(CANCER RESEARCH54, 2582-2590, May 15, 19941

ABSTRACT

Apoptosis(programmed cell death) servesas a commonmechanismofinterphase cell death after radiation exposure in thymic, lymphoid, andhematopoletic cells but has infrequently been documented in other adultmammalian cell types. The present study demonstrates that apoptoticinterphase cell death occurs in endothelial cells after exposure to clinicallyrelevant radiation doses and that basic flbroblast growth factor (bFGF)protects endothellal cells against this mode of the lethal effects of radia

tion. Radiation exposure produced heterologous double-stranded DNAbreaks in endothelial cells, but the cells exhibited a similar competence forrepair of this damage in the presence or absence of bFGF. However,subsequent to the completion ofthis repair process, a sacond pro@ OfDNAfragmentation hocanseapparent@whithwastWected@ in the absenceofbFGFand was associated with a DNA ladder of oligonucleososnalfragments diaracteti@k ofapopto@. The apoptotic DNA degradation o@ireed mainly hi G@-G@—cellsandwasinhibitedbybFGFslimulalion.G3H/HCJmiceexposedtolethal doses ofwhole lung hmdlalion exhib@edsimilar apoptotic changes in theendothelial cell lining of the pulmonary microvasculalure within 6—8h afterradiaUon expoenre. bFGF given Lv. immediately before and after Inadladoninhibited the developmentofapopto@sIn these eeLsand pntected mice againstthe development of lethal tadladon pneumonitis. These Andingsns@ed thatinterphaseapoptoshmsy representa blologicalyrelevantmechanismof radladon-Inducedcell ldll in nonlymphold mnmmnlinnce& both in vilmand in vMiand thatnatural protectionmeehaninasagainstthLsefl@tmaybeamocited withthe levelofnidlatioii re@tance In normal and malignnnt tissuesin vim.

INTRODUCTION

The prevailing hypothesis on the mechanisms of radiation-inducedcell kill identifies the cell nucleus as the primary target for the lethaleffects of ionizing radiation and heterologous DNA double-strandbreaks as the most common type of radiation lesions that lead tomammalian cell death (1, 2). Such lesions are produced as result of directinteraction of the DNA with X-rays or with the short-lived and chemically unstable free radical, produced in the cell by the radiolysis of waterand other molecules (1, 2). The outcome ofthis damage is not necessarilycell death since mammalian cells are proficient in the capacity to repairradiation-induced DNA breaks (3—5).However, not all DNA doublestrand breaks are repaired. In some instances, the load DNA lesions mayexceed the repair capacity of the cell, and some structurally complextypes of double-stranded breaks may be irreparable (1, 6). Thus, residualDNA double-strandedbreakswhich the cell fails to repairare regardedasthose that eventually lead to radiation-induced cell death (4, 5).

Cell death associated with unrepaired DNA breaks is not instantaneous because the cell continues to function after X-ray exposure andmay undergo one to several cell divisions before viability is lost (7, 8).As the cell undergoes mitotic divisions, it progressively developschromosomal aberrations and DNA dysfunction (3, 9) and eventually

Received 1/12/94; accepted 4/5/94.The costs of publication of this article were defrayed in part by the payment of page

charges. I'his article must therefore be hereby marked advertisement in accordance with18 U.S.C. Section 1734 solely to indicate this fact.

I Supported by NIH Grant CA-52462.

2 To whom requests for reprints should be addressed, at Department of Radiation

Oncology,MemorialSloan-KetteringCancerCenter,1275YorkAvenue,NewYork,NY10021.

succumbs to necrotic or apoptotic cell death associated with failure tomaintain its structuraland metabolic integrity (10, 11). When a populalion ofcells is exposed to ionizing irradiation, the probability ofcell deathis a function of the radiation dose and is commonly assessed by loss ofthe clonogemc capacity (12-14). Clonogenic dose-survival curves consistof an exponential component (the “linearcomponent―when plotted onsernilogarithmic coordinates) and a gradually curving component at thelow dose range, frequently referredto as the “shoulderregion―(12—iS).It is generally believed that the slope of the linear component (D0)represents the inherent radiation sensitivity of a tested cell population,while the width of the shoulder (Dq) represents its capacity to repairradiation lesions (12—14).

Whereas the postmitotic mode of radiation-induced cell kill is theprevalent mechanism expressed in the majority of mammalian cells, ithas been shown that thymocytes, lymphocytes, and hematopoieticcells respond to the lethal effects of radiation with an interphase modeof cell death associated with activation of the programmed cell death(apoptotic) pathway (16, 17). Apoptosis is known as a fundamentalphysiological mechanism of cell death that is genetically regulatedduring embryonal development and adult mammalian life and contributes to the balance between cell growth, differentiation, and thewear and tear maintenance of normal mammalian tissues (18, 19). Itis conceptualized as a preprogrammed pathway of sequential biochemical events that are only partially known but which eventuallylead to activation of a calcium/magnesium-dependent endonucleasethat cleaves the nuclear chromatin at selective internucleosomal linkersites (20). The damaged cell exhibits typical morphological changesconsisting of nucleoplasmic condensation, nuclear budding, and formation of apoptotic bodies (ii), and agarose gel electrophoresis ofDNA extracted from such cells reveals a typical pattern of a DNAladder of oligonucleosomal fragments (20), both regarded as hailmarks of the apoptotic process. In addition to thymic, lymphoid, andhematopoietic cells (17), radiation-induced apoptosis has been reported in stem and undifferentiated progenitor cells of testicular,intestinal, renal, neuronal, and oligodendrocytic lineage (17). Matureand differentiated progenies of nonlymphoid maimnalian cells rarelyexhibit apoptosis in response to ionizing irradiation (17), although ithas been reported in the epithelium of intestinal crypts (21), the aciniof the protid and lacrimal glands (22), and in an experimental modelsof transplanted murine ovarian and hepatocellular carcinomas (23).

In a recent study (24), we reported that bFGF3 protects endothelialcells against the lethal effects of ionizing irradiation in vitro inaddition to serving as a potent growth factor for these cells. The bFGFeffect was observed mainly in G0-G1 phase cells, was expressed onlyif bFGF stimulation was provided during the first several h afterirradiation, and was neutralized by specific monoclonal antibodiesagainst bFGF (24, 25). The present study demonstrates that theradiation protection associated with bFGF stimulation is not mediatedvia an effect on the repair of DNA breaks but rather through inhibition

3 The abbreviations used are: bFGF, basic fibroblast growth factor; hrbFGF, human

recombinant bFGF; BAE@ bovine acetic endothdlial cells; PBS, phosphate-bufferedsaline;DMEM,Dulbecco'smodifiedEagle's medium;IL, interleukin;PKC,proteinkinaseC.

