AD-A2 6 6 380

First Quarter1993

O3TIC

AFRRI Reports

93-149819 6 3 0 10 i1llrfi'!i~ltlilll

F~rr'• ,A:,prowe•

REPORT DOCUMENTATION PAGE ivn, Ao oN4aec

O0ý0c '000,r-9 0,t I afl'Q "'1 CC' CO c 1ýOfl dl O at01~4.Q Sý *.1 .es- 0 t~ l~o 8" age .-ow ge tCC'qW nCWOI.0~g tt e d' *- -4fl w~ . t - Sea .. I . a" , - a 0. v

gjW t tý'lg v.0 w- ,.n . "a~r JAW aaw n. d and cQ'v.*f-Q 4-d '*.*.t-gl I" C wv.m!on Inonve 5'tnd e~l. g -! '.a -4 - 0, 3-. *VI.c - .'An a 40

Cd~tCOflOf"ltflUlOn C'.d tgoebo to, 1400.0n9 Of'e . lCn D-Oudtn 10 6cr' rglon 'Catdcvl flt0 's SeWýnce, 0O'mN~oe'os .0 sn-, 4"K~''Qa ' o

0.1., .t Vt. S-1 t.204 A".,tgon VA 12202 4302 aneond 111t40"Ce ofk mnagu..~e ndl4Qý0* 5og ~~o 0 ' .. cnP...O A'5 6n.go C

I AGENCY USE ONLY (Leave blank? 2 REPORT DATE 3 REPORT TYPE AND DATES COVEREDMayv 1993 Repr..int, s,/1'ec,,h i (-I

4 TITLE AND SUBTITLE 5 FuNDING NUMBERS

AFRRI Reports, First Quarter 1993 PK: NW,,D Q)AX,1.

6 AUTHOR(S)

7 PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8 PERFORMING ORGANIZATIONREPORT NUMBER

Armed Forces Radiobiology Research Institute

8901 Wisconsin Avenue SR93-1 - SR93-i2Bethesda, MD 20889-5603

9 SPCNSOR1',4, MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10 SPONSORiNGi MONITORING

AGENCY REPORT NUMBER

Defense Nuclear Agency6801 Telegraph RoadAlexandria, VA 22310-3398

11 SUPPLEMENTARY NOTES

12a. DISTRIBUTION/AVAILABILITY STATEMENT 1 2b DISTRIBUTION CODE

Approved for public release; distribution unlimited.

13 ABSTRACT (Maxmum 200 words)

This volume contains AFRRI Scientific Reports SR93-1 through SR93-12 for

January-March 1993.

14 SUBJECT TERMS 15 NUMBER OF PAGES

8016 PRICE CODE

17 SECURITY CLASSIFICA.TION 18 SECURITY CLASSIFICATION 19 SECURITY CLASSIFICATION 20 LIMITATION OF

OF REPORT OF THIS PAGE OF ABSTRACT ABSTRACT

IJNCLASSI Fl ED UNCLASS T FT ED UNC(,ASS 1 Ft ED SAR

NSN 7540-01ý280-5500 Standard Form 298 iRov 2 89)ptUrw bad It AP4SI Ste 239 1829O '02

CONIT:N'I'S

Scientific Reports

SR93-I: Baicer-Kubiczek EK, Harrison GH, Hill CK, Blakely WF. Effects of WR-1065 and WR-151326on survival and neoplastic transformation in C'31,1101-l/2 cells exposed to I-RIGA or JANUS fissionneutrons.

SR93-2: Collins DL. Behavioral differences of irradiated personls associated with the Kvs htym.Chelyabinsk, and Chernobyl nuclear accidents.

SR93-3: Kandasamy SB, Kumar KS, Harris Al-. Involvement of superoxide dismutase and glulathioneperoxidase in attenuation of rad iation- induced hyperthermia by interleukin-lIn in rats.

SR93-4: Kandasamy SB, Stevens-Blakely SA, Dalton T'K, Harris AH. Implication of nitric oxidesynthase in radiation -induced decrease in hippocampal nor-adrenaline reieas": in rats.

SIR93-5: Kearsley E. Energy deposition in a spherical cavity of arbitrary size and composition.

SR93-6: Neta R, Williams D, Selzer F, Abrams J. Inhibition of c-kit ligand/steel factor by antibodiesreduces survival of lethally irradiated mice.

SR93-7: Patchen ML, Fischer R, MacVittie Ti. Effects of combined administration of interleukin-6 andgranulocyte colony -stimulating factor on recovery from radiation -induced hemopoictic aplasia.

SR93-8: Peristein RS, Whitnall MH, Abrams iS, Mougey EH, Net-a R. Synergistic roles of interleukin-6.interleukin-1, and tumor necrosis factor in the adrenocorticotropin resýponse to bacterial lipopO~ysaccharidein vivo.

SR93-9: Steel-Goodwin L, Arroyo CM, Gray B, Carmichael Al. Electron paramagnetic resonancedetection of nitric oxide-dependent spin adducts in mouse jejunum.

SR93-10: Vaishnav YN, Swenberg CE. Radiolysis in aqueous solution of dinucleoside monophosphatesby high-energy electrons and fission neutrons.

SR93-1I1: Winsauer PJ, Mele PC. Effects of sublethal doses of ionizing radiation on repeated acquisitionin rats.

SR93-12: Xu R, Birke S, Carberry SE, Geacintov NE, Swenberg CE, Harvey RG. Differences inunwinding of supercoiled DNA induced by the two enantiomers of anui-benzotajpyrene diol epoxide.

Accesion For

NTIS CRA&I__ýDTIC TAB

- .-UnnnlOunced IJ

Justification

Disti ibution~y.........

Availability Codes

Avail idfdIorDist Sp!cial

A-1 I

INT. J. RADiiAl. BIM... 1993. VOL., 63, No. 1,371 *ARMEDFORCES RAOtOOLtOCY,

Etl'ects ot W\\R- 1065 and XVR- 15' 1 326 M] SurlVd~l and necoplastictranflifrniatiofl in (.:3H/1I0I1" cells exposcd t) TYR I (A or JAN USfISisson neuLtrons

E'. K. G~.IRK BI/.~.(. 11. l1.XkkiS0Nt. C. K. HI1LL'+ and WV. F BLA.KI-lY§

.Reti e . d 1) M ar, h1, 992. iv o on re(, I al I )_ -/ l am- 1992, a1( elb~d PI II, u i~ I /992i-1

Abstract. \\c (ittaontrat-d rthe abilit% of arnmiotfitok ý\R- lah~iti 'I al. 19)78. Nlori tiat. 1983, M\Iilas ot al. 1981.110)65 aind k\R- F)l 1 2b. vatf h t iiton(ltitraition I hiM. to protc) t Grdfina ea a!. 1 985 a ,bh I 989a. 1. 1 ')(1 ,.Nit\ III(duC ill 101 4(1 cr agaitist1 rhv (idIsf ti~llinlgl -hi ,11(1 fII lo 11'1 Gr(rdina~ 1986, Hill r (it. 1986. Nai. a et l1986, Co((in

rnclitt %kith k\R- l0WS %%crv pvrlfornwd with Ntaftonrirl( uhilti.L el elt 1987 , (1irdina anid Sigdesraul 1989,.as- dl1'4of (C311 101)k (elts and at rRcAealtor-genermAre tssion (it. V989, Schw~sartz el at 14988 ( arties and (I'ttittidbiutton herld at dii Arnud Fotr'c, Radibl ologv Rescar, It 1992 . Earlier protcc fion stialics fOloundth hI thci i o~stI nstitu0tc USA . Experimntcns %% it WR-1Id1326 wv rc per- ('fleciv( I c(ompoundfl(s vvrvt pltos;)l0o loat es t hatformerd wNith prolifi-rating lI ltlif(5 of C311 10'1* ((1k and hapovae olr~lt Oatiefe*JA N ILS real to - geticia td isa ion nt1VUtrIit hield at tile Argontinc ru IIccpohr\atd()rdce11t11IrvNational ILaborators I SA . Radioprottc~tors %&(r(, tre~vtit thloj 'Mori -I a!. I 98.3. Sinoluk el (at 1988 , Ihcbetore. during. and aftir iirrdatlinon t~r total periodls of 35 111111 he nchitark radioproteetor remnains \\R-2721 mitrd its*WR-1IS)326: 14)min prec-incubatit ai or I h W\NR-1I065; 30miii associated Free thiol 'A'R - 10365. I1 l ver.i WR Rpte-ito ubation . Bioavajlability of' WR-I1065 arid WR- 151 1Th 1 .51 32 7. a st ruc t u ral an a logueI( of ' WR -272 1 hoein extracellular medium under experimental conditions simu- prms .gis high-LEATIatiniig those of' the( tra nsf ormnat ion ex periments was studied bý as rbef iC inS[itg C( I uall), (i'fhcmeiasurfing oxidation rates ii the presenie ofimachded ('31( radlations 'Sigdestad el al. 1981;. Steriel at. 198710"F! cells in plateau anid exponenitial phase of' growt h ft WR- 15 1326, one of' the co mpou nds Ii vcstigawtd Inperiods of' uip to .5 h. Estiniated half-lives fior autoxidatioti of' the present studyý, ]is an associa ted thiol ofiWR- 1065 or WR- 151326 were approximatcek 8nmii or I Ii X'R- 15 1327.re-gard less of the, proliferativc stat us of cells. ini the a bsetoc of Pors npoetrrsac a e oanmc'AR-compounds. dose response data for transfibrmation intluc- Pors npoetrrsae a e oattnition by neutrotis from, TRIG,\ and JANIS reactors were fitte'd of postulated mechanisms of' action, I ncluding freeto a common CUrve with at linear coefficient of about 7 x 10 -4 radical scavenging, chelation of' metld Ions, oxsgcn(iv. AR- 151326 and WR- 1065 were fbund to provide signili- depletion, restoration of* damaged target molecules(ant radioprotection by factors of 1-79+0-408 arid 3-23 +019, by hydrogen atom tor elect roni donation, atnd thert'spectivelv. against fission neutroni-induced ncoplastit tranis- ehneeto NArpi rcl eoeypofo~rmat ion. *No significanrit protect ion against itentro n-inrd needcell lethalitv was observ-ed. cesses, or a combinationi of thesc mtechanisms Livesy

and Ro' ed 1987., Aguilera Pl at. 1992 . These mechati-isms are generally consistentI with the abilitv tif thiolsto reduce several types of' )NA lesions linked to cell

1. Introduction lethality, and with better protection against lowi-LET! radiations than high-LET radiations ;Nagv

Th, use of' chemical agents for protection against and( (;rdina~ 1986, Si'gdestad et at. I 986. AI-zal anodthe acute effects of' irradiatioo v -as first investigated Ainsworth 1987, Srhvvarrize (it. l988, van Ankervin elhy Patt et at. (1949), but it was only within the past a!. 1989, Mrtrrav el al. 1990. Oni the other hand, cellseveral years that antimutagenic and antineoplastic mutation and transformation are caused b\, n(,:i-properties were discovered (Marquardt el al, 1974, lethal types of'DNA damage. Although the support-

Ing data are spa r~f- som" tn -! of 011(,j~~poeto

*Author for (orresponden~e. ilir tihe endpoint of radiation-induced cell killing1'Department of Radiation Oncology, U~niversity of' Mary- appear to d ifler significantly f'ront I hosc ('or

larid St hool of Medicine, Baltimore NMI) 21201) .IA. mutation, transformnat ion or I urour Inductlion.lDepartrment of Radiation Oncology, Southern Califbrnia Sonw oifI these dlifferenles a rt: t hat, !or these Ia te

Cancer Research Labo~ratory, U1Tniversity of'Southern Calilor- (fetp~( ii at r r sal ihrP temia School of Medic ine. I o%'Angle CA 90015, prtSor fatr '\,ulkhge Pt e

§Ratdiation Biophvsi s D~epartment, Armed Forces; Radio- (it. 19,19, MaVIin el at. 197H. Milas, ti ell. 198 1, ( ;rdina~biology Rvaca r lnstituict, Bf-thesda MD, 20889-SI 45. USA. el at. 1 985a ,b, I 989a. 6, 1991, (rditia and S igdecstad

postirradiationu (Crolmnaetiat. 1989h I pnmT~t'toni al Ii111l1m'1 III (clký peri tL~ik [(-'.I hed .t '~.111ttul I atjo ft1it\

occur at dfrugý coiicvntratitnits hlx%('r thant thio'm td,() 8x If) I(vI~ .uIS per ILAI \ju'linilnti %Tc P~11'[Tepiilr(! fi prKOTt2.tiout t'ouin killing 1, clieniicak or fioui nid dax s alter the Iast nwudmiur dwic.ug Fhalradiatin NI ani oa"t el al. 1971, Viii el ala. 19 i(5 nt n IAwnimniit'it- ittica~.tt'( that ntoav OwNaigv i/ tit. 1986' On)re of' us Hill it ait. 1986 917", of lit- telils m. cen in G, phaM'Epein '.t

previously reported the anti-neoplastic efli~ct of WVR- purfrn~ictd it 37 C.1(065 foloidnng losI l irradiation of activcl\ 0ell survival \%as (If-teiiiivietl l h% tlS-II iiigrowsing (:3H,10iI(Y cells. W\hile this stirdi ctaItxi'' aIbilitv. wshile neopla~tu~ll 11,1 kl trif~ Incud foik~ %s'rC(1erions1ltrawd'( 1 )roltt''~ik atptuist I-t'a\-itcl"W('~ il(( iden'utifiedl ac'(cwdiilg to pudishli-he (hiritei 07"ibi1iplastic transtorinatioui in thle (lii; IA0 assay there e (ita. I 973.h I A R( NC I LiA'A Worki n,, ( 'Ii phas b)(een noi corresponidinig poblishiCd Stutlv On WVR 1985 .'11w endpo~into trAnnsinra""InT Nnr 55isnig

protection against high-LFIl irridiation - cell was ('alculated b\ t he null1 inc-i hod of Hall andIII this report we pres.ent thle results oft twoi mde- Elkind 1979 wt ncr r i e e -~ e ' r

pendentn I invtigations ot the potential of WARL prt- i og to ouir r'.NIdIe anialvsis, Bal ler- K rbm /-k -I al,tectors to mu ntga ic at the eel Iolr level the 1987" Iii add it ion, ss e determiief I thle a eratge(carcinogenic action of neutror(ics Iin porta nt appli- tnumb~er of I rainforma iis per CAsl a% r cci tini riiendecations rnia be estab~lished for the-se agenits in itet- bv I AR ( C I:EPA 1985 withI u ncert ai nt its (Ai-tron therapy. imiitry or occupational exposures ('ulated acecordinrg to the published menit-hd Hcs ii. hi-and space exploration. clcpeitding on the neutron et at. 198 7dlose range within which protection may lbe deleon-stl ated. Protective effiects of' WR- 106.5 on fissionneuLItron-i rid uIced lethality and neoplastic transliir- 2.2. IRadiaprolrc/urnmaton were studied at the Armied Forces Radiobio-logs' Research Institute AFRR I TRIGA reactor 5-2- 3ainopuonpvlatnino cdiv! 1J11(spliulciliitifacility, using plateau cultures of C3H .101' t'elk. acid WXR- 1065, lot BK 71 365 vas ob~taine'd fro ntI ndependently, the protective efh('cts tof 'AR-151326 the )i~vision of' Cancer Treatmniit. Nationali insti-on fission neo tron-induced lethality and ii oplastic tutes of' Health. Bethesda, NID1) 3- 3-Mietlwlamnino-transfor-mation were Studied at the Argonn prXlri(Vrpntho ivrclrd R-

National Lahorators ; ANO JANUS reactor, using 151 326. lot BLAHI.) 1 -"as olntijiid fromn the 1INi&exp~onential cultures of' C3H ,I ffY c'els In addition. sion of ExpednIriu me tllIhrapwf ics, "\t A'!' R reecsomne supporting pharmacokinctic data tur WR- Arniy Medcical (:enter, Washington. DO )C lu ta-1065 and IVR4 51326 were obtained at A"RR I bv thiorie andI cysteinec for pharmiacokint-tic studies wereone (of ts ,A IFL B. Although the same t ansforma- obItained from Sigmia Chemical Comnpany. St LookN.tion assay a nd protector concentration off 1~m~ were MIO. Structuores and select ed Inpvroptie of thest-used for 1both p)rotectors, the studies were not cormdi raditoprotctous. includ inrg halkf", liesfw a uutmIici ationnated. leading to differences in treatmient lproto('os measured in the p~resenrt study, are listed in 'F1abl l.

detaied blow.2.2.1. Treatment ait/i fUR-1513,26 XVR- 1 5 1326treatmient in the present experimients was simiilar to

2. Materials and methods that previously described for VR- 1065 comibinediwith 6 0 Co, y-irradiatio~n Mill ai at. 1986 .BrAelN

2.1. C'ells, cut lure conditionv, and bioa 5say, \\VR- 151:326 stock wvas freshly prepatredl on the da.vof each experimnent as a I -m s tt c k sc it tn in 3nil I(f-

So bconfllmten parent tcul tutres of C31H; lOT! cells phospha tr-liuereci saine PBS andl srt-ieri d b\Iwere used to irritiv t a t xponential or platau cuItru rrs fil trat ion thro~ugh a 0-2 pm filute. It was added ltousing previously reported ruateria Is and miethods prewarmecd complete mcedi urn at a final ioncen Ira-Ila ii _. c El kind 1 979, EHill Ai at1982 Balcer- tic i of' I nmM andI Used to fill flasks l (1 iii befi IreK uhiczek el at. 1987. 1988, 1991 , Balcer-Kubiczck i rrad K lionj. Nlediitm (, 11,4u t :-P noII~5 • t0

aim! laii ni 0 91,. in *JANULS neu tron experr- remnover] utnt il the cells were harvested after the endments with or wit houit WRI- 131326, 3 -4-day-old of' the irsdi'a t ion. The total exposure tinut' to \\'R-

activl guwiri ci I uesat1 ()6 cells per flask were 15 i 26 was kept cou~ita nt at 35 ini it. After corn plt-used. Confi ient mnono lay ers for' IR I A neutitron tic n i f' irradiation and or NV\R -1 51 326 treatniutiriexper-im-ents wvith or without XVR- 1065 wetrc (clls were rinsedI twice with PBS,5 the hit'trypsinized.obtained by groiwing cells Pw 18 days with imleli urn coun ted, andr pla ted for sorviva I andt] ia uoristtualit It

lilt;/ ,t" jt I'll,'1. !~a ], n

(:)(I a!. 11J82 1 ~ ld' 1

Briifly to- siult tý)e cxpnilitioii', ofktklaik tl

-- -. -lien/mih(ltul aIldm in -I 37 C I, lmie oic (LVbor' JA ilI. S o~'ilr A, Illk ypI rie rls iuctial" or"M ' Jia \Q-aeahcxiwme , R 111i wand -ýlmigd Filt, sealed cult uI I"- of\110' c~ er Ia e runder IIitill trogilenI indic i0ll trbeand tre t aliv b ras1iritiocxn-ieit.I Niri e

-20( . Irimclitel prorto se.~\R I065waBtliois, tocr added latr the cinditcieonsIra tionir of it

dissolved ini warni miediumi xitlid rawri firomn at o serumi-coitainiiig c ustomn-made nit-diuni i10"I),mcdiVuIn111-fitlled flask. 'Ilhi thio stock soldution was fr-tal lnwiine scrunin. 90", Eagle's basal uined iurn.sterilizedl by fit tra Lion as abi ive, and then added stanidardco Imiposition,~ brill Nit liont phenol ri'4 inidi-back to flasks to obtain 1 rn.N final concenitratiomn ui (a or. C ultutres ýs Crc inc ubhated in the( presence oIfflasks designated as drug-treatmenrt groups. The d rugs at 37 (C. At varying, intervalds of* up[ to 5 h., 0- 1IVR- 165 conocentration in individuat flasks basedl or I int mdit-urn sarnptes x\ cre -, it hicrawn froinupon weight was routinely checked by titration flasks. The sul phyd ry Ol cotent rinlairiini in"with Ettinan's reagent; the methods are described ineidiuni\-wais decterminend by ithe adtdition of' 10141 of'below in § 2.23. Thioil level ini mediumn determined tO rus IYI'NB iii phosphate b~uffer to t he reactionspectrophotornetricatll agreed well to within uneer- samiple. Ihet( total volumei of' l( t-e a eation solutiontainties with the thiol concentration based uponi was I nil with at p1- of' 7-0 7-4. Alter 6 min, during,weight. which the colour dlevelops after reaction with

.Neutron irradiations were initiatedl 30 miii aftetr DTINB duei to the libterated p-nitrotliiophrenol anlion,the beginning of exposure to IVR- 1065; the total thre increase in abisorbanrce at 4 12 nmrr ais me(asuiredexposure timei to M\- 1065 was I h for all samples. with a spectmrphotonetrnu ElIinan 159 Si.ukli etAfter completion of neutron1 irradiation and~jr drurg at. 199(11 tnar. ~ ir MsOM cum rc g('rated withtreatment, mediumn was remioved, (cellular mon~o- glutathlionle. The thiol content. inl meldiuml front ntio-layers were rinsedi twice with grow-th rnediurn and drug control flask v as measured arid folund tol beflasks refilled with fresh warmi medium before lbeing negligible corripJared to mccl mm that did nlot comereturned to Baltimore, where tells were dissociated ini 'onitact xvidth ce'lls.and plated for survival anid t raiisf'oriation detcr-ininat ion as described previously Bah er-K ubiciekai at. 1987, 1988, 1991, Baleer-K-ubiL iek and Harri- 2.3. .eutrwi irradialinruson 1991.

(iharnirerist ic of' the- AN, I. aiid A F'R f.P issio n2.2.3. 1/ziol djxida/mfl asstay. Biolavailabililv ol thiolis neutron sources were previously described Ma rshiall,it the nine(ia was mevasured using Ell1inan's assay, anid \Villiamlsonl 1985. Z.eman el at1 1988 . Bothwhiich is based on the( reaction of' sulphyd ryl- JANUVS and IR IGA reca(tors gcuriri~i iicit nis

c'ontLain ing radliat ion protectors wvit i -5.5'-di t Ii olis- with1 average energy of' < I \ I ceV, and y{-nitroben,,oiccidl IDNII or Ellinan's re-agent, conitarniinat ion oh- <5", nors iiso'terdo

Sigma, St fQuis, N10 ). Our prlyoctocl involve a prnteIr Oke 'l s of %VR -106.5 \vith Ii 1k A lin--

410 E. A'. I ihrcr-Aubil .zck ct Id.

troll ,, i rradi tai 0 ns we're pe'rformeid using a (h T'-raet 3. Resultsof0'. I G'., mil Iat doses (f <0.9 (;' anid a dose-rateof 0-3 ( [111 1 at (l(s5(* >0,9 (; ;o t.hat all e'xpo- 3. 1. 7'kjul uxt~hijot ýtudlo's

suretuft' X~cr < I0) mill. InI'I et(iiC ~radloi pr( dccmc iv leefcts ot'W\'R - 151I326) %~it h *JA\NUS The sponIItanieous oxidation (& o a %ariet'. of ,jul-ticitt'in., all irradiationis Wetec perfIormed ict singi at phv~drvl compoun~ds in) tnetia li' has bee studied b\.di se-ratt'of (&02,5G milsti (I Gr-lita t-f al. I989a. othler investigait irs -Held andl M ekltr 1%87. Suzu ki 4f

ti!. 1990 ,hut nonie f, (ties(. dIata wer prvicusloibtainted for (- C'I fI' cclfs andi or \\R- I1065 orI

2.41. Da/a ania/v.us W k- I5 I326. '[hel( rate (it' thioll) oxidatio~n i . rifztt';.which afi~cts1 protecttor efli( at'V mfay. dependO on

The C hio.l oxidlation clata a poioled f'r( m individual several fatctors, Inctluding ty pe of'rmed ia. huller. (ellexperiments wNere anialvsed by log-linear regression, den'isity, proliferative state arid type of cells, as wecll atsanti half-I hie timies . , for au toxidat i ni of' each the nature and conten C ration iifthe thh f Biagfov.s etcompound In miediumn" were expressed as a reciprocal a!. 1984. Held and Melder 1987 :. Our purpose %%xasof C he slope timies Ini 2. The standard errors for the T4 to estimCate thiol levels in nmedi umi abOVe theill' UIs atvalues were calculated Iromf the slope varianlce, the time of iieuC ron irradiat ion, replicating ats ( losl-I\

Pooled sets of dose- response datat onl cell survival ats possible the( conditions described above fr- t ranis-with or wit hoot the 'AR thiol were analysed using a formation experiments. Two other well-st udicdsinCgle litiear-q uad ra Cit equation with at least-squtares Chiols, glu tathione arnd cysteivi. were inclutded forfit. co.mparison wvith the lite ,rature data Biaglow ef (it.

Quantitative analysis of' dlose -response curves for 1984, Held and Melder 1987 ýcell transfornmat ion was perfbOrmed by the sirnulta- Although the between-experiments alue." vari( driCOUs fitting of' transformation dlata obtained with considcrably, die hierarchy of autoxidation rates of'ne'utronis alone atnd inl combination withi a 'AR each thiof, show&n Iin Figure 1, wýas retained Iin allcompou nd, T[he proitectioin against neutrons, experiments. There was no clear differenice betwee-nexpressed as at dose-miodify~ing factor D)M1", , was the rates measured for a given thiof in the presence of'calculated fo~r WR- 1065 atic4 WR- 151326 from data e~xponenitiaj or plateau cultures, indicating a second-obtained with and without a radioprotector present. ary importance of' the prolif'erative state of' cells.We used a modification of* the three-parametermethod. originally. described for determining oxygenenhancement ratios f'romn surviuval data (Pike andAlper 1965. Harrison and Balcer-Kuhiczek 1980).Briefly. sets of'observations with or without a radia-0tion protector were treated as if' there were a com-MOnL intercept for both sets of data. Two linear Uclose --response curves were then described by three z

paramneters (a common intercept and two slopes), 0.instead of' the four (two intercepts and two slopes),

needed fo~r fitting a straight line to each set of'0.1Robservatioors separately by the standard method. LThe simultaneous three-parameter fit is a statisti- DUncally rigorous and efficient means for determining 007

protection factors as the ratio of slope for theradiation-only curve to the slope of each radiopro- 0 1 2 3 4 5

tector curve. The standard error for DMF was TIME (h

determined by a slope variance ratio. Details ofrequired procedures, including a test for the suffi- Figujre 1. Spontancous oxidation of' glutathiiine A.WR-

ciency of' the common intercept, were similar to 112 ~>csen n R 6 ~.a 7thos desribd inPikeandAlpe (194).Accod- easuired in the prest'nc(' ol( C3K l(fJ cells, D~rugs were

fos dsrie inPoolead Ale(16.Acod prepare'd at I TmM in comiplete mediuiim wtboiji phienolingly, freach poldset of the data with and red indicator. Snlphydryf concenTtraitions we-re merasured

without a protector, the three-parameter fit was at limnes indicated on the abscissa. Show.n arc valuescompared with the four-parameter fit by the P'-ratio). relative to thte maximumin. Solid lfines are the best lti to

the dat a pooled fi oin five to s'vci' i'xfer ni en ts wit IiFor the data presented here-, no gain in the precision t'xpuinentiallvY growing or density'inhibitted (ultres.of' the slope estimates resulted by adding a f'otrth fHf f-lives calctihited from these data are shown in 'l't.blparameter. I.

I ~ ~ (cit %itl h fife\ot', i t-.u. t, pi irt(-(l wcpairatc t l x 6

100 1 ANVS a 11(1 I Rl(A. nctIroiti-lrradliatedl (Sil If I(elils lull el al. 1 982, Bdic('-Kiilfl(/ckand1( fl~atrsi

U 1983, Bakecr- l\ it Ibic/.k e al. 1 988, P09 I . '1' wh la( k ,if

Fi <r 2. _aji'ctableiept riipoteultiiallx Itia rsuhd

cor 010 dof expo cntial\ growlng (, .111Fi 1195,rNagasaw III a!. 199

exposed~~~ ~~ toARInorn wt rwtot3.3 %%fas prcIiotR ihw, n-on tcl/ lotn sc/ar~nian \tincR- 1)6 an-1 101 C.li Marugs 1971 flan prprend nt

th bes 10 to the poo9e dat inIbe2 rtc ls aotrteas t. ornii~ n chu c iA sea to r thei l ts

NEUTRO DOSE(Gy) esponsenta cultresobagrvee wdin it corrlentponding4ofConsistent ~ ~ ~ ~ ~ ~ ~~CH T wihtepctrnrpre yHldad dt o latea culturs. the p-resen Tresult uprtsal

of' o-106 tor simiarton tthevlue f Aor ryste~ine ' edn ,JNSdt o laeuclue rwhl terates26 for WRcit-15 ibt 1326OT HIawysbtee vill frcmprsnth posed foyte Fine and fori glutathione wtheu haf-Rf The nur-onl dat were poonle riatOr acufor 6 eac th/o show inTale1gs thee weightevd mealns i inonoemdfighitr iue n ;of 11pcc theT 1 vaues obtainedg from r the tie-ouresen of Cnsistenrtr wis(th ourprevous dtransformto r3- 02 eulls.oxidaton meauredg ain sfequenaditial samplesin foromthe tedsoepnecre-itdt h obnddsaeolask Ofspecial Finterest3an wasc a 7 solid diffrnce is obtained in thpeismsu ntws comat ib-lve prith lierityrinthe ct faitestor WR-06 poedaandW-53 in thee l roow-ol at deat rangeNU of- nR euctron dosesexam-

whoe cemial truturs ae vry imiar Tabe 1. md.ar 'e cummrvzedpoidda inTbe2 igia slope-etiate lt orCompriso of ur pesen daa wih th itratue xaouetia 7 culturansfgrmeet per surivth orellnding

renssulnts fowgutthtione pandcyteinreponfirm tyHedatd thea aoreementawithlthoese rheprtedn presiuslt fuportthiolde autxidtio rlateson vronsidiedraboly, whithte JAU ou r TRrevnetros fining aeotdfri telkn 1978,r(chemical compoitiond ofidy Tisse cultues mexdium,) Hills (i a!92 Balcer-Ku hickel e al. 1988 1991 .N ie

dhoemfornstraingte aned For gtthese mesueent hl-ine Ahenexaintion-ofl dth efret pole fWr 151 326individachl mhodel sysTemsleg Iithed anihtd melern tVRo 6 predsen mdurying iraditiotsi Figures 3 and4i

of he ,-vlue otaied romthetie-curs of Cosigniicnt wit reuce theviincidencefofrmansonrmsuatst,

3..xifdatof mesre ihil n ellquential sapermte texpoe-riesptse with e fRite netronsh prombined more

Naeutrsk.On survival datas a a7f dfee obtained in txermnti512 s~rtudy with comatile wth ineartryinthat involved poroetn W~R1c065 are shown326 in (the lo-t moert 79ng +frcto 008.).Figure 2;mia strradditional e in ermaio simea (Table 2). A nd h uv rvddar nta lp siaeC)ie omparison ofouprsn datapons with ore witeroute aot7X1- rnfrat e uvvn el IWreut forpound, indicated rtioe appaire stht prtecie 4greemenussion toerpotdpeiosyf

dmntaigtenefothsmesrmnsi Aneaiaino'teeflect of' WR- 151326 or W-16 t1mi hsresvulsagrewt modl sstemvts from. reldated experi These6 present~ iduings suppraitithocntralgurest3ltnd

mens wthC3H I '1! cll reortdiearier(Hll i ill C3H/. 198) thats Todrae tliiol6 prtherap can

aN.u186).neurnol survival curvobaies wnexerientsis mitigat proitool wiidthe cell S tr nefurtrotns

412 1F. K. liabtf-KuAbo -e4 ct al.

ui-ulru and( call do( so at drug (lttss Itwcr thati that IcironttiI kt-~c( onIII, I~i ndI\ Idual d ~ta (-I, I thlr 2rveiniiTel fif IMIttcutnon itoaiit; cell killing. F Owtil-, ohtajllr1 InI tin-' per-a, .11 ctttiincit-"x it.,-o I-

protection ag-AIIt'r tar.TIIO( (gr11it rhiruts III "Iz Ma\Iair rclk, ý cvc \ iIrtlall\ lilt h .m 1i a- ill- Initial ~iopcel al. 1078, Crditia ct al. 199 1. ( dr11Si I R i(l iT(:thI Iitd 1O'I~ld 161i I.Xp mcit-nial IIIl~t I I týo wI Iht~ Ik tIII Ifi Ai--

1992. lo TI IttIiit II( -x pt- r I I IItI.It-I- a Itdi 0(1 at .\\I. 'a

InI the presentI c~xpvrimlt-nts, protection itt \\-R- .\FR RI hbctoc HIill fal . 1982, Bale i--Kuiihh /ck and1065 and VR - 15 1326 against icopl itstic cd I t rails- IaIitrri-ai I 1 183 , . I I ie It i iI I\ Ball vti- Ku it I rk- al.hformation milas charater~tized h)\ 'lopo (ItlgIcs oft 1991 , and aihcr hiah-r-Khidmirk In Intprallat-tl1i

transforiltat~io inductionl (IliACS Nugi cstini that, In lite xpelillinrt- kN oh radljttjrtttot-.)[ \\i-n tt-cotitlh~iitttiln with flissI)lt tirtitf1,ltt. X\\R- I1065 andI dlill-l-irit dal~a 'tlhi)r'' %(.ItI ,talxk scdI Ow\ih liticra-

SR - 1.31 :3 2 a Ire I )s c-Ii iIo:I I i Igý: t I ItIs, no oss, o t Ii : idIat i I Iith(l. d d, IIII(rrncrs11 ario I I ItI-VI It-atI1) 1o tCint 14)IISho1UId br I ItIt-I. ICII (I1 10% low I neuton It-lins %% ri-c rss; t han 11 imt stanidard (I or ann ,ti du(fo.tcloses. A Ithiougih this result IS at CuiSClsqulicc o f tilh- q nadra ic term1s rr lw lci4l ivihtir. lIIIN tiInI

conistra'inilt tha t till dalta m, rrr lit to a1 Ii tilar II(dlitud. atrI- thla t, w ItIont m rad IoproirTir( IIffs. dIrI do "cniex-cri-tless In lile~ pt-csvnt cast. this 1iltid( I canl hr respoiise It1iiitittit 10r trai~st,0rniatiuii hx AN 1, orac-cepted ats at good tpp)roxi ilia tWitt. For our- previous AFRR I iwtitrtiis obitainted x ith pndlih-raiiii orneut ron data tithr qunadrati mc (1 mponilrtt was nrg- otati ar\ nU r r iiiar

ligihlc [fill et al. 1982, Balcrr-Kuhiczrk el a!. 1991 With ergard to thr rtlati\ vi calnir o WR - I1065

For the limited nei n-ls-iortrrata, oiily a and \\R- 151 326. stonir limited i silirhs ranl hrlinear fit was I IIIamjustified "in the( statistical sense. ated b\ ((~lliularnig (the Cirrrsporinding, Iihi-

The sltipc v.alues fotund t0r jANL'S and IR IGA phosphate tlerivativcs. WR-2721 iand WR- IS I327.

Table 2. Pooled dat it trmi vxperimcrt-iis oni ntwplastiic trancsformnation atnd surit val of 3 [1 tt i vIls rxposrd to I tum \%R (ibio

and OTr fiSsi~tit ne u t rins

Conldition Dosc GN. SU' Pfh C+ S Xd y- 1 d f, +- S- 1f j)4 I-

AFFRI 1 RIGA fission neultrons

No protector Co :ou-rrciit 0 -63 280 0 280 23(9 of <0- I3conirol

0-3 0-67 + 0-12 174 11 163 2 25 6-3 ± 2,0 P-it + 0-870-45 0-43 + 0.08 191 14 19 -1o :307 7-3 + 2-,t 2.14 -+-0-660- 75 01-35 + 0-08 229 :31 199 282 I 351+2-2 5_00 + 0 930 9 0.27 +010M 33 7 513 288 299 15 -7 +2 -3 5-26 + 0781-5 0-11 +0.02 138 34 I0N 228 24.6±+4-8 10718-2-10

+ WR 1065' 0t -0-58. 142 0) 1412 218 0 < 0.2804.5 0-64 + 0t-I1 203 2 20 1 247 0(-98 + 01-701 0-40 +0.200-7.5 0-28+0)1)8 103 8 96 397 7.8+2-7 1,17 +0.7120-9 10-26 + 0-06 3101 12 299 317 3-9)+ -I -1 1-14+0 361-5 0.12 -c0-03 335 -11 296 353 12.2 + 1-9 3-5 1 + 0-592-t0 0-035 + 0-It1 230 38 202 :31 1 12-2+ 2,3 t- 17 + 0t-79

Argonnec Laboratory JA-NUtS fissýion necutrons

No protet tmt Historic eonniol (1-7(1 2425 6 24 P) 244 0-3 + 0-1 0-110 + 0,0140-21 (0-82 +0-12 20)7 7 2010 228 3A + 13 1 -51 +01.5 70-41 01-89 +0t 12 158 11 14!9 227 7-0 -+ 2-I 2-58 + 0-8 7(078 (1-63 + 0-08 I146 2-4 124 275 It)-14-4. j 51 4 4- 1 27

WR 1:51326" 0-21 0-88+ 0-37 81 2 79 294 2-5 + 1-8 0-8.5 + 0-320rt -077 +0-25 51 2 52 216 3-7+2-6 1-_75 + 1 -33

0-78 01-6:1 + (0-21 49 6 413 378 12.2 + 5-0 3-4t7 + 1-.11

'I rmt WVR -116.5 add ed 3(1mitt prior to n-n itron irra diat ion.- ( t- Is kct-t at 37 C , total x psta tS~t irtit I It.-')] mmt 'R- 151326) adlded It)mitt prior toeutrtoiitt irradiatijont. ( els ket-p at 37"C. totial expostire time 35mln.,Su" PE - Sitrviv ing fracitontr plating 1-f liit-ic \ is -"A'N. Af, 1-. P. toratI nttiinfr of dish-cs, innb-wr of' tranorisaenits tof type 2 tr 3 : Re-iikolf P/ (if- 1973h,. IARC!N(I:F.P- Working

Group 1I8- tii(( -nmber tof disbics without irarisf'oriermtn, tm itir of cutlls per dish t-rruct--d for platiing ,fliP-wieitv of tin-eadiancd -ells, ortso iv i vi , Ira( tio t and pl atin tg i-fu it Ptt- of irradiate-r i tl s

"F~rati-iruttiuaitirt rat- pe-r dlisli JA:R :!N(: TI:PA Witrkintg (Croup 198(5 and its staendardl i-no Ilicbtr it al!. 19137-- Irarnsforrrtmf oin rate- ptr so C\ vfr -la n aned V Ikind 1479 7) antd its siat rte ~cel riror~ e-Ki)i to- alt 19087 -

tti)r protec itt Ion atttmt fttt.h-l;F radiatiotns. Fill thecitiidpoitint o atilwaf Ictthaditý,. Steel i! (it'. 1087 ?

OfiM'rý, d no) marked difit-renieT In pritteciti onbetvs ce WR -2721 an ml\\R- 151 327 alit-tr .\FRRI R: 6~ A I

nieutIron Irradliatioln. I [()%\ cv'r -- WR- 151 27 %%I,;~

mot re vf ectivc than \\'R-2721 Ii in fit stud., v of SitZdf,, i(e a/ . I 986) \w'i I tt) t I~ (I t t o u ( e to

fields I)( S.R. lea ft h I hvsic' ResearchI Reactor ,it , "

O ak Fi(Itig Nationa 11,1 tforat (UX. O ak Ridge TIX, ~and( JAN L.' S and x% it h hgt-nrxc lt tp)rod1uced neultronIs I'LRM1II..B 0

We do no( t fiax direct evidence tiat. at th 0w

Crjiol0ar contentr'atmots. ViR- 1065 Is motire 'fkeC- NE; TRC'4 DOS.E Gy)

tI\ci ll it t n \\R- 151 3261 A *ttt-nijts ti( cxplaiiit the 1-.Il 1. 1-Ih Iti l ., - I Il'- ct li IIt' ý4i.it I''i b\(Ifser\ ed chiflcetttes Iin 1)N1- for' WR- 106 amb \Viu R- t1~Ir ~~tii~Itt~itiiI~.I liijciaaii"I .t1,'w~ai1 51326 Iin our1 txperimnttllit are- hindered hýl toweCr- wi Iit .\FR RI tttill 1 a\ Ril R- 106 if I0 mm %i'

tailiits inc txpr possiled contig but ions1 on f' RI th peodtI)l irifulltht ' , It Ifit di a llw, (ip- tlit. tl-Iir'a aflc "It

Of thetse factors. the proliferative status of ccelk secents III Itiqnb~itimo %ith k\R- ]KIf' it, 7 1- IT - 0 Itthe less irnportantt, sitice no coIrrelIationt betvernu-t the ( o a 2 11) + 1) 22 ý lit ( 1%.r'p tioi- Ix ldit 4 ' ~eflect of' thiols and the cell cycle distributionl was tItI kIi-1(Itli~i r 2 .di- 162, (. tick dim - I p -I xohbserved Iit the( Stuldy Of' MUrray el al. I 990( . 'Ihe(tIAttras siS 1tidiedl the( developmen t of' raidl( pr()tec-

1(111 in cilro ats at itt nction of drug pretreatmen t timeit prtetion dowlvt yJmeit Ifor thesw IV R r hiot \N afo(.r varilous thiols. including WR- 1065 and( WR- similar: at mittiri ut '30-nin pret reatmenomt \%x aN1.51 326, at 37 (C. lhev fouind that the time-course of* required f'or each of' the drulgs lC1)(I(ir reach mlig a hitt

protection lee.Co mpa red \ ilth W R- I (65. I -5 -fold higher ext raceil a r vo nt en tratioIn if' \\ R -

____________________ 151326 was needed to a(+c~ict thle samc lexel of'

0 4-AWR1136]otek-tion. These cliflt'renees, xxii 11I arc co nsistetn0 ~~ overall Wxith our resuilts. could be atttribuited. at feast

Z Ili part, tol di flerreces Iin the( kinetics, of'thec red istFnbaI-

JANU NEROS tori(f the thiol~s T'lable I and Figure I . hose- drug's

6'- that oxidize rapidly elistetine. WR-'106,5 also reachZ anl eqnih Inium of' inracel lular and eXt raCelIl uar

ftjl comptfartmenets rapididl. InI sonric cases eqifuilib~rium i.,:Fmpet within I mnd nCalabro-Jotsc ! 8

"V) Dennis et al. 1989,. While tite support itg data are

0 ____ limited, the correspoIndinjg trend for thiols that oxi-'0 05 10 1 5 ~~duze slowly glutathione, \\R-1.12ýcnb

NEUTRON DOSE 16~expected. 'l'hesc ctonsiderat ionts make It temrpt ing tospeculate that the lowler IMhF we observed for WVR-

Figa ri 3. 'If'l (efftect of t\R-15 I 21) ont cell transformaItfionI by~ I,51326 coulId b)e dute to at lowxer In tra-et lutar d ruttfitiioran tc at ori .. I h'll ric Ut im-o n-Iv (lose -response curvc co ncen tra tiotn, (lie itt turn to1 tne pnarmautogxtý of'represents it fit ito data tor cx-<oirt('tltiatl\ growing (:311 thle 'om pounrd cofu pled with the sitortcr preirradia-10T tll iIs cxpoed to JANtUS rti'ltrolis t., and toi ti11 drug exposu re ti me.

detI'iit-itih 'itiwdc (:3111T cells~ e''1 xposed toi AFR RI Frcl iln n nna ehltt~o rtctwiiltrars , A = tranflstrinatrio ireu trciacits ini'100ri -l illo n nmllehlt.tiiwvithl JANM'S ncutrons wixhen WR-I 51 326 at Imm tive ca pabilityv was reported to d curt-atse v ithf anlWva, Madvdi 1(1mi beifore the' irradiattion. Solution increase in LI'Y Ii somle sttidies Sigdestad it a!for the( Itliri'i-pairaintctr mrrodel: common ti.Titer- I1986, Steel Pt oal. 19187 , hiltt a mitre compltex lmi t fire-

r -jt l! 0. 1 + 04 t(2 x If)t 4 rn~rminssixv emerged from Ithc studtx of' the proiteci(Int b\y 1 R-i-l., slope's tar n('Itroitt-or~ i T i or t~fitt(rn iatioit 'ixih ,

WR- ~ ~ ~ ~ ~ ~ ~ ~ ~ 7 TI aT r 081t06~ o~~; r ~ I gaintst t Ih eflects ill high--llrurgv heavx_85+f)_-31 X x( 4i G (s risfpl'tixils (IiI,(dtt's of fit: chred patrticl IbeItamts :\AC il and AiutisxxtIhtf 1987

(Iii-'iitairf - 21. (tigrees oif reeitint Ip.-0.7,), A s~im1ilr 'onctlusionit I'tgardlitg jptt itcctioit ill ft-ga I'd

41 A'. K. Ha rA'br;tAi at.

v; LET d catr'tatt IWuld bexi~ic noub by "npaiiig ourit B.-x 1:1 it-Kr I 41cii KIL. K .""d I I %RSUNitiirs H, !W'i . WIo

peetnentron data in Fig~ure t vith pt'xrviousi guii55 11r1rrrirrrrior U I (2)A !0I ji rilk ir X-raýr'". irla

detieriniiied transhrrimattrio da~ta liii' \\R-10(i) andic tp; Irrl ncuo'n, % l itý Y''-i -,ill 44 fit itr"IS i y

trrw-LFLI raulratioIi Hill el at. HM96 .Although WK- Kit j -Kfs/K L K uid 11,AHRiVN, (, If Pill!, tour

1 06.5 atuutrdlect sigiili('ti t'ddiorr1 )otcttiii agains lis- Oivkm Tachlif qair rjrar r mu \rf i,-l. o tr\ IxtirtI(,tt

sinor neutrtton radiatirit DM1' ~3: a pt'ictiou fak- Ed.itorr irý \INI, Fkisund nudi. 'rni' ýIrdwavsrr

toro about 6 was re'ported tot' this thiuri, based ott Bvii 59, 117 1182

Hllt"'' t ransformuatio n results with t oxx r LEA xIV :j 4117,:/<K RI;.K i Hbm. .I ~ir ium.iN msoutirce ( tin a and SpIgt iad HM ,)~ W're nra itt', hr iw- HH 01 of, welyl wit~ d t" jImr"o "dr "ati radrmarever, that scexorat celul at' s tudieijs Alimvwd WR - 165 lrsu'nmar lna! Irrnr 4 IRadwimsnIh 51, 21" 22f)

1o gi\C anl CeqUatt\ o'let'ti%'C prrIte'tnOTr againist BAtImm it'ICIFIij . L. K.. HithuidlN (i . ZIt W if%r G, !I

Inutiation after ex~posresi to t an- hii1tigh-1LEA' ji. NI A I ION. P j . arid Kr NK, A.\., PI8. Ian k a1 iuorV'lr-

r'adia tiotns (.r din t0 at 1i.I985min) 198 9 ,1,i Nagy arnd .nrrsrr rb :i iimn , Wo *,ctia,na/ laniAu Razat

( irdina 1986, H itt (ii . 1986. Napv el al. 1 986, monI of (3H 514 534r.Itemoa ,wdqRuivl

Grdina arnd igdetad t1989 Cteartly the vilitur'v 4 Bmt, ERHKviiurtFK. E', K , II aHRtsil. G It. AND 16LI I ,K.

'AR rcomitpountds ut r )r'tc t agaittist neoptasic' Ctlr'(ts 149L Nemrutir (lose-At, v'\pr'r aut'r i' Al TR1R

should he further itiv'esti atet rover mm utn,~~t suuarr'l Radraism~ Wa~nk 128, S65a SAI.

L ET.', dw. -Btnm aig f HAGItAow.j. E., Istvo. Rý IV GI-Hr\ýc% L, E,.. \ 1.i ~XjAr;OHsr.'-, B., \Ivwirr.i Irj B. ..st( Rv~so .\.. P4'84.

IInUrnular\, Ohe ('nIt'rgting C% idCttICC tQt Pt'MCO 01'f F'actorrr inflrsr'tilrrs' till oixidationi 411' r\'tvir'rrnim aridagainst inutatiot artd cat'(ititgenesis, p)rovides a othecr tihkrl: implicationi toi. hpvprtfrrhr'rm s'CIS311aiu/srr

mo re'(t Ot imist ic picture fi r arni n it Ii t a pptic at ion a r d iatadjrii pm tro tion riIradiatr'n Xwf'a m/ 100,than (lid the earl ier diata for pri lect'ion against acuite 298 3 12.

cv totoxicit v. Futrt her itix'cstigtO sarne of CA RJrNEiS. 1) XL. AojiiFR'A,. JA., \\AHRD_. F._' Slmort Kýcv ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ G t1t~ni arnitdo riio). alnn FAtmiy, R C_. 1988. Uptake' of \\R -2721

Whether radkrprotector can be administered at tow. dem~alivi' 1, cell ini rulture: idrittra~th~ru -0 thne trumbduises so t hat sidne-icets tan be avoided. ai rd jprrtd t6rr oF' th drrug. (.'anr Reoain/. 48, 3634 1640,wvhethtir t he catn atso be admin itstered tiller UnCX- ( ARNiýS. B. A- and (,RDNrA, 1 J .. 19¶92, Ipn ti, pi' in-u on Iý dit)'

peCtted exoo1st-res, aroinnwhjo VR -2721 aginsfats trerrirri-iriduned cmgr'ti'.'s. Inilernatuma/ Journal ,/ Radifitiwi Riobn,gl. 61,

CoRN. B. XW., 1.ui':-R, If, L. anid ,I I rij. B., 1987, DiP' 'rciliialAcknowledgements Il'ofrtradical ucax(n'igl'rs on X-r rx'iidrn'dudmtto

a nd~ vtotrxirt\ itýin Irurirair (r'lk Radi~awin Rntarch. 109,

This wiork w~as sutppor'ted byV Grants CA' 50)629 1NNs 4)4 1 8.and A 1808frot te Ntionl Cnce Intitte.XI. F_. STRATroRD. NI. R. I... WAKRIAIsAN. P). arlidand ( 4288 tro tin' Natiu mt Caner Itistit u o'. AIPA, R. V. 1989, In~cm ii ~i- ittnt rai'i'lular uxt'nr

and by- the( Armerd Fonrces Radibiotogy Researc'h .ii~nr r'xpiursir tin rithiorlthrvitr[ nimpikatinuis fii radio-Institute. iDet'esc Nuictlear A\gencyv. uindce' wwrk unit Iliigs . Ifhatwalmnal Jirw jntrr/ o Raditairn J/oor56,

4620. 'The authos thatnk Dr Erie Kearsttv frw his 877 88:.

hielpfult dIiscu ssionts arid commienti s. arid Dr' Ni ncila LtiANMAN. G.. L... 1959, Fkii. 82,1111d~ 70 Oljl 77.ýo

Lonx i~r!up~n NR 06 ath tuyW GRtiINA. 1). .1. aiid SnuAWSi-'rs. C . P., 1989, Radiation proter'vvish to ackn'owledge Ni s 1) .M \ N i Nrsbrook rA FR RI ý, tors' the n' rrxpnnti'd lir-rifits. l)rugi Afel'talma/irr Ruzew,,',

M~r \V. A. NIcCready arnd Mir J. Klaff IiNIAB 20, 13 42.School of Mledicine for extpett techinical assistanee. (;RDNtA. 1). J., Pt.RAiNr. C . (;Aiuýs, H. ýA. arid 111ii._ C K.,

Mir G. Hohmblab for the c'f'ic'iert runninrig of' ttn' 1985a, Protrto t vfir- tfhi of"' S-2- 3-a ;i not p rr pvlarninror'i~lt phosphirrrurhioir acid agaitnst inidrr tionl of altered

JAN US reartor, and Mis B. A. Williams fir the hcrpator'nru fori in rats tr'iu'cd irrir r sithi rrirtnha radia-dursimectrir' support at .\IRRI.ii (ion nttr one drn ativ kithbi . Cancer Rewarrauh. 45,

5-3 79-538 1.

References ti'.~PRATNOr C~., 198.51h. ''llre radii rotiriie~t' r XVRI 065Reeecsred uces rad iatio nn-i ndmoeid mu rta rtion at t hi' It vpo xani-

fuint, 1ihisphiiriboisx'l transiclrasi' liui'i~o iii X7!) tells. ('Or-A'FZAI. -S,\. J. arid \NArN-oR'ur n I'..J. 19'87, Raliiopr'rr' lcntiol rrnoenpvrt 6, 929 943 1

of rtouriri crilin\ fo riming uniris-splvr'rt agairnt' hr'anv' ( DIrA. 1). j_ Sii I'SAti.AD C . I. arid (:ARNFo. BA. %_ ;910aratgr 'P parr itivIdarta, a n "i iXR 2721. Radio/un Rpwnorh, Ptit Uionnb I, %R 114 arid XX' RI 5 132ti agalrsnin st fiii-

109, 11 H 126. nieuitri ititnr (d ruustatirri at tit' Iigprt univ it \'ns.\It 11A. .1 \_ NtFUwir, G L., [sir -i'. R C . arid %\At,_ ce Jrlls. Radralmnr Reverir'/r 117, 5001 .510).

I. 1Ti'2. *fill, t urn ptaki' tin (hines,' hamirnrs (en'rilk anir (R[Itir'A, 1), J. NAGY, B.. tinii r C. K. anrd Sio;It'Iorx . C, R,

arrol rirtnir taIi ir~t'rtirir In ; a furs, tinri of rtse lict 'hr~r gi' 1989'4b Priri rtionr .rgairrs r'adiniturioninirdi id mmina-(tn the niriol Raduawin Rreearih_ 130, 194 201, gvrur is, sri \'79 i illk h 2-1 unmoroirrrrpJ.

lt'I op it o I nei? III II ' tio p la t I/Ii' I I an ýJr ma/i 1)t I

(p 11 11. j.. a\nu H I Iri f'v i~ H- 1: y''i I',o I-sit I' I (I/w 1 /- 1 a:f t .! 58,11ý ý1 I II

Rr&if ýn RI -6 0?,n 11 ,.1 .111, _`)8 ' ~ ~ ii'7 i~ 57 571

Cp 1) 81.,\R aw I a m A f- H A,s.U ' 1 I OU. P"Mt- 83,al rI li8 58, n I 2 15 I

mmi- 1. n \% a Ri -i Ii Y 07 a a. I'6 11 s Uvu- tit, -, a ,ifLc tt fiutsV iN~s ~ H.. l) , " R I' lu I it i ( ls .1 1) 1 1 1 6, i

ii. in n T ~11' 11o! leiw~ t i n , pilx'nirff8, (git I flo o ( iinwin-4nli-,11 hua'qulaticiciii i I ,I hytitisO

hI'v' A a -nd , M s N I' - 1 Ki,'io I /h/ mA, ma, Hii of69 In( i~ I ti I X 0I I kipi N 1 1- j- a. Mi pI ii- t M 1k .A Mao '-i io.IO I.

Ca ill .i. F` NI. 111662 [i-iimand-ipuioirrd ul-itons ii 1o0 X6. 21dwm Baq 71 4

ar tI I(~ I I.~ a- rIIIII it ln l ro I in tR ,acisori~rih 39, ~ i X.,(% B and \i D'a< I . (8 ýPf Illl I. At-l~ frd , n,1ii2-

*12 3 ' 298 11 16 utflcil-iim li iiojnl afn mrs hoic i /l .Isyruc /wn' l~~oit !i37Hi wwt,(. (K It AM-v H m I Bp Pit K.-i L1 .cn (dPOMi~ 1) and '181,.~v 2-I 1516 ill

.\tgclnil ,p rii flicnli-tfall,) Rl lfV (181 - B inic aro andtl R i.Kicll-. ( . He srill Kn w.t I) X iit iu- -lI/iH

Ill I I i. -in)nd \ aIt -I t x, 1)i . i g`117 c hwlc it thg 1((1 iu i-n m W9 6; ~, P lli-llu. l and hiiai 1) Mwt W!of m O!,l

.5 " it I dir ill. ,pline ("I 7, in i i t,,! Rwhazi. nIii oll' 4 I-i- alfls! ,-d if-4 iI- it-I-n , , . f

Re a L 1 12, V A lt~a Al and-i' 18665.ý fli i-Iiiyc illcc in l, 7";iliciu (id I -i'l by i-ioi, ram ical " .1% ch 33,

aIA,1 oI idan ) ,ii I' I inc icii ý I", o n -. tlIjsccd l ( Ina n of - 46,-uii- C .13i 2r.'u .1 -, w~ias. 1 1

D_' Ii ~i-cIk,-I 1 ou Rdmim-A(.nh h. 128 ,r/, 45,I 238 1 1 38l.-.de N"OM I ili 1. MIJ~iri~oiiiii oT ( i and cc N inn-m Rf

Ilis HEk. 1. IN .111( PI in (. I f ,J. IN. S '17 (:66,-rij L ,ol i-tjc1989 m ous i'f-mbro Ills I il i.c w nt 1ý i In l ififlah'i I- uH, tcli,

cgci nI H Ii ki.IA . \l,, !`Mi ''i .111 Io an, dzlc I i ll" -, C IIC lzI'uv tIll flyhoti-l denl6 1t-i ndi 6an'-d R,,,ooiz 33liat and fnoI il

13,ii' -2, 851. 869.v~urt 11 X 1 , (IuNR p FL - B . 5. I N I, , R I II tI S m I 1. j1i.- 1

NI fisi C. K . B1'r i-i F.MiM EN. C. 1'.d HA, Axr-.-o -Ind.178 acif (i-rnINjju I). J.. lion, -JLal' \cIicr~adiairn I Am Iicoj

fill I II_ B n,- 1f RAIN1 Cl, l andn GI?,ArIN . 74, 1. 1 65. dAHiai \7)cll.R ili2.'ac.13, 5 1N~psc.. . . 8. ard Wr~ ~ `u\. 1 5. f1G`. iNlor noplas i n Rft- 7N;WIAI)P,. A..P BR(;HDINA. I). \\. arl M IDI- 1.-,id I

r.a to i rlci-ariol.re el~ 7d 1. 07N W' l'lt)8 . ailanX...In( of C166 FroA (tmcaloi f, alm rl itin, to~l ooI.

MAR91A ii NC .I ST Wl)RKI\K. NI. I. ar 191C i-)i. Cellula an un7-6 17ni.62. inhibitim n /?of (I/i. G06, 221eh 233.lIdai tilt ai, inv suoiiin-I c CI ira ,nornia geo -- i- an iadruucd Slu- ~ K 32. 323.. Miv . .i.sic-i~-.I.N o

ty~1XR it iil lilt- rin ta it cits ail-er Icd,',ach 34vt, I CA, A..-\ a lX~il .S Bv_1616. R ln 1) W. %lilt

1187~itol fI r llvli 6.60 ficll-l al' ciriv% aus HyDR-272FR 0. M Ni (ja iiiai~ fand q l ilo-r~ii- ornITs. . ~'17 prvis .~ k It. mrn Rmin,ud1 ýt r -Ih s 45, r2305 231m 1ndrug rsof lsli,-l taifor nait~ihiollo (7ue l Icd filemof 48,

jt C an .I VRi-lit r1) ri . PC'e,iiica Chunia ra i onco "Ius Radbr-,l ',ml sewiii to .6647.iticin ill

a.ak~tlI iorI.llzin radiatioS.Aimf In 110)l Riil l~i w i- (.anrer t'gc of carldiv,. ji 109,, Rr16'1h 336761. ,2

13, i~.0 44, 55,7.56 II,i AR. J.. Ln . 1-> Q I ZI , S M Klu K,,s ARRI(,ls 1'.. Coiio .KM1Rl R_ . i ExIA (:.Aand.IImmKAW -CM..N IKAI)M. 1978. anid D~iwr. J_ 1988, 2-18 Annouxp iropI(~l~ amtidl

\i~vR22 tk v-bits dirvative u arid ti-irridiopronreing5- iadits tiran i-s,;nioni-,scc1 ( 15 t-iviltor ii rdu~,lti (sf hrl.6no leton. 1 Cm a n,(,hrif drals i an itirivi. lntrverehc7,n 11 115.rl2 darnitgritn1Iil \'79 cells.tlWas. Wa M 113,,'ca 7,211515 ,8

-I_ 1. R.u anrds'v 44LAM1_ -S, ;485 56MI UI" SMI-,rx . ( IT v-si . N. . 11i , ( 1 s m A. A and I-ui. I -l'91.

NIRVIM. Ff. - Kr INI. M. I ~yNI t an A nu iX. N iSs. .194 51o 27. Raidi,.,I 6n Reeri-i 106, i 21, 2i33.rnlcc~l-s II '~i'). try icr yll;ltcan ,fn- d tCI thoi l s (uie iv is " reintoL Imo ~i K. G D..~ g F ~iiiI RlCdo. 9, M_1 Aw81

1)%I~AIB - ~in cod t aff.i6 -IS 61ing sic Crinue Rsar chd 34lI, J.Hsls I A. tlAnd WARDN. 1,1(. PiN ti, Hadipoc I-615 io -f

/., & I /,~ -A ubi' c1 t a I.

I\a ' .\NIHI~tA I N Iia •t \. I. .Aiid Wt Ay I). !cW<. at \IR RI Ikl t , I I I/I/t<.

Ill, ,-tl oll i t ut tih ti (A ii'," hill'io- At, A , Owt "., I ,' ; 23, i 2i.i i t it , It liti d tW R - 2 '0Jt 1 ll h htn II , '!, i .ita tid D NI)X

ll'.llli.{ll, ~lt~ i il 1111(mll llll, hr!"~ f, 1,,l .; ,wl, o i'( ,,f

RE$,EARCii INSTýUTEMILITARY MEDICINE. 157. 1)OR 9 -12 *c11m4s1ic RIpoat

SR93-2

Behavioral Differences of Irradiated PersonsAssociated with the Kyshtym, Chelyabinsk, and

Chernobyl Nuclear Accidents

LTC Daniel L. Collins, USAF

Three nuclear accidents besides Chernobyl have occurred in ....the former Soviet Union. The accidents occurred aroundKyshtym and Chelyabinsk in the Ural Mountains between 1949and 1967 and contaminated over one-half million people. Thehealth ministries are now interested in the data previously col- '95 7 -•-lected on these irradiated populations in order to examine thehealth (e.g., psychological, hereditary, genome damage, etc.)implications of long-term radiation exposure.

Introduction *$'erdvsk " yy s

In September 1991, 3 weeks after the attempted putsch, I rep-resented the Defense Nuclear Agency/Armed Forces Radio-

biology Research Institute (DNA/AFRRI) as part of an officialcontingent to Moscow and St. Petersburg, Russia. I met numer-ous scientists who were and are currently involved in all thenuclear accidents that occurred in their country. The majority ayof the data contained in this report comes from discussions andbriefings. In addition, on December 2, 1991. several scientists Kajs4from the Institute of Biophysics of the USSR Ministry of Health.Chelyabinsk Branch Office, who have studied the Kyshtym ,.Me ,and Chelyabinsk nuclear accidents for decades, presented hu- Kshtman data that have never before been released to the West.Consequently. the radiation units of measurement reflect thoseprovided by different scientists and vary (sieverts, roentgen. r7echca Rivweretc.) throughout the paper. 194,2- 5Q.2

The radiation situations in the area of Kyshtym and CAeyaln sk * 0 io 20 1 .Chelyabinsk are unique because over the last 40 years masses o -,- 200 .........of people have been exposed to -"°Sr in the food and water Lttchain. Basic dosimetry investigations have focused on thedoses of 9OSr. The first estimates of activity were from nuclide Fig. I, TheKyshtym/Chetyabinsk regionof the FSU.

measurements taken from the river sediment. These analysesbegan in 1951. 1.5 years after the accident occurred.

from reprocessing. stored for not more than 6-7 months. fromNuclear Accidents in the Former Soviet Union (FSU) which not only uranium and plutonium but also cesium had

In 1948, the USSR began operating a plutonium production been removed.plant called Mayak in the Kyshtym/Chelyabinsk region (Fig, 1), Due to the confined nature of the blast, the majority (90%1 ofIn 1949-1951, an accident released 3 million Ci of radiation the nuclear waste was dispersed near the tanks in the form of a

into the Techa River, A second accident occurred in 1957. liquid pulp3 However, a plume cloud with an activity of 2 mil-southeast of Kyshtym. when improperly ventilated storage lion Ci dispersed its radioisotopes over the area shown on thetanks exploded, and 20 million Ci of radioactive waste were map in Figure 1. The contaminants from the plume cloud werereeanksed epodhed at0m ilosph . Te stofrageioamtlex waste we confined to the Chelyabinsk and the Sverdlovsk provinces. Be-released into the atmosphere. The storage complex was lo-foetecidn.mrthn2,0poleivdnte38i-

cated 1.5 km from the reprocessing plant and consisted of 60 fore the accident. more than 28.000 people lived in the 38 OFunderground storage tanks. The radionuclide composition of (ages along the Techa River. Contamination levels of '•1Sr ex-

ceeded 0.01 Ci/km 2 and were distributed over 23.000 km2. Thethe fallout indicated it was comparatively fresh nuclear waste highest concentration of'"Sr was located in an area known as

Metlino. located near Kyshtym. The doses in the Metlino areaArmed Forces Radiohiology Resach ln haveraged 3 Sv/km2 . The dispersion of radioactivity from the

Arrn~d orrs Rdioiolgy eserchinstitute. tBethesda. MDl 20989-This manuscript was received for review in February 1992 The revised plume cloud is shown in Figure 1.

manuscript was accepted for publication in June 1.992 In the spring of 1967. further contamination of the TechaReprpnt & Copyright ',, by Association of Military Surgeons of U.S. 1992. River occurred when a severe drought caused the highly radio.

Military Medicine, Vol. 157, October 1992 548

Nuclear Accidents in the FSU 549

active Lake Karachay to recede. which allowed the wind to Relocation of Irradiated Victimsblow the contaminated silt and sand particles over theKysht3m, Chelyatinsk, and Techa River areas. This lake is Of the 38 villages along the Techa River belore the accident.located west of Argayash. A total of 600 Ci 1_37Cs and 'K'Si was only 4 are safe to inhabit today. Analyses of the people alongreleased by the air transfer of sand particles froni the Lake the Techa River revealed decreased leukocvte and immunreKarachay beach. Radiation was primarily gamma radiation system functioning. This is attributed to the time pernod whenalong the Techa River and reached 5 Rlhour. The average activ, people drank the radioactive water and ate contaminated foodity along the Techa River was 10- ) Ci/I of water. A massive daily before their relocation. The time range for relocation3-month clean-up effort occurred between the summer and au- spanned 1 week to II y'ears for the evacuation of all residcntstumn of 1967 and resulted in a 10-fold decrease in radiation from the 34 contaminated villages along the Techa River.I Thelevels. This occurred because the Soviets removed a consider- total numbr of people relocated during this time exceededable amount of the water from the contaminated reservoir 1o- 10.500. After the people were relocated, the villages were incin-cated near the Metlino area. cerated to ensure that no human habitation would occur in this

highly radioactive area. However. the heat and smoke createdby the incineration process further spread the contamination

Discussion over the streams. rivers, and lakes, which exacerbated the al-ready contaminated food chain for animals and humans

In the aftermath of the Kvshtym accident. if a village was Although some of the residents living in the contaminatedfound to be radioactive it was scheduled for evacuation as polit- areas experienced a protracted relocation, others were quicklyical circumstances dictated, In addition to the area dosimetry, evacuated out of necessity. Following the Kyshtym accident.which measured river sediment for contamination, human 1,154 residents of Kasli, ne;ir Kvshtvm. were removed 7-10dosimetry began in the summer of 1951, measuring body ex- days after the accident due to extremely high *'"Sr levels. Peo-crement and contaminated clothes. It should be noted that dur- pie who were removed during this time now hate twice theing this 1.5-year hiatus between the accident and the begin- acute mveloid leukemias that the control groups have exhib-ning of dosimetry measurements. the people were neither ited.informed of the accident nor of their internal or external expo- On a historical note. the construction of the radioactive facil-sure to any form of ionizing radiation. Consequently, all the ities was conducted between 1945-1948 by approximatelyfood and water consumed during this time period was contami- 70,000 inmates from the nearby gulag (prisonl. The Kyshtymnated with radiation isotopes, location is N 55-44, E 60-35; the Kyshtym restricted area coy-

In an attempt to further quantify dosimetry from the ered 2.700 sq km and contained eight lakes with interconnect-Kyshtym and Chelyabinsk accidents, the bones of deceased ing watercourses. The Kyshtym atomic plant was built in apersons were exhumed during the 1960s and resulted in the tunnel, which extended beneath a river, with only a smoke-creation of a data base that reflected their "lifetime" exposures stack visible from the air or ground. During the constructionto 9°Sr. In 1974, Soviet scientists created a data base using live process, one lake was drained, a building was built on its lake-subjects to determine their whole-body doses of 9Sr. The bed with cement. rubber, and lead, and the lake was refilled.'methodology used to make these determinations from the liv- During the Cold War. several high-altitude reconnaissance air-ing Techa River residents included dosimetric examinations of craft routinely photographed this area. In 1960. MAd Francisurine samples and frontal lobes. The urine samples were ob- Gary Powers was shot down by a surface-to-air missile whiletained from 1,500 residents living beside the Techa River. The flying over the Kyshtym and Chelyabinsk atomic facilities.12.500 living subjects involved in the frontal lobe study alsoresided along the Techa River, The teeth of 15,000 living sub-jects in this area were also examined for 9'°Sr doses. The loca-tion from which these residents were evacuated is known to- Control Groupsday as the "Post Box Chelyabinsk-40" area.i

The dosimetry revealed that 1,000 people living by the Two control groups were selected for comparison purposesTecha River had greater than I taCi of (!)Sr in their bones. Fur- for this longitudinal field study of irradiated humans. The con-ther analyses showed that those people born in 1932-1933. trol groups were located just south of this area and were notwho were teenagers during the first accident, accumulated exposed to the radiation. The first control group consisted ofthree to five times more •Sr than did those who were adults at 34,000 persons of the same socioeconomic status as the vic-the time of the first accident. The maximum reading of 6 yCi tims. They did not have access to the contaminated Techaper person occurred in those individuals who were teenagers River and were not contaminated by the plume cloud or theduring the time of the first accident. Measuring the metabolism other accidents. The second control group consisted of all non-measurements of 9Sr and overlaying them with all three pre- irradiated people in the greater Chelyabinsk province (i.e..vious accidents provided the following results. The dose levels those not living in or near the contaminated city of Chelya-averaged 0.42 Sv along the Techa River to the southwest, 0.52 binsk). The control groups within the nonradioactive portionsSv along the middle portion of the Techa River. and 2 Sv near of the Chelyabinsk province consisted of 1.5 million people.Chelyabinsk (Akleev. 1991, personal conversation). The three The Chelyabinsk Ministry of Health's data base has resultedaccidents affected 437,000 people. Of these, 1,200 people ob- from an ongoing research project over the last 40 years. Thetained 200 reins over a 2-year period. In addition, some people following information is a summary compiled from 33 years ofreceived doses of up to 400 rems to their bone cells. data collec ion (Akleev, 1991. personal conversation).

Military Medicine, Vol. 157, October 1992

550 Nuclear Accidents in the FStU

Impact of Ionizing Radiation Accidents 8-9% of the persons who began work before 1958 and receiveddoses greater than 100 rem died of cancer.ý' Cancer mortality

Results showed significantly increased death rates along the was approximately 88% higher to, those who received moreTecha River over the last 33 years. Coefficients were described than 100 rem than for those persons receiving less than 100in terms of excessive risk per Gy. Stomach cancer rates were rem.3

4

found to be two to three times greater than in survivors ofNagasaki and Hiroshima, breast cancer two times greater, andlung and esophageal cancers two to three times greater (Deg-teva, 1991, personal conversation). Research is in progress that Contamination within the Chernobyl Nuclear Plantexamines the progeny of irradiated mothers for stillborn chil- In addition to the early incidents of ionizing radiation expo-dren, abortions, and miscarriages. Analyses thus far indicate a s n a t to and ear incide more io rmation egpr-significantly greater number of birthing complications in the sure at Kyshtym and Chelyabinsk. more informaton regard-irradiated mothers than in the control groups. Today. near the ing irradiated persons comes from the accident at the Cher-Kyshtym reservation, where the town of Kasli used to be, the nobyl nuclear power plant. Following the accident at Unit 4.ground surface still contains from 1.000 to 2,000 Ci/km 2 of which occurred on April 26. 1986, the Soviets were faced with-0Sr,1 It was from this area that 1.154 previous residents of decisions that would disrupt and redirect the lives of personsKasli were rapidly evacuated in a 7- to 10-day period following living in the contaminated areas but also those working withinthe accident. In the past, several types of military training the damaged Chernobyl plant. In order to minimize the dangermaneuvers were routinely conducted in these contaminated of more contamination to the residents living in the greaterareas. Chernobyl area, several thousand persons working within the

In addition to the ionizing radiation doses, the victims of plant would be exposed to significant doses of ionizing radia-these three radiation accidents received little, if any, human- tion.itarian support. The people also lived on less than a well-bal- In order to remove the highly radioactive debris from theanced diet and received only rudimentary medical care due to roof of Unit 4, it was necessary to use humans since the me-a pervasive lack of medical equipment. There were and still are chanical robots ceased to function due to the high radiationonly 50 hospital beds to care for the 500,000 irradiated people levels or became stuck in the radioactive debris scatteredfrom the three irradiation accidents. across the roof. Consequently, approximately 3,200 humans

called "biorobots" were used to run onto the roof and throw theradioactive chunks of debris into the radioactive chasm left bythe exploded reactor. The biorobots accomplished this task "by

Contamination within the Kyshtym and hand" since no other method worked. Their voluntary effortsChelyabinsk Nuclear Facilities allowed for access to this area and the sarcophagus construc-

tion to proceed.In addition to the radiation exposure to the populace outside Clean-up work needed to be carried out in other places that

the atomic plants, intra-plant contaminations also occurred.3.4 were much closer to the melted reactor core. Workers transitedDuring these formative ycars of discovering how to work with areas that were 3-4 m from the damaged core being protectedradioactive material, several personnel were exposed to vary- by existing concrete and lead shielding. These workers wereing doses of ionizing radiation. Those individuals primarily at exposed to between 100 and 500 R/hour in the Block 4 area.5risk were maintenance and reactor personnel. This period was Overall, more than 600.000 workers were involved in clean-characterized by an operational focus that minimized safety up activities and the sarcophagus construction. Those in-concerns. 3 Several persons manifested the symptoms of acute volved in consolidating the radioactive waste from the "hotradiation exposure following the start-up of these atomic facili- zone" into waterproof trenches were called "liquidators:* andties. Many persons also suffered symptoms of chronic radiation worked 2-week (or/oft) shifts in the "hot zone."sickness, which resulted in the creation of a dosimetry pro- It is interesting to note that the anticipatory stress of thesegram.4 people was high but so was their motivation and goal orienta-

The dosimetry programs recorded that numerous persons tion. Many willingly and repeatedly exposed themselves toreceived 25 rem or more performing routine repair operations, ionizing radiation by being a biorobot. a liquidator, or explor-The 25 rem or more doses were received in different time inter- ing the damaged reactor to discover the extent of damage andvals (e.g., day, week, month, or year). Consequently, the rota- locate the missing core section. When the melted core sectiontion of these irradiated persons to other jobs not requiring fur- was finally located, the solidified mass of radiation resembledther exposure resulted in an ever-increasing demand to fill the an "elephant's foot" and registered 10,000 R/hour.5 These peo-vacancies created by the safety-mandated transfers. The irra- ple had a modicum of control over whether or not they woulddiated populations consist of approximately 6-7,000 persons be exposed (volunteering or not) and could largely regulatewho worked in these atomic plants before 1958.4 The irradi- their exposure in this manner. The presence of this "perceptionated persons were divided into four groups: < 25 rem, 25- 100 of control" is very important to the maintenance of goal-ori-rem, 100-400 rem, and >400 rem. 4 These irradiated individu- ented behavior. As we will see later on. the absence of thisals have been monitored for the last several decades to detect perception of control results in feelings of victimization and thefor the presence of any health anomalies, cascade of negative psychological feelings that resulted in

One-half the workers at Chelyabinsk routinely received 100 heated demonstrations, even though their exposure to ionizingrem during the late 1940s and early 1950s. Of this population, radiation was far less than the Chernobyi workers.

Military Medicine, Vol. 157, October 1992

Nuclear Accidents in the FSU 551

Psychological Aspects of Nuclear Accidents Nevsich, Iput. Besvoad. Braginka, Kolpita. and Pokot Riv'rs

It is noteworthy that people living in the Kshtv and carrying radioactive silt into the Dncpr River. The entiregrid of power stations on the[ Dnepr River down to the Black

Chelyabinsk areas are antinuclear as evidenced by the volun- Sea is threatened with 60 million tons of radioactive silt, as aretary shutdown of two nuclear electrical generating reactors the 60 million people in these regions.during June-July 1989.1 Also, although billions of rubles The anticipatory stress of these people is readily apparent(equivalent to millions of U.S. dollars) had been spent on th.e and can be easily understood. Interestingly, according to anconstruction of a breeder reactor, it too was shut down approxi- assessment by 200 scientists from 25 countries and 7 multina-mately 3 years ago. an aftermath of the Chernobyl accident in tional organizations done for the United Nations International1986.' The reason given for shutting down both of these "nu- Atomic Energy Agency. stress-related illnesses are caused byclear" facilities after years of use and construction was the psy- lack of public information about the disaster and the masschological animosity that existed in the people living in this evacuations that follow. 7 The psychological stress that fol-contaminated region. A similar psychologically induced result lowed in the surrounding areas outside the radioactive hotoccurred in Moscow. where a new. ready-to-be-used nuclear zones was "wholly cisproportionate to the biological signifi-power plant was prevented from opening due to the public cance of the radioactive contamination:*"7

outcry. In addition, a nuclear power plant approximately 40 Even when no radiation is released from a nuclear accident,km south of St. Petersburg was closed to appease the psycho- but only threatens, as was the case at Three Mile Island (TMI).logically upset populace. The impact that individual percep- the Kemeny Commission,' 4 and other documents.6' 7 15It, con-tions of ionizing radiation have on society is now being real- eluded that mental stress would be the main effect. The psy-ized, 6.7

In other parts of the FSU. specifically around the contami- chological findings could be criticized, if used alone, for theirpotential self-serving function. To avoid this perception neuro-

nated Chernobyl region, 35,000 people in the Belorus city of chemical analyses that measured individual stress values wereGomel went on strike on April 26, 1991 to protest the fifth employed. 6. Vi.7 By using this multidisciplinary approach. weanniversary of the Chernobyl accident. Similarly, 60,000 peo- further clarified the adverse effects that exposure or potentialple waving "nationalist" flags packed the square in front of the exposure to ionizing radiation has on humans. Consequently.Sofia Cathedral in Kiev. the capital of the Ukraine, and de- since increased stress and associated behavioral alterations oc-manded punishment for those responsible for the world's worst cur from anticipation as occurred at TMI. and when actualnuclear accident.8- 9 The Rukh press agency reported similar exposure and deaths occur from radiation. the psychological.antinuclear demonstrations in Kiev. the western Ukrainian and behavioral actions of the FSU residents are easily under-city of Lvov, and in the Belorus city of Minsk, as well as demon-strations elsewhere in the two republics that were the worst hit Thus, the workers within the Ktshtvmo Chelyabinsk, andby fallout from the accident. 9 Thus, hewrears ithien thed yt Chlyabink and

Reasons for the psychological outcry among the Chernobyl Chernobyl nuclear facilities continued to voluntarily functionvictims are numerous. 0 "1 The delays caused by scientific and despite knowing they would receive a significant dose of radia-pictimscal disussionsfinaThe relaysculted in theacitionofi adtion. Conversely, the populace of the FSU was extremely con-political discussions finally resulted in the evacuation of cerned about being exposed to any dose of radiation. and vocif-40000 Furthesid rents from 00peoplelivingn an areawe contamin n w erously demonstrated to close nuclear plants that "threatened"Cl/km 2. Furthermore, 10.000 people living in an area contami- them with potential exposure.

nated with 15-20 Ci/km 2 and 60,000 people living in an area the differencin ehvoret

with 5-15 Ci/km 2 were not allowed to relocate. It has been The difference in behaviors between persons who were vol-

estimated that the rate of thyroid cancer will be 5 to 10 times untarily exposed and those threatened with exposure to radia-

the normal rate expected for 1.5 million Soviet citizens, leuke- tion isane"mind se ss we ing principle to explain the

mia iates among children in some areas of the Ukraine are 2 to chological "mind sets. " The unifyse proups to ain

4tmsnormal levels, and the death rate for people working in behavioral differences between th~ese two groups is a function4 times what it of their control-oriented coping strategies when faced withbefore the accident)° Furthermore, the scientific director of their ionizing radiation stressor. The importance of this coping

and control construct was first uncovered at TMI regardingthe zone surrounding the damaged Chernobyl power station individuals who experienced anticipatory stress to potentialestimated that the disaster has currently claimed many more exposure to ionizing radiation. (Note SeesCollins, 1984 for anlives than previously reported.' 2 The best estimates to date eposuretorionzingradation (Not See Clis 18foan31indepth doctoral dissertation that describes this "human" con-regarding the Chernobyl death toll is that approximately 321 trol-oriented coping concept). The knowledge obtained aboutpersons have perished as a direct result of radiation exposure. the nuclear accidents in the FSU has extended our understand-

Perhaps the worst aspect of these nuclear accidents is the ing of this construct to parsimoniously explain the radicallyomnipresent invisible threat and the continuing fear that the different behaviors of voluntarily irradiated persons from thosefuture is marred by irreversible cancer or genetic defects. This involuntarily irradiated or experiencing potential radiation cx-may have increased since the accidents. when radioactive fall- posure.out contaminated the environment, animals. and people. Anundeniable and continual reminder for the residents is that allthe timber in the affected areas is radioactively contaminated Conclusionand cannot be used for furniture, for construction, or even forfirewood.''3 The compilation of knowledge from this article regarding

In addition to the forests, the waters of the Pripyat, Sozh. the different behaviors of irradiation-threatened or -exposed

Military Medicine. Vol. 157. October 1992

552 Fraclurc'd (lfiring Up(-raliun *lu',! au

persons results in wisdom to be used appropriately in the fu- Referencesture by mission planners, medical personnei. response teamn I St f f Br I h,haa k,-1 I n a- a Z. w.'1.members, special forces personnel. Federal Emergency Man- ~a 26 :12 4991agemnent Agency personnel, and others. This psy ,chological 2 fiurudzý,n AlJ IL slsIiihriii :''' -

and behavioral information can be appropriately used to our ''w ~urU 11aa~aa iI

advantage in the future to better understand and predict the Ow' Kii:Utpai1mudi iRr.,ooi' P47'4 r, f-' o'

behaviors of highly trained and goal-oriented persons vs. the 1a1 l ") i aisiu~i~!~ 1 '1d'r' ' ~ 7

behaviors of a population that perceives itself as being vic- 4 Nikipvlv I% izloý AF Kushuaikoia %A A;1f 0;, !1!'!.

timized by radiation exposures. Failure to properly assimilate itiu:rradwitonos,-dind UU jx,nnn 1;.! if 1. lU.

and apply these key psychological constructs. which explain hnir.lpHrnIi iv~i, \'kljlli HA,',!, MA

otherwise counterintuitive behaviors. especially as they apply dixrlins )1 values in Mhe Mcda sl Hav oairndioi ý4ý1i 1 v' t 'lIi : 1 -ý

to potential nuclear adversaries le~g.. terrorists), will risk the loi a I 1,ser .c Edad it, hr kit. k, R( N r4, M E V~ 1, P" -. t

use of these complex aspects of human behavior being used ppl)7-;,

against us. 7 l nterim Iiwia Jihrrirfo Ir r I Im , -Ir(I I r 1; .1 "rt I-li', 112 2'11'. I

Consequently, the situation in the FSU is likely to become of IAtrniEnvf-4 Agent ; 19,41 pl)277 41.5 [ialt týI 'r(ll n C. Top I herirti sl ra .-rn'11t'd nw Or 'A It: ! iý h

major concern in future y-ears as the future of 60 million people sewnt tlaolrt rarnp Thc'Aas~hrinkiton listw 'lW 140 7 pA I

is adversely affected by t e hrnobvl accident, and another 9 kr-ri-us Norts lstrrr - hrrnrrlrrl tr ali litr , ,.,:&1.

million people ( 4.- are adversely affected byl the three accidents Aprzl27 1490 pA6

in the Kvshtx'm and Chelvabinsk areas. The psychological. 10) BarInkt-r F Four Years later- Sovi-Isr-trWa Aid' r st. q- Iih', mstrr

physiological, and epidemiological implications of these disas- I I TxcmG A cprob rAial ol*x-n &I c f dban ein!isf

ters require further study. 'ost -Julv8 1987 p A I

12 %Vise %Z L N n-port biatnus .ir(-s,.u rnirad lw , ýrr(taIh r~ i -.iogton 1si Mar 22. I 191 p A25

A 3 Malukovskv N Th, ;, *)nsof (hito inbvi li%!-.tr Ma,,u h 26111'$10PiAcknowledgments 114 1i-nmei J(' BR-port of th, President C(,mmrniss.! (,n ! h A( ,r id v 11f1 h'

Island The Need for C hangtp. Thu Legat -, o TMI Vrura IAhrfo 14711I wish to thank the numerous scientists in the FSL for sharing 15 Coln DL Persistent v d~iefret-n,. tya-wIrvn Thrv W', Wyatri and

the historical impact of these nuclear accidents. Thanks go to the group AD-A145.%7. National Tec hnical Irournaidriil 'A rVI spurrin VA ilkiAFRRI Director and Scientific Director, CAPT Robert L. I, Davidson I.M Baurn A. A-oiLst DI. Slurs.. and wrood urr o rln~a'U

Bumrgarner. MIC IUSN, and E. John Ainsworth. Ph.D.. for funding Island .J Appl fsvi roi 12 349-359, 19N2this international research effort. 17 Collins DL Baum A. Singer )E Io~o r ,urrir th orn~ sit 1li. Vu~ 'x;x p ,~

t hiilogcal and b~uu hernral e% idf-rcr, Hetalth Psi Purr 2 144- 16t 1' 4 i

' P 1 Is f l Iun t I Ihi s c - I ! R , %) i i W , h f d ) I I~ I '-1

ARMED FORCES RADIOBIOLOGY

lRIFS Ishl' RESEARC14 INS¶tTUTEOCIENTIFIC REPORT

SR93-3

Involvement of superoxide dismutase and glutathione peroxidase inattenuation of radiation-induced hyperthermia by interleukin-I a in rats

Sathasiva B. Kandasamy ", K. Sree Kumar " and Alan H. Harris "

Bvia t ;ral Sr , 101l( Wid t ' Radit•fiir'n Ii0t lioINItrnr X'4i-i.mt•itls, .4rmcd Fori-s tRahotwitihei RI-at h Insiauw. lithsc.da. . ) M I) s,5'/ If S -1. ;

(Accepted 0 ( )clober 1992 )

Kcx wiords Antioxidant enz.nmcn I Ipcrthcrmia: I •hp0th3ilatni'.. IntcrleukIi: Radiatioin

PretreatmQnt 'A ith recomhinant human intcrlctukin-- (rhiL.- In( 20 h beftore irradiation att•nnuatic, radi oniti -indu•cd hIVprItC mrmia I xpciimenlt t'Aere conducted to determine the role of antioxidant enzymes such as superoxidc dismuase (SOD) and glutathionc pcroxidae (WSi1 IN) inrhIL-I -induccd at•enuation of radiatlion-induced hyperthermia. Radiation exposurc increaNed SOD) and decreased (iSllfIx lccels in theh,,pothalamus, while treatment 'Ailth rhli-<Inr increased (GSHP lcvcl and had no effect on SOD le,.ls. l Io'cver. rhldL- -I and irradiatfontogether increased hypothalamic SOD level but prcented thc fall in (iSllPx level. Our result,, suggc.t that attenuati,,riIon raaion-inducedhypcrthermia bh rhiL.-|I nla. involve stimulation of SOD and GS(iIPx because rhl[.- In treatment and irradiation together increascdhypothalamic GS1I Px and SoD levels, and intracerelrovcntricular administration of SOD) and (iSIIPx attenuated the radiation- idijcdhyperthermia.

INTRODUCTION mice from the lethal effects of ionizing radiation and ithas been reported that the radioprotectant effect of

Exposure of mammals to ionizing radiation causes rhIL-lat may involve induction of endogenous

the development of a complex, dose-dependent series MnSOD 4 '.of potentially fatal physiologic and morphologic changes Exposure of rats to ionizing radiation induces hyper-

known as acute radiation syndrome. Critical cellular thermiatJ. and the hypcrthermic response appears tohiomolecules are directly damaged by ionizing radia- be centrally mediated because irradiation of the head

tion or indirectly damaged by radicals generated from alone causes these effects whereas irradiation of thethe radiolysis of cellular water''. The free radicals thus trunk only does not". The hypothalamus is the most

produced can further interact among themselves or important site for thermoregulation. Radiation-in-with cellular oxygen to perpetuate the radiation effects. duced hyperthermia is mediated by prostaglandin E,Protection and/or recovery from the consequences of (PGE,)ta. Preliminary experiments demonstrated that

ionizing radiation have been investigated at different when rats were pretreated with rhlL-la 20 11 beforecellular and molecular levels. DNA repair mechanisms radiation exposure, rhiL-la pretreatment attenuatcd

and chemical radioprotection afforded by thiol com- radiation-induced hyperthermia". Experiments con-pounds have been studied extensively"'2 Radioprotec- firmed rhIL-la attenuation of radiation-induced liv-

tion has also been reported to be conferred by endoge- perthermia, and the hypothalamic levels of ) SOD andnous radioprotective mediators such as interleukin and GSHPx were thict measured to determine if the atten-immunomodulatory radioprotectors that elicit inter- uation is due to the stimulation of these substances.leukin production ". Antioxidant enzymes such as su-

peroxide dismutase (SOD), glutathione peroxidase MATERIALS AN) METH)ODS

(GS1-Px) and catalase, offer protection against ionizingradiation-induced oxidants''. Pretreatment with re- RhlDl.,t was a generous gift from Dr.ý P. 1 omedt of oltmant

combinant human interleukin-l (rhlIL-I a) protects LaRoche (Nutlcy Ni) and was diluted to the desired c-nic-ntrati•r

(orre~pondenne S.B. Kandasamy, Behavioral Sciences, Department. Armed Forces Radiohiology Rescarch lnttitutc. licthcida, -D) 20tSS-51-15.USA. Fax: ( I ) (311I) 2956S52

I I

III J'%rigeeItAiree ',litte hct01tC injliCL101. ['(iI ., ( isiIv l SOD5 . aind all I)r10I iiiii BI let %,Ict li. h M11i11i,1 %11 IIc I -I'l~l III I i'tIf iillv I I

oihc]he m i caiiti.l,'1 iii eICjktti', uised weiIC of tII, lite Clt ,mitahstatilk nun 10iic 'tatl ttI II tI \)It . i IIIIltte ~ 11111d IklI A~.'51 M II A 1 1%, I

Inr.5 1c. sCId nbtiricdi hornl Slk,,1t1i (SI I outs. Ml( )). jiild N te All, tIcIIII tci ýl l 11L71ftlk~ ttciitasiiI p Ic 'YSI W1KN ()Is S111111'.l,,1scl t p ii t-te saline hetorc Injection. 11t l ite I 'feltl s pitStle'. ( I I I 111Iseicit1 ill 5111tiH]i

tip 11)vul, aiet iifiil wCId Ci) itite111d MIttcltiit Cflt ' M~i I lu ills% 1111 AHi (teL-c i

Mlae Spr~tucu Dawles rats WCrl (l)tSl)BI);1 (hines Rusei CilIiNAT) dioxidt! [IN 1111a,11,1it 1

Bittccdinc I-hor atornes. Knigts n, NN I weightniog2 -31 acr

iiiarantined il in-maln andil scriecnci tfor c~idence (tdliwl V.(110 ilclisi ito lpiti/iidssCriilol.!N and ttistiipatliilogs, The iiats %Nder housed ndi dii~tdul it Prti oi taa ecnic h h td iiIltilripitlicarlonait Mwiro-Isolator cage, (I -ill ProductN. Miss', ool. N.l onl tiin biri citi iuri il f ilte sialitlild Siaitst1k,1 te tlUAIdIoiiilsaIC.iiiCta idh,lrdisood contaict bed-ding Meti ('hip. Nirihcastcrrt ' utidetkC 'lciit'i i itfc tl in~i~fe cProduct, Cor-p.. Warrenshurit. NY)I and were proMIdd COMITInCtA~il (it P' 0.0 )h tcitroziip ciii slsSi!cpI li e 151 trodent chow (\Vax ne Roident Bliik. C ontinentail G raiin ('Ao., (hicaiin.It,) and i5 11cr ad Uhbiumn. Aninual hoilding riiiims acre kep-l't tet

21 ,I C with ý11 *Ill"- rclatityc hurtrudlx oin a 12-h lieht =dark cielwll ith)n tiilight. 4,1 S U I IS

Rats we re placed Iin clear platitci well1-ventilated container, foll When 11) Arg/kg o1'rh I F- I (:i %as adm inistclred ipl 20 happroximatels 5 miii before irradiation or shamn in ardiat ion. The eoesar raito t tcesdI~otaancl:.arnimAs aecr unilatcrally irradiated (irradiation of rills when thle bfr hmirdaini nrae yptaaiiradiation incidents from one 'ide) to At total dose ot 10 G (ivmidline cls of' total and scle ni Um-depe udent (iSH Px (Fig. ItissueC) Of I18.5 NleVI (niominal ) electron,, using it linear arccelerator aind had no effect onl SOD) (Fig- 2), Radiation exposure(disc rate: If) (%,is w 1.8ttI75 puss' I tiposo re of ratls it) 1- 1 (jGlnNnrae d tm ertr Fg ) erae

of ), raxs or elecctro ns i nduces hype ri lermia anid at sabnaxiial clos aln nraeweýtmertr Fg ) cracoh It) Gy was selected to study the mnechanismis in olveI in radiation- hypothalamic levels of' total and seleniumn-depenldentInduced hxelcm herefore. thle radiation dose it)li II ('1y, (~s SHPx (Fig. 1), and increased hypothalamic SOD (Hig.selected for this study. Prior to Irradiation. the radiation duse ailewas meaisured using. tilt 0 cc tISSueI-eqliiialent chamber (nianufac- 2). However. rhill-AIn treatmecnt and irradiation to-lured hy, Exradin Inc) plaiced at the nridline of at 5 cm diameter gether attenuated radiation-induced hyperthermia (Fig.acrxlic rat phantom, I XPOSUrc tiniformity measurements showed that 3) and increased hypothalamic SOD (Fig. 2) but pre-the dose rate s rin byh itt most 4 " along the length of' thephaintomn. All ionizationt chambers that were used have ealibration vented the fall in total and seleniuIm- dependent (iSH11xlaclOrs traceajble tio the National Institute of Standards and Techniil- (Fig. 1). When rhIL-Itit was administered I h beforeo gy. Do~sirnectry measurements were performed foillowing the AAPM rdainepsri a oefc i ~ohlruTaisk Group 21 Protocol for the D~etermination of the Absorbed rdainepsr.i a oefc nhptaaiD~ose friim Hligh-Energy Photon and Flectron Beamns". GSF1Px andi SOD levels (Tahle 1) and it did not attenu-

Thcmeairenewof OD ntlGISII~ inh ipotaliif llate radiation-induced hyperthermia (Table 11). Intrac-

Groups, of ratts %ere pretreaited ip with Ill iggkg of rhIL, Ini Airsaline 21) h prior to radiation exposure (thle dose of rh IL- 1 waschos~en from previous data> and at pilot study). The ratls weredecapitated aifter shami-irradiation or irrandiation with high-energy ~Dprda SP

electrons: the braiins were remnoved, aind thle hypothalamus dissected. Toa*SPfroz'en, and stored at - 70'('. All Ihe tissues were stored for 2 weeks 6.0 L hiGI~

while analyses wecre in progress. For analysis. the tsuswere sus-pendcd in 10.25 NM sucrose and homoigenized using at Potier-Elvebjtemhomogenizier. The homogenate wras centrifug~ed for 1ii min at 7001 X g. 6.0-and the suipernatant wats used for assaying SOL) and (iSIIPx activity. C

Sol) activity %Nas assayed with tile xant hinte oxidase -eytiiehri tite c ,tilehod" an nl I ttil andi selen i umni -iepeitent fiSt!PX aet ivilts were 4.07-dectermined in a coupled assay with glutathione redIC~iasC And 0)

Ridts sNIer Anesheifliedi wNith I mi/lkg lit as mixture (if ketlamine 20

151) ng,'kg). xvlainine 15 mg/kig). and iiceproma/,ine HI mg/kg) gisen

tnt anif were placed in aI stercotaxic aippairatus I [asid Kopf Itistrir- X i .merits. No. 321)). A single cannula wasý inserted atsepticailly Into tile (D 1.lateral Ycise lt I iveacord irn to the coitidin ales Het ircit from the atlasof Peltigri-rti et l. (1i8 trm posterior lii bregmai 2 5 ntn lateral.The catinula was bisycred until cerebnispinal fluid rose in the cal- 0o L0... -- --

nula. D~ental jet'slie cemnent %as used to secure the caninula. Anlittitas Control Radiation RhIL- la RhIL 1

(,

Were illiiwei to recover fior 2 days, hefo re they were used fur TreatmentRaatoexiperimeitils. At the end of the experiments, itecclioir sites were Fig. 1. iftfect i11 20) I-ipreticattrctitl wiit) -It p n, kg (if Hill liti p.veit ieil histi'lugiralls. atlone. iii comintitatiuon w~ith lit (1 of li adimliin. Ailii W GN Aif raIiliatott

adlone oil hypothilarlaic seleniinilldepeitilenit and totl;I( ;SiIs [P\lsMen mn-eiount ii/ h'id tempe~iraturei Values are expiessei ;is nicaln (it 'ictis It rim) ii rats S I- NI Sig-

All cxpcrintentis were performed at in etivirnicinttcual Icinperi- rtiicaintlv ditlerentl front ci-iuil sahiics. P I 00' " sucitificitti',(litlure ofl 22 +I andmi body tenmperature was ineasiirci as describeid teirnt trout irrtiiatcifsaic I' it iSý

I O8

fABLE I

tEffi(' of I itp. pretreatment with 10 i•g / kjg of rhiL- l. alone. m (-umihation with 10 (; •f radiation, or 10 (;' oj) radiution ahe on hIip-,halamtn.leh'nium-depenent (GSItPv, total (;SttP, and SOD let ev (tiat.i I/ ntn, protein)

Values are expressed as mean of activity from 6 rats S.E.M.

An1o.0Adan I Shamt radiation Radiation alone Rh!L- Ia• ,one Rh!l-) loenttn}te, radiatiot

Selenium-dependentGSHPx 4.1 +0.1)0 2.8 + 0.15 4.5+11.11) 3.1) ,(l.15

Total GSHPx 5.2 +t 011) 4.0 +0. 10 5.5 ±0.15 43 _+0 15SOD 21.0+1 0.15 28 ±0.10 22 +0.15 29) 0 "1)

"Significantly different from sham radiation values: P < (1.1)5.

IF ....... ....... -I7 TA BLE 11- Effect of I h i.p. pretreatment wah 10 AX/gkg of rhiL.hIa or valine

321- giren i.p. on radiation-induced (10 Gy) hyperthermia

* Values are expressed as mean of activity from 6 rats ± S.E.M.c 28i-•

- Treatment Mean change inS24 ,S•_temperatureE 2o0 oc ± SE-A-..)

>* Saline + radiation 1.1 + (1.10"16" RhIL-I a +radiation 1.3±+0.15

"1 RhIL-la +sham radiation 0.3+ O1.10

I- I

44

oi l - erebroventricular administration of 1-5 units of GSHPxCon Rad RhIL-1a RhIL-l1÷ (Fig. 4) or 1-5 Ag of SOD attenuated radiation-in-

Radiation duced hyperthermia (Fig. 5).Treatment

Fig. 2. Effect of 20 h pretreatment with 10 ug/kg of rhll-la i.p.alone, in combination with 10 Gy of radiation, or 10 Gy of radiationalone on hypothalamic SOD levels. Values are expressed as mean ofactivity from 6 rats + S.E.M. " Significantly different from controlvalues: P < 0.05. ' Significantly different from irradiated values: 2F- 5-

P <0.05.Rad~aton

S iiadP •I /

T I I1 1 E2.. Radiation/Sham Radiation . .. .-..

<• _a._J__.J_ .- J.X =L A .•._L--t... .L_ 1-. A..J • -..L _1L --- I-J

0 30 0 30 60 90 120 30 0 30 60 90 120

-30 0 30 60 90 120 Tm ~nTime (Min) Fig. 4. Effect of (GSHIt•, WV on rad iation -induced hyperlhermia. A:

non irradiated controls given I unit (',) 3 units (a) or 5 units ( v ) of

Fig. 3. E'ffect of 20 h pretreatment with rhlL la or saline given i.p. GSHIPx. B: 10 Gy of irradiation alone (0). in the presence of I uniton radiation-induced hyperthermia. 10 GJy of irradiation alone W.l in (,rA) 3 units (n•), or 5 "nit% of GSlIPx ( v ). Each point represent,, [ticthe presence of 10i ;kg/kg rhll-la C•,) or I10 /glkg rhll.-la• alone mean + S,'.M. of observation of five animals, Zero on oidinm'tev ). Each point represents mean-+ S.E.M, of observation of six represent% the temperature at the time of second injection (AI()

animal%. Zero on ordinate represents the temperature of (a) 38.3 +- 38.0 f 0I.10, (M ) 38,0:4- O0,(5, and 3T• )37 4 0,l10, and at the time (if!0 1, (,;) 38.1 + 0.05, and (A38.0 +0,1 at the time of radiation/sham irradiation (B) We 38,2 + 0.t0, (,) 38. J + 0,.15, (2) 3,.11 t- 0. 10, and

radiation, ( v ) 37.9 0 0.05.

109

to radiation-induced superoxide radical production by,.increased activity in the hypothalamus, thus providing

an efficient quenching of superoxide radicals. Ac the$hse..., ..... •.o., present time we have no experimental data to explain

I I . the decreased GSHPx level after radiation exposureSo SoIoosl

I - and the failure of rhlL-la to increase hypothalamic'o SOD levels when given alone; however, we would like

to suggest that GSHPx is utilized to scavenge free

radicals caused by radiation exposure, resulting in de-creased GSHPx levels. We also do not have the expla-

L L___L to attenuate radiation-induced hyperthermia and the

30 0 30 60 90 120 -30 0 30 60 90 120 changes in GSHPx and SOD levels.Time (Min) Several mechanisms have been proposed for radio-

Fig. 5. Effect of SOD, ICV on radiation-induced hyperthermia. A:non-irradiated controls given I 1g (o), 3 pg (u) or 5 gg (v) of protection by IL-I. Two of these mechanisms are in-SOD. B: 10 Gy of irradiation alone (0) in the presence of I jig (o), 3 duction of PGs and acute phase proteins such as ccru-gg (a), or 5 pg of SOD (v). Each point represents the mean+ 2oplasmin and metallothionem- 4'.'17 , which are freeS.E.M. of observation of five animals. Zero on ordinate representsthe temperature at the time of second injection (A) (o) 38.0+±0.05, radical scavengers. However, experimental evidence(N) 37.9+0.10. and (v) 37.8±0.10, and at the time of irradiation does not support mechanisms based on PG or metal-(B) (e) 38.1+0.05, (o) 37.9+0.15, (S) 37.9+0.10, and (v) 37.9+ lothionein induction for radioprotection by IL-IF4 . It

0.10.has been reported that, in mice, the radioprotectanteffect of rhIL-la may involve induction of endogenous

DISCUSSION MnSOD in liver, which was confirmed by the inductionof MnSOD-mRNA in the liver"9 . SOD might be a

Considering the role of free radicals in radiation naturally occurring compound with a radioprotectiveinjury, it is apparent that mechanisms of protection effect' because the iv administration of SOD provokedinvolve detoxification of radicals produced by radia- an increase in SOD level in various tissues of experi-tion 20 . Protectors of this type include xenobiotic scav- mental animals and led to an enhanced resistance toengers of free radicals and inducible endogenous an- ionizing radiation 27

,28 .

tioxidant defense mechanisms such as antioxidant en- Although the mechanisms that underlie the attenua-zymes5 '2. Cell damage induced by superoxide radicals tion of radiation-induced hyperthermia by rhIL-l a re-and related oxygen species, such as singlet oxygen and main speculative, our results suggest that one of thehydroxyl radicals, has recently been tentatively con- mechanisms of rhIL-la-induced attenuation of radia-nected with the etiology of a number of pathological tion-hyperthermia is stimulation of hypothalamic SODconditions in the brain 4

,6 1"3.

29, Enzymatic defense and GSHPx, because rhiL-la and irradiation togetheragainst activated oxygen species involves a cooperative increased both hypothalamic GSHPx and SOD levelsaction of several enzymes. The major defense against and central administration of SOD and GSHPx attenu-toxicity of superoxide radicals is conferred by SOD. ated radiation-induced hyperthermia. HypothalamicThis enzyme catalyses the dismutation of superoxide PGE 2, corticotropin releasing factor (CRF), and plasmaradicals to hydrogen peroxide and oxygen. Although adrenocorticotropic hormone (AC`TH) levels are in-the resulting hydrogen peroxide is relatively less toxic creased after pretreatment with IL-I 7. CRF and ACTH(scavenged by catalase and GSHPx) a highly toxic are antipyretics and have been implicated in thermo-hydroxyl radical is produced when hydrogen peroxide regulation 22 . Recent experiments with CRF and ACTHreacts with superoxide or with transition metals, such in normal, hypophysectomized, and adrenalcctomizcdas iron or copper. Therefore, catalase and GSHPx, rats suggest that the attenuation of radiation-inducedwhich scavenge hydrogen peroxide, act in concert with hyperthermia by IL-I is mediated by an increase inSOD"t . CRF and ACTH in addition to the stimulation of

Radiation exposure increased SOD and decreased antioxidant enzyme levels'.GSHPx levels in the hypothalamus. It has been re-ported that there is an increase in SOD levels in bone REFERENCESmarrow of mice I h and 24 h and 72 h in rats afterirradiation3 " 8 . Superoxide radicals and hydroxy radi- I Bartosw. G_. tLcyko, W. and Fried. R., Is superoxide dimn,,utac a

cals are produced by irradiation. SOD reacts promptly physiological radioproteclor.'. Evperi'nt'a. 35 (1979) 1194,

2 B3,aelos. 1.1:.. V;trioes, MTl.. Flpp 1: R ,'1.1k.k I-. ,iffd ANtOr. IIs' Ki wala. J.- Stoklaiso%,a .. \ &,ls %k a!4 nd Ilc& o,,! %I.I lM-. IRekto\ CnI/NiCllesard Tl lakl ~iic,ll'. lIn A. lBrceeia, NIA. Iler C ,i2T1 y-/,rradliatJ~I;old, ly5 ,14011- 11 OIL AIClII~ide 11liIIt.1t2'Rodgeers and 0 . Scincianio U kd' ) .) rtiec,, and tii 'o,6 Rath~aIN tit I ~IsIt I [lie boItic 01.1!tOss and cwhos Ic oi~S~ t I us. Wit 1(il 16-

heoz"MM1 told .lI'djitol. Io Sen ihco. Itals. 198h. pp 89 1012. 91 (1982) 5117 *; 15.3 lBraloid. MAI. A rapid and scoolixr encthod tOr the (11111n1ital loll 1) Kunvir. K.S- \,oshnais ) N - Nniih. oiad (itI. T . k-1,

O) plOIIlnitililIntiti the p~ltiCipIC AltI jl1tCilII-51

5 hilldjllC Atim! [ion inIulrs n aill~ iotvi~dai it tllliiiisll~' Ot ploictiuun Ill S NJTIhocitm , 724 (976) 24$S 254, KN. 1 lonn. I A Marnlli. *iIII Il, Walden i , id, I L u,,d

4 ( adel. .11 [ .ohr. J.1B mid Jeste. ON.V. Frce radicajls and tarTdi\C: toi oth, Moat0 tit e' Lip,p6 ot! * 'our, Intlopim'aoun1 '00 Mhoilai ojcf~skuintsua. f dSVliv flni Neo 1,1 ( 1980i) HIS - 101). bIjirri, Klusser. Bostoni. 1)90, 11pP.3 3 .3

.1115 (:hipmmi. J.D.1) Rleuscis .ARP. Iloisa, J. and (ireenstock. (FL.. 21) Kumnar. K.S.. Vaishnias. Y.N. and Weiss, .11< . Rild It 0flcl~k. 111111

lcicriAl radioprotection and radiows itslization of nuanlmalian antjosidaint ClnA me'c and eli/S11 11e noolcikN . Phi ,-os'L /hr?cctIs grosiming Ili 5 1(0o. Radiait. Revý. 5ý6 ( 1973) 291 -3016. (1 9.8S) 30 1 31I)9.

I, IDcxer, DLI.. ( arter. (AJ.. Wellk. F. R.. Jaxov-Agid. F. Agid Y.. 2 1 L~asrencc. R.A and Buirk, R. . ( lul ii. Itnc pvi \tiisu,3 is t'Ix, Ve . Millner. P. an~d Marsden. (.D), Basal lipid peroxidailion Iin selnium-deticienirt il ikci Mr. hIt, ini /uqihi R, ( 'iii,outi -

Ii subihsa ntima n igra is j lcrC asec Ilit Parkinson's diseiise 1. N.ew- 71 (19'7o) 9)52-1)5"i.

riiihon .5 381 ;I--389 21 2 LiptonI. JIM- anld Clark. Xk (j.. NCaliotransmiltcl' lbt ltllpv[.,tIAlc

7 )inarello. N. Bioloes of interleukin 1. I-ASh! .1 , 2 411988) control, Apitti,. Re,. P/ic-aol ,48 ( 198il) (113.1018 -11 . 23 Mc('ord, I.M. and FridoisCl. I.. Saperoxide: dismualise A\n en-

-S Elkind. MA.NI.. Repair processes Iin ladi at ion biology. Radio,. Res. iymi fuIn Iction to r crtrocll0ucipre ii (11U hcro CLapre In 1, 1 U1111 i,'11(1) (1k984) -42ý-4t I. 244 19169)) 61)1149155.

1) ( lass. GN. aniild Sitmnles . J-C.. In Staoisomo Methodls tin Eihaltioo 24 Ne Ia. R.. ()ppe p limit. Ii.. I ouclle'. S.!).. ( "ICLa.1' PC-. nihi ..andt Michohgi's. Pi-rie 1C-liall. F nglewOod (liffs. Nj. 1970- RT an d Kazrin M-.Radinpri 9CC) on A ilIh irIt criu kill- I ( IinpaIi -

I1l Goldsicin. I NI. aind Charo. I.E.. ('CrUlutplasmin:n An actitol phase son ss ith other cylokines. PnI'?o. I~naioiiii. VI II 980 1)illv909s.

reactaint and antioxidant. Ini E. Pick ( Ed.). Lvrmphokitw%, IVo. S. 2i Nela. R-. Douches. S- and ()ppenhleimn. IT). Inicileuikin I I, IAcadem-ic Press. New York. 1983. pp. 373-4 11. radisiprotectiir. 1. Initimlg~oI.. 136 (1 980u) 248~3 .2-Isi.

11 1 fal. Fi.. T1he physics and chemistry of radiat ion absorption. In 26 Pelligrino. I.S.. Pelligrino. AS, and ('ushman- AT.J. A4 Sleweaiu,1F.T l 11:1 (Ed.). Riaboio~log% tot the Radto/agw u.3rd edn.. Lippin- ,4dos o~fthe Rat Braim. Plenum. Ncks York. Iý97'.colt . Philadelphia, I1988, pp. 1 -- 16. 27 PCI kan. A.. Radialion protect ion Il' su peroxsi sde sn.4'

1 2 HaIllI. LiT. R adioprolecri rs, In LiJ. [fall (Ed.) Radhoboiohgv for tocht'in. F'Jotobio(. . 28 (I1978) 705 ' 774.uthe Raitio/o~izt.%t. 3rd edn.. Lippinocott. Philadelphia. 1988. pp. 28 Pelkitu. A., (hclack. S. and PlesNkachi. .841).. Prolectioti 4n ýuper-

2(11 -209'. Oxide diSMU I aS of 'White blood eel Is ill X-irrad iaed mIice. Ia 6('13, tblhlvl. BJ. and (itutteridge. JJM.. Oxygen radicals and thle ner- Sc,., 22 (1978) 867-882.

%OLNs NNICIII. htends Xv\eloaue..( 1985 22-26, 29) Sacggu. Ii.. Cookse-v. J.. 1)exter. D).. Wells,. ER.. Lces. A., .Icinnr.14 Kanidaisamv. S.B. and Hunt. W.A.. Involvenment of prostaglandins P. and Marsden. C.D.. A selective incre;ase in particulate silperý

and histamine in i acialion-indrlced teniperalure responses in oxide dismutlase activity ill parkinsollian substantia nigol. J, \en-rats. Radial, R,'s.. I12I (1990)) 84 -90. (U(w hem. 53 119)89) 692 - 697

1IS Kandasamly. S.13.. Kumiar. K.S.. Harris. All. aind Weiss. IT-. 301 Task Group 21. Radiation Therapy Committee AAP\I. A prolo-Effect o1 interleuikin-lIa on radiat ion -i ndcl d hyperthermia in coI for the cdetermination of absorhed dose fromi high cnricg'rats.o8m. Neu'rosit. .4hivi. 16 (1)91)0)496.10,I photon and electron beams. Mfel. Ph~l . 11)4(1983) 74 1.

16 Kiindasanty. S.B.. Stesens-Blakelv. S.A.. D)alton. T.K. alid H arris. 31 Weiss. I.T. and Kujiar. K.S.. Anlioxidalll mneehanisnls in radia-A. I.. Mec hani sms i n'olved in atlen nation of radiationi-indLIcCd t iin injury and rad iiprotection. In C. K. ('ho" (I' d.). (,'Ihit/rbvperthermia in rats b% interletikin. l-ASEB I.. Abstr. 6. Part 1. Anlooxidlant IDefrens, Methani.bol. 'oL. 11. CRC' Press. Bloca Raton.( 19912) [Abstract 22961. FL, 1988, pp. 163-189.

[7 Karin. MI.. Metallothioncins: protein in search of function. (CeY,41 (1985) 9-I10.

252 Biology of ýtric oxide

ARUEC FORCES RADIOOIOI.OGYReS&AC" INSTITUTE

I:The Biology of Nitric Oxide. IETXWNS. Moncada, M.A. Marietta, L.B. SR93-4Hibbs, Jr., and E.A. Higgs, eds.Portland Press, London, 1992.

Implication of nitric oxide synthase in radiation-induceddecrease in hippocampai noradrenaline release in rats

S. B9. Kandasamy, S. A. Stevens-lakely, T. K. Daltonand A. H. HarrisBehavioral Sciences Department. Armed Forces Radiobiology ResearchInstitute, Bethesda, MD 20889-S145, U.S.A.

Introduction

The hippocampus is important in critical functions such as learning, memory, andmotor performance, and these functions are impaired after exposure to ionizingradiation [1]. Noradrenergic systems are important in mediating arousal, foodintake and to some extent motor functions; histofluorescencc and immunohisto-chemical techniques have shown noradrenergic pathways in the hippocampus [2].Several factors can contribute to acute nervous system damage in arivo: systemicblood pressure is reduced following exposure to 25-100 Gy y radiation [3, 41, andcerebral blood flow decreases in a variety of brain regions. including thchippocampus [51; the ischaeMia produced by decreased blood flow is likely to affectneuroinal activity 161; ionizing radiation generates free radicals, and resultingoxyg~en radicals havc been implicated in cell damage following ischaemnia. brainischaemia induce, the release of an excessive amount Of glutamate in thehippocampus, and glutamate acts on nitric oxide (NC)) synthase to form NNOthrough '\-mcth%-1-i-aspartatc receptors, causing toxic effects [71. The purpose ofrhiý stud% was io examine the cffect of ionizing radiation on hippocampalnoradrcnaline (N\A) release in vitro, stimulated b%- IK0 0.5, 24, 48, and '2 h aftert rradiat ion /sham- irradiat ion and too determine the role of NO svnthase in therad iat ion- induced decrease in NA release.

Methods

Male Sprague lDawlc% rats, weicghing 2041-3(X) g, were used in these experiments.Rai, were killed b\ thcap~titaimn, and brains were remuoved, The hippocampus wasdt~secied using the mecthod of Glowinski & Ivecrscn 181. NA was measured byh.pc. Io. wpled \oiih elect rochemnical detection. Rclease of hippocampal N A in rinrowiaý stimulated h\ KC.l and the irradiation procedures were carried out as describedPre\ P)uSlk 191. Raiý were exposed bilaterally to varving doses of y rays using a '"Cosource at ~i rate of 10 6% min'. Statistical analysis \%a,, performed using Student'sitcst- Multiple c4ompari%o)ns with a shamr-irradiated control were done hy analysis

of v ariance and Dunnemtt\ test, Data were identified as significant if P < 0.05.

Central and peripheral nervous system 2S3

NA release (pmol mg-' of protein) Table 1

Radiation dose/sham-irradiation 0.5 h 24 h 48 h 72 h

Sham 15.8±0.3 15.2+0.4 15.5+0.2 15.3_+0.4

5 Gy 16.5±0.8 14.5±0.9 12.0+0.5* 12.0+0.2*

10 Gy 16.5+0.6 12.4+0.8 8.4+0.65* 8.0+0.5*30 Gy 16.7±0.5 12.0±0.4* 8.0±0.85* 8.0±0.6*

Effect of exposure to y rays (10 Gy min') on NA releaseValues are means + S.E.M. of three separate experiments. * Significantly different from

sham-irradiated values; P < 0.05.

NA release (pmol mg-' of protein) Table 2

L-NA L-NA+concentration Sham Saline+ sham- L-NA +(mg/kg) irradiation irradiation irradiation irradiation

0 16.0+0.3 8.4_+ 0,6* 16.9+0.9 8.6±0.4*3 15.8+0.6 8.0 +0.4* 17.5+0.6 12.0+0.4**

5 15.4+0.4 8.2±0.3* 21.5±0.4* 16.5+0.8**

10 15.5+0.65 8.0+0.3* 24.0+±0.6* 19.5+0.6**

Effect of pretreatment with L-NA on NA (pmol mg-I of protein) release

48 h after exposure to 10 Gy of y raysValues ore means ± s.r.m, of three separate experiments. * Significantly differtnt from

sham-irradiation values; P < 0.05. **SiZnificantly different from irradiated values;P < 0.05.

Results

There was no significant difference in NA release between irradiated and sham-irradiated rats when the hippocampal NA concentration was determined 0.5 h after

3adiation exposure (5-30 G" at 10 Gv mini). However. dhere were significantdecreases in hippocampal NA release 48 and -2 h after exposure to 5 and 10 Gv and

24. 48. and -2 h after exposure to 30 (;V of y rays (Table 1). Based on the above

data. a post-irradiation time period of 48 h and a y-radiation dose of 10 Gv at

10 G\ minn were choýsen for further studies with an NO synthase inhibitor, No'-ntr,-.i-arginc ft -iN . Pretreating rats with I mg kg I of L-NA administered i.p.

I h beforc, rra,!iation or sham-irradiation had no effect on the radiation-induced

decrease in N\ (data not shown). Howe\e-, 3 mg kg' of L-NA prevented theradiation-induced decrease in NA release, and 5 and 10 mg, kg ' i.-NA not onlv

prevented the decrease in NA release in irradiated rats but also enhanced NA release

in sham-irradiated rats (Table 2). Similar pretreatment with 1-10 mg kg-t of L

arginine before irradiation or sham-irradiation had no effect on either the radlat:on-induced decrease in NA release or the ha~al hippocampal NA release (data not

shownj

254 Biology of nitric oxide

Discussi --

This stLdy demonstrates that radiation had no effect on hippccampal NA release0.5 h after exposLre but decreased NA release 24, 48 and 72 h after exposure,depeiding on the radiation dose. At the present time we have no data to explain thedifferences in NA release at post-irradiation time intervals. However, it has seensuggested that the blood-brain barrier could be disrupted by ionizing radiation[10, 111. This would allow radiation-released neurotransmitters such as prosta-glandins (PGs), histamines, and 5-hvdroxvtrvptamine, as well as other circulatinlgfactors, (abnormal) access to neurons that modulate hippocampal NA release 112).It has been shown that PGs of the E series inhibit the release of NA fromsympathetic nerves both in the periphery and the central nervous system;conversely, inhibition of PG svnthesis leads to an increase in NA release 1131.

Immunochemical localization of NO synthase has been demonstrated inmost areas of the rat brain, including the hippocampus. NO synthase forms NOfrom L-arginine [7]. Pretreatment with L-NA (a selective inhibitor of brain NOsynthase) reversed the radiation-induced decrease in NA release, suggesting that.,.0 svnthase is involved in this phenomenon. Pretreatment with L-NA alsoenhanced NA release in sham-irradiated rats, suggesting that NO is involved i, theregulation of NA under normal conditions. However, inhibition of PG synthesisby L-NA (thereby increasing NA levels) should not be excluded at this time.Although no results are currently available on the measurement of glutamaterelease in the hippocampus following exposure to radiation, our results support thehypothesis that toxic overstimulation of glutamate receptors by ra ation orexcitotoxicitv contributes to overproduction of NO that can be toxic to !urons.

In conclusion, these results suggest that ionizing radiation induces adecrease in hippocampal NA release 24, 4F and 72 h after exposure, and NOsvnrthase is implicated in this radiation-induced decrease in ,,A release.

ReferencesI Kimm drt. 1) 1 & Hun.. 1. 1. '196ý in Ionii:ng Radutrn• \cural I-uncT,.n and kcha w.,r

Ki-cid,,rt, 1) & Ifunt. IF L ,cd. . pp 166 213, \cademic PrcN.' Nc• ) r,2 \ An i)neti,J . 1 .\ \1 19051, Prý,g. Ncur,,b.l 16, 1 " 141

n.hapana. P II & S un.:. R I 01968 RAdiat Rc. 35. $" M4 f .- ch.,an. 1. 0 ( .:nc , T I& |i mpt.n. I. I) 119M6) N at. :pa.c I~nl rmn l Id 57. ,• '12

"( (ýkvrham I G. 'Pau tkr. I| & Ilampi,,n. I) D 195-t I-d Pc 44, •-G Kirn. I . TAmura. V & Shun,. K 1195M3 Pr'e Brain Rc, 63. •9 58

- n~di,. . If B%131~. 1) 094)t, T..~.d. Pha.rmacol *- 12. Ia.12; Q

(,%% -nki, & crI c n.. 1 ( I ( \cut,s chrn) 13. t,,; (o.1 I,pn. N I.urm. b . Hunt, \X\ . taiti,, T K & StMc•n'. .i (19PKH% thuT-m3CA

B.,chcm Bi~h• 29. t;;: "Il

l , ,,h( r \\ ý s & TI,. V. \\ \ I'F- ( Ancc, 40. 11W() IfIllI I nhol.,Ic & n. 'na) ' IV- 1~ N-,.urt: 32. 314 9412 Kndj,,..n, I It , . Pu , NN \ t1'41)4' Riuhda Rc' 121. 84 9),

I t ; .r- r, . I .rmb•i, I ( & I uý. K i 9-31 Iur. L PhI rm,1u ol "1. 1,2 lu.'

ARI•ED FORES RANOITOLOGYRE•I.ARCH I&STITUTE

,IE i , jCIIUTIIiUIIPORT

I Ii ''I~ ] SR93-5 1

ENERGY DEPOSITION IN A SPHERICAL CAVITY OFARBITRARY SIZE AND COMPOSITIONE. KearslevRadiation HiophS sics DepartmentAmied Forces Radiohiology Research InstituteBethesda. MD- X89-5145. USA

Abs.tract The dose distribution inside a spherical cavity is calculated usinyu analytncal cv rk-1on,.r both ,ithe 1ho,,topping power and the starting energy distributions for elastically scattered secondarN chaired parli,.le, C'.gcfleraed ,secondaries are treated separately from secondaries generated in the surrounding medium Phe result i, ananalytical expression for the ratio of the dose to the ca',ty to) the do.e to the ,urrounding medium. Ihbi expressionsometimes referred to as the 'effective stopping power is in a form similar to the Burlin general Caytn theor forphotons.

INTRODUCTION by neutron interactions in a differential )olume

The energy deposited in a spherical cavity element, dV, located at a distance, r. from the

irradiated in a neutron field depends on the origin can be written a,composition of the cavity and surrounding N1dV (I )medium, the size of the cavity, and the energy of 4Trr2the neutrons. Caswell'' analysed this problem interms of 'insiders, starters, stoppers, and crossers'. where N is the numher of secondary chargedreferring to the trajectories of secondary charged particles per unit volume.particles relative to the volume of the cavity. His If it is assumed that the range of a secondaryobjective was to provide a detailed understanding can be written as R = A E", where A and rn areof the pulse height distributions obtained from constants that depend on the particle type. then tihemeasurements using tissue-equivalent propor- stopping power for a secondary generated in dVtional counters. Rubach and Bichsel" "' applied with an initial range. R. after travelling a distance.these same techniques to the study of the response r. can be writtenof ionisation chambers with a variety of wall-gas dE I /combinations and cavity volumes. This approach ( ) (R -- r) 12)provides insight into the total energy deposition in d× ina volume but little information about the spatial The contribution to the dose at the origin of thedistribution of the depo,,ited energy, which may be coordinate system is determined by consideringimportant to our understanding of the biological the separate contributions fromn neutronresponse of certain tissues of the body after interactions in the cavity material tie. the cavityneutron irradiation. This paper describes a contribution) and in the surrounding wall (i.e. thecalculation of both the dose distribution and the wall contribution). The complete calculationaverage dose within a spherical cavity of arbitrarysize and composition from neutron interacionswith both the cavity material and the surrourndingmedium.

THE CALCULATION

The origin of a spherical coordinate system isplaced at a distance. x. from the centre of a sphere aof radius, a (Figure I . The coordinate r mayextend to wny point inside or outside the cavityvolume. The dose at x is the product of the fluenceof secondary charged particles and their stoppingpower. Assuming an isotropic source of secondaries t1 igure 1 . ( eotnet ry 6 firlte calculatioln Ot t tc esp T ntseand neglecting any scattering effects at the (f a spherical cavit, in a neutron field. A spheitrialinterface for secondaries generated outside the co-ordinate system is cenired at at distanLce ", irom thccavity, the charged particle Iluence at x, generated cenlre of a sphere of radios, a

61

L KEARSLUY

considers a series of cases that depend on the boundary of the cavil, at a given angle 0; 0_ is therange of the secondaries and the boundaries of the azimuthal angle at which the range ol a secondarycavity. Only the case in which the maximum range is equal to r. The sum of Equations 3. 4. and tof the secondary. Rm, is less than the cavity radius, represents the total dose to the cavity, froma. will be illustrated in detail, interactions with the cavity material producing

secondaries with ranges less than the cavityThe cavity contribution radius.

Every point in the region, 0 < x < (a - R ). is The wall contributionsurrounded by a thickness of cavity materialgreater than Rm. The dose in this region can be To account for the fact that some fraction of thewritten as path of the secondary is in the wall, the residual

E fo fR'' N, range appearing in the stopping power becomes

D. 1(x)df dt dE d0Jdr 4 rEsin0(• mA) (R-r)--) R-fi(r-r),-r, (7)where fi is the ratio of the range in the cavity to

(R-r) i/,•- (3) the range in the wall material. The dose at x fromsecondaries produced from the wall can then be

Both # and 0 have their usual meanings in a writtenspherical coordinate system. The integrand is theproduct of the secondary fluence and stopping f-21 f(D f f E m.,, Npower, divided by the density of the cavity D.I(x) = d•J dE o d Jdr 4,,material. The radial integration is limited by the 0 LI (, ,,

range of the secondary which depends on itsenergy. The energy integral is over the starting sine m(r)energy distribution, which is assumed to be asimple step function. The subscript, c, refers to thecontributions to the dose to the cavity from where r, is the maximum range of a secondary

neutron interactions within the cavity; the starting in the wall at an angle 0, correcting for the

numerical subscript is an index to distinguish range differences in the two materials. That is.

between different components of the dose. R,- rFor the region (a-Rm) < x < a, secondaries rma, = r, + Ri (9)

generated within the cavity with a range less than(a-x) will contribute The Brass simplification

2i~ H fRC, i For the special case in which the cavity and

D,,2(x) = dD IdE f dO f dr l(rE,0) (4) wall material are identical, the sum of the cavity0 0 0 o contribution and the wall contribution at any point

x within the cavity must be equal to thewhere l(r,E,0) is the same integrand used in equilibrium dose to the material:Equation 3. El is the energy for a secondary with arange equal to (a-x): 1.0

Ex) aA )'m(5) 0.8-

For a secondary with a range greater than (a-x). 0.6the integral becomes

0.4-f2m F.r em RiI t

D,.3(x)=J dlJ dEJ dOf dr l(r,E,O)F, 1 0 f) 02 -

+dD J dE J dO dr l(rE.,) (6) 0 2 4 6 8 to0 em I Or 0 R,,ia

Figure 2. A plot of d as a function of R,/a for a value ofwhere r, is the distance between the origin and the m = 1.75 (appropriate for protons).

62

ENERGY DEPOSfION IN A SPItRERICAL CAVITY

NE absorption coefficients in the Burlin theory)D + Dx) N ,2p (10) multiplied by (l-d).

The function d(R,,,a.m) is illustrated in Figure 2Therefore, the cavity contribution can be for recoil protons (m=1.75). The exact expressiondetermined from the wall contribution and the for d(Rm,am) is a complicated, multi-termedtotal dose at any point x within the cavity can be function. Space limitations do not permit a fullwritten: listing of the analytical expression for d(R,,,a.m).

However, an approximate expression that can be

DT(x) = DN(x) + DD(x) = N - D(x)l+,, D,(x) used to determine values of d(R,,.a~m) to within2p. " 1% of the exact value is

This procedure, first suggested by Bragg'5', is d(Rma) l- (a (14)

useful because the expression for the wallcontribution, D,(x), is always much simpler to Recommended values for a, 1, and y are providedderive than the cavity contribution, D,(x). The in Table I. For a practical calculation involvingaverage dose to the cavity can then be determined complex materials such as muscle tissue or A-150by integrating over all values of x. The result can plastic, the contribution of each secondary must bebe put into the simple form: calculated and summed to determine either the

dose at x as in Equation I I or the average dose to<DT > = N Em (I -d) (12) the cavity, Equation 12.

2pc 2pDISCUSSION

where d = d(Rm.a,m) The calculation rests on several assumptions.

First, it is assumed that the secondaries generatedDividing both sides of this expression by the by neutron interactions either inside or outside theequilibrium dose to the wall of the cavity, we cavity are produced isotropically. This assumptionobtain an expression for what is sometimes restricts the use of the model to obtain dosereferred to as the 'effective stopping power' distribution information to cases in which the

neutron fields can be considered to be isotropic.f, Pw d + NcEmRp (1-d) (13) However, relations involving the average dose top -- i Nd.+E•p the cavity such as Equation 12 or Equation 13, are

general for any neutron field because of the

Note that this is in the form of the Burlin general spherical symmetry of the cavity. Second. thecavity theory for photons 6 '. The first term is simple parameterised form for the range-energyeffectively the mass stopping povwer ratio for the relationship and therefore the stopping power,secondaries multiplied by d(Rm,a,m). The second Equation 2, is not strictly correct at the end ofterm is effectively the neutron kerma factor ratio the track of the secondary. As before, this(equivalent to the ratio of the mass energy simplification has little effect on the total energy

deposited in the cavity as long as very little of thetotal energy of the secondary is involved. This

Table I. Recommended parameters for Equation 14. simplification will have a greater impact on theR./a < 2 dose distribution, because the stopping power at

the end of the track will be underestimated. Third,m Y[, it is assumed that the secondary energy

distribution is a simple step function appropriate0.868 (ions) 0.07724 0.2359 I for secondaries generated via elastic scattering1.500 (alphas) 0.08511 0.2511 1 interactions. At high neutron energies. this can1.750 (protons) 0.08117 0.2501 1 introduce substantial errors in the calculation.

S>Fourth, it is assumed that the ratio of ranges for aa2 particular secondary in two different media is

m a independent of the energy. This is approximatelycorrect over a wide range of energies.

0.868 (ions) -0.000902 0.6257 0.5298 The results of this calculation have been1.500 (alphas) --0.001315 0.6151 0.6202 compared with the calculations by Rubach and1.750 (protons) -0.002566 0.5727 0.6908 BichselO3A)for a wide range of neutron energies

(0.760 - 14 MeV), three cavity--wall combinations

63

E_ KFARSLEY

(TE-TE, TE-air, and C-CO,) and four decades of distribution within a spherical cavity of arbitrarygas-filled cavity volumes ((.01 - 10 cm'). At size and composition surrounded by a medium of2 MeV and below, the maximum difference arbitrary composition. The expression was averbetween the ratio of the dose to the cavity to the aged over the cavity volume to determine the ratioequilibrium wall dose was less than 3%. At higher of the total dose to the cavity to the equilibriumenergies, the differences between the two dose to the surrounding medium. The tonn for thecalculations are much larger, probably as a result latter expression is identical to the form of theof the assumed shape of the secondary starting Burlin general cavity theory for photons.energy distribution (i.e. neutron interactions are nolonger dominated by elastic scattering). ACKNOWLEDGEMENT

CONCLUSION This work was supported by the Arned ForcesRadiobiology Research Institute. Defense Nuclear

An expression has neen derived for the dose Agency, under Work Unit 4610.

REFFRENCES

I. Caswell, R. S. Deposition of'Energy by Neutrons in Spherical Cavities Radiat. Res. 27, 92-107 (1966).2. Rubach. A. and Bichsel. H. Neutron Dosimetry with Spherical Ionizanon Chamher.s I. Theory oj the Dom,

Conversion Factors r and W. Phys. Med. Biol. 27, 893-904 (1982).3. Rubach, A. and Bichsel, H. Neutron Dosimetry with Spherical Ioniation Chambers Ill. Calculated Remtulf for

Tissue-equivalent Chambers. Phys. Med. Biol. 27, 1231-1243 (1982).4. Rubach, A. and Bichsel, H. Neutron Dosimetrv with Spherical Iomization Chambehr.s 1IV Neutron Sensit; aite.s

for CICO, and Tissue-equivalent Chambers. Phys. Med. Biol. 27,1455-1463 (1982).5. Bragg, W. H. Studies in Radioactivity (London: Macmillan) (1912).6. Burlin, T. E. A General Theorv of Caiity honization. Br. J. Radiol, 39, 727-734 (1966).

64

Inhibition of c-kit Ligand/Steel Factor by Antibodies Reduces Survivalof Lethally Irradiated Mice ARMED FORCES RADiOBIOLOGYRESEARCH ,.STiTUTE

By Ruth Neta, Douglas Williams, Faith Seizer, and John AbramsSR93-6

Survival after irradiation with LD100030 (radiation dose lethal similarly abrogates LPS- and IL-1 -induced radioprotection.to 100% of mice in 30 days) is based on recovery of impaired Furthermore, administration of this antibody to unmani-hematopoietic function. Our previous studies using anti- pulated mice increased LD, 0 30 radiation lethality from 50%bodies to interieukin-1 receptor (IL-1R), tumor necrosis to 100%. Such an effect was not obtained using anti-IL-3,factor (TNFI, and |L-6 d~mons,rated that endogenous pro- anti-IL-4, or anti-gr3nulocyle-macrophage colony-stimu-duction of these three cytokines is required for untreated lating factor antibody. Thus, like IL-1, TNF, and IL-6, SIFmice as well as mice protected with lipopolysaccharide is required for survival from lethal irradiation.(LIPS), IL-1, or TNF to survive lethal irradiation. In this report This is a US government work. There are no restrictions onwe show that anti-c-kit ligand/steel factor (SIF) antibody its use.

R ADIATION-INDUCED destruction of the hemato- interaction of all three cytokines is required for radioprotec-poietic system was documented to be the primary cause tion.

of septicemia and death based on the findings that transfer In this study. we evaluated antibodies to SIF. IL-3, IL-4.

of normal bone marrow cells prevents death from lethal ir- and GM-CSF to assess the relative contribution of these HGFs

radiation (with LD 10 0 /30 , a dose that causes death of 100% of to radioprotection with LPS and IL- I and to innate resistance

animals within 30 days). Transplantation of bone marrow of untreated mice to radiation. The results indicate that SIF

can be replaced in part by the administration of inflammatory is absolutely necessary for innate as well as LPS- and IL-I -bacterial lipopolysaccharide (LPS) as well as the proinflam- induced protection from the lethal effects of radiation.matory cytokines, interleukin-I (IL-I) and tumor necrosis

factor (TNF), which when administered before lethal irra- MATERIALS AND METHODSdiation enhance the percentage of surviving mice by accel- AMice. CD2FI female mice, 8 to 10 weeks old, were purchasederating the recovery of the hematopoietic system.'I2 This effect from the Animal Genetics and Production Branch. National Canceris attributed in part to the ability of these agents to stimulate Institute, National Institutes of Health (Frederick, MD). B6D2FIthe production of hematopoietic growth factors (HGFs), in- female mice, 8 to 10 weeks old, were purchased from Jackson Lab-cluding IL-6, granulocyte-macrophage colony-stimulating oratories (Bar Harbor, ME). Mice were handled as previously de-factor (GM-CSF), granulocyte-CSF (G-CSF), and macro- scribed.2

phage-CSF (M-CSF) 3 Several additional cytokines are Antibodies. A rat monoclonal tgG , anti-IL-I receptor antibody

thought to contribute to the growth and differentiation of (35F5), was previously described." A rat monoclonal antibody

cells of the hematopoietic lineages. These cytokines include (MoAb) to 8-galactosidase (GLI 13) was used as a control. Chro-

T-cell-derived pluripotent IL-34 and IL-4' and the more re- matographically purified rat IgG (Sigma. St Louis, MO) was used ascently cloned, stromal cell-derived, steel factor (SIF).67 an additional control. Polyclonal anti-SIF antibody (P2) was raised

cdentlycl tioned stromf cell-derivthatared esteen or theF)6 -7 in rabbit against purified recombinant yeast-derived murine SIF asIdentification of the HGFs that are essential for the res- described.6 This antibody at 1:40 dilution neutralized 1.25 ug of re-

toration of sufficient hematopoiesis to result in survival after combinant murine SIF in an MC 6 cell proliferation assay. As alethal irradiation may be achieved using neutralizing anti- control, equivalent concentrations of normal rabbit serum or preim-bodies to these cytokines. Indeed, we have previously dem- mune serum was used. Rat monoclonal antimouse IL-4 ( 11 B 11) wasonstrated that antibodies to IL-I and TNF abrogate the ra- prepared as described." One nanogram of this antibody neutralizesdioprotective effect of LPS.8 This indicates that the ability of 15 pg of IL-4. Antimouse GM-CSF (22E9.1 1) was prepared as de-LPS to enhance survival of mice depends entirely on its ability scribed. I' Twenty micrograms of this antibody completely neutralized

to induce IL-i and TNF. The radioprotective effects of IL-1 44 U of GM-CSF in the BCLI proliferation assay.t2 Rat monoclonal

and TNF, in turn, depend on their induction and interaction anti-IL-3 antibody (8F8.1 1) was prepared as described."'Treatment. Recombinant human IL-I (rHulL-Io; 117-271 Ro

with IL-6, because antibody to IL-6 blocked the radiopro-t 24-5008, lot IL-I 2/88; activity, 3 X 10 U/mg) was kindly provided

tective effects of IL- and TNF.9 These results suggest that by Dr Peter Lomedico (Hoffmann-La Roche. Nutley, NJ). Bacterial,

protein free LPS, prepared from Escherichia cob K235 by the phenol-From the Department of Experimental Hematologi, Armed Forces water extraction method, was kindly provided by Dr Stefanie Vogel

Radiobiology Research Institute, Bethesda, MD; the Department of (Uniformed Services University for the Health Sciences. Bethesda.Experimental Hematology, Immunex Corp, Seattle, WA; and the MD). The antibodies and recombinant cytukines were diluted inDepartment of Immunology, DNAX Research Institute, Palo Alto. pyrogen-free saline on the day of injection. Antibodies or control IgCA. were administered intraperitoneally (IP) 6 to 20 hours before IP in-

Submitted May 11, 1992, a~cepted September 10. 1992. jection of 100 ng/mouse of IL-I or I gg/mouse of LPS. Mice wereAddress reprint requests to Ruth Neta. PhD, Department of Ex- irradiated 18 to 20 hours after IL-I or LPS treatment. In an additional

perimental Hematology, Armed Forces Radiobiology Research In- series of experiments, untreated mice were first irradiated and I tostitute, Bldg 42, NNMC, Bethesda, MD 20889. 2 hours later received IP injections of antibody, control protein, or

The publication costs of this article were defrayed in part by page vehicle. In each case, the inoculum was 0.5 mL/mouse.ch`targe payment. This article must therefore be hereby marked Bone marrow cellularity, colony-forming units (CFUs) assay, and"advertisement" in accordance with 18 U.S.C. section 1734 qclely to CSF assay. Bone marrow cells were obtained from groups of miceindicate this fact. irradiated with LDwwo and LD,0tov doses 8 days after irradiation (3

This is a US government work. There are no restrictions on its use mice/group) in two separate experiments and counted in hemacy-0006-4971/93/8102-0007S0.00/0 tometer. For CFUs determination, bone marrow cells from unirra-

324 Rfood. Vol 81, No 2 (January 15). 1993 pp 324-327

ANTIBODY TO c-Kir LIGAND REDUCES RADIOPROTECTION 325

dialed mice (donors) treated with saline only, anti-SIF antibody, or Table 2. Effect of Anti-HGF Antibodies on Survivalcontrol rabbit serum 20 hours before killing were obtained and 5 x of LD50/30 Irradiated Mice10"cells were injected intravenously to lethally (1,050 rad) irradiated

Treatment Oead/ toui % SDervjioairecipients (5 mice/group). Splenic colonies were evaluated 8 dayslater. CSF levels in the serum of mice receiving anti-SIF antibody or Saline 28603rabbit control serum and subsequently treated with IL-I were deter- Anti-IL-1R (100 Mg) 19/20 5"

mined as previously described.'" Anti-SIF 11 10) 40/40 0.

Irradition Mice were randomized, placed in Plexiglass con- rl9 1800 mg) 10120 50

tainers. and received whole-body radiation at 40 cGy/min midline Anti-IL-3 (400 Mo) 4'1 0 78

tissue dose by bilaterally positioned 'Co elements. The radiation Anti-IL-3 I800 Mg) 5/18 72

field was uniform within ±2'. The number of surviving mice was Anti-GM-CSF (800 ig) 6/18 66

recorded daily for 30 days. Anti-lL-4 (400 gg) 11/28 60

Statistical analysis of the results was performed using a contingency Anti-lL-4 (800 mg) 14/28 50

table analysis. CD2F 1 mice received 825 cGy radiation followed by IP administration

of saline, antibody, or control Ig in doses as specified in 0 5 mL total

RESULTS volume.* Different (P < 05) from the control, saline treated mice

The dffect a! anticvtokine antihod" on LPS- and IL-l-in-duced radioprotection. To assess the contribution of GM-CSF, IL-3, IL-4A and SIF to IL-I - and LPS-enhanced survivalfrom lethal irradiation, mice were treated with the antibodies vival time of anti-SIF antibody treated mice was 13.0 t 1.2

or control proteins before administration of LPS or IL-I and days versus 16.5 ± 2.8 days for control mice. In contrast.

LD 85/30 irr,,diation. The results (Table 1) indicate that, as treatment with anti-IL-3, anti-IL-4, or anti-GM-CSF anti-with anti-IL-I R antibody, anti-SIF antibody blocked IL-1- body did not affect the survival of mice. Therefore, in addition

induced protection from radiation lethality. Furthermore, this to its critical role in survival of ILPS and IL-I -treated mice.

antibody also completely blocked LPS-induced radioprotec- SIF is required for survival of untreated, lethally irradiated

tion. In contrast, 800 ug doses of anti-IL-3, anti-IL-4, or mice.

anti-GM-CSF antibody had no effect on IL-I- or LPS-in- Assessment of biologic effi'cts ofanti-SIF anthodr. We

duced radioprotection. Thus, SIF is critical to protection from have tested the anti-SIF antibody in unirradiated as well as

radiation induced lethality by IL- I and LPS. irradiated, untreated, and IL- I-radioprotected mice for theirThe efLect ofanticytokine antibody on innate resistance to effect on bone marrow cells. Mice received a single injection

radiation. The antibodies were next administered to mice of 1:10 dilution ofanti-SIF antibody or control rabbit serumthat were otherwise unmanipulated to test for the effect of and their bone marrow cells were examined for CR/s in

the treatment on survival from irradiation. The results in lethally irradiated recipients. Whereas 5 X 10' bone marrow

Table 2 indicate that treatment with anti-SIF antibody, as cells from saline alo.- treated mice yielded 17.2 t 2.8 splenicwith anti-IL-I R antibody, increased the incidence of mor- colonies, the same number of cells from anti-SIF-treated or

tality of LD50/30 irradiated mice. Furthermore, the mean sur- control rabbit serum-treated mice yielded 14.6 ± 1.4 and13.4 ± 4.4 colonies, respectively. Thus, a single dose of theantibody in normal mice did not reduce the number of he-

Table 1. The Effect of Anti-HGF Antibodies on Survival matopoietic progenitor cells.Table 1.iThe Effe t o WAnt ithG Antiis or SurLHowever, similar treatment followed by irradiation resultedin a reduced number of recovering bone marrow cells. Thus.

IPS IL-1 mice receiving LDo01•0 irradiation and anti-SIF antibody hadDead/ % Dead/ % 2.2 X 106 bone marrow cells/femur at 8 days postirradiation,

Treatment Total Survival Total Survival whereas the control antibody receiving mice had 4.5 X 106Saline 25/30 17 34/40 15 cells/femur. Similarly. after LDI(ii 3o irradiation, mice thatRat Ig + 8/20 60 12/28 58 had received IL-I had 2.4 X 10' cells/femur, mice receivingControl antibody 4 1/20 95 13/44 70 control antibody and IL-I had 1.8 × 10' cells/femur, andAnti-lL-1R + 15/20 25" 35/40 12.5° mice receiving anti-SIF antibody and IL-I had 1.0 X 106Antp-SIF + 19/20 5" 27/30 i0" cells/femur. Additional assays showed that anti-S1F serumRabbit Ig + - - 5/30 83

Anti-GMI-CSF + 7/19 63 6/20 70 in normal mice did not affect the titers of IL-I-induced CSF

Anti-lL-4 + 6/20 70 4/20 80 in the serum.

Anti-IL-3 + 5/20 75 2/15 87

CD2F 1 or B6D2F 1 mice received IP 100 ,g of anti-ILL 1 R antibody, 1: DISCUSSION

10 dilution of polyclonal rabbit anti-SIF serum, or preimmunized rabbit These results represent the first demonstration that en-

serum; 800 ,g of anti-GM-CSF, anti-IL-3, or anti-IL-4 antibody; 800 These result s r t st destration thatien-

,ug GL 113 or rat Ig; or saline in 0.5 mL total volume. Six to 20 hours dogenously produced SIF is absolutely necessary for survivallater, mice received 100 ng of IL-1 or 1 1Ag of LPS IP, and 1 day later from lethal irradiation of unmanipulated mice as well as L.PS-

received 950 cGy gamma radiation. and IL- I-radioprotected mice. SIF has been reported to act" Different (P < .051 from the treatment groups, but not different from as a most potent comitogen, in combination with 11-6. IL-

the control, saline-only-treated mice. 3, or IL- I, for hematopoietic stem cells (HSC5s). ' Although

326 NETA ET AL

treatment of mice with antibody to c-kit/SIF receptor resulted available in the bone marrow. Perhaps local cell-to-cell in-in elimination of hematopoietic progenitor cells,") the re- teraction in a tissue such as spleen allows for utilization ofquirement for c-kit ligand/SIF for fetal hematopoiesis was IL-3 and IL-4 in hematopoiesis. However, such a local eltlectquestioned by a recent report showing that absolute numbers might reduce the accessibility of the cylokine to the neutral-of hematopoietic progenitor cells nevertheless still increase izing antibody.in SIF-deficient SI/SI mice."? Thus, our results indicate that SIF is critical for survival

Our results showing that untreated, as well as IL-I- and from lethal irradiation and is generated in limiting quantities.LPS-radioprotected mice, do not survive lethal irradiation because a single dose of only 0.05 ml[/mouse of immuneafter receiving SIF neutralizing antibody suggest that the ca- serum precluded recovery.pacity of an animal to survive lethal hematopoietic syndromedepends on the availability of SIF. The question of whether ACKNOWLEDGMENTSIF is also required for constitutive hematopoiesis remains We thank William Jackson for statistical analysis of the resultsto be addressed. Constitutive expression of messenger RNA and DrsJoost Oppenheim, David Ledney, and Dov Pluinik for critical(mRNA) for SIF was detected in the bone marrow of un- comments on this manuscript.manipulated normal adult mice and was further upregulatedin the bone marrow of 5-FU-treated mice (D.W., unpublished REFERENCESresults), suggesting that, indeed, constitutive and emergency 1. Ainsworth EJ. Hatch MH: Decreased x-ray moriality in en-hematopoiesis may depend on the supply of SIF. The ad- dotoxin-treated mice. Radiat Res 9:96. 1958ministration of the anti-SIF antibody to IL-I-treated. unir- 2. Neta R. Oppenheim JJ. Douches SD: Interdependence (,f theradiated mice did not affect the levels of serum CSF and the radioprotective effects of human recombinant IL-I. TNF, G-CSF.numbers of CFUs, but significantly reduced the recovery of and murine recombinant G-CSF. J Immunol 140:108, 1988

bone marrow cells in IL- I-treated, as well as untreated. le- 3. Neta R, Sayers T. Oppenheim JJ: Relationship of tumor necrosisthally irradiated mice- These results suggest that SIF may be factor to interleukins, in Vilcek J, Aggarwal B (eds): Tumor Necrosis"y rFactor: Structure. Function and Mechanism of Action. New York.required for emergency hematopoiesis. NY, Marcel Dekker. 1991, p 499

The requirement for SIF is similar to the previously ob- 4. Metcalf D: The multipotentiat colony-stimulating factor. multi-served requirement for IL-I and TNF, because antibody to CSF (IL 3). Lyomphokines 15:183. 1988IL-1 R and TNF, each administered separately, abrogated ra- 5. Peschel C. Paul WE. Ohara J. Green I: Effects of B-cell stim-dioprotection by IL-I, TNF, or LPS and also enhanced the ulatory factor-I/interleukin-4 on hematopoicmc progenitor cells.rate of mortality of unmanipulated mice.8 Similarly, anti- Blood 70:254, 1987IL-6 antibody blocked IL-I- and TNF-induced radioprotec- 6. Williams DE, Eisenman J. Baird A. Rwiu Ness KV, Marchtion and increased the numbers ofunmanipulated mice dying C0, Park LS, Martin U, Mochizuki DY. Bosskell tIS. Burgess GS.

after LD5eso/0 irradiation.9 Taken together, our results indicate Cosman D, Lyman SD: Identification of the ligand for the c-kit proto-

that the presence of each of four cytokines (IL-I, TNF. IL- oncogene. Cell 63:167. 19907. de Vries P, Brasel KA, Eisenman JR. Alpert AR, Williams DE:

6, and SIF) is absolutely required for hematopoietic recovery The ef•ect of recombinant mast cell growth factor on purified murinefrom the lethal effects of radiation. hematopoietic stem cells. J Exp Med 173:1205, 1991

In contrast, systemic administration of antibodies to each 8. Neta R, Oppenheim JJ, Schreiber RD, Chizzonite R. I.edneyof the cytokines (GM-CSF. IL-3, and IL-4) did not reduce GD, MacVittie TJ: Role of cytokines (interleukin 1. tumor necrosisthe number of surviving mice, although these three cytokines factor. and transforming growth factor b) in natural and lipopoly-are known stimulators of hematopoietic cells. Repeated ad- saccharide-enhanced radioresistance. J Esp Med 173:1177, 1991miiistration OfGM-CSF, IL-3, and IL-4 after irradiation with 9. Neta R, Perlstein R, Vogel SN. Ledney GD. Abrams J: Roleless than LD95g/3, doses increases the number of surviving of IL 6 in protection from lethal irradiation and in endocrine responses

animals.i8 "o There is an apparent inconsistency between the to IL I and TNF. J Exp Med 175:689, 1992It). Chizonetti R, Truitt T. Kilian PL. Stern AS, Nones P. Parker

two sets of observations above. However, because the effects 10. Ka na R. Chua T. luaD PLubSer u: P. h arkntof hes thee ytoine oneary pogeito cels ay verap, KPR Kaffka KL, Chua AO, Lugg DK. Gubler U: 'Two high-affinity

of these three cytokines on early progenitor cells may overlap. interleukin-I receptors represent separate gcne products. Proc Natltreatment with a single antibody may be insufficient to coun- Acad Sci USA 86:8029, 1989teract the effect of the remaining cytokines. Alternatively. 11. Finkelman FD, Katona IM, Urban JF, Itolmes J, Ohara J.GM-CSF, IL-3, and IL-4 may be produced in amounts too Tung AS, Sample JG. Paul WE: II[ 4 is required to generate andhigh to be neutralized by the amount of antibodies used in sustain in vivo IgE responses J Immunol 141:2335. 1988this study. However, in vivo treatment with 0.5 mg/mouse 12. O'Garra A. Barhis D. Wu J. Hodgkin PD. Abrams J. Howardofanti-GM-CSF antibody significantly reduced the survival M: The BCLI B lymphoma responds to IL. 4. It. 5 and GM ('SF,ofC neoformans-infected mice,"' suggesting that the quantity Cell Immunol 123:189, 1989

of the antibody used in our experiments (0.8 mg/mouse) 13. Abrams JS, Pearce MK: Development of rat anti-mouse in-terleukin 3 monoclonal antibodies which neutralize hioactisity in

should have been sufficient. Of the three cytokines (IL-3, IL- tro. 3 mmnol a4ti1 die w e eivitro. J Immunol 140:131, 1988

4, and GM-CSF). GM-CSF is likely to be more prevalent, 14. Vogel SN. Douches SI3. Kaufman FN. Neta R: Induction ofbecause it is produced by many cell types, including T cells, colony stimulating factor in vivo by recombinant interlcukin-I amonocytes, fibroblasts, and endothelial cells, and is also pres- and recombinant tumor necrosis factor v. J Immunol 138:2143, 1987ent in circulation after challenge with LPS or IL-I."4 In con- 15. Metcalf D, Nicola NA: Direct proliti'rative actions of stemtrast, IL-3 and IL-4 are T-cell products and may be less readily cell factor on murine bone marrow cells in vitro: Fflicts of combi-

ANTIBODY TO cK/r LUGAND REDUCES RADtOPROTECTION 327

nation with colony-stimulating factors. Proc Natl Acad Sci USA 88: 19- Neta R, Wong GHW, Pilcher M: LIF and IL 4 used after6239, 1991 lethal irradiation protect mice from death. Lymphokine Res 9:568,

16. Ogawa M, Matsuzaki Y, Nishikawa S, Hayashi SI, Kunisada T, 1990Sudo T. Kina T, Nakauchi H. Nishikawa SI: Expression and function 20. Monroy RI, Skelly RR. MacVittie TJ, Davis I A, Sauber JJ,of c-kit in hematopoietic progenitor cells. J Exp Med 174:63, 1991 Clark SC, Donahue RE: Fhe effect of recombinant GM-CSF on the

17. Ikuta K. Weissman IL: Evidence that hematopoietic stem cells recovery of monkeys transplanted with autologous bone marrow,exprem mouse c-kit but do not depend on steel factor for their gen- Blood 70:1696, 1987eration. Proc Natl Acad Sci USA 89:1502, 1992 21. Collins HL, Bancroft GJ: Cytokine enhancement of comple-

18. Kindler V, Thorens B De Kossodo S, Allet B, Eliason JF, ment-dependent phagocytosis by macrophages. Synergy of tumor'hacher D. Farber N, Vassalli P: Stimulation of hematopoiesis in necrosis factor-a and granulocyte-macrophage colony-stimulating

vivo by recombinant bacterial murine interleukin 3. Proc Natl Acad factor for phagocytosis ofCryptococcus neoformans. Eur J ImmunolSci USA 83:1001, 1986 22:1447, 1992

ARMED FORCES RADIOBIOLOGYRESEARCH INSTITUTE

Experimental Hematology 21:338-344 (1993) [ CIENIFC REPORT~j ExpeInifetal01993 International Society for Experimental Hematology S R93-7 Hmatolog

Effects of combined administration ofinterleukin-6 and granulocytecolony-stimulating factor on recovery fromradiation-induced hemopoietic aplasiaM.L. Patchen, R. Fischer, T.J. MacVittieDepartment of Experimental Hematology, Armed Forces Radiobiology Research Institute, Bethesda, MDOffprint requests to: Myra L. Patchen, PhD, Department of Experimental Hematology, Armed Forces Radiobiology Research Institute,Building 42, NNMC, Bethesda, MD 20889-5145(Received 22 May 1992; in revised fomi 1 September 1992; accepted 14 September 1992)

Abstract. Hemopoietic aplasia is the primary limitation of IL-6 and G-CSF are two cytokines that, individually, stimu-drug and radiation cancer therapies. We have previously late hemopoiesis in vivo in normal animals [8-15] anddemonstrated that, individually, both interleukin-6 (IL-6) and enhance hemopoietic recovery when administered after radia-granulocyte colony-stimulating factor (G-CSF) can accelerate tion-induced hemopoietic injury [11,16-19]. While G-CSF hasrecovery from radiation-induced hemopoietic aplasia. In vitro been shown to selectively stimulate the proliferation of prog-studies suggest that IL-6 affects cells early in the hemopoietic enitor cells committed to myeloid differentiation [6,7], IL-6hierarchy, while G-CSF affects more committed progenitor has been reported to elicit numerous effects 18,20-26], includ-cells. Because these cytokines may also affect different cell ing the production of platelets 19-111. As opposed to acting onpopulations in vivo, we hypothesized that the use of these committed progenitor cells, IL-6 appears to act on multipo-agents in combination may further enhance recovery from tential cells more proximal to the hemopoietic stem cells,hemopoietic aplasia. Female B6D2FI mice were exposed to a synergistically enhancing the responsiveness of these cells tohigh sublethal 7.75 Gy dose of 6(Co radiation. Following irra- additional hemopoietic cytokines. In vitro data suggest thatdiation, mice were administered subcutaneous injections of IL-6 accomplishes this by shifting uncommitted cells fromeither saline, 500 Mg/kg of recombinant human IL-6 once the GO to the GI stage of the cell cycle where they becomedaily on days 1-6, 125 Mg/kg of recombinant human G-CSF more responsive to additional hemopoietic factors 120-241,once daily on days 1-17, or both cytokines as described, perhaps, via cytokine receptor upregulation. Because IL-6 andPeripheral white blood cell (WBC), red blood cell (RBC), and G-CSF appear to affect hemopoiesis at distinct levels withinplatelet (PLT) counts, as well as femoral and splenic granulo- the hemopoietic heirachy, we hypothesized that the use ofcyte-macrophage colony-forming cell (GM-CFC) and day-12 these agents in combination may be more effective atspleen colony-forming unit (CFU-S) contents were evaluated enhancing hemopoietic recovery in myclosuppressed animalson days 7, 10, 14, 17 and 21 postirradiation. IL-6 treatment than the use of these agents individually. In these studies, wealone slightly accelerated postirradiation recovery of most have evaluated the ability of IL-6 plus G-CSF therapy tohemopoietic parameters, while G-CSF treatment dramatically enhance recovery from radiation-induced hemopoietic injury.enhanced recovery of all hemopoietic parameters evaluated.Co-administration of IL-6 and G-CSF further enhanced the Materials and methodshemopoietic recovery. The most notable effects in combina- Mice. B6D2F1 female mice (-20 g) were purchased fromtion-treated mice were on recoveries of bone marrow and Jackson Laboratories (Bar Harbor, ME). Mice were maintainedsplenic CFU-S, which were significantly enhanced above in an accredited AAALAC (American Association forthose in G-CSF-treated irradiated mice as early as day 10 Accreditation of Laboratory Animal Care) facility in Micro-postirradiation. Although by day 14 postirradiation, splenic Isolator cages on hardwood-chip, contact bedding and wereGM-CFC and CFU-S recoveries in both G-CSF- and combina- provided commercial rodent chow and acidified water (pHtion-treated mice had surpassed unirradiated control values, 2.5) ad libitum. Animal rooms were equipped with full-spec-combination-treated mice exhibited a greater overshoot, trum light from 0600 to 1800 hours and were maintained atThese studies demonstrate the ability of IL-6 treatment to 211C ± 1°C and 50`Y0 ± 10% relative humidity with at least 10enhance G-CSF-mediated acceleration of multilineage recov- air changes per hour of 100% conditioned fresh air. Uponery following radiation-induced hemopoietic aplasia. arrival, all mice were tested for Pseudomonas and quarantined

until test results were obtained. Only healthy mice wereKey words: IL-6---G-CSF-Radiation--Myelosuppression- released for experimentation. All animal experiments were

Therapy approved by the Institute Animal Care and Use Committeeprior to performance.

Introduction. Hemopoietic stem and progenitor cell injuryand the resulting depletion of functional white blood cells IL-6 and G-CSF. Recombinant human 11,-6 and recombinantand platelets are critical problems associated with both human (;-(CSF were provided by Amgen (Thousand Oaks,chemotherapy and radiation exposure [1-31. Sustained hemo- CA). L,-6 (lot #012789) had a specific activity of IS2x 107

poietic recovery following chemotherapy or radiation expo- U/mg and G-CSF lnot #600)) had a specific activity of 1)0sure requires surviving pluripotent stem cells to self-renew as U/mg. Endotoxin contamination was less than 0.5 ng/mgwell as to differentiate into multipotent and committed prog- protein based on the limulus amebocyte lysate assay.enitors capable of giving rise to functional mature cells [4-71.

ML Patchen et al.: IL-6 and G-CSF Therapy for Radiation Aplasia 339

20 | i 1 i i a i i i a i i1ay Postirradiation s a

18 - 0O 14 17 21 10 7 10 t, 17 21

o16 -G.CSF s .Sai • al"l*I I- ;,L.6v G-sC vs Salm,, •I . Gi-* • ..I* -,

14 GICSF G-CS •: .1IL4i,

ESGIL-6 S

I- Eftf G-CSF n IL-6 plus G-CSF onbone

exoe t .5 yC adsbctnoul d initre F-Srcveyi iraiae B .F mie Mi.. ereithe saie IL6 50 igk/d y .on day 1-6), (#"F(2 xoe o 7 y 6 oadsbutnosyamnsee

ggody 7o - pu C. D rps enh etr sln, .. G-CSF. oscC ..... .-6 + G-CSF.'

m e a n s -o-f S a lin e 3 e x e r m e t. C F - vre m o f v u o t

"L 4 ... ."I 1 •••.weLL 23 se .p"O

6 8 10 12 14 16 18 20 22 06 8 10 12 14 16 18 20 22

Day Postirradiation (7.75 Gy) Day PosRo oi be a c ti on (7.75 Gy)

FRg. 1. Effect of IL-6, G-CSF, and IL-6 plus G-CSF on bonebarrow CFU-S recovea in irradiated B6D2F1 mice. Mice were Finl. 2. Effect of n L-6, G-CSF, and IL-6 plus G-CSF on splenic

exposed to 7.75 Gy caCo and subcutaneously administered CFU-S recovery in irradiated B6D2FI mice. Mice wereeither saline, IL-6 (w0a pg/kg/day on days 1-6), G-CSF (125 exposed to 7.7h Gy a rCo and subcutaneously administeredag/kg/d on days 1-17), or IL-6 plus G-CSF. Data represent the either saline, I-6, G-CSF, or IL-6 plus G -ClSF as described inmeans of values obtained from 3 experiments. CFU-S values Figure 1. Data represent the means of values obtained from 3in nonirradiated control mice were 6837 th 198 per femura experiments. CFU-S values in nonirradiated control micea p<0.0ms were 2378 m0 197 per spleen. *p<0.05.

Irradiation. The "Co source at the Armed Forces Radiobiology nized by cervical dislocation and their spleens were removed.

Research Institute (AFRRI) was used to administer bilateral The spleens were fixed in Bouin's solution, and the grossly

total-body lCo gamma radiation. Mice were placed in venti- visible spleen colonies were counted. Each treatment grouplated uelexiglas containers and irradiated at a dose rate of 0.4 consisted of f mice and experiments were repeated 3 times.

Gyemin. Dosimetry was performed using ionization chambersas previously described 127], with calibration factors traceable Granulocyte-iacrophage colonyfminyg cell assaya. pcemopoietic

to the National Institute of Standards and Technology. Before progenitor cells committed to granulocyte and/orexperiments were initiated, the dose rate at the midline of an mcohg eeomn eeasyduigadul-ae

Cell uspesion. Thecellsuspnsios use foreachassa rep- blo loymclWCredbood ge ll (fRBC)nd plaelet (PITk)

acrylic mouse phantom was measured with a 0.w cmr tissue- cont wreuperor ha Colon ter.equivalent ionization chamber manufactured by Exradin assay [29t . Mouse enplotoxin serum (ent v/v) was added to

(Lisle, 1L). Before each experimental irradiation, th e prate feeder layers as a source of colony-stimulating factors.

at the same location with the phantom removed was mea- Colonies (>-0 cells) were counted after 10 days of incubation

sured with a s5 cmr ionization chamber fabricated at AFRRI. in a 37lC humidified environment containing dif, COf 2The ratio of these 2 dose rates, the tissue-air ratio (TAR), was Triplicate plates were cultured for each cell suspension, and

then used to ensure delivery of the midline dose desired for experiments were repeated 3 times.

each animal exposure. The TAR in these experiments wasrneral d ce were exosd to7.5iv fholnIzd by c a halothane-anesthetized mice by cardiac puncture using a

Cell suspensions. The cell suspensions used for each assay rep- heparinized syringe attached to a 20-gauge needle. White

resented tissues from 3 normal, irradiated, or cytokine-treated blood cell (WBC, red blood cell (RBC) and platelet tLT)and irradiated mice at each time point. Cells were flushed counts were performed with a Coulter counter.fro m fe m u rs w ith 3 .0 m l , o f M cC o y 's S A m ed iu m (F lo w L a .b s,St t s i .R e u s of r p ca e x e i m n s w e p ol d r d thMcLean, VA) containing 10%t heat-inactivated fetal bovine meanposi ±suttis of cpedidata were calculateserum (Hyclone Labs, Logan, UT). Spleens were pressed mas_ tnaderr fpoe aawr acltd

through a stainless steel mesh screen, and the cells were Student's ha est was used to determine statistical differences.washed from the screen with 6.0 mL medium. The number of Significance level was set at p<0.05, and statistical differences

nucleated cells in the suspensions was determined by Coulter between treatment groups are indicated on the figures.

counter. Femurs and spleens were removed from mice eutha- stem el dburn. in one expeient 7[bl II holevnized by cervical dislocation. body 60Co radiation. Postirradiation, (y'tokines we ,re adminis-

Spleen colony-forming unit assay. Exogenous spleen colony- tered subcutaneously (s.c.) in a 0.1 nil- volume at S(00 pg/kg/dforming units (CFU-S) were evaluated ,by the method of Till for IL-6 and at 125 pg/kg/d for G;-CSF. G-CSF was adminis-

and McCulloch [281. Recipient mice were exposed to 9.25 Gy tered on days I through ]17 postirradiation. IL-6 was generally

of total-body radiation to eradicate endogenous hemopoietic administered only on days I through 6 post irradiation for

stem cells. Three to 5 hours later, bone marrow or spleen cells fear that prolonged treatment with this cytokine twhich is

were intravenously (IV) injected into the irradiated recipients. known to act on early hemopoietic stem cells) might resullt in

Twelve days after transplantation, the recipients were eutha- stem cell "burnout." In one experiment (F[able 1), however,

340 ExpenmeUtal Hematolog vol. 21 (19931

40 1 1 1 " 24 r I

D'y P W F 7 0 14 17 21

Yo 20 ii -l2L ,, 6 ,• - - :i... 20 G -CSF •, sa" . . ._ 30 IL 6 . G CSF Sa, * .... .... GCSF SaM ,, -I*l*I*l*'

i.CSsF L- ** *U * * .GCSR .- "S16 CILr6 G-CSF -SF - * !* " " /

1. -6 G-CSF . G-CSF - -- -- I." 1 C4 F m r C W/.

LL 20 -IL-6 CL12 - L-6SG-CSF 12 G-CSF

CL IL-6 +G-CSF f. 10 IL-6S+ G-CSF"Saline --8 Saline

10- ILL0• .. .--- O 6.-i /J.6-

0 ... " I I I I I 26 8 10 12 14 16 18 20 22 0

Day Postirradiation (7.75 Gy) 6 8 10 12 14 16 18 20 22Day Postirradiation (7.75 Gy)

Fig. 3. Effect of IL-6, G-CSF, and IL-6 plus G-CSF on bone Fig. 4. Effect of IL-6, G-CSF, and IL-6 plus G-CSF on splenicmarrow GM-CFC recovery in irradiated B6D2F1 mice. Mice GM-CFC recovery in irradiated B6D2FI mice. Mice werewere exposed to 7.75 Gy 60Co and subcutaneously adminis- exposed to 7.75 Gy •"(Co and subcutaneously administeredtered either saline, IL-6, G-CSF, or IL-6 plus G-CSF as either saline, IL-6, G-CSF, or IL-6 plus G-CSF as described indescribed in Figure 1. Data represent the means of values Figure 1. Data represent the means of values obtained from 3obtained from 3 experiments. GM-CFC values in nonirradiat- experiments. GM-CFC values in nonirradiated control miceed control mice were 11,478 ± 406 per femur. *p<0.05. were 3036 ± 180 per spleen. *p<0.05.

mice were administered IL-6 on days 1 through 17. Irradiated Effects on granulocyte-macrophage progenitor cell repopulation. Inmice received either saline, IL-6, G-CSF, or IL-6 plus G-CSF. comparison to saline treatment, all cytokine treatments accel-On days 7, 10, 14, 17 and 21 postirradiation, 3 mice from erated bone marrow (Fig. 3) and splenic (Fig. 4) GM-CFCeach treatment group were euthanized to evaluate hemopoi- recovery. G-CSF therapy was again significantly more effec-etic recovery based on bone marrow and splenic CFU-S and tive than IL-6 therapy. Combination therapy, although itGM-CFC content and on peripheral blood WBC, RBC and again appeared to be more effective than G-CSF therapy, sta-PLT numbers. Nonirradiated mice (normal controls) were also tistically offered an advantage over G-CSF therapy only at dayevaluated at each time point. 14 postirradiation. In spite of initial recovery delays, splenic

GM-CFC numbers in all cytokine-treated mice dramaticallyResults overshot splenic GM-CFC numbers in normal (nonirradiated)Effects on stem cell repopulation. The ability of IL-6, G-CSF, and control mice. By day 17 postirradiation, GM-CFC numbers inIL-6 plus G-CSF therapies to accelerate bone marrow and combination-, G-CSF- and IL-6-treated mice were, respective-splenic CFU-S recovery in irradiated mice is illustrated in ly, 758%, 614% and 302% of normal control values.Figures 1 and 2, respectively. As early as day 7 postirradiation,bone marrow CFU-S numbers in all cytokine-treated mice Effects on mature peripheral blood cell repopulation. The reap-were significantly greater than in saline-treated mice. In gen- pearance of peripheral WBCs, RBCs and PLTs indicated thateral, G-CSF therapy was more effective than IL-6 therapy, and all 3 cytokine treatments could facilitate multilineage hemo-combination therapy was more effective than G-CSF therapy. poietic repopulation (Figs. 5-7). Combination- and G-CSF-Combination therapy produced significantly greater bone treatment led to production of WBCs within 14 days postirra-marrow CFU-S recovery than either IL-6 or G-CSF therapy on diaton, and production of RBCs and l'LTs within 17 daysdays 10, 14, 17 and 21 postirradiation. postirradiation. Although evidence of RBC recovery in IL-6-

In the spleen, only the combination therapy statistically treated mice also occurred on day 17 postirradiation, elevatedenhanced CFU-S recovery within 7 days postirradiation. At WBC and PLT recovery was not observed until day 21 poxstir-later times, significantly enhanced splenic CFU-S recovery radiation.was observed in all cytokine-treated mice, with combinationtherapy again being significantly more effective than G-CSF Efficts of prolonged IL-6 administration on hernopoietic repoplda-therapy, and G-CSF therapy being significantly more effective tion. In our fin,, experiment, additional groups of mice werethan ;L-6 therapy. In contrast to the gradual cytokine- incorporated to evaluate the hemopoietic effects of 6-day vs.induced CFU-S recovery observed in the bone marrow, 17-day IL-6 treatment. In this experiment, deaths weresplenic CFU-S recovery in all cytokine-treated mice exceeded observed in mice treated long-term with IL-6 alone such that,that in nonirradiated mice by day 17 postirradiation. At this by day 21 postirradiation, no L1-6-treated mice remained alivetime, IL-6-, G-CSF- and combination-treated mice, respective- to be evaluated. Although mice receiving 11-6 for 1 7 days plusly, exhibited 206%, 354% and 422% of number of normal G-CSF survived better than mice receiving the 1 7-day IL-6control splenic CFU-S. By day 21 postirradiation, however, treatment alone, these mice exhibited less CRF-S and (;M-the number of splenic CFU-S in cytokine-treated mice CFC recovery than mice receiving only the b-day II.-b treat-decreased toward normal levels. ment combined with (G-CSF treatment (Table I). Peripheral

blood values also reflected the suppressed progenitor celleffects. At 21 (lays postirradiation, peripheral WB(', RB(C and

ML Patchen et al.: IL-6 and G-CSF Therapy for Radiation Aplasia 341

1 I I i I I a I I I I I I I I I I I I I I I I j I I I IDay Posarradat bo Day poslu ,awon7 10 14 17 21 8.0 7 10 14 17 21

3.0 16 " .s,*r- - - - IL*6 s Sahne 0

GýCSFýSauna .G CSF - SaunalL. G-CSF vs Saline -, 7. 7n - --tG-CSF v9SILi6 - G CSF ý ,I6

'- 2"016-GCSi--G-CF--------"--,L• + G-CSF VS IL8 -- ILA G -- F .'L.• .6X•IL 6 G -CSF vs r 'IF . . .-,•v I]L6 -•• * •S ' "

_2.0 ... .•6.o0 "•- ;E

-IL-6 ".------.G-CSF

IL-6 + G-CSF N*"-- - Saine .i

0.0 3.0 -L I L6 8 10 12 14 16 18 20 22 6 8 10 12 14 16 18 20 22Day Postirradiation (7.75 Gy) Day Postirradiation (7.75 Gy)

Fig. 5. Effect of IL-6, G-CSF, and IL-6 plus G-CSF onl white Fig. 6. Effect of IL-6, G-CSF, and I1.-6 plus 6-CSIF oil redblood cell recovery in irradiated B6D2F1 mice. Mice were blood cell recovery in irradiated B6D2Ft mice. Mice wereexposed to 7.75 Gy "'Co and subcutaneously administered exposed to 7.75 Gy ý"Co and subcutaneously administeredeither saline, IL-6, G-CSF, or IL-6 plus G-CSF as described in either saline, IL-6, G-CSF, or IL-6 plus G-CSF as described inFigure 1. Data represent the means of values obtained from 3 Figure 1. Data represent the means of values obtained from 3experiments. WBC values in nonirradiated control mice were experiments. RBC values in nonirradiated control mice were5.39 ± 0.35x l06 per mL. *p<0.05. 9.06 ± 0.12x10 , per mL. *p<0.05.

PLT numbers in 6-dav IL-6-treated mice, respectively, were lations committed to granulocytic differentiation 1221.3.90xl0h/mL, 8.16xli)/mL and 522x10 6/ml., while WBC, Data presented in this paper confirm previous dataRBC and PI.T numbers in 17-day IL-6-treated mice, respective- describing the abilities of 11-6 and G-CSF, individually, toly, were only I.90x I06/mL, 5.94x 10O/mL and 244x106 /mL. accelerate hemopoietic recovery in myelosuppressed mice. In

this study, IL-6 and G-CSF induced regeneration of multipleDiscussion hemopoictic cell lineages, including WBCs, RBCs and PuAs,Morbidity and mortality associated with high-level presumably, via the ability of each to enhance regeneration ofchemotherapy or radiation exposures can be directly attrib- multipotent CFU-S as well as committed GM-CF(: progenitoruted to infectious and hemorrhagic complications resulting cells. G-CSF was generally more effective at enhancing recov-from therapy-induced neutropenia and thrombocytopenia. ery than was IL-6. The differences in effectiveness may haveRecovery from potentially lethal effects of hemopoietic deple- occurred because IL-6 was administered for only 6 days postir-tion requires both the generation of functional granulocytes radiation, while G-CSF was administered for 17 days postirra-and platelets that will prevent sepsis and hemorrhage, and diation. This is unlikely, however, since we have observedthe self-renewal of pluripotent and multipotent stem cells that 17-day IL-6 treatment is not as effective as 6-day Il-6that will lead to long-term reconstitution of the hemopoietic treatment in stimulating hemopoictic recovery in irradiatedsystem. mice.

In recent years, at least 5 cytokines (G-CSF, GM-CSF, ILA-, Based on the fact that G-CSF in vitro has been shown toIL-3 and 11.-6) have been evaluated for the ability to stimulate selectively induce granulocytic proliferation and differentia-hemopoietic regeneration following radiation- or chemother- tion 16,171, the multilineage effects we observed in vivo fol-apy-induced myelosuppression t 11,14,16-19,30-371. The lin- lowing G-CSF administration might seem unusual. It is inter-cage-specific cytokine G-CSF, particularly, has shown promise esting to note, however, that recently multilineage (granuLo-114,16-19,30,35-371. In preclinical studies involving irradiated cyte, erythrocyte and platelet) effects have also been observedcanines, G-CSF dramatically accelerated granulocytic recovery in some myelodysplastic patients following (G-CSF adminis-and reduced infections; however, platelet recovery was unaf- tration 138,391. The discrepancy between in vitro and in vivofected, and without platelet transfusions, irradiated animals effects of G-CSF suggests that some in vivo effects may beremained at risk for spontaneous hemorrhage. Promise of indirectly mediated, possibly through tht: induction ot addi-treatment for thrombocytopenia has recently come from tional hemopoietic cytokines (as has been shown for (,M-studies demonstrating the ability of 11-6 to enhance platelet CSF) 140,411, or through the regulation of cytokine receptorrecovery following suppressive radiation or chemotherapy in expression. Although no direct evidence of (;-(:SF-inducedmice I 11,311. Based on these effects, we hypothesized that hemoptietic cytokine production has yet been reported, G-combinations of cytokines capable of inducing the produc- (1SF has been demonstrated to increase II,-1 receptor expres-tion of multiple cell types necessary for survival following sion on hone marrow cells both in vitro and in vivo [421.chemotherapy and radiation exposures might be more benefi- Since 1l:- is known to be a potent hemopoietic regulator incial than individual agents. The combination II,-6 plus G(-CSF combination with G-CSF (34), an alteration of 11.-I receptorwas chosen fo! evaluation not only because of the desirable expression may contribute to the heniopoictic effectseffects each agent induced on granulocyte and platelet pro- observed following G-(CSF" administration in irradiated mice.duction but also because in vitro data indicated an ability of Based on in vitro data 120-22,24-261, the multilincageII-6 to synergize with G-CSF to expand progenitor cell popu- effects observed following Il-6 ad(ninistration were not sur-

342 Experimental Hematology vol. 21 119931

600 L " * I I I neous factors were necessary or in vitro sv nrg. 122_, and onOay Postradiabion our experience with iII vivo col)hinatitiwt ol It-, and 0,t-7 ¶0 ¶4 17 215 L0 - 1- 1 (2SF, where simultaneous admhinistration proved supurior to500 Iu-• n S~ine i-- Ip --I*

G-CSF vs Saline - .**sequential administration in I Velosuppresstd animlsIL-6 ÷ G-CSF vs Saline - - - * -- * (MacVittie et al. unpublishedl. lhe 11. - plus t- Sl therapy

40 ,L-6 . G.CSF v ,L6 I -- - I- - .iaccelerated hemopoictic recovery more than (,-( SF therapyIL-6 . G-GSF vs G-CSF - - - - * (the next best treatment) based on all parameters evaluated

300 . (i.e., femoral and splenic CFUj-S and -(Al( values is %%lIl asIL-6 WV(" RBC and PIT values). .he earliest and most striking

T 00 .....-- 6--- G-CSF .effects were obser.,ed on CFU-S recovery, which wds evident i

a-- -- Saline to 4 days earlier in conmbination-treated mice than in (P-( S-100 .. treated inice. Interestingly, pl*,pheral blood p, ramitcrs

revealed the least significant effects. lecause the generation of"-nature functional end cells from progenitor cells may take0 I I I I I t ! I I I I ..

6 8 10 12 14 16 18 20 22 several days, it is possible that etfects on recovery of periphih r-Day Postirradiation (7.75 Gy) al blood elements may have been more apparent ' ter than

day 21 postirradiation, which was the last time point evaluat-ed in our studies.

Fig. 7. Effe-t of 1L-6, G-CSF, and IL-6 plus G-CSF on peripher- Recent studies evaluating E-kit ligand used in combination

al blood platelet recovery in irradiated B6D2FI mice, Mice with G-CSF have demonstrated synergistic interactions in t,.e

were exposed to 7.75 Gy "•Co and subcutaneously adminis- production of blast-cell colonies 146] and in vitro expanded

tered either saline, IL-6, G-CSF, or IL-6 plus G-CSF as CFU-S [471. In addition, c-kit ligand in combination with ( -

described for Figure 1. Data represent the means of values CSF has been shown to synergistically enhance granulopoiesis

obtained from 3 experiments. PILT values in nonirradiated in vivo [48, 49]. Our -tudies have demonstrated that 11-6 plus

control mice were 904 /38x10" per mi.. *p<0.05. G-CSF produces a similar in vivo effect. Since t-kit ligand, li,,,,IL-6, has been rep,,rted to be capable of shifting multipo)tentprogenitor cells from G0, to G, of the cell cycle, wh, re they

prising. of particular interest, however, was the ability of 11,6 become more responsive to additional cvtokines 146], similar-

to stimulate RBC regenera t ion. Previous studies havebIt emon- ities observed following treatment with these two cytokine

strated anemia following IL-6 adninistration [10,431. These combinations may be related to a common mechanism.In conclusion, we have demonstrated the abilit' of thera-studies, however, were performed in normal mice and pri- peutically administered It-6 plus G-C'" to enhance CUU-S

mates. We can only speculate that the mechanisms regulating and GM-CFC repopulation and to acelerate the productionlineage commitment may be very different in normal vs. of mature WBCs, RBCs and PI.Ts in .adiation-injured micehemopoietically depleted animals. in fact, recent studies in moe ffectively than It-6 or G-'SF alone. Whether these

monr laboratoryy than45 have orostae that aloe.shetergherour laboratory [44,45] have demonstrated that messenger effects are directly or indirectly mediated following in vivo IL-ribonucleic acid (mRNA) expression for a variety of hemopoi- 6 and G-CSF administration remains to be determined.etic cvtokines is increased in hemopoictic tissues following aradiation exposure such as the one used in the studies pre-sented here. In particular, bone marrow and splenic 11.-!. GM- AcknowledgmentsCSF, and c-kit ligand expression have been observed to be dra- 'The authors are grateful to Ms. Ruth Seemnann for technicalmatically increased; depending on the specific cytokine and assistance, to Ms. Donna Solyan for editorial assistance and totissue, increases can be observed as early as 6 hours postexpo- [r. Larry Souza asure and persist for up to 10 days postexposure. Hence, it is used in these studies, This work was supported by the A\rmedalso possible that in vivo IL-6 and G-CSF may interact with thes stdies. R kswa s stituted byfthe Nrmedadditional cytokines being endogenously produced in the Forces Radiobiology Research Institute, Defense Nuclear

irraiatd aima toindce he ultplehempoitic Agency, Linder Research Work Unit 00132. Re'ea,j h was con-irradiated animal to induce the multiple hemopooetic ducted according to the principles enunciated in the Guideresponses observed. for the Care and Use of L.aboratorv Animals prepared by the

Because IL-6 is known to act on early multipetent progeni- Institute of Laboratorv Animal Resources, National Rt- "archtor cells 120-261, It-6 treatment was intentionally restricted in C(buncil.length for fear that continual 11-6 administration, especiallyin combination with simultaneous G-CSF administration. Referencesmight induce multipotential cells to differentiate at tic I. Benacerraf B 11961) Influenc' of irradiation on resistanceexpense of self-renewal and result in "stein cell burnout. to infection- Bacteriol Rev 24:35Although this did not appear to be a problem with the 6-day 2. Haammond CW, omtpkins M, Miller (I1 I19.54) Studies onII-6 protocol used in out studies, the detrimental effectsobserved following prolonged I1.-6 treatment alone or in tibe to infection ollogin g radiation

combination wlih (G-CSF, suggest that further cautious inves- time of onset and duration of endogenous bacteremias in

tigation of this issue is warranted to determine whether ,hese m.ice.. jxp Med 98:4)h5effects reflect 11-6 toxicity or stern cell burnout. uack 11.1, Ietion in compromiced hosts ltl.achman n PJ, Peters l)K I edsi (ili nical Aspe'cts (ot

In addition to individual effects, data presented in this immnlogy. Oxforu. Blackwell Scientific Ptblishers.paper provide evidence of the ability of 1l-6 and (J-CSF to 1713interact in vivo to further enhance multilineage hemopoietic 4. Metcalf I) o'd) (177117 llemiopoietic colonies. New N ,rk:regeneration in irradiated mice above that induced by 11-6 or Springer VerlagG-CSIF individually. Siimultaneous 11-6 and (-CSF adminisira- S. (;f)de I)W, (line MJ, Metcal I1), Fox (t (eds) (l1978)tion was chosen for evaluation instead of sequential adminis-tration based on Rennick's data demonstrating that simulla- llemopoieti( cell differontiation. New Yorl,: \cadcic

ML Patcluen et al.: IL.4 and G-CS Therapy for Radiation Aplasia 343

Table 1. Hemopoietic regeneration following 6-day andi 17-day 11:6 treatment combined with 6-( SF treaitment(day 21 postirradiation I

11.-6 treatment 1,FU-S per lemur IJFU-S per spleen G3M-CFC per femur (IM-(T(, per spleen

6-day 18o0±124 .3338±288 305,S±282 62 51 + 63

I17-Day 458±32* 15 16±136* 603±SS* 2842±3 16*

*p<o.05 with respect to 6-day values

Press 18. MacVittie TJ, Monroy RI., Patchen ML., Sou/.a LM I 19:40)6. Robinson BE, Quesenberry P (1990) Review: Hemopoietic Therapeutic use of recombinant human G~~:(rh(1.( .SF)

growth factors: overview and clinical applications, part 1. in a canine miodel of sublethal anc e~thal whole-bodY ifra-Am J Med Sci 300:163 diation. IntJ Radiat Biol.57:723

7. Robinson BE, Quesenberry P (1990) Review: Hemopoietic 19. Tanikawa S, Nakao 1, I'suneoka K, Nara N l1989) Effects. ofgrowth factors: overview and clinical applications, part 11. recombinant granulocyte colony-stimulating factor (rC;-Am I Med SOi 300:237 CSF) and recombinant granulocvte-miacrophiage colony-

8. Suzuki C, Okano A, Takatsuki F, Miya saka Y, Hirareo T, stimulating factor (rGM-CSiI on acute radiation hernopoi-Kishimoto T, Ejima D), Akivarna Y (1989) Continuous per- etic injury in mice. Lxp Hematol 17:883fusion with interleukin-6 (IL-6j enhances production of 20. Ikebuchi K, Wong GG, Clark SC, Ihle JN, Hirai Y, Ogawahematopoietic stem cells týCFU-S). Biochemn Biophys Res MI (1987) lnterleukin-6 enhancement of intcrleukin-3-Commun 159:933 dependent proliferation Of mnultipotential heinopoictic

9. Ishibashi 1, Kimura H, Shikama Y, Uchida T, Karivone S, progenitors. "froc Natl Acad SOi USA ýý4:9035Hirano T, Kishimoto T'. Takat~uki F, Akiyama Y (1989) 21. Leary AG, Ikebuchi K, Hirai Y, Wong GG, Yang YC., C'larklnterleukin-6 is a potent thrombopoietic factor in vivo in SC, Ogawa M (1988) Synergism between i-erleukin-6mice. Blood 74:1241 and interleukin-3 in supporting proliferation Of humnan

10. Asano S, Okano A, Ozawa K, Nakahata T, Ishibashi T, hematopoietic stem cells: comparison with interleukin- IKoike K, Kimura H, Tanioka Y, Shibuya A, Hirano T, alpha qlo-od 71.1759Kishimoto T, Takaku F, Akiyama Y (1990) In vivo effects 22. Rennich D), Jack-son J, Yang G, Wideman J. Lee F, IlIuda k Sof recombinant human interleukin-6 in primates: stimu- (1989) lnt,:rleukin-6 interad:s with interleuikin-4 andlated production of platelets. Blood 75:1602 other hemnatopoietic growth factors lo selectively enhance

11. Patchen ML, MacVittie TJ, Williams JL, Schwartz GN, the growth of meg.akaryocytic, erythroid, myeloid, andSouza L.M 11991) Administration of interleukin-6 stimu- multipotential progenitor cells. Bloo'd 73:1828'lates multilineage hematopoiesis and accelerates recover 23. Bruno L, Hoffman R (1989) Lffect of interleukin-6 on in,rom radiation-induced htn etc depression. Blood vitro human megakaryocytopoiesis: its interaction with77:472 other cytokines. Exp Hematol 17: 1038

12. C~ohen AM, Zsebo KM, Inoue H, Hines D, Boone TC, 24. Okatto A, Suzuki C, Takatsuki F, Akivamra Y, Koike K,Chazin YR. Tsai I., Ritch T, %:uza LM (198/) Itn vivo stim- Ozawa K, Hirano T, Kkýhimoto T, Nakahata F, Asn Sulation of granulopoiesis by recombinant human granu- ( 1989) In vitro expansion of the murine pluripotentlocyte colony-stimulating factor. Proc Natl Acad Sci USA hemopoietic stern cell population in response to inter-84:2484 leukin-3 and interleukin-6: application to bone m,--w

13. Fujisawa MI, Kobayashi Y, Okabe I, Takaku F, Kon~atsu Y, transplantation. Transplantation 48:495Itoh S (1986) Recombinant human granulocyte colony- 25. Wong GG, Witek-Gianr")tti JS, Toinple P.A', lKriz R, Ferenzstimulating factor induces gratiulocytosis in vivo. Jpn J C, Hewick RM, Clark SC, Ikebuchi K, ()gawa M (19881Cancer Res 77:866 Stimulation of murine heniopoietic colony format ion bn

14. Welte X, Bonilla MA, (Gillio AP', Boone TC, Potter GK, human Il_-6.J Immunol 140)304t0Gabrilove JI., Moore MAS, O'Reiily RT, Souza L.M (1987) 26. Suda T, Yamaguchi Y, Sudla j1, Miura Y, COkano A, AkivamaRecmnbinant human granulocyte colony-stimulating fac- Y (1988) Effect of interlcukin-6 (11,-61 on the differentia-tor: effects on hematopoiesis in normal and cyclophos- tion and proliferation rf murine and hurnan hernopoicticphamide-treated primates. J Exp Med 165:941 progenitors. Exp Hemrnaol 16:891

15. Fujisawa M, Kobayashi Y, (Ikabe F, Takaku F, Komatsu Y, 27. SchulzJ, Almond P1R. Cunningham IR, floltJ IG, Loevinge:itoh S (1986) Recombinant human granulocyte colony- R, Suntharalingam N, Wright KA, Nath It, tempert F)stimulating factor induces granulocytopoiesib in vivro. JPN (1983) A protocol for the dletermination of absorbed dost,J C:ancer lkes /7:866 for high-energy photon an(I electron beams. Med Physiol

16. Patchen MLI, MacVittie TJ iolberg B1D, Souza L.M (1990)) 10:741[herapeutic administration of recombinant human gran1- 28. JAIl IF, McCulloch EA ( 196]) 'direct nwasurcmnent of ',heulIo(yte colony-stitnulating factor accelerates hetnopoietic radiation sensitivity of notimat mou0Lse bone marrow cellIs.regeneration and enhances survival in a murine model of Radiat Res 14:213radiation-induced myclosuppression. Int J Cell Cloning 29. P-tchen Mi., Mac'.'ittie IJ (1986) Hertopoictic effects, of8:107 intraventous soluble glucan adilninistialion. I

17. Koibayashi Y, ()kabe T. lrabe A, Suzuki N, Takaku F (1987) linautiopharmacol 8:407Human granulocyte colotiv-stimulating factor produced] 30. Matsunmoto M, Matsubara S, Matsuno 1, larnura %A,by Escherichia (oh shortens the period of granulocy~orwo-ia ltatti ri K, Nomura 1I, ()no ýM, Yokota 1 (1987) Protectivein~duced by irradiation in mice. JPN J Cancer Res 7h t63 ('ffect of human graniulocyte (oloilv-stimiilating facli ir (

344 Expenmental Hematolog vol. 211(1"3)

crturfobal 111tt-ttionl III neitlropenli( on~ce. Intt-% I InImuruts'I1 40. l.indumnain A,. Kit-d- I I) ( )\,tcr \\ \f1(ut-iVe ,d flohun 1)S 5: 27 15SM it-Osmrrani H BII, I icr rwlatii 1 91188~ 1 (rariu~o I f t-

31. 'Tdkatsuki I., tOkano A, Sutnki I� M.VdlsaskJ Y, Hliram)i 1, nicrphgetlottit NIIIIatiiig 1,(tot foriilidw t~ " 1111kKishinioto F, [jiima 1), -Xki,-inia Y 11990)) Interleukin-0 leukiri- I procita tion h% fiuniarit psiInorpl1(iW(iu leVir HLIe-perfus~ionl stimlaltet-t reconstituftion Ot the Ifl)rnutie and tuijituls. I Irnururiol 140:81-hemiopoiet ic s% Istems atter i- t ui or ift il t rL ITIn-nt. 41. Mooxre RN, )ppt-uherrri 11, 1a rarm 1 , ( jativi abeti.Cancer Res 50:2885 'shadduck RK Iy8(OI ProduliKonl of lsrrijintiate-a fixating

32. Kindler V. Fhorves B, lDeKossodo S, Ablet B, [liason JI-. tactot iintcrlcukiri- I Os mat inroptiagL- at tivatckl v~ith[hatcher 1). Farther N. Vassalli 1I' 1986) Stimulation of kololx--stitmutating flkcuirs Ilinraunol 1251 11)02hewmopoiesis in mi- by recombinant bacterial miurine 42. Shieli J11, Pevtersoin R11,. Moore MA I )Q') I i(,ratiiot!) Steintierteukin- 3. Proc Nail Acad Sci USA 81:100I tolon. -,tiinutating fac~tor od1Keration it (vtu ukiii r-tcl)-

33. \lonrov RI, Skvlls RR, 1lasor P, DInboid A, lDonahne RL, tors in inurinc bonie niarri w. ct-ev li ýin o and Iin %itiroMa~iti I 188 Rcoers frorn severe henlopoiet ic studies. I 11Immuno 14 7.2984

suppression using recont hi narnt human gra nulocyte- 43. McI orwIld 111, ( ittrelt MBI. !swearingvign(i, lift Id( 11t )1 kmnacropjhage colonsy-st nil Lila t i ng faictior. Ex PI lemiatot C omparative- effects ()I throitih upowut in anid interleukin -t16-344 on murine rnegakaryopoiesios aind pIaft-let produt ton

34. Mo -e M0A, Warren t)J ( 1987) Synergy of interleuk11in- I Blood 77:735sand gra nulocyte COIClon-StItn ut.1ating factor: Iin v~ivo stimnu- 44. Baker \%"If, Limianni A, ihang (,'%. kWitiani II t Paietnlation of stern-cell recoverx and iteratopoietic regvnera- Mi., 119921 C om pari son o) in terten kin- I flphaj gmi i

tion following 5-fluorotiracil treatment of mice. Proc Nat) expression Ii n urinte spteeni after let hal ant) sublethOWAcad ScO USA 94:7 134 cobalt-60 irradiatio n. Lxp I lerinatot 20:7171

35. Shimiamura \M, Kobayashi Y, Yuo A, Iiiahe A. Okabe T, 4S. Chang ('M%, Baker %VI I, Limianni A, Williarns 11,I Iragoso I ,Kom iatsu Y, Itohi 1, Ila kaku F ( 1987) Effect of Ounman Patchen M01 (1992) In li w ~gunc expression it if itt-rtcukin-recombinant granu IocvIte cotoly-sti mutating factor on 3i, gra n uOCtocvte/macrti)phagc- c0Olti-st ilni nilatirg tat for.hematopoictic injury in mnice induced Os .S-fluorouracil, and t-kit ligand in murine bone marrow, and splt-co aflvrBlood 69:3,53 sublethal irradiation. I xp I lemiafol 201:77S

36 Mreen (,Cmpbell L., Souza 1.M,, Alton NK, Keech J 46. Tsuji K, /sebo K<. ()gava %,I !199 11 1 tilanriettient o)f('en M, Sheridan W, Metcalf 1), Fox B (1988) Effect of rnurine blast cell colony formnation in culture Os rec~ombi-

granutocyte colony-stimulating factor ton neutropenia nanit rat stern cell factor. tigand for i -kit. Blood 7_8:122induced byV cy-totoxic chemotherapy. Lantcet 1:667 47. Bodine D%1, ()rlic 1), Birket-t NC, Seidel NI-, /Aeb(u KM

37. (;abrilove JE., jak-ubowskj A, Scher H, Sternberg C. Wong (1992) 'item cell factor increase,, toloinv-tortiiig unit(i (rous J, Yagoda A, Fain K, Moore MAS, Clarkson B, spleen colony- numbers in vi tro Iii sv.*nergx with inter-

Oettgen Hf F, Alton K, Welte K, Souza LM (1988) Effect of leukin-6, and in v-ivo inl Sf51 mice as at single, tiatfor.a granulocyte colony-stimulating factor on neutropenia Blood 79:913and associated mrbulidity due to chemotherapy for transi- 48. Ulrich TR, lDel(Iastillo j, NcNiece 1K, Yi E-S, Al mia ( 1), Yintional-celt carcinoma of the urotheliurn. N E~ngI J Med S, /sebo KM (1991) Stemn cell factor in combination wvith3 18:1414 granulo:cyte colony-stimrulating factor WFt A or grantiti -

38. (Greenberg lIT, Negrin B, Nagler A (199011 The use of cyte-macrophage (CSF- synvrgisticallv increadses grarin-I aemopoietic growth factor iii the treatment of lopoiesis in vivo. Blood 78: 1954rnyelodysplastic syndromes. Cancer Surv 9:199 49. Mlolineux G, Migdatska A, Szmitkowvski M, /eh KMI

39. Washimuka F, Koike T, To~ba K, Nagai K, I akahashi M, D)exter TM 11991) Effects oti hernatopoiesis of ret oini-Shibata A (1992) A rise of ervthrocytes and platelets in a nant stem cell factor (ligantl for c-kitu admiinistcrt-d ul viVopatient with myelodysplastic syndrome during adminis- to mice either alone or in (-ombinat ion wvith granuflocytetratitln of G-( SF, Am J Hematol 39: 153 colony-stimulating factor. Bloodl 78:901

ARMED FORCES RADIOBIOLOGY003 "7227/91111323 0946503.00:0 RESEARCH INSTIrTU TEEndocrinIolo gA' ,N":

CEopright c CIENTIFIC REPORT VT c r13, NCS A

SR93-8

Synergistic Roles of Interleukin-6, Interleukin- 1, andTumor Necrosis Factor in the AdrenocorticotropinResponse to Bacterial Lipopolysaccharide in Vivo*

ROBERT S. PERLSTEIN, MARK H. WHITNALL, JOHN S. ABRAMS, EDWARD H. MOUT GEY,AND RUTH NETA

Department of Experimental Hematology (R.S.P., R.N.) and the Department of Physiology (M.H. W.), ArmedForces Radiobiology Research Institute, Bethesda, Maryland 20889-5145; Neuroendocrinology andNeurochemistry Branch, Depat •ment of Medical Neurosciences (E.H.M.), Walter Reed Army Institute ofResearch, Washington, D.C. 20307-5100; and the Department of Immunology, DNAX Research Institute(J.S.A.), Palo Alto, California 94304

ABSTRACT release, we used blocking antibodies to !L-6, TNF, and the IL-AAdministration of lipopolysaccharide (LPS) results in activation of receptor. Our results demonstrate that anti-IL-6 antibody abrogated

the hypothalamic-pituitary-adrenal axis. LPS induces the release of a ACTH induction throughout the course of the response both 2 and 4 hnumber of proinflammatory cytokines, i.e. interleukin-l (IL-1), IL-6, after LPS challenge. In contrast, anti-IL-I receptor and anti-TNFand tumor necrosis factor (TNF), which activate the hypothalamic- antibody, given individually, blocked ACTH production at 4 h, but not

at 2 h. Only combined administration of these two antibodies dimin-pituitary-adrenal axis as well and may mediate the effects of LPS. ished, but did not eliminate, ACTH release at 2 h. This is the firstVariations in the kinetics of appearance of IL-1, TNF, and IL-6 after demonstration that all three inflammatory cytokines are obligatory' forLPS challenge suggested that these cytokines may play different roles LPS-induced elevation of plasma ACTH. In addition, these resultsat different times. To elucidate the mutual dependence and contribu- suggest that IL-i, IL-6, and TNF play different roles in LPS-inducedtion of individual cytokines in the course of LPS-induced ACTH ACTH release. (Endocrinology 132: 946-952, 1993)

NFLAMMATION and/or infection lead to activation of suggested that IL-A may need to interact with the IL-6 itthe hypothalamic-pituitary-adrenal (H-P-A) axis (1, 2). induces endogenously in stimulating ACTH release. This

For many years, this phenomenon was studied in models hypothesis was further supported by our finding that pre-employing lipopolysaccharide (LPS), a component of bacte- treatment with murine monoclonal anti-IL-6 antibodyrial cell walls of gram-negative bacteria (3-6). More recently, blocked the IL-I-induced ACTH response (11).a number of proinflammatory cytokines, i.e. interleukin-1 LPS induces the release of IL-A, TNF, and IL-6 (14), which(IL-t), IL-6, and tumor necrosis factor (TNF), were shown to may mediate its stimulatory effect on the H-P-A axis. There-similarly activate the H-P-A axis both in vivo and in vitro (7- fore, the use of cytokine blocking antibodies to modulate the9). The finding that these biochemically distinct cytokines LPS-induced ACTH response should aid in elucidating thehad similar effects suggested redundancy. Our previous mutual dependence and contribution of endogenously pro-work, however, indicated that interaction of these cytokines duced individual cytokines. Indeed, Rivier et al. (5) reportedwas required for ACTH induction (10, 11). that monoclonal anti-iL-1 receptor antibody partially blocks

More specifically, we demonstrated in C3H/HeN mice the H-P-A response to LPS in mice (5). In addition, depletionthat within 2 h of ip administration, IL-1 is a potent inducer of cytokines, in particular IL-I, by destruction of macro-of ACTH, whereas pharmacological amounts (up to 10 ug) phages using liposome-encapsulated dichloromethylene di-of IL-6 induced only a negligible response (10). However, phosphonate blocks the H-P-A response to subpyrogenicthe combination of IL-1 and IL-6 produced a synergistic amounts of LPS in rats (6).response within 30 min of injection (10), and IL-1 induces Moreover, variations in the kinetics of appearance of IL-IL-6 within 2 h of injection (11-13). Together, these results 1, TNF, and IL-6 after LPS challenge have been observed

(13, 15-21), suggesting that these cytokines may play differ-Received July 28, 1992. ent roles at different times. TNF levels were consistentlyAddress all correspondence and requests for reprints to: Dr. Robert found to peak approximately I h after LPS administration

S. Perlstein, USAF MC, EXH, AFRRI, Bethesda, Maryland 20889-5145 and then rapidly declined (13, 15-20), in part probably" This work was supported by the Armed Forces Radiobiology Re-

search Institute, Defense Nuclear Agency, under work units 00129 and because TNF release is especially sensitive to negative feed-00105. The views presented in this paper are those of the authors; no back by the glucocorticoid end product of H-P-A activationendorsement by the Defense Nuclear Agency or the Department of (18, 19). In contrast, IL-I and IL-6 levels were found to peakDefense has been given or should be inferred. Research was conducted somewhat later (within 2-4 h) and were sustained longeraccording to the principles enunciated in the Guide for the Care andUse of Laboratory Animals prepared by the Institute of Laboratory (15, 17- 19, 21). It was, therefore, postulated that TNF initiAnimal Resources, National Research Council. ates, while IL-I and IL-6 sustain, H-P-A activation after LPS

946

CYTOKINES AND ACTH RESPONSE TO LPS 947

exposure (8). The ACTH antibody used in this assay is derived from rabbits immuIn this report, we present results which indicate that IL-I, nized against ACTHi-(1-24), a region that is identical in human and

murine~~~~~~~ A fIsThthehlsestvtofhiassay was 8 pgiml.IL-6, and TNF are required for LPS-induced ACTH induc- mutne ACTls. The threshold sensitivity of this ation, and their relative contributions depend on the timeinterval after LPS cla!!enge. Statisticud adrwiysils

In Figs. I and 3, evaluation of the results was carried out usinganalysis of variance, followed by the Scheffe F test. In Figs. 2 and 4,

Materials and Methods comparison of the response to each cytokine treatment at each timepoint with the response to simultaneously injected vehicle was made

Experimental animals using Student's t test. Comparison of the response to combined cytokinetreatment with the sum of the responses to each cytokine treatment

Female C3H/HeN mice were purchased from the Animal Genetics given separately at each time point was made as a I degree of freedomand Production Branch, NCI (Frederick, MD). Mice were handled as contrast. For each time point, each P vdlue stated reflects a Bonferronipreviously described (10). correction for the number of tests run.

In the first set of experiments, groups of four to six mice were injectedip with vehicle (0.5 ml pyrogen-free normal saline), control antibody, orantibodies directed against the IL-I receptor, IL-6, or TNF at 1630 h onday 1. At 0800 h the next morning (day 2), LPS was administered ip to Resultsall of the pretreated groups as well as a group that had not received anypretreatment. Either 2 or 4 h later, unanesthetized mice were decapitated LPS-induced A CTH release(model 130 Rodent Decapitator, Harvard Apparatus, South Natick, MA)with minimal stress to obtain plasma samples for ACTH. The ACTH levels in the plasma of mice receiving various

In the second set of experiments, groups of four to six mice were amounts of LPS at 2, 4, and 6 h are presented in Table 1.injected ip with vehicle, recombinant human IL-la (rhiL-la), rhlL-6, The administration of all doses of LPS resulted in a maximalrecombinant human TNFa (rhTNFa), or combinations of these cytokines ACTH response at 2 h, which progressively diminished at 4and decapitated 30-180 min later. In a final set of experiments, groups and 6 h. All maximal ACTH responses at 2 h were similar.of four to six mice were pretreated with vehicle and antibodies on day1, as described in the preceding paragraph, injected with a combination Therefore, we chose 1 jug LPS to study the modulation of theof rhIL-la and rhTNFa at 0800 h on day 2, and decapitated 120 rmin 2 h ACTH response to LPS. The 4 h ACTH response waslater. similar after 5-50 Ag LPS. Therefore, we chose 5 ug LPS to

In addition, 5-10 noninjected control mice were killed on the day of study the modulation of the 4 h ACTH response to LPS. Theeach experiment. 6 h ACTH response after all doses of LPS injected was not

substantial enough to allow further study. Thus, the magni-Cytokines and LPS tude of the ACTH response to LPS in C3H/HeN mice is less

rhIL-la (117-271 Ro 24-5008, lot IL-I 2/88; SA, 3 x 108 U/mg) was than that observed in BALB/c mice by Rivier et al. (5). Thisgenerously provided by Dr. Peter Lomedico, Hoffman LaRoche, Inc. probably is related to genetic differences between these two(Nutley, NJ). rhlL-6 (SDZ 280-969, batch PPG 9001; SA, 5.2 x I07 U/ strains.mg) was a gift from Dr. E. Liehl, Sandoz, Vienna, Austria). rhTNFa (lotCP4026P08; SA, 9.6 x 106 U/mg) was provided by Biogen (Cambridge,MA). LPS (protein free; prepared from Escherichia coli K235 by the Effect of a:ntibody pretreatment on the plasma level of ACTH 2phenol-water extraction method) was kindly provided by Dr. Stefanie h after challenge with LPSVogel, Uniformed Services University of the Health Sciences (Bethesda,MD). The recombinant cytokines were diluted in 0.5 ml pyrogen-free Figure IA demonstrates the effect of pretreatment withsaline on the day of injection. anti-IL-6 antibody, anti-ILiR antibody, anti-TNF antibody,

the combination of anti-IL-1R antibody and anti-TNF anti-Antibodies body, or antigalactosidase antibody on the 2 h ACTH re-

Rat monoclonal antibody to mouse rIL-6 (MP5 20F3) was prepared sponse to 1 jig LPS. Pretreatment with anti-IL-6 antibodyusing semipurified Cos-7 mouse IL-6 as an immunogen, as previously completely blocked the response to LPS, while the combi-described (22). Rat monoclonal antibody to 0-galactosidase (GL 113) nation of anti-TNF antibody and anti-IL-1R antibody onlywas used as an isotype control. Rat monoclonal immunoglobulin GI, partially blocked the response. In contrast, pretreatment withantimurine IL-I receptor (anti-IL-IR) antibody (35F5) (23) was gener-ously provided by Dr. R. Chizzonite, Hoffman LaRoche. Hamster mono- TABLE 1. Plasma ACTH levels after ip injection of LPSclonal antibody to murine TNFa (TN3.19.12) (24) was a kind gift fromDr. R. Schreiber, Washington University (St. Louis, MO). The antibodies LPS dosewere diluted in 0.5 ml pyrogen-free saline on the day of injection. The (,g) 2h 4h 6hamount of antibody injected (anti-IL-6, ani-IL-IR, and anti-TNF) wasapproximately the same as the quantity we used in earlier work to block Vehicle 69.2 ± 8.8 66.3 ± 8.51 68.6 ± 3.86LPS-, IL-I-, and TNF-induced radioprotection (11, 25). Moreover, the 1 185.7 ± 6.0 109.8 ± 8.4 65.6 ± 6.3amount of anti-lL-IR antibody used (250 Aug) was similar to the quantity 2 211.0 ± 18.7 102.0 ± 5.8 65,0 ± 3.0of the same antibody (200 gg) found to be effective by Rivier et al. (5) 5 165.8 ± 15.8 142.9 ± 7.6 69.4 ± 5.3in partially blocking LPS-induced ACTH release and reducing by 90% 10 170.8 ± 7.3 130.4 ± 4.3 61.0 ± 2.9IL-I-induced leucocytosis. Noneof the antibodies injected by themselves 25 178.2 ± 12.2 135.6 ± 8.7 102.0 ± 17.9had an effect on ACTH release. 50 170.0 ± 15.0 145.0 ± 11.1 105.0 ± 9.4

Values are expressed as picograms per ml. Female C3H/HeN mice

Measurement of ACTH in plasma received various amounts of LPS ip and then were decapitated toobtain plasma for ACTH measurements 2, 4, or 6 h later. Each value

ACTH was assayed in plasma from decapitated mice using an "'I shown is the mean ± SEM for 5 animals, except for the vehicle values.RIA kit (INCSTAR Corp., Stillwater, MN), as previously described (10). which represent 10 animals each.

948 CYTOKINES AND ACTH RESPONSE TO LPS End,,.M iVAI Q No 1

2a A administration of 10 ng rhlL-la and I ug rhTNFx resultedin a significant increase in circulating ACTH at 30, 60, 120,and 180 min compared with the response to simultaneously

E io injected vehicle (Fig. 2). The responses to simultaneouslyinjected vehicle were inconsequential (Fig. 2). When the

S.e responses to the rh[L-I/rhi NF combination were compared

0 with those achieved with 10 ng rhlL-hI or I ug rhTNFa00 ogiven separately, th~e responses to the combined injection

were significantly greater than the sum of the responses toeach cytokine injected alone at 120 and 180 min (Fig. 2).

50

Effect of antibody pretreatment on plasma ACTH 2 h after thecombined injection of rhJL-Jo and rhTNFa

V0h-c-* LIPS ,A-A. aIL-1R+ aiL-IR. aTNF, aCd, Figure 3 demonstrates the effect of pretreatment with anti-LPS aTnF. LPS LPS LPS IL-6 antibody, anti-IL-IR antibody, or antigalactosidase an-

LIPStibody on the 2 h ACTH response to the combined injection

160.8 a of 10 ng rhlL-la and I pg rhTNFa. Pretreatment with dnti-IL-6 antibody was as effective as anti-IL-1R antibody inblocking the ACTH response to the combined rhlL- 1/rhTNF

120 injection. Pretreatment with either of these antibodies pro-duced a significant decline compared to pretreatment with

b control antibody.

Release of ACTH after the injection of a combination of40. rhTNFa and rhIL-6

We previously observed that suboptimal amounts of rhlL-0- 1 a and rhlL-6 synergistically stimulate the release of ACTH

VehiC1. LPS aIL- ,,L-AR. , aTNF, a", (12). To determine whether a similar interaction occurs be-tPS LPS LPS LPS tween rhTNFa and rhlL-6, we evaluated the effect of the

FIG. L C3H/HeN mice received ip injections of antibody 1600 pg anti- combined injection of rhTNFa and rhlL-6. After the com-IL-6 (alL-6), 600 tpg a-galactosidase (aGal), 250 jg alL-1R, 100 Ag bined administration of 1 tg rhTNFa and 1.25 pg rhlL-6, aaTNF, or 250 pg aIL-IR and 100 pg aTNF combined] 16 h before ipchallenge with 1 pg LPS (A) or 5 ug LPS (B). Other mice were significant increase in circulating ACTH was observed at 30,administered vehicle, 1 pg LPS (A), or 5 pg LPS (B) without antibody 60, 120, and 180 min compared with the response to simul-pretreatment. Blood samples were obtained 2 h (A) or 4 h (B) after taneously injected vehicle (Fig. 4). The responses to simul-LPS or vehicle alone. Each bar represents the mean ± SEM for 8-34 taneously injected vehicle were inconsequential 'Fig 4).animals. a, P < 0.05 vs. vehicle alone; b, P < 0.05 vs. aGal plus LPS. avl

ISO a.banti-TNF, anti-IL-iR, or control antibody did not attenuate ---the ACTH response. -0-. ,gl•

16 -0••-1 V•,t

**g*-VehIC5* .Effect of antibody pretreatment on the plasma level of ACTH 4 E 140

h after challenge with LPS S120.Figure 1B demonstrates the effect of pretreatment with

anti-IL-6 antibody, anti-IL-IR antibody, anti-TNF antibody, 00or antigalactosidase antibody on the 4 h ACTH response to5 lg LPS. Pretreatment with any of the three anticytokine so. -

antibodies alone substantially blocked the ACTH response -------------..

to LPS, while pretreatment with control antibody had no 60effect. All of the anticytokine antibodies were equally effc-tive. 40

20 40 60 60 100 120 140 160 160 200

Release of ACTH after the injection of a combination of rhIL- Time (ndn)

la and rhTNFa Fio. 2. Comparison of the time course of increase in plasma ACTH inC3H/HeN mice after ip injection of 10 ng rhL -la combined with I ug

Preliminary experiments indicated that rhTNFa adminis- rhTNFa, 10 ng rh[L-l., or 1 pg rhTNF-,. The mean vehicle responsesat each time point are also shown. Each time point represents the meantered ip to mice by itself induced a minimal ACTH response. ± SEM of hormone determinations for 6 28 animals. a, P < 0.05 vs. the

Therefore, we examined the effect of the combined injection response to simultaneously injected vehicle; b, 11 < 0.05 us. the sum ofof suboptimal amounts of rhlL-la and rhTNFa. Combined the responses to rhIL-la and rhTNFa injected separately.

CYTOKINES AND ACTH RESPONSE TO LPS 949

160 a 35). Our results demonstrate that IL-6 plays a fundamental

140 role in LPS-induced ACTH release, but the participation andinteraction of IL-1 and TNF are also required. In addition,

E 120 a.b the relative importance of these three cytokines varies at?• 1o different times after LPS challenge.

Pretreatment with anti-IL-6 antibody completely abro-- 0 s gated the ACTH response to LPS 2 and 4 h after injection.

TFurthermore, the synergistic induction of ACTH after the

60 combined injection of rhTNFa and rhIL-la was blocked by40i anti-IL-6 antibody.

Inexplicably, although we were able to eliminate LPS-20 induced ACTH responses by pretreatment with anti-IL-6

antibody, ip administration of large doses (10 pg) of IL-6 to

VO ---- mice elicited only a minimal response (10). Vhis suggests thatT+I Tr++ IL-6 in the circulation may require an additional factor(s) to

FIG. 3. C3H/HeN mice received ip injection of antibody [600 jg anti- induce ACTH release. Alternatively, it is possible that sys-IL-6 (aiL-6), 600 mg a-galactosidase (aGal), or 250 jug aIL-iRI 16 h temic IL-6 does not reach the necessary local site(s) in thebefore ip injection of 10 ng rhIL-la combined with 1 jg rhTNFa (T+I). brain, whereas the anti-IL-6 antibody neutralizes LPS- or IL-Other mice were administered vehicle without antibody pretreatment. I-stimulated IL-6 produced in the hypothalamus and/orBlood samples were obtained 2 h after T+I or vehicle alone. Each barrepresents the mean ± SEM for 13-15 animals. a, P < 0.05 vs. vehicle pituitary gland (36-39). If a cofactor(s) is required for IL-6 toalone; b, P < 0.05 vs. aGal plus T+I. stimulate ACTH release, it is not clear at what level the

interaction takes place, e.g. at the target cell level or due to1 ,0 facdlitated transport across the blood-brain barrier.140 Our results suggest that both IL-I and TNF play important

-o-1ug M .1.2.ugL-4 roles as IL-6 cofactors. The ACTH response 2 h after LPS

1*0 challenge was not blocked by pretreatment with anti-IL-1R20•o. .. ,W.veie or anti-TNF antibody given separately, but was diminished

C 1 by the combination of these antibodies; moreover, pretreat-ment with either anti-IL-1R or anti-TNF antibody alone

< totally blocked the 4 h ACTH response. Our previous obser-"9. • avations that IL-I and IL-6 synergize in inducing ACTH

go release (10) and that the 2 h ACTH response to IL-I may bedependent upon an obligatory interaction between IL-I and

.. ... the IL-6 it induces endogenously (11) further suggest that60 . IL-1 is an important cosecretagogue for IL-6 in stimulating

20 40 60 80 100 120 140 160 180 200 the H-P-A axis. The coinjection of rhTNFa and rhIL-6 re-Time (min) suited in a greatly augmented (but not synergistic) ACTH

FIG. 4. Comparison of the time course of increase in plasma ACTH in response, suggesting a lesser role for TNF-IL-6 interaction inC3H/HeN mice after ip injection of 1 jug rhTNFa combined with 1.25 ACTH induction.;Ag rhIL-6, 1 jig rhTNFa, or 1.25 jig rhIL-6. The mean vehicle responses Since anti-IL-6 antibody totally abrogated ACTH release,at each time point are also shown. Each time point represents the mean while the combination of anti-IL- I R and anti-TNF antibodies+ SEM of hormone determinations for 7-28 animals. a, P < 0.05 vs. the only partially blocked the ACTH response 2 h after LPSresponse to simultaneously injected vehicle; b, P < 0.05 vs. the re-sponses to rhTNFa or rhIL-6 injected separately. administration, it is possible that in addition to IL-I and TNF,

other factors cooperate with IL-6. Among these, the arachi-When the early responses to the rhTNF/rhIL-6 combination donic acid cascade metabolites, i.e. prostaglandins, leuko-were compared with those to 1 jg rhTNFa or 1.25 pg rhIL- trienes, and epoxygenase products, which have been shown6 given separately, the responses to the combined injection to modulate CRH release from the hypothalamus (40) andwere significantly greater than the responses to each cytokine ACTH release from the pituitary (41) in vitro, seem likelyinjected alone (but not significantly greater than the sum of candidates. Other possible factors are histamine (3) and IL-2the responses to each cytokine injected alone; Fig. 4). (8, 9, 42).

In addition to directly stimulating the hypothalamus andDiscussion pituitary in conjunction with IL-6, LPS-induced IL-I and

TNF also contribute to stimulation of the H-P-A axis byPrevious studies demonstrate that IL-1, IL-6, and TNF inducing IL-6 production. In contrast to observations with

each stimulate the H-P-A axis in vivo via a CRH-dependent anti-IL-6 antibody, pretreatment with either anti-IL-IR ormechanism (7-9, 26-32) and in vitro at the level of the anti-TNF antibody blocked the 4 h, but not the 2 h, ACTHhypothalamus and pituitary (7-9, 27, 33, 34). On a molar response to LPS. The greater efficacy of these antibodies atbasis, IL-1 is a more potent stimulator than TNF or IL-6 (31, 4 h may be due in part to their ability to interfere with TNF/

950 CYTOKINES AN[) ACTH RESPONSE TO LPS :v.,.- ,

IL-A induction of IL-6. There is ample evidence that LPS- addition, may play an interactive role with IL-6 at both timeinduced elevation of IL-6 depends upon IL-I and TNF in- points. The definitive explanation of how these cytokinesduced by LPS. LPS stimulates the release of IL-I and TNF mediate the activation of the Ht-P-A axis by LI'S will havein vitro (43, 44), including the production of IL-I in the to take into account the contribution of cytokines producedhypothalamus and pituitary (45, 46) and TNF in central in the hypothalamus and pituitary gland, and how anj ifnervous system microglial cells (47). In vivo, serum levels of they are induced by circulating cytokines originating in theTNF peak before IL-I and IL-6 after LPS administration (13, periphery.15-21). TNF and IL-i, in turn, both stimulate the release ofIL-6 (11-13, 43, 44, 48-50). TNF is a much less potentinducer of IL-6 than IL-1 in mice (11). This may help to Acknowledgments

explain why ip injection of TNF stimulated only a minimal We thank Drs. T. ), MacVittie and C D. Lednev for critically re' iewingACTH response in mice, in contrast to reports of more this manuscript; Drs, P. Lomedico, E. Liehl, S. Vogel, R Chizzonite, andsubstantial ACTH responses after iv TNF administration to R. Schreiber and Biogen for generously providing cvtokines and anticy-rats (32, 51). At the local level, subpopulations of nonneu- tokine antibodies; Dr W. !ackson for assisting with the statistical analysis

of the data; and Petty Officr Sam Tom, Miss Faith Seizer, and Mr. Clintronal cells in both the hypothalamus (36) and pituitary (37, Wormley for their technical assistance.38) of rats spontaneously produce IL-6, and IL-i enhancesthe release of IL-6 from cultures of ra :-iterior pituitary cells(39). References

The synergistic induction of ACTH after coinjection of 1. Dinarello CA 1984 Interleukin- 1 and the pathogenesvr ol the acuterhIL-ln and rhTNFn was completely blocked by pretreat- phase response. N EngI J Med 311:1413-1418ment with anti-IL-6 antibody, suggesting that these cytokines 2- Sibbald W), Short A, Cohen MP, Wilson RF 1977 Vanations insynergistically induced IL-6 to produce the ACTH response. adrenocortical responsiveness during severe bacterial infectionsIndeed, recent in vivo (13) and in vitro (52) studies have Unrecognized adrenocortical insufficiency in severe bacterial infec-tions. Ann Surg 186:29-33demonstrated that IL- I and TNF can synergistically stimulate 3. Nakano K, Suzuki S, Oh C 1987 Significance of increased secretionIL-6 production. of glucocorticoids in mice and rats injected with bacterial endotoxin

Previous work employing blocking antibodies to TNF or Brain Behav Immunol 1:159-172IL-I provides further support for the hypothesis that in- 4. Yasuda N, Greer MA 1978 Evidence that the hypothalamus me-pve n diates endotoxin stimulation of adrenocorticotropic hormone ,cre-creased IL-6 levels during inflammation are dependent on tion. Endocrinology 102:947-953TNF and IL-1. Results from our own laboratory (unpub- 5. Rivier C, Chizzonite R, Vale W 1989 In the mouse, the activationlished), as well as reports by a number of other investigators of the hypothalamic-pituitary-adrenal axis by a lipopolysaccharide(13, 15, 17, 22) show that pretreatment with anti-TNF anti- (endotoxin) is mediated through interleukin-l Endocrinology

bodysubtanialy dminshe IL- 2- h fte adinitraion 125:2800-2805body substantially diminished IL-6 2-4 h after administration 6. Derijk R, Van Rooijen N, Tilders FJH, Besedovsky HO, Del Reyof LPS [as well as LPS-induced increases in IL-1 (15, 17)]. In A, Berkenbosch F 1991 Selective depletion of macrophages pre-

addition, we observed that pretreatment with anti-IL-IR vents pituitary-adrenal activation in response to subpyrogenic, butantibody markedly diminished the IL-6 response 4 h after not to pyrogenic, doses of bacterial endotoxin in rats. Endocrinology

129:330-338LPS administration (unpublished). Similarly, pretreatment 7. Whitnall MH, Regulation of the hypothalamic corticotropin-releas-with anti-IL- IR antibody significantly attenuated the plasma ing hormone neurosecretory system. Prog Neurobiol, in pressIL-6 response to a turpentine-induced sterile abscess in mice 8. Eskay RL, Grino M, Chen HT 1990 Interleukins, signal transduc-(53). tion, and the immune system-mediated stress response. Adv Exp

Med Biol 274:331--343It is also likely that LPS directly stimulates the release of 9. Inura H, Fukata J, Mori T 1991 Cytokines and endocrine func-IL-6, especially 2 h after injection. The ability of LPS to tion-an interaction between the immune and neuroendocrine sys-induce IL-6 in various cell cultures has been observed by a tems-review. Clin Endocrinol (Oxf) 35:107-115number of investigators (54, 55), including nonneuronal cells 10. Perlstein RS, Mougey EH, Jackson WE, Neta R 1991 Interleukin-

1 and interleukin-6 act synergisticatly to stimulate the release ofin the hypothalamus (36) and pituitary (37, 39), and Romero adrenocorticotropic hormone in vivo. Lymphokine Cytokine Reset al. (56) reported that IL-I receptor antagonist blocks IL- 10:141-14610-induced, but not LPS-induced, IL-6 release from cultures 11. Neta R, Perlstein R, Vogel SN, Ledney GD, Abrams J 1992 Roleof rat anterior pituitary cells. In addition, in a recent in vivo of interleukin 6 (IL-6) in protection from lethal irradiation and instudy, IL-1 receptor antagonist did not block IL-6 induction endocrine responses to IL-I and tumor necrosis factor. J Exp Med

175:689-694after the administration of sublethal amounts of LPS (57). 12. Neta R, Vogel SN, Sipe JD, Wong GC, Nordan RP 1988 Compar-

In summary, in mice injected with sublethal amounts of ison of in vivo effects of human recombinant IL I and humanLPS, IL-6, IL-1, and TNF play different roles in initiating and recombinant IL 6 in mice. Lymphokine Res 7:403 -412

sustaining an ACTH response. The presence of IL-6, derived 13. Shalaby MR, Waage A, Aarden L, Espevik T 1989 Endotoxin,tumor necrosis factor-a and interleukin I induce interleukin ofrom the direct effects of LPS and/or its induction by TNF/ production in vivoi. Clin Immunol Iramunopathol 53:488-498

IL-1, is obligatory at both time points studied. However, to 14. Vogel SN, Hogan MM 1990 Role of cytokines in endotoxin-meelicit an ACTH response, an interaction with another factor diated host responses. In: Oppenheim Ii, Shevach EN (eds) Immumay be required. IL-I and TNF appear to be essential in nophysiology, Role of Cells and Cytokines in Immunity and Inflam-

mation. Oxford University Press, Oxford, pp 238-258sustaining the IL-6 levels required to maintain an ACTH 15. Fong Y, Tracey KJ, Moldawer LL, Hesse DG, Manogue KB,response, especially 4 h after LPS administration and, in L,enney JS, Lee AT, Kuo GC, Allison AC, Lowry SF, Cerami A

CYTOKINES ANT) ACTH RESPONSE TO LPS 951

1989 Antibodies to cachectin/tumor necrosis factor reduce interleu- the release of prostaglandin E, from rat hypothalamic explants inkin 1fi and interleukin 6 appearance during lethal bacterernia. J Exp vitro. Neuroendlocrinology 56:61-68Med 170:1627-1633 35. Warren RS, Fletcher H, Starnes 1, Alcock N, Calvano S, Brennan

16. Waage A, Haistensen A, Shalaby MR, Brandtzaeg P, Kierulf P, MF 1988 Humoral and metabolic response to recombinant humanEspevik T 1989 Local production of tumor necrosis factor alpha, tumor necrosis factor in rat: in vitro and in vivo. Am J Physiolinterleukin 1, and interleukin 6 in meningococcal meningitis. Rela- 255:E206-E21 2tion to the inflamnmatory response. J Exp Med 170:1859-1867 36. Spangelo BL, Judd AM, MacLeod RM, Goodman DW, lsakson

17. Zanetti C, Heumann D, Gerain J, Kohler J, Abbet P, Barras C, PC 1990 Endotoxin-induced release of interleukin-6 from rat medialLucas R, Glauser M-P, Baumgartner J-D 1992 Cytokine production basal hypothalami. Endocrinology 127:1779-1785after intravenous or peritoneal gram negative bacterial challenge in 37. Spangelo BL, MacLeod RM, Isakson PC 1990 Production of inter-mice. J Immunol 148:1890-1897 leukin-6 by anterior pituitary cells in vitro. Endocrinology 126:582-

18. Chensue SW, Terebuh PD, Remick DG, Scales WE, Kunkel SL 5861991 In vivo biologic and immunohistochemical analysis of interleu - 38. Vankelecomn H, Carnieliet P, Van Damme J, Billiau A, Denef Ckin-I alpha, beta and tumor necrosis factor during experimental 1989 Production of interleukin-6 by folliculto- steliate cells of theendotoxemia. Am J Pathol 138:395-402 anterior pituitary gland in a histiotypic cell aggregate culture system.

19. Zuckerman SH, Shelthaas J, Butler LD 1989 Differential regulation Neuroendlocrinology 49:102-106of lipopolysacchanide-induced interleukin 1 and tumor necrosis 39. Spangelo BL, Judd AM, lsakson PC, MacLeod RM 1991 lnterleu-factor synthesis: effects of endogenous and exogenous glucocorti- kin- I stimulates interleukin-6 release from rat anterior pituitary cellscoids and the role of the pituitary-adrenal axis. Eur J Immunoil in vitro, Endocrinology 128:2685-269219:301-305 40. Bernarditti R, Chiarenza. A, Calogero AE, Gold PW, Chrousos

20. Michie HIR, Manogue KR, Spriggs DR, Revhaug A, O'Dwyer S, GP 1989 Arachidlonic acid metabolites modulate rat hypothalamicDinarello CA, Cerami A, Wolff SM, Wilmore DW 1988 Detection corticotropin -releasing hormone secretion in vitro. Neuroendocri-of circulating tumor necrosis factor after endotoxin administration. nology 50:708-715N EngI J Med 318:1481-1486 41. Cowell AM, Flower RJ, Buckinghams JC 1991 Studies on the roles

21. Fang Y, Moldawer LL, Marano K, Wei H, Tatter SO, Clarick RH, of phospholipase A2 and eicosanoids in the regulation of corticotro-Santhanami U, Sherris D, May LT, Sehgal PB, Lowry SF 1989 phin secretion by rat pituitary cells in vitro. I Endocninol 130:21-32Endotoxemia elicits increased circulating 02-IFN/]L-6 in man. j 42. Cambronero JC, Rivas Fl, Borrell J, Guaza C 1992 lnterleukin-2Immunoil 142:2321-2324 induces corticotropin-releasing hormone release from superfused rat

22. Starnes HF, Pearce MK, Tewari A, Yim. 11, Zou J-C, Abrams JS hypothalami: influence of glucocorticoids. Endocrinology 13 1:677-1990 Anti-IL-6 monoclonal antibodies protect against lethal Esche- 683richia coli infection and lethal tumor necrosis factor-a challenge in 43. Dinarello CA 1991 Interleukin-I and interleukin-l antagonism.mice. J Immunoil 145-.4185-4191 Blood 77-.1627-1652

23. Chizzonite R, Truitt T, Kilian PL, Stern AS, Nunes P, Parker KP, 44. Neta R, Sayers TJ, Oppenheimn JJ 1992 Relationship of TNF toKaffka KL, Chua AO, Lugg DK, Gubler U 1989 Two high-affinity interleukins. In: Aggarwal BB, Vilcek J (eds) Tumor Necrosis Factors:interleukin-l receptors represent separate gene products. Proc NatI Structure, Function, and Mechanism of Action. Marcel Dekker, NewAcad Sai USA 86:8029-8033 York, pp 499-566

24. Sheehan KCF, Ruddle NH, Schreiber RD 1989 Generation of 45. Rettori V, Dees WL, Hiney JK, Milenkovic L, McCann SM.hamster monoclonal antibodies that neutralize tumor necrosis fac- Interleukin-1 alpha (IL-ler)-irnmunoreactive neurons in the hypo-tars. J Immunol 142:3884-3893 thalamus of the rat are increased after lipopolysaccharidle (LPS)

25. Neta R, Oppenheim JJ, Schreiber RD, Chizzonite R, Ledney GD, injection. 74th Annual Meeting of The E.idocrine Society. SanMacVittie TJ 1991 Role of cytokines (interleukin 1, tumor necrosis Antonio TX, 1992, p 185 (Abstract 534)factor, and transforming growth factor 0) in natural and lipopoly- 46. Koenig JI, Snow K, Clark 81D, Toni R, Cannon JC, Shaw AR,saccharide-enhanced radioresistance. J Exp Med 173:1177-1182 Dinarello CA, Reichlin S, Lee SL, Lechan RM 1990 Intrinsic

26. Spangelo BL, MacLeod RM 1990 Regulation of the acute phase pituitary interleukin-1 beta is induced by bacterial lipopolysaccha-response and neuroendocrine function by interleukin 6. Progr Neu- ride. Endocrinology 126:3053-3058roendocrin Immunoil 3:167-175 47. Ricciardi-Castagnoli P, Pirami L, Righi M, Sacerdote P, Locatelli

27. Lyson K, McCann SM 1991 The effect of interleukin-6 on pituitary V, Bianchi M, Sassano M, Valsasnini P, Shamnsah S, Panerai AEhormone release in vivo and in vitro, Neuroendocrinology 54:262- 1990 Cellular sources and effects of tumor necrosis factor-a on266 pituitary cells and in the central nervous system. Ann NY Acad Sci

28. Naitoh Y, Fukata J, Tominaga T, Nakai Y, Tamai S, Morl K, 594:156-168Imura H 1988 Interleukin-6 stimulates the secretion of adrenocor- 48. Libert C, Brouckaert P, Shaw A, Fiefs W 1990 Induction ofticotropic hormone in conscious, freely-moving rats. Biochem Bio- interleukin 6 by human and mutine recombinant interleukin I inphys Res Cornrun 155:1459-1463 mice. Eur J Immunoil 20:691-694

29. Whitnall MH, Perlstein RS, Mougey EH, Neta R 1992 Effects of 49. McIntyre KW, Stepan Ci, Kolinsky KD, Benjamin WRt, Plocinskiinterleukin- I on the stress-responsive and -nonresponsive subtypes JM, Kaffka KL, Canipen CA, Chizzonite RA, Kilian PL 1991of corticotropin- releasing hormone neurosecretory axons. Endocri- Inhibition of interleukin 1 (IL-I) binding and bioactivity in vitro andnology 131:37-44 modulation of acute inflammation in vivo by ILAI receptor antago-

30. Dunn AJ 1990 lnterleukin-1 as a stimulator of hormone secretion. nist and anti-IL-i receptor monoclonal antibody. I Exp MedProg Neuroiendlocrin Immunol 3:26-34 173:931-939

31. Besedovsky HO, Del Rey A, Sorkin E, Dinarello CA 1986 Im- 50. Mengozzi M, Bertini R, Sironi M, Ghezzi P 1991 Inhibition bymunoregulatory feedback between interleukin- I and glucocorticoid interleukin 1 receptor antagonist of in vivo activities of interleukinhormones. Science 233:652-654 1 in mice. Lymphokine Cytokine Res 10:405-407

32. Bernardini R, Kantillaris TC, Calogero AE, Johnson EO, Gomez 51. Sharp BM, Matta SG, Peterson PK, Newton R, Chao C, McAllenET, Gold PW, Chrousois CP 1990 Interactions between tumor K 1989 Tumor necrosis factor-alpha is a potent ACTH- secretagogue:necrosis factor-alpha, hypothalamic corticotropin-releasing hor- comparison to interlieukin-1 beta. Endocrinology 124:3131-3133mone, and adrenocorticotropin secretion in the rat. Endocrinology 52. Denveniste EN, Sparacio SM, Norris JG, Grenett HE, Fuller CM126:2876-2881 1990 Induction and regulation of interleukin-6 gene expression in

33. Lyson K, McCann SM 1992 Involvement of arachidonic acid cas- rat astrocytes. J Neuroimmunol 30:201-2 12cade pathways in interleukin-6-stimulated corticotropin -releasing 53. Gershenwaid JE, Fang YM, Fahey TJ, Calvano SE, Chizzonite R,factor release in vitro, Neuroendocrmnology 55:708-713 Kilian PL, Lowry SF, Moldawer LL 1990 Interleukin 1 receptor

34. Navarra P, Pozzoli C, Brunetti L, Ragazzoni E, Besser M, Gross- blockade attenuates the host inflammatory rrsponse. Proc NatI Acadman A 1992 lnterleukin-i# and interleukin-6 specifically increase Sci USA 87:4966-4970

952 CYTOKINES AND ACTH RESPONSE TO LPS Endo. 1993Vol 132. No 3

54. Kotloff RM, Little J, Elias JA 1990 Human alveolar macrophage saccharide (LPS) stimulated IL-6 secretion by rat anterior pituitaryand blood monocyte interleukin-6 production. Am J Respir Cell Mol cells. 73rd Annual Meeting of The Endocnne Society, WashingtonBiol 3:497-505 DC, 1991, p 150 (Abstract 479)

55. Jirik FR, Podor TJ, Hirano T, Kishimoto T, Loskutoff DJ, Carson 57. Fischer E, Marano MA, Van Zee K), Rock CS, Hawes AS, Thomp-DA, Lotz M 1989 Bacterial lipopolysaccharide and inflammatory son WA, DeForge L, Kenney JS, Remick DG, Bloedow DC,mediators augment IL-6 secretion by human endothelial cells. J Thompson RC, Lowry SF, Moldawer LL 1992 lnterleukin-1 recep-Immunol 142:144-147 tor blockade improves survival and hemodynamic performance in

56. Romero LI, Lechan RM, Clark BD, Dinarello CA, Reichlin S, IL- Escherichia coli septic shock, but fails to alter host responses to1 receptor antagonist inhibits hRL- I beta but not bacterial lipopoly- sublethal endotoxemia, J Clin Invest 89:1551-1557

U Biology of nitric oxideAR5ED FORCES RADIOsIOLOGY

RESEAKH INSTITUTEIn: The Biology of Nitric Oxide. SCITIC 11101T

S. Moncada, M.A. Marietta, J.B. SR93-9Hibbs, Jr., and E.A. Higgs, eds.Portland Press, London, 1992.

Electron paramagnetic resonance detection of nitricoxide-dependent spin adducts in mouse jejunumL Steel-Goodwin', C. M. Arroyo', B. Gray3 and A. J. Carmichael1

'Radiation Biophysics and 2Radiation Biochemistry Departments. ArmedForces Radiobiology Research Institute, Bethesda. Maryland 20889-5145,U.S.A.

Abstract

The electron paramagnetic resonance (EPR) spectra of aqueous solutionscontaining nitric oxide (NO) and the spin trap 3,5-dibromo-4-nitrosobenzene-sulphonate (DBNBS) have previously been described [1). Similar NO-derivedDBNBS adducts plus the DBNBS oxidation product were observed in 50 mmphosphate-buffered incubation media (pH 7.0) upon addition of NO. Moreover,the above adducts plus additional DBNBS radicals were observed in incubationmedia following suspension of mouse jejunal slices (- I cm length) for 20 min at37 TC when NO was added. Less intense DBNBS spin adduct spectra wereobserved in analogously treated jejunum slices that had no NO added. This resultsuggests a continuous production of NO by jejunum which may play a role inperistalsis. Moreover, the basal level of NO production was stimulated in thepresence of the radioprotectant, N-(2-mercaptoethyl)-1,3-diaminopropane fWR-1065).

Introduction

Endothelium-derived relaxing factor (EDRF) has been identified as NO orcompounds derived from this labile gas [2]. There have been numerous attempts toisolate this ephemeral gas from a variety of cells and tissues. One approach employsspin traps which form free radical adducts with half-lives sufficiently long to permitEPR detection [1]. Papers reporting the presence of NO-derived radical adductsproduced by cell systems have appeared recently [3, 41. Therefore, it is reasonableto expect that NO-derived radical adducts may be trapped and detected by EPR intissue slices using spin traps. The gut is an appropriate source of tissue slices whichmay be expected to elaborate NO. EDRF has been demonstrated in the ileocolonicjunction of dogs 151 and the guinea pig stomach 161. Also, the NO synthase enzymehas been immunologically detected in myenteric plexi throughout the gut of the rat[71. Finally, the peristaltic, rhythmic contractions and relaxations occurring in thegut could reasonably be expected to result in part from the effect of NO on gutsmooth muscle. This paper reports results of spin trapping experiments carried outwith gut tissue slices which elaborate NO or free radicals derived from NO.

Enzymology and biochemistry I1

Materials and methods

Mice were euthanized by cervical dislocation, the small intestine was removed andflushed with 50 mm sodium phosphate-buffered (pH 7.0) 0.9 % sodium chloridesolution (PBS) and a slice of jejunum (- 1 cm) was placed in incubation medium.The incubation medium (pH 7.0 ± 0.1) was PBS supplemented with I I mm-glucose, 2 mu-KCI, 26 mu-NaHCO3 , 1.18 mm-KHPO, and 2 mu-CaCl,.2HO18). N-(2-mercaptoethyl)-1,3-diaminopropane [WR-1065, HsN-(CH,),-NH-(CHO),-SHI was hydrated in incubation medium immediately before addition tosolutions. The spin trap DBNBS and bovine copper/zinc superoxide dismutase(SOD) were weighed and added to test tubes. These materials were hydrated withincubation medium within one hour of their experimental use. NO was deliveredfrom a compressed gas cylinder. Incubations having 10 mm-DBNBS and 300 unitsmli' SOD were carried out in polyethylene test tubes. Spectra were made withfluids preincubated for 20 min at 37 °C before transfer to a quartz flat cell which wasthen installed on a Varian E-109 X-band spectrometer. The EPR instrument wasoperated at a magnetic field set at 338.8 mT, microwave frequency about9.500 GHz, microwave power 20 mW, receiver gain 2 x 10', modulation frequency100 KHz, modulation amplitude 2 mT, time constant 4 s, scan time 16 min, scanrange 0.1 T and temperature 25 *C. Hyperfine values were determined by directmeasurement from the spectra.

Results and discussion

Figure 1 shows EPR spectra of DBNBS radical adducts observed followingexposure of materials to NO gas. Figure Ia, the incubation medium, has DBNBSadducts (open arrows) with a hyperfine coupling constant, a, = 0.959 mT, whichcorresponds to the previously reported value for aqueous solutions exposed to NOgas [1]. In addition, the oxidation product DBNBS also appears to be present witha coupling constant, a, = 1.25 mT [9]. Figure lb is a spectrum made withincubation medium after 20 min at 37 *C without NO bubbling. This spectrum isrepresentative of random noise. Figure lc is the spectrum resulting from a jejunumtissue slice incubated for 20 min at 37 *C and then gently bubbled for 5-10 s withNO. This spectrum has both NO-derived adducts (open arrows) and the oxidationproduct DBNBS (solid arrows) seen in Figure Ia. In addition, there appear to beless intense EPR shoulders in the spectrum that are not seen in the incubationmedium alone.

Figure 2 shows EPR spectra resulting when incubation mediumcontaining jejunum slices was treated with various compounds. Figure 2a was madeafter 0.1 ml of NO-saturated, anaerobic water was added to a jejunum slicefollowing a 20 min incubation at 37 *C. In addition to DBNBS oxidation product(solid arrows) plus No-derived adducts (open arrows) there are two maxima whichcorrespond to shoulders seen in Figure lr. The quantity of NO introduced and themethod of its delivery may have subtle influences on the spectral results. Figure 2bshows the spectrum resulting from incubation of a jegunum slice for 20 min at3"7 'C. The analogous control lacking tissue (Figure lb) indicates that unstimulatedmouse iciunum slices do elaborate No. Figure 2c is a spectrum made with jejunumincubated with 5 m.i-W'R-1065 for 20 min at 37 'C. The DBNBS oxidation productplus the NO-derived maxima are present. WR-1065 appears to stimulate NOproduction in Ieunum slices because these maxima are more intense than thoseobserved with unstimulated slices (Figure 2h). It is important to note that the

Ii Biology of nitric oxide

Figure 1 (a) 4

(b)

(C)1.0 mT

EPR spectra of DBNBS radical adducts following exposure of materialsto NO gas(a) Incubation medium suspended for 20 min at 37 0C and bubbled with NO gas. (b)Incubation medium suspended for 20 min at 37 0C without NO bubbling. (c) Jejunumslice incubated for 20 min at 37 *C and bubbled with NO gas. Filled arrows indicatethe DBNBS oxidation product and open arrows indicate the NO-derived DBNBS spinadduct.

control spectrum (Figure 2d) made with incubation medium and WR-1065 doeshave EPR-detectable radicals present. The spectrum is not random noise (Figurelb). However, the pattern does not correspond to any of the other spectracontaining gut tissue slices. Minor spectral details seen in Figure 2c could resultfrom the contribution of W1R-1065 as well as the jetunal slices. It is possible thatWR-1065 reacts with DBNBS to produce the spectrum observed in Figure 2d.

These results support tne conclusions that mouse jejunum slicescontinuously produce a low level of NO. Additionally, WR,1065 stimulates thebasal level of NO production by jeunal slices. It is possible that this production ofNO plays an important role in the rhythmic relaxations of peristalsis.

Spported by the Armed Forces Radiobiolog-, Research Institute, Defense Nuclear Agena.I 'Jews presented in this paper are those of the azthors; no endorsement by the Defense NsclearAgexgf has been gmen or should be inferred. Research was condmcted according to the principles

Enzymology and biochemistry 83

(a) Figure 2

(b)

(d)

1.0 mT

EPR spectra following treatment of Incubation medium containingjejunum slices with various compounds(a) Jejunum slice incubated for 20 min at 37 °C to which 0. 1 ml of NO-saturatedanaerobic water was added. (b) Jejunum slice incubated for 20 min at 37 *C. (c)jejunum slice incubated for 20 min at 37 0C with 5 mt-WR-1065. fd) Controlincubation medium plus 5 mm-WR- 1065 incubated for 20 min at 37 *C. The arrows

indicate the same products as in Figure 1.

enunriated in tr * Guide jor the Care and Use of Laboratory Animals' prepared by theInsttute of Laboratory A-tnimal Resources. National Research Council.

ReferemesI Amrfi,,. ( M & K M.. N. (1990) Htev Radical Rce. (.immun 14, 141 IBS2 M ,,ncada. , Palrm r. R \1 I & FIogg%, l A (1991) PharnAcoI. Re'. 43, 109 142

, Airr• N \1 & ,,rraw. ( (19911 I.ur I Pharmacil M00. 15- 161

4 PreinaI. l. chhm ,rl. k , ',tinki. II . .akzwa ... II . Okin -o. II . (-armtchacl. A. J. & Arruvii. G K1.(161) -.ut ) I,,whcm 202. 9ý21 91,

S Suit,,11 I.cckxacn,. ( I.-. Pcckmrna . P A . J,)rdaen.. 1:. I v, ian Macckc, YN M. & Herman,

A (D vaI. ) %aturI . .,mdn .) 34nS. 34( ) a3(I Dk$al, K,. M. ,CI:SA.\\ (. & \ anc;. I R. (1991) N•atu.re (LA)ndon) 351, 4"77-4"79

84 Biology of nitric oxide

7. Bredt. D. S.. Hwang. P. M. & Snyder, S. H. (j1990) Nature (L'ndon) 347, 768-7708. Ganhwaite, J., Garthwaite, G., Palmer, R. M. J+ & Moncada, S. (1989) Eur. ). Phanmacol. 171.

413-4169. Nazhat, N. B., Yang, G., Allen, R. E.. Blake, D. R. & Jones. P. (1990) Biochern Biopvs. Res.

Commun. 166. 807-812

RADIATION RESEARCH 133, 12-19 (1993) TARME FOPCES RAD1OOIOVOGYRESEARCH INSTITUTESCIENTIFIC REPORT

I SR93-10

Radiolysis in Aqueous Solution of Dinucleoside Monophosphatesby High-Energy Electrons and Fission Neutrons

Y. N. VAISHNAV AND C. E. SWENBERG

Radiation Biochemistry Department. Armed Forces Radiobiology Research Institute, Be.hesda, Maryland 20889-5145

ofd-[TpT] and X-irradiated "purine-pyrimidi ne" dinucleo-VAISHNAV, Y. N., AND SWENBERG, C. E. Radiolysis in tide sequence isomers were reported in the pioneering re-

Aqueous Solution of Dinucleoside Monophosphates by High- search of Cadet, Box, and co-workers (6-11).Energy Electrons and Fission Neutrons. Radiat. Res. 133, 12- Our present study examines effects of different qualities19 (1993). and quantities of radiation on dinucleoside monophos-

The radiation chemistry in aqueous solution of the dinucleo- phates of "nyrimidine-pyrimidine" systems,' which hope-side monophosphate d-[CpT] and its sequence isomer d-ITpCI fully will improve our perspective on the radiation chemis-in air or nitrogen was examined using different qualities and try of native DNA. To compare and correlate the effects ofquantities of radiations. High-performance liquid chromatogra- different radiations with different values of linear energyphy and gas chromatography-mass spectrometry were used to transfer (LET), aqueous solutions of 2'-deoxycytidylyl-(3'-analyze the high-energy electron (13.2 MeV) exposure products 5')-thymidine (d-[CpT]) and its sequence isomer thymi-or fission-neutron exposure products of d-ICpTI and d-ITpCI. Acomparson of product profiles obtained from irradiated d- dylyl-(Yr-5)-2'-deoxycytidine(d-[TpCo ) (see Fig. i forchem-ICpTI and d-1TpCj suggests that, at relatively low radiation ical structures)were exposed to different doses of high-en-luzes (50-250 Gy), products are formed by N-glycosidic or ergy electrons (13.2 MeV, low LET) or fission-neutrons

phosphodiester bond-cleavage, while at higher doses (500-1000 (high LET), and the product profiles were analyzed by high-Gy) additional products were detected as a consequence of ring- performance liquid chromatography (HPLC). Individualt,-o'ification mechanisms. The plots of radiation dose-yield and products were isolated ard characterized by gas chromatog-".or:esponding calculated G values of the released undamaged raphy/mass spectrometry (GC/MS) and their G values werebases and nucleosides from d-ICpT] and d-ITpCI suggest a base- ( etermined.sequence dependence and a quality- and quantity-dependent re-sponse to ionizing radiation. Although the product quantitiesformed from sequence isomers were slightly different, we found MATERIALq AND METHODSn, qualitative differences in the product formed at the lowestdoses examined C 1993 Academic Pre•s, Inc. The dinudeoside monophosphatesd-[CpT] and d-[TpC] and samples of

cytosine, thymine, 2'-deoxycytidine, thymidine, thy midine-3'-phosphate,thymidine-5'-phosphate, 2'-deoxycytidine-3Y-phosphate, and 2'-deoxycyti-dine-5'-phosphate were purchased from Sigma Chemical Co. (St. Louis,

INTRODUCTION MO); HPLC-grade acetonitrile was obtained from Aldrich Chemical Co.(Milwaukee, WI). N,O-bis(Trimethylsilyl)trifluoroacetamide (BSTFA)was purchased from Supelco, Inc. (Bellefonte, PA). All the HPLC analyses

Cytotoxic, carcinogenic, and mutagenic effects in cells were performed using a Kratos Analytical Spectroflow 400 solvent deliveryexposed to ionizing radiation are thought to be primarily system (Ramsey, NJ) on-line with a N el SP 4100 computing integrator

the result of damage to DNA (1-3). Recent interest in radio- and an Applied Biosystems (Ramsey, NJ) Modd 783 absorbance detectorgradient controller. Mass spectral analysis was performed using the Kratos

therapy with heavy particles and radioprotection against Analytical 25RFA mass spectrometer systems (Manchester. UK). Samplessuch particles encourages the search for the molecular basis we'e alialy7ed by a direct insertion probe at 70 eV in an electron impactof their action (4). Understanding the fundamental bio- mode or were converted into their corresponding trimethylsi!yl (TMS)chemical pathways involved in radiation ;ensitivity in rela- ethers and injected into the gas chromatography column and analyzed by

tion to radiation damage to the primary target molecule, mass spectrometry (14). The TMS derivatives were prepared by dissolving50 ,g of each HPLC-purified material in 25 j.I of BSTFA. Samples were

presumably DNA, in cells requires identification of the sen- allowed to react and equilibrate at room temperature for 18 h before analy-sitive and reactive sites of DNA. The identification of such sis. A Carlo Erba (Strada Rivoltana, Italy) high-resolution gas chromato-sensitive and reactive sites in cells is exceedingly difficult, ascellular systems are highly complicated (5). One approach 'Y. N. Vaishnav and C. E. Swenberg, Radiolysis ofdinucleoside mono-is to study model systems. L'nucleotides are suitable, small, phosphate and its sequence isomer by high-energy electrons and fissiontwo-base models for inN,;stigating the effectsofionizing radi- neutrons. Abstract p32-18, 9th International Congress of Radiation Re-ation on single-strand DNA. Products of gamma radiolysis search, Toronto, Canada, 1991.

0033-7587/93 $5.00 12Ccpyright © 1993 by Aidcrnic Press, Inc,All rights of reproduction in any form reserved.

RAIDl0l YSIS (W1 DI )NI ( Ii )Sllf )l M0)\ WIt( SPI x I IS I11

H4% N.I4 %hith) ItI. iJaS .11d 1ilAg1)CfiUni IMed "soh 11Y011 Ti to JfI Tt c Uh k,11~i '01d

OH 2N HThe Ii PL( product p)rorik's- In aqueLIouIS~l~ lolut l (fill 7.0)N H generated fromn the high-energN electron and lission-nicu-

H ' . H0 45H trnirradiation of' di( p I anJ d-Yp(' i n air wecre re-0V~ corded. F-igure 2 is anl example of'an IIPl.C product profile

0- P0 Hobtained after 1 3.2-MeV electron Irradiation oft dI( I.0 OH O-P-0 [he individual Components. of' the Irradiation proiducts

114." OH from the product protlecs labeled a it) p are listed in I able 1.HNA %~CH H-.NH quimolar concentrations ol'd-[C'pTI and d-[lIp(] A scre ex-

SN ~ posed to total radiation doses of I) 50. 25(, 51X), or 1(0(0I -Lulvoue l-I aliquot) for irradiated and uin-

114 N, 0 irradiated samples were injected into fihe IPIC columni

H H 0 4under elutiun conditions identical to those: described underH H_ H Materials and Methods. Major components were isolated

H O H H 2 3 H H and identified bý comparing IIP[lC retention times %kith

H OH available standards and also bN isolating the indisý iduaf com-ponents. freeze-dryNing, and subwequentlN treating "fill

d-Eco-i d-tpclBSTFA. SilvI derivativ es so 6ormed were further anaiwzed

FIG. I. Chemical structures of din ucleoside monophosphates. b CM n/radrc neto rb/assetoetry (L)IP/MS) as descrihed under Materials and Methods.The observed HPLC retention times and prominent massspectral fragmentation patterns arc listed in Table 1.

graph %as interfaced wiih the mass spectrometer. (IC/MS analyses "ere To quantify radiation-induced release of ýhe free basesperformed using a fused silica capillary column (50 ), 0.32mm i.dA 'T'hetone() doxvuin dl.h-injection port and the ion source were monitored at 2700 C wiih ihe (ACl n ulesdscNISinterfaced at 290'C. Helium-was used as carrier gas at an inlet pressure mine (T), and thymidine WdT) (products labeled c. 17, h. andof' I1) kPa and the split mode was used for (MI/MS analysis. The mass m, respectively). equimolar mixtures of vary ing concentra-spectrometer was calibrated using perfluorokerosines for a mass range of18 -800 AM U.

Sample preparation and irradiation. In a closed polypropylene con-tainer, solution (I mmol dm 3) of either d-ICpTI or d-lTpC] was preparedin deionized water in the presence of either air or nitrogen. Samples (200MIl) were exposed to 0, 50, 250), 5f)0. or 1000 (Is of I 3-2-MeV electrionirradiation (3.2 Gy/pulse) or fission neutrons at '4'C. Irradiated samptes Vwere subspquently frozen in liquid nitrogen until HPL-C analysis. A re-----.

verse-phase analytical column (Spheris, 250 mm X 4 6 mm) w&as used in Csolvent gradient mode ofO. I mot dm 'ammonium acetate versus aeetoni-trite. The HPL-C product profiles of the irradiated samples and unirra- )-ei I r _ ivdiated controls were recorded by optical (detection at 254 rim (or in some W _

cases 220 nm), and the individual components were isolated from themixture product. Aliquots from individual component~s were reinjectedinto the HP-C column, using identical HPL( elution conditions. and theobserved retention times were compared with available standards. For fur- Jjg~)

ther characterization of the HI-PC-purified individual components. indi- ccvid!ual components were freeze-dried and treated with RsTrFA. and result-ing s~lyl ethers were characterized by (IC/MS. a

L)osArneirli Samples were exposed to either I 3.2-,MeV elecirons or fits- .~-.J

sion neutrons. For experiments, with high-energý electron irradiation, alinear accelerator (LINAC) was used with the average dose rate 3.2 (lv!pulse (4 -Mes pulse). Dosimetry measurements were performed b'. measuring 0 5 10 15 20 25 30the dose to individual pulses using lithium fluoride thermolumineseni do- Retention time (min)simeters (ri.DI. Trhe tiotal dose was determined by the product olithe num-ber of pulses and the dlose per pulse. The AFRRI _TRI6iA Mark-F reactor FIG. 2. I Iill[.( prolduct proliles; of high-tenerg% clectron 4(;3 2 WN iwas used to) obtain a mixed neutron to -(-ray field with a total dose rate of irradiated d-[(p I I in aqueous oL' and in aur I ti) V rcler to radiationabout SO (,v minm and a ratio) of neutron to -y-ray kerma in free air of1 doses: coinirol. 50, 2,0, 500. and IOM4 (\ .ý respeetive Ib ric allor irradia.about I5- 18. Dosimetry measurements wcý.e done using 0.5 cm 'ioniza- tion prolduct cni(iiienisC., labeled a through p., %%ere isolated. quitinfichd.tion chambecrs constructed of A-IS tisljsue-equivalent 4 [II plastic filled and characteriz-d b'. I IPI ( and ( iC( I and are listid in t able I

14 VAIStHNAV AND SWENB-FR(

TABLE IIsolation and ('haracterization of the Products from Irradiation of d-ICp'i'l and d-ITpCI

HPI.C Retention Prominent MS peakspeak time (min) Compound tm/z, percentage intensity)

a 4.5 2'.Deoxyc.tidine-3'-phosphate 595 M0,O. 505 (0.31,449 (0.3), 447 (10.3). 331 (1t0). 313 (15).299 (60), 285 (30J0)

b 4.7 2'-Dtoxctidine-5'-phosphate 667 (0). 652 (0. 1). 595 (0.!). 429 (0.2). 341 (0.4), 313 to 3).299 (0.2). 255 (0.4), 241 (0.6)

c 51 (Cytosine 255 (55). 254 (90t) 240) (100). 183 (20). 168 (45). 147 (60).125 (10), 98 (28)

d 5.5 Thy midine-Y-phosphate 538 (0.1), 523 (0.2). 429 (0.1), 369 to 1), 341 (30), 299 125),147 (1(0), 120 (20)

c 6,0 Ihymidine-5'-phosphate 538 (0.4), 523 (0,4), 241 (50). 211 (1it. 81 130)f 6.5 2'-Deoxscytidine 433 (3.0), 428 (2.0). 240 (5 0). 183 (10), 184 (10). 170 (50).

103( 20), 59) 15)g 7.0 5,6-Dihydro-5.6-dihydroxyth,,midine 448 (25), 433 1 !Il. 318 (50. 30) (10), 259 ( I00), 203 (5.0,

174(5) 13;

h 8.1 Thymine 270 (100). 25- ý. 170 (10), 147 (30). 131 (5). 113 (30). 81(5). 59 (13)

8.4 5,6-Dihvdro-5.6-dihydroxyuraciI 434 (5,0). 419 ((0), 362 (30), 347 (18). 331 170). 245 (40),130(10). 73 (100(

8.8 5,6-Dihydroxy-uracil 432 (15). 417 (30). 343 (5.0), 147 (60). 73 (100)9.3

k 10.2I 12.0m 13.A Thymidine 458 11). 443 (2t. 270 (5), 255 (3). 183 t10), 170 (35). 147

(30). 129 (10) 81 (25). 59 (10)n 14.4 2'-Deoxycytidylyl-(3'-5')-5,6-dihydro-5,5-

dihydroxythymidineo 15.2 b ,

p 19.2 5.6-Dihydro-5,6-dihydroxythmidvlyl-(3'-5')-2'-deoxycytidine

HPLC-purified samples were transformed into their TMS derivatives: the TMS derivatives were subsequently analyzed by GC/MS and/or DIP/MS."Not identified.HPLC-purified samples were hydrolyzed by acid. and hydrolysates were treated with BSTFA. Subsequently the TMS preparations were analyzed by

GC/MS. and corresponding base moieties were characterized.

tions of authentic samples of C, dC, T, and dT in water were irradiated d-[TpC] (Fig. 3B-l). Product p from neutron-irra-eluted through a reverse-phase analytical HPLC column. diated d-[TpC] (Fig. 3B-1) is not detectable in high-energyNo differences were observed in the retention times or peak electron-irradiated d-[TpC] (Fig. 3B-1I). Product o fromareas whether the mixtures of C, dC, T. and dT or the indi- high-energy electron-irradiated d-[TpC] (Fig. 3B-11) wasvidual bases or nucleosides were injected into the HPLC completely absent in neutron-irradiated d-[TpC] (Fig.column. Calibration curves were determined by integration 3B-1).of peak areas as a function of concentration. These calibra- Figure 4 provides an example illustrating the effects oftion curves were used for quantification of radiation-in- different gaseous environments on product yields. Whenduced release of undamaged base and nucleoside moieties d-JCpT] and/or d-[TpC] is irradiated using fission neutronsfrom irradiated d-[CpT] and d-[TpC] in aqueous solution (500 Gy) in the presence of either air or nitrogen in deion-as a function of dose. ized water (pH 7.0), no significant qualitative differences

Figure 3 is a representative comparison of products gen- are observed in the product profiles. However, severalcrated in aqueous solution from high-energy electron or quantitative differences are evident: for example, the prod-fission-neutron irradiation (500 Gy) of d-[CpT] or d-[TpC] uct labeled m resulting from fission-neutron-irradiated d-under aerobic conditions. The majority of the products [CpTJ in air (Fig. 4A-1) versus nitrogen (Fig. 4A-ll) andformed are qualitatively similar for the two LETs investi- product label, J p from fission-neutron-irradiated d-IlpC]gated, although the quantities of the products formed differ in air (Fig. 4B-I) versus nitrogen (Fig. 4B-11) arc attributedslightly- there are, however, several significant differences. to oxygen-dependent reactions.Specifically, the products labeled i and k are not detectable To compare and correlate quantitatively the efficienciesin neutron-irradiated d-[CpTJ (Fig. 3A-1) or in neutron- of the postirradiation release of free bases and free nucleo-

RADIOLYSIS OF DINUCLEOSIDE MONOPHOSPHATES 15

A B

O •I' :--i

(50 ec

Sg I I m b h

•k I m /

I k Im

; ¢o ..... /5 2'o 2's 30 o ; t'o ¢5 2'o 2's 30

Retention time (rain)

FIG. 3. HPLC profiles. Fission-neutron versus high-energy electron (500 Gy each) irradiation of d-[CpT1 and d-[Tl•'] in air: (A) I, neutron-irra-diated d-[CpT]; (A) 11, high-energy electron-irradiated d-[CpT]; (B) I, neutron-irradiated d-[TpC]; and (B) !!, high-energy electron-irradialed d-[TpCI.Representative examples of HPLC product profiles were obtained after irradiation of d-{Cp/l or d-(TpC1 in aqueous solution.

sides, aqueous solutions (I mmoi dm-3, pH 7.0) were ex- lary column GC/MS and DIP/MS properties under our ex-posed to either fission-neutron or high-energy electron radi- perimental protocol. However, several of the TMS prepara-ation in the presence of air or nitrogen, and the yields in tions ofnucleotides underwent thermal degradation at leastumol dm-3 were calculated at radiation doses of 50, I00, to some extent: hence DIP/MS was used to characterize theand 250 Gy according to the procedures of Roots and co- TMS derivatives ofnucleotides. Electron impact mass spec-workers (14). The average G values (the number of mole- tra and fragmentation patterns of TMS derivatives of nu-cules of product formed per I00 eV energy absorbed) and cleic acid components have been extensively studied andcorresponding SEMs are listed in Table II. reported by McCloskey and co-workers (15. 16), and mass

Structural elucidation of the irradiation products from spectral data of both modified bases (g and i) have beend-[CpT] and d-[TpC], labeled a to p in Fig. 5, was per- discussed by Dizdaroglu (17. 18). Our structural assign-formed using data obtained from GC/MS and/or DIP/MS ments for the TMS derivatives of products labeled a to p areafter treating purified individual components with BSTFA. consistent with those already reported (15-18). In charac-Prominent mass spectral peaks (m/z) and corresponding terizing modified intact d-[CpT} and d-[TpC], HPLC-puri-relative intensities are listed in Table II. The TMS deriva- fled products peak-labeled n and p were hydrolyzed using 6rives of the radiation-induced intact free bases C and T and N HCOOH and hydrolysates were subsequently dried andnucleosides dC and dT (c, h, f, and m, respectively, in Table treated with BSTFA according to procedures described byI) and the modified bases thymine glycol and uracil glycol Dizdaroglu (17). The presence of thymine glycol and intact(g and i, respectively, in Table I) exhibited excellent capil- cytosine was revealed by GC!MS assay of the TMS deriva-

A B

m pT• •, f{dl€" c nS, i i i ' t I

t•

._.> d-•!So• . -gT] , i"• b , in-. " " r " i i

Slb !'s 2'o d5 3o • •'o •'• •o is 3'0

Retention time (rain)

FIG. 4. HPLC profiles. Fission-neutron irradiation of d-jC'pT] and d-{Tl•'] in air versus nitrogen (radiation dostr: 5(X) (}•, eachl: {/'.) I. •rradtatedd-{CpT1 in air: (A) II, irradiated d-(CpT] in nitrogen; {BI I, irradiated d-{TpCl in air; and (B) II, irradiated d-lTP("I in nitrogen, Rcpresentati• c examples {•tHPL£" product profiles were obtained after irradiation of d-[CpT] or d-ITpC] in aqueous solution.

16 VAISHNAV AND SWENBERG

TABLE II

G Values for Po-tirradiation Release of Undamaged Nucleobases and Nucleosides from d-ICpT] and d-ITpCI

G values t SEM"

Dinucleotide Released product H-E (e-) in air Neutrons in air Neutrons in nitrogen

d-[CpTI Cytosine 0.008 ± 0.001 0.019 t 0.002 0.009 1 0.001Thymine 0.008 ± 0.001 0.024 t 0.005 0.014 .- 01X)02-d-Cytidine 0.025 ± 0.002 0.028 ± 0.003 0,008 ± 0,001Thymidine 0.008 ± 0.001 0.015 ± 0.002 0X005 ± 0.001

d-[TpC] Cytosine 0,013 ± 0.001 0.010 ± 0.002 0.007 - 0001Thynine 0.014 ± 0.002 0.009 t 0.001 0.007 _ 0.0)02-d-Cytidine 0.015 ± 0.004 0.019 ± 0.001 0.004 0.001Thymidine 0.005 ± 0.002 0.007 t 0.001 0.010 t 0.001

'The release of intact nucleobases and nucleosides is measured after irradiation of 0. 50, 100, 250. 500, and 1000 Gy; the G values for the free bases andnucleosides were calculated after irradiation with 50, 100, and 250 Gy and are means of at least three independent experiments. The G values are definedas the number of molecules of product formed per 100 eV energy absorbed.

tives in both samples, confirming the chemical structures of terized. Comparisons of the product profiles generatedn and p as modified d-[CpT] and d-[TpC], respectively, and from the fission-neutron or high-energy electron irradiationthat the modification had taken place at 5,6-positions of the of d-[CpT] and d-[TpC] suggest that independent of thethymine moieties of d-[CpT] and d-[TpC]. quality of the radiation, at relatively low radiation doses

(50-250 Gy), most of the detectable products originatedDISCUSSION from N-glycosidic or phosphodiester bond-cleavage.

whereas at higher doses (500-1000 Gy) additional productsTo obtain an improved perspective ofthe radiation chem- appear to be formed by ring modification. For example,

istry of native DNA and to determine whether nucleoside product n in Fig. 2 and product p in Fig. 3B-I are ring-mod-sequence is important in determining the types and amount ified products detectable at doses 500 Gy or greater- prod-of damage caused by radiations of different LET, we se- ucts m and f(Figs. 2, 3, and 4) are present at all radiationlected a "pyrimidine-pyrimidine" system consisting of the doses examined independent of the sequence isomer. At thedinucleoside monophosphate d-[CpT] and its sequence lower doses, we found no qualitative differences in productisomer d-[TpC]. Aqueous solutions of d-[CpT] or d-[TpCI formation although the quantities of the product formedwere exposed to either fission neutrons or high-energy elec- were different; however, product p, intact modified d-trons, and products were analyzed. The chemical nature of [TpC], in Fig. 3B-I was not apparent in Fig. 3B-ll, therebythe postirradiation release of free and modified bases, free signifying a radiation quality-dependent response to ioniz-nucleosides, and free and modified nucleotides were identi- ing radiation. Teoule and Cadet (7) had previously investi-fled and quantified by standard I IPLC and by GC/MS or gated radiation damage to DNA monomer. Based on theDIP/MS. The HPLC and GC/MS methodologies for the studies of Teoule and Cadet and Box and co-workers (6-11)isolation, purification, and structural elucidation ofthe radi- of dinucleoside monophosphates, certain expected prod-ation-induced products were selected since both the meth- ucts were not observed in our studies. For example, in theodologies are quite specific and sensitive, having detection case of dinucleotides containing thymidine, the principalcapability of approximately I-5 pmol. In most instances known modification is that in which the base moiety de-these procedures allow assaying radiation damage in small grades to formamide. It is possible that the yields are smallDNA oligomers without the necessity of further degrada- and that these expected products are among the severaltion by acidic conditions. Acid hydrolysis is known to lead minor products not identified in our investigation. Mecha-to artifactual lesions (19). For the structural characteriza- nistic rationales for product formation have been describedtions, we first isolated individual components from the radi- and discussed elsewhere (5, 20).ation product mixtures by HPLC, freeze-dried the isolated High-performance liquid chromatography in conjunc-individual components, and subsequently treated them tion with proton magnetic resonance (PMR) spectrometrywith BSTFA. The resulting silyl ether derivatives were ana- is an excellent analytical technology for the direct determi-lyzed by GC/MS. To minimize the possibilities of thermal nation of absolute stereochemistry including anantiomericdegradation, several TMS ethers of nucleotides isolated forms (1/, 12) of organic compounds, for example, glycolsfrom the product profiles were further characterized by of thymine and uracil as well as 5.6-dihydro-5,6-dihydroxyDIP/MS. The nucleotide dimers were exposed to varying modification of d-(CpT] or d-[TpC] obtained in this inves-doses of radiation, and the detectable products were charac- tigation. However, at least 100 jtg of purified glycol is re-

RADIOLYSIS OF DINUCLEOSIDE MONOPFtOSPHATES 17

H, H H, H H, N H H

N N'114

4-1 3. 154-, N H.N' I CH 3

N3 OI N*O€ N

dR-3' -P dR-5'-P H IdR-3" -Pa b c d

0 0"4O H H a CH 3 _C H 3"t4 C3 N N H

H N 1>t413 -0 HO( QN N5 CH3 1 2 N H

t'4HN" H°J HN"'

6 H OH HI H

dR-5"-P dR

e t ghO HHN,,,'IOH

O-J-i,,_N 0 o

"114 OH 114H H'N CH3 H.N 5CH 3HI !-OH C

N OH OH O2uO O0 4 5" H 0 4 51

H H1 H HH H H23-H H H 2. 3 . HH

OH- PH .H 0-4' 5'0

H H H HA11 OH m OHINH

H HH 2 S H90-- 0 50-ý

S. -0 (1o6H O-P•4O" OH H j-N OH " OH

14 0H H.

,HH H H 0--1HpH

H O 3 H HI H 0 H* 4.4 H j

H OHn

FIG. 5. Chemical structures of the identified irradiation products from d-[CpT] and d-[TpC]. Labels a to p correspond to those in Table 1.

quired for the acceptable PMR spectra for full characteriza- The TMS derivatives of hydrolysates of products n and p,tion. Our experimental procedure yielded only nanomolar modified dinucleoside monophosphates, also revealed thequantities of purified materials which was sufficient for presence of thymine glycol in cis configuration in both ofGC/MS analysis but not for detailed direct configurational the glycol moeities upon GC/MS analysis and comparisonassignments by PMR. The stereochemistry of glycols of with the reference mass spectra. The trans isomer was notthymine and cytosine (uracil glycol-deamination product detected in either of the cases. It is possible that the yields ofof cytosine) observed in the present studies, both had cis trans isomer of thymine and cytosine were low and thatconfigurations with respect to 5,6-hydroxy groups. These they were also among the several minor products not iden-configurations were verified indirectly by comparison of tified in our investigation.mass spectral peak intensity ratios, with the corresponding A recent report by Fox and McNally (22) suggest thatreference cis-oriented TMS derivative of glycol mass spec- fission neutrons produce fewer single-strand breaks but ap-tra. The reference massspectra forcis-and trans-TMSderiv- proximately the samc number of double-strand breaks inatives of thymine glycol were generously provided by Dr. Chinese hamsterV79 cells after exposure to equal doses ofSwartz of Wake Forest University (Winston-Salem, NC) X or -y rays. This suggests there is a difference in the detailsand mass spectra of cis-uracil glycol have been reported by of the molecular types of damage produced by differentDizdaroglu and co-workers (21). qualities of radiations (23). For an end point such as cell

18 VAISHNAV AND SWENBERG

killing, neutrons are approximately 1.5 to 6 times more intact bases released might correlate with the extent ofeffective than either X or -y rays (24). In the cells, radiation DNA strand breaks, we compared the yields of release ofdamage is generally considered to be brought about by a intact nucleobases (C and T) and nucleosides (dC and dT)combination of indirect and direct effects as described by from d-[CpT] and d-[TpC] after radiation exposure. The GRoots and co-workers (25). The indirect effect consists of values calculated according to the equation of Roots andthe interactions of target molecule (DNA) with the prod- co-workers (14) are given in Table II. The G values foructs of water radiolysis, hydroxyl radical (OH), hydrogen released intact C, dC, T, and dT after fission-neutron irra-atom (H) and solvated electrons (eq). The direct effect con- diation in air from d-[CpT] are significantly greater thansists of excitation and ionization of target molecules by the the G values of high-energy electron-irradiated d-[CpT] inparticle and its secondaries. In low-LET radiation, DNA air. On the other hand, fission-neutron-irradiated d-[T pC]damage occurs primarily via indirect effect. In the case of in air showed a somewhat different trend: only G values forhigh-LET radiation, for example, fission-neutron damage the release of C and T were slightly higher when d-[TpC]occurs primarily via high-energy recoil protons (26). Direct was exposed to high-energy electron irradiation (in air).and indirect action are not completely distinct because ra- Our G values do not include the effect of radiation-induceddiolysis produts produced by indirect action can react with labile bonds formed since we did not determine the G val-radicals formed in the target molecules by either indirect or ues after incubating irradiated samples at 37°C for a spe-direct action. In the present studies, several quantitative cific time (i = 3-6 h) (14, 27). The initial G values obtaineddifferences were observed in the product profiles obtained in this study clearly suggest a dependence on base sequencefrom high- and low-LET irradiated (equal doses) d-[CpT] and radiation quality for the amount of radiation damage.and d-[TpC] may have arisen due to some combinations of The yields of primary radicals in aqueous solutions ex-direct and indirect effects. As expected, these differences posed to ionizing radiation can be altered by saturating thewere not so large since irradiation were performed on rela- solution with different gases prior to irradiation (31). Sev-

tively small sized target molecules, and in dilute aqueous eral radiobiological end points (for example, enhancementsolution. of radiosensitivity of cells in saturated air compared to ni-

During the course of our studies, we observed generation trogen-saturated conditions) can be altered by such modifi-of free bases, nucleosides and mononucleotides along with cations (32). At the molecular level, these yield differencesfree glycols of bases, and intact 5,6-dihydro-5.6-dihydroxy- in DNA base products are known to be a function of oxygenthymine modification of d-[CpT] and d-[TpC] prior to acid concentration (33). We examined the effects of fission-neu-hydrolysis. The chemical nature of the products detected tron irradiation for air-saturated versus nitcogen-saturatedfrom irradiated d-[CpT] and d-[TpC] are consistent with solutions of d-[CpT] and d-[TpC]. Our observed product Gthose observed by Ward and Cuo (27) and Teoule and co- values were significantly lower when dinucleotides wereworkers (28). On th, other hand, we did not observe intact irradiated in nitrogen-saturated solution as opposed to oxy-glycol derivatives or any other base-modified thymidine, genated solution, an effect attributed to the presence of2-deoxycytidine or uridine (modified intact nucleoside), or oxygen.corresponding intact modified nucleoside monophosphates Although the dinucleoside monophosphates studied hereprior to acid hydrolysis, which is consistent with the obser- are crude DNA models, they are nevertheless useful sincevations made by Belfi and co-workers (8). This rules out the the fundamental reaction pathways are presumably similarpossibility of two chemical effects, namely, base modifica- to those occurring in longer DNA segments. In the absencetion and strand scission, having occurred simultaneously of repair processes, our data demonstrate that the "pyrimi-on the same target molecule. Teoule and co-workers (28) dine-pyrimidine" system, composed of only two nucleo-have also reported that, even at relatively low radiation bases, responds differently to ionizing radiation of differentdoses, bases such as thymine and cytosine are released from qualities. Future objectives will be to extend these investiga-DNA. Detection of free thymine and uracil glycols could tions to the radiation chemistry of larger oligomers havingalso be a result of secondary radiation products, at least in specific lengths and base sequence, to quantify the amountpart. and type of products generated, and to determine whether

A recent report by Roots et al. (14) suggest a possible multicenter damage sites exist.

correlation of strand break yields with release of intact bases ACKNOWLEDGMENTin SV40 DNA. Several other authors have also reportedrelease of intact bases after irradiation under nitrogen (29), This research was supported by the Armed Forces Radiohiology Re-search Institute. Defense Nuclear Agency, under work unit 00)145.in air (oxygenated solution) (27), and in cultured mamma-

lian cells (30); the release ofintact nucleosides from dinucle- RECEIVED: April 8. 1992: ACCEPTED: August 31. 1992otides has been verified by Box and co-workers (IH). Tocompare and quantify the radiation-produced damage in REFERENCESmodel compounds exposed to fission neutrons or high-en- 1. R. Teoule. Radiation-induced DNA damage and its repair. Int J1ergy electrons and to test the assumption that the amount of Radial. Rio[. 51, 573-589 (1987).

RADIOLYSIS OF DINUCLEOSIDE MONOPHOSPHATES 19

2. M. Raha and F. Hutchinson. Deletions induced by -y rays in the i8. M_ Dizdaroglu, Use of capillary gas chromatography for identifica-genome of Escherichia coli. J. Mol Biol. 220, 193-198 (1991). tion of radiation-induced DNA base damage and DNA base-amino

3. M. Erreraa, Les effects des radiations nucleaires a faibles doses. La acid cross-links. J Chromatogr. 295, 103-121 (1984).Recherche 16, 959-968 (1985). 19. N. E. Geacintov and C. E. Swenberg, Chemical, molecular biology,

4. M. Sotheim-Maurizot, M. Charlier, and R. Sabattier, DNA radiolysis and genetic technique for correlating DNA damage induced by ioniz-by fast neutrons. Int. J. Radiat. Biol. 57, 301-313 (1990). ing radiation with biological end-points. In Physical and Chemical

5. D. Schulte-Frohlinde, Studies of radiation effects on DNA in Mechanisms in Molecular Radiation Biology(W. A. Glass and M. N.aqueous solution. ICRU NEWS 4-15, December (1989). Varma, Eds.), pp. 453-474. Plenum, New York, 1991.

6. J. Cadet and L. Voituriez, Seperation par chromatographie sur 20. F. Hutchinson, Chemical changes induced in DNA by ionizing radia-couche minceet par chromatographie liquide a haute performance de tion Prog. Nucleic Acid Res. Mol. Biol. 32, 115-154 (1985).dinucleoside-monophosphates modified par action du rayonnment 21. M. Dizdaroglu, E. Holwitt, M. P. Hagan, and W. F. Blakely, Forma-gamma. J. Chromatogr. 178, 337-343 (1979). tion of cytosine glycol and 5,6-dihydroxycytosine in deoxyribonu-

7. R. Teoule and J. Cadet, Radiation-induced degradation of the base cleic acid on treatment with osmium tetroxide. Biochem. J. 235,component in DNA and related substances-final product. Mol. Biol. 531-536 (1986).Biochem. Biophys. 27, 171-203 (1978). 22. J. C. Fox and N. J. McNally, Cell survival and DNA double-strand

8. C. A. Belfi, A. V. Arakali, C. R. Paul, and H. C. Box, Radiation break following x-ray or neutron irradiation of V79 cells. Int. J. Ra-chemistry of a dinucleoside monophosphate and its sequence isomer. diat. Biol. 54, 1021-1030 (1988).Radiat. Res. 106, 17-30 (1986). 23. K. M. Prise, S. Davies, and B. D. Michael, The relationship between

9. C. R. Paul, A. V. Arakali, J. C. Wallace, J. McReynolds, and H. C. radiation-induced DNA double strand breaks and cell kill in hamsterBox, Radiation chemistry of 2'-deoxycytidylyl-(3LS')-2'-deoxy-guan- V79 fibroblasts irradiated with 250 kVp x-rays, 2.3 MeV neutrons ofosine and its sequence isomer in nitrous oxide and oxygen saturated 238-Pu alpha particles. Int. J. Radiat. Biol. 52, 893-902 (1987).solutions. Radiat. Res. 112, 464-478 (1987). 24. G. W. Barendsen, Radiobiology of neutrons. Int. J. Radiat. Oncol.

10. C. R. Paul, C. A. Belfi, A. V. Arakali, and H. C. Box, Radiation Biol. Phys. 8, 2103-2107 (1982).damage to dinucleoside monophosphates: Mediated versus direct 25. R. Roots, A. Chatterjee, P. Chang, L. Lommel, and E. A. Blakely,damage. Int. J. Radiat. Biol. 51, 103-114 (1987). Characterization of hydroxyl radical-induced damage after sparsely

11. H. C. Box, C. R. Paul, and J. Przybyszewski, Studies on radiation and densely ionizing irradiation. Int. 1. Radiat. Biol. 47, 157-166damage using DNA oligomers. Free Radical Res. Commun. 6, 123- (1985).126 (1989). 26. D. E. Watt, Absolute biological effectiveness of neutrons and pho-

12. Y. N. Vaishnav, E. Holwitt, C. E. Swenberg, H.-C. Lee, and L.-S. tons. Radiat. Prot. Dosim. 23,63-67 (1988).Kan, Synthesis and characterization of stereoisomers of 5,6-dihydro- 27. J. F. Ward and 1. Kuo, Strand-breaks, base release and postirradia-5,6-dihydroxy-thymidine. J. Biomol. Struct. Dyn. 8, 935-951 (1991). tion changes in DNA gamma-irradiated in dilute oxygen-saturated

13. M. Dooley, L. J. Goodman, G. H. Zeman, R. B. Schwartz, C. M. aqueous solution. Radiat. Res. 66, 485-498 (1976).Eisenhauer, and P. K. Blake, Ionization chamber intercomparison in 28. R. Teoule, A. Bonicel, C. Bert, J. Cadet, and M. Polverelli, Identifica-mixed neutron and -y ray radiation fields by National Bureau of Stan- tion of radioproducts resulting from the breakage of thymine moietydards and Armed Forces Radiobiology Research Institute. Technical by gamma irradiation of E. coli DNA in an aerated aqueous solution.Report TR 86-3, Armed Forces Radiobiology Research Institute, Radiat. Res. 57, 46-58 (1974).Bethesda, MD, 1986. [Available from National Technical Informa- 29. M. Ullrich and U. Hagen, Base liberation and concomitant reactionstion Services, Springfield, VA, Accession number A 178336.] in irradiated DNA solution. Int. J. Radiat. Biol. 19, 507-517 (197 1).

14. R. Roots, E. Henle, W. R. Holley, and A. Chattedjee, Measurements 30. M. R. Mattern, P. V. Hariharan, B. E. Dunlap, and P. A. Cerutti,of nucleic bases released after '1 irradiation of DNA in solution in air. DNA degradation and exicision repair in gamma-irradiated ChineseRadiat. Res. 125, 288-292 (1991). hamster ovary cells. Nature 245, 230-232 (1973).

15. J. A. McCloskey, Mass spectrometry. In Basic Principles in Nucleic 31. A. F. Fuciarelli, B. J. Wegher, W. F. Blakely, and M. Dizdaroglu,Acid Chemistry ( P. O. P. Ts'o, Ed.), Vol. I, pp. 209-309. Academic Yields of radiation-induced base products in DNA: Effects of DNAPress, New York, 1974. conformation and gassing conditions. Int. J. Radiat. Biol. 58, 397-

16. J. A. McCloskey, A. M. Lowson, K. Tsuboyama, P. M. Krueger, and 415(1990).R. N. Stillwell, Mass spectrometry of nucleic acid components. Tri- 32. C. J. Koch, A thin-film culturing technique allowing rapid gas-liquidmethylsilyl derivatives of nucleotides, nucleosides and bases. J. Am. equilibrium with no toxicity to mammalian cells. Radiat. Res. 97,Chem. Soc. 90, 4182-4184 (1968). 434-442 (1980).

17. M. Dizdaroglu, Application of capillary gas chromatography-mass 33. A. F. Fuciarelli, B. J. Wegher, W. Gajewski, M. Dizdaroglu, andspectrometry to chemical characterization of radiation-induced base W. F. Blakely, Quantitative measurement of radiation-induced basedamage of DNA: Implications for assessing DNA repair processes, products in DNA using gas chromatography-mass spectrometry. Ra-Anal. Biochem. 144, 593-603 (1985). diat. Res. 119, 219-231 (1989).

ARMED FORCES RAOIOSIOLOGYRESE.ARCH INSTITUTE

Ia CIE1iTuFtC REPORTSRg3-11

Pharmacology Biochemistry and Behavior, Vol. 44. pp. 809-814, 1993 0091-3057/93 S6.00 + .00Printed in the U.S.A. All rights reserved. 1993 Pergamon Press Ltd.

