Marked Induction of the IAP Family Antiapoptotic Proteins Survivin and XIAP by VEGF in Vascular Endothelial Cells

MPi

JR*24aC

R

teaccwaatPaoneip(edsGsgppvpi©

5

Biochemical and Biophysical Research Communications 264, 781–788 (1999)

Article ID bbrc.1999.1589, available online at http://www.idealibrary.com on

arked Induction of the IAP Family Antiapoptoticroteins Survivin and XIAP by VEGF

n Vascular Endothelial Cells

ennifer Tran,* Janusz Rak,* Capucine Sheehan,* Samuel D. Saibil,* Eric LaCasse,†obert G. Korneluk,†,‡ and Robert S. Kerbel*,1

Division of Cancer Biology Research, Sunnybrook and Women’s College Health Sciences Centre, S-218 Research Building,075 Bayview Avenue, Toronto, Ontario M4N 3M5, Canada; †Apoptogen Inc., Suite R306, CHEO Research Institute,01 Smyth Road, Ottawa, Ontario K1H 8L1, Canada; and ‡Department of Biochemistry, Microbiology,nd Immunology, University of Ottawa and C. Karsh Molecular Genetics Laboratory,hildren’s Hospital of Eastern Ontario Research Institute, Ontario, Canada

eceived August 30, 1999

cwhafiaw(vcttflci

icItmsdvvwSmtwed(

Vascular endothelial growth factor (VEGF) is a po-ent angiogenic factor that has been shown to act as anndothelial cell mitogen as well as a vascular perme-bility factor. Several recent reports have also impli-ated VEGF as a major survival factor for endothelialells during angiogenesis and vasculogenesis alongith other growth factors such as bFGF andngiopoietin-1. VEGF has been shown to mediate thisdditional function, at least in part through the induc-ion of bcl-2 and the activation of the PI3 kinase-Akt/KB signaling pathway. We report here that VEGF canlso mediate the induction/upregulation of membersf a newly discovered family of antiapoptotic proteins,amely the Inhibitors of Apoptosis (IAP), in vascularndothelial cells. We show that VEGF165 leads to thenduction of XIAP (2.9-fold) and survivin (19.1-fold)rotein in human umbilical vein endothelial cellsHUVECs). In contrast, bFGF had little effect on XIAPxpression, but produced approximately a 10-fold in-uction on survivin. VEGF-dependent upregulation ofurvivin could be prevented by cell cycle arrest in the1 and S phases. These findings implicate that the

urvival and mitotic functions of VEGF in an angio-enic context may be more intrinsically related thanreviously anticipated. Moreover, they also raise theossibility of therapeutically targeting XIAP or sur-ivin in antiangiogenic therapy as a means of sup-ressing tumor growth, in addition to directly target-

ng tumor cells which express these survival proteins.1999 Academic Press

1 To whom correspondence should be addressed. Fax: (416) 480-703. E-mail: [email protected].

781

Tumor angiogenesis is thought to result from ahange in the net balance of the local concentrationithin tumor masses of endogenous angiogenesis in-ibitors and stimulators (1). Among the best known,nd characterized, angiogenesis stimulators are basicbroblast growth factor (bFGF), angiopoietin-1 (ang-1)nd vascular endothelial cell growth factor (VEGF),hich is also known as vascular permeability factor

VPF) (2, 3). VEGF is expressed at high levels by theast majority of human and animal cancers, both byancer cells themselves, and at least in some cases, byumor infiltrating normal stromal cells (2, 3). In con-rast, the VEGF receptor tyrosine kinases known ask-1/KDR and flt-1, are expressed by the endothelialells of newly formed, immature blood vessel capillar-es, including such vessels found in tumors (3).

Activation of VEGF receptors, especially KDR/flk-1,nduces a number of phenotypic responses in endothelialells which contribute to the angiogenic phenotype (3).nitially one of the most important of these was thoughto be mitogenesis, as the term “VEGF” denotes. However,ore recent results have implicated VEGF as a major

urvival (anti-apoptotic) factor for the newly formed en-othelial cells found in immature, newly formed bloodessels (4–6). This anti-apoptotic function, along withascular permeability, are now believed to be the majorays in which VEGF contributes to angiogenesis (4–6).uch findings raise the obvious question of how VEGFediates its prosurvival/antiapoptotic function in endo-

helial cells. Several possibilities have been put forward,hich are not necessarily exclusive. For example, Gerber

t al. and Carmeliet et al. have reported evidence that itoes so through the phosphatidyl-inositol 39 kinaseP13K)/Akt signal transduction pathway (7). In addition,

0006-291X/99 $30.00Copyright © 1999 by Academic PressAll rights of reproduction in any form reserved.