2582

Basic Fibroblast Growth Factor Protects Endothelial Cells against Radiationinduced Programmed Cell Death in Vitro and in Vivo'

Zvi Fuks,2 Roger S. Persaud, Alan Alfieri, Maureen McLoughlin, Desiree Ehleiter, Jeffery L. Schwartz,Andrew P. Seddon, Carlos Cordon-Cardo, and Adriana Haimovitz-Friednian

Departments ofRadiation Oncology fZ. F., R. S. P., A. A., M. M., D. E., A. H-F.J and Pathology [C. C-C.J, Memorial Sloan-Kettering Cancer Center, New York New York10021; Argonne National Laboratory, Argonne, Illinois [J. L S.]; and Lederle Laboratories, American Cyanami4 Pearl River, New York [A. P. S.]

on June 8, 2020. © 1994 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

bFOF INHIBITION OF RADIA11ON-INDUCED APOFrO5I5

of interphase apoptosis, regularly triggered in endothelial cells byradiation exposure. Furthermore, a similar protection was observed inthe endothelial cell lining of the pulmonary microvasculature in vivoafter i.v. injection of bFGF into mice receiving lethal doses of wholelung irradiation. These findings suggest that apoptosis may representa biologically relevant mode of the lethal effects of radiation innonlymphoid tissues in vivo and that natural protection mechanismsagainst this effect may exist and may affect the level of tissuesensitivity to radiation exposure at the clinically relevant dose range.

MATERIALS AND METhODS

Cell CUltUresand Radiation Dose-Survival Curves. Clonedpopulationsof BAEC were established from the intima of bovine aorta, cultured andirradiated as described previously (24, 25). Colony formation assays forassessingtheclonogenicsurvivalwerecarriedout in culturedishescoatedwiththe HR9 extracdllular matrix as described previously (25). The survivingfraction was calculated at each dose level as the ratio of the colony numbersobserved in the irradiated versus unirradiated controls (24). Survival curveswere generatedby a computer-assistedprogram(26) which calculatesthe bestfit curves according to the single-hit multitarget model (13) by iterativeweighted least square regressionanalysis of all data points, estimating thecovariance of the survival curve parameters and the corresponding confidencelimits.

Evaluation of DNA Double-Strand Breaks. DNA double-strand breakswere quantified by the nondenaturing neutral elution method, which measuresthe size of double-strandedDNA fragmentsby elution profiles of extractedDNA throughpolycarbonatedifiters(27). BAEC(2.0 X l0@in 100-mmdishes)were radioactively labeled by overnight incubation with [3H]thymidine(finalconcentration,0.20 @Ci/ml)in DMEM containing 10% bovine caff serum.After a 1-h chase with fresh medium (without radioactive thymidine), themedium was replaced with cold PBS, and the cells were irradiated on ice in the

presence or absence of 0.75 ng/ml of human recombinant bFGF (R&D Systems, Minneapolis, MN). Immediately after irradiation or following reincubation at 37°Cin complete mediumwith or without0.75 ng/ml bFGF,the cellswere collected, washed, and applied onto 20-mm polycarbonated filters (2 @mpore size). Extraction of the DNA and its elution through the filter wereperformed as described previously (27). The fraction of DNA retained on thefilterrepresentseitherprimarilyintactand/orrepairedDNA. Whencells wereirradiated on ice and tested immediately after irradiation, the fraction of DNAretained on the filter decreased from 86% recorded in unirradiated controls to29% recovered from cells exposed to a radiation dose of 20 Gy. The latter dosewas chosen to evaluatethe effect of bFGFon the rejoiningof DNA doublestrand breaks because at lower doses, the magnitude of the measured effect wastoo low to assure statistical significance.

Terminal Transferase Nick-End Translation Assay in Vitro. The presence of DNA breaks was also evaluated by a modification of the nick-endtranslation method for in situ labeling ofDNA breaks (28, 29). Briefly, 1 X 106cells were removed from the extracellular matrix-coated culture dish by a2-mm incubation in 0.05% trypsin and 0.02% EDTA in PBS, washed twice inHanks' balanced salt solution, and collected in 10% BCS-DMEM. The cellswere fixed with 4% bufferedformaldehydefor 15 mm on ice and stored in70% ethanol overnight at —20°C.The cells were then rehydrated by doublewashing in PBS and incubatedin TDT buffer (25 mMTris-HC1,200 mMpotassium cacodylate, and 5 mMcobalt chloride at pH 6.6), 0.2 unit/@dterminaldeoxynucleotidyl transferase, and 2 aM biotin-11-dUTP (both purchased from

Boehringer Mannheim Biochemicals, Indianapolis, IN) at 37°Cfor 30 mm.The reactionwas stopped by additionof ice-cold PBS, and the cells wereincubated for 30 min at room temperature in the dark in a solution containing4X SSC buffer(Sigma ChemicalCo., St. Louis, MO), 15 pg/mi fluoresceinated avidin (fluorescein isothiocyanate-avidin; Boehringer Mannheim), 0.1%Triton X-100 (v/v), and 0.5% nonfat milk. After one additional wash inice-cold PBS containing 0.1% Triton X-100 (v/v), the cells were suspended ata concentrationof 10@cells/mI in PBS containingpropidiumiodide (5 @g/ml)and 0.1% RNase. The red (propidium iodide) and green (fluorescein) fluorescence were measured using a FACSCAN flow cytometer equipped with adoubletdiscriminatingmodule(BectonDickinson,SanJose, CA), andthedatawere analyzed using the LYSYS II software (Hewlet Packard, Palo Alto, CA).

An analysis region was set to exclude cell aggregates, and the green channelwas set to score less than 1% of the signals from unirradiated control cells.

Agarose Gel Electrophoresis of DNA. The presence of apoptosis in vitrowas evaluated by agarose gel electrophoresis of DNA extracted from BAEC asdescribed previously (20). Briefly, 5 X 10@cells were washed twice with coldPBS andlysed with Nicoletti lysis buffer[10 mMTris-HCI(pH 7.5), 100 mMNaCI, 1 mM EDTA, 1% sodium dodecyl sulfate, and 50 @g/mlproteinase K]at 43°Covernight. The DNA was extracted with an equal volume of phenol

following a chloroform:isoamyl alcohol wash. The DNA solution was thenincubated for 1 h at 37°Cwith 50 g@g/mlRNase followed by a chloroform:isoamyl alcohol wash. The DNA was then precipitated with 0.5 M NaCI and95% ethanol at —20°Covernight. The pellets were resuspended in 50 g.dH20,and the DNA concentration was determined from the absorbance at 260 nm.Each DNA sample (10 @.tg)was electrophoresed at 80 V for 4 h through a 2.0%agarose gel in Tris-borate buffer. DNA bands were visualized under UV lightafter staining with ethidium bromide. HaeIH digests ofdouble-stranded 4X174

DNA were used as molecularsize markers.Histochemlcal Staining ofApoptotic Cells. Morphological changes in the

nuclear chromatin of cells undergoing apoptosis were detected by staining withthe DNA-bindingfluorochromebis(benzimide)trthydrochloride(Hoechst33258)as describedpreviously(30).Briefly,confluentmonolayersofBAECwere washedwith cold PBS and irradiated as described above. Prior to staining, the cells weretrypsinizedwith 0.05%trypsinand 0.02%EDTA in Hanks' balancedsalt solution,collected in 10% @S-DMEM,resuspended for fixation in 200 @iJof 3% pamformaldehyde,and incubatedfor 10ruinat roomtemperature.The fixativewasthen removed,the cells were washedwith PBS, resuspendedin 20 @dof PBScontaining8 p.g/mlof Hoechst33258, and incubatedat room temperaturefor 15min. Aliquots of 10 @dwere then placed on glass slides coated with 3-aminopropyltriethoxysilane, and 500 cells were counted and scored for the incidence ofapoptotic chromatin changes under an Olympus BH-2 fluorescencemicmscopeusing a BH2-DMU2UV Dich. Mirror Cube filter.

