Received for publication, March 13, 2007, and in revised form, September 24, 2007 Published, JBC Papers in Press, October 1, 2007, DOI 10.1074/jbc.M702175200
Md. Ruhul Abid‡1, Katherine C. Spokes‡, Shou-Ching Shih§, and William C. Aird‡
From the ‡Division of Molecular and Vascular Medicine, Department of Medicine and §Department of Pathology, Beth IsraelDeaconess Medical Center and Harvard Medical School, Boston, Massachusetts 02215
Vascular endothelial growth factor (VEGF) and reactive oxy-gen species (ROS) play critical roles in vascular physiology andpathophysiology. We have demonstrated previously thatNADPH oxidase-derived ROS are required for VEGF-mediatedmigration and proliferation of endothelial cells. The goal of thisstudy was to determine the extent to which VEGF signaling iscoupled to NADPH oxidase activity. Human umbilical veinendothelial cells and/or humancoronary artery endothelial cellswere transfected with short interfering RNA against the p47phoxsubunit of NADPH oxidase, treated in the absence or presenceof VEGF, and assayed for signaling, gene expression, and func-tion. We show that NADPH oxidase activity is required forVEGF activation of phosphoinositide 3-kinase-Akt-forkhead,and p38 MAPK, but not ERK1/2 or JNK. The permissive role ofNADPH oxidase on phosphoinositide 3-kinase-Akt-forkheadsignaling is mediated at post-VEGF receptor levels and involvesthe nonreceptor tyrosine kinase Src. DNAmicroarrays revealedthe existence of two distinct classes of VEGF-responsive genes,one that is ROS-dependent and another that is independent ofROS levels. VEGF-induced, thrombomodulin-dependent acti-vation of protein C was dependent on NADPH oxidase activity,whereas VEGF-induced decay-accelerating factor-mediatedprotection of endothelial cells against complement-mediatedlysis was not. Taken together, these findings suggest thatNADPH oxidase-derived ROS selectively modulate some butnot all the effects of VEGF on endothelial cell phenotypes.
Vascular endothelial growth factor (VEGF)2 plays a criticalrole in endothelial survival, migration, and proliferation. VEGF
has been implicated in wound repair, angiogenesis of ischemictissue, tumor growth, microvascular permeability, vascularprotection, and hemostasis (1–8). VEGF binds to two recep-tors, VEGF receptor (VEGFR)-1 (Flt-1) and VEGFR2 (Flk-1/KDR). VEGF has been shown to activate several downstreamsignaling pathways, including protein kinase C, phosphoinosit-ide 3-kinase (PI3K), Akt, extracellular signal-regulated kinase-1and -2 (ERK1/2), mitogen-activated protein kinase (MAPK)p38, and phospholipase C (PLC)-� (9–13). VEGF may alterendothelial cell phenotypes through transcriptional and post-transcriptional mechanisms. The effect of VEGF on gene tran-scription may be explained, at least in part, by its ability toactivate nuclear factor (NF)-�B, early growth response factor-1,nuclear factor of activated T cells-1, Ets-1, and signal transduc-ers and activators of transcription-3/5 (14–18). VEGF inducesphosphorylation of forkhead transcription factors (e.g. FKHR),resulting in their nuclear exclusion and transcriptional inacti-vation (19).Reactive oxygen species (ROS) have long been implicated in
the pathogenesis of cardiovascular diseases, including athero-sclerosis, hypertension, and diabetes (reviewed in Ref. 20). Inaddition, there is a growing appreciation for the role of ROS inphysiological signaling inmany cell types, including endothelialcells. NADPH oxidase is the primary source of superoxide inendothelial cells (21–24). NADPH oxidase was originally iden-tified and characterized in phagocytes, where it contributes tohost defense. The NADPH oxidase complex consists of twomembrane-bound components, gp91phox (also known asNox2)and p22phox, and several cytosolic regulatory subunits, includ-ing p40phox, p47phox, p67phox, and the small GTPase Rac (Rac1or Rac2). Upon enzyme activation, the cytoplasmic units trans-locate to the cell membrane where they are assembled withgp91phox/p22phox. The resulting multisubunit complex trans-fers electrons fromNAD(P)H to molecular oxygen. Each of thecomponents of the neutrophil NADPH complex has been iden-tified in endothelial cells (25–27). In addition, endothelial cellsexpress a homologue of gp91phox/Nox2, termed Nox4 (28, 29).The endothelial NADPH oxidase differs from its leukocytecounterpart in several importantways. First, it is pre-assembledand displays constitutive low level activity. Second, agonist-me-diated stimulation of NADPH oxidase results in slower and lesspotent activation of the enzyme (22). Third, the enzyme com-
* This work was supported by American Heart Association Grant SDG0453284N and National Institutes of Health Grants 5R01HL077348 and5R01HL082927. The costs of publication of this article were defrayed inpart by the payment of page charges. This article must therefore be herebymarked “advertisement” in accordance with 18 U.S.C. Section 1734 solely toindicate this fact.
□S The on-line version of this article (available at http://www.jbc.org) containssupplemental Figs. 1–5.
1 To whom correspondence should be addressed: Molecular and VascularMedicine, Dept. of Medicine, Beth Israel Deaconess Medical Center, Har-vard Medical School, E/RW-663, 330 Brookline Ave., Boston, MA 02215. Tel.:617-667-1025; Fax: 617-667-2913; E-mail: [email protected].
plex is localized in the perinuclear region of endothelial cellsand produces an intracellular influx of ROS (30, 31).Many different agonists have been shown to induce endothe-
lial NADPH oxidase activity, including hemodynamic forces(32–34), VEGF (21, 22, 24), angiopoietin-1 (Ang1) (21, 35, 36),tumor necrosis factor-� (37), thrombin (38), angiotensin II (37,39), endothelin-1 (40), transforming growth factor-� (41), oxi-dized low density lipoprotein (27), and high potassium (42).Activation ofNADPHoxidase ismediated by post-translationalmodification and/or increased transcription of the regulatorysubunits.We previously reported that VEGF induces NADPH oxidase
activity and that NADPHoxidase activity is required for VEGF-mediated migration and proliferation of endothelial cells (21,22). These results were confirmed and expanded upon by othergroups. For example, PI3K and Rac1 were shown to be requiredfor the VEGF-dependent oxidative burst in porcine aorticendothelial cells expressing VEGFR2/KDR (23). Ushio-Fukai etal. (24) demonstrated an important role for gp91phox in medi-ating the effect of VEGF on endothelial cell migration and pro-liferation in vitro and in promoting angiogenesis in a mousesponge implant model. Similar findings were observed in ahindlimb ischemia model (43). Moreover, we showed thatNADPH oxidase-derived ROS are required for VEGF stimula-tion of manganese superoxide dismutase, activation of NF-�B,and inactivation of FKHR in endothelial cells (44).The mechanisms by which NADPH oxidase-derived ROS
influence VEGF are poorly understood. ROS may result in oxi-dation of cysteine residues on receptor and nonreceptor pro-tein kinases and phosphatases. Indeed, previous studies havedemonstrated a role for ROS in VEGFR2 autophosphoryla-tion (23, 24). Here, we show that NADPH oxidase activity isrequired for some, but not all, downstream effects of VEGFin endothelial cells. These data argue against a global sensi-tivity of VEGF signal transduction pathways to the redoxstate of the cell, and suggest that therapeutic modulation ofROS will selectively influence the effect of VEGF on endo-thelial cell phenotypes.
EXPERIMENTAL PROCEDURES
Cell Culture and Reagents—Human coronary artery endo-thelial cells (HCAEC) and human umbilical vein endothelialcells (HUVEC) were obtained from Clonetics and grown inEndothelial Growth Medium-2-MV (EGM-2-MV) BulletKit(Clonetics, San Diego) at 37 °C and 5% CO2. Endothelial cellsfrom passage 3 to 6 were used for all experiments. Cells wereserum-starved in 0.5% fetal bovine serum for 12–16 h prior totreatment with 50 ng/ml human VEGF-A165 (PeproTech Inc,Rocky Hill, NJ). Where indicated, cells were preincubated for30 min with 100 nmol/liter L-NAME (Calbiochem), 50 �mol/liter LY294002 (Calbiochem), 50 �mol/liter PD98059 (Calbio-chem), or 10 �mol/liter SB203580 (Calbiochem). Phenazinemethosulfate and S2366 were from Calbiochem.Inhibition of NADPH Oxidase Activity by Targeting Endoge-
nous p47phox—HCAEC and HUVEC were grown to 70–80%confluence in 6-cm plates and transfected with 100 nM p47phoxantisense oligonucleotide (5�-TTTGTCTGGTTGTGTGT-GGG-3�; Sequitur, Natick, MA), scrambled antisense (Scram-
AS), siRNA against p47phox (5�-GGUCAUUCACAAGCUC-CUGtt-3�, Ambion, Austin, TX), or scrambled siRNA (Scram-si) in Opti-MEM containing 10 �g/ml Lipofectin (Invitrogen)for 4 h. The cells were then incubated in EGM-2-MV mediumfor 24 h and serum-starved in endothelial basal medium(EBM-2; Clonetics) containing 0.5% serum for 12–16 h prior toVEGF treatment. HP-validated siRNA against Src (catalognumber SI02223928) and FAK (catalog number SI00287791)kinases were from Qiagen (Valencia, CA). ON-TARGETplussi-RNA against PLC�-1 was from Dharmacon (Lafayette, CO).Assay for NADPH Oxidase Activity in HCAEC and HUVEC—
HCAEC andHUVECwere washed with ice-cold PBS, collectedby a cell scraper, and Dounce-homogenized in buffer contain-ing 20 mM KH2PO4 (pH 7.0), 1� protease mixture inhibitor(Sigma), 1 mM EGTA, 10 �g/ml aprotinin (Calbiochem), 0.5�g/ml leupeptin (Sigma), 0.7 �g/ml pepstatin (Sigma), and 0.5mM phenylmethylsulfonyl fluoride (Calbiochem). NADPH oxi-dase activity of the cell lysate was measured using a modifiedassay (45). Briefly, photon emission from the chromogenic sub-strate lucigenin as a function of acceptance of electron/O2
� gen-erated by the NADPH oxidase complex was measured every15 s for 20 min in a Berthold luminometer. NADPH oxidaseassay buffer containing 250mMHEPES (pH 7.4), 120mMNaCl,5.9 mM KCl, 1.2 mM MgSO4(7H2O), 1.75 mM CaCl2(2H2O), 11mM glucose, 0.5 mM EDTA, 100 �MNADH and 5 �M lucigeninwas used. The data were converted to relative light units/min/mg of protein, using a standard curve generated with xan-thine/xanthine oxidase. Lucigenin activity (light units/min/mgof protein) of control cells (Scram-AS-transfected) was arbi-trarily set at 100%. Total intracellular levels of ROS were deter-mined by FACS analyses of the oxidative conversion of cell-permeable 2�,7�-dichlorofluorescein diacetate (DCFH-DA;Molecular Probes Inc., Eugene, OR) to fluorescent dichlo-rofluorescein as described previously (22).Oligonucleotide Microarray Analysis of Gene Expression—
The transcriptional profile of control or VEGF-treatedHCAECwith siRNA against p47phox was characterized by oligonucleo-tide microarray analysis using the human U133A AffymetrixGeneChip, according to previously described protocols fortotal RNA extraction and purification, cDNA synthesis, in vitrotranscription reaction for production of biotin-labeled cRNA,hybridization of cRNA with U133A Affymetrix gene chips,scanning of image output files, analysis of gene expression data,and hierarchical and functional clustering algorithms (46). Thescanned array images were analyzed by dChip (47). In thedChip analysis a smoothing spline normalization method wasapplied prior to obtaining model-based gene expression indi-ces, also known as signal values. There were no outliers identi-fied by dChip so all samples were carried out for subsequentanalysis.When comparing two samples (groups) to identify thegenes enriched in a given phenotype, we used the lower confi-dence bound (LCB) of the fold change between the experimentand the base line. If the 90% LCB of the fold change between theexperiment and the base line was above 1.2, the correspondinggene was considered to be differentially expressed. An LCB of�1.2 typically corresponds with an “actual” fold change of atleast 3 in gene expression. GOTree machine was used to iden-tify gene ontology categories for the input gene set (48).