Effects of Sublethal Doses of IonizingRadiation on Repeated Acquisition in Rats

PETER J. WINSAUER' AND PAUL C. MELE

Behavioral Sciences Department, Armed Forces Radiobiology Research Institute,National Naval Medical Command, Bethesda, MD 20889-5145

Received 6 July 1992

WINSAUER, P. J. AND P. C. MELE. Effects of sublethal doses of ionizing radiation on repeated acquisition in rats.PHARMACOL BIOCHEM BEHAV 44(4) 809-814, 1993. - To extend previous research on the effects of ionizing radiationon learning, dose-effect data with 6WCo y-rays were collected for individual rats responding under a repeated-acquisitionprocedure. Under this procedure, subjects acquired a different three-response chain each session by responding (nose push)on one of three transilluminated response keys in the presence of each of three sequentially ordered colors. The responsechain was maintained under a second-order fixed ratio (FR) 2 schedule of food presentation. An error produced a 5-s timeoutbut did not reset the three-response chain. Acquisition of each response chain was defined by a decrease in errors as thesession progressed (i.e., within-session error reduction). Each session ended after 200 reinforcements or 90 min, whicheveroccurred first. When day-to-day acquisition for all four subjects reached a steady state, the effects of three or four doses of'v-rays were assessed. In general, radiation doses of 1, 3, 4.5, and 8 Gy of gamma radiation delivered at a dose rate of 2.5 Gy/min produced a dose-dependent decrease in the overall response rate for 24-72 h after exposure in all four subjects. Radiationexposure also produced an increase in percent errors but only at doses that substantially decreased overall rate of responding.Unlike the effects on response rate, which were relatively consistent over a 72-h period, the effects on accuracy were greaterat 72 h than at 24 h in three of four subjects. The results indicate that the repeated-acquisition procedure may be particularlyuseful for quantifying the effects of ionizing radiation on acquisition behavior or learning and that 'v-rays can differentiallyaffect behavioral measures of rate and accuracy over a 72-h period following exposure.

Repeated acquisiti"" Oe-•nt behavior 'v-ravy' Rats

FEW studies have examined the effects of ionizing radiation One possible explanation may be related to the fact thaton learning, or the acquisition of behavior. Moreover, these there are large individual differences in rates of acquiring be-studies, which involved both monkeys and rats, have pro- haviors. This aspect of learning, along with the well-duced widely differing results ranging from improved to dis- established finding (11) that subjects reacquire specific tasksrupted learning [see (15) for review]. In the studies involving at a faster rate (i.e., subjects "learn to learn"), have consis-monkeys, for example, the reported inconsistencies can often tently posed problems for the study of learning and thosebe attributed to different tasks or doses and types of radiation variables affecting it. For these reasons, as Anger and Setzeror simply to the lack of data in general. However, in the rat (I) state in their article examining the effects of a pesticide onstudies where the effects of radiation have in general been learning, certain types of learning such as maze learning maytested on only one type of learning (i.e., maze learning), the not be the most'appropriate baseline for the evaluation ofinconsistencies in the data are still marked. Two reports (3,9), toxic agents because it can be a one-time learning phenomenonfor example, found that rats exposed to either 4.8 Gy of who- and is not repeatable in the same subject [cf. (17)J. The impor-le-body X-rays or 24.2 Gy of cranial X-rays actually per- tant distinction between testing the effects of a toxic agent onformed better on a maze task than nonirradiated control rats. learning and testing the effects of a toxic agent on retentionAlthough the effects in these studies have been explained in of learning was demonstrated in a radiation study by Urmervarious ways (e.g., irradiated subjects may be less distracti- and Brown (24). They found that sublethal doses of ionizingble), the failure to find convincing evidence that ionizing radi- radiation in rats produced no disruptive effects on the reten-ation affects the acquisition of behavior in a consistent man- tion of response patterns acquired preirradiation but did pro-ner is somewhat surprising. duce a decrement in the capacity of subjects to reorganize the

'To whom requests for reprints should be addressed.For ease of comparison with previous research, all the radiation doses cited (either roenglen or rad) have been converted to Gray (Gy). The

conversion factor for roengten to tad is 0,966 to account for the difference between air and muscle. Rad can be converted to Gray by multiplyingthe dose by 1/100.

809

810 WINSAUER AND MELE

preirradiation response pattern into a new response pattern produced an audible click of the feedback relay. Each cham-postirradiation. ber was enclosed within a sound-attenuating cubicle equipped

To deal with some of the problems involved in studying with a fan for ventilation. White noise was continuously pres-learning and the effects of drugs on learning, for example, ent in each chamber to mask extraneous sounds. The chain-Thompson (21,22) adapted a repeated-acquisition technique bers were connected to a PDPI 1/73 computer (Digital Equip-and used each subject as its own control. The repeated acquisi- ment Corp., Bedford, MA) programmed in SKED-1Ition of behavioral chains provided a baseline under which software (State Systems, Inc., Kalamazoo, MI) and to cumu-the effects of various drugs on learning could be examined lative recorders (Gerbrands Corp., Arlington, MA) located inrepeatedly, and the use of each subject as its own control a nearby room.helped deal with the problem of intersubject variability. Theresults indicated that repeated-acquisition procedures provide Procedurea sensitive baseline for assessing the effects of drugs on learn- Baseline. During each session, all three response keys wereing in individual subjects and that learning was more sensitive illuminated at the same time with one of three colors, eitherto drug effects than a performance condition where learning green, red, or yellow. The rodent's task was to respond (nosewas not required. Shrot et al. (18), for example, developed push) on the correct key in the presence of each sequentiallythis procedure in rats to examine the effects of microwave illuminated set of colors (e.g., keys green, center correct; keysradiation on learning. Although there were methodological red, left correct; keys yellow, right correct; reinforcement).differences (e.g., only a single condition and auditory cues vs. The same chain (in this case, center-left-right (CLR)i was re-visual cues), they found that microwave radiation was highly peated throughout a given session. The three-response chaindisruptive to learning at certain power densities. Since then, was maintained by food presentation under an FR 2 schedule,other investigators have effectively used this procedure with a that is, every second completion of the chain illuminated thevariety of species to examine the effects of many toxic agents pellet trough and produced a 45-mg pellet. A completion of[e.g., carbon monoxide (19), carbaryl (1), and lead (7)] on the response chain that did not produce food was followed byacquisition behavior, a 0.4-s presentation of the pellet trough light. When the sub-

To examine the effects of ionizing radiation on learning in ject pressed an incorrect key (in the example, the left or rightindividual rats, the present research used a repeated-acquisi- key when the green light was presented), the error was fol-tion task similar to that used by Thompson (21,22). More lowed by a 5-s timeout. During the timeout, the key lightsspecifically, the subject's task was to acquire a different three- were turned off and responses had no programmed conse-response chain each session by responding sequentially on quence. An error did not reset the three-response chain, thatthree keys in the presence of three colors. Acquisition of each is, the stimuli were the same before and after the timeout.response chain was defined by a decrease in errors as the To establish a steady state of repeated acquisition, thesession progressed. Following baseline stabilization where ac- three-response chain was changed from session to session. Aquisition for each subject reached a steady state, dose-effect typical set of five chains was CLR, RCL, LRC, CRL, anddata were obtained for three or four doses of 6CCo y-rays. RLC, with the order of the color presentations always green,

red, yellow (reinforcement). The chains were carefully selectedMETHOD in several ways and there were restrictions on their ordering

Subjects across sessions. More specifically, each sequence was sched-uled with equal frequency and adjacent positions within a

Four adult male Sprague-Dawley rats (R-9, R-13, R-14, sequence for a given session were different. Occasionally, aand R-16) maintained at approximately 80% of their free- correct sequence position for a given color was the same twofeeding weights (373, 404, 373, and 393 g, respectively) served sessions in a row.as subjects. Food was earned during the experimental session Sessions were conducted Monday through Friday betweenand, if necessary, was provided in the home cage after the 9:00 a.m. and 2:00 p.m. Each session was terminated aftersession to maintain subjects at their 80% weight. All subjects 200 reinforcements or 90 min, whichever occurred first. Thewere housed individually in plastic Microisolator cages con- data for each session were analyzed in terms of a) the overalltaining sterilized hardwood-chip bedding. The housing room response rate (total responses/min, excluding timeouts) andwas maintained at 21 ± I C with 50 ± 10% relative humid- b) the overall accuracy, expressed as percent errors [(errors/ity on a 12 L: 12 D cycle, which began at 6:00 a.m. each day. total responses) x 100]. In addition to these measures basedAcidified water (pH 2.5-3) was available in the home cage upon session totals, within-session changes in responding werethroughout the experiment to minimize the possibility of op- monitored by a cumulative recorder and the computer. Forportunistic bacterial infection. Each subject had an extensive example, acquisition of a response chain was indicated byhistory of repeated acquisition of three-response chains under within-session error reduction, that is, a decrease in the num-fixed-ratio (FR) schedules. ber of errors between food presentations as the session pro-Apparatus gressed.

Radiation testing. Following baseline stabilization, dose-Four identical modular test chambers (Coulbourn lnstru- effect data were obtained for multiple doses of gamma radia-

ments, Inc., Model EI0-10TC) configured specifically for ro- tion. Subjects received bilateral, whole-body, midline tissuedents were used. The front wall of each chamber contained a doses of 1, 3, 4.5, and 8 Gy of y-rays administered at a fixedhouselight, speaker, auditory feedback relay, pellet trough (10 rate of 2.5 Gy/min fro,a a "Cobalt source. These doses werecm above the floor and centered), and three response keys selected from a range of doses previously found to have analigned horizontally (8 cm apart, center to center, and 4.5 cm effect on other schedule-controlled operant behaviors (e.g.,above the floor). Each response key could be transilluminated fixed-interval (FI) and FR schedules (13)]. In general, dosesby three Sylvannia 28ESB indicator lamps, one with a red of radiation were given in a mixed order except for the 8-Gyplastic cap, one with a green, and one with a yellow. Response dose, which was administered last in three of four subjects.keys required a minimum force of 0.15 N for activation and Subject R-16 received the 8-Gy dose as a third exposure and

RADIATION EFFECTS ON LEARNING 811

was allowed 13 weeks of baseline recovery before being given as large as those for overall response rate (i.e., accuracy wasa final 4.5-Gy exposure. The minimum time between expo- comparable for all four subjects). In general, exposure to radi-sures for doses less than 8 Gy was 4 weeks. This interval was ation produced dose-dependent increases in percent errors inchosen to allow for a) complete baseline recovery [cf. (13)] each subject. These effects were most evident at the 8-Gy doseand b) the collection of sufficient control data prior to the for both the 24- and 72-h sessions. In this regard, the effectsnext exposure. on accuracy were unlike those on overall response rate, where

For irradiation, subjects were placed in well-ventilated, an intermediate dose (4.5 Gy) of radiation produced an effectclear plastic restraining tubes. A clear plastic stand, which in three of four subjects. Note that exposure to 4.5 Gy onlyheld the tubes in a stacked position, allowed all four subjects increased percent errors in RP-16 and RP-9 at 72 h. In otherto be exposed at one time, if desired. Dosimetry for bilateral words, decreases in response rate tended to occur at lowerirradiations was completed prior to the actual animal irradia- doses than those required to increase percent errors and in-tions. Standard Task Group 21, Radiation Therapy Commit- creases in percent error occurred only at doses that substan-tee of the American Association of Physicists in Medicine tially decreased overall response rate.(AAPM), protocol procedures (20) were used. A 50-cm 3 ion- Interestingly, the effects of the 8-Gy dose of radiation onization chamber was used to obtain the free-in-air (FIA) tissue percent errors were different from the effects on response ratedose rate at the appropriate exposure position. A 0.5-cm 3 tis- in that percent errors were affected more at 72 h than at 24 hsue-equivalent ion chamber was also used to obtain the tissue in three of four rats. In these three rats, the increase in percentdose rate in the same position in a tissue-equivalent rat phan- errors at 72 h occurred at a time when the overall responsetom. The tissue-to-air ratio (TAR) was calculated by dividing rates were comparable to those at 24 h. Moreover, in RP-9the nominal 0.5-cm3 chamber reading by the 50-cm3 chamber this noticeably larger effect at 72 h occurred at the 4.5-Gyreading. The administered dose (midline tissue at the abdo- dose of radiation when the decrease in overall response ratemen) to each animal was determined by the TAR value, the was less than that which occurred at 24 h (i.e., at a time whenFIA value, duration of the irradiation, and other factors such response rate appeared to be returning to control levels).as temperature and pressure. Each irradiation was conducted Figure 2 illustrates some of the within-session effects of anon a Monday and required approximately 20 min. This time 8-Gy dose of radiation in subject RP-9. Each cumulative re-included the actual exposure and the time necessary for trans- cord represents a complete session from a different day. Asporting rats to and from the exposure area. Sham irradiations, can be seen in the control record, errors decreased in fre-which also included transport, consisted of subjects' being quency while the number of correct completions of the chainplaced in the restraining tubes for a comparable amount of increased in frequency as the session progressed. This within-time. session error reduction, which occurred shortly after the start

of the session, reflects acquisition of the response chain. Fol-

RESULTS lowing acquisition, the pattern and rate of correct respondingremained relatively constant throughout the session. Although

Under baseline conditions, stable responding in the repeat- some pausing did occur toward the end the session, the rated-acquisition task was obtained in each of the four subjects. did obtain the total number of available reinforcers. At 24 hStability was reflected in the consistent levels of overall re- after an 8-Gy dose of radiation (middle record), there was asponse rate and accuracy for each subject from session to substantial decrease in the overall rate of responding, a de-session. Acquisition of response chains was characterized by a crease in the number of correct sequence completions, andsteady state in terms of stable within-session error reduction, an increase in pausing. Note that periods of no respondingthat is, the number of errors decreased as each session pro- occurred earlier in the session and were longer in durationgressed. than under control conditions. In addition, there was little

Figure 1 shows the overall response rate and percent errors evidence of any within-session error reduction, as indicatedfor each subject during control sessions and sessions at 24 and by the relatively constant error rate that occurred on the event72 h following -y-ray irradiation. Although all subjects were pen when the subject was responding. At 72 h after the 8-Gytested daily after radiation exposure, the data from these two exposure (bottom record), similar within-session effects onsessions most clearly illustrate the effects obtained. In general, overall response rate occurred. As can be seen, long pausesthe effects at 48 h after exposure were comparable to those at occurred throughout the session and the total number of rein-24 h; and in three of four cases, those at 96 h were comparable forcers obtained during the session was substantially reduced.to those at 72 h. For the sessions presented, a dose of radiation In this session, however, the total number of errors waswas considered to have an effect to the extent that the postirra- greater than that in the 24-h session even though the overalldiation data fell outside the control range. As shown in the response rate was comparable. This difference in accuracy isupper panel, despite relatively large individual differences in evident in the pattern and frequency of errors indicated bythe control ranges, radiation exposure dose dependently de- the event pen in both records. In general, the within-sessioncreased the overall response rate in each of the four subjects effects of this dose of radiation were replicated in two of theduring both the 24- and 72-h sessions. Although there was an remaining three rats. Although RP-16 did show an increase ininstance in which the effect of radiation at 72 h was larger percent errors and a decrease in overall response rate at thisthan that at 24 h (i.e., the rate-decreasing effect obtained at dose, there was little or no difference in the within-sessionthe 8-Gy dose in RP-16), the effects on response rate during effects at 24 and 72 h.the 24-h session were in general larger than or consistent with Three of four rats (RP-9, RP-13, and RP-16) received athose that occurred 72 h after exposure. This was in particular total -y-ray dose of 16.5 Gy. The other rat, RP-14, received aevident in RP-16 and RP-9 at the 4.5-Gy dose, where the total dose of 13.5 Gy. This total dose was lethal in RP-9,rate-decreasing effects obtained 24 h after exposure were nota- RP-13, and RP-14. RP-9 and RP-13 died within 2 weeks ofbly larger than those obtained 72 h after exposure. their final 8-Gy exposure, whereas RP-14 (the subject that

As can be seen in the lower panel of Fig. 1, the differences received the smallest total dose) died more than 2 months laterin control ranges for percent errors across subjects were not after completely recovering baseline levels of responding. The

812 WINSAUER AND MELE

60 RP-16 RPNM RP-13 RP-14

-45- 2 ir

cc 30-wLUz0. 15-U')

a: --

0 1 r T-r -T-T- T f ' T " IF I I T I IIl I F I I I I F

0

zI

C 1 3 45 8 C 3 4.5 8 C 3 45 8 C 1 3 4.5 8GAMMA-RAY DOSE (Gyl

FIG. 1. Effects of varying doses of 6OCo -y-rays on the overall respoise rate (upper panel) and percent errors (lower panel) for each subject. The

unconnected filled point and vertical line at C indicate the mean and range of the control data (i.e., data from the 10 baseline sessions

immediately preceding the first exposure and 2-5 sessions between irradiations where sham exposures were conducted). The filled points indicate

the data obtained 24 h after each exposure; the open points indicate the data obtained 72 h after each exposure.

RP-9

control

5mi

8 Gy

24 hrt.

72 two.

FIG. 2. Within-session effects of 'Co -7-rays during a control session (top record) and two sessions following exposure to an 8-Gy dose (middleand bottom records) in subject RP-9. Each of the three cumulative records is from a different day and shows a complete 90-min session. In eachrecord, the response pen stepped upward with each correct response and ws deflected downward each time the three-response chain wascompleted. Errors are indicated by the event pen (below each record), which was held down for 5 s during each timeout.

RADIATION EFFECTS ON LEARNING 813

surviving subject, RP-16, received the 8-Gy dose and a 4.5-Gy tive than other tasks for examining the effects of radiation ondose 13 weeks afterward. Except for small decreases in food acquisition behavior. Studies investigating the effects of bothintake for several days following exposure, this subject re- ionizing radiation (8,24) and microwave radiation (10,18) onmained in good health with baseline recovery after each ex- accuracy have noted that complex behavioral tasks are moreposure. sensitive to radiation than simple behavioral tasks. This view

has already been demonstrated to be true for a variety of drugDISCUSSION effects (22,26). In regard to radiation effects, for example,

Urmer and Brown (24) reported that the effects of 4 Gy ofUnlike many of the previous studies examining the effects gamma radiation on learning were most notable when the

of ionizing radiation on acquisition behavior in rats, the pres- subject was rechallenged or asked to "reorganize" a preirradia-ent experiment found substantial effects on both rate and ac- tion response pattern into a new response pattern. If proce-curacy of responding after administration of 1-8 Gy of -y-rays. dural manipulations of this type are critical to showing radia-The large dose-dependent effects on overall response rate were tion-induced deficits in learning, as the authors suggest, asimilar to effects found in other studies using either X-rays or repeated-acquisition task may be particularly sensitive because3y-rays and schedule-controlled operant behavior (4,5,12, subjects are rechallenged (i.e., required to learn a different13,25). In these studies, whole-body irradiation with compara- sequence) daily.ble doses produced dose-dependent decreases in responding A simple explanation could be that many prior studiesregardless of the schedule (fixed ratio or variable interval) or (2,3,6,9) failed to examine the effects of radiation at the criti-positive reinforcer (food or water) used. Wicker and Brown cal times postexposure. In the present study, the effects on(25), for example, found that repeated exposures with 1.9, percent errors were greater at 72 h than at 24 h in three of3.9, and 7.7 Gy of gamma radiation decreased responding four subjects. These effects on accuracy were unusual in thatunder an FR 1 schedule of water presentation. More recently, peak effects on percent errors were obtained long after theMele et al. (13) reported that 4.5 and 6.5 Gy of whole-body onset of effects on overall response rate. Different time-y-ray exposure produced marked decreases in overall respond- courses for radiation-induced decreases in response rate (anding under an FR 50 schedule of milk presentation. As in the for recovery from such decreases) have already been shown topresent study, the effects on response rate in that study tended be partially dependent upon the schedule of reinforcement,to peak 24 h after exposure and remained evident for extended type of reinforcer (positive or negative), and even the particu-periods of time (e.g., 5 days) depending upon the dose admin- lar behavior being tested (12-15). However, these factors haveistered. Also of note was the fact that repeated exposures not yet been shown to influence radiation-induced effects onat 6- to 9-week intervals showed no evidence of cumulative accuracy. One factor that has been shown to play a role in thebehavioral effects. More specifically, repeated exposures in effects on response rate, but been restrictive to establishingthe same subjects with the same doses produced reliable rede- effects on accuracy, is radiation sickness (i.e., a syndrometerminations whereas a subsequent exposure after these rede- generally characterized in a variety of species by weakness,terminations with a different dose produced dose-dependent fatigue, lethargy, and a decrease in food intake). The influ-effects. ence of radiation sickness has previously been so problematic

The effects of gamma radiation on accuracy in the present in many maze studies (3,6,9) that investigators have tested forstudy were in general different from those found in previous effects on learning after long postirradiation intervals (e.g.,studies involving other complex acquisition tasks (i.e., maze 20-60 days). Unfortunately, all of these studies failed to findacquisition tasks). Only a few studies have reported error-in- any significant disruptions of learning following radiation ex-creasing effects (6,8,24), and only two found effects of whole- posure.body irradiation at comparably low sublethal doses and post- The present study does seem to indicate that decreasedexposure days. Fields (8), for example, reported that maze motivation resulting from radiation sickness cannot solely ac-errors in an elevated T-maze were greatest 72 h after exposure count for the disruptive effects on accuracy and that motiva-to 3.5 or 5.8 Gy of X-rays. Other maze studies, however, tional deficits associated with radiation sickness may have ahave found acquisition behavior relatively unaffected (2,9) or time course independent of that for radiation-induced de-enhanced (3) after radiation exposure. Although the disrup- creases in accuracy. At the doses tested (1-8 Gy), onset oftion of acquisition behavior under the repeated-acquisition radiation sickness in rats would in general be expected withintask in this study occurred only at doses that substantially the first 24 h after radiation exposure. Jarrard (12), for exam-decreased overall response rate, the effects on percent errors pie, has shown that decreases in food consumption werewere clearly evident in all four subjects. In one of the few clearly evident 24 h after exposure and beginning to return toother rodent studies using a repeated-acquisition procedure baseline 72 h after'exposure depending upon the dose. In theirand a within-subject design, error-increasing effects were re- study on the effects of radiation on schedule-controlled per-ported with microwave radiation. Schrot et al. (18) found that formance, Mele et al. (13) also stress that there are severemicrowave exposure at varying power densities dose depen- limitations on attempts to relate radiation-induced changes indently increased errors four to six times over control levels performance to specific food intake changes that occur duringwhile decreasing sequence-completion rates. In addition, radiation sickness. They found little correlation between thehigher power densities of microwave radiation completely dis- magnitude and time course of disruption in performance andrupted the pattern of acquisition in each of the subjects tested. whether or not postsession chow was consumed. Moreover,

The effects on accuracy in the present study may have been repeated-acquisition studies with both monkeys (23) andmore extensive than those found in other studies involving pigeons (21) have specifically shown that prefeeding manipu-ionizing radiation for a variety of reasons. First, use of a lations (which decrease deprivation level and, thereby, the ef-within-subject design may have helped eliminate some of the fectiveness of food reinforcement) generally produce rate-variability that can often obscure effects on learning in group decreasing effects but little or no effects on accuracy.studies [cf. (21)]. Another possibility is that the repeated- As was the case with another 6°Co gamma radiation studyacquisition task used in the present study may be more sensi- involving operant schedules of reinforcement and food pre-

814 WINSAUER AND MELE

sentation (13), lethality in this study did not appear to be individual rats responding under a repeated-acquisition proce-strictly the result of the total cumulative dose received. dure. Although overall response rate was more sensitive toRather, lethality seemed to depend upon the order of doses disruption than percent errors (i.e., decreases in response rateadministered and the interirradiation interval. Given the small tended to occur at doses lower than those required to increasenumber of animals used, it is impossible to say with any cer- percent errors), presentation of three or four graded doses oftainty to what degree lethality was influenced by radiation 6Co -y-rays did produce dose-dependent disruptive effects orhistory. However, it should be noted that the lethal effects of both overall response rate and accuracy in all subjects tested.radiation on individual subjects in this study were in accor- Of further importance was the finding that percent errors at adance with previously established effects of 7-rays or X-rays given dose were differentially affected during a 72-h periodon groups of rodents. Specifically, dose fractionation in- postexposure. Unlike the effects on response rate, which werecreased the total cumulative dose that could be tolerattu with- relatively consistent (or in some cases lessened) over the 24-out producing lethality (13,16), and recovery from behavioral and 72-h periods after exposure, the effects on percent errorsdisruptions did not necessarily reflect recovery from the long- were notably larger at 72 h than at 24 h. Together, theseterm physiological effccts that produce lethality (13,15). This findings emphasize the need for extending the evaluation oflast point was reflected in the present study, and in the Mele the effects of ionizing radiation in both rati and other specieset al. study (13), by the fact that there was often a temporal using complex operant proceduirs such as repeated acquisi-separation between the more acute behavioral disruptions and tion.the lethal effects. In the present study, for example, two ofthe three rats that died did so many days after showing more . KNOWLEDGEMENTSimmediate behavioral disruptions. The third rat, RP-13, died The authors thank Dr_ Donald M. Thompson for helpful com-over 2 months after completely recovering baseline levels of ments on the manuscript and Ens. James F. Verrees for expert techni-responding. Moreover, the data from the surviving subject cal assistance in conducting this experiment. This work was supported

(RP-16) seems to indicate that future studies that incorporate by the Armed Forces Ridiobiology Research Institute (AFRRI), De-fense Nuclear Agency. Research was conducted according to the prin-longer interirradiation intervals after high doses could poten- ciples e..unciated in the Guide for the Care and Use of Laboratory

tially provide the same information about the acute effects of Animals pr.pared by the Institute of Laboratory Animal Resources,-y-rays on learning without obtaining a lethal effect. National Research Council, DHEW Pub. No. (NIH) 85-23, 1985.

In summary, the present research found that acute suble- AFRRI is fully accredited by the American Association for Accredita-thai exposure to ionizing radiation readily disrupts learning in tion of Laboratory Animal Care.

REFERENCES

1. Anger, W. K.: Setzer, J. V. Effects of oral and intramuscular physiological changes with exposure to ionizing radiation. In:carbaryl administrations on repeated chain acquisition in mon- Zaitchuk, R.; Bellamy, R. F.; Ingram, V. M., eds. Textbook %,'keys. J. Toxicol. Environ. Health 5:793-808; 1979. military medicine. Medical consequences of nuclear warfare. Falls

2. Arnold, W. J. Maze learning and retention after X-radiation of Church, VA: TMM Publications; 1989:105-151.the head, J. Comp. Physiol. Psychol. 45:358-361; 1952. 16. Paterson, E.; Gilbert, C. W.; Matthews J. Time intensity factors

3. Blair, W. C. The effects of cranial radiation on maze acquisition and whole body irradiation. Br. J. Radiol. 25:427-43.1; 1952.in rats. J. Comp. Physiol. Psychol. 51:175-177; 1958. 17. Peele, D. B.; Baron, S. P. Effects of scopolamine ua iepeated-

4. Brown, W. L. Response rate during X-irradiation and recovery acquisition of radial-arm maze performance by rats. 3. Exp.following irradiation. 1. Genet. Psychol. 108:117-120; 1966. Anal. Behav. 49:275-290; 1988.

5. Brown, W. L.; Overall, J. E.; Logie, L. C.; Wicker, J. E. Lever- 18. Shrot, J.; Thomas, J. R.; Banvard, R. A. Modification of thepressing behavior of 1lbino rats during prolonged exposures to repe3ted acquisition of response sequences in rats by low-levelX-radiation. Radiat. Res. 13:617-631- ;1960. microwave exposure. Bioelectromagnetics 1:89-99; 1980.

6. Burt, D. H.; Ingersoll, E. H. Behavioral and neuropathological 19. Shrot, J.; Thomas, J. R.; Robertson, R. F. Temporal changes inchanges in the rat following X-radiation of the frontal brain. J. repeated acquisition behavior after carbon monoxide exposure.Comp. Physiol. Psychol. 59:90-93; 1965. Neurobehav. Toxicol. Teratol. 6:23-28; 1984.

7. Dietz, D. D.; McMillan, D. E.; Mushak, P. Effetts of chronic 20. Task Group 21, Radiation Therapy Committee AAPM. A proto-lead administration on acquisition and performance of s,:rial posi- col for the determination of absorbed dose from high energytion sequences by pigeons. Toxicol. Appl. Pharmacol. 47:377- photon and electron'. oean',. Med. Phys. Io:741; 1983.384; 1979. 21. Thompson, D. M. Repeated acquisition as a behavioral base line

8. Fieids. P. E. The effect of whole-body X radiation upon activity for studying drug effects. J. Pharmacol. Exp. Ther. 184:506-514;drum, straightaway, and maze performance of white rats. J. 1973.Comp. Physiol. Psychol. 50:386-391; 1957. 22. Thompson, D. M. Repeated acquisition of response sequences:

9. Furchtgott, E. Effects of total body X-radiation on learning: An Stimulus control and drugs. J. Exp. Anal. Behav. 23:429-436; 1975.exploratory study. J. Comp. Physiol. Psychol. 44:197-203; 195 1. 23. Thompson, D. M.: Moerschbaecher, J. M. An experimental anal-

10. Galloway, W. D. Microwave dose-response relationships on two ysis of the eflCct, of d-amphetamine and cocaine on the acquisi-behavioral tasks. Ann. NY Acad. Sci. 247:410-416; 1975. tion and performance of response chains in monkeys. J. Exp.

II. Harlow, H-. F. The formation of learning sets. Psychol. Rev. 56: Anal. Behav. 32:433-444; 1979.51-65; 1949. 24. Urmer, A. H.; Brown, W. I.. ihe effect of gamma radiation on

12. Jarrard, L. E. Effects of X-irradiation on operant behavior in the reorganization of a complex maze habit. J. Genet. Psycholthe rat. J. Comp. Physiol. Psychol. 56:608-611; 1963. 97:67-76; 1960.

13. Mele, P. C.; Franz, C. G.; Harrison, J. R. Effects of sublethal 25. Wicker, J. E.; Brown, W. L. The effect of gamma radiation upondoses of ionizing radiation on schedule-controlled performance operant water-reinforcement b,'havior. J. Genet. Psychol. 106:in rats. Pharmacol. Biochem. Behav. 30:1007-1014; 1988. 295-299; 1965.

14. Mele, P. C.; Franz, C. G.; Harrison, J. R. Effects of ionizing 26. Winsauer, P.J..; Thompson, D. M.; Moerschbaecher, J. M.radiation on fixed-ratio escape performance in rats. Neurotoxi- Comparison of drug effects on fixed-ý atio perf-rmance and chaincol. Teratol. 12:367-373; 1990. performance maintained under a second-order fixed-ratio sched-

15. Mickley, G. A.; Bogo, V.; West, B. R. Behavioral and neuro- ule. J. Exp. Anal. Behav. 44:367-376; 1985.

Nucleic Acids Research, 1992, Vol. 20, No 23 6167-6176

Differences in unwinding of supercoiled DNA induced bythe two enantiomners of anti-benzojia]pyrene diol epoxide

Rong Xu, Sheryl Birke, Susan E.Carberry+, Nicholas E.Geacintov*, Charles E.Swenborg' andRonald G.Harvey 2

Chemistry Department, New York University, INiew York, NY 10003, 'Radiaiion BiochemistryDepartment. Armed Forces Radiobiology Research Institute, Bethesda, MD 20814 and 2 Ben MayInst~ute, University of Chicago, Chicago, IL 6063i', USA

ARMED FORCES RADIOSIOLOGYj RESEARCH INSTITUTE

Received October L7, 1992, Accepted October 30. 1992 CEircRPT

ABSTRACT

The unwine~ig of supercoiled oX174 RFI DNA induced N2 -guanosine covalent adduct is situated at externalby the tumorigeric (+) and non-tumorigenic (-) binding sites and the mech,,nisms of unwinding areenantiomers of trc. is-7,8-dihydroxy-anti-9,1O-epoxy- therefore different from those relevant to noncovalent7,8,9,10-tetrahydrobenzo[alpyrer. (BPDE) has been intercalative BPDE-DNA complexes or za classicalinvestigatet~. by agarose slab-gel and ethidiumn titration intercalating drug molecules; a flexible hinge joint andtube gel electrophoresis. The differences in adduct a widening of the minor groove at the site of the lesionconformiations were verified by flow linear f$ichroism may account for the observed unwinding effects. Thetechniques. Both enantiomers cause a reversible more heterogeneous (- )-BPDE-nucleoside adductsunwinding by the formation of noncovalent intercalative (involving cis and trans t42-gUano'~.ne, and adenosinecomplexes. The effects of covalently bound BPDE adducts) are less effective in causing unwinding ofresidues on the electrophoretic mobilities of the RIF I supercoiled DNA for reasons wvhich remain to beDNA form in agarose gels were investigated in detci' elucidated.in the range of binding ratios rb ý'0.0-0.06(covalently bound BPDE residties/nucleotide). In this INTRODUCTIONrange of rb values, there is a stiriking difference in themobilities of (+)-BPDE- and (-)-BPDE-adducted <5X1 74 Benzolalpyrene and related polvcyclic aromatic hydrocarbonsDNA in agarose slab-gels, the covalently bound (PAl-) are metabolized in vivo to potent mutagenic and(+ )-BPDE residues causing a significantly greater tumorigenic 'liol epoxide derivatives [II]. These hydrophoiiic.retardation than (- )-BPDE residues. Increasing the ciectrophilic. and highly reactive compounds can bind bothlevel of covalent adducts beyond rb 0.06 in the case noncovalently and covalently to DNA in aqueous solution.. Theof tha (+ )-BPDE enantiomer, leads ttý further unwinding formation of non-covalent complexes is important in deter-miningand a minimum in the mobilities (corresponding to co- the chemical reaction kinetics of' these diol epoxide derivativ.esmigration of the nicked form and the covalently closed with nucleic acids 121. while the covalent binding to DNA is arelaxed modified form~) at rb 0.10 _+ 3 01; at still higher critical step in the expression of their mutagenic and turnorigenicrb values, rewinding of the modified DNA in the potentials [31- The ultimatc biologically active metabolite ofopposite sense is observed. From the minimum in the benizo[a]p~rene, is the bay region 7.8-dihydroxv-9, 10-epoxy-mobility, a mean unwinding angle (per BPDE residue) derivative. Among the four different stereoisomners ofof 0 = 12:11.50 is determined, which is in good 7.8-dihydroxy-9, 10-epoxy-7,8,9. 10-tetrahydrobenz~o[alpyrenc.agreement the value of *9 = 11±_+1.80 obtained by the the 1+1-enantionier of the diastereomer trans-7,8-dihydroxv-tube gel titration method. Using t0 is latter method, anti-9, I 0-epoxy-7,8,9, 10- tetrahvdrobenzol alpyrcioe (( )_13PDL.values of *9 = 6.8 ±+ 1.70 for (- )-BP[.-_-X1 74 adducts known also as ( ±)-anti-B3PD!-. or ( !)-BPDE 2) is the most activeare observed. It i- con'tluded tha' agarose slab gel tumorigen and displays the highest mutagenic ac~iv'ity intechniques, are not suitable for determining unwinding mammalian cells 14-81. In contrast. the (- 1-enantiorner,angles for (-)-BPDE-modified 0174 DNA because the (---)-BP[)[- (or (-)-arni-B3PDE, or (---BPT)E 2) is,alterati ,ons in the tertiary structures for rb < 0.06 are nontumorigenic 16.71. and its mnutagenicity is quite different fromtoo small to cause sufficiently large changes in the that of the (4 )-enantiomner [4,5,81. Differences in covalent adductelectrophoretic mobilities. The major trans (+ )-BPDE- conformations and extent of alteration of the local DNA structure.

-TIo whomn (orresp''ndcrwc ~houtd he .ifdrc',.ed

.Prescnt addre", tDeparnnu'nt of Cheitti-trN Huntcr Coluiicc (it the City iii NcA York, Ne% York. N Y I (W I. 1 NA

6168 .Viucleic .Acids Research. /992, V01. 20. No. 2.3

are believed to be two of the critical kai,-ors which distinluish 40. mMV 'I ris. H-UI, 10) mM MgIp 1 . andj 12 n1M\ ( .( [the biological activities of' i - )-BPDF anid (-13fBr)[ 12,4,1) 101. 7.8) buffer so~lutionl at 37 C he adducis %Acrc Wrnhcr di1'estcdIThe two enantiolners 'B~.of -PI are characrerized bx differences wNith snake %enoiri phospho klisicrase i arid jlkilninc phospliatascin their chemical binding patterns with 1)NA 14- 11 -- 131, and Pharmnacia I .KB Boitcehrioltwý InII,_ l'icat&Aaw a NJ ii i I(the conformations of the chemical i)NA adduetIS which arc miM lris 1-10. 1(X ImM NJCI aind 15 HIM Mtg(l.T sN IL1tin iorformed [revieAed In 2,9.,101. Most recent)y. the solution 24 hour,. atl 37 C . The 1-1P1 .( elutio n mles of thc dif fercent

conformat ions of the most abundant. stereochem ical I definied IWID -D N A digests were comiiiard "with those oI BPI )l- N: d(Irians-adducts deriv ed fromt thre binding, of I -+ )- and -- )-BPD[- and F3PI)-N`-dA standards Using ice'.erse phase 1-1I.(' ineihodisto the exocsclic amnino group Of guan~os\I moi0eties InI as describewd elsewhere 124.251 f, sing molar cmtlnmionoligonulcleotides has been determined bsý hiehI r'esolution NMR coefficient,, of' 290MX) M I cm:n at the 340 inaminiun !22: tormethods: the bulk,, pyrenvl resiclies deried from these two i +- 1-BPDI-'DtNA adducts. and li 5() MI at 31S2 3S4 rnin forstereoisorners are positioned in the minor groove of DNA. anid t -- i-BP[)h-DNA adducts (the absorption niasinia aire broaderare oriented in nearly opposite directions with respect to the and less well defined 12.1.261), agreeiment be-tween thie lilI'(-5' -3' strand polarity: all hydrogen bonds are intact at 5<C( and and spectrophotonicitric miethodis wkas, within ± 2

iW; The low erthe B-DNA structure is maintained near and at the lesion site extinction coefficients for adducts witah absorpt ion niaxmini at

[1.IL352 -- 354 nuni is InI agreement wo ith recent result,, obtajined w ithDifferences in changes Iin the tertiary, structures of DNA site-specific and stereospecific 131Loioulo Id dducits

induced by the two eniantiomners of BPDE. lor example bends 125,27].or flexible hinge joints at the site oft the lesion, and unw indingof supercoils in closed circular DNA. could also play a role in G;el electrophoresisdefining the differences in the biological effects induced by these The electrophoretic inob11ilte's of UnIntlikifiCd and BPI'D-, 1.ni slititwo isemrers. The unwinding of supekrcoiled DNA by racemlic supercoiled [)NA were compared UsingI I '; agarose wkedge-BPDE (( ±- )-BPDE) via noncoxalent complex (116,.171 and shaped slab-gefs Iin 89 iN-I Tris ba~se. 89) ni1M sodium borate.covalent adduct t18-211 formation. has been Investigated andi 2 rnMi EDTA solution (VBF- butter) The final dimensionsprev iously. In this work, the effects of the resolved enantiomiers of' thre eels \,ere 15 X 20 cmn " ith the thick arid thin ends abo~ut

+- -BPDF and ( -iI-BPDF. arc described for the first time. anid I I IS1 and 3 --5 mmn thick. respectisel\ - Bet we the san11iples wereTunwinding angles 0 are estimated by. different methods. added to the gels, at loading buffer ( I miM I-DA. 2,5"; Ficoll.

and 0.WX5% broniophenol blute) was added. Flectrophorcsis was

MATERIALS AND METHODS Performied (thick - thin direction) using an LKB 20i12 MaxiphorSubmarine electrophoresis unit iWharniacia-I.KB. Pisc:atawkaN. NJ)

Preparation of BPDE-DNA adducts connected wo a 13uehley Model No. 3- I5kX po-wcT ,U~pINThe (-+- t-BPDE and (- )-BPDE enantiomiers w'ere obtained from (Haakebuchler. Saddle River. NJ) at 40 -45 V and 24 mA forthe National Cancer Institute Chernical Carcinogen Reference 22 hours. The gels were then stained with ethidiuni bromide it) 75Standard Repository. Supe rcoiled OX 174 RFI DNA was pg,'ml) for 30) mm. and destained for 31) min with wýater.purchased front Bethesda Research Laboratories (Gaithersburg. photographed under UV illumination, and the film negatises wecreMD). The covalent binding reactions of the BPDE enantiomners thenr subjected to densmitoiet rN analysis. The densitomneter waswith DNA were carried out in 5 mnM Tris (Tris(hydroxy- a Fisher Scientific (Springfield. NJ) Model [('91It. manufacturedmnethyl anminomethane) buffer containing I mM sodium by E-C Apparatus Corporation (St. Petersburg. FL... andi wasethylenediaminetetraacetate (EDTA) at PH 7.9 (TE buffer). coupled to at computer systemi. In the range oft DNAVarying anmounts of a concentrated BPDE-tetrahvdrofuran concentrations utili/ed. it w.as verified that thle areas, uinder tnesolution (2 -4 mM BPDE) were added to 5WK ti aliquots of buffer denisitomneter Peaks were proportional to the amiount of DN Asolutions containing 12 gg, of supercoifed DNA (7.5 x 10 1M added to the gels (40 ng DNA in 10) pi, per well)in concentration of nucleotides) to initiate the chemical binding In order to relate the decreases Iin mobilities of the partiall'.reaction: the concentration oftettrahydrofuran did not exceed 3 % relaxed super-coiled DNA bands to the number of superhielicalby volume in any of the experiments. The samples were placed turns removed by the covalentlN bound BPDF residues. differenton a shaker for four hours at room temperature- Subsequently. topo~isomier distributions ofo~Xl74 DNA were prepared. 1ýeactmonthe reaction mixtures were exhaustively dialyied against TE mixtures were prepared containing 3.2 A~g of'oX 174 1)NA andbuffer to remove the 7,8.9, 1 0-tetrahydroxvtetr".- 30 units of topoisonterase I ((Bethesda Research Laboratoriesihydro-benzo[alpyrene (EPYT) hydrolysis products. in 50 miM Tris-- HCI, 501 mM Nd.It) 10iM' MgCl -,. 0. 1 mM~

EDTA , 0.5 muM dJihiothreitol. and 30 jig bovine serum albumnin.Determination of extent of modification and product The reactions were allowAed to proceed at 37V' for time inter-,alsdistribution by HPLC analysis of I. 2., 4. 6, 10 anid 18. The reactions, were stopped b\ addingSpectroscopic methods based oin the UV absorbance of the a 25:24:1 solution of pheniol: chloroform:v isoam' II alcohol.adducts in the 346 -354 nmnl region were employed inl most cases fiilinm and subsequent extract Ion wxith thlis soluttit on i tw i.cI . andto estimate the level of' covalent binding [22.23]: the accuracy one extraction with efiloroflirni.of this rapid and convenient methodl was verified by entvmaticall-vdegrading the covalent BPDE-DNA adducts to nucleosides andl Tube gel muethod.- tiltration of sizp'rhellical turns 4Aith ethiditumquantitatively deterrmining the fraction of unmodified and BPDFE- bromidemodified nucleosides, by reverse phase high performance liquid This approach 128,291 is particularlyý Useful Iior deterniining thechromatography flHPL-C) by established methods [ 131 and used titratable superheflical dcnsitN Iin arnon-isild NA 3 -0]routinely in our laboratory 124,251. Briefly, the BPDL- Briefly. agarose tube gels were prepared] bN poluring hot U-it) C)supercoiled I)NA adducts were first digested with I)Nase I inl I '7( agarose solutions, in 0I 018 M NaCI Il-F1 buffer Containing

Nucleic Acids Research, 1992, VtK-. 20, ,Nu, 23 6169

different amounts of ethidium bromide (0.001 --0. l( ug/L) into as the concentration of EB is increased 117,35]. A typical EB20 cm-long glass tubes with one of their ends closed with titration curve obtained by the flow linear dichroism method isparafilm. After 45 min, the solidified tubular gels were cut to shown in Fig. 1. The minimnum in the LI) signal, correspondinga length of 18 cm and inserted into a vertical tube gel apparatus to a superhelical density of zero 117.35J. is observed at the cnticalwith the upper and lower chambers filled with 0.018 M NaCt ratio of rc = (EB added /(DNA nucleotidc) = O048±0.0()3.TBE buffer solution. About 200 ng of modified or unmodified Assuming that - 95•% of the EB molecules are bound to the DNAsupercoiled DNA in 10 aL of solution containing 4.5• ficoll at this equivalence point 136]. the critical binding ratio U(and 2.4 ,uM ethidium bromide EB/(DNA nucleotide) ratio (bound EB moleculesinucleotide) is equal to =0,95r. =I= 0.2) were added to the tops of ttte tubes. Electrophoresis was 0.046:± 0.003. Defining the superhelical density a in terms ofperformed at 95 - 100 V (2.5 mA/tube, up to 16 tubes at a time) the unwinding angle 0 = 26' (per bound EB molecule) accordingfor 3 hours at 20'C. After electrophoresis, the gels were extruded to the formula a = -(t,(c/1 8 )0 [37], we obtain a value of o =into separate large test tubes and the DNA bands were developed -0.066 c0.0M4 tinder our own conditions of ionic strengthby adding a 0.7 •ggL EB solution for 30 min with shaking. and (0.071) and temperature (24'C) which, after correcting for thewere subsequently photographed under UV illumination, differences in temperature and ionic strength. is close to the

standard published values of -0.057 ±0.010 at 37'C and 0.2Flow linear dichroism measurements of relative adduct M NaCI concentration 137]. Since the number of base pairs, inconformations OX174 DNA is 5386 1381, this value of a corresponds to 35 ± 3Linear dichroism (LD) measurements provide information on the superhelical turns per molecule at 24°C.relative orientations of the long axis of the pyrenyl residue relativeto the average orientations of the planes of the DNA base.. [31 ]. Unwinding due to formation of noncovalent. unstable BPDE-Briefly, the aqueous DNA solutions are placed within the annular DNA complexesspace of a Couette cell consisting of two concentric suprasil When (+)-BPDE (or (-)-BPDEI molecules are added to ancylinders, with the outer cylinder (23 mm inner diameter) aqueous solution of <X 174 DNA, there is an immediate increaseremaining stationary, while the inner one (22 mm outer diameter) in the magnitude of the LD signal which returns approximatelyis rotated at speeds of 400 RPM. The resulting hydrodynamic to its initial value after some 12- 15 min with exponential deca)flow gradient (900 s , 1) causes a partial orientation of the DNA kinetics (Fig. 2). The first-order decay constant under thebases perpendicular to the flow direction. The linear dichroism experimental conditions used here is 0.0020 s 1. Because of thesignal is defined as LD = A -A where A, and A are tendency of the planar DNA bases to orient with their normalsabsorbances measured with the linearly polarized light E-vector parallel to the flow lines, the linear dichroism signal of theoriented parallel and perpendicular with respect to the flow supercoiled DNA solution is negative in sign below 3(X) nm [351.direction, respectively. Additional details concerning the LDapparatus may be found elsewhere [32,33]. Linear dichroism characteristics and conformatioi.. of

covalent BPDE-DNA adducts

RESULTS After removing the BPT molecules from the equilibrated reactionmixtures by exhaustive dialysis, the LD spectra of covalent

Superhelical density adducts shown in Fig. 3 are obtained. These spectra are analogousThe superhelical density of the OX174 DNA used in these to those obtained with linear DNA 122,23.26.39-411. thusexperiments can be most conveniently evaluated from ethidium indicating that the adduct conformations in supercoiled DNA arebromide (EB) titration-unwinding data [34]. Because the similar to those observed in linear DNA.hydrodynamic volume of supercoiled DNA molecules increases The (-)-BPDE-OXI74 covalent adducts (Fig. 3B) displa, aas the degree of superhelicity is decreased, the flow linear red-shifted (relative to BPDE or BPT in DNA-free aqueousdichroism method can be used to monitor the degree of unwinding solution) negative LD spectrum resembling in shape the LD (and

I.D LD

-0.0010 0.000

-0.0015-0.001

-0.0020Q

-0.0025 .002

-0.0030.01"0 0.02 0.04 0.06 0.08 0.1 -0.003 500 t 150 2000[EB/Nucleolidel lIME.

