig. 6A). These results strongly suggested that HIF-1α expression in PGE₁-treated cells is dependent mainly upon ongoing protein synthesis.

With the aim of determining the rate of HIF-1α synthesis, serum-starved HUVECs were pretreated with PGE₁ or DFX for 30 min and then pulse-labeled with [35S] Met-Cys for 40 min, followed by immunoprecipitation of HIF-1α (Fig. 6B). In contrast to control serum-starved cells, 35S-labeled HIF-1α clearly increased in PGE₁-treated cells (lane 2), whereas the amount of labeled HIF-1α protein did not increase in cells treated with DFX (lane 3), relative to the control. Thus, both CHX addition and metabolic labeling experiments provide evidence for increased synthesis of HIF-1α in response to PGE₁ treatment. Immunoblotting analysis against the same lysates with anti-β-actin antibody was indicated as loading control of the lysates (Fig. 6B).

To examine the signaling pathways leading to HIF-1α protein induction in HUVECs and HASMCs, we focused on various kinases, including PI3K, mitogen-activated protein kinase (MAPK), protein kinase C (PKC) and mammalian target of rapamycin (mTOR) (Fukuda, Kelly & Semenza, 2003).

HIF-1 activity induced by the stimulation of receptor tyrosine kinases or G protein-coupled receptors has been reported to require MAPK and PI3K signaling (Fukuda, Kelly & Semenza, 2003; Hirota et al., 2004). Cells were pretreated with LY294002, PD98059 or GF109203X, which are selective inhibitors of PI3K, MAPK and PKC, respectively, for 30 min, and then treated with 10 µM PGE₁ for 4 h and harvested for immunoblot assay in both HASMCs and HUVECs (Fig. 6C). All the inhibitors inhibited the induction of HIF-1α protein expression by PGE₁. HIF-1β protein expression was not induced nor affected by any of the kinase inhibitors. Consistent with kinase inhibitor-mediated inhibition of HIF-1α protein expression, LY294002, PD98059, or GF109203X also suppressed PGE₁-induced VEGF and GLUT1 mRNA expression (Fig. 6D) in HASMCs.

PGE₁ increases permeability of vascular barrier reconstituted with HUVECs

With the aim of determining whether PGE₁ altered vascular barrier function, in vitro vascular permeability assay using HUVECs were performed (Fig. 5). HUVECs were seeded, allowed to form a monolayer for 72 h, and treated with or without 1 µM PGE₁, the EP1-specific ligand ONO-DI-004 or the EP3-specific ligand ONO-AE-248 for a further 18 h under 20% O₂ conditions and the extent of permeability was then determined as described in Materials and Methods. As shown in Fig. 5A, PGE₁ increased the permeability of vascular barrier attained by HUVECs. The EP1 agonist ONO-DI-004 and the EP3 agonist ONO-AE-248 also decreased in transepithelial resistance as well as PGE₁. The increase of the permeability was partially prevented by treatment with the HIF-1 inhibitor YC-1 or siRNA against HIF-1α (Fig. 7A). In addition, we also examined the permeability adopting transepithelial electrical resistance measurements. Transepithelial resistance was decreased 24 h after treatment with conditioned media derived from HUVECs treated with 1 µM PGE₁, ONO-DI-004, or ONO-AE-248 treatment and the treatment with YC-1 or siRNA against HIF-1α prevented the increase of the permeability (Fig. 7B).