Selective Sensitivity of VEGF Signaling Pathways to Redox
35374 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282 • NUMBER 48 • NOVEMBER 30, 2007
Quantitative Real Time PCR—Real time PCRwas carried outas described previously (49). Briefly, total RNA was preparedusing the RNeasy RNA extraction kit with DNase-I treatmentfollowing the manufacturer’s protocol (Qiagen, Valencia, CA).To generate cDNA, total RNA (100 ng) from each of triplicatesamples was converted into cDNA using random primers andSuperscriptIII reverse transcriptase (Invitrogen). All cDNAsamples were aliquoted and stored at �80 °C. Primers weredesigned using the Primer Express oligo design software(Applied BioSystems, Foster City, CA) and synthesized byIntegrated DNA Technologies (Coralville, IA). All primersets were subjected to rigorous data base searches to identifypotential conflicting transcript matches to pseudogenes orhomologous domains within related genes. Amplicons gener-ated from the primer set were analyzed for melting point tem-peratures using the first derivative primer melting curve soft-ware supplied by Applied BioSystems. The SYBR Green I assayand the ABI Prism 7700 sequence detection system were usedfor detecting real time PCR products from the reverse-tran-scribed cDNA samples, as described previously (50). 18 SrRNA, which exhibits a constant expression level across all theHCAEC samples, was used as the normalizer. PCRs for eachsample were performed in duplicate, and copy numbers weremeasured as described previously (50). The level of target geneexpressionwas normalized against the 18 S rRNA expression ineach sample, and the data were presented as mRNA copies per106 18 S rRNA copies.Adenoviruses—HCAEC were transduced with replication-
deficient adenoviruses encoding the cDNAs of �-galactosidase(Adv), triple mutant (TM)-FKHR (49), constitutively active(CA) Akt, or dominant-negative (DN) Akt (19) as describedpreviously (49). The triple mutant version of FKHR containsT24A, S256A, and S319A and is resistant to agonist-inducedphosphorylation.Western and Northern Blot Analyses—HCAEC were har-
vested for total protein, and Western blots were carried out asdescribed previously (19). Anti-Akt, anti-FKHR, anti-p38, anti-VEGFR2, anti-PLC�-1, and anti-ERK1/2 antibodies and phos-phospecific antibodies against Ser-256 FKHR, Ser-473 Akt,ERK1/2, JNK, p38MAPK, Src, FAK, PLC�-1, Tyr-951VEGFR2,and Tyr-1175 VEGFR2 were purchased from Cell Signaling(Beverly, MA). Anti-�-actin antibodies were from Sigma.Western blots were performed in triplicates using cell extractsprepared from three independent experiments. RNA extrac-tion andNorthern blot assayswere performed as described pre-viously (22). Immunoprecipitation was carried out as described(19). The protein kinase assaywas carried out using theHTScanSrc kinase assay kit with a slight modification of the manufac-turer’s protocol (Cell Signaling). Briefly, HCAEC were lysed inbuffer (60mMHEPES (pH 7.5), 1%TritonX-100, 150mMNaCl,1 mM EGTA) plus protease and protein-tyrosine phosphataseinhibitors. Reaction mixtures, containing 12.5 �l of cell lysatesand 37.5 �l of kinase buffer (60 mM HEPES (pH 7.5), 5 mMMgCl2, 5 mM MnCl2, 3 �M Na3VO4, 1.25 mM dithiothreitol, 20�M ATP, 1.5 �M biotinylated peptide substrate), were incu-bated in a 96-well plate at room temperature for 30min. 50�l ofSTOP buffer (50mM EDTA (pH 8)) were added per well to stopthe reaction. 25-�l aliquots from each reaction were added to
the well of a DELFIA streptavidin-coated, 96-well plates(PerkinElmer Life Sciences) containing 75 �l of distilled H2O,and the resulting mixture was incubated at room temperaturefor 60 min. Each well was then washed three times with PBS/T.100 �l of primary mouse monoclonal antibody to phosphoty-rosine (P-Tyr-100; 1:1000 dilution) were added to each well.After 60 min of incubation at room temperature, each well waswashed three times with PBS/T. 100 �l of secondary anti-mouse IgG (1:500 dilution in PBS/T with 1% BSA) were addedto each well and incubated for 30 min at room temperature.Wells were washed with PBS/T five times, and 100 �l ofDELFIA enhancement solution (PerkinElmer Life Sciences)were added for 5 min. Fluorescence emission was detected at awavelength of 615 nm.Immunolocalization Studies—HCAEC were grown in 4-well
chamber slides (Lab-Tek, Christchurch, New Zealand) andtreated with or without VEGF for the times indicated. Subcel-lular localization of FKHR and ERK1/2 was determined usinganti-FKHR and anti-ERK1/2 antibodies, respectively, and aCy3-conjugated secondary antibody as described previously(19). Quantitative analyses were carried out by counting 200cells per time point.Complement-mediated Cell Lysis Assay—HCAEC were
plated in 24-well plates and incubated at 37 °C for 24 h priorto siRNA transfection. 24 h following siRNA transfections,the cells were serum-starved overnight and then incubatedwith or without VEGF for 24 h. 7 �mol/liter calcein ace-toxymethyl ester (Molecular Probes, Leiden, Netherlands)were added to each well. 30 min later, cells were washed withserum starvation medium containing 1% BSA and then incu-bated for 30 min with 250 �l of monoclonal anti-endoglin(CD105) antibody (IgG2a) (Covance, Berkley, CA) to opso-nize the cells. HCAEC were washed with HBSS containing1% BSA and incubated with 250 �l of 5–20% baby rabbitcomplement (Serotec, Oxford, UK) at 37 °C for 30 min. Thesupernatant from each well was transferred to a 96-wellmicrotiter plate. The remaining HCAEC in the 24-well platewere washed with HBSS plus 1% BSA, and the calceinremaining in the cells was released by incubation with 250 �lof HBSS containing 1% BSA and 0.1% Triton X-100. Thelysate was then transferred to another 96-well plate, and thecalcein released by complement and detergent was esti-mated using a fluorescence plate reader (model 680; Bio-Rad). Percent specific lysis in triplicate wells was calculated as((complement-mediated release � spontaneous release)/max-imal release � spontaneous release)) � 100%, where maxi-mal release � complement-mediated release � detergent-mediated release.Assay for Thrombomodulin-dependent Activation of Protein
C—Functional assay for cellular thrombomodulin was carriedout as described previously (51), with slight modifications.Briefly, HCAEC transfected with 75 nM siRNAs were seeded on24-well tissue culture plates. After 24 h, HCAEC were serum-starved in EBM-2 containing 0.5% fetal bovine serum for 18 hand treated with 50 ng/ml VEGF for 14 h. Cells were washedwith PBS and incubated in a reaction mixture containing 2.5mM CaCl2, 0.15 M NaCl, 5 nM thrombin, and 5 nM protein C for3 h. After withdrawing 100 �l at each time point, hirudin was
Selective Sensitivity of VEGF Signaling Pathways to Redox
NOVEMBER 30, 2007 • VOLUME 282 • NUMBER 48 JOURNAL OF BIOLOGICAL CHEMISTRY 35375
added to inhibit thrombin. Equal volume of the chromogenicsubstrate, 400 �M S2366, was added to the supernatant, andabsorbance was measured at 405 nM.Statistical Analyses—All values are presented asmean� S.D.
where appropriate. Statistical significance between two groupswas determined by use of a paired t test, and values of p � 0.05were considered significant.
RESULTS
NADPHOxidase Activity Is Abrogated byDown-regulation ofp47phox Expression in Primary Human Endothelial Cells—Toinhibit NADPH oxidase in endothelial cells, HCAEC orHUVEC were transfected with antisense or siRNA againstp47phox. Antisense (AS-p47phox) and siRNA (si-p47phox)resulted in significant (�90%) reduction in p47phox protein lev-els, compared with scrambled antisense (ScramAS) and siRNA(Scram-si), respectively (supplemental Fig. 1, A and B). Trans-fection of HCAEC and HUVEC with AS-p47phox or si-p47phoxalso resulted in significant inhibition of NADPH oxidase activ-ity as measured by lucigenin assay (Fig. 1A, HCAEC). As a neg-ative control, treatment of endothelial cells with the nitric-ox-ide synthase inhibitor, L-NAME, had no effect on NAPDHoxidase activity (Fig. 1A). To determine the effects of NADPHoxidase inhibition on intracellular ROS production in HCAEC,FACS analyses were performed using DCFH-DA. Transfectionwith si-p47phox significantly reduced intracellular levels of ROS(to 58.6 � 6.5%) (Fig. 1B). Consistent with the findings of thelucigenin assays, ROS levels were unaffected by L-NAME. As apositive control, the superoxide donor, phenazine methosul-fate, resulted in 180% increase in ROS content (Fig. 1B). Thus,AS-p47phox and si-p47phox are each effective in reducingNADPH oxidase activity in endothelial cells.
NADPH Oxidase Activity IsRequired for Late but Not EarlyVEGF-mediated Tyrosine Phospho-rylation of VEGFR2/KDR—VEGFtreatment of HCAEC resulted in atime-dependent increase in tyrosinephosphorylation of VEGFR2/KDRstarting as early as 1 min and peak-ing at 10 min (data not shown).siRNA-mediated knockdown ofp47phox resulted in partial inhibitionofVEGF-mediated phosphorylationof VEGFR2/KDR at 10 min (to 69 �6.6%) but had no effect at 1 and 5min (Fig. 2). Similar results wereobserved with AS-p47phox (supple-mental Fig. 2A). These findings sug-gest that NADPH oxidase activity isnot required for early phosphoryla-tion of VEGFR2/KDR by VEGF butmay play a role in maintaining thereceptor in the phosphorylatedstate.Inhibition of NADPH Oxidase in
Primary Human Endothelial CellsHas a Differential Effect on VEGF-
FIGURE 1. Inhibition of NADPH oxidase activity in HCAEC. A, HCAEC trans-fected with control antisense oligonucleotide (Scram-AS), control siRNA(Scram-si), AS-p47phox, or si-p47phox or pretreated with L-NAME (100 nM) wereassayed for NADPH oxidase activity as measured by a low concentration-based lucigenin (5 �M) assay. Control (Scram-AS) activity was arbitrarily set at100%. B, flow cytometric analyses (FACS) of intracellular ROS production inHCAEC were carried out using DCFH-DA. Dichlorofluorescein fluorescence ofcontrol cells (Scram-si) was arbitrarily set at 100. The superoxide donor, phen-azine methosulfate (PMS) (5 �M), was used as positive control for ROS induc-tion in HCAEC. All experiments were performed in triplicate, and the datashown are means � S.D. *, p � 0.05, relative to control.
FIGURE 2. NADPH oxidase activity is required for late but not early VEGF-mediated phosphorylation ofVEGFR2/KDR. Immunoprecipitation (IP) of HCAEC lysates by anti-VEGFR2/KDR antibody followed by immunoblotswith anti-phosphotyrosine (p-Tyr) antibody. A, HCAEC were transfected with Scram-si or si-p47phox and treated withor without VEGF for 1 and 5 min. The right panel shows quantitative analysis (mean � S.D.) of VEGFR2/KDR phos-phorylation at 5 min based on three independent experiments. B, same as in A except HCAEC were treated withoutor with VEGF for 10 min. Right panel shows quantitative analysis of three independent experiments with mean�S.D.*, p � 0.05, relative to control.
Selective Sensitivity of VEGF Signaling Pathways to Redox
35376 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282 • NUMBER 48 • NOVEMBER 30, 2007
mediated Phosphorylation of Sig-nal Intermediates—Incubation ofHCAECwithVEGFresultedintime-dependent phosphorylation of Aktat serine 473 and threonine 308 res-idues, with maximal levels occur-ring at 10 min (Fig. 3A, Ser-473).VEGF-mediated phosphorylation ofAkt was inhibited by transfectionwith si-p47phox (to 12% of controllevels) (Fig. 3A). Similar results wereobserved with AS-p47phox (supple-mental Fig. 2B). VEGF also inducedphosphorylation of ERK1/2, withpeak levels occurring at 5 min(Fig. 3A). However, VEGF-mediatedactivation of ERK1/2 was unalteredin p47phox-deficient endothelialcells (Fig. 3A and supplemental Fig.2C). Similarly, VEGF-induced phos-phorylation of JNK was unaffectedby p47phox knockdown (Fig. 3B). Incontrast, siRNA against p47phoxreduced VEGF-mediated phospho-rylation of p38 MAPK by �75% atall time points tested (Fig. 3B).Consistent with a role for PI3K-
Akt in mediating VEGF-induciblephosphorylation of the forkheadtranscription factor, FKHR, si-p47phox abrogated the effect ofVEGF on FKHR phosphorylation atSer-256 (Fig. 4). Together, thesefindings suggest that VEGF triggersboth ROS-sensitive (PI3K-Akt-FKHR and p38MAPK) andROS-in-
sensitive (ERK1/2 and JNK) signaling pathways in endothelialcells.Phosphorylated ERK1/2 is localized in the nucleus of endo-
thelial cells (52). In keeping with the phosphorylation studies ofERK1/2, si-p47phox-mediated inhibition of NADPH oxidasehad no effect on nuclear translocation of ERK1/2 in VEGF-treated cells (64% nuclear in VEGF-treated control cells versus68% nuclear in VEGF-treated si-p47phox-transfected cells)(supplemental Fig. 3, A and B). In accordance with the effect ofNADPH oxidase inhibition on VEGF-induced FKHR phospho-rylation, si-p47phox blocked VEGF-mediated nuclear exclusionof FKHR inHCAEC (27%nuclear inVEGF-treated control cellsversus 72% nuclear in VEGF-treated si-p47phox-transfectedcells) (supplemental Fig. 4, A and B).Inhibition of NADPH Oxidase in Primary Human Endothe-
lial Cells Blocks VEGF-mediated Phosphorylation of Src andFAK—The above results indicate that NADPH oxidase activityis required for VEGF-mediated phosphorylation of Akt prior toany observable effect on VEGFR2/KDR phosphorylation (i.e. at1 and 5 min). Based on these findings, we hypothesized thatNADPH oxidase-derived ROS influence VEGF signaling at alevel distal to VEGFR2/KDR and proximal to Akt. The nonre-
FIGURE 3. NADPH oxidase activity is required for VEGF-mediated phosphorylation of Akt, p38 MAPK, but notERK1/2 or JNK in HCAEC. Western blot analyses using lysates of HCAEC treated with VEGF (50 ng/ml) for the timesindicated. A, HCAEC were transfected with either control siRNA (Scram-si) or si-p47phox, serum-starved, and thentreated with or without VEGF for the times indicated. Western blots were carried out using anti-phospho-Akt (p473-Akt) antibody. The membranes were stripped and reprobed with anti-phospho-ERK1/2 antibody. Anti-Akt antibody(Akt) was used as loading control. B, same as in A except Western blot analyses were carried out using anti-phosphoJNK (p-JNK) antibody. The membrane was then stripped and probed with anti-phospho-p38 antibody (p-p38) andanti-p38 antibody (p38). Bar graphs show quantitative analyses of three independent Western blot experiments(mean � S.D.). *, p � 0.05, relative to time-matched control.
FIGURE 4. VEGF-induced phosphorylation of the forkhead transcription fac-tor, FKHR, is dependent on NADPH oxidase activity. Upper panel, HCAEC weretransfected with either control siRNA (Scram-si) or si-p47phox, serum-starved, andthen treated with or without VEGF for the times indicated. Western blot analysiswas carried out using anti-phospho-Ser-256 FKHR antibody (p-FKHR). Lowerpanel, the membrane was then stripped and probed with anti-FKHR antibody(FKHR). Bar graphs show quantitative analyses of three independent Western blotexperiments (mean � S.D.). *, p � 0.05, relative to time-matched control.
Selective Sensitivity of VEGF Signaling Pathways to Redox
NOVEMBER 30, 2007 • VOLUME 282 • NUMBER 48 JOURNAL OF BIOLOGICAL CHEMISTRY 35377
ceptor tyrosine kinases Src and FAKhave been shown to play a role in theactivation of PI3K-Akt (3, 53–56).VEGF treatment of HCAECresulted in time-dependent phos-phorylation of Src at Tyr-416 andFAK at Tyr-925/861 (Fig. 5A).These effects were blocked bysi-p47phox (Fig. 5A). Phosphoryla-tion of Src has been correlated withthe activity of this nonreceptorkinase (57–59). Consistent with thephosphorylation data, VEGF-medi-ated induction of Src activity wassignificantly reduced in cells trans-fected with si-p47phox (Fig. 5C).PLC�-1 and Shb, through their
interactionwith the phosphorylatedTyr-1175 residue on VEGFR2, havealso been shown to play a role in theactivation of the PI3K-Akt signalingpathway in endothelial cells (60).VEGF treatment of HCAECresulted in time-dependent phos-phorylation of PLC�-1 (Fig. 5B).These effects were not affected bysi-p47phox (Fig. 5B), suggestingthat, unlike Src and FAK, VEGF-mediated phosphorylation ofPLC�-1 is not dependent onNADPH oxidase activity inHCAEC. Together, these resultssuggest that NADPH oxidase-de-rived ROS are critical for mediatingthe effects of VEGF on Src/FAK sig-naling intermediates.To determine whether the ROS-
sensitive Src and/or FAK play a rolein VEGF-mediated activation ofAkt, siRNAs were employed toinhibit expression of one or theother kinase in HCAEC. In theseexperiments, siRNA against Srcresulted in �70% reduction of Srcprotein, whereas FAK siRNAresulted in �90% inhibition of FAK(Fig. 6A). Knockdown of Src signifi-cantly attenuated VEGF-mediatedphosphorylation of Akt (by 58 �5.4%) but not ERK1/2 (Fig. 6A).Inhibition of FAK had no effect onVEGF-mediated phosphorylation ofAkt or ERK1/2 (Fig. 6A). Similarly,siRNA knockdown of PLC�-1(�75%) failed to inhibit Akt phos-phorylation (Fig. 6B). These find-ings suggest that NADPH oxidasemay exert its permissive effect on
FIGURE 5. VEGF-induced phosphorylation of Src and FAK, but not PLC�-1, is sensitive to NADPH oxidaseactivity in HCAEC. A, Scram-si or si-p47phox-transfected HCAEC were serum-starved overnight and then treatedwith VEGF (50 ng/ml) for the times indicated. Western blot analysis was carried out using anti-phospho-Tyr-416 Srcantibody (p-Src). The membrane was stripped and reprobed using anti-phospho-Tyr-925/861 FAK (p-FAK) antibody,followed by phospho-specific anti-ERK1/2 antibody (positive control for VEGF response, negative control for NADPHoxidase inhibition), and total anti-ERK1/2 antibody (as control for loading). Lower panel shows quantitative analysisof three independent experiments (mean�S.D. of fold changes versus control). *, p�0.05, relative to time-matchedcontrol. B, same as in A except Western blots were carried out using anti-phospho-PLC�-1 antibody, followed bystripping and reprobing of the membrane using anti-PLC�-1 antibody as loading control. C, VEGF-induced Srckinase activity was measured in lysates from Scram-si or si-p47phox-transfected HCAEC as described under “Experi-mental Procedures.” *, p � 0.05, VEGF-treated versus untreated in Scram-si transfected endothelial cells.
FIGURE 6. VEGF-induced phosphorylation of Akt is dependent on Src but not FAK or PLC�-1. A, Scram-si, si-Src,or si-FAK-transfected HCAEC were serum-starved overnight and then treated with VEGF (50 ng/ml) for 10 min.Western blot analysis was carried out using anti-phospho p473-Akt antibody. The membrane was stripped andsubsequently reprobed using anti-phospho-ERK1/2 (p-ERK1/2), anti-FAK (FAK), and anti-Src (Src) antibodies. Totalanti-Akt (Akt) antibody was used as a loading control. Bar graphs show quantitative analyses of three independentWestern blot experiments (mean � S.D.). *, p � 0.05, relative to control. B, Scram-si or si-PLC�-1-transfected HCAECwere serum-starved overnight and then treated with VEGF (50 ng/ml) for 0, 5, and 15 min. Western blot analysis wascarried out using anti-phospho p473-Akt antibody. The membrane was stripped and subsequently reprobed usinganti-PLC�-1 antibody. Total anti-Akt (Akt) antibody was used as a loading control. Bar graphs show quantitativeanalyses of three independent Western blot experiments (mean � S.D.).
Selective Sensitivity of VEGF Signaling Pathways to Redox
35378 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282 • NUMBER 48 • NOVEMBER 30, 2007
VEGF-PI3K-Akt signaling, at least partially, at the level of thenonreceptor tyrosine kinase Src.Dependence of Src Kinase on NADPH Oxidase Activity
Occurs at a Post-receptor Level—Phosphorylation of the tyro-sine residue, Tyr-951, in the kinase insert domain of VEGFR2has been shown to activate Src through its interaction with theT cell-specific adapter (TSAd/VRAP) molecule (61, 62). How-ever in immunoprecipitation assays, VEGF-induced phospho-rylation of Tyr-951 was unaffected by inhibition of NADPHoxidase (Fig. 7A). These results suggest that NADPH oxidase-derived ROS modulate the activity of Src kinase (and thus,PI3K-Akt) independently of VEGFR2 Tyr-951 in endothelialcells.Phosphorylation of the tyrosine residue, Tyr-1175, in the
C-terminal region of VEGFR2 promotes activation of PLC�(63). Activated PLC�, in turn, generates inositol trisphosphateand diacylglycerol by hydrolyzing phosphatidylinositol 4,5-bisphosphate, a common and rate-limiting substrate for bothPLC� and PI3K. In addition, Tyr-1175 provides a binding sitefor Shb, an adapter molecule that has been implicated in theactivation of the PI3K-Akt signaling pathway in porcine aorticendothelial cells (63). Inhibition of NADPH oxidase by siRNA
did not inhibit VEGF-mediated phosphorylation at Tyr-1175(Fig. 7B), arguing against a role for this residue in mediatingredox sensitivity of the PI3K-Akt signaling pathway.Inhibition of NADPH Oxidase in Primary Human Endothe-
lial Cells Has a Differential Effect on VEGF-mediated GeneTranscription—We next wished to determine whether theexistence of NADPH oxidase-dependent and -independentVEGF signaling pathways is associated with differential sensi-tivity of downstream target genes. HCAEC were transfectedwith scrambled siRNA (control) or si-p47phox and then incu-bated in the absence or presence of 50 ng/ml VEGF for 4 h.Total RNA was extracted and processed for DNA microarraystudies. The clusters were analyzed for a pool of VEGF-respon-sive genes that were inhibited by si-p47phox. A total of 1486genes were induced by VEGF in scrambled siRNA-treatedHCAEC (LCB 1.5). Of these, 402 were blocked by si-p47phox (atLCB 1.5; data not shown). The existence of distinct ROS-sensi-tive and -insensitive classes of VEGF-inducible genes was vali-dated by real time-PCR of selected genes (Tables 1 and 2). Asexpected, VEGF treatment of HCAEC also resulted in reducedexpression of certain genes (Table 3). In each case, VEGF-me-diated gene repression was blocked by si-p47phox (Table 3).Selected examples of these three gene classes were verified byNorthern blot experiments and are shown in Fig. 8.VEGF Induction of NADPH Oxidase-sensitive Genes Is
Mediated by ROS-dependent Signaling Pathways—We askedwhether the sensitivity of VEGF-responsive genes to NADPHoxidase activity could be explained by the involvement of one ormore of the ROS-dependent signaling pathways elucidated inthis study. Indeed, VEGF induction of twoNADPHoxidase-de-pendent genes, VCAM-1 and E-selectin, was significantlyblocked by LY294002 (PI3K inhibitor) or SB203580 (p38MAPK inhibitor) but not PD98059 (ERK1/2 inhibitor) (Fig. 9,Aand B). Similarly, VEGF-mediated inhibition of GADD45Awas attenuated by si-p47phox (see Table 3), LY294002, andSB203580 but not PD98059 (Fig. 9C). In contrast, expression of
FIGURE 7. VEGF-induced phosphorylation at the tyrosine residues 951and 1175 of VEGFR2 is not dependent on NADPH oxidase activity inHCAEC. Immunoprecipitation (IP) of HCAEC lysates by anti-VEGFR2/KDR anti-body followed by immunoblots with anti-phosphotyrosine 951 (A) and 1175(B) antibodies are shown. Scram-si- or si-p47phox-transfected HCAEC wereserum-starved overnight and then treated with VEGF (50 ng/ml) for the timesindicated. Immunoprecipitation followed by Western blot analyses were car-ried out as described under “Experimental Procedures.” The membrane wasstripped and reprobed using anti-VEGFR2/KDR antibody (as control forloading).
TABLE 1NADPH oxidase-dependent genes that are induced by VEGF in endothelial cells as determined by RT-PCR analyses (-fold induction)The basal levels of expression for each gene in scrambled siRNA-transfected, unstimulated, serum-starved HCAEC were arbitrarily set at 1 (-fold) per 106 18 S mRNAcopies. Numbers are expressed as -fold induction (or reduction if less than 1) over the basal levels. The abbreviations used are as follows: ACE-1, angiotensin-convertingenzyme 1; ADAMTS9, A disintegrin and metalloproteinase with thrombospondin motifs 9; MGLAP, matrix Gla protein; Endocan, endothelial cell-specific molecule-1;HMOX1, heme oxygenase 1; SOD2, superoxide dismutase 2; INS-R, insulin receptor; PDGF-D, platelet-derived growth factor D; PCDH12, protocadherin-12; E-selectin,endothelial selectin; VCAM-1, vascular cell adhesion molecule 1; PMCH, pro-melanin concentrating hormone; TM, thrombomodulin; UBE2J1, ubiquitin-conjugatingenzyme E2, J1.
a p � 0.05 is relative to scrambled siRNA-transfected uninduced/basal level.b p � 0.05 is relative to uninduced/basal level.c p � 0.001 is relative to basal level.dp � 0.05 is relative to si-p47phox-transfected basal level.
Selective Sensitivity of VEGF Signaling Pathways to Redox
NOVEMBER 30, 2007 • VOLUME 282 • NUMBER 48 JOURNAL OF BIOLOGICAL CHEMISTRY 35379
anNADPHoxidase-independent gene, DAF, was unaffected byinhibition of PI3Kor p38MAPK (Fig. 9D). These results suggestthat relative sensitivity of PI3K, p38MAPK, and ERK1/2may inpart account for the differential effect of NADPH oxidase onVEGF-mediated gene expression.Inhibition of Akt Phosphorylation in VEGF-treated NADPH
Oxidase-depleted Endothelial Cells Is Sufficient for ModulatingDownstream Gene Expression—We have demonstrated previ-ously that VEGF reduces mRNA expression of GADD45A inendothelial cells through a mechanism that involves phospho-Akt-mediated exclusion of forkhead from the nucleus (49).Consistent with the redox sensitivity of Akt, this effect wasreversed by si-p47phox, DN-Akt, or constitutively active triple-mutant (TM) FKHR (Fig. 10). In contrast, expression of a CA-Akt alone or in the presence of si-p47phox mimicked the effectof VEGF on GADD45A expression (Fig. 10). Data demonstrat-ing a role for Akt-FKHR inmediating the expression of another
redox-dependent gene, MnSOD, were obtained (data notshown). Taken together, these findings suggest that activationof the redox-sensitive signaling intermediates may modulateactivity of downstream transcription factors, which in turn leadto alterations in target gene expression.Inhibition of NADPH Oxidase in Primary Human Endothe-
lial Cells Has a Differential Effect on VEGF-mediated Endothe-lial Cell Function—VEGF has been shown previously to inducethe expression of thrombomodulin and DAF in cultured endo-thelial cells (49, 64, 65). VEGF-mediated induction of thrombo-modulin, but not DAF, was blocked by si-p47phox (Tables 1 and2 and Fig. 7). To determine whether differential redox sensitiv-ity occurred at a functional level, we carried out in vitro assaysfor thrombomodulin and DAF activity. The thrombomodulinassay relies on its ability to activate exogenously supplied pro-tein C, which in turn catalyzes a substrate S2366 (measured at405 nMwavelength). Treatment ofHCAECwithVEGF resulted
TABLE 2VEGF-inducible, NADPH oxidase-independent genes in endothelial cells as determined by real time-PCR analyses (-fold induction)The basal levels of expression for each gene in scrambled siRNA-transfected, unstimulated, serum-starved HCAEC were arbitrarily set at 1 (-fold) per 106 18 S mRNAcopies. Numbers are expressed as -fold induction (or reduction if less than 1) over the basal levels. The abbreviations used are as follows: BMP-2, bone morphogeneticprotein 2; DSCR1, Down syndrome critical region 1; DAF, decay-accelerating factor (CD55); PDGF-A, platelet-derived growth factor A; VEGF-C, vascular endothelialgrowth factor C; Egr-1, Early growth response 1; KLF5, Kruppel-like factor 5; ICAM-1, intercellular cell adhesionmolecule 1; PAI-1, plasminogen activator inhibitor 1; t-PA,plasminogen activator, tissue; SPRY4, Sprouty protein 4.
a p � 0.001 is relative to scrambled siRNA basal level.b p � 0.05 is relative to si-p47phox-transfected uninduced basal level.c p � 0.05 is relative to scrambled siRNA basal level.dp � 0.05 is relative to scrambled siRNA-transfected, uninduced/basal level.e p � 0.05 is relative to scrambled siRNA-transfected, VEGF-treated level.
TABLE 3NADPH oxidase-dependent genes that are repressed by VEGF in endothelial cells as determined by real time-PCR analyses (-fold reduction)The basal levels of expression for each gene in scrambled siRNA-transfected, unstimulated, serum-starved HCAEC were arbitrarily set at 1 (-fold) per 106 18 S mRNAcopies. Numbers are expressed as -fold induction (or reduction if less than 1) over the basal levels. The abbreviations used are as follows: ADFP, adipose differentiation-related protein; BTG-1, B cell translocation gene 1; CDKN1B, cyclin-dependent kinase inhibitor 1B;GADD45A, growth arrest andDNAdamage-inducible 45�; GADD45B,growth arrest and DNA damage-inducible 45�; CASP8, caspase 8; INPP5D, inositol polyphosphate-5-phosphatase D;MAP2K7, mitogen-activated protein kinase 7; NFIB,nuclear factor I/B; THSD1, thrombospondin, type 1; TNFRSF1B, tumor necrosis factor receptor superfamily, member 1b; TNFSF10, tumor necrosis factor (ligand)superfamily, member 10; TXNIP, thioredoxin-interacting protein.
a p � 0.05 relative to scrambled siRNA-transfected, uninduced/basal level.b p � 0.05 relative to scrambled siRNA basal level.c p � 0.001 relative to scrambled siRNA basal level (downregulated).dp � 0.001 relative to scrambled siRNA basal level (upregulated).
Selective Sensitivity of VEGF Signaling Pathways to Redox
35380 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282 • NUMBER 48 • NOVEMBER 30, 2007
in�2-fold induction of thrombomodulin activity, an effect thatwas abrogated by si-p47phox (Fig. 11A), suggesting VEGF-me-diated induction in the activity of thrombomodulin wasNADPHoxidase-dependent. TheDAF assaymeasures comple-ment-mediated lysis of endothelial cells (49). Consistent withthe gene expression data, VEGF-mediated protection againstC3b complementwas unaffected by si-p47phox (Fig. 11B). Thesefindings suggest that VEGF induction of thrombomodulin
activity but not DAF is sensitive to NADPH oxidase in endo-thelial cells.
DISCUSSION
Vascular NADPH oxidase is believed to play an importantrole in both physiology and pathophysiology. NADPH oxidaseactivity varies between vascular cells and betweendifferent sitesof the vasculature. For example, vascular smooth muscle cellsexpress high levels of Nox1 andNox4, whereas endothelial cellsexpress high levels of gp91phox/Nox2 and Nox4 (reviewed inRef. 67). Cultured human microvascular endothelial cells dis-play higher NADPH oxidase activity compared with HUVEC(68). Human saphenous veins were shown to express moreNox2 and p22phox, whereas internal mammary arteriesexpressed relatively high Nox4 (69). Many extracellular factorshave been reported to induce NADPH oxidase activity in cul-tured endothelial cells. The nature of input signals varies atdifferent sites of the vascular tree and over time. Together,these observations underscore the remarkable complexity ofNADPH oxidase signaling in the intact vasculature.Adding to this complexity is our finding that NADPH oxi-
dase activity selectively modulates downstream signalinginduced by a single agonist, namely VEGF. Differential effectsof NADPH oxidase inhibition on VEGF signaling wereobserved at the level of signal transduction, transcription factoractivation, target gene expression, and cell function.Previous studies have implicated a role for ROS inmediating
reversible receptor autophosphorylation in response to suchligands as insulin, epidermal growth factor, and platelet-de-rived growth factor (70, 71, 72). In addition, angiotensin II-mediated transactivation of the epidermal growth factor recep-tor and platelet-derived growth factor receptor in vascular
smooth muscle cells was shown tobe redox-sensitive (73, 74). Not allprotein-tyrosine kinase receptorsare similarly sensitive to ROS. Forexample, the NADPH oxidaseinhibitor diphenyleneiodonium(DPI) or apocynin had no effect onAng1-mediated phosphorylation ofTie2 (35). In this study, we demon-strated that si-p47phox–mediatedknockdown of ROS had no effect onVEGFR2/KDR tyrosine phospho-rylation at 1 and 5 min followingVEGF treatment and only partiallyreduced phosphorylation at 10 min.Similar results were obtained withDPI (data not shown). Two previousstudies have shown a role for ROS inautophosphorylation of VEGFR2/KDR. In one case, pretreatment ofporcine aortic endothelial cells withcatalase or the antioxidant NDGApartially inhibited VEGF inductionofVEGR2phosphorylation between10 and 15min (23). These results areconsistent with our 10-min data. In
FIGURE 8. Northern blot analyses demonstrate two distinct sets of VEGFtarget genes in HCAEC. Scram-si or si-p47phox-transfected HCAEC wereserum-starved overnight prior to incubation with VEGF (50 ng/ml) for 4 h. Leftpanel shows VEGF target genes that are sensitive to NADPH oxidase activity.Right panel shows VEGF target genes that do not require NADPH oxidaseactivity for expression. The Northern blots shown are representative of threeindependent experiments.
FIGURE 9. PI3K or p38 inhibition but not MEK/ERK1/2 inhibition can reverse VEGF-mediated induction ofVCAM-1 and E-selectin and VEGF-mediated repression of GADD45A. Fold induction or reduction of thegenes in HCAEC by quantitative real time-PCR analyses of total RNA is shown. HCAEC were preincubated for 30min with LY294002 (50 �M), PD98059 (50 �M), or SB203580 (10 �M) and then treated in the absence of presenceof VEGF (50 ng/ml) for 4 h. The basal levels of expression for each gene in unstimulated HCAEC were arbitrarilyset at 1 (-fold) per 106 18 S mRNA copies. Numbers are expressed as fold change over basal levels (mean � S.D.).*, p � 0.05, relative to basal levels; †, p � 0.05, relative to VEGF treatment without inhibitor.
Selective Sensitivity of VEGF Signaling Pathways to Redox
NOVEMBER 30, 2007 • VOLUME 282 • NUMBER 48 JOURNAL OF BIOLOGICAL CHEMISTRY 35381
another study, preincubation of HUVEC with chemical inhibi-tors of ROS or transfection with antisense against gp91phoxresulted in partial attenuation of VEGFR2/KDR autophospho-rylation at 5 min (24). The reason for the discrepancy betweenthe results of the latter report and the current study is not clear.Nevertheless, in neither case did ROS inhibition cause com-plete loss of VEGFR2/KDR autophosphorylation. Our resultsdemonstrating insensitivity of VEGFR2 to NADPH oxidaseactivity at early time points are in accordance with the findingsof Berk and co-workers (75). Our findings also suggest thatinhibition of VEGFR2/KDR phosphorylation at later timepoints (10 and 30 min) was insufficient to completely abrogatesignaling (as evidenced by the existence of a subclass ofNADPHoxidase-independent pathways). The observation that NADPHoxidase is not required for VEGF-mediated phosphorylation oftwo major phosphorylation sites (Tyr-951 and Y Tyr-1175)argues against a tyrosine residue-specific effect of ROS onVEGF autophosphorylation. Although it is formally possiblethat other sites within the receptor are differentially sensitive tothe redox state, amore likely explanation for our findings is thatearly and transient activation of VEGFR2/KDR is adequate forpropagating downstream signals.That VEGF activation of PI3K/Akt is redox-sensitive is con-
sistent with previous findings. For example, in HUVEC, VEGF-mediated phosphorylation of Akt was blocked by DPI, apocy-nin, and catalase (76). Studies of Ang1 have yielded conflictingresults. Chen et al. (35) demonstrated that Ang1-dependentphosphorylation of Akt at Ser-473 in porcine andmurine endo-
thelial cells was dependent on NADPH oxidase-derived ROS.In another study, Ang1-mediated phosphorylation of Akt inHUVEC was unaffected by overexpression of superoxide dis-mutase, catalase, or dominant-negative Rac1 (Rac1N17) (36).We have found that hepatocyte growth factor-mediated activa-tion of Akt in HCAEC occurs independently of NADPH oxi-dase activity (supplemental Fig. 5). These findings suggest thatthe redox sensitivity of PI3K/Akt in endothelial cells is ligand-and receptor-dependent.In time course experiments, siRNA-mediated inhibition of
p47phox attenuated VEGF activation of Akt prior to any observ-able effect on VEGFR2/KDR phosphorylation. These data sug-gest that NADPH oxidase-derived ROS exert their positiveeffects on Akt at site(s) distal to the VEGF receptor. We haveshown si-p47phox blocks VEGF-mediated phosphorylation andactivity of Src. This finding is consistent with a previous inves-tigation in which antioxidants attenuated VEGF-induced Srcphosphorylation in HUVEC (77). Importantly, we also demon-strated that Src knockdown attenuates VEGF stimulation ofAkt phosphorylation. Thus, NADPH oxidase influences Aktsignaling via an effect on Src (Fig. 12). The precise mechanismunderlying redox sensitivity of Src requires further study.
FIGURE 10. Redox-sensitive signaling intermediate Akt modulates activ-ity of the downstream transcription factor FKHR that in turn regulatesgene expression. A, HCAEC were transduced with replication-deficient con-trol (Adv), dominant-negative Akt (DN-Akt), or constitutively active myristoy-lated Gag-Akt (CA-Akt) expressing adenoviruses as described under “Experi-mental Procedures.” Whole-cell lysates (20 �g/lane) were subject to Westernblot analysis, and the membrane was probed with anti-Akt antibody. Adeno-virally expressed DN-Akt was of the same mobility as native Akt (indicated onthe left side of the panel), whereas Gag-Akt showed a slower mobility (indi-cated on the right side of the panel). B, HCAEC were either transfected withsiRNAs (Scram-si or si-p47phox) or transduced with adenoviruses (Adv, DN-Akt, or CA-Akt) as indicated. The cells were then serum-starved overnight andtreated without (�) or with (�) VEGF (50 ng/ml) for 15 min. Western blotswere carried out using anti-phospho-Ser-256 FKHR (p-FKHR) antibody. Themembrane was then stripped and reprobed using anti-�-actin antibody(�-actin) as loading control. C, Northern blot analyses demonstrate that VEGF-mediated down-regulation of GADD45A was reversed by si-p47phox, DN-Akt,or constitutively active triple mutant FKHR. On the contrary, expression of aconstitutively active (CA) Akt mimicked VEGF effects and counteracted theeffects of si-p47phox on GADD45A expression. HCAEC were transfected withsiRNA or transduced with adenoviruses as described in B. Cells were serum-starved overnight followed by VEGF treatment for 4 h. Total RNAs wereextracted and subjected to Northern analysis as described under “Experimen-tal Procedures.” Lower panel shows ethidium bromide-stained 28 S as loadingcontrol. A, control adenoviruses; D, DN-Akt; C, CA-Akt; Si, si-p47phox; SC,si-p47phox plus CA-Akt; SD, si-p47phox plus DN-Akt; T, TM-FKHR.
FIGURE 11. VEGF-mediated thrombomodulin-dependent activation ofprotein C is NADPH oxidase-dependent, but protection from comple-ment-mediated endothelial cell lysis is not. A, activation of protein C wasmeasured as a function of thrombomodulin activity as described under“Experimental Procedures.” Scram-si or si-p47phox-transfected HCAEC wereincubated with VEGF for 14 h and assayed for thrombomodulin-dependentactivation of protein C (aPC) using chromogenic substrate S-2366 at 405 nmwavelength. *, p � 0.01 (VEGF-treated versus untreated); †, p � 0.05 (si-p47phox-transfected plus VEGF versus Scram-si-transfected plus VEGF). B, C3bcomplement-mediated lysis of HCAEC was assayed as described under“Experimental Procedures.” Scram-si or si-p47phox-transfected HCAEC wereserum-starved overnight and treated with or without VEGF for 24 h prior tocell lysis assay using calcein acetoxymethyl ester. *, p � 0.05 (VEGF-treatedversus untreated). The data are obtained from three independent experi-ments and analyzed as mean � S.D.
Selective Sensitivity of VEGF Signaling Pathways to Redox
35382 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282 • NUMBER 48 • NOVEMBER 30, 2007
We have demonstrated a role for NADPH oxidase activity inmediating the effect of VEGF on p38 MAPK but not ERK1/2 orJNK. These data contrast with those ofWu et al. (78) who showedthat VEGF-mediated phosphorylation of JNK was oxidase-dependent. Moreover, previous studies in HUVEC have demon-strated a role for ROS in VEGF-mediated activation of ERK1/2 inHUVEC and porcine aortic endothelial cells (23, 36). The rea-son for the discrepancies between these latter reports and thepresent study is not clear. However, our findings that both AS-oligonucleotides and siRNA against p47phox had no effect onERK1/2 or JNK phosphorylation in HUVEC and HCAEC inmultiple independent time course experiments strongly argueagainst a role for NADPH oxidase-derived ROS in VEGF acti-vation of these signal intermediates.It is interesting to note that similar discrepancies exist
regarding the role for ROS in Ang1-mediatedMAPK signaling.One group demonstrated that overexpression of superoxidedismutase or Rac1N17 in HUVEC resulted in potentiation ofAng1-mediated phosphorylation of ERK1/2 but not p38MAPK(36). Chen et al. (35), employing porcine coronary artery endo-thelial cells, as well as wild type and p47phoxmouse heartmicro-vascular endothelial cells, demonstrated a role for NADPHoxi-dase in Ang1-stimulated phosphorylation of ERK1/2.These discrepancies notwithstanding, the effect of ROS on
the MAPK family appears to be dependent on the agonist andcell type. For example, whereas the current results argue againsta role for NADPH oxidase in mediating VEGF activation ofERK1/2 in endothelial cells, previous studies support a role forthe enzyme in angiotensin II-mediated activation of ERK1/2 inendothelial cells (79). Urotensin-II-mediated activation ofERK1/2, p38 MAPK, and JNK was inhibited by DPI and anti-sense to p22phox in human pulmonary artery smooth musclecells (80). Thrombin-mediated activation of p38MAPKbut notERK1/2 in vascular smooth muscles cells was p22phox-depend-
ent (81). Endothelin-1-mediated activation of JNK, but notERK1/2, in vascular smooth muscle cells was inhibited by DPI(82). In murine cardiac microvascular endothelial cells, hypox-ia-reoxygenation-mediated phosphorylation of ERK1/2 andAkt was NADPH oxidase-dependent (83). Another study dem-onstrated a critical role for p47phox in tumor necrosis factor-�-mediated activation of ERK1/2 and p38 MAPK in cardiacmicrovascular endothelial cells (66).Previously, we have shown that NADPH oxidase-derived
ROS play an important role in mediating the effect of VEGF onthe activity of the transcription factors, NF-�B and forkhead(44). The current study extends these observations by demon-strating an inhibitory action of si-p47phox on VEGF-mediatednuclear exclusion of FKHR. This study has also established alink between the relative redox sensitivity of theVEGF signalingintermediates (e.g. PI3K-Akt, p38 MAPK/ERK1/2) and the dif-ferential effects ofNADPHoxidase-derived ROSon the expres-sion of several VEGF-dependent genes. The data also sup-ported the notion that redox-dependent alterations of asignaling molecule (e.g. Akt) by VEGF may result in the modu-lation of the activity of a downstream transcription factor(s)(e.g. FKHR), which in turn leads tomodification of gene expres-sion (e.g. GADD45A). Additionally, based on the results of theDNA microarrays, we predict the existence of a class of tran-scription factors whose activity is not sensitive to NADPH oxi-dase activity. Identification of those ROS-insensitive factorsrequires further studies.Prior studies from our own group (21, 22), as well as others
(23, 24, 43, 76), have demonstrated an important role forNADPHoxidase inmediating the effect ofVEGFon endothelialcell migration and proliferation and angiogenesis. Here, wehave shown that NADPH oxidase is also necessary for VEGF-mediated induction of thrombomodulin activity. However,inhibition of NADPH oxidase had no effect on VEGF-stimu-lated DAF function. These findings are in accordance with thedifferential effect of NADPH oxidase on downstream signalingpathways, transcription factors, and gene expression, and theyprovide compelling evidence for the existence of NADPHoxidase-dependent and -independent VEGF functions.The finding that NADPH oxidase is necessary for some, but
not all, functions of VEGF have important biological and ther-apeutic implications. The data suggest that basal and/or induc-ible ROS are not required for global VEGF signaling but ratherserve as signal intermediates in selected downstream pathways.In this way, alteration of NADPH oxidase activity either inresponse to agonists or drugs will have highly selective effectson VEGF signal transduction and endothelial cell phenotypes.Such an effectmay be leveraged for therapeutic gain. For exam-ple, it is plausible that the therapeutic inhibition of NADPHoxidase would inhibit VEGF stimulation of migration and pro-liferation, while retaining certain protective effects of VEGFsignaling, e.g. protection against complement lysis.
Acknowledgment—We are grateful to Lewis C. Cantley for helpfuldiscussions.
FIGURE 12. Proposed model for the bifurcation of VEGF signals intoredox-sensitive and redox-insensitive pathways downstream ofVEGFR2. VEGF-induced activation of Src kinase is dependent on NADPHoxidase activity in endothelial cells. The redox-sensitive VEGF signalingpathway (shown in colored box) appears to deviate from the redox-insen-sitive pathway (PLC�, ERK1/2) at the level of Src kinase. IP3, inositol 1,4,5-trisphosphate; DAG, diacylglycerol.
Selective Sensitivity of VEGF Signaling Pathways to Redox
NOVEMBER 30, 2007 • VOLUME 282 • NUMBER 48 JOURNAL OF BIOLOGICAL CHEMISTRY 35383
REFERENCES1. Dvorak, H. F. (2000) Semin. Perinatol. 24, 75–782. Zachary, I., Mathur, A., Yla-Herttuala, S., and Martin, J. (2000) Arterio-
scler. Thromb. Vasc. Biol. 20, 1512–15203. Matsumoto, T., and Claesson-Welsh, L. (2001) Sci. STKE 2001, RE214. Ferrara, N., Gerber, H. P., and LeCouter, J. (2003) Nat. Med. 9, 669–6765. Mukhopadhyay, D., Zeng, H., and Bhattacharya, R. (2004)Mol. Cell. Bio-
chem. 264, 51–616. Zelzer, E., and Olsen, B. R. (2005) Curr. Top. Dev. Biol. 65, 169–1877. Folkman, J. (2004) APMIS 112, 496–5078. Stupack, D. G., and Cheresh, D. A. (2004) Curr. Top. Dev. Biol. 64,
207–2389. Xia, P., Aiello, L. P., Ishii, H., Jiang, Z. Y., Park,D. J., Robinson,G. S., Takagi,
H., Newsome,W. P., Jirousek,M. R., and King, G. L. (1996) J. Clin. Investig.98, 2018–2026
10. Guo, D., Jia, Q., Song, H. Y.,Warren, R. S., and Donner, D. B. (1995) J. Biol.Chem. 270, 6729–6733
11. Kroll, J., and Waltenberger, J. (1997) J. Biol. Chem. 272, 32521–3252712. D’Angelo, G., Struman, I., Martial, J., and Weiner, R. I. (1995) Proc. Natl.
Acad. Sci. U. S. A. 92, 6374–637813. Rousseau, S., Houle, F., Landry, J., and Huot, J. (1997) Oncogene 15,
2169–217714. Chen, Z., Fisher, R. J., Riggs, C. W., Rhim, J. S., and Lautenberger, J. A.
(1997) Cancer Res. 57, 2013–201915. Korpelainen, E. I., Karkkainen, M., Gunji, Y., Vikkula, M., and Alitalo, K.
(1999) Oncogene 18, 1–816. Johnson, E. N., Lee, Y. M., Sander, T. L., Rabkin, E., Schoen, F. J., Kaushal,
S., and Bischoff, J. (2003) J. Biol. Chem. 278, 1686–169217. Marumo, T., Schini-Kerth, V. B., and Busse, R. (1999) Diabetes 48,
1131–113718. Mechtcheriakova, D., Wlachos, A., Holzmuller, H., Binder, B. R., and
Hofer, E. (1999) Blood 93, 3811–382319. Abid, M. R., Guo, S., Minami, T., Spokes, K. C., Ueki, K., Skurk, C., Walsh,
K., and Aird, W. C. (2004) Arterioscler. Thromb. Vasc. Biol. 24, 294–30020. Li, J. M., and Shah, A. M. (2004) Am. J. Physiol. 287, R1014–R103021. Abid, M. R., Kachra, Z., Spokes, K. C., and Aird, W. C. (2000) FEBS Lett.
486, 252–25622. Abid, M. R., Tsai, J. C., Spokes, K. C., Deshpande, S. S., Irani, K., and Aird,
W. C. (2001) FASEB J. 15, 2548–255023. Colavitti, R., Pani, G., Bedogni, B., Anzevino, R., Borrello, S., Walten-
berger, J., and Galeotti, T. (2002) J. Biol. Chem. 277, 3101–310824. Ushio-Fukai, M., Tang, Y., Fukai, T., Dikalov, S. I., Ma, Y., Fujimoto, M.,
Quinn, M. T., Pagano, P. J., Johnson, C., and Alexander, R.W. (2002)Circ.Res. 91, 1160–1167
25. Jones, S. A., O’Donnell, V. B., Wood, J. D., Broughton, J. P., Hughes, E. J.,and Jones, O. T. (1996) Am. J. Physiol. 271, H1626–H1634
26. Bayraktutan, U., Draper, N., Lang, D., and Shah, A. M. (1998) Cardiovasc.Res. 38, 256–262
27. Gorlach, A., Brandes, R. P., Nguyen, K., Amidi, M., Dehghani, F., andBusse, R. (2000) Circ. Res. 87, 26–32
28. Ago, T., Kitazono, T., Ooboshi, H., Iyama, T., Han, Y. H., Takada, J.,Wakisaka, M., Ibayashi, S., Utsumi, H., and Iida, M. (2004) Circulation109, 227–233
29. Van Buul, J. D., Fernandez-Borja, M., Anthony, E. C., and Hordijk, P. L.(2005) Antioxid. Redox. Signal. 7, 308–317
30. Li, J. M., and Shah, A. M. (2003) J. Am. Soc. Nephrol. 14, S221–S22631. Li, J. M., and Shah, A. M. (2003) J. Biol. Chem. 278, 12094–1210032. De Keulenaer, G. W., Chappell, D. C., Ishizaka, N., Nerem, R. M., Alex-
ander, R. W., and Griendling, K. K. (1998) Circ. Res. 82, 1094–110133. Zhuo, L., Sun, B., Zhang, C. L., Fine, A., Chiu, S. Y., andMessing, A. (1997)
Dev. Biol. 187, 36–4234. Howard, A. B., Alexander, R. W., Nerem, R. M., Griendling, K. K., and
Taylor, W. R. (1997) Am. J. Physiol. 272, C421–C42735. Chen, J. X., Zeng, H., Lawrence, M. L., Blackwell, T. S., and Meyrick, B.
(2006) Am. J. Physiol. 291, H1563–H157236. Harfouche, R., Malak, N. A., Brandes, R. P., Karsan, A., Irani, K., and
38. Holland, J. A., Meyer, J. W., Chang, M. M., O’Donnell, R. W., Johnson,D. K., and Ziegler, L. M. (1998) Endothelium 6, 113–121
39. Landmesser, U., Dikalov, S., Price, S. R., McCann, L., Fukai, T., Holland,S. M., Mitch, W. E., and Harrison, D. G. (2003) J. Clin. Investig. 111,1201–1209
40. Duerrschmidt, N., Wippich, N., Goettsch, W., Broemme, H. J., andMorawietz, H. (2000) Biochem. Biophys. Res. Commun. 269, 713–717
41. Hu, T., Ramachandrarao, S. P., Siva, S., Valancius, C., Zhu, Y., Mahadev,K., Toh, I., Goldstein, B. J., Woolkalis, M., and Sharma, K. (2005) Am. J.Physiol. 289, F816–F825
42. Al-Mehdi, A. B., Zhao, G., Dodia, C., Tozawa, K., Costa, K., Muzykantov,V., Ross, C., Blecha, F., Dinauer, M., and Fisher, A. B. (1998) Circ. Res. 83,730–737
43. Tojo, T., Ushio-Fukai,M., Yamaoka-Tojo,M., Ikeda, S., Patrushev, N., andAlexander, R. W. (2005) Circulation 111, 2347–2355
44. Abid, M. R., Schoots, I. G., Spokes, K. C., Wu, S. Q., Mawhinney, C., andAird, W. C. (2004) J. Biol. Chem. 279, 44030–44038
45. Griendling, K. K., Minieri, C. A., Ollerenshaw, J. D., and Alexander, R. W.(1994) Circ. Res. 74, 1141–1148
46. Wada, Y., Otu,H.,Wu, S., Abid,M. R., Okada,H., Libermann, T., Kodama,T., Shih, S. C.,Minami, T., andAird,W. C. (2005) FASEB J. 19, 1914–1916
47. Barash, Y., Dehan, E., Krupsky,M., Franklin,W., Geraci,M., Friedman,N.,and Kaminski, N. (2004) Bioinformatics (Oxf.) 20, 839–846
48. Zhang, J. J., Bledsoe, G., Kato, K., Chao, L., and Chao, J. (2004) Kidney Int.66, 722–732
49. Abid, M. R., Shih, S. C., Otu, H. H., Spokes, K. C., Okada, Y., Curiel, D. T.,Minami, T., and Aird, W. C. (2006) J. Biol. Chem. 281, 35544–35553
50. Shih, S. C., Robinson,G. S., Perruzzi, C. A., Calvo, A., Desai, K., Green, J. E.,Ali, I. U., Smith, L. E., and Senger, D. R. (2002) Am. J. Pathol. 161, 35–41
51. Kerlin, B. A., Yan, S. B., Isermann, B. H., Brandt, J. T., Sood, R., Basson,B. R., Joyce, D. E., Weiler, H., and Dhainaut, J. F. (2003) Blood 102,3085–3092
52. Osawa, M., Itoh, S., Ohta, S., Huang, Q., Berk, B. C., Marmarosh, N. L.,Che, W., Ding, B., Yan, C., and Abe, J. (2004) J. Biol. Chem. 279,29691–29699
53. Claesson-Welsh, L. (2003) Biochem. Soc. Trans. 31, 20–2454. Alavi, A., Hood, J. D., Frausto, R., Stupack, D.G., andCheresh, D. A. (2003)
Science 301, 94–9655. Kanno, S., Oda, N., Abe, M., Terai, Y., Ito, M., Shitara, K., Tabayashi, K.,
Shibuya, M., and Sato, Y. (2000) Oncogene 19, 2138–214656. Ashton, A. W., and Ware, J. A. (2004) Circ. Res. 95, 372–37957. Hunter, T. (1987) Cell 49, 1–458. Calalb, M. B., Polte, T. R., and Hanks, S. K. (1995) Mol. Cell. Biol. 15,
954–96359. Schlaepfer, D. D., Hanks, S. K., Hunter, T., and van der Geer, P. (1994)
Nature 372, 786–79160. Matsumoto, T., and Mugishima, H. (2006) J. Atheroscler. Thromb. 13,
son, P., Wang, L., Wikner, C., Qi, J. H., Wernstedt, C., Wu, J., Bruheim, S.,Mugishima, H., Mukhopadhyay, D., Spurkland, A., and Claesson-Welsh,L. (2005) EMBO J. 24, 2342–2353
62. Olsson,A. K., Dimberg, A., Kreuger, J., andClaesson-Welsh, L. (2006)Nat.Rev. Mol. Cell Biol. 7, 359–371
63. Holmqvist, K., Cross, M. J., Rolny, C., Hagerkvist, R., Rahimi, N., Mat-sumoto, T., Claesson-Welsh, L., andWelsh, M. (2004) J. Biol. Chem. 279,22267–22275
64. Mason, J. C., Steinberg, R., Lidington, E. A., Kinderlerer, A. R., Ohba, M.,and Haskard, D. O. (2004) J. Biol. Chem. 279, 41611–41618
65. Calnek, D. S., and Grinnell, B. W. (1998) Exp. Cell Res. 238, 294–29866. Li, J. M., Fan, L. M., Christie, M. R., and Shah, A. M. (2005)Mol. Cell. Biol.
25, 2320–233067. Chen, K., and Keaney, J. (2004) Endothelium 11, 109–12168. Li, J. M., and Shah, A. M. (2001) Cardiovasc. Res. 52, 477–48669. Guzik, T. J., West, N. E., Black, E., McDonald, D., Ratnatunga, C., Pillai, R.,
and Channon, K. M. (2000) Circ. Res. 86, E85–E90
Selective Sensitivity of VEGF Signaling Pathways to Redox
35384 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282 • NUMBER 48 • NOVEMBER 30, 2007
70. Bae, Y. S., Kang, S. W., Seo, M. S., Baines, I. C., Tekle, E., Chock, P. B., andRhee, S. G. (1997) J. Biol. Chem. 272, 217–221
71. Catarzi, S., Biagioni, C., Giannoni, E., Favilli, F., Marcucci, T., Iantomasi,T., and Vincenzini, M. T. (2005) Biochim. Biophys. Acta 1745, 166–175
72. Mahadev, K., Zilbering, A., Zhu, L., and Goldstein, B. J. (2001) J. Biol.Chem. 276, 21938–21942
73. Heeneman, S., Haendeler, J., Saito, Y., Ishida, M., and Berk, B. C. (2000)J. Biol. Chem. 275, 15926–15932
74. Ushio-Fukai, M., Griendling, K. K., Becker, P. L., Hilenski, L., Halleran, S.,and Alexander, R. W. (2001) Arterioscler. Thromb. Vasc. Biol. 21,489–495
75. Tanimoto, T., Jin, Z. G., and Berk, B. C. (2002) J. Biol. Chem. 277,42997–43001
76. Yamaoka-Tojo, M., Ushio-Fukai, M., Hilenski, L., Dikalov, S. I., Chen,Y. E., Tojo, T., Fukai, T., Fujimoto,M., Patrushev, N. A.,Wang,N., Kontos,
C. D., Bloom, G. S., and Alexander, R. W. (2004) Circ. Res. 95, 276–28377. Lin, M. T., Yen, M. L., Lin, C. Y., and Kuo, M. L. (2003) Mol. Pharmacol.
64, 1029–103678. Wu, R. F., Gu, Y., Xu, Y. C., Nwariaku, F. E., and Terada, L. S. (2003) J. Biol.
Chem. 278, 36830–3684079. Xie, Z., Pimental, D. R., Lohan, S., Vasertriger, A., Pligavko, C., Colucci,
W. S., and Singh, K. (2001) J. Cell. Physiol. 188, 132–13880. Djordjevic, T., BelAiba, R. S., Bonello, S., Pfeilschifter, J., Hess, J., and
Gorlach, A. (2005) Arterioscler. Thromb. Vasc. Biol. 25, 519–52581. Brandes, R. P., Viedt, C., Nguyen, K., Beer, S., Kreuzer, J., Busse, R., and
Gorlach, A. (2001) Thromb. Haemostasis 85, 1104–111082. Simoncini, T., Maffei, S., Basta, G., Barsacchi, G., Genazzani, A. R., Liao,
J. K., and De Caterina, R. (2000) Circ. Res. 87, 19–2583. Chen, J. X., Zeng, H., Tuo, Q. H., Yu, H., Meyrick, B., and Aschner, J. L.
(2007) Am. J. Physiol. 292, H1664–H1674
Selective Sensitivity of VEGF Signaling Pathways to Redox
NOVEMBER 30, 2007 • VOLUME 282 • NUMBER 48 JOURNAL OF BIOLOGICAL CHEMISTRY 35385
Legend to supplementary figures: Supplementary Fig 1. Antisense and siRNA against p47phox results in significant inhibition of p47phox expression in primary endothelial cells. (A) Western blot of lysates prepared from control antisense oligonucleotide (Scram-AS) or antisense p47phox (AS-p47phox)-transfected HCAEC or HUVEC followed by immunoblots using anti-p47phox antibody. The membrane was stripped and reprobed using anti-β-actin antibody as loading control. (B) Western blots of lysates prepared from HCAEC transfected with control siRNA (Scram-si) or two different si-p47phox, using anti-p47phox antibody. The membrane was stripped and reprobed using anti-β-actin antibody as loading control. The blots shown are representative of at least two independent experiments. Supplementary Fig 2. NADPH oxidase activity is required for late VEGFR2/KDR phosphorylation, Akt phosphorylation at all time points, but not for ERK1/2 phosphorylation. HCAEC were transfected with control antisense oligonucleotide (Scram-AS) or AS-p47phox, serum starved overnight, and then treated with VEGF (50 ng/ml) for 10 min. (A) HCAEC were transfected with Scram-AS or AS-p47phox, serum starved, and then treated with or without VEGF for 10 min. Immunoprecipitation of HCAEC lysates by anti-VEGFR2/KDR antibody followed by immunoblots with anti-phospho-tyrosine (p-Tyr) antibody was carried out. The membrane was stripped and reprobed with anti-VEGFR2/KDR antibody (KDR) as loading control. (B) Same as in (A) except Western blots were carried out using anti-phospho p473-Akt antibody. The membrane was stripped and reprobed with anti-Akt antibody (Akt) as loading control. (C) Same as in (B) except Western blots were carried out using anti-phospho ERK1/2 (p-ERK1/2) antibody, and the membrane was stripped and reprobed using anti-ERK1/2 antibody as loading control. Bar graphs show quantitative analyses of three independent Western blot experiments (mean ± SD). ∗p <0.05, relative to control. Supplementary Fig 3. VEGF-mediated cytoplasmic to nuclear translocation of ERK1/2 is independent of NADPH oxidase activity. (A) A total of 50,000 Scram-si- or si-p47phox-transfected HCAEC were cultured in 4-well chamber slides, serum starved for 24 h, treated in the absence or presence of VEGF (50 ng/ml) for 15 min, fixed with paraformaldehyde, and then incubated with primary antibody to ERK1/2, followed by Cy3-conjugated anti-IgG secondary antibody. The cells were incubated with DAPI for nuclear staining and observed under fluorescence microscope. (B) Quantitation of the distribution of ERK1/2 in the nucleus (N) and cytoplasm (C) of HCAEC treated with or without VEGF are shown. All experiments were carried out three times independently. Mean±S.D. from 200 cells for each condition are shown. ∗p<0.05 (VEGF-treated vs. control, in both Scram-si and si-p47phox-transfected HCAEC). Supplementary Fig 4. VEGF-mediated nuclear to cytoplasmic translocation of FKHR is dependent on NADPH oxidase activity. (A) A total of 50,000 Scram-si- or si-p47phox-transfected HCAEC were cultured in 4-well chamber slides, serum starved for 24 h, treated in the absence or presence of VEGF (50 ng/ml) for 30 min as described above, and then incubated with primary antibody to FKHR, followed by Cy3-conjugated anti-IgG secondary antibody. DAPI were used for nuclear staining and observed under fluorescence microscope. (B) Quantitation of the distribution of FKHR in the nucleus (N) and cytoplasm (C) of HCAEC treated with or without VEGF are shown. Mean±S.D. from 200 cells for each condition are shown. All experiments were performed in triplicate. ∗p<0.05 (VEGF-treated vs. control); † p<0.05 (si-p47phox-transfected plus VEGF-treated vs. Scram-si-transfected plus VEGF-treated).
34
Supplementary Fig. 5. NADPH oxidase activity is not required for hepatocyte growth factor (HGF)-mediated phosphorylation of Akt and ERK1/2 in HCAEC. Western blot analyses using lysates of HCAEC treated with HGF (20 ng/ml) for the times indicated. HCAEC were transfected with control scrambled siRNA (Scram-si) or siRNA against p47phox subunit of NADPH oxidase (si- p47phox), serum starved, and then treated with or without HGF. Western blots were carried out using anti-phospho ERK1/2 (p-ERK1/2) antibody, and the membrane was stripped and reprobed using anti-phospho-Akt (p473-Akt) antibody. The same membrane was then stripped and reprobed using anti-Akt antibody (Akt) as loading control. The data shown are representative of two independent experiments.
35
36
37
38
39
40
Md. Ruhul Abid, Katherine C. Spokes, Shou-Ching Shih and William C. AirdFactor Signaling Pathways