Figure I. Linear dichroisnim gnal mcasured "xithin the DNA ahsorptton hand Figure 2. Kinetic tra•.c of the Ittear dithroi',m signal measorei at 2t'i omn I Diat 20) nrn as a ftuni(.tiin ol the ethidiutn hromide concentratiOn expressed in tertns reflecting the uni inding and rckwndi• g 1of the suer, oitled D)NA i I q 0 '•M)of the molar rat•o [: B added kI )DNA nucleolidej] The DNA concentration *Ai', alter the addition (a[ time I i) i) ot f lt BPI)I (8 6 j•i, • im) ilir ec|i, I,10 ) M, the hullcer soluioii in was identtica] ti the tine emphoycl in the tithe gel Aere i hinsr.ed in the c.ae it the t cnart totier ItliI 11ot 'h,'s " I I he ttli• tnlitexperimnents te . Matterials and Methixksi. the temperaturc was 24"C oI the I t(+-) BP e- nantio met is also hslt),.s

6170 Nucleic Acids Research, 1992, Vol. 20, No. 2-3

0.Ak)CW)-a Pr)___ tAt i+)-8PDE4'-Xt74 ONA M)t i--BPDIE-4X t74 DSA

LD-0.001

110-0.002-

3.!1)(-)-SPDE Figure 4. Photographs ot ethidiunin-Liined .igarose ýIah VON ýonip..ring d,'tain~exof rnigdtinon of cmnalefltly imajdified BPt)E-eX 174 t)NA 'smiple' at ditterent Iceclkofrt, covalentls hound BP)E residuesnuj:eo~tide, A) ,i-BPI)LOX 74 l)NA,adducts. ,(B) -)-BPDE-oXl741DNA adductis. r, saluev IA lane'. Oi))'. i2)0.036,13) 0.054.(4)0.10,(510 12.(6i 0 12.C7)0 14, iit 1 6 i111 c ]n I

LD 170.0. (2) 0.020, (3) 0 030). (4) 0.03S, (5)0 045. (6)0 05.3 All %aluc' ) f,, art:within zi 20's

:E 6. and 18 ±E 3% of all covalent BPDE products. How~ever.only overall adduct levels are reported here since, in any c',eru.the effects of different types of adducts on the unwainding of'supercoiled DNA cannot be resolved in our experiments.

0.00 I 1.02 260 300 340 360 400 Gel electrophoresis

Wavelength, nm Photographs of ethidiurn bromide-stained agarose slab-gelsshowing some typical electrophoresis data obtained with covalent

Figure 3. Linear dichroism spectra of covalent )±)-BPDE-oX 174 and -) (-BPDE- (+ )BPDE-tX 174 and (- -BPDE-oX 174 DNA adducts areoX 174 adducts in solution (after exhaustive dialysis of the equilibrated reaction shown in Fig. 4. Typical densitometeir tracings obtained frommixtures). In (A). the initially ad,;ed BPDE concentration was as in the legend the negatives are displayed in Fig. 5. The relative proportion~sto Fig. 2. w hereas in (B) all experimental conditions were as in (A). except that of the supercoiled RF I and relaxed (nicked) RF If DNA formsthe initial reaction mixture contained 26x 10 6 NM I-)-BPDE.weesm havribefo smpetsml.

The electrophoretic mobilities of' the supercoiled l)NAcovalently modified with (+ )-BPDE are markedly, slow.ed relative

absorption spectrum) of noncovalently bound intercalated DNA to the unmodified DNA: approximate co-migration of the[23,32]. These types of quasi-intercalative binding sites have been supercoiled DNA and the nicked form is observed for relativelydesignated as site 1 [2,9,10.39]. high binding ratios rh = 0,09-0.11 (10± I/( of the bases

The LD spectrum of the (+)-BPDE-OX174 DNA adducts is modified). This behavior is attributed to the removal of left-mostly positive in sign above 300 nm and is characteristic of handed superhelical turns and the concomitant increase in theexternal site 11 adducts, although a negative LD signal due to overall hydrodynamic size and lower electrophoretic mobilitiesthe contribution of smaller fractions of site I adducts is also visible of the BPDE-DNA molecules.(Fig. 3A); in site 11 adducts, the pyrenyl residues are at least The mobilities of covalent adducts derived from thepartially exposed to the aqueous environment 126,42.43] and are (- 1-enantiomner are barely affected in th,2 range of r *tilted ( > 50') away from the planes of the DNA bases 0.00-0.06 (because of' the low reactivities of' the ( - (-BPDE[22,26,32,39.401. enantiomer with DNA. solubility problems at r, > I1.5. a,, well

Distibuton o reationprodctsas the relative scarcity and high cost of the BPDE enantioniers.Distibuionof eacton rodctsno attempts were made to obtain adducts with higher levels ot

Using initial reaction ratios r, ([initial BPDE]/[DNA nucleotidel) modification with this isomer.in the range •- 0.5. it was found that 22 + 4% of the initially Values of the relative mobilities as a function of the numberadded (+)-BPDE binds covalently to DNA. This fraction tends off +)-BPDE residues per genome are show~n in Fig, 6: the datato dlecline with increasing values of r, -e o.5 In the cause off (- - points represent the maxima in the distributions which areBPDE. 3.5 -6.6% of the diol epoxide molecules were converted observed in the densitometer tracings (Fig. 5vý At high ( + )-BPr)Eto covalent DNA adducts in the range of r, =0.ý3 -1.5, the adduct levels, these distributions are quite broad, especiallyhighest fraction being obtained at the lower r, values. In the case beyond the value of r6 ý 0. 10. above thts value of r,. rewkindingof the (-s-.-BPDE-DNA adducts, only the trans-BPDE-N 2-dG of the closed circular D)NA gives rise to positive stiperblieticty.adduct was detectable (> 90% of all adducts), consistent with and is reminiscent of the effects observed with LB i Fig. 1).previous findings u.~ing racemic BPDE and linear [4,11,131 and In contrast to the behavior of the ,upercoided band, thesupercoiled [19,441 DNA. In the ease of the (-)-BPDF-DNA electrophoretic mobility of the slower, nicked band, increase".adducts, the adduct distribution is more heterogeneous [4,11,13]. somewhat upon modlification with )i-BPI)[,'i iFig. 5). [In the caseThe (-)-trans N2-dG, (- )-cis N2-dG, (---)-trans N6-dA, and of the (- )-BPDEOX 174 DNA adducts. [the mobilit tito theother, unidentified adducts, accounted for 35 + 5, 18 + 2, 29 nicked band does not seem to change visibiti as a function oI- r,.

Nucleic Acid. Research, /992, Vol. 20, No 23 617J

rb (+)-BPDE rb (-)-BPDE 120b b S4100 1.,

0.00 0.00

60 -

0.04 400.020 2

20

0 0 0.05 0). 1 0.15 0.2

0.030 ADD U'TS/N 1.IEOTIDE

Figure 6. Relative mnobilities of BPDE-supercoded DNA adtuicts in el.ixi'ophretu.agarose slab-gels as a fuinction of the covalent modification level r . - ( +-

0.10 BPDE-bX!74 DNA adducts. U ..... .- )-BPDE-oX174 DNA adducts The0.038 values provided correspond to maxima in the mobility distributions eoaluated from

densitometer tracings.

0.12 A0,045

__ __ __ _1 2 3

0.160.053

DISTANCE MIGRATED

Figure 5. Densitometer tracings of photograph negatives (same data as in Fig. 4).

Comparisons of electrophoretic mobilities of (+)-BPDE-OX174 DNA adducts with mobilities of unmodifiedtopoisomers

Samples of untreated supercoiled DNA were partially relaxed 120with topoisomerase I for varying amounts of time, and thesesolutions were then subjected to gel electrophoresis. Typical 100photographs of slab gels are shown in Fig. 7A. In our OX174RF I samples. we cout led 19 bands, including the fully relaxedRF 11 and supercoiled RF I bands. This approach is not capable 6of revealing all of the bands corresponding to molecules with 40different number of superhelical turns, since some of the highestand lowest mobility bands cannot be resolved using agarose slab- 20gels; however, the resolved adjacent bands differ from one 0another by one superhelical turn [451. The relationship between 0 5 10 15 20the band number (starting from the fastest bands) and relative BAND NUMBERelectrophoretic mobilities (measured as a fraction of the distanceon the slab-gels between the RF I and RF II bands in theunmodified samples) is plotted in Fig. 7B. Figure 7. (A) Electrophoretic slab-gel separation of opoisomcrase I-treated

unmodified OXI74 DNA. Lanes: (1) reference, untreated DNA. t2) treatmentThe relative mobilities of the (+)-BPDE-•X174 DNA adducts for 2 min. and (3) 4 min. (B) Positions of the different bands observed ufmn(Fig. 6) were then matched with those in Fig. 7B, and the treating OX174 DNA with topoisomerase I. The distance of each band %kitscorresponding band positions were plotted as a function of the measured relative to the distance between the fast RF I and slow RF i1 bandsnumber of covalently bound (+)-BPDE residues/OX 174 DNA of untreated DNA measured on the same gelsmolecule (Fig. 8) as previously described for other bulky adducts[30,46). A reasonably good straight line with a slope of The experimental data points corresponding to 40 < [BPDEJAT/!ABPDE] = 0.017 + 0.001 turnsiadducts is obtained < 200 adducts/genomc lie well below the straight line in Fig. 8.([BPDEJ = adducts/genome); the effective unwinding angle Apparently, at low adduct levels, it is difficult to detect anydetermined by this method is 3600×Axr/A[BPDE - changes in the mobilities due to remi,,; of the first few6.2 ± 0.5', superhelical turns.

6172 Nucleic Acids Research, 1992, Vol. 20, No. 23

25 gel was employed in order to determine the approximate valueof C', (2) the experiments were then repeated with the same DNA

20 - sample using a narrower range of C near the value of C' estimatedfrom the first experiment. In this way, a more accurate value

S15 of C' was obtained.z Examples of such experiments are shown in Fig. 9. Using aa 1coarse range of C values, the value of C' for unmodified OX 174"• RF I DNA is found to be in the range of 0.040 -0.050 ggiml

5 (Fig. 9A); using a narrower range of C values, C' is found to

0 L, lie in the range of 0.042-0.046 jug/ml: the corresponding value0 " 200 '400 '600 800 "1,000 1,200 1,400 of ov( is 0.046±0.002, yielding values of a = -0.066 ± 0.(X)3

ADDUCTS/GENOME and 35 ± 2 superturns/DNA molecule, in excellent agreementwith the data obtained from the linear dicbroism EB-titrationcurve (Fig. 1).

Figure 8. Comparisons of band positions of (+)-BPDE-OX 174 DNA adducts Examples of similar experiments with (+)-BPDE-(pX 174 andwith those of topoisomerase I-treated unmodified OX 174 DNA as a function of (-)-BPDE-OX 174 DNA adducts are shown in Figs. 9B and 9C.covalently bound BPDE residues per DNA molecule (see text). respectively. In both cases, rewinding of the DNA in the opposite

sense is observed at ethidium concentrations C > C'. In the case

Unwinding measured by the tube gel method of the (+)-BPDE adduct with a level of modification of ri =0.05, the value of C' is 0.021 ±0.003 jug/ml (Figs. 93). whichChanges in electrophoretic mobilities of carcinogen-modified is considerably smaller than in the case of the unmodified DNA.

supercoiled DNA on agarose slab-gels may, in principle, also In the case of the (-)-BPDE-uX174 DNA sample with a similar

reflect alterations in the tertiary structures of the molecules other level of modification (rb 0.045), C' 0.030± 0.003 gmil.

than those produced by changes in the number of superhelical Thus, both enantiomers, upon binding covalently with DNA, give

turns. Other effects may include bends [47] or flexible hinge joints rise to measurable degrees of unwinding as measured by the tube

[2,23,31,41] at the site of binding. The adducts derived from gel method.

the binding of (-)-BPDE appear to be predominantly stacked Particularly striking are the apparently small effects of the

(or site I, quasi-intercalated [22,401) with neighboring bases as covalently bound (-)-BPDE residues on the mobilities of the

judged from linear dichroism experiments with linear DNA

[22,23,26,40,411 and with supercoiled DNA (Fig. 3) , while modified DNA as measured by the agarose slab-gel method (Figs.4 and 5), and the definite unwinding effect detected in the same

(+)-BPDE residues appear to2be3situated tly we range of binding by the gel method (Fig. 9C)s the loss ofexternal site II binding sites [22,23,26,32,40-43]; thus, we superhelical turns increases with increasing levels of (-i-BPDEinitially expected to find more unwinding in the case of ( +)- modification, and values of 0 in the range of 6-8 ' are obtainedBPDE- than in the case of (-)-BPDE-4Xl74 DNA adducts; (al )

however, the exact opposite conclusion can be derived from the (Table I).

agarose slab gel results (Figs. 4 & 5). Therefore, another methodwas sought to determine the losses of superhelical turns. The tube DISCUSSIONgel technique is convenient and suitable for this purpose. In this Unwinding of supercoiled DNA by noncovalent binding ofmethod [28,29], samples of the same DNA adduct solution are BPDEsubjected to agarose gel elec'rophoresis in individual tubes, each Noncovalent BPDE-DNA complexes are formed on time scalescontaining EB molecules at different concentrations, thus of milliseconds after the addition of the PAH diol epoxides [2].producing different degrees of unwinding in each tube. The and cause the rapid initial increase in the magnitude of the linearnumber of superhelical turns can be estimated from the EB dichroism signal at 260 nm (LD>6) shown in Fig. 2. The decayconcentration C' in the tube in which the fully relaxed covalently kinetics of the LD,6 signal are identical to the kinetics ofclosed circular DNA molecules co-migrate with the nicked form, disappearance of intercalated PAH diol epoxide molecules [ 17].using the equation [28,29]: The LD 2,6 signal decays because the BPT hydrolysis products

vc = Urn KC'/(I + KC) are characterized by a -6 times smaller binding affinity(noncovalent complex formation) than BPDE [48]; the overall

where K is the equilibrium binding constant of EB and ur, is the number of noncovalently bound intercalated molecules thusmaximum molar ratio of bound EB/nucleotide. Since the decreases with time.magnitude of K depends on the ionic strength and other factors, Both enantiomers of BPDE cause the unwinding of supercoiledwe determined its value under our own experimental conditions DNA by noncovalent complex formation. Most 116.23.48 - 501using standard equilibrium dialysis methods. A Scatchard plot though not all [49] non-covalent PAH diol epoxide-and otherwas constructed (data not shown), which yielded values of K = PAH metabolite-DNA [511 complexes appear to be intercalative(3.0:± 0. 1) x 106 M - and Un = 0. 18, similar to those reported in nature. Meehan et al. [161 have shown that the unwinding angleby other workers [28,291. The superhelical density is then associated with (± )-BPDE intercalation is only 13', about onecalculated from the values of v(- using the relationship a = half of the value observed with EB 137,521.- I .44vc. In these tube gel experiments, the levels of covalentBPDE adducts were held below rh = 0.06, since the association Slab-gel electrophoretic mobilitiesconstant for EB binding was not measurably affected in this range Supercoiled firm (RF I). There is a striking diflfrence in theof BPDE binding levels (data not shown). electrophoretic mobilities of adducts derived from (he binding

The tube gel experiments were run in two stages: (I) a coarse of (+)-BPDE and ( -)-BPDE to 4.X 174 DNA at similar valuesscale of incremental increases of the EB concentration C in each of rh ( - 0.06, Figs. 4-6). At the highest level of moification

Nucleic Acids Research, 1992. Vol. 20, No. 2.? 6173

(A) OX174 DNA1 2 3 4 5 6 7 8 9 10 I1 12 13 14 15 16 I 2 3 4 5 6 7 8 9 10 I1 12 13 14 15 16

(B) (+)-BPDE-cIXI74 DNA

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

(C) (-)-BPDE-AX174 DNA

I 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 I 2 3 4 5 6 7 8 9 10 Ii 12 13 14 15 16

Figure 9. Photographs of tube gels of (A) untreated OX174 DNA, (B) (+)-BPDE-OX174 DNA adducts, rt, 0.05, and (C) (-)-BPDE-tX174 DNA adducts,rh = 0.045. Left: coarse Ethidium bromide concentration intervals; right: finer intervals. EB concentrations in pg/ml: (A) Left: (t) 0.00)1. (2) 0.005. 3) 0.010.(4) 0.020, (5) 0.025, (6) 0.030, (7)).035, (8) 0.040. (9) 0.050, (10) 0.060, (11) 0.070, (12) 0.090, (13) 0.090. (14) 0.110, (15) and ( 16) 0.00 (A) Right: ( 1 ) 0 022.(2) 0.026, (3) 0.030. (4) 0.034, (5) 0.038, (6) 0.042, (7))0,046. (8) 0.050, (9) 0.054. (10) 0.058. (11) 0.062. (12) 0.066. (13) 0.070. (14) 0.074. and t15) and(16) 0.000. (B) Left: (1) 0.001, 2) 0.005, (3) 0,010. (4) 0.015, (5) 0-020. (6) 0.025, (7)).030, (8) 0.035. (9) 0.040, (10) 0.045, (II) 0,050(. (12) 0.060, (13) 0.070.(14) 0.080. and (15) 0.00. and 16) 0.00, unmodified DNA. (B) Right: (1) 0.010. (2) 0.012, (3) 0.014. (4) 0.016, (5) 0.018, (6) 1.020. (70.022. (8) 1.024. 19)0.026, (10) 0.028. (11) 0.030, (12) 0.032. (13) 0.034, (14) 0.036, and (15) 0,00, and (16) 0.00, unmodified DNA. (C) Left: (1 0.0015. (2) 0.010. (3) 0015. (4)0020. (5) 0.025. (6) 0.030, (7)).035. (8) 0.040. t9) 0.045, (101 0.050. (1I) 0.055. (121 0.060. (13) 0.065, (14) 0.070. and (15) 0.00, and (161 0(M), unmodifiedDNA. (C) Right: (1) 0.018. (2) 0.021, (3) 0.024, (4) 0.027. (51 0.030, (6) 0.033, (7)1).036, (8) 0.039. (9) 0.042. (10) 0.045. (11) 0.048. (12) 0.051. (•13) 0.054,(14) 0.057. and (15) 0.00, and (161 0.00, unmodified DNA,

with (-)-BPDE, the average mobility of adducts is decreased Table I. Unwinding angles determined for covalent (- -BPDE-OX174 DNA

by only =7%, while in the case of the adducts derived from adducts at different values of r,.

(+)-BPDE the mobility is decreased by -45% (Fig. 6). The r5,- C'`' (,Ug/mt) 0(.1' Unwinding anglecovalent adducts arising from the binding of (+)-BPDE are more (I) (2) (3) (4)effective than (-)-BPDE in decreasing the mobilities of covalent 0(004 0.046±0.0015 26±2"BPDE- supercoiled DNA adducts. Thus, decreases in the 0019±0.002 0.038 0.041±0.0017 6.8A:09'clcctrophoretic mobilities of racemic anti-BPDE-SV40 0.030j± 0.002 0.035 0.039 ± 0.0(18 6.1 ±0 7'

supercoiled DNA adducts previously observed in agarose gels 0.038±0.003 0.033 0.037 ±0.0019 6.2 ±0 7'[18-211, are predominantly associated with the (+) enantiomer. 0).046±0.(X)3 0030 0.034±0.0020 6.9 ±-0 7'

0.053 ± 0.(0M u,026 0.030 0.0021 7.8 ±(0.99

Nicked form (RF ll. The increase in the mobility of the nicked 'Covalcntly bound BPDE residucs/nuclCoIde.closed circular RF II band with increasing rh for both BPDE iEthidium bromide concentration at which the covalently closed and nicked D)NA

co-migrate. Central value taken Ironi the higher resolution tutbe gel experiments,isomers (Fig. 4) have also been observed with racemic anti- the maximum uncertainties in these values is +0.002 g/nml (ýsee text)

BPDE-DNA adducts [19,201. The higher mobility ofthe moxlified 'Critical EB concentration ratio which leads to the complete relaxation of

form I1 DNA was attributed to a flexible hinge-like behavior at supercoiled modified DNA; the uncertainties, reflect the extreme ranges of ,

6174 Nucleic Acids Research, 1992, Vol. 20, No. 23

the alkylation sites [ 19]. The observed reductions in the mobilities Table II. Summarii of unvinding angle, obtained h% ditfcrcnt nictho.d,,

of short linear DNA fragments [473, and decreases in the degrees ADDUCT Slab Gd. Slab (ii. Slab (d. 'ut• tkof alignment of high molecular weight linear DNA molecules FROM: ArAN, Relative ('Complelcmodified with racemic BPDE and (+)-BPDE in hydrodynamic Moh1illtlc%'Y' unwXinding ,flow gradients [23,31,41]. are consistent with this hypothesis. l (2) (3) (41 ,,

Estimates of unwinding angle (+I-BPDE 6.2 ± 0 5' > 7' 12 II IX

Slab- gels. The unwinding angle 0 can be estimated using the (-)-BPD>I > 1is 1'

relationship [37]: :a'From topoisomer experiments. Fig 8=hh2 Estimated froim data in Fig. 6, see text

0IAJX360°)/(ir[ 10,772) (2) ''Estimated from the minimum in Fig 6 at r,, (o(11)) ± (•1 (1

where Ar is the number of superhelical turns removed at a givenvalue of rb, while 10,772 is the number of bases per genome ethidium bromide titration tube gel method. This suggests thatofOX 174. In principle, unwinding angles in carcinogen-modified the slab gel electrophoresis method is not suitable 1ior estimatingsuperhelical DNA can be determined by comparing the changes unwinding angles in those cases in which the unwindingin electrophoretic mobilities (Fig. 6) with those observed with anglelligand is small, and the degree of modification is ahothe unmodified topoisomers (Fig. 7). However, only a lower limit relatively low. In these cases, the removal of' the first fewof the unwinding angle 0 can be obtained in this way, as is shown superhelical turns does not appear to be sufficient to cause an%here. From Fig. 4, using rh values of 0.05, the relative mobility significant changes in the electrophoretic mobilities in agaroseof (+)-BPDE-DNA adducts is estimated to be • 55 %, while for slab gels.the (-)-BPDE-DNA adducts it is -93%, corresponding to a The value of 0 estimated from the data in Fig. 7 in the rangeloss of 1 10 and - 1.5 superhelical turns respectively. The fastest of superhelicities between N and N + 18 (Table 11, column 21 ismobility RF I band is known to consist of several highly wound smaller than the overall, mean vaiu,. 'f 8 obtained by othertopoisomers [45]. The loss of the first few N superturns is not methods (Table II, columns 4 & 5). This suggests that theexpected to exhibit any noticeable changes in the observed unwinding angle/[BPDE] residue must be higher than the averageaverage electrophoretic mobilities. Thus, at binding levels of rh value 0 = 11.5 + 2.4' for highly supercoiled <ýX 174 isomers,= 0.05, Ar = N + 10, or = N + 1.5, for (+)-BPDE- and Such a phenomenon was observed by Gamper ct al. 119] for(-)-BPDE-0X 174 DNA adducts. respectively. Since the values racemic BPDE-SV40 DNA adducts, who reported decreasingof Ar = 10 and 1.5 represent lowest limits for the numbers of unwinding angles with increasing levels of modification in thesuperhelical turns removed when rb = 0.05, unwinding angles range of 330' -30' , the highest values being obtained for highlNof 0 > 70 and 0 > 1 for (+)-BPDE-OX174 DNA and supercoiled topoisomers. These values (or the inferred mean(-)-BPDE-OX 174 DNA adducts, respectively, are estimated by value) are substantially higher than those obtained by us withthis method and Eq. 2. (+)-BPDE using a different kind of supercoiled DNA (Table II).

The observance of a minimum in the relative mobilities as a Gamper et al. [19] worked with a 10% dimethyl sulfoxidefunction of rb in the case of adducts derived from (+)-BPDE (DMSO)-90% buffer mixture, while the organic solvent(Fig. 6) provides yet another method for estimating the average concentration (THF) in our reaction mixtures was less than 3/%.unwinding angle per bound BPDE residue. With v, = 0.10 + However, we found that the levels of binding of ( + )-BPDE to0.01, and using Eq. 2 with r = 35 ± 3 (complete unwinding), OX174 DNA (and the relative electrophoretic mobilities) wereyields a value of 0 = 12 :k 1.5'. the same in the absence or presence of 10% dimethylsulfoxide

during the preparation of the adducts (data not shown). Therefore.Comparisons of unwinding angles 8 determined by different differences in reaction conditions do not appear to be the sourcetechniques. The results are summarized in Table II. The average of this discrepancy.value of 0 for (-)-BPDE-OX174 DNA adducts obtained by the There is a difference in the adduct level (r5 ) thresholds attube gel method is 6.8 ± 1.70 . Good agreement is obtained which unwinding is first observed in our experiments on agarosein the case of (+)-BPDE-6XI74 DNA adducts by the complete slab gels and those of Gamper et al. [191. These latter authors.unwinding-agarose slab gel method (column 4) and the tube gel based on the retardation of individual topoisomer bands inducedmethod (column 5); the average value of 6 is 11.5 4- 2.4'. This by (±)-BPDE in a different gel system, observed significantvalue is, however, about twice as large as the value of 0 estimated decreases in mobilities of SV40 DNA at levels of modificationby the band counting method (column 2). This latter value (6.2 of only 2-40 BPDE residues per genome. Drinkwater et al. I 181+ 0.5') was obtained by comparing the slab-gel mobilities of stated that the binding ratio rh in agarose slab gelshad to be at(+)-BPDE-OX 174 DNA adducts with those of unmodified least 0.015 (> 160 adducts/genome) in order to observe changesisomers with the number of superturns/molecule between N and in electrophoretic mobilities on agarose slab gels. In ourN+ 18 (Fig. 7). The value of N can be estimated from Eq. 2 experiments, changes in electrophoretic mobilities of modifiedusing 0((+)-BPDE) = 11.5 + 2.4', and 0 ((-)-BPDE) = 6.8 OX174 DNA are not observed unambiguously on slab-gels unless± 1.7' and Ar = N + 10 and Ar = N + 1.5 for (+)-BPDE- there are at least =-200 (+ )-BPDE adduct residuesi/enomc t(rDNA and (-)-BPDE-DNA adducts, respectively. Values of N >_ 0.02, Fig. 6). which is more consistent with the threshold= 7.2 ± 2.7 and N = 8.7 ± 2.5 are obtained which are in established by Drinkwater ct al. than the one lound by Gasperreasonable agreement with one another, and also with the et al. In these latter two investigations. racemic BPDE (ratherexpected number of unresolvable fast electrophoretic bands [45]. than the resolved enantiomers) were used. and the sources of

In the case of (-)-BPDE-OX 174 DNA adducts, the changes the supercoiled DNA were different from ours, It seems unlikcly,in the relative electrophoretic rmobilities observed on agarose slab though not certain, that these factors arc the sources of thegels are too small to reflect the value of 0 measured by the differences,

Nucleic Acids Research, /992, Vil 20, NM. 23 6175

Our value of the mean unwinding angle for (+ )-BPDE-DNA flexible hinge joint at the site of binding. Flow linear dichroismadducts (11,5 ±2.4°) is smaller than the value of 22±03 experiments [23,31.411 and molecular dynamic simulation studiespublished by Drinkwater et al. [181. Because the ( +)-isonmer of [601 suggest that such effects are predominantly associated withBPDE is about 3-4 times more reactive with respect to (+)-BPDE. rather than with (-)-BPDE. Furthermore. usingsupercoiled DNA than the ( -) enantiomer. the unwinding angle ligation and gel electrophoresis techniques 1611 with covalentshould be. all other things being equal, somewhat higher in the oligonucleotide adducts derived from the trawis and cis additioncase of (+ )-BPDE than in the case of racemic BPDE. This of both I +)- and (- )-BPDE. we have shown that only the ( + )-discrepancy is traceable to the differences in the values of rt, at trans-BPDE-N2 -dG adduct gives rise to a flexible hinge joint atwhich minima in the elctrophoretic mobilities are observed (rb the site of binding (B.Mao, B.Li and N.E.Geacintov. in(nin) = 0.05 ±0.008 in the case of Drinkwater et al., and rh preparation).(min) = 0. 10 +0.01, this work). Given the + 20% uncertainty Adducts derived from the binding of (-)-BPDE are chemicallyin our values ot r. . the disc )ancy is not large. Inaccuracies more heterogeneous. Little is known about the conformations ofin adduct level determinations could account for this difference. BPDE-dA adducts at this writing. The (- )-trans-BPDE-N 2 -dGAnother possible source of error is the absolute superhelical adducts are also situated in the minor groove, but their orientationdensity which is used to calculate 0 bas ,d on the experimentally are different from those of the analogous (+)-adducts [15]-determined v%,ues of r, (min). In this x, ork, we have made unwinding due to the mechanism discussed above is thereforeefforts to reduce the errors associated with these problems by also expected. However, the absence of a prominent hinge joint(I) determining the superhelical density of the OX174 DNA in (-)-trans-BPDE-N 2-dG-oligonucleotides suggests that thesamples under our own experimental conditions, and (2) basing overall unwinding may be lower than in the case of theadduct level determinations on two different methods. structurally related (+ )-trans adducts. The (-)-cis adduct.s appear

to be situated inside the helix and are at least partially base-stackedAdduct conformations and mechanisms of unwinding with neighboring bases [25,27]. However, their conformationClassical intercalation involves the insertion of planar aromatic may be different from those of classical intercalative structures:molecules between adjacent DNA base pairs, which leads to an the modified guanosyl residues may be displaced by the pyrenylextension of the phosphodiester backbone, an untwisting of base moieties in a quasi-intercalative carcinogen-base-stackedpairs. and thus to unwinding [53]. Flow linear dichroism studies conformation, as proposed recently by Singh et al. based onindicate that the noncovalent binding of (+)-BPDE and (-)- potential energy minimization studies of analogous, higher energyBPDE to DNA leads to adduct conformations c-)nsistent with trans-adducts [601. The degree of unwinding associated with suchthose of intercalative complexes [23]" however. noncovalent conformations may be lower than for classical intercalationracemic BPDE-SV40 supercoiled DNA complexes are complexes.characterized by unwinding angles of 13' [16], which is much Collectively, the heterogeneous (-)-BPDE adduct species givesmaller than the value of 26' [37,521 ot_ -- ', for the classical rise to a smaller unwinding angle than (+ )-BPDE-N2-dG lesionsintercalator ethidium bromide. While the unwinding of (Table I1). The reasons for these differences are presently notsupercoiled DNA by covalently bound (+)-BPDE is also close well understood and must await further developments in theto 13 ° (Table 11), the mechanisms of unwinding must be different structural determinations of site-specific and stereospecific BPDE-in the noncovalent complex and covalent adducts since their oligonucleotides adducts [14,15], and the analysis of theconformations are quite different from one another [23,32]. While associated structural effects involving unwinding of superhelicalthe unwinding of supercoiled DNA by drug molecules is often turns and other alterations in the tertiary structure of DNA.taken as evidence for an intercalative binding mechanism, Bauerhas emphasized that the demonstration of duplex unwinding ACKNOWLEDGEMENTSshould not be considered as incontrovertible evidence forintercalation [37]. Other causes of duplex unwinding not involving This work was supported by the Department of Energy (Grantsintercalative binding have been documented, including the effects DEFGO2-86ER60405. and in part by Grant CA 20851 from theof covalently bound cis- and trans-dichlorodiammineplatinum (II) US Public Health Service, Department of Health and Human[54,551, dehydration [56], the binding of histones [57] and metal Resources, awarded by the National Cancer Institute. Theions [581, and photochemical thymine dimer formation [59]. assistance of Dr. Y.Mnyukh with the linear dichroisni

In the case of ( +)-BPDE, over 90% of the adducts formed measurements is gratefully acknowledged.involve trans-addition (at the C 10 position) to the exocyclic aminogroup of guanine in native [i I - 13] and in supercoiled 144) DNA. REFERENCESRecent high-resolution NMR experiments with stereospecific andsite-specific ( +)-BPDE-deoxyoligonucleotide adducts have shown . Conncy, AH (12) Caner Rc. 42.4875 417+ ~hav snwn Geac nt',' N.E I (9588 Palt'cix' ,4rw Artmtti Ut driu'arhan ( a lr• /t~2'trn.' ,•

that the pyrenyl residue is situated in a widened m inor groove tn4t-ur -A tiitRet98n .Ahipc (K. Y atng and R D Sihenrnan, C n. ). V R('R

of B-DNA as gauged from the larger than normal distances Press. Biwa Raton. FL. V,+ II. pp, 181 -206

between adjacent negatively charged phosphate moieties [K14151. 3. Singer. B. and Gninhergcr. 1). i ,w50 M,,h'ular ffiolo.gx )/ ,lutae,rs and

Possible mechanisms of unwinding based on this effect can be G(ar,'i,n,,'e,n.. Plenum Press. New' Yink,. Brookes. P and Oshorne. NM R. (1982) (Cartonoev.t•est 3. 122, i 22fi

discussed in terms of a model recently proposed by Wilson et 5. waoI. A W.. Chang. R 1.. l'e%,in. W . Yagp. H . I hAker. t) R . Jrina.

al. [531 for classical intercalators: because of the lowered D.M. and Connev. A H. (1977) Bi'wrm. Bi,,pih, R,- (',rmnmn 77.

repulsive interactions between neighboring phosphate groups due 1399 -- 1396

to the minor groove widening. an untwisting of the base pairs 6 Buening, M.K.. VisIo.ki. P.VG . LC'n. iW . Yagi. H . Thakker. I) R . Ak.gi.

may occur which increases the favorable van der Waals contacts H.. Koreeda . 3. Jrina .M. anti . onne.. A I!. (197) Pro" Nat hai.S.i WSL•Aý 75, 5,358 5161.

due to better overlap between adjacent base pairs. This 7 Slaga.TJ - Bracke.n. W I., (lcaon. ( . ILo in. %W . Yfgi. IH .. crina, 1) M

mechanism of unwinding could be further complemented by a and conney, A H 1l979) (Cmcr Res .39. 67 71

6176 Nucleic Acids Research, 1992, Vol. 20, No. 23

8. Stevens, C.W., Bouck. N., Burgess. J.A. and Fahl. W.E. (1985) Murat. 48. Shahbaz, M., Geacintov, N.E_ and Harvey, R.6 (1986 Bitohetnistrr (U(SAiRes. 152. 5- 14. 25, 3290-3296.

9. Harvey, R.G. and Geacintov, N.E. (1988) Accs. Chem. Res. 21, 66-73, 49. Carberry, S.E., Shahbaz, M., Gcacintov. N.E and Harvey. R.G, (1987)10. Gr~islund, A. and Jernstr6m. B. (1989) Quart. Rev. Biophys. 22, 1-37. Chem.-Riol. Interaction, 66. 121-- 14511. Meehan, T. and Straub. K. (1979) Nature 277, 410-412. 50, Carberry, S.E . Geacintov. N.E. and Har-NcN. R G. (19893 Car(inow,'n,,t'12. Jeffrey. AM. (1985) in: Polvcvclic Aromatic Hvdrocarbons and 10. 97-103.

Carcinogenesis, ACS Symposium Series No. 283 (R.G. Harvey. Ed.). The 51. LeBreton, P. (1985) in P uly ' flthroarbons and Can inl•,on•cnc (R G

American Chemical Society, Washington. DC, pp, 187-208, Harvey, Ed.) ACS Symposium Steries No. 283. American Chicnn~al S4Kiete.

13. Cheng, S.C., Hilton. B.D., Roman. J.,M. and Dipple, A. (1988) Chein. Res. Washington. DC, 2109-238.Toxicol. 2. 334-340., 52. Waring. M.I. (1981) Ann. Rev. Biochem. SO. 159- 192.

14. Cosman, M.. de los Santos, C., Fiala. R., Hingerty. BE., Ibanez. V., 53. Williams, L.D., Egli, M.. Gao. Q, and Rich. A. (1992) in Sirniiurc &

Margulis, L.A., Live. D., Geacintov, N E., Broyde. S.. and Patel, D.J. Function, Volume /: Nucleit Acid, (Sarnia. RH. and Sanna. MI H , Eds.(1992) Proc. Natl, Acad. Sci. (USA) 89, 1914-1918. Adenine Press. Guilderiand. NY. pp. 107--125.

15. de los Santos, C., Cosman, M.. Hingerry, BE., Ibanez, V.. Margulis. L.A., 54. Scovell, W.M. and Collart, F. 11985) NVut lei- Acidi Rei. 13. 2881 -2895Geacintov, N.E.. Broyde, S.. and Patel, D.J. (1992) Bichemistry 31. 55. Cohen, G.L.. Bauer, W.R., Barton, J.K and Lippard. S.J (19783S¢'acntc5245 -5252. 203. 1014 - 1016.

16. Meehan. T.. Gamper, H. and Becker, J.F, (1982) J. Biol. Chem. 257, 56. Lee. C.-H.. Mizusawa. H. and Kakefuda, T. (1981) Pro•. Nail. Ad•d. Sit.10479-10485. (USA) 78. 2838-2842.

17. Yoshida, H., Swenberg, C.E. and Geacintov. N.E. (1987) Biochemistry tUSAI 57. Germond. I.E., Rouviere-Yanuv, J., Yaniv. M. and Brutlag, D. ( 19791 Pro(.26, 1351 -1358. Nail. Acad. Sci. (USA) 76, 3779-3783.

18. Drinkw-ater, N.R.. Miller, J.A., Miller, E.C. and Yang. N.C. (1978)Cahner 58. Anderson, P. and Bauer, W. (1978) Bioc•hemidstr (USA) 17. 594 -601.Res, 38. 3247-3255. 59. Boullard. A. and Giacomoni, PU. (1988) J. Pholochem, Photabiol..

19. Gamper. H.B., Straub, K.. Calvin, M. and Bartholomew, J.C. (1980) Prto. B:Biology, 2. 491 -501.Natl. Arai. Si. (USA) 77, 2000-2004. 60. Singh, S.S.. Hingerty, B.E.. Singh, U. C., Grecnberg. J.P., Geacinto%, N.E.

20. Agarwal, K.L.. Hrynio. T.P. and Yang, N.C. (1983) Biochem. BiophS. and Broyde, S. (19911 Cancer Re,. 51. 3482-3492.Res. Commun. 114, 14-19. 61. Schwartz, A.. Marrot. L. and Leng, M. (1989) . Moi. Biol. 207. 445-450.

21. Kakefuda, T. and Yamamoto, H.-Y. (1978) Proc. Nall. Acad. Sci. (USA)75, 415-419.

22. Geacintov, NE.. lbanez, V.. Gagliano, A.G.. Jacobs. S.A. and Harvey.R.G. (1984)1 J. Biomol. Structure & Dynamics 1, 1473-1484.

23. Roche, C.J., Geacintov, N.E.. lbanez, V. and Harvey, R.G. (1989) Biophys.Chem. 33, 277-288.

24. Cosman, M.. Ibanez, V., Geacintov. N.E. and Harvey. R.G. (1990)Carcinogenesis 11, 1667-1672.

25. Geacintov, N.E.. Cosman, M.. Mao, B., Alfano, A_. Ibanez. V. and Harvey,R.G. (1991) Carcinogenesis 12, 2099-2108.

26. Zinger. D., Geacintov, N.E. and Harvey, R.G. (1987) BiophysS. Chem. 27.131-138.

27. Cosman. M. (1991) Ph.D. Dissertation, New York University.28. DeLeys, R.J. and Jackson, D.A- (1976) Biochem. Biophys. Res. Commun.

69. 446-454.29. Espejo. R.T. and Lebowitz, J. (1976) Analyt. Biochem. 72, 95-103.30. Lang. M.C.E., Freund, A.M., de Murcia. G., Fuchs, R.P.P. and Daune,

M.P. (1979) Chem. -Biol. Interactions 28, 171 - 180.31. Norden, B.. Kubista, M. and Kurucsev, T. (1992) Quart. Rev. Biophys.

25, 51-170.32. Geacintov, N.E., Yoshida, H., Ibanez, V., Jacobs, S.A. and Harvey. R.G.

(1984) Biochem. Biophys. Res. Commun. 122, 33-39.33. Geacintov, N.E., Ibanez, V.. Rougee. M. and Bensasson, R.V. (1987)

Biochemistry (USA) 26, 3087-3092.34. Waring , J. (1970) J. Mo!. Bio!. 54, 247-279.35. Swenberg, C.E.. Carberry, SE. and Geacintov, N.E. (1990) Biopalymers

29, 1735- 1744.36. Crawford, L.V. and Waring.. M.J. (1967) J. Mol. Biol. 25. 23-30.37, Bauer, W.R. (1978) Ann. Rev. Biophys. Bioeng. 7, 287-313.38. Sanger. F., Air, G.M.. Barrell, B.G.. Brown, N.L., Coulson, A.R., Fiddes,

J.C., Hutchinson III, C.A., Slocombe, P.A. and Smith, M. (1977) Nature265, 687-695.

39. Geacintov, N.E., Gagliano, A.G., lbanez, V. and Harvey. R.G. (1982)Carcinogenesis 3, 247-253,

40. Jernstr6m, B., Lycksell, P.O., Graslund, A. and Norddn, B. (1984)Carcinogenesis 5. 1129-- 1135.

41. Eriksson, M.. Nord~n, B., Jernstr6m. B. and Graslund, A. (1988)Biochemistry (USA) 27, 1213-- 1221.

42. Kolubayev, V., Brenner, H.C, and Geacintov, N.E. (1987) Biochemistry(USA) 26, 2638-2641,

43. Kim, S.-K., Brenner. H.C., Soh. B.J. and Geacintov. N.E. (1989)Photochem. Photobiol. 50, 327-337.

44. MacLeod, M.C. and Tang, M, (1985) Cancer Res. 45, 51-56.45. Keller, W. (1975) Proc. Natl. Acad. Sci. (USA) 72, 4876-4880.46 Wiesehahn, G. and Hearst, J.E. (1978) Proc. Nail. Acad. Sri. (USA) 75,

2703 -2707.47. Hogan, M.E.. Da)nagupta, N and Whitlock, JP.. Jr. (1981)J. BioL Chem.

256, 4504-4513.

DISTRIBUTION LIST

DEPARTMENT OF DEFENSEU.S. ARMY RESEARCH INSTITUTE OF ENVIRONMENTAL MEDICINE

ARMED FORCES INSTITUTE OF PATHOLOGY ATTN: SGRD-UE-RPPATTN: RADIOLOGIC PATHOLOGY DEPARTMENT

U.S. ARMY RESEARCH OFFICEARMED FORCES RADIOBIOLOGY RESEARCH INSTITUTE ATT-N: BIOLOGICAL SCIENCES PROGRAM

ATTN: PUBLICATIONS DIVISIONATTN: LIBRARY WALTER REED ARMY INSTITUTE OF RESEARCH

ATTN: DIVISION OF EXPERIMENTAL THERAPEUTICSARMY/AIR FORCE JOINT MEDICAL LIBRARY

ATTN: DASG-AAFJMLDEPARTMENT OF THE NAVY

ASSISTANT TO SECRETARY OF DEFENSEATTN: AE NAVAL AEROSPACE MEDICAL RESEARCH LABORATORYATTN: HA(IA) ATTN: COMMANDING OFFICER

DEFENSE NUCLEAR AGENCY NAVAL MEDICAL COMMANDATTN: TITL ATTN: MEDCOM-21ATTN: DDIRATTN: RARP NAVAL MEDICAL RESEARCH AND DEVELOPMENT COMMANDATTN: MID ATTN: CODE 40C

DEFENSE TECHNICAL INFORMATION CENTER NAVAL MEDICAL RESEARCH INSTITUTEATTN: DTIC-DDAC AT-N: LIBRARYATTN: DTIC-FDAC

NAVAL RESEARCH LABORATORYFIELD COMMAND DEFENSE NUCLEAR AGENCY ATTN: LIBRARY

ATTN: FCFSOFFICE OF NAVAL RESEARCH

INTERSERVICE NUCLEAR WEAPONS SCHOOL ATTN: BIOLOGICAL SCIENCES DIVISIONATTN: TCHTS/RH

SURGEON GENERAL OF THE NAVYLAWRENCE LIVERMORE NATIONAL LABORATORY ATTN: MEDICAL RESEARCH AND DEVELOPEMENT

ATTN: LIBRARY

UNDER SECRETARY OF DEFENSE (ACQUISITION) DEPARTMENT OF THE AIR FORCEATTN: OUSD(A)/R&AT

BOLLING AIR FORCE BASEUNIFORMED SERVICES UNIVERSITY OF THE HEALTH SCIENCES ATTN: AFOSR

ATTN: LIBRARYBROOKS AIR FORCE BASE

ATrN: ALUOEBSCDEPARTMENT OF THE ARMY ATTN: USAFSAM/RZ

ATTN: AL.DEBLHARRY DIAMOND LABORATORIES

ATTN: SLCHD-NW NUCLEAR CRITERIA GROUP, SECRETARIATATTN: SLCSM-SE ATTN: OAS/XRS

LETTERMAN ARMY INSTITUTE OF RESEARCH SURGEON GENERAL OF THE AIR FORCEATTN: SGRD-ULY-OH ATTN: HO USAF/SGPT

ATT-N: HO USAF/SGESSURGEON GENERAL OF THE ARMY

ATTN: MEDDH-N U.S. AIR FORCE ACADEMYATTN: HO USAFA/DFBL

U S. ARMY AEROMEDICAL RESEARCH LABORATORYATTN: SCIENTIFIC INFORMATION CENTER

OTHER FEDERAL GOVERNMENTU S. ARMY ACADEMY OF HEALTH SCIENCES

ATTN: HSMC-FCM ARGONNE NATIONAL LABORATORYATTN: ACQUISITIONS

U.S. ARMY CHEMICAL RESEARCH, DEVELOPMENT, ANDENGINEERING CENTER BROOKHAVEN NATIONAL LABORATORY

ATTN: SMCCR-RST ATTN: RESEARCH LIBRARY, REPORTS SECTION

U.S. ARMY INSTITUTE OF SURGICAL RESEARCH CENTER FOR DEVICES AND RADIOLOGICAL HEALTHATTN: DIRECTOR OF RESEARCH ATTN: HFZ.110

U S. ARMY MEDICAL RESEARCH INSTITUTE OF CHEMICAL DEFENSE GOVERNMENT PRINTING OFFICEATTN: SGRD-UV-R ATTN: DEPOSITORY RECEIVING SECTION

ATTN: CONSIGNED BRANCHU S ARMY NUCLEAR AND CHEMICAL AGENCY

ATTN: MONA-NU

LIBRARY OF CONGRESS REStLARCH AND OTHER ORGANIZATIONSATTN: UNIT X

BRITISH LIBRARY (SERIAL ACQUISITIONS)LOS ALAMOS NATIONAL LABORATORY ATTN: DOCUMENT SUPPLY CENTRE

ATTN: REPORT LIBRARY/P364CENTRE DE RECHERCHES DU SERVICE DE SANTE DES ARMEES

NATIONAL AERONAUTICS AND SPACE ADMINISTRATION ATTN: DIRECTORATTN: RADLAB

INHALATION TOXICOLOGY RESEARCH INSTITUTENATIONAL AERONAUTICS AND SPACE ADMINIS'RATION, GODDARD ATTN: LIBRARYSPACE FLIGHT CENTER

ATTN: LIBRARY INSTITUTE OF RADIOBIOLOGYARMED FORCES MEDICAL ACADEMY

NATIONAL CANCER INSTITUTE ATTN: DIRECTORATTN- RADIATION RESEARCH PROGRAM

KAMAN SCIENCES CORPORATIONNATIONAL DEFENSE UNIVERSITY ATTN: DASIAC

ATTN: LIBRARYNBC DEFENSE RESEARCH AND DEVELOPMENT CENTER OF THE

NATIONAL LIBRARY OF MEDICINE FEDERAL ARMED FORCESATTN: OPI ATTN: WWDBW ABC-SCHUTZ

U S. ATOMIC ENERGY COMMISSION NCTR-ASSOCIATED UNIVERSITIESATTN: BETHESDA TECHNICAL LIBRARY ATTN: EXECUTIVE DIRECTOR

U.S. DEPARTMENT OF ENERGY RUTGERS UNIVERSITYATTN: LIBRARY ATTN: LIBRARY OF SCIENCE AND MEDICINE

U.S. FOOD AND DRUG ADMINISTRATION UNIVERSITY OF CALIFORNIAATTN: WINCHESTER ENGINEERING AND ATTN: LABORATORY FOR ENERGY-RELATED

ANALYTICAL CENTER HEALTH RESEARCHATTN: LAWRENCE BERKELEY LABORATORY

U.S NUCLEAR REGULA rORY COMMISSIONATTN: LIBRARY UNIVERSITY OF CINCINNATI

ATTN: UNIVERSITY HOSPITAL, RADIOISOTOPELABORATORY