lodination and Blotinylationof bFGF. lodinationof recombinanthumanbFGF (R&D Systems, Minneapolis,MN) was performedusing lodogen kits(Pierce, Chicago, IL) as described previously (31). Biotin conjugation tohuman recombinant bFOF was performed as previously described (32).Briefly, bFGF (1 mg dissolved in 0.4 ml PBS) was mixed with 0.1 ml of 0.2M Na2H@03, 0.15 M KCI (pH 8.8), and 5 @dof biotin N-hydroxy-succinamide

ester solution (Calbiochem, San Diego, CA) in dimethylformamide (12 mg/ml). After 15 min of incubation at room temperature with gentle shaking, 50@.dof 1 M NH4CI (jH 6.0)were added to stop the reaction, and the excess biotin

N-hydroxy-succinamide ester was dialyzed out against four changes of PBS.Biotin conjugationto the bFGFdid not affect its mitogenic activity as demonstrated by [3H]thymidine incorporation into the DNA of BAEC (data notshown).

Survival after Whole Lung Irradiation in Vivo. Adult (4-month old)C3H/HeJ male mice were anaesthetized with phenobarbital (35 mg/kg) andpositioned in a lucitejig designed to expose both lungs but to shield the mediastinumfromthe effectsofradiation.A PhilipsX-rayunit (MGC-20model,320KV,10 mA, 03 mm Cu ifitration, 140 cOy/mis, 50 cm source to skin distance) wasused to irradiate the mice. Some animals received i.v. injections of the humanrecombinantbFGF analogue glu3―-hrbFOF(33) (kindly provided by Dr. PeterBohien, American Cyanamid).The glu3@'-hrbF0Fwas shown to have the samebiological properties as native bFOF (33). In addition, experiments performed byus (data not shown) revealedthat giu3@'-hrbFGFhad an identicalmitogeniceffecton BAEC in vitro as observed with hrbFOF (purchased from R&D Systems,Minneapolis,MN). The hrbFOF was delivered at doses of 400 ng immediatelybeforeandafterirradiationandat 1 and2 h thereafter.Timeadjustedsurvivalwascalculated from the date of irradiation using the psoduct limit Kaplan-Meiermethod (34).

Detection of Apoptosis in Vivo by the Terminal Transferase Nick-EndTranslation Assay. The detection of apoptotic changes in vivo was performedaccording to the in situ DNA nick-end translation method as described previously (28). Lung and thymic tissue specimens were fixed by overnight incubation in 4% buffered formaldehyde and embedded in paraffin blocks. Tissuesections (5 p.m thick) were obtained by microtomy from the paraffin blocks,adhered to slides, and deparaffinized by heating at 90°Cfor 10 min and at 60°Cfor 5 min, followed by a xylene wash twice for 5 min. The section-mountedslides were washed with 90 and 80% ethanol (3 min each), rehydrated, andstained by the DNA nick-end translation method as described previously (28).

2583

on June 8, 2020. © 1994 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

the measured effect was too low to assure statistical significance.Fig. 2 shows that neither the initial number of radiation-induced DNAdouble-strand breaks nor the rejoining process of such breaks wasaffected by the presence or absence of bFGF, with repair proceedingto its maximum within the first 90—120mm after irradiation, regardless of bFGF stimulation. Similarly, there was no effect of bFGF onthe rejoining of radiation-induced DNA single-strand breaks, meanured by the alkali-unwinding technique (35), with repair reaching nearmaximum levels at about 30 mm after irradiation, regardless of bFGFstimulation (not shown).

Although these data indicated that bFGF stimulation does not affectthe rejoining of radiation-induced DNA breaks, it was still assumedthat either the structural or the functional integrity of the DNA wereimpaired in bFGF-deprived cells because of the significant decrease inthe clonogenic survival associated with the delay in bFGF stimulationshown in Fig. 1. To evaluate the structural integrity of the DNA overthe course of time after the repair process of the primary radiationinduced DNA breaks is completed, the terminal transferase nick-endtranslation technique for in situ labeling of DNA breaks (28, 29) wasused. This method is based on nick-end translation of 3'-OH DNAtermini by the enzyme terminal deoxynucleotidyl transferase introduced into permeabiized cells after fixation. Biotin-conjugated dUTPwas used as a substrate for the nick-end translation, and fluoresceinated avidin, which conjugated with the incorporated biotin, was usedto report the presence of 3'-OH DNA termini. The fluorescein-labeledcells were identified by flow cytometry. Fig. 3A shows that thismethod was not sensitive enough to detect the primary radiationinduced DNA double-strand breaks produced immediately after cxposure to a single dose of 500 cGy, apparently because the incidenceof specific 3'-OH break terniini was too low to be detected by thismethod. However, starting at approximately 4 h after irradiation(>2 h after the presumed completion ofthe repair of primary radiationinduced breaks, as disclosed in Fig. 2), a secondary fragmentationprocess of the DNA could be detected in bFGF-deprived cells, cvidenced by a time-dependent increase in the number of cells exhibitingintense fluorescence in the terminal transferase assay (Fig. 3A). The

S

11@

C 0.80

SCS

@ 0.6C0

0S11@

z

0.0

Fig. 2. Effects ofbFGFon the generation and repair ofDNA double-strand breaks afterradiation exposure. Plateau phase BAEC were exposed to a single dose of 20 Gy in thepresence and absence of 0.75 ng/ml bFGF. The DNA was extracted at the indicated timepoints and evaluated for DNA double-strand breaks by the neutral filter elution method(27).Thefractionof intactor repairedDNAretainedon thefilteris shownas a functionof time after irradiation. Points, mean ±SE (bars) obtained from four consecutiveexperiments.

bFGF 1NHIB@ON OF RADIATION-INDUCED APOPTOSI5

C0

,@ 10_I

C

c@io2

0 2 4 6 8

Radiation Dose (Gy)

Fig. 1. The effect of bFGF on the clonogenic survival of irradiated plateau phaseBAEC. Human recombinant bFGF (0.75 ng/ml) was added to the media of the colonyformation assays either immediately (0) or at 72 h (is) after irradiation. Each data pointrepresents results obtained from 3-6 dishes ±1 SEM (bars).

RESULTS

Effect of bFGF on the Clonogenic Survival of BAEC. Fig. 1shows the clonogenic survival curves of BAEC exposed to gradeddoses of radiation. The cells used in these experiments were in plateauphase 7—9days after reaching confluence, shown to consist mainly ofcells in G0-G1 (25). The control cells in Fig. 1 received delayed bFGFstimulation at 72 h after irradiation to assure that bFGF-sensitiveradiation lesions were either fixed or repaired and were no longeramenable to modulation by bFGF (24). Under these conditions,BAEC exhibited a significant sensitivity to ionizing irradiation with aD0 value of 83.7 ±4.3 cOy but only a limited capacity for recoveryfrom radiation damage, as evidenced from the minimal shoulderregion (Dq of 34.4 ±2.5 cGy). Addition of bFGF to the culture mediaimmediately after irradiation (the 0 time curve) resulted in a significant increase in the clonogenic survival at each dose level. This wasalso expressed as an increase in the shoulder region of the dosesurvival curve (Dq of 205 ±3.1 cOy; P < 0.05 as compared to the72-h curve), but there was no significant change in the slope of thelinear component of the curve.

Effect of bFGF on DNA Fragmentation in Irradiated BAEC.To explore the effects of bFGF on the initial yield and rate of repairof primary radiation-induced DNA double-stranded breaks, DNA wasextracted from irradiated plateau phase BAEC, and the proportion ofDNA double-stranded breaks was quantified by the neutral filterelution method (27). This nondenaturing technique fractionates double-stranded DNA according to size by elution through polycarbonated filters. The DNA retained on the filter represents primarily intactand/or repaired DNA, while the fraction that elutes through the filterrepresents smaller DNA fragments produced by double-stranded fragmentation (27). When BAEC were irradiated on ice and tested immediately after irradiation, the DNA fraction retained on the filterdecreased as a dose-dependent function (not shown), reaching aminimum level of 29% (as compared to 86% in unirradiated controls)in cells exposed to a dose of 20 Gy. In the present study, the latterdose was chosen to evaluate the effect of bFGF on the rejoining ofDNA double-strand breaks because at lower doses, the magnitude of

I .0 2.0 3.0

Time (hours)

2584

on June 8, 2020. © 1994 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Control.@

@ -•M& ••4@i@@@@ -st@@

—

LSOOcGy

Otinie.

@ @ate@@ @eèi •@eM@SOOcGy+bFGF8hrs00

SOOcGy81us.97•8

BAEC to the antiapoptotic effect of bFGF. Staining of irradiated cellswith bisbenzimide trihydrochloride (Hoechst 33258) revealed that inconjunction with the induction of internucleosomal DNA fragmentation, the cells also expressed extensive morphological features ofapoptosis, including nucleoplasmic condensation, nuclear budding,and formation of apoptotic bodies (Fig. 5A). There was a dosedependent increase in the proportion of cells undergoing apoptosis(Fig. 5B) and similar to the other expressions of apoptosis (Figs. 3Band 4), it was inhibited by bFGF stimulation (Fig. SB). These datasuggest that the process of programmed cell death is activated uponradiation exposure in G0-G1 endothelial cells and that bFGFstimulation rescues these cells by inhibiting the apoptotic pathway,thus creating a shoulder pattern in the clonogenic dose-survivalcurve (Fig. 1).

Distribution ofbFGF after i.v. Injection In C3H/HeJ Mice. Ourprevious immunohistochemical studies (36) showed that bFGF isubiquitously detected in the basement membranes of large and intermediate size blood vessels, but capifiaries have significant segments oftheir basement membranes with diminished or absent bFGF deposits. Toevaluate whether i.v. applications of bFGF can enrich the vascular endotheium in vivo with bFGF, we performed pharmacokinetic studies ofthe distribution of ‘@I-bFGFinjected i.v. into C3H/HeJ mice. Fig. 6Ashows that most of the injected bFGF disappeared from the circulation@within1-2 mm after injection. The rapid clearance suggests an intravas

cular binding of bFGF, probably to the heparan sulfate that coats theintraluminal surface ofthe vascular endotheium (37) known for its bFGFbinding affinity (31). To explore this possibility, we injected biotinconjugated bFGF i.v. into C3H/HeJ mice (1 pg/mouse), removed tissuespecimens, and stained fresh-frozen tissue sections with avidin-biotinperoxidase complexes and the diaminobenzidine chromogen (36).As shown in Fig. 6B, this study confirmed that the injected bFGF remainsbound to the endotheium with minimal if any leakage into theperivascular space.

Effect ofbFGF on the Development ofRadiation-induced Pneumonitis in C311/HeJ Mice. Having established the availability ofi.v.-injected bFGF to the vascular endothelium, we tested its clinicaleffect on C3HIHeJ mice receiving whole lung irradiation. A previousstudy reported that the LD501180for whole lung irradiation in C3HIHeJmice was 1600 cGy (38). However, whole lung irradiation in thisstudy was delivered by a whole chest irradiation technique (38),without protection provided to the mediastinal organs (i.e., esophagus

bFGF 1NH1B@ON OF RADIATION-INDUCED APOPTO5IS

A

S

C0S20

U-

CS2S

0@

Time (hours)B

@_. 45 @øS @NS@ 151@@@@ lii

DNA ContentFig. 3. DNA strand breaks detected in plateau phase BAEC by the in situ terminal

transferase nick-end translation assay. A, effect of bFGF (0.75 ng/ml) on the number offluorescingcells as a function oftime afterexposure to a single dose ofSOOcOy. The cellswerecountedby flowcytometry,gatingthewindowfor thegreenfluorescenceto permitthe recording of fluorescence signals from 0.7% of the unirradiated control cells. B,contour plots of the the frequency of cells exhibiting fluorescence emitted from thedUTP-biotin-avidin-fluorescein isothiocyanate complexes versus cellular DNA content asdetected by flow cytometry. This analysis makes it possible to correlate the presence ofDNAbrCaICSin individual cells with their DNA ploidy and their position in the cell cycle.Plateau phase BAEC were irradiated with a single dose of 500 cOy and tested for thepresence of fluorescing cells after 8 h of incubation with or without 0.75 ng/ml bFGF.Note that fluorescing apoptotic cells (Apo) appear only among bFGF-untreated cells andhave a DNA ploidy of G0-G1cells.

- + + + + + + + +

10 0 0 2 2 6 6 10 10- . + - 4 • 4 - +

Radiadoa -TIme(bn) 0bPGF -

formation of this new type of DNA break was inhibited by bFGFwhen added to the medium immediately after irradiation. Flow cytometry analysis of the fluorescence emitted from dUTP-avidin labeled cells versus cellular DNA content (Fig. 3B) revealed that thefluorescing cells were mostly G0-G1 cells, involving approximately50% of the population of G0-G1 cells after a dose of 500 cGy (Fig.3A).

To furtherexplorethenatureof thesecondaryDNA fragmentationprocess, neutral agarose gel electrophoresis was performed with DNAextracted from irradiated BAEC. These studies revealed that thesecondary DNA degradation process produced a typical DNA ladderof oligonucleosomal fragments (Fig. 4), characteristic of apoptosis(ii, 20). The apoptotic DNA degradation was inhibited by bFGF (Fig.4), althougha minimal degreeof ladderingcould be detectedat 10 hin the presence of bFGF, suggesting a heterogeneity among the tested

Fig. 4. Agarose gel electrophoresis of DNA extracted from plateau phase BAECirradiated with a single dose 500 cOy and incubated after irradiation with (0.75 ng/ml) orwithout bFGF.

2585

@T-;@

on June 8, 2020. © 1994 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

D Control—bFGFLa@

c)

0

0.00.

.,t@

0

I..

04

20

0 50 lt@O 200 500 1@

Dose (cGy)

Fig. 5. Morphological changes associated with radiation-induced apoptosis in plateauphase BAEC. A, the cells were irradiated with a single dose of 500 cOy and wereincubated without or with bFGF (0.75 ng/ml) for 8 h. The cells were then fixed withparaformaldehyde and stained with Hoechst 33258. Note that while in the control(bFGF-untreated) cells (a) approximately one-third of the cells in the field shown haddeveloped chromatin changes of apoptosis; there was only a single cell with apoptoticchanges among the bFGF treated (b) cells. B, the effect of bFGF (0.75 ng/ml) on thepercentage of cells exhibiting morphological changes of apoptosis 8 h after exposure toincreasing doses of irradiation. A total of 500 cells were counted for each data point.

and heart), which may contribute to the lethal syndrome of wholechest irradiation. We initially confirmed that the LD50,iso for wholechest irradiation in our strain of C3HIHeJ mice was indeed 1600 cGy(not shown). However, when a lead block was introduced to shield themediastinal organs from the effects of radiation, the LD50,iso in thetested strain of C3H/HeJ mice was 2050 cGy. bFGF protected suchtreated animals from lethal radiation pneumonitis at all dose levels.Fig. 7 shows the protection conferred in mice receiving a whole lungdose (with mediastinal shielding) of 2250 cGy. In the bFGF-treatedgroup, treatment consisted of four i.v. injections of the human recombinant bFGF analogue glu35-hrbFGF (33), delivered at doses of 400ng each immediately before and immediately after irradiation and at1 and 2 h thereafter, but no additional maintenance therapy was givensubsequently. This arbitrary dose schedule of bFGF therapy may besuboptimal in terms of providing the full potential therapeutic effectof bFGF, but it still demonstratesthat bFGF therapy significantlydecreases the mortality from radiation pneumonitis. The actuarialsurvival at 180 days was 57% in bFGF-treated mice compared to 35%

bFOF IN}flB@ON OF RADIATION-INDUCED APOPTO5IS

in control bFGF-untreated mice (P = 0.001). At 60 days after irradiation, the cumulative mortality from radiation pneumonitis in thecontrol (bFGF-untreated) group was 29 of 45 (64%) compared to 12of 42 (28%) in bFGF-treated mice (P = <0.0001).

Since bFGF was given only immediately before and within the first2 h after irradiation, without maintenance therapy given afterwards,these data suggested that bFGF may affect an early critical radiationinduced event that triggers the pathogenesis of acute radiation pneumonitis. Fig. 8 shows the effect of hrbFGF therapy on the tissuedamage that evolves in the irradiated lungs by 8 h after exposure to adose of 2000 cOy. In the control (hrbFGF-untreated) group, stainingwith hematoxylin and eosin (Fig. 8a) revealed severe radiation-induced changes typical of capillary leakage consisting of atelectasis,vascular congestion, hemorrhage, and extensive interstitial edema.Concomitant in situ terminal transferase nick-end staining for thepresence of apoptotic nuclei (28) revealed extensive apoptoticchanges in the nuclei of the capillary endothelium (Fig. 8c), but otherpulmonary tissue elements, including the pneumocytes, the bronchialepithelium, and the endothelium oflarge and intermediate-sized bloodvessels, did not show similar apoptotic changes. Treatment with i.v.hrbFGF nearly completely abrogated the radiation-induced apoptoticchanges in the capillary endothelium (Fig. 8d), although examinationof thymic tissue, incidentally present in the same sections, showedapoptotic changes in more than 80% of the thymocytes (not shown).Concomitantly, there was a nearly complete prevention of the capillary leakage-related changes (Fig. 8b) observed in the control group.

The apoptotic changes in control bFGF-untreated mice peaked at 12 hand nearly completely disappeared (presumably due to phagocytosisof apoptoticcells)by 24h (not shown).Thesedatasuggestthat,in thisexperimental system, the therapeutic effect of bFGF may be associated with inhibition of programmed cell death, selectively induced inthe pulmonary capillary endothelium by the radiation exposure.

DISCUSSION

The data reported in this study demonstrate that interphase apoptotic cell death represents a significant mechanism of the lethal effectsof ionizing radiation in plateau phase endothelial cells and that theradiation protection conferred by bFGF in these cells is associatedwith down-regulation of the apoptotic pathway. A characteristic feature of the plateau phase BAEC used in the present experiments is thatmore than 90% of the cells were in G0-G1 (25). This is analogous tothe reported cell cycle distribution of vascular endothelial cellsin vivo, shown to be mostly quiescent and to exhibit a low incidenceof cycling cells, with a labeling index ofless than 1% and an estimateddoubling time of 50—1000days (39). The present data indicate thatthe cells which exhibit sensitivity to radiation-induced interphaseapoptosis are G0-G1 cells (Fig. 3B) and that the bFGF-mediateddown-regulation of the apoptotic pathway in this subpopulation ofendothelial cells is associated with the appearance of a shoulderpattern in the clonogenic dose-survival curve (Fig. 1). On the otherhand, bFGF did not affect the slope (D0) of the linear component ofthe dose-survival curve (Fig. 1), believed to be associated with unrepaired primary DNA double-strand breaks. This notion is supportedby the lack of effect of bFGF on the repair of DNA double-strandbreaks, demonstrated by the neutral elution experiments (Fig. 2).

The ability to modify the shoulder region of the dose-survival curvehas traditionally attracted a great deal of attention, because the doserange in which the shoulder phenomenon occurs corresponds with thedose range commonly used for the management of human tumors. Inhis classical publications on the response of mammalian cells toradiation exposure, Elkind (12, 15, 40) conceptualized the shoulderphenomenon as representing an ability of irradiated cells to accumu

2586

B40

30

10

on June 8, 2020. © 1994 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

bFOF INHIBITION OF RADIATION-INDUCED APOFFO5I5

A

Ba

0.C.,

Fig. 6. Distribution of bFGF injected iv. intoc:3H/HeJ mice. A, pharmacokinetic studies of i.v.injected 1@I-bFGF. The radioactivity of wholeblood samples (150 @d)is shown as a function oftime after i.v. injection of400 ng of @l-bFGF.Theexpected initial cpm/150 gd ofwhole blood sampleafter reaching equilibrium and assuming no removal from the serum was approximately 5000cpm. B, tissue localization ofbFGF after i.v. injection of 1 gagof biotin-conjugated bFOF per mouse.Fresh frozen lung and brain tissue specimens wereobtained, frozen at -20°C, and cut using a cryostatat a 5-@smslice thickness. Binding of biotinylatedbFOF to tissue elements was identified by secondaT), conjugation to avidin-biotin-peroxidase cornplexes,and colordevelopmentwas identifiedbydiaminobenzidine chromogen as described previously (43). Sections were counterstained with hematoxylin. A lung specimen (a) was obtained 5ruin and a brain specimen (b) 60 mm after bFGFinjection.

Time After Injection (minutes)

I

V

late sublethal quantities of radiation lesions and to repair them overthe course of time. Although Elkind did not characterize the biochemical nature of this type of radiation damage, he noted that the kineticsof sublethaldamagerepairwasappreciablyslowerthantherepairofradiation-induced DNA single- and double-strand breaks (40). Thisconcept has gained a broad acceptance and has constituted the biological basis for the design of fractionated radiation schemes inclinical radiotherapy.