NccVv(slat

lac1(at1mhotouic9

aXtai(VtIpd

iebvhvtpp

M

lv

for experiments. Tissue culture dishes (Nunc, Kamstrup/Roskilde,DLwM(notsiiuam

wtc

gmtMM2cwfmmPwtrrlpHm(tpColdQsHftaarp

cuuaan

eW

Vol. 264, No. 3, 1999 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

or et al. (8) and Gerber et al. (9) have found that VEGFan induce high levels of bcl-2 in vascular endothelialells, while Gupta et al. have reported evidence thatEGF suppresses apoptosis in human micro- or macro-ascular cells via induction and suppression of the MAPKp42)/ERK (p44) and SAPK/JNK signaling pathways re-pectively (10). Moreover the VEGF-mediated endothe-ial survival and angiogenesis may be dependent on thedhesion molecule VE (vascular endothelial)-cadherinhrough both Akt kinase and bcl-2 (7).

In addition to the bcl-2 family of apoptosis regu-ators, a new and different family of modulators ofpoptosis has been recently discovered, which arealled the “IAPs”—inhibitors of apoptosis proteins (11,2). They include neuronal apoptosis inhibitory proteinNAIP), X-chromosome linked IAP (XIAP) and HIAP-1nd HIAP-2 (human IAP-1 and -2). Also included inhis family is the newly discovered “survivin” gene (13,4). The IAPs have in common two defining structuralotifs, one being the so-called “BIR” (Baculovirus In-ibitor of apoptosis Repeat) domains (15) and the sec-nd being a RING zinc finger domain near the carboxyerminus (12, 15, 16). Whereas anti-apoptotic membersf the bcl-2 family, such as bcl-2 itself or bcl-XL, blockpstream caspases, the IAPs and survivin appear to

nhibit apoptosis by blocking the terminal, effectoraspases such as caspase 3 and 7 as well as procaspase(12, 15, 16).Given the potency of the IAPs as inhibitors of

poptosis (12, 15, 16), and recent reports showingIAP can be induced in endothelial cells, and protect

hem from tumor necrosis factor-alpha (TNFa)-inducedpoptosis (12, 16), and that inflammatory cytokines cannduce HIAP-1 in human umbilical endothelial cells17), we decided to address the question of whetherEGF might be an inducer of one or more members of

he IAP family of proteins in vascular endothelial cells.f so, such a relationship could represent yet anotherotential survival mechanism that such cells utilizeuring the process of new blood vessel formation.The main purpose of the present study was to exam-

ne the effects of VEGF, in comparison to bFGF, on thexpression of bcl-2, XIAP, and survivin, in human um-ilical vein endothelial cells (HUVECs) grown underarious conditions in cell culture. We show that VEGFas a powerful inductive effect on both XIAP and sur-ivin protein expression, especially the latter. As such,he possibility is raised that induced or targeted sup-ression of these proteins might be considered as aotential antiangiogenic therapeutic strategy.

ATERIALS AND METHODS

Cell lines and culture conditions. Human umbilical vein endothe-ial cells (HUVECs) were purchased from Clonetics Corp. (Walkers-ille, MD) and cells between the second and sixth passages were used

782

enmark) were precoated with 1% gelatin (Sigma Chemical Co., St.ouis, MO) for approximately a half hour in 37°C after which gelatinas removed. HUVECs were maintained as monolayer cultures inCDB131 base medium supplemented with 10% fetal bovine serum

FBS) (Gibco Canadian Life Technologies, Grand Island, NY), 1g/mL basic fibroblast growth factor (R & D Systems Inc., Minneap-lis, MN) and 10 units/mL heparin (Gibco BRL). On the day ofreatment, cells were repeatedly washed with phosphate-bufferedaline (PBS) to remove any residual growth factors. Cells were thenncubated for various time intervals in 5% FBS supplemented median the absence of growth factors or with either 1 ng/mL bFGF and 10nits/mL heparin or 50 ng/mL VEGF (R & D Systems Inc., Minne-polis, MN). These conditions were derived through a series of opti-ization experiments.

