Baicalein Induces Functional Hypoxia-Inducible Factor-1 and Angiogenesis

Targeting the oxygen-sensing mechanisms of the hypoxiainducible factor (HIF) pathway provides pharmacological ways of manipulating the HIF response. Because HIF-1 -specific prolyl-4 hydroxylases (PHDs) prime degradation of HIF-1 , we have made an effort to find a small molecule capable of upregulating the HIF pathway by inhibiting prolyl hydroxylation. Through an in vitro high-throughput screen, we have discovered a PHD2 inhibitor baicalein, which is also found to abrogate asparaginyl hydroxylation of HIF-1 . Such inhibitory effects are reversed by the addition of excess 2-oxoglutarate and iron(II), suggesting the involvement of baicalein’s binding at the enzyme active sites, which has also been corroborated by spectroscopic binding assays between baicalein and enzyme. In addition, baicalein suppresses ubiquitination of HIF-1 , which works in concert with the inhibition of the HIF-specific hydroxylases to increase the HIF-1 content, leading to induction of HIF-1-mediated reporter gene activity and target gene transcription in tissue culture cells, whereas it induces HIF-independent activation of other genes. Furthermore, in vivo organ models based on the chick chorioallantoic membrane assay demonstrate that baicalein promotes new blood vessel formation. Together, our results indicate that baicalein possesses a proangiogenic potential and thus might have the therapeutic utility in the treatment of ischemic diseases. The hypoxic response in animals involves coordinated regulation of the expression of numerous genes that participate in development, physiology, and pathogenesis of cancer and ischemic disease (Pugh and Ratcliffe, 2003; Semenza, 2004). At the transcription level, hypoxia-inducible factor (HIF)-1, composed of an / heterodimer, mediates cellular adaptive processes to low oxygen availability as a pivotal transcription factor. Although both subunits are constitutively expressed, HIF-1 but not HIF-1 is rapidly degraded during normoxia through the ubiquitin-proteasome system. HIF-1 proteolysis is controlled by hydroxylation of specific proline residues in HIF-1 , which renders it able to recognize the von HippelLindau tumor suppressor protein (VHL) in a protein complex of VHL-Elongin B-Elongin C (VBC) (Epstein et al., 2001; Ivan et al., 2001). Prolyl hydroxylation is catalyzed in an oxygen-sensitive manner by HIF-1 -specific prolyl-4 hydroxylases (PHDs), which require oxygen and 2-oxoglutarate (2-OG) as cosubstrates, and iron(II) and ascorbic acid as cofactors (Bruick and McKnight, 2001). Among PHDs, PHD2 is known to serve as the critical oxygen sensor setting the low steady-state levels of HIF-1 in normoxia (Berra et al., 2003). HIF-1 also possesses a highly conserved site of asparaginyl hydroxylation within the C-terminal transactivation domain This work was supported by the InterFrontier Project from the Functional Proteomics Research Center of the 21st Century Frontier Research Program funded by the Korean Ministry of Science and Technology and from the Intelligent Microsystem Center of the 21st Century Frontier R&D Program sponsored by the Korean Ministry of Commerce, Industry and Energy, a Korean Institute of Science and Technology grant, and a grant (2004-01969) from the Neurobiology Research Program of the Ministry of Science and Technology (to H.P.). H.C. and H.-Y.L. contributed equally to this work. Article, publication date, and citation information can be found at http://molpharm.aspetjournals.org. doi:10.1124/mol.107.040162. □S The online version of this article (available at http://molpharm. aspetjournals.org) contains supplemental material. ABBREVIATIONS: HIF, hypoxia-inducible factor; VHL, von Hippel-Lindau tumor suppressor protein; VBC, von Hippel-Lindau protein-Elongin B-Elongin C; PHD, hypoxia-inducible factor-1 -specific prolyl-4 hydroxylase; 2-OG, 2-oxoglutarate; FIH-1, factor-inhibiting hypoxia-inducible factor-1; HRE, hypoxia-response element; VEGF, vascular endothelial growth factor; FP, fluorescence polarization; GST, glutathione transferase; DMSO, dimethyl sulfoxide; B-P564, biotin-DLDLEALAPYIPADDDFQLR; MALDI, matrix-assisted laser desorption ionization; TOF, time of flight; RT, reverse transcriptase; PCR, polymerase chain reaction; CAM, chorioallantoic membrane; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium; MG132, N-benzoyloxycarbonyl (Z)-Leu-Leu-leucinal; CQ, clioquinol; shRNA, short hairpin RNA; bFGF, basic fibroblast growth factor; B, baicalein; H, hypoxia. 