BrJCancer197735439EFFECTSOFDEXAMETHASONEANDBETAMETHASONEONINVITRO - PDF document

1 Br.J.Cancer(1977)35,439EFFECTSOFDEXAMETH
Br.J.Cancer(1977)35,439EFFECTSOFDEXAMETHASONEANDBETAMETHASONEONINVITROCULTURESFROMHUMANASTROCYTOMAM.GUNER*,R.I.FRESHNEY,D.MORGAN,M.G.FRESHNEY,D.G.T.THOMASANDD.I.GRAHAMFromtheBeatsonInstituteforCancerResearch,Bearsden,Glasgow,andtheInstituteofNeurologicalSciences,GlasgowReceived12October1976Accepted16 M.GUNERETAL.MATERIALSANDMETHODSCultureswerederivedfrom5differentastrocytomas(Table)bydissociationincollagenase(Freshney,1972).Asepticallycollectedbiopsyspecimenswerechoppedfinely,washedandtransferredtocultureflasks.Growthmedium(Ham'sF12plusEagle'sMEMaminoacids(FlowLabora-tories),non-essentialaminoacids(FlowLaboratories)andsupplementedwith20%foetalbovineserum(Gibco-Biocult))wasTABLEAge2445686025SexFMFMMCulturedesignationACHWEWHLRWLYPTAHistologicalclassificationAnaplasticastrocytomaAnaplasticastrocytomaAnaplasticastrocytomaAnaplasticastrocytomaIntermediategradeastrocytomaaddedtogiveapproximately5-20mgtissueperml.Sinceprimarycultureandsub-sequentcloningwereperformedinsealedbottleswithairasgasphase,andbufferingwasachievedby20mMHEPES,thebicar-bonateconcentrationwaslow(4mmatthetimeofpreparation).Collagenase(CLSgrade,Worthington)wasaddedat200u/mlfinalconcentration,andthetissuefragmentswereincubatedfor24-48hat365°C.Thetissuewasthendissociatedwithgentlepipetting,centrifugedtoremovecollagenase,andresuspendedinfreshgrowthmediumat5X104to5X105cells/ml,dependingontheyield.Thecellsusuallyattachedwithin24-48handwerereadyfortrypsinizationinabout1-2weeks.Cellsforassayweretakenafterthesecondtrypsinization(1-2weeksafterthefirst)andwereclonedbydilutingto50-100cells/mlandinoculating120-cm2glassbottleswith2500-5000cellsperbottlein50mlgrowthmediumcon-tainingarangeofconcentrationsofdexa-methasoneorbetamethasone.Acromegalicpituitaryandnormalbrainculturesweretreatedasforastrocytoma.MDHcells(ratminimaldeviationhepatoma(Pitotetal.,1964)weregrowninthesamemediumandclonedat25cells/mlin50-cm2glassbottles.MRC5cellsdidnotclonesatis-factorilyundertheseconditions,andwereclonedin5-cmPetridishes(Lux)inmediumMCDB104(McKeehanetal.,1976)with2%foetalbovineserum(FlowLaboratories).DexamethasonewasobtainedfromMerck,Sharpe&DohmeLtdasDecadron.Beta-methasoneasBetnesolwasobtainedfromGlaxoLaboratoriesLtd.TheywerebothdilutedinHanks'balancedsaltsolutionandaddeddirectlytotheculturemedium.Noaddition