Cell synchronization. HUVECs were treated overnight at 37°Cith 400 mM mimosine (Biomol, Plymouth Meeting, PA), 2 mM

hymidine (Sigma) or 0.4 mg/mL nocodazole (Biomol) to induce a cellycle arrest in G1, S, or G2/M phase, respectively (18).

Survival, proliferation, and apoptosis assays. The effect of angio-enic growth factors on HUVEC viability was assessed by measure-ents of cumulative mitogenic activity ([3H]thymidine incorpora-

ion), by cumulative cell survival in the MTS assay (Promega,adison, WI) and the rate of apoptotic cell death (cell death ELISA).TS assay was conducted by using the Celltiter 96 AQ plate. Briefly,

4 h prior treatment 6000 cells per well were seeded in 96-wellulture plates (Nunc). Upon overnight attachment, cells wereashed with PBS and were then treated with or without growth

actors for the indicated time periods. On the day of the assay, aixture (30:1) of MTS (3-(4,5-dimethylthiazol-2yl)-5-(3-carboxy-ethoxyphenyl)-2-(4-sulfophenyl)2H-tetrazolium) (Promega) andMS (phenazine methosulfate) (Sigma) was added to each to eachell. MTS is bioreduced by viable cells into a brown colored formazan

hat is soluble in tissue culture medium. The absorbance of thiseaction product was then read at 490 nm on the Bio-Rad ELISAeader (Bio-Rad, Hercules, CA) and expressed as relative cell cumu-ative viability. To assay for cell proliferation, [3H]thymidine incor-oration was measured following addition of bFGF and VEGF toUVECs as described previously. In brief, after indicated treat-ents with growth factors, 50 mL of 40 mCi/mL of [3H]thymidine

Amersham, Oakville, Ontario, Canada) was added to each well for 4o 6 h after which plates were either frozen at 220°C until furtherrocessing or immediately harvested onto dry filters using a Titertekell Harvester 530 (Flow Laboratories, Irvine Scotland). The amountf incorporated [3H]thymidine was then measured by a Beta Plateiquid scintillation counter (Wallac, Turku, Finland). The cell deathetection ELISA Plus Kit (Roche Molecular Biochemicals, Laval,uebec) was used to estimate apoptotic death by quantifying nucleo-

omes released into the cytoplasm by dying endothelial cells.UVECs grown in 5% FBS in the presence or absence of growth

actors were harvested, lysed and incubated with a mixture of bio-inylated antihistone antibody and peroxidase-conjugated anti-DNAntibody, both of which bind to histone DNA complexes and result incolor reaction in the presence of ABST substrate. The absorbance

eflective of the apoptotic rate was read at 405 nm and plotted as aercentage of control values obtained for untreated cells.

Antibodies. The rabbit polyclonal anti-bcl-2 antibody was pur-hased from Santa Cruz Biotechnology (Santa Cruz, CA) and wassed in 1:250 dilution in Western blotting. The anti-XIAP antibodysed at 1:2000 dilution and was generated as a rabbit polyclonalntiserum raised against a full-length XIAP GST fusion protein. Theffinity purified rabbit anti-survivin (Alpha Diagnostics, San Anto-io, TX) antibody was used at 1 mg/mL.

Western blot analyses. The effect of the various treatments onxpression of survival genes in endothelial cells was assessed byestern blotting. Briefly, cells were cultured in 100-mm dishes un-

der indicated conditions until they reached 60 to 80% confluence.CE5mmp(w((temmapignetD(tmt

ivPd(rRrfiS

R

lacttsofimbmVotstbTts

c(fviaaiwrc1pmeEau(pcrH

tsbtesnfontbrawVbVLeit4vsV1

ppa


ells were then harvested with a brief incubation in 10% Trypsin/DTA (Gibco). Cell pellets were then lysed in lysis buffer containing0 mM Hepes (pH 7.4), 150 mM NaCl, 1% IPEGAL, 1 mM EDTA, 20g/mL aprotinin/leupeptin, 1 mM PMSF and 1 mM Na3VO4 for 30in on ice. Lysates were centrifuged at 12,000g for 10 min at 4°C and

rotein concentrations were quantitated by Bio-Rad protein assayBio-Rad). Fifty micrograms of total protein lysate for each treatmentas resolved by SDS–PAGE (12% gel) under reducing conditions

100 mM DTT) against Rainbow molecular weight size markersAmersham Corp). Proteins were subsequently transferred to a ni-rocellulose filter overnight after which equal loading and transferfficiency were verified by staining with Amido black for 5 min. Theembranes were then blocked in 10% dry skim milk in TBS-T (20M Tris, 137 mM NaCl, 0.1% Tween 20) for 1–2 h at room temper-

ture and probed with primary antibodies for 1 h at room tem-erature. Following several washes in TBS-T, membranes werencubated for 1 h in horseradish peroxidase-conjugated secondaryoat anti-rabbit antibody at a dilution of 1:5000 (Jackson Immu-oResearch Laboratories, Westgrove, PA). Signals were detected bynhanced chemiluminescence (Amersham Corp.). Signal quantita-ion was performed by using a Personal Densitometer (Molecularynamics, Inc., Sunnyvale, CA) coupled to ImageQuant software

Molecular Dynamics, Inc.). The Amido black staining was also quan-ified similarly. The fold induction in signal intensity was deter-ined by normalizing the signal obtained by each antibody with

heir respective Amido black stain intensity.

FACS analysis. Cell cycle profile was determined by propidiumodide staining. Approximately 3 3 106 HUVECs subjected to thearious treatments described above were harvested and washed withBS. Cells were then fixed in 80% ethanol overnight. The followingay, cells were washed once with PBS, then with PI staining bufferPBS, 0.12% Triton X-100, 0.12 mM EDTA). HUVECs were thenesuspended in PI staining buffer containing 100 mg of DNase-freeNase for 45 min at 37°C. Cells were stained for 1 h in the dark atoom temperature with propidium iodide (50 mg/mL). Cell cycle pro-les were analyzed by FACScan flow cytometer (Becton–Dickinson,an Francisco, CA) and analyzed using CellFit software package.

ESULTS

The relative potency of VEGF and bFGF as endothe-ial survival factors. VEGF and bFGF are potent pro-ngiogenic factors, which can stimulate endothelialell proliferation (19). Both of these polypeptides, par-icularly VEGF, have also been implicated as media-ors of endothelial cell survival under a variety oftress conditions (5). To evaluate the relative potencyf VEGF and bFGF as endothelial survival factors, werst decided to use their pro-mitogenic effects on hu-an umbilical vein endothelial cells (HUVECs) as a

iological reference point. As shown in Fig. 1A, treat-ent of HUVECs with 1 ng/ml of bFGF or 50 ng/ml ofEGF165 produced a comparable rate of DNA synthesisver a wide range of exposure times (24–72 h). Underhese conditions, incorporation of [3H]thymidine byemiconfluent endothelial cell cultures constituted upo 200% of control values obtained without addition ofFGF or VEGF (5% fetal bovine serum (FBS) only).he same concentrations of the respective growth fac-ors were then selected to evaluate endothelial cellurvival (Figs. 1B and 1C). Thus, cumulative survival

783

urves of HUVEC cultures supplemented with bFGFFig. 1B) showed sustained metabolic activity over aour day incubation period (up to 200% of controlalues at 96 h). An even greater effect (300–400%ncrease in viability) was observed when VEGF wasdded to the medium (Fig. 1B). Results of the MTSssay were interpreted as a reflection of differentialmpact of bFGF and VEGF on cell survival since theyere in agreement with direct measurements of the

ate of apoptotic cell death (Fig. 1C) as opposed toorresponding results of cell proliferation assays (Fig.A). Release of nucleosomal complexes to cellular cyto-lasm, a hallmark of incipient apoptotic cell death wasarkedly reduced in HUVECs under the influence of

ither growth factor, as measured by a cell deathLISA (Fig. 1C). However, VEGF treatment producedgreater delay in the rate of cell death (12.1% of

ntreated cells after 48 h) than that induced by bFGF34.1% after 48 h) particularly at the earlier timeoints (Fig. 1C). Thus, despite an equivalent impact onell proliferation, VEGF and bFGF differed in theirelative potency as endothelial cell survival factors forUVECs in cell culture.