0026-895X/08/7401-70–81$20.00 MOLECULAR PHARMACOLOGY Vol. 74, No. 1 Copyright © 2008 The American Society for Pharmacology and Experimental Therapeutics 40162/3355785 Mol Pharmacol 74:70–81, 2008 Printed in U.S.A. 70 http://molpharm.aspetjournals.org/content/suppl/2008/04/21/mol.107.040162.DC1 Supplemental material to this article can be found at: at A PE T Jornals on A uust 7, 2017 m oharm .aspeurnals.org D ow nladed from catalyzed by the additional iron(II)and 2-OG-dependent enzyme factor-inhibiting HIF-1 (FIH-1) (Lando et al., 2002). This modification prevents recruitment of the transcriptional coactivator cAMP response element-binding protein-binding protein/p300. Thus, the presence of sufficient oxygen inactivates the HIF system by a dual process involving proteolytic destruction and inactivation of transcriptional activity. In contrast, the reduction in the activities of HIF-specific hydroxylases under hypoxia leads to stabilization of HIF-1 , which forms a dimeric complex with HIF-1 . The complex then translocates to the nucleus where it binds to the hypoxia-response element (HRE) in the 5 -flanking regions of the target genes, including vascular endothelial growth factor (VEGF) and other cytoprotective proteins, thereby promoting hypoxic tolerance (Pugh and Ratcliffe, 2003; Semenza, 2004). Efforts to target key components of signal transduction pathways have resulted in therapeutic successes, often through inhibition of upstream signal mediators. In signaling of tissue oxygen homeostasis, the master mediator HIF-1 regulates antiangiogenic and proangiogenic factors. Thus, considerable interest has been shown in the inhibition of angiogenesis for the treatment of cancer and conversely in the activation of angiogenesis for the treatment of ischemic disorders (Carmeliet, 2000; Maxwell and Ratcliffe, 2002; Brahimi-Horn and Pouysségur, 2007). Because of the primary role of the HIF-specific hydroxylases in regulation of HIF-1 protein levels, inhibition of the hydroxylases has been predicted to produce activation of downstream signals in the HIF system with enhanced transcription of the angiogenic factor VEGF, which might be beneficial in therapeutic angiogenesis. We previously developed a convenient fluorescence polarization (FP)-based assay for quantitative measurements of prolyl hydroxylase activities (Cho et al., 2005). The present study uses this FP-based assay to screen a collection of 1040 biologically active compounds. Among the hit compounds as potential inhibitors for PHD2, we chose baicalein, which is a major component of the dried root of Scutellaria baicalensis, which is widely used in traditional Chinese medical applications. It has been found to exhibit antibacterial, free radical scavenging, antioxidant, anticancer, anti-inflammatory, and lipoxygenase-inhibitory activities (Ding et al., 1999; Gao et al., 1999; Shen et al., 2003; Zhang et al., 2003a), in addition to antiangiogenic activity shown in vivo and in vitro (Liu et al., 2003). Despite various interesting reports on the beneficial effects of baicalein, its proangiogenic potential has not yet been demonstrated. Accordingly, we explore the effects of baicalein on the HIF pathway in molecular, cellular, and organ levels, which unravel its unrecognized biological activity. Materials and Methods Cells, cDNAs, Proteins, and Reagents. Human HepG2 hepatoma cells, human SH-SY5Y neuroblastoma cells, and mouse 3T3-L1 fibroblast cells were obtained from American Type Culture Collection (Manassas, VA) and maintained as recommended. The cells were incubated in a hypoxic incubator (Thermo Fisher Scientific, Waltham, MA) at 37°C to induce hypoxia. HIF-1 (GenBank accession number U22431), VHL (GenBank accession number NM000551), Elongin B (GenBank accession number NM007108), Elongin C (GenBank accession number NM005648), PHD2 (GenBank accession number AJ310543), and FIH-1 (GenBank accession number AF395830) human cDNAs were used as expression vectors. The p(HRE)4-luc reporter plasmid (Hur et al., 2001) was used for reporter assays. Truncated GST-PHD2 and His-FIH-1 proteins for enzymatic reactions were expressed and purified as described previously (Cho et al., 2005, 2007). Baicalein, 2-OG, ascorbic acid, and iron(II) sulfate heptahydrate were purchased from SigmaAldrich (St. Louis, MO), cinnamyl-3,4-dihydroxy-cyanocinnamate was from BIOMOL Research Laboratories (Plymouth Meeting, MA), and anti-human HIF-1 was from BD Biosciences Pharmingen (San Diego, CA). Culture media were obtained from Invitrogen (Carlsbad, CA), and fetal bovine serum was from Lonza Verviers SPRL (Verviers, Belgium). All other chemicals were of the highest grade of purity commercially available. Screening for PHD2 Inhibitors. NINDS Custom Collection II (MicroSource Discovery Systems, Gaylordsville, CT) was used for screening potential inhibitors against PHD2 activities. An FP-based assay was performed using proteins and peptides prepared as described previously (Cho et al., 2005). A peptide substrate F-P564 (fluorescein isothiocyanate-aminocaproic acid-DLDLEALAPYIPADDDFQLR) at the final concentration of 1 M was incubated with 0.2 g/ l recombinant GST-PHD2 in NETN buffer (20 mM Tris, pH 8.0, 100 mM NaCl, 1 mM EDTA, 0.5% Nonidet P-40, and 1 mM phenylmethylsulfonyl fluoride) containing 200 M ascorbic acid and 20 M 2-OG in the presence of 20 M compounds for 1 h at 25°C. The reaction mixtures containing DMSO alone with or without GSTPHD2 were included as positive or negative controls, respectively. The reactions were terminated by heating the reactants for 1 min at 95°C, followed by dilution to a final peptide concentration of 100 nM in EBC buffer (50 mM Tris, pH 8.0, 120 mM NaCl, and 0.25% Nonidet P-40) in the presence of 250 nM GST-VBC. The mixtures were gently mixed and transferred to 96-well plates, and their FP values were measured by a Wallac 1420 VICTOR (PerkinElmer Life and Analytical Sciences, Boston, MA). After subtracting the FP value of the negative control, the percentage of inhibition was calculated based on the FP value of the positive control. Mass Spectrometric Analyses of PHD2 and FIH-1 Activities. B-P564 (biotin-DLDLEALAPYIPADDDFQLR) synthesized by Anygen (KwangJu, Korea) was used to analyze prolyl hydroxylation of HIF-1 by mass spectrometry. The B-P564 peptide at 3 M was mixed with 0.20 g/ l GST-PHD2 in NETN buffer containing 400 M ascorbic acid and 100 M 2-OG. For asparaginyl hydroxylation, 4 M F-HIF-1 -(788-822; fluorescein isothiocyanate-aminocaproic acid-DESGLPQLTSYDCEVNAPIQGSRNLLQGEELLRAL) was mixed with 0.55 g/ l His-FIH-1 in hydroxylation buffer (20 mM Tris-HCl, pH 7.5, 5 mM KCl, and 1.5 mM MgCl2) containing 400 M ascorbic acid and 100 M 2-OG. After incubation for 2 h at 25°C, the reactants were desalted with ZipTipC18 (Millipore Corporation, Billerica, MA), and eluted from the tip by addition of -cyano-4-hydroxycinnamic acid in 50% acetonitrile/50% water containing 0.1% trifluoroacetic acid. The purified peptide solutions obtained after extensive washing with 0.1% trifluoroacetic acid in water were transferred to a MALDI sample plate, and MALDI-TOF measurements were performed with a Voyager analyzer (Applied Biosystems, Foster City, CA). UV-Visible Spectroscopic Measurements for Iron(II)-Baicalein Binding. Oxygen was removed from ultrapure water by sonication for 30 min, which was used immediately for solution preparations to minimize the oxidation of baicalein. Baicalein at 24 M was dissolved in 10 mM phosphate buffer, pH 7.4, and mixed with varying concentrations of iron(II) sulfate, followed by recording of absorption spectra from 230 to 500 nm using a Libra S22 UV-visible spectrometer (Biochrom, Berlin, Germany). Immunoblotting. HepG2, SH-SY5Y, and 3T3L1 cells were grown to 80% confluence on 60-mm tissue culture plates, and whole cell extracts were prepared as described previously (Hur et al., 2001). For immunoblotting, whole cell lysates were resuspended in SDS sample buffer, boiled for 5 min, and run on SDS-polyacrylamide gel electrophoresis gels, followed by transfer of the proteins to nitrocellulose membranes by semidry