2 wasmadetocontrols.Clonalgrowthcultureswe
wasmadetocontrols.Clonalgrowthculturesweremaintainedat36.50Cwithoutmediumchangefor2-3weeks.TheywerethenwashedinHanks'BSS,fixedfor10mininmethanolandstainedinGiemsafor10min.Thecolonieswerecountedunderadissectingmicroscopeusingapreparedmaskwiththreestandardfieldsof900mm2atdifferentpartsofthebottle.Allthecolonieswithineachfieldwerecountedandanextrapolationmadetothetotalsurfaceoftheflask.Thisfigurewasthenusedtocalculatethecloningefficiency(numberofcoloniesperbottleX100numberofcellsinoculatedAmicroscopewasusedtocountcolonies,toovercomethedifficultyofcountingsmalldiffusecoloniesfoundparticularlyincontrols.Onlycoloniesover16cellswerescored.Thisfigurewaschosenarbitrarilytoincludetheproductofatleast4cellgenerationsandavoidproblemsarisingfromcountingsmallercolonieswhichmaylaterabort.Colonysizedeterminationswereper-formedbycountingthecellspercolonyinapproximately50coloniesfromdifferentpartsoftheflask.Thiswasdoneinoneexperimentwithdexamethasoneandinthreewithbetamethasone,butallexperi-mentsperformedwithastrocytomacellshaveshownincreasesincolonysizewhicharereadilydetectedbythenakedeye.Chromosomepreparationsweremadebytreatingcoverslipsbearingcellsinexpo-nentialgrowthwith0.004%colchicinefor4h.Theywerethenfixedinaceticmethanol,driedandstainedwithGiemsaandcounted.Atleast20spreadswerecountedforeachcellstrain.Autoradiographswereperformedoncover-slipculturesgrowninLeightontubesandlabelledforupto96hwith[3H]thymidine,0-01,Ci/ml(2.0Ci/mmol).Theywerewash-edinBSS,fixedinmethanolanddried.Aftermountingthecoverslipsonslides,acid-solubleprecursorswereextractedin10%ice-coldtrichloroaceticacidandtheslidesrewashedthoroughlyincolddistilledwater.440 STEROIDSANDCULTUREDASTROCYTOMAStrippingfilm(KodakAR-10)wasappliedandtheslidesexposedfor3weeksat40C.TheyweredevelopedinKodakD19for10min,washed,fixedinIlfordHypamfor2min,washedfor15minanddried.TheywerethenstainedinGiemsaandtheper-centagelabelledcellsdetermined.Electromticroscopy.Afterthesecondtryp-sinization,about5x106cellswerefixedwithchilled2%glutaraldehydein0-2Mcacodylatebufferfor20min.Theywererinsedinthesamebufferpriortopost-fixationwithcacodylate-buffered1%osmiumtetrox-idefor1h,dehydrationingradedethanolsolutions,clearinginpropyleneoxide,andembeddingaspelletsinaraldite.ThinsectionswerecutwithaglassknifeonaLKBultramicrotome,stain

3 edwithuranylacetateandleadcitrate,andexa
edwithuranylacetateandleadcitrate,andexaminedinaPhilipsEM201electronmicroscope.RESULTSIdentityofcellsAllcultureswerederivedfromastro-cytomabycollagenasedigestion.Al-thoughcultureshavebeenobtainedbyusfromnormalbrainbythismethod,itisunusualtoobtainayieldorinitialgrowthratecomparabletothatfoundwithtumourmaterial.Thecytologyofthecellsusedforclonalgrowthstudieswastypicalofastrocytomastrainscul-turedhere.Thecellswerespindle-shapedwithverylongprocessesformingareticularnetwork(Fig.Ia).Therewasnoindicationoffibroblasticcontamination.Electronmicrographsshowedthatthemajorityofthecellsincultureweremorphologicallysimilar(Fig.2aandinset).Thenucleuswasirregularinoutline,withclumpsofheterochromatinattachedtotheinnermarginofthenuclearmembrane,andtherewasoftenaprominentnucleolus.ThecytoplasmofthesecellscontainedaGolgicomplex,moderatequantitiesofsmoothandroughendoplasmicreticulum,avariablenumberofmultivesicularbodiesandfreeribo-somes,7-10-nm-thickfilaments,theoccasionaldensebodyandmultiple"condensed"mitochondria.Pinocytotic31vesicleswerenotprominentandtubularbodieswerenotseen.Aminority(15-20%/)ofthecellscontainedlargenumbersofvesicles,aprominentperinuclearGolgi,clustersofJ.astiocytoma.(a)Livingprimaryculture1wveekafterdissociationincollagenase.Phasecontiast.(b)Partofacloneinalowdensitysecondaryculture.Giemsastained.(c)Similarctulturetreatedwith10,ug/mlbetamethasone.(a)x70,(b)and(c)x:30.441 4M.GUNERETAL.FiG.2(a).Electronmicroscopyofsecondaryculturesfromastrocytomabiopsies.Uranylacetateandleadcitrate.Majoritycelltype.Cellsvariedinsizeandshape.Thenucleus(N)oftencontainedanucleolusandthecytoplasmaperinuclearGoldicomplex(G),"condensed"mito-chondria(M),segmentsofendoplasmicreticulum(RER),clispersedlribosomesandmultivesicularbodies(MV).x16,500.InsetCytoplasmofsimilarcellshowinganabundanceofinterlacingfilaments(F),Mitochondrion=M.x25,000.freeribosomes,bundlesoffinefilamentsand"condensed"mitochondria.Insome,thecisternaeoftheendoplasmicreticulumweredilatedandcontainedgranularmaterial.Althoughnoneofthecellscontainedrod-shapedcytoplasmicinclusions,largeunitmembrane-boundvacuoleswitharelativelyelectron-lucentmatrixandcontainingafewtubuleswereseeninthiscomponent(Fig.2b).Onlytheoccasionalsmooth-musclecellwasseenandthereappearedtobeacompleteabsenceoffibroblasts.Thecytoplasmof