VEGF induces expression of the anti-apoptotic pro-eins bcl-2, XIAP, and survivin. VEGF has beenhown to mediate its survival function, at least in part,y induction of bcl-2 (8). However, the contribution ofhe IAP family of apoptosis inhibitors has not yet beenvaluated. We therefore decided to examine XIAP andurvivin expression using bcl-2 expression as an inter-al positive control. As above, HUVECs were trans-erred to 5% FCS supplemented media in the absencef growth factors, or in the presence of either bFGF (1g/mL) or VEGF (50 ng/mL). Over a 72-h time period,otal protein lysates were analyzed for expression ofcl-2 as well as members of the IAP family of apoptosisegulators such as XIAP and survivin (12, 16). Ingreement with previous reports (8) bcl-2 expressionas strongly upregulated (4.2-fold at 72 h) followingEGF treatment (Fig. 2A). In our hands, however,FGF (at a concentration that was “equimitotic” toEGF) was unable to increase the expression of bcl-2.ikewise, treatment with bFGF had little effect onxpression of XIAP, which was in fact down regulatedn the presence of this growth factor (Fig. 2B). In con-rast, XIAP expression was strongly induced within8–72 h by treatment with VEGF (2.9-fold over controlalues; Fig. 2B). Most strikingly, the expression ofurvivin was dramatically increased by both bFGF andEGF treatment, especially the latter (approximately0.3- and 19.1-fold, respectively; Fig. 2C).

Expression of survivin in HUVECs is cell cycle de-endent. As shown recently by Li et al., survivin ex-ression is regulated in a cell cycle dependent mannernd is believed to play a role in maintaining cellular

iowipgcodtpaeiilGppu

arptiodatastsa4Vopd

sVpta


ntegrity at the critical G2/M boundary (18). In light ofur aforementioned results an obvious question washether, similar to tumor cells, expression of survivin

n activated endothelial cells is regulated by cell cyclerogression. To address this point, HUVECs wererown in 10% FCS with or without addition of selectiveell cycle inhibitors (18) such as mimosine, thymidiner nocodazole. Such treatments were expected to pro-uce a growth arrest at G1, S, or G2/M phases respec-ively. Upon treatment cells were stained with pro-idium iodide to determine cell cycle profile andnalyzed for expression of survivin. Indeed, HUVECsxpressed the highest levels of survivin when arrestedn G2/M by nocodazole (39.52%), i.e., showed a 3.7-foldnduction of this anti-apoptotic effector at the proteinevel compared to cells arrested in G1 (18.24% cells in2/M; Fig. 3). These results suggest that survivin ex-ression in activated endothelial cells follows the samerinciple as in transformed cancer cells in that it ispregulated in G2/M phase of the cell cycle.

FIG. 1. VEGF and bFGF induce proliferation and survival of humynthesis in cultures of HUVECs treated for different lengths of timEGF, 50 ng/mL) as measured by [3H]thymidine incorporation. (Bresence of either VEGF or bFGF (same concentrations as in A)—MTreated with bFGF or VEGF. Apoptosis was quantitated by cell deathnd Methods). All values are expressed as percentage of control (un

784

Induction of survivin by VEGF is attenuated by G1rrest. The cell cycle specific regulation of survivinaises the question whether the action of VEGF on ex-ression of this protein is also linked to transitionhrough the G2/M boundary. It could be speculated thatf that was the case a pharmacological G1 arrest couldbliterate the effect of VEGF on survivin expression. In-eed, as shown in Fig. 4A, both mimosine (G1 blocker)nd thymidine (S phase inhibitor) strongly attenuatedhe induction of survivin expression by VEGF. Similar,lbeit somewhat less intense attenuation was also ob-erved in the context of bFGF treatment (Fig. 4A). Fur-hermore, inhibition of VEGF dependent upregulation ofurvivin in the presence of mimosine was associated withn increase in rate of apoptotic death of HUVECs (Fig.B). Thus, endothelial cell specific pro-survival effects ofEGF, at least in part, may be executed through increasef survivin levels at the transition through the G2/Mhase of the cell cycle and reduction in the rate of cell lossuring accompanying cell division.

umbilical vein endothelial cells in culture. (A) Similar rates of DNAith growth factors at “equimitotic” concentrations (bFGF, 1 ng/mL;ifferential cumulative cell survival in cultures of HUVECs in thessay. (C) Comparison of the rates of apoptotic cell death in HUVECsISA detecting release of nucleosomes to the cytoplasm (see Materialsted HUVECs).

ane w) DS aELtrea

D

aeh

i(adiB

twM


ISCUSSION

VEGF was originally described as a vascular perme-bility factor, called VPF (20) and as a highly specificndothelial cell mitogen (21). These two functionsave, until recently, been the major focus of attention

FIG. 2. Expression of anti-apoptotic proteins bcl-2, XIAP and surreated with growth factors (bFGF, 1 ng/mL, VEGF, 50 ng/mL) for inas evaluated by Western blotting. Signal intensities (bottom panelsaterials and Methods) and expressed as fold induction over untrea