Our observations provide a possible alternative explanation for theshoulder phenomenon in the context ofthe pleiotropic lethal effects ofradiation in mammalian cells. According to our model, the preprogrammed pathway of apoptosis may be triggered by radiation exposure in mammalian cells more frequently than has been appreciated inthe past. Down-regulation of apoptosis by cell-type specific mechanisms (such as bFGF stimulation in endotheial cells) may rescueapoptosis-sensitive subpopulations of irradiated cells, thus creating ashoulder pattern of the dose-survival curve. The shoulder region may,therefore, represent a dose range in which interphase apoptotic death

may prevail (if antiapoptotic mechanisms are not operational) over thepostmitotic mechanism of cell death. Furthermore, radiation-inducedapoptosis may not be restricted to the low-dose range of the shoulderregion only. Fig. 1 demonstrates a survival advantage for bFGFstimulated BAEC at the high dose range as well, where postmitoticcell death from unrepaired DNA double-strand breaks dominates asthe major mechanism of cell death. Thus, we suggest that both theinterphase apoptotic mechanism and the postmitotic mechanism ofcell death contribute to cell killing after exposure of mammalian cellsto ionizing irradiation. The relative contribution from each mode ofcell death may differ with dose and from one cell type to another,relative to their inherent and inducible capacities to overcome each ofthese types of lethal radiation damage.

The observations on the protective effect ofbFGF against radiationinduced apoptosis in endothelial cells are analogous to the reportedinhibition of glucocorticosteroid- and cytotoxic agent-induced apoptosis conferred by hematopoietic cytokines (i.e., IL-2, IL-3, IL-6, andgranulocyte-macrophage colony-stimulating factor) in normal and

2587

b

%

on June 8, 2020. © 1994 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

bFGF INHIBITION OF RADIATION.INDUCED APOFrO5I5

radiation resistance. In this context, it is of interest to note that recentstudies have reported that radiation directly activated membrane PKCwithout apparent stimulation by cytokines or growth factors in humansarcoma, small cell carcinoma, and B-lymphocyte precursor cells(48—SO).It thus seems that some cells activate PKC spontaneouslyafter irradiation, while other cells, such as endothelial cells, lack thisproperty and require cytokine or growth factor stimulation to elicitthis response. Furthermore, recent studies have reported that PKCinhibitors sensitized the radiation killing of such cells without affecting their capacity to repair primary radiation-induced DNA doublestrand breaks (49, 51). Although it remains unclear whether apoptosisis involved in the enhanced radiation killing of PKC-inhibited cells,the resemblance of these observations to our current findings on theeffect of bFGF deprivation on irradiated endothelial cells (Figs. 1 and2) is striking. Taken together, these observations provide a basis for ahypothesis on the existence of a balance between radiation inductionof programmed cell death and its down-regulation by natural suppressor mechanisms through PKC. According to this model, spontaneousradiation activation of membrane PKC or its activation by growthfactors and cytokines may play an important role in the homeostaticcontrol of radiation resistance in many types of cells. When antiapoptotic mechanisms are activated, a shoulder becomes apparent on thedose-survival curve, associated with increased resistance to the lethaleffects of radiation. Such a mechanism may be of special relevance atthe clinically applicable low dose range, where primary radiationinduced DNA double-strand breaks are less prevalent than at higherdose levels.

To explore whether this putative mechanism is of biological relevance in experimental or clinical models in vivo, we studied theeffects of bFGF on the radiation response of the endothelial cell liningof the pulmonary microvasculature in mice exposed to lethal doses ofwhole lung irradiation. Previous histopathological and ultrastructural

100

DlC>

5-

Cl,

C05-00@00@

MonthsFig. 7. Modulation of the response of C3H/HeJ mice to whole lung irradiation by

bFGF.Actuarial(Kaplan-Meier)survivalcurvesof f3WHeJmicereceiving2250cOywholelungirradiation;deliveredwitha mediastinalblockto shieldthe esophagusandheart from the effects of irradiation. The differences between the time-adjusted incidencerates were evaluated using the Mantel log-rank test for censored data.

malignant hematopoietic cells (19, 20, 41). Activation of PKC hasbeen suggested as a mechanism for the cytokine-mediated protectionagainst glucocorticosteroid-induced apoptosis (42, 43). Thymic andlymphoid cells are similarly protected by PKC activation againstradiation-induced apoptosis (44—46),although the agonists for PKCactivation in these cells remain unknown. As to the BAEC system, wehave recently reported (47) that PKC activation mediates the protection conferred in endothelial cells by bFGF against radiation-inducedapoptotic DNA degradation, such as demonstrated in the presentstudy. These data suggest that PKC activation may provide an antiapoptotic mechanism and may constitute a generic mechanism of

@.@ . .. .@ .: ‘

‘.m “@ )@-— I

‘..,

.-@

:-@ ,@

-.@\..@‘.:@ J_@ -

0 6 12

•@ :@ :

@@@@ ••@@ . . :@@ T

b@@e 4,,,4•5 S /

•.@•-.•.. —HI@

•-@4'. @m.

@.;@ 4$

@, @ts@@@ !,•

S.' ,@@ I. ‘

,11._*. p @..,. .t@. .@

.•#‘@ , -Fpy@@ â€@̃:#@@@%. @.@ -,t@•

Fig. 8. Effect ofglu3-'-hrbFGF on the radiation-induced histopathological changes in formaldehyde-fixed and paraffin-embedded pulmonary tissues 8 h after whole lung irradiationwith a dose of2000 cOy. Hematoxylin and eosin staining oflung tissue is shown in a confrol(hrbFGF-untreated a)and hrbFOF(b)treated mice. C,staining by the terminal transferasenick-end translation method ofapoptotic nuclei ofthe capillary endotheium within the parenchyma ofan irradiated control(hrbFGF-untreated)lung. Note the insertwhich demonstratesa further magnification (X 400) of a region within the section shown in c. 4 the results of nick-end labeling in the lung of a hrbFOF-treated mouse.

- e.@@‘ @1-

-@ ‘Is, p .@

C'@@ • “4.@

F,@ ‘.s @:‘@t'@i@

2588

@_, ‘d

f

on June 8, 2020. © 1994 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

bFGF LNHIB@ONOF RADIATION-INDUCEDAPOPTO5IS

studies revealed that damage to the microvascular endothelium ispredominant in radiation injury to several normal tissues (e.g., skin,heart, lung, liver, and central nervous system) (51, 52). Our recentimmunohistochemical studies (36) showed that bFGF is ubiquitouslydetected in the basement membranes of large and intermediate-sizedblood vessels. However, capillaries were found to have significantsegments of their basement membranes with diminished or absentbFGF deposits (36). It was, therefore, suggested that bFGF-deflcientsegments of the microvascular endothelium may represent vulnerablesites for damage by radiation exposure (25). Based on these findings,we postulated that i.v. applications of bFGF may saturate bFGFdeficient sections of the capillary endothelium and would thus serveas a radioprotective agent against radiation damage in vivo. Thepharmacokinetic studies of the plasma levels of i.v. injected ‘@IbFGF (Fig. 6A) and the studies on the binding of biotinylated bFGFto thevascularendothelium(Fig. 6B) confirmedthatinjectedbFGFremains bound to the endotheium and may be available for protectionagainst radiation-induced damage to the microvasculature. Furthermore, i.v.-injected bFGF did indeed inhibit the development of radiation pneumonitis in mice exposed to lethal doses of whole lungirradiation (Fig. 7). It should be emphasized that the bFGF therapywas given immediately before and within the first 2 h after irradiation(without maintenance bFGF treatment thereafter), but it nonethelessprevented the clinical syndrome of radiation pneumonitis, which inbFGF-untreated mice developed over the course of weeks to 1 monthlater on. These data suggested that bFGF may affect an early andcritical radiation-induced event that triggers the pathogenesis of radiation pneumonitis. Fig. 8 suggests that this early event may beradiation-induced apoptosis, selectively induced in the pulmonarycapillary endothelium by the radiation exposure.