4 mostcellscontainedmoder-atelysizedclumps
mostcellscontainedmoder-atelysizedclumpsofglycogen.ThechromosomenumbersoftwolinesarepresentedinFig.3.Althoughdiploidcellsarepresent,thereisevidenceofconsiderableaneuploidy,withchromo-somenumbersrangingfrom28to56.Chromosomecountswerenotavailablefromallthecultures,but5othersexaminedherehaveshownsimilarhypodiploiddistributionstothosereportedabove.Althoughtheseresultsarenotconclusive,thereisconsiderableevidenceelsewherethatnormalhumancellsremaindiploidinculturewhileneoplasticcellsbecomeaneuploid(Jones,1974).Chromosomalanalysisofaculturefromnormalbrainhandledinexactlythesamemannerastheastrocytomaculturesgaveamodalnumberof46withagreatlyreducedspread.GrowthpotentialofsecondaryculturesPopulationdoublingtimeshavebeenestimatedwith10differentsecondaryculturesfromgliomasbycountingthenumberofcellsperwellinmicrotitration442ol'IIVrglAff,..:.-1.1m.1.."Ol STEROIDSANDCULTUREDASTROCYTOMAFIG.2(b).-Minoritycelltype(15-20%ofpopulation).Predominanteuchromatinnucleus(N),largeperinuclearGolgicomplex(G)withmanyvesicles,"condensed"mitochondria,occasionaldensebody(LY),dilatedsegmentsofendoplasmicreticulum(RER)andlargevacuQleswithsmalltubularstructuresembeddedinamorphousmatrix(V).x19,800.8(fl7=650432am86s43226283032343638404244464850525456ChromosomeNumberFIG.3.Chromosomedistributionoftwolines.Coverslipculturesweretreatedasin"Methods".(a)34intactspreadsofWLY,and(b)35ofWEW.aplateculturesgrownfrom2x104cells/mluptoabout2x105cells/ml.Culturesfromdifferenttumoursgavedifferentdoublingtimeswhich,takenatmid-logphase,variedfrom30to60h,withmostbeingaround40h.Sixdifferentcultureslabelledfor3and4dayswith0-01jtCi/ml[3H]thymidinegavelabellingindicesbetween88and97%,implyingthatthebbulkofthepopulationinsuchcultureisincycle.Althoughlowerlabellingindicesmightbeanticipatedwithprimarycultures,secondarycultureswereusedforthisstudy,witharesultantincreaseintheproportionofproliferatingcells.DexamethasoneandbetamethasonetreatmentThreeexperimentswereperformedwithdexamethasoneand3withbeta-methasone;HLRwastreatedwithbothsteroids,andtheotherswithonesteroid11111111-11I-.--1---1I.-.-I.-----1.I443 444~~~~~M.GUNERETAL..r-250?E2000C)$15014)1L.1004)5001-1iIjg/mlIDexamethasoneBetamethasoneSteroidConcentrationFiG.4.Cloningefficiencyinthepresenceofsteroids.Secondaryculturesofthelinesindicatedweretrypsinized,

5 dilutedto50or100cells/mlandinoculatedint
dilutedto50or100cells/mlandinoculatedintoglassbottlesof120CM2surfaceareain50mlSF12/20containingtheconcentrationofsteroidindicated.Cloningefficienciesweredeterminedafter3weeks'cultureandareplottedas%stimulationorinhibitionofcontrolsclonedwithoutsteroids.Thecloningefficiencyofcontrolsisquotedatthetopofeachhistogram.Molarconcn.abouttwicetheg/lconcn.only.Determinationofcloningefficiencyshowedasimilarpatternineachexperiment(Fig.4)withmaximumcloningefficiencyat12-5jtg/ml(,-.25mm)withbothhor-mones.Cytotoxicitycouldonlybede-monstratedat50/ag/ml(4cases,3withdexamethasoneandwithbetametha-sone)and25/kg/ml(1case,dexametha-sone).Thecloningefficiencywasmorethandoubledin4casesinthepresenceof12-5/tg/mlsteroid.Visualexaminationofthecoloniesshowedacleardistinctionbetweentreatedanduntreatedcolonies(Fig.lb,c).Untreatedcolonieswerediffuseandcon-tainedfewercells,whiletreatedcolonies(exceptatmaximumsteroiddoses)weremuchdenserandcontainedmanymorecells.Whenthenumberofcellspercolonywascounted(Fig.5),anincreaseincolonysizewasobservedwithincreasingconcentrationofbetamethasone,reachingamaximumat12-5/ag/ml,whenthebulkofthecoloniescontainedmorethan60cells(6doublