P. Bendeck and Duncan J. StewartSaeid Babaei, Krystyna Teichert-Kuliszewska, Juan-Carlos Monge, Farida Mohamed, Michelle

FactorRole of Nitric Oxide in the Angiogenic Response In Vitro to Basic Fibroblast Growth

Print ISSN: 0009-7330. Online ISSN: 1524-4571 Copyright © 1998 American Heart Association, Inc. All rights reserved.is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation Research

doi: 10.1161/01.RES.82.9.10071998;82:1007-1015Circ Res. 

http://circres.ahajournals.org/content/82/9/1007World Wide Web at:

The online version of this article, along with updated information and services, is located on the

  http://circres.ahajournals.org//subscriptions/

is online at: Circulation Research Information about subscribing to Subscriptions: 

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

  document. Permissions and Rights Question and Answer about this process is available in the

located, click Request Permissions in the middle column of the Web page under Services. Further informationEditorial Office. Once the online version of the published article for which permission is being requested is

can be obtained via RightsLink, a service of the Copyright Clearance Center, not theCirculation Researchin Requests for permissions to reproduce figures, tables, or portions of articles originally publishedPermissions:

by guest on April 3, 2014http://circres.ahajournals.org/Downloaded from by guest on April 3, 2014http://circres.ahajournals.org/Downloaded from

Role of Nitric Oxide in the Angiogenic Response In Vitro toBasic Fibroblast Growth Factor

Saeid Babaei, Krystyna Teichert-Kuliszewska, Juan-Carlos Monge, Farida Mohamed,Michelle P. Bendeck, Duncan J. Stewart

Abstract—Angiogenesis is a complex process that involves the activation of quiescent endothelial cells (ECs) to aproliferative and migratory phenotype and, subsequently, their redifferentiation to form vascular tubes. We hypothesizedthat NO contributes to angiogenesis by terminating the proliferative action of angiogenic growth factors and initiatinga genetic program of EC differentiation. Human umbilical vein ECs (HUVECs) and calf pulmonary artery ECs(CPAECs) were grown directly on plastic dishes or on three-dimensional fibrin matrices. In the absence of fibrin,treatment with NO-donor compounds, such asS-nitroso-N-acetylpenicillamine (SNAP, 0.1 and 0.4 mmol/L), produceda dose-dependent inhibition of proliferation in both cell lines, whereas the inhibition of endogenous NO productionusing NG-nitro-L-arginine methyl ester (L-NAME, 1 mmol/L) orNG-monomethyl-L-arginine (L-NMMA, 1 mmol/L)significantly increased proliferation of the CPAECs. The addition of basic fibroblast growth factor (bFGF, 30 ng/mL)increased the expression of endothelial NO synthase mRNA and the production of NO in both cell types when culturedon three-dimensional fibrin gels and produced profound morphological changes characterized by the appearance ofextensive capillary-like vascular structures and the loss of EC monolayers. These changes were quantified by measuringtotal tube length per low-power field (3100), and a differentiation index was derived using the ratio of tube length overarea covered by residual EC monolayer. In the absence of additional angiogenic factors, the differentiation index waslow for both HUVECs and CPAECs (control, 1.1660.19 and 2.0760.87, respectively). Treatment with bFGF increasedthe differentiation index significantly in both cell types (10.5962.03 and 20.0265.01 for HUVECs and CPAECs,respectively;P,.05 versus control), and the addition of SNAP (0.4 mmol/L) mimicked the angiogenic response to bFGF(8.5761.34 and 12.2063.49 for HUVECs and CPAECs, respectively;P,.05 versus control). Moreover, L-NAMEinhibited EC tube formation in response to bFGF in a dose-response manner, consistent with a role of endogenous NOproduction in EC differentiation in this angiogenic model. These findings suggest that NO may act as a crucial signalin the angiogenic response to bFGF, terminating the proliferative actions of angiogenic growth factors and promotingEC differentiation into vascular tubes.(Circ Res. 1998;82:1007-1015.)

Key Words: nitric oxiden angiogenesisn fibrin gel n proliferationn differentiation

Angiogenesis, the formation of new capillaries frompreexisting vascular structures,1 is of paramount impor-

tance in the maintenance of vascular integrity both in therepair of tissue damage in wound healing and in the formationof collateral vessels in response to ischemia.1–3 Angiogenesisis a complex process that is orchestrated by a multitude ofcytokines and growth factors.2 In its broadest sense, angio-genesis cannot be viewed as a single process; it involves aspectrum of cellular events, beginning with stimulation ofquiescent nondividing ECs to a state of rapid proliferationand migration and terminating with the redifferentiation ofactivated ECs into vascular tubes.2,4–6 Thus, it is likely thatdifferent mediators are involved in different phases of angio-genesis. Inflammatory cytokines such as tumor necrosisfactor-a2 and, more recently, soluble E-selectin6 promote EC

migration and matrix dissolution, which are essential for theinitiation of capillary tube formation. During later stages ofangiogenesis, various angiogenic growth factors such asVEGF and bFGF promote growth and may mediate theorganization of ECs into complex vascular tubes.1–4,7

It is now recognized that in addition to its many otherfunctions, the endothelium produces a number of potentvasoactive factors that play a crucial role in regulatingvascular tone and growth. One of the most important endo-thelial vasoactive factors is the free radical gas NO,8 which isproduced from the amino acidL-arginine by the action ofNOS.9 Three isoforms of NOS have been identified: nNOS,iNOS, and eNOS.10 Both nNOS and eNOS are constitutivelyexpressed, and their activity is tightly regulated by calciumand calmodulin,10,11whereas iNOS is expressed in response to

Received December 17, 1997; accepted March 2, 1998.From the Terrence Donnelly Heart Centre, Division of Cardiology, St. Michael’s Hospital, Department of Medicine, University of Toronto, Ontario,

Canada.Correspondence to D.J. Stewart, Director, Division of Cardiology, University of Toronto, and Head, Division of Cardiology, St. Michael’s Hospital,

30 Bond St, Toronto, Ontario, M5B 1W8, Canada.E-mail stewartd@smh.toronto.on.ca© 1998 American Heart Association, Inc.

1007 by guest on April 3, 2014http://circres.ahajournals.org/Downloaded from

stimulation by cytokines.11 NO not only is a potent vasodila-tor and antiplatelet factor12 but also inhibits proliferation andmigration of smooth muscle cells and fibroblasts13; thus, it isthought that NO plays a crucial role in the maintenance ofvascular homeostasis.

Angiogenesis is predominantly an EC process and can beinitiated in vivo by factors that have nearly complete selec-tivity of action for ECs, such as VEGF.7 Moreover, “angio-genesis” can be demonstrated in vitro in EC monoculture.14

Thus, the endothelium is both necessary and sufficient fornew capillary vessel formation. Not surprisingly, it has beensuggested that endothelial factors, such as NO, may play akey role in angiogenesis, although reports to date are con-flicting, showing both angiogenic15 and antiangiogenic activ-ity.16 We hypothesized that NO might contribute to ECdifferentiation in response to angiogenic growth factors,possibly by inhibiting proliferation of activated ECs andpromoting the formation of vascular tubes. The aim of thepresent study was to determine the importance of NO in thewell-characterized three-dimensional fibrin gel model of invitro angiogenesis.

Materials and MethodsEC CultureHUVECs and CPAECs were obtained from the American TypeCulture Collection or isolated from fresh umbilical veins by 0.2%collagenase digestion.17 HUVECs were grown to confluence inHam’s/F-12 medium (GIBCO) and supplemented with 15% FBS,penicillin (500 U/mL), streptomycin (50mg/mL), and heparin (100mg/mL) (all from GIBCO) and EC growth factor (20mg/mL,Boehringer Mannheim) equilibrated with 95% air/5% CO2 at 37°C.CPAECs were grown in DMEM supplemented with 10% FBS andantibiotics as described above. Confluent EC cultures between the5th and 18th passages were washed with HBSS, harvested using0.05% trypsin/0.53 mmol/L EDTA, and counted using a hemocy-tometer. ECs were resuspended in F-12 medium or DMEM andplated on six-well dishes, either directly on plastic or on dishescoated with fibrin matrices. ECs were incubated for 48 hours in thepresence or absence of the following agents singly or in combination:bFGF (30 ng/mL, Boehringer Mannheim), L-NAME (0.3 to3.0 mmol/L), its enantiomer D-NAME (1 mmol/L), L-NMMA(1 mmol/L), SNAP (0.1 and 0.4 mmol/L), SNP (0.1 to 10mmol/L),

GSNO (1 to 100mmol/L), and SIN-1 (0.1 to 10mmol/L, SigmaChemical Co).

Preparation of Fibrin GelsEndotoxin- and plasminogen-free human and bovine fibrinogen (10and 5 mg/mL, respectively; Calbiochem NOVAbiochem Corp) wasdissolved in serum-free medium and filtered through 0.2-mm filters(Millex GS, Millipore). Fibrin matrices were prepared by polymer-izing the fibrinogen solution using a low concentration ofa-throm-bin (2.5 U/mL, Sigma). After polymerization, gels were soaked inculture medium containing 10% FBS for 2 hours at 37°C toinactivate the thrombin. Where appropriate, bFGF was added to boththe fibrinogen solution before polymerization as well to the mediumbathing the cells. ECs were plated on the surface of the three-dimensional matrix and cultured for periods up to 48 hours, in thepresence or absence of study agents as described above.

RNA Extraction and Northern Blot AnalysisTotal cellular RNA was isolated using the method of Chomczynskiand Sacchi.18 Total RNA (20mg) from each sample was treated with2.2 mol/L formaldehyde in 50% formamide and 200 mmol/LHEPES, pH 7, at 65°C for 15 minutes, and RNA was separated byelectrophoresis in 1.2% agarose/2.2 mol/L formaldehyde gels inMOPS buffer containing 20 mmol/L 3-(N-morpho)sulfonic acid,8 mmol/L sodium acetate, 1 mmol/L EDTA, and 5 mmol/L NaOHand transferred by capillary blotting to Gene Screen Plus nylonmembranes (NEN Research Products) following the method sug-gested by the manufacturer. The membrane was optimally cross-linked with UV light (1200 J, UVXL-1000, Fisher Scientific) orbaked at180°C in a vacuum oven. Membranes were hybridized for24 hours at 42°C with specific cDNA probes radiolabeled with[a-32P]CTP (Amersham) by the random primer technique to aspecific activity of at least 13109 cpm/mg in hybridization buffercontaining 50% formamide, 53 SSC (750 mmol/L NaCl and 0.03mol/L sodium citrate), 10% dextran sulfate, 1% SDS, and 100mg/mL denatured salmon sperm DNA. After hybridization, the filterswere washed under conditions of high stringency in 23 SSC (0.3mol/L NaCl and 0.03 mol/L sodium citrate 2H2O, pH 7.0) containingl % SDS according to the manufacturer’s instructions. Autoradiog-raphy was performed using double intensifying screens (Cronex) andX-OMAT AR film (Eastman Kodak Co). Signal intensity wasquantified as integrated areas using scanning densitometry. Densito-metric data were standardized against the level of GAPDH mRNAexpression in each sample. The cDNA probe for eNOS was producedas described previously.19 GAPDH cDNA, a constitutively expressedgene, was obtained from the American Type Culture Collection (No.57091), and a 0.78-kbPstI-XhoI fragment was used as a cDNAprobe.

Reverse-Transcription Polymerase Chain ReactionTotal RNA (2mg) was reverse-transcribed, as previously reported,20

in 20 mL of reaction volume containing 250 ng of random primers,0.5 mmol/L of each dNTP, 50 mmol/L Tris-HCl (pH 8.3),75 mmol/L KCl, 3 mmol/L MgCl2, 10 mmol/L dithiothreitol, and200 U Moloney murine leukemia virus reverse transcriptase(GIBCO/BRL). After 1 hour of incubation at 37°C, samples wereheated to 95°C for 5 minutes and then chilled on ice. Thermalcycling of 5mL of the cDNA from the reverse transcriptase mix wasperformed using specific primers for bFGF: sense, 59-GCCTTCCCGCCCGGCCACTTCAAGG-39; antisense, 59-GCACACACTCCTTTGATAGACACAA-39. Coamplification of the same cDNA wasperformed for GAPDH as an internal standard with the followingprimers: sense, 59-GGTGAAGGTCGGAGTCAACGGATTTGG-39;antisense, 59-GGCCATGAGGTCCACCACCCTGTT-39. Primerswere obtained from the Pharmacia Biotechnology Service Center atthe Hospital for Sick Children (Toronto, Canada). Primer (1mmol/Lof each) was added to the reaction mixture containing 50 mmol/LTris-HCl (pH 8.3), 1.5 mmol/L MgCl2, 50 mmol/L KCl, 0.2 mmol/Lof each dNTP, and 2.5 U of Taq polymerase (Boehringer Mannheim)in a final volume of 50mL. Amplification was performed in a

Selected Abbreviations and Acronyms

bFGF 5 basic FGFBrdU 5 59-bromo-29-deoxyuridine

CPAEC 5 calf pulmonary artery ECEC 5 endothelial cell

eNOS, iNOS, nNOS5 endothelial, inducible, and neuronal NOSFGF 5 fibroblast growth factor

GSNO 5 S-nitrosoglutathioneHUVEC 5 human umbilical vein EC

L-NAME 5 NG-nitro-L-arginine methyl esterL-NMMA 5 NG-monomethyl-L-arginine

NOS 5 NO synthaseRT-PCR5 reverse-transcription polymerase chain

reactionSIN-1 5 3-morpholinosydnonimineSNAP 5 S-nitroso-N-acetylpenicillamine

SNP 5 sodium nitroprussideVEGF 5 vascular endothelial growth factor

1008 Nitric Oxide in Angiogenesis

by guest on April 3, 2014http://circres.ahajournals.org/Downloaded from

thermocycler using the following parameters: for bFGF, 94°C for 1minute, 63°C for 1 minute, and 72°C for 1 minute (35 cycles); forGAPDH, 94°C for 1 minute, 60°C for 1 minute, and 72°C for 1minute (35 cycles). The resultant products were separated on a 1.5%agarose gel, together with a 100-bp DNA ladder, and bands werevisualized with ethidium bromide. The sizes of the amplificationproducts were 179 and 983 nt for bFGF and GAPDH, respectively,and the intensity of the bands were measured by densitometry.

Quantification of EC ProliferationCell replication was determined by quantifying the incorporation of[methyl-3H]thymidine into trichloroacetic acid–insoluble macromol-ecules as described previously.21,22 ECs were seeded at a density 1 to23104 cells per well in 24-well dishes. After cell attachment, bFGFand other study agents were added to the medium, and the cells wereincubated for 24 hours. After a preincubation period of 18 hours, atracer amount of [3H]thymidine (1 mCi per well; specific activity,1.85 TBq/mmol; Amersham) was added, and the cells were incu-bated for a 6-hour period. Subsequently, the cells were washed threetimes with PBS and fixed with ice-cold 10% (wt/vol) trichloroaceticacid for 20 minutes. The resulting precipitate was washed withice-cold 10% trichloroacetic acid and 95% ethanol and solubilized bymixing with 200 mL of 0.3N NaOH at room temperature for 20minutes, followed by neutralization with the same volume of 0.3NHCl. 3H radioactivity was measured in a liquid scintillation counter(Liquid Scintillation System, Beckman Instruments Inc).

In addition, CPAECs were seeded as above on culture slides andcultured for 6 hours before being pulse-labeled with 50mmol/L ofBrdU for 1 hour, washed extensively, and then fixed with 4%paraformaldehyde. The slides were then blocked with 0.3% H2O2 andpermeabilized with 2N HCl at 37°C for 15 minutes. Anti-BrdUmonoclonal antibody (mouse IgG, DAKO) at 1:40 dilution in 1%normal horse serum and 1% BSA/PBS was applied to each slide for60 minutes at 37°C. BrdU staining was detected with a biotinylatedhorse anti-mouse IgG (dilution, 1:250; Vector Laboratories). Finally,the slides were incubated with avidin-biotin-peroxidase complex(1:100 dilution, Vector Laboratories) in PBS for 45 minutes, fol-lowed by incubation with diaminobenzidine in 0.03% H2O2 and50 mmol/L Tris buffer (pH 7.2). The slides were then counterstainedwith hematoxylin, and the total number of cell nuclei stainingpositive for BrdU was counted under each condition and expressedas a percentage of total number of nuclei. Cytotoxic effects ofNO-donor compounds were determined by trypan blue exclusion incell suspension after 48 hours of incubation.