785

n terms of its functional contributions to angiogenesis21). However, in 1995 Alon et al. reported VEGF couldlso function as a potent inhibitor of apoptosis of en-othelial cells comprising newly formed retinal vesselsn neonatal rats exposed to hypoxia (5). Subsequently,enjamin et al. (6) and Jain et al. (22) reported VEGF

in in endothelial cells treated with bFGF and VEGF. HUVECs wereted periods of time after which the expression of respective proteinsere quantitated by densitometry, normalized to loading control (seecontrol. (A) bcl-2 (B) XIAP (C) survivin.

vivdica) wted

ciaftgj(

miptlabsitIr

iaurcwrbsvfmVssJt(sass

c(tes(

litndFaqe


ould also protect endothelial cells of newly formedmmature tumor blood vessels. This survival/anti-poptotic function of VEGF is now considered crucialor its pro-angiogenic function, and may help explainhe observation of the profound deficiency of vasculo-enesis and angiogenesis in embryonic mice lackingust one VEGF allele, such that the mice die in utero23, 24).

The potency of VEGF to inhibit apoptosis and pro-ote survival of endothelial cells could be explained by

ts ability to activate not one, but several independentathways, each of which can suppress apoptosis. Inhis regard, our results implicate what could be (ateast) a third major mechanism to account for the anti-poptotic function of VEGF. The first is upregulation ofcl-2 (7–9), which our results have confirmed, and theecond is activation of the P13 kinase-Akt/PKB signal-ng pathway, which we did not examine (7, 25). Thehird is upregulation of at least two members of theAP gene family, namely XIAP and survivin. In thisegard, our Western blot results showed a rather strik-

FIG. 3. Cell cycle dependent expression of survivin in endothe-ial cells. (A) Western blotting analysis of survivin expressionn HUVECs exposed to cell cycle inhibitors: mimosine (G1 arrest),hymidine (S arrest) and nocodazole (G2/M arrest). The asynchro-ous population is also shown (no treatment). Cells subjected to eachrug treatment were stained with propidium iodide and analyzed byACScan for cell cycle profile. Percentage of cells at the G2/M bound-ry is shown. (B) Intensities of survivin signal in each sample wereuantitated by densitometry, normalized to the loading control andxpressed as fold induction as described previously.

786

ng upregulation of these two proteins—between 10-nd 20-fold—could be induced by VEGF in HUVECsnder the conditions employed, while bcl-2 was up-egulated by about 4-fold under the same conditions. Inontrast, no induction of XIAP by bFGF was observed,hereas survivin was induced by 10-fold. The VEGF

esults were not restricted to tissue culture as we haveeen able to detect pronounced survivin protein expres-ion in human endothelial cells of newly formed bloodessels under a variety of circumstances, e.g., in newlyormed vessels comprised of human microvascular der-

al endothelial cells growing under the influence ofEGF within a sponge implant in SCID mice (8), or inome vessels found within spontaneous human tumorpecimens, such as breast and bladder cancer (Tran,othy, Polverini, and Kerbel, unpublished observa-ions). While we have not yet proven that the survivinor XIAP) expressed by endothelial cells in such ves-els, or endothelial cells in culture, provides a survivaldvantage, the high levels induced that we have ob-erved along with their known potency to block apopto-is under many different conditions (12, 16) is clearly

FIG. 4. VEGF-dependent upregulation of survivin in endothelialells is dependent on transition through G2/M phase of the cell cycle.A) Arrest of HUVECs in G1 (mimosine) or S phase (thymidine) ofhe cell cycle prevents VEGF and bFGF from upregulating survivinxpression. (B) Induction of G1 arrest attenuates VEGF-dependenturvival of HUVECs as measured by the rate of nucleosomal releasecell death ELISA).

sVaanb(

lmnsatfhesgctcto

cswomrsbraccVtfd

tboipsaiai3sX

A

tNCNRH

R

1

1

1

1

1

111

1

12

22

2


uggestive of such a role, particularly in relation toEGF, and perhaps with respect to other pro-ngiogenic growth factors as well, e.g., bFGF orngiopoietin-1. In this regard, Karsten et al. haveoted that bFGF can inhibit endothelial cell apoptosisy both bcl-2 dependent and independent mechanisms26).

Our finding that survivin can be induced at highevels in endothelial cells is also of interest since hu-

an survivin has not yet been shown to be expressed inon-transformed cells. Indeed, previous reports haveuggested that human survivin, while absent in mostdult differentiated tissues, is detected in almost allransformed cell lines and cancers in vivo tested thusar, as well as in fetal tissue (13). Thus, survivin be-aves as an “oncofetal” protein. Our endothelial cellxpression results therefore raise the interesting pos-ibility that therapeutically targeting survivin at theene expression or protein expression/function levelsould result in suppression of angiogenesis. Survivinherefore becomes a potentially attractive target forancer therapy because of high levels of expression byumor cells and the tumor vasculature, but not byther normal cells.Another interesting aspect of these findings was the

ell cycle dependency of VEGF induced upregulation ofurvivin expression. Li et al. have suggested in theirork that survivin may counteract a default inductionf apoptosis in the G2/M phase by associating to theitotic spindle at the beginning of mitosis. In this

egard, it could be speculated that VEGF’ functions inurvival and mitogenesis of endothelial cells may note mutually exclusive. Indeed, survivin may be up-egulated at the G2/M interface in order to counteractn apoptotic signal and thereby allowing endothelialells to survive (i.e., VEGF as a survival factor) cellycle progression and to finally proceed to mitosis (i.e.,EGF as a mitogen). This could therefore represent a

ighter association between the mitotic and survivalunctions which VEGF contributes to sprouting vesselsuring the process of tumor angiogenesis.Given these results it will also be of interest to de-

ermine whether agents which block VEGF, e.g., anti-odies (27), or VEGF receptor function, e.g., antibodiesr small molecule drugs (28, 29), inhibit angiogenesis,n part, by the simultaneous downregulation of multi-le effectors of endothelial cell survival such as bcl-2,urvivin, and XIAP. Furthermore, certain other anti-ngiogenic agents or regulators which are known tonduce endothelial cell apoptosis, e.g., tubulin-bindinggents such as combretastatin-A4 (30), or endogenousnhibitors of angiogenesis such as angiopoietin-2 (31,2) may do so, at least in part, by interfering with theurvival functions of IAP proteins such as survivin andIAP expressed by such cells.

787

CKNOWLEDGMENTS

We thank Cassandra Cheng and Lynda Woodcock for their secre-arial assistance. This work was supported by a grant from theational Cancer Institute of Canada and the Medical Researchouncil to R. S. Kerbel. R. S. Kerbel is a Terry Fox Scientist of theational Cancer Institute supported by funds from the Terry Foxun, Canadian Cancer Society. Dr. R. G. Korneluk is a Howardughes Medical Institute International Research Scholar.

EFERENCES

1. Hanahan, D., and Folkman, J. (1996) Cell 86, 353–364.2. Klagsbrun, M., and Soker, S. (1993) Curr. Biol. 3, 699–702.3. Neufeld, G., Cohen, T., Gengrinovitch, S., and Poltorak, Z. (1999)

FASEB J. 13, 9–22.4. Benjamin, L. E., Golijanin, D., Itin, A., Pode, D., and Keshet, E.

(1999) J. Clin. Invest. 103, 159–165.5. Alon, T., Hemo, I., Itin, A., Pe’er, J., Stone, J., and Keshet, E.

(1995) Nature Med. 1, 1024–1028.6. Benjamin, L. E., and Keshet, E. (1997) Proc. Natl. Acad. Sci.

USA 94, 8761–8766.7. Carmeliet, P., Lampugnani, M. G., Moons, L., Breviario, F.,

Compernolle, V., Bono, F., Balconi, G., Spagnuolo, R., Oostuyse,B., Dewerchin, M., Zanetti, A., Angellilo, A., Mattot, V., Nuyens,D., Lutgens, E., Clotman, F., de Ruiter, M. C., Gittenberger-de,G. A., Poelmann, R., Lupu, F., Herbert, J. M., Collen, D., andDejana, E. (1999) Cell 98, 147–157.

8. Nor, J. E., Christensen, J., Mooney, D. J., and Polverini, P. J.(1999) Am. J. Pathol. 154, 375–381.

9. Gerber, H. P., Dixit, V., and Ferrara, N. (1998) J. Biol. Chem.273, 13313–13316.

0. Gupta, K., Kshirsagar, S., Li, W., Gui, L., Ramakrishnan, S.,Gupta, P., Law, P. Y., and Hebbel, R. P. (1999) Exp. Cell Res 247,495–504.

1. Liston, P., Roy, N., Tamai, K., Lefebvre, C., Baird, S., Cherton-Horvat, G., Farahani, R., McLean, M., Ikeda, J. E., MacKenzie,A., and Korneluk, R. G. (1996) Nature 379, 349–353.

2. LaCasse, E. C., Baird, S., Korneluk, R. G., and MacKenzie, A. E.(1998) Oncogene 17, 3247–3259.

3. Ambrosini, G., Adida, C., and Altieri, D. C. (1997) Nat. Med. 3,917–921.

4. Ambrosini, G., Adida, C., Sirugo, G., and Altieri, D. C. (1998)J. Biol. Chem. 273, 11177–11182.

5. Miller, L. K. (1999) Trends Cell Biol. 9, 323–328.6. Deveraux, Q. L., and Reed, J. C. (1999) Genes Dev. 13, 239–252.7. Horrevoets, A. J., Fontijn, R. D., van Zonneveld, A. J., de Vries,

C. J., ten Cate, J. W., and Pannekoek, H. (1999) Blood 93,3418–3431.

8. Li, F., Ambrosini, G., Chu, E. Y., Plescia, J., Tognin, S., Marchi-sio, P. C., and Altieri, D. C. (1998) Nature 396, 580–584.

9. Folkman, J. (1995) N. Engl. J. Med. 333, 1757–1763.0. Senger, D. R., Galli, S., Dvorak, A. M., Perruzzi, C. A., Harvey,

V. S., and Dvorak, H. F. (1983) Science 219, 983–985.1. Ferrara, N., and Davis-Smyth, T. (1997) Endocr. Rev. 18, 4–25.2. Jain, R. K., Safabakhsh, N., Sckell, A., Chen, Y., Jiang, P.,

Benjamin, L., Yuan, F., and Keshet, E. (1998) Proc. Natl. Acad.Sci. USA 95, 10820–10825.

3. Carmeliet, P., Ferreira, V., Breier, G., Pollefeyt, S., Kieckens, L.,Gertsenstein, M., Fahrig, M., Vandenhoeck, A., Harpal, K.,Eberhardt, C., Declercq, C., Pawling, J., Moons, L., Collen, D.,Risau, W., and Nagy, A. (1996) Nature 380, 435–439.

24. Ferrara, N., Carver-Moore, K., Chen, H., Dowd, M., Lu, L.,

2

2

2

2

29. Rockwell, P., Neufeld, G., Glassman, A., Caron, D., and Gold-

3

3

3


O’Shea, K. S., Powell-Braxton, L., Hillan, K. J., and Moore,M. W. (1996) Nature 380, 439–442.

5. Gerber, H. P., McMurtrey, A., Kowalski, J., Yan, M., Keyt, B. A.,Dixit, V., and Ferrara, N. (1998) J. Biol. Chem. 273, 30336–30343.

6. Karsan, A., Yee, E., Poirier, G. G., Zhou, P., Craig, R., andHarlan, J. M. (1997) Am. J. Pathol. 151, 1775–1784.

7. Kim, K. J., Li, B., Winer, J., Armanini, M., Gillett, N., Phillips,H. S., and Ferrera, N. (1993) Nature 362, 841–844.

8. Fong, T. A. T., Shawver, L. K., Tang, L. S. C., App, H., Powell,T. J., Kim, Y. H., Schreck, R., Wang, X., Risau, W., Ullrich, A.,Hirth, K. P., and McMahon, G. (1999) Cancer Res. 59, 99–106.

788

stein, N. (1995) Mol. Cell. Different. 3, 91–109.0. Iyer, S., Chaplin, D. J., Rosenthal, D. S., Boulares, A. H.,

Li, L. Y., and Smulson, M. E. (1998) Cancer Res 58, 4510 –4514.

1. Maisonpierre, P. C., Suri, C., Jones, P. F., Bartunkova, S., Wie-gand, S. J., Radziejewski, C., Compton, D., McClain, J., Aldrich,T. H., Papadopoulos, N., Daly, T. J., Davis, S., Sato, T. N., andYancopoulos, G. D. (1997) Science 277, 55–60.

2. Holash, J., Maisonpierre, P. C., Compton, D., Boland, P., Alex-ander, C. R., Zagzag, D., Yancopoulos, G. D., and Wiegand, S. J.(1999) Science 284, 1994–1998.

Documents

Marked Induction of the IAP Family Antiapoptotic Proteins Survivin and XIAP by VEGF in Vascular Endothelial Cells