The existence of in vivo correlates to the in vitro data supports theconcept that interphase apoptosis may represent a fundamental dcment of the pleiotropic effects of radiation in nonlymphoid mammahan cells. It is becoming increasingly apparent that molecular eventsmediated by autocrine, paracrine, and endocrine growth and inhibitoryfactors play a major role in tissue responses to ionizing radiation (53).Recent studies (54) have provided an insight into mechanisms bywhich messages exchanged between cells via paracrine loops maymediate specific cellular and tissue responses that lead to the eventualdevelopment or prevention of radiation damage at the tissue level.Understanding the molecular basis and the temporal evolution ofapoptosis and antiapoptotic effects during a course of fractionatedirradiation may have important clinical implications. The evolution ofradiation injury in normal and tumor tissues is thus placed within acellular and molecular context that provides a background for testableexperimental hypotheses at dose ranges relevant to clinical radiotherapy. As mechanisms of radiation-induced apoptosis in normal andtumor tissues are unraveled, it may become possible to activelyintervene in specific pathways to favorably alter the therapeutic ratioin the radiation treatments of several types of human cancer.

ACKNOWLEDGMENTS

We acknowledgeLauraKane for technical assistance in carryingout thealkali unwinding experiments, Dr. Neil Goldstein for the biotinylation ofbFGF,andDr. RuthGalili for the assistancein the cytochemicalexperiments.

REFERENCES

1. ward, j. F. DNA damage produced by ionizing radiation in mammalian cells:identities, mechanisms of formation and reparability. Prog. Nucleic Acid. Mol. Biol.,35: 95—125,1988.

2. Radford, I. R. Evidence for a general relationship between the induced level of DNAdouble-strand breakage and cell-killing after X-irradiation of mammalian cells. Int. J.Radiat. Biol., 49: 611—620,1986.

3. Bradford,J. S. Sublethaldamage,potentiallylethaldamage,and chromosomalabberations in mammalian cells exposed to ionizing radiation. Int. J. Radiat. Oncol.Biol. Phys., 21: 1457—1469,1991.

4. Steele, 0. G., McMillan, 1. J., and Peacock, J. H. The picture has changed in the1980s. tnt. J. Radial. Biol., 56: 525—537,1989.

5. McMillan, T. J. Residual DNA damage: what is left over and how does this determinecell fate. Eur. J. Cancer, 28: 267—269,1992.

6. Brenner, D. J., and Ward, J. F. Constraints on energy depletion and target size ofmultiply damaged sites with DNA double-strand breaks. Int. J. Radial. Biol., 61:737—748,1992.

7. Thompson, L. H., and Suit, H. D. Proliferation kinetics of X-irradiated mouse L cellsstudied with time-lapse photography. H. mt. i. Radiat. Biol. Relat. Stud. Phys. Chem.Med., 15: 347—362,1969.

8. Hurwitz, C., and Tolmach, J. T. Time-lapse cinemicrographic studies of X-irradiatedHeLa S3 cells. Biophys. J., 6: 607—633,1969.

9. Dewey, W. C., Miller, H. H., and Leeper, D. B. Chromosomal aberrations andmortality of X-irradiated mammalian cells: emphasis on repair. Proc. NatI. Aced. Sci.USA, 68:667—671,1971.

10. Judah, J. D., Ahmed, K., and Mclean, A. 0. Pathogenesis of cell necrosis. Fed. Proc.Am. Soc. Exp. Biol., 24: 1217—1221,1965.

11. Kerr, J. F. R., and Harmon, B. V. Definition and incidence of apoptosis: an historicalperspective. In: L D., Tomei and F. 0. Cope (eds.), Apoptosis: the Molecular Basisof Cell Death, pp. 5—29.Cold Spring Harbor: New York: Cold Spring HarborLaboratory Press, 1991.

12. Elkind, M. M., and Sutton, H. 0. X-ray damage in recovery in mammalian cells inculture. Nature (Lond.), 184: 1293—1295,1959.

13. Curtis, S. B. Lethal and potentially lethal lesions induced by radiation—aunifiedrepair model. Radiat. Rca., 106: 252—270,1986.

14. Chadwick, K. H., and Leenhouts, H. P. A molecular theory of cell survival. Phys.Med. Biol., 18: 78—87,1973.

15. Elkind, M. M. Repair processes in the treatment and induction of cancer withradiation. Cancer (Phila.), 65: 2165—2171,1990.

16. Yamada, T., and Ohyama. H. Radiation-induced interphase death of rat thymocytes isinternally programmed (apoptosis). mt. J. Radiat. Biol., 53: 65—75,1988.

17. Allan, D. J. Radiation-induced apoptosis: its role in a MADCAT (mitosis.apoptosisdifferentiation-calcium toxicity) scheme of cytotoxicity mechanisms. lnt. 3. Radiat.Biol., 62: 145—152,1992.

18. Raff, M. C. Social control on cell survival and cell death. Nature (Lond.), 356:397—399,1992.

19. Vaux, D. L Toward an understanding of the molecular mechanisms of physiologicalcell death. Proc. Nati. Aced. Sci. USA, 90: 786-789, 1993.

20. Wylie, A. H. Glucocorticoid-induced thymocyte apoptosis is associated with endogenous endonuclease activation. Nature (Land.), 284: 555-556, 1980.

21. Potten, C. S. Extreme sensitivity of some intestinal crypt cells to X and ‘yirradiation.Nature (Lond.), 269: 518—521,1977.

22. Stephens, L C., Schultheiss, T. E., Price, R. E., Mg, K. K., and Peters, L J. Radiationapoptosis of serous acinar cells of salivary and lacrimal glands. Cancer (Phila.), 67:1539—1543,1991.

23. Stephens, L C., Ang, K. K., Schuitheiss, T. E., Muss, L, and Myen, R. E. Apoptosisin irradiated murine tumor. Radial. Res., 127: 308—316, 1991.

24. Haimovitz-Friedman, A., Vlodavsky, I., Chaudhuri, A., Witte, L, and Fuks, Z.Autocrine effects of fibroblast growth factor in radiation damage repalr in endothelialcells. Cancer Res., 51: 2552—2558,1991.

25. Fuks, Z., Vlodavsky, I., Andreeff, M., McLoughlin, M., and Haimovitz-Friedman, A.Effects of extracellular matrix on the response of endothelial cells to radiation in vüro.Eur. J. Cancer, 28/A: 725—731,1992.

26. Albright, N. Computer programs for the analysis of cell survival. Radiat. Res., 112:331—340,1987.

27. Bradley, M. D., and Kohn, K. W. X-ray induced DNA double strand break productionand repair in mammalian cells as measured by neutral filter elution. Nucleic AcidsRes., 7: 793—804,1979.

28. Gavriei, Y., Sherman, Y., and Ben-Sasson, S. A. Identification of programmed celldeath in situ via specific labeling of nucleic DNA fragmentation. J. Cell Biol., 119:493-501, 1992.