Quantification of Endothelial DifferentiationCulture plates containing HUVECs and CPAECs were assessed after48 hours of incubation under study conditions using an OlympusBX50 inverted microscope (3100). Images were digitized using aSony CCD-IRIS/RGB camera (Cohu Inc) and analyzed using acomputer-assisted morphometric analysis system (C Imaging,Compix Inc) by observers blinded with respect to the experimentalconditions. Tubelike structures ($30 mm) were identified, and totaltube length was derived for each of four randomly chosen fields. Atthe same time, the total area of the culture surface covered by ECswas determined in the same fields. The differentiation index wascalculated as the ratio of the total tube length over cell area for eachfield, and a mean differentiation index value was obtained for eachculture well.

Nitrite AssayNitrite, a stable end product of NO, was measured in the culturemedium using the colorimetric Griess assay. An aliquot of phenolred–free medium (80mL) from each culture well was mixed with 20mL of nitrate reductase for conversion of nitrate to nitrite, followedby 100 mL of the Griess reagent (1% sulfanilamide and 0.1%naphthylenediamine dihydrochloride in 2% phosphoric acid) (AlexisCorp). The mixture was incubated for 10 minutes at room tempera-ture to allow the color to develop, and the absorbance at 540 nm wasmeasured in a microplate reader (Vmax 250, Molecular Devices).Concentrations were determined by comparison with a sodium nitritestandard curve. Results were expressed per 104 cells.

Statistical AnalysisStatistical analysis was performed using SYS-STAT software. Sig-nificance of differences between groups was performed using Stu-dent t test or a two-tailed Wilcoxon signed rank test and a one-wayANOVA for dose-response experiments as appropriate. Values arepresented as mean6SD. Differences were considered statisticallysignificant atP,0.05.

ResultsThe effect of bFGF on EC morphology was dependent on theconditions of cell culture, as shown in Figure 1. In theabsence of bFGF (panels a and c), cultured CPAECs exhib-ited a classical cobblestone appearance regardless of whetherthey were cultured directly on plastic or on fibrin matrices for

Figure 1. CPAECs cultured for 48 hoursdirectly on plastic dishes in the absence(a) or presence (b) of bFGF formed a typ-ical cobblestone monolayer. A similarappearance was seen when ECs wereplated on a fibrin matrix in the absenceof the angiogenic factor (c), whereas theaddition of bFGF to ECs on three-dimen-sional fibrin gels (d) produced dramaticchanges in morphology characterized bythe appearance of abundant tubelike en-dothelial structures. Original magnifica-tion 3100.

Babaei et al May 18, 1998 1009

by guest on April 3, 2014http://circres.ahajournals.org/Downloaded from

48 hours. The addition of bFGF had little effect on morphol-ogy of ECs grown on plastic (panel b), whereas this growthfactor induced profound differentiation of ECs grown onfibrin (panel d), with the appearance of capillary-like tubesand loss of area covered by EC monolayers. Similar resultswere obtained with HUVECs (data not shown).

Figure 2A shows representative Northern blots demonstrat-ing the changes in endothelial gene expression in response tobFGF in both HUVECs and CPAECs grown on plastic or onfibrin matrices. Under nonangiogenic conditions (ie, cultureon plastic), bFGF had little effect on eNOS mRNA expres-sion (lanes 1 and 2). ECs cultured on fibrin alone demon-strated a consistent decrease in the basal expression of eNOS(lane 3); however, under conditions that favored vascular tube

formation, addition of bFGF resulted in a significant increasein the expression of eNOS in both cell types (lane 4). Theseexperiments were repeated three times with similar results.Thrombin at the concentration used to polymerize fibrinogen(2.5 U/mL) had no direct effect on eNOS mRNA expression(data not shown).

The effect of bFGF on NO production by HUVECs andCPAECs is shown in Figure 3. Addition of bFGF to ECs onplastic had no significant effect on the accumulation of thestable degradation products, nitrite and nitrate. However,under angiogenic conditions (ie, ECs grown on fibrin matri-ces), NO production was significantly increased by bFGF inboth cell types. Addition of L-NAME (1 mmol/L) to theculture medium significantly attenuated nitrite accumulationin response to bFGF.

The effect of various manipulations of NO production onrates of EC proliferation is shown in Figure 4. Under controlconditions in the presence of FBS, both CPAECs andHUVECs demonstrated relatively high basal levels of thymi-dine incorporation, which were further increased by theaddition of bFGF (Figure 4A). Inhibition of endogenous NOproduction using L-NAME also caused a slight increase inproliferation rates, which achieved statistical significance forCPAECs. Identical results were obtained for L-NMMA (datanot shown). In contrast, the pharmacological generation ofNO using SNAP produced a dose-dependent inhibition of

Figure 3. Effects of bFGF on NO release from HUVECs andCPAECs. Cells were incubated for 48 hours on plastic or fibrinmatrices with phenol-free medium in the presence or absenceof bFGF (30 ng/mL) and L-NAME (1 mmol/L). After the incuba-tion, medium was collected and stored at 220°C until assay.Nitrite concentration in the medium was determined asdescribed in “Materials and Methods.” Values are presented asmean6SD of three different experiments.

Figure 2. Alterations in endothelial gene expression were evalu-ated by Northern analysis. A, Representative blots for eNOS (4.2kb) compared with GAPDH (1.7 kb) to correct for differences inRNA loading. Each lane was loaded with 20 mg of total RNAfrom HUVECs or CPAECs, grown on plastic or fibrin (ie, nonan-giogenic and angiogenic conditions, respectively) in the pres-ence or absence of bFGF (30 ng/mL) for 48 hours as indicated.B, Relative density of bands from ECs on fibrin (lanes 3 and 4,highlighted in the box) normalized for GAPDH and expressed asa ratio of the density of bFGF vs control for HUVECs (shadedbars) and CPAECs (solid bars) (mean6SD for threeexperiments).

Figure 4. A, The effect of bFGF and various manipulations of theNO pathway on [3H]thymidine incorporation in ECs cultured onplastic dishes was determined in CPAECs (shaded bars, left ordi-nate) and HUVECs (solid bars, right ordinate). Cells were culturedon plastic in the presence or absence of bFGF (30 ng/mL),L-NAME (1 mmol/L), and SNAP (0.1 and 0.4 mmol/L) for 24 hours.B, BrdU labeling is shown for CPAECs cultured for 48 hours in thepresence or absence of test agents as described in panel A.

1010 Nitric Oxide in Angiogenesis

by guest on April 3, 2014http://circres.ahajournals.org/Downloaded from

proliferation in both cell lines. Identical results were observedin CPAECs using BrdU nuclear staining as a marker of cellsengaged in the DNA synthesis (Figure 4B). A direct cytotoxiceffect of the NO-donor compounds on EC viability was ruledout by in vitro experiments showing no difference in trypanblue exclusion between treated and control cells after 48hours of incubation at the concentrations shown in Table 2and Figure 6.

Figure 5 shows representative fields from CPAECs (panelsa through d) and HUVECs (panels e through h) depicting theeffect of altering the production of NO on EC differentiationin the fibrin matrix model using a similar pharmacologicalapproach. Compared with cells cultured under control condi-tions (panels a and e), the addition of bFGF resulted in theextensive formation of capillary-like structures (panels b and

f). In the absence of bFGF, the exposure of cells to theNO-donor compound, SNAP, resulted in identical morpho-logical changes (panels c and g), whereas the inhibition ofendogenous NO production by L-NAME (1 mmol/L) (panelsd and h) partially prevented EC differentiation in response tobFGF in both cell types. Summary data for three experimentsare shown in Figure 6. The differentiation index and tubelength were very low for both HUVECs (Figure 6A) andCPAECs (Figure 6B) grown on fibrin under control condi-tions. Addition of bFGF produced significant increases indifferentiation index and tube length in both cell types,although the magnitude of differentiation was greater forbovine than for human ECs. Inhibition of endogenous NOproduction using L-NAME significantly reduced differentia-tion in response to bFGF. The inhibition of EC differentiation

Figure 5. Photomicrographs of CPAECs(a through d) and HUVECs (e through h)cultured on fibrin gels for 48 hours in theabsence of angiogenic factors (a and e),in the presence of bFGF (30 ng/mL, band f) or SNAP (0.4 mmol/L, c and g), orwith the combination of bFGF (30 ng/mL)and L-NAME (1 mmol/L) (d and h). Origi-nal magnification 3100.

Babaei et al May 18, 1998 1011

by guest on April 3, 2014http://circres.ahajournals.org/Downloaded from

in response to bFGF by L-NAME was dose dependent (Table1) and nearly complete at the higher concentrations of theinhibitors. In contrast, D-NAME (1 mmol/L) had no effect onthe angiogenic response to bFGF, suggesting that the effectsof L-NAME were related to the inhibition of NO production.Similar results were obtained using another inhibitor of NOsynthase, L-NMMA (1 mmol/L) (Table 1). Addition ofexogenous NO (ie, SNAP [Figure 6] or SNP, SIN-1, andGSNO [Table 2]) produced a dose-dependent increase inindices of EC differentiation to levels not different from thoseobserved in response to bFGF.

The effect of the angiogenic mediators on expression ofbFGF mRNA was studied in HUVECs by RT-PCR. Bands ofthe expected size were obtained for bFGF (179 nt) andGAPDH (983 nt) as shown in Figure 7. The density of thebands normalized for GAPDH was not significantly alteredrelative to control by exposure to bFGF (30 ng/mL) andSNAP (0.4 mmol/L) or the combination. This experiment wasrepeated twice with an identical result.

DiscussionIn the present study, bFGF produced differentiation of ECscultured on three-dimensional fibrin matrices into tube-likestructures while increasing the expression of eNOS mRNAand the production of NO. The pharmacological generation ofNO inhibited the proliferation of ECs on plastic and promoteddifferentiation of ECs in the fibrin matrix model. Moreover,inhibition of endogenous NO synthesis significantly reducedEC differentiation in response to bFGF in this model. Theseresults are compatible with a central role of endogenous NOin angiogenesis.

Angiogenesis is a highly complex process that involves atleast two distinct phases.1,4 The initial phase is characterizedby activation of quiescent ECs to a proliferative and migra-tory phenotype.1,5,6 Among the mediators of this transition areinflammatory cytokines such as tumor necrosis factor-a andinterleukin 823 and, possibly, soluble adhesion molecules suchas E-selectin.6 Growth factors such as bFGF and VEGF arealso potent stimulators of EC proliferation and migration1–3

and likely contribute to EC activation in the initial stages ofthe angiogenic response. A later phase of new vessel forma-tion involves the redifferentiation of the migrating andproliferating ECs into vascular tubes. This necessitates areversal of EC phenotype, back to a more quiescent statetypical of mature vascular structures.1 It is perhaps somewhatof an enigma that the same mediator, ie, bFGF, could beinvolved in distinct phases of the angiogenic process, whichappear to require opposite actions. The present data suggestthat NO, produced in response to bFGF, may act as amolecular “switch,” counteracting the growth-promoting ac-tions of this angiogenic mediator and initiating a program ofEC differentiation. This is in agreement with very similarobservations in a neuroblastoma cell line24 in which NO wasreported to mediate termination of the proliferative responseto nerve growth factor.

Figure 6. The effect of bFGF and various manipulations of theNO pathway on the differentiation of ECs grown on fibrin gels,HUVECs (A), and CPAECs (B). EC differentiation was quantifiedby both the total length of tubelike structures per low-powerfield (LPF) (solid bars, ordinates on right) and the differentiationindex (shaded bars, ordinates on left), as described in “Materialsand Methods.” Bars represent mean6SD for three separateexperiments.

TABLE 1. Effect of NOS Inhibitors on EC Differentiation in Response to bFGF

HUVECs CPAECs

DI TL, mm/LPF DI TL, mm/LPF

bFGF (30 ng/mL) 18.2963.91 2.8760.44 18.1963.75 3.6460.80

bFGF1D-NAME (1 mmol/L) 15.7563.51 2.7060.12 16.5561.13 3.5860.12

bFGF1L-NAME (0.3 mmol/L) 11.4060.67 2.3160.36 12.4762.57 3.2160.40

bFGF1L-NAME (1 mmol/L) 7.0060.94* 1.7260.31 6.6360.94* 2.5360.49

bFGF1L-NAME (3 mmol/L) 3.3760.91* 1.0460.28* 4.9061.29* 1.8560.61*

bFGF1L-NMMA (1 mmol/L) 6.4961.67* 1.5160.38* 5.5061.77* 1.6860.58*

DI indicates differentiation index; TL, tube length; and LPF, low-power field (3100). Values are mean6SD of threeseparate experiments performed under identical conditions.

*P,0.05 vs bFGF alone.

1012 Nitric Oxide in Angiogenesis

by guest on April 3, 2014http://circres.ahajournals.org/Downloaded from

In the present study, we did not see changes in eNOSmRNA expression in response to bFGF in ECs grown onplastic (Figure 2), although another group has reported thatbFGF increased eNOS expression in these conditions.25 Incontrast, in the angiogenic conditions with ECs cultured on afibrin matrix, bFGF did increase the expression of eNOS andthe production of NO in differentiating ECs. Under theseconditions, inhibition of NOS activity prevented tube forma-tion, confirming the importance of increased endogenous NOproduction in mediating the angiogenic response to bFGF.Interestingly, the fibrin matrix alone (ie, in the absence ofbFGF) reduced basal eNOS expression. This is in keepingwith a further “activation” of ECs grown on the fibrin matrix,which may be an important element of this in vitro model,possibly mediated by matrix–adhesion molecule interactions.Indeed, specific interaction of endothelialav/b3 integrin withfibrin may be critical for EC differentiation in this model.26

We ruled out the possibility that the very low concentrationsof thrombin used to polymerize the fibrinogen solution haddirect effects on eNOS expression, since in separate experi-ments, thrombin at these concentrations had no effect ofendothelial factor gene expression. Although the highest

quality fibrinogen was used in the present experiments, it ispossible that there was low-level contamination of the fibrin-ogen solution with cytokines. We have previously demon-strated that tumor necrosis factor-a, even at concentrations aslow as 10 U/mL, could profoundly downregulate expressionof this gene19 by a posttranscriptional mechanism involvingthe destabilization of eNOS mRNA. Despite the lower levelsof eNOS mRNA expression in cells cultured on fibrinmatrices, there was little change in basal NO production,suggesting an increased level of NOS activity under theseconditions.

Previous studies of the role of NO in the angiogenicresponse have provided conflicting results. Ziche et al15 havereported that NO mediated the angiogenic effects of sub-stance P in the rat cornea. In another report, tumor cellstransfected with inducible NOS grew more slowly in vitro butexhibited increased metastatic spread in vivo,27 associatedwith evidence of increased tumor vascularity. However, it hasalso been reported that the NO-donor compound SNP inhib-ited angiogenesis in the chick chorioallantoic membrane16,28

and reduced vascular tube formation of HUVECs grown onMatrigel.16 One can only speculate about the reasons for theseapparent discrepancies. Certainly, there may be importantdifferences between the Matrigel and fibrin matrix models ofangiogenesis. ECs grown on Matrigel form cords spontane-ously; this occurrence is likely due to contamination of thetumor basement membrane components with cytokines andgrowth factors.14,29 In contrast, in the fibrin gel model used inthe present study, exogenous angiogenic factors must beadded for full EC differentiation to be apparent. Furthermore,in fibrin gels it has been shown that ECs form tubes withidentifiable lumina,14,30 and this process requires new geneexpression,14 which does not always appear to be the case inthe Matrigel model.14

However, it is possible that the apparent disagreement maybe related not only to differences in the various models of

TABLE 2. Effect of NO Donors on Capillary-like Tube Formation in ECs Cultured onFibrin Matrices

HUVECs CPAECs

DI TL, mm/LPF DI TL, mm/LPF

Control 1.1660.19 0.4660.03 2.0760.87 0.7660.25

SNP

0.1 mmol/L 4.0860.40 1.5260.23* 7.6160.32 2.3960.11*

1 mmol/L 7.1761.41* 2.0360.44† 10.6960.32* 2.7860.23*

10 mmol/L 16.1962.84† 2.9360.07† 14.6261.05* 3.6560.50*

SIN-1

0.1 mmol/L ND ND 5.0760.65 1.8760.19

1 mmol/L ND ND 6.2261.07 1.8860.22

10 mmol/L ND ND 9.8160.92* 2.2960.14*

GSNO

1 mmol/L ND ND 6.6060.82 2.0960.19

10 mmol/L ND ND 10.5961.08* 2.7060.32*

100 mmol/L ND ND 13.2161.40* 3.2460.57*

DI indicates differentiation index; TL, tube length; LPF, low-power field (3100); and ND, not done. Values aremean6SD of three separate experiments.

*P,0.05 and †P,0.001 vs control.

Figure 7. The effects of exogenous NO production on bFGFexpression was investigated by RT-PCR. HUVECs wereexposed to SNAP (0.4 mmol/L) or bFGF (30 ng/mL) singly or incombination for 48 hours under angiogenic conditions (ie, onfibrin matrices), and bFGF mRNA expression was quantified byRT-PCR analysis.

Babaei et al May 18, 1998 1013

by guest on April 3, 2014http://circres.ahajournals.org/Downloaded from

angiogenesis but also to differences in the actions of NO,depending on the preexisting conditions. Thus, the directeffect of NO in most cell systems is that of growth inhibition,as was observed in both HUVECs and CPAECs grown in theabsence of the fibrin matrix. In the early phase of angiogen-esis, which is characterized by EC activation, this effect ofNO could reduce the initial proliferative response. However,during the later stages of the angiogenic process, the presentdata would suggest that increased production of NO may bea differentiating signal acting to terminate the proliferation ofactivated ECs and facilitate the initiation of a program of ECdifferentiation. Recently, Papapetropoulos et al31 have dem-onstrated that inhibitors of NOS largely prevented differen-tiation of ECs in response to another angiogenic factor,transforming growth factor-b, in a collagen matrix model ofangiogenesis in which EC proliferation was greatly inhibited.These findings are in agreement with those of the presentstudy and suggest a general role of NO in the formation ofvascular tubes in different angiogenic models and in responseto different angiogenic factors.

Of particular interest are the observations of Ziche’sgroup15,32,33 suggesting that rather than inhibiting EC prolif-eration, NO might actually mediate the mitogenic effects ofVEGF on postcapillary coronary ECs. Again, this is incontradistinction to the present study, which found that NOinhibited EC proliferation in response in two different ECtypes, using two different methods to quantify mitogenicactivity. It is possible that NO may have opposite effects inmicrovascular compared with macrovascular ECs. However,others have reported consistent growth inhibition by a varietyof NO-donor compounds in both macrovascular34,35 andmicrovascular34 EC lines. In coronary venular microvascularECs, Ziche et al36 have recently reported that the proliferativeeffects of NO are indirect, mediated by increased expressionof bFGF. However, such a mechanism cannot account for theangiogenic effects of NO described in the present report, asthere was no increase in bFGF expression induced by SNAP.Moreover, inhibitors of NOS reduced EC differentiation inresponse to exogenous bFGF, suggesting that in our systemNO mediated the effects of this angiogenic growth factorrather than the other way around. The same group has alsorecently reported data suggesting that NO may mediate theangiogenic response to VEGF but not bFGF.37 However,these results are not directly comparable to those of thepresent report. Ziche et al37 studied proliferation and migra-tion of coronary venular ECs cultured on plastic, which likelyreflect events of the early stages of angiogenesis involvingEC activation, whereas in the present study we quantified ECvascular tube formation in the three-dimensional fibrin gels,which is more representative of the later stages of theangiogenic response.

Taken together, our results are consistent with the conclu-sion that increased production of NO might act as a crucialmolecular “signal” in the angiogenic response to bFGF,terminating EC proliferation and initiating the formation ofvascular tubes in the fibrin matrix model of angiogenesis. Theobservations that capillary-like tube formation in our modelcan be largely prevented by inhibition of endothelial NOSactivity and that exogenous NO from NO donor compounds

reproduced the angiogenic effects on bFGF strongly supporta central role of NO in the initiation of EC differentiation.

AcknowledgmentsThis study was supported by a grant from the Medical ResearchCouncil of Canada. Dr Stewart is the Dexter Man Chair of Cardiol-ogy, University of Toronto. Dr Bendeck is a Scholar of the Heart andStroke Foundation of Canada.

References1. Folkman J. Seminars in medicine of the Beth Israel Hospital, Boston:

clinical applications of research on angiogenesis.N Engl J Med.1995;333:1757–1763.

2. Battegay EJ. Angiogenesis: mechanistic insights, neovascular diseases,and therapeutic prospects.J Mol Med.1995;73:333–346.

3. Folkman J. Angiogenesis in cancer, vascular, rheumatoid and otherdisease.Nat Med.1995;1:27–31.

4. Zimrin AB, Maciag T. Progress towards a unifying hypothesis for angio-genesis [editorial].J Clin Invest.1996;97:1359.

5. Ferrara N. Leukocyte adhesion: missing link in angiogenesis [news,comment].Nature.1995;376:467.

6. Koch AE, Halloran MM, Haskell CJ, Shah MR, Polverini PJ. Angio-genesis mediated by soluble forms of E-selectin and vascular celladhesion molecule-1.Nature. 1995;376:517–519.

7. Ferrara N. The role of vascular endothelial growth factor in pathologicalangiogenesis.Breast Cancer Res Treat.1995;36:127–137.

8. Moncada S, Palmer RMJ, Higgs EA. Nitric oxide: physiology, patho-physiology, and pharmacology.Pharmacol Rev.1991;43:109–142.

9. Palmer RMJ, Ashton DS, Moncada S. Vascular endothelial cells syn-thesize nitric oxide from L-arginine.Nature.1988;333:664–666.

10. Stuehr DJ. Purification and properties of nitric oxide synthases.MethodsEnzymol.1996;268:324–333.

11. Forstermann U, Kleinert H. Nitric oxide synthase: expression and expres-sional control of the three isoforms.Naunyn Schmiedebergs ArchPharmacol.1995;352:351–364.

12. Wu KK, Thiagarajan P. Role of endothelium in thrombosis and hemo-stasis.Annu Rev Med.1996;47:315–331.

13. Lloyd-Jones DM, Bloch KD. The vascular biology of nitric oxide and itsrole in atherogenesis.Annu Rev Med.1996;47:365–375.

14. Zimrin AB, Villeponteau B, Maciag T. Models of in vitro angiogenesis:endothelial cell differentiation on fibrin but not Matrigel is transcrip-tionally dependent.Biochem Biophys Res Commun.1995;213:630–638.

15. Ziche M, Morbidelli L, Masini E, Amerini S, Granger HJ, Maggi CA,Geppetti P, Ledda F. Nitric oxide mediates angiogenesis in vivo andendothelial cell growth and migration in vitro promoted by substance P.J Clin Invest.1994;94:2036–2044.