transfer (Trans-Blot SD; Bio-Rad, Hercules, CA). Proteins were then reacted with anti-human HIF-1 antibody and/or with anti-Hsp70 or anti-14-3-3 antibody, and visualized by Baicalein as an HIF and Angiogenesis Inducer 71 at A PE T Jornals on A uust 7, 2017 m oharm .aspeurnals.org D ow nladed from enhanced chemiluminescence according to the manufacturer’s instructions (Pierce Chemical, Rockford, IL), with anti-mouse Ig conjugated with horseradish peroxidase as a secondary antibody. Weak signals from protein bands on Western blots were visualized with a luminescence image analyzer (LAS-3000; Fuji, Tokyo, Japan). Reverse Transcriptase-PCR. HepG2 and SH-SY5Y cells were treated with baicalein or clioquinol (CQ) at indicated concentrations in normoxic conditions for 16 h. One-microgram aliquots of isolated total RNA were used for reverse transcription. To amplify and visualize the cDNAs of VEGF and CA9 mRNA or 18S RNA, two sets of primers were used: for VEGF (AF022375), forward primer (ccatgaactttctgctgtctt) and reverse primer (atcgcatcaggggcacacag); for CA9, forward primer (ctgtcactgctgcttctgat) and reverse primer (tcctctccaggtagatcctc); for 18S rRNA (X03205), forward primer (accgcagctaggaataatggaata) and reverse primer (ctttcgctctggtccgtctt). After an initial melt at 95°C for 10 min, 27 cycles of amplification (95°C for 45 s, 56°C for 45 s, and 72°C for 60 s) were performed with 1 M concentrations of each primer. For PCRs, a final 5-min extension at 72°C was performed, and the amplified products were analyzed on 1% ethidium bromide-stained agarose gels. The expression levels were measured with a luminescence image analyzer (LAS-3000; Fuji). Quantitative Real-Time RT-PCR. Quantitative real-time RTPCR was performed on the iQTM SYBR Green Supermix using MyiQ single color real-time PCR detection system (Bio-Rad). Oligonucleotide primers were designed as reported previously (Zhang et al., 2003b; Murakami et al., 2007) for VEGF, CA9, and BINP3 mRNA or 18S rRNA: for human VEGF, forward primer (aaccatgaactttctgctgtcttg) and reverse primer (ttcaccacttcgtgatgattctg); for mouse VEGF(120), forward primer (gccagcacatagagagaatgagc) and reverse primer (cggcttgtcacatttttctgg); for human CA9, forward primer (cagttgctgtctctcttgga) and reverse primer (tgaagtcagagggcaggagtg); and for mouse BINP3, forward primer (ttaaagggtgcgtgcgggttatct) and reverse primer (aaggcgagaatcctcatcctgcaa). All PCRs were performed in triplicate, and the expression levels normalized to that of 18s rRNA are presented as the averages of at least three experiments. Transient Transfection and Luciferase Assay. Transfections were carried out using the Lipofectamine Plus reagent according to the manufacturer’s instructions (Invitrogen). For luciferase assays, 5 10 HepG2 or 3T3-L1 cells in 24-well plates were transfected with 200 ng of hypoxia reporter plasmid p(HRE)4-luc containing four copies of HRE and 50 ng of plasmid pCHO110 for -galactosidase (Promega, Madison, WI). Cells were treated with baicalein and clioquinol at indicated concentrations or hypoxia 16 h before harvest. Cell extracts were prepared 48 h after transfection and analyzed with a luminometer (Turner Designs, Sunnyvale, CA) using the Luciferase Assay System (Promega). Measured luciferase activities were normalized for total protein concentrations, as determined by Bradford assay (Bio-Rad). Chorioallantoic Membrane Assay. Fertilized chick eggs were maintained in a humidified incubator at 37°C with 55% humidity. After incubation for 10 days, a hypodermic needle was used to remove 2 ml of egg albumin, and the chorioallantoic membrane (CAM) and yolk were allowed to drop away from the shell membrane. On day 10.5, the shell (1 cm) was punched out and the shell membrane was peeled away. For angiogenesis testing, sterile circular filter paper (0.5 cm in diameter; Whatman, Maidstone, UK) loaded with 10 l of sample was air-dried and applied to the CAM surface on the Y-shaped vascular branch point. Ten percent fat emulsion (1–2 ml) was injected into the chorioallantois 3 days later, and it was observed under a microscope. When the CAM showed an avascular zone of 3 mm in diameter, the response was scored as positive. Compounds 0 150 300 450 600 750 90