29. Gorczyca, W., Gong, J., and Darzynkiewicz, Z. Detection of DNA strand breaks inindividual apoptotic cells by the in situ terminal deoxynucleotidyl transferase and nicktranslation assay. Cancer Rca., 53: 1945—1951,1993.

30. Oberhammer, F., Wilson, J. W., Dive, C., Morris, I. D., Hickman, J. A., Wakeling,A. E., Walker, R. P., and Sikorska, M. Apoptotic death in epithelial cells: cleavage ofDNA to 300 and/or 50 kb fragments prior to or in the absence of internucleosomalfragmentation. EMBO J., 12: 3679—3684, 1993.

31. Bashkin, P., Klagsbrun, M., Doctrow, S., Svahn, C. M., Folkman, J., and Vlodavsky,I. Basic fibroblast growth factor binds to subendothelial extracellular matrix and isreleased by heparanase and heparin-like molecules. Biochemistry, 28: 1737—1743,1989.

32. Kan.Mitchell, J., Imam, A., Kempf, R. A., Taylor, C. R., and Mitchell, M. S. Humanmonoclonal antibodies derived against melanoma tumor@associatedantigens. CancerRes., 46: 2490-2496, 1986.

33. Seddon, A. P., Decker, M., Muller, T., Armellino, D., Kovesdi, I., Gluzman, Y., andBohlen, P. Structure/activity relationship in basic FGF. Ann. NY Aced. Sci., 638:98—108,1991.

34. Kaplan, E. L., and Meier, P. Nonparametric estimation from incomplete observations.J. Am. Statist. Assoc., 53: 457—816,1958.

35. Jorgensen,T. J.,Prasad.S. C.,Brennan,T. P.,andDritschilo,A. Constraintsto DNAunwinding near radiation-induced strand breaks in Ewing's sarcoma cells. Radiat.Rca., 123: 320—324,1990.

2589

on June 8, 2020. © 1994 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

bFGF lNHIB@ON OF RADIA11ON-INDUCED APOFFOSIS

36. Cordon-Cardo, C., Vlodavsky, 1.,Haimovitz-Friedman, A., Hicklin, D., and Fuks, Z.Expression of basic fibroblast growth factor in normal human tissues. Lab. Invest.,63: 832—840,1990.

37. Simionescu, M., Simionescu, N., Silbert, J. E., and Palade, 0. E. Differential microdomains on the luminal surface of the capillary endothelium. II. Partial characterization of their anionic sites. J. Cell Biol., 90: 614—621, 1981.

38. Sharplin, J., and Franco, A. J. A. Quantitative histological study of strain-dependentdifferences in the effects of irradiation on mouse lung during the early phase. Radiat.Res.,119:1—14,1989.

39. Hobson, B., and Denekamp, J. Endotheial proliferation in tumours and normaltissues: continuous labelling studies. Br. J. Cancer, 49: 405—413,1984.

40. Elkind, M. M. Repair processes in radiobiology. Radiat. Res@,100:425-449,1984.41. Oren, M. The involvement ofoncogenes and tumor suppressor genes in the control of

apoptosis. Cancer Metastasis Rev., 11: 141-148, 1992.42. Vazquez, A., Auffredon, M. A., Chaouchi, N., Taieb, J., Sharma, S., Galanaud, P.,

and Leca, 0. Differential inhibition of interleukin 2- and interleukin 4-mediatedhuman B cell proliferation by ionomycin: a possible role for apoptosis. Eur. J.Immunol., 21: 2311—2316,1991.

43. Rajotet, D., Haddad, P., Haman, A., Cragoe, E. J., and Hoang, T. Role of proteinkinase C and the Na'/H@ antiport in suppression of apoptosis by granulocytemacrophage colony-stimulating factor and interleukin-3. J. BioL Chem., 267: 9980—9987, 1992.

44. Kizaki, H., Tadakuma, T., Okada, C., Maramatsu, J., and Ishimura, Y. Activation ofa suicide process of thymocytes through DNA fragmentation by calcium and phorbolesters. J. Immunol., 143: 1790—1794,1989.

45. McConkey, D. J., Hartzell, P., Jondal, M., and Orrenius, S. Inhibition of DNAfragmentation in thymocytes and isolated thymocyte nuclei by agents that stimulateprotein kinase C. J. Biol. Chem., 264: 13399—13402,1989.

46. Story, M. D., Stephens, L C., Tomasovic, S. P., and Meyn, R. E. A role for calciumin regulating apoptosis in rat thymocytes irradiated in virro. Int. J. Radiat. Biol., 61:

243—251,1992.47. Haimovitz-Friedman, k, Balaban, N. A., McLaughlin, M., Ehleiter, D., Michael, J.,

Vlodavsky, I., and Fuks, Z. Protein kinase C mediates basic fibroblast growth factorprotection of endothelial cells against radiation-induced apoptosis. Cancer Res., 54:2591—2597,1994.

48. Hallahan, D. E., Virudachalam, S., Schwartz, J. L, Panje, N., Mustafi, R., andWeichselbaum,R. R. Inhibitionof proteinkinasessensitizeshumantumorcellstoionizing radiation. Radiat. Rca., 129: 345—350,1992.

49. Kim, C. Y., Giaccia, A. J., Strulovici, B., and Brown, J. M. Differential expression ofprotein kinase C e protein in lung cancer cell lines by ionising radiation. Br. J. Cancer,66: 844—849,1992.

50. Uckun, F. M., Schieven, 0. L, Tuel-Ahlgren, L M., Dibirdik, I., Myers, D. E.,Ledbetter, J. A., and Song, C. W. Tyrosine phosphorylation is a mandatory proximalstep in radiation-induced activation of the protein kinase C signaling pathway inhuman B-lymphocyte precursors@Proc. NaIl. Acad. Sd. USA, 90: 252-256,1993.

51. Hallahan, D. E., Virudachalam, S., Ordina, D., and Weichselbaum, R. R. Theisoquinoline sulfonamide H7 attenuates radiation-mediated protein kinase C activation and delays the onset ofX-ray-induced 02 arrest. tnt. J. Radiat. Oncol. Biol. Phys.,24: 687—692,1992.

52. Reinhold, H. S., Fajardo, L. F., and Hopewell, J. W. The vascular system. Adv.Radiat. Biol., 14: 177—226,1990.

53. Fuks, Z., and Weichselbaum, R. R. Radiation tolerance and the new biology: growthfactor involvement in radiation injury to the lung. mt. j. RadiaL Oncol. Biol. Phys.,24: 183—1944,1992.

54. Rubin, P., Finkelstein, J., and Shapiro, D. Molecular biology mechanisms in radiationinduction of pulmonary injury syndrome: interrelationship between the alveolarmacrophage and septal fibrosis. Int. J. Radial. Oncol. Biol. Phys., 24: 93—101,199@

2590

on June 8, 2020. © 1994 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

1994;54:2582-2590. Cancer Res   Zvi Fuks, Roger S. Persaud, Alan Alfieri, et al.   in Vivo

and in Vitroagainst Radiation-induced Programmed Cell Death Basic Fibroblast Growth Factor Protects Endothelial Cells

  Updated version

  http://cancerres.aacrjournals.org/content/54/10/2582

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/54/10/2582To request permission to re-use all or part of this article, use this link

on June 8, 2020. © 1994 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from