16. Pipili-Synetos E, Sakkoula E, Haralabopoulos G, Andriopoulou P,Peristeris P, Maragoudakis ME. Evidence that nitric oxide is an endog-enous antiangiogenic mediator.Br J Pharmacol.1994;111:894–902.

17. Jaffe EA, Nachman RL, Becker CG, Minick CR. Culture of humanendothelial cells derived from umbilical veins.J Clin Invest.1973;52:2745–2756.

18. Chomczynski P, Sacchi N. Single-step method of RNA isolation by acidguanidinium thiocyanate-phenol-chloroform extraction.Anal Biochem.1987;162:156–159.

19. Mohamed F, Monge JC, Gordon A, Cernacek P, Blais D, Stewart DJ.Lack of role for nitric oxide (NO) in the selective destabilization ofendothelial NO synthase mRNA by tumor necrosis factor.ArteriosclerThromb.1995;15:52–57.

20. Teichert-Kuliszewska K, Hamilton BS, Deitel M, Roncari DAK.Decreasing expression of a gene coding for a protein related to basicfibroblast growth factor during differentiation of human preadipocytes.Biochem Cell Biol.1994;72:54–57.

21. Teichert-Kuliszewska K, Hamilton BS, Deitel M, Roncari DAK. Aug-mented production of heparin binding mitogenic proteins by preadi-pocytes from massively obese persons.J Clin Invest. 1992;90:1226–1231.

22. Greenstein LA, Nissley SP, Moses AC, Short PA, Yang W, Lee L,Rechler MM. Purification of multiplication-stimulating activity. In:Barnes DW, Sirbasku DA, Sato GH, eds.Methods of Preparation ofMedia, Supplements and Substrata for Serum Free Animal Cell Culture.New York, NY: Alan R Liss Inc; 1984:111–138.

1014 Nitric Oxide in Angiogenesis

by guest on April 3, 2014http://circres.ahajournals.org/Downloaded from

23. Norrby K. Interleukin-8 andde novomammalian angiogenesis.CellProlif. 1996;29:315–323.

24. Peunova N, Enikolopov G. Nitric oxide triggers a switch to growth arrestduring differentiation of neuronal cells.Nature.1995;375:68–73.

25. Kostyk SK, Kourembanas S, Wheeler EL, Medeiros D, McQuillan LP,D’Amore PA, Braunhut SJ. Basic fibroblast growth factor increases nitricoxide synthase production in bovine endothelial cells.Am J Physiol.1995;269:H1583–H1589.

26. Stromblad S, Cheresh DA. Cell adhesion and angiogenesis.Trends CellBiol. 1996;6:462–468.

27. Jenkins DC, Charles IG, Thomsen LL, Moss DW, Holmes LS, Baylis SA,Rhodes P, Westmore K, Emson PC, Moncada S. Roles of nitric oxide intumor growth.Proc Natl Acad Sci U S A.1995;92:4392–4396.

28. Pipili-Synetos E, Sakkoula E, Maragoudakis ME. Nitric oxide is involvedin the regulation of angiogenesis.Br J Pharmacol.1993;108:855–857.

29. Baaout S. Endothelial differentiation using matrigel.Anticancer Res.1997;17:451–456.

30. Koolwijk P, van Erck MGM, de Vree WJA, Vermer MA, Weich HA,Hanemaaijer R, Van Hinsbergh VWM. Cooperative effect of TNFa,bFGF, and VEGF on the formation of tubular structures of human micro-vascular endothelial cells in a fibrin matrix: role of urokinase activity.J Cell Biol. 1996;132:1177–1188.

31. Papapetropoulos A, Desai KM, Rudic RD, Mayer B, Zhang R, Ruiz-Torres MP, Garcia-Cardena G, Madri JA, Sessa WC. Nitric oxide

synthase inhibitors attenuate transforming-growth-factor-b-stimulatedcapillary organizationin vitro. Am J Pathol.1997;150:1835–1844.

32. Ziche M, Morbidelli L, Masini E, Granger H, Geppetti P, Ledda F. Nitricoxide promotes DNA synthesis and cyclic GMP formation in endothelialcells from postcapillary venules.Biochem Biophys Res Commun.1993;192:1198–1203.

33. Morbidelli L, Chang CH, Douglas JG, Granger HJ, Ledda F, Ziche M.Nitric oxide mediates mitogenic effect of VEGF on coronary venularendothelium.Am J Physiol.1996;270:H411–H415.

34. Gooch KJ, Dangler CA, Frangos JA. Exogenous, basal, and flow-inducednitric oxide production and endothelial cell proliferation.J Cell Physiol.1997;171:252–258.

35. Lopez Farre A, De Miguel LS, Caramelo C, Gomez-Macias J, Garcia R,Mosquera JR, De Frutos T, Millas I, Rivas F, Echezarreta G, Casado S.Role of nitric oxide in autocrine control of growth and apoptosis ofendothelial cells.Am J Physiol.1997;272:H760–H768.

36. Ziche M, Parenti A, Ledda F, Dell’Era P, Granger HJ, Maggi CA, PrestaM. Nitric oxide promotes proliferation and plasminogen activator pro-duction by coronary venular endothelium through endogenous bFGF.Circ Res.1997;80:845–852.

37. Ziche M, Morbidelli L, Choudhuri R, Zhang H-T, Donnini S, Granger HJ.Nitric oxide synthase lies downstream from vascular endothelial growthfactor-induced but not basic fibroblast growth factor-induced angio-genesis.J Clin Invest.1997;99:2625–2634.

Babaei et al May 18, 1998 1015

by guest on April 3, 2014http://circres.ahajournals.org/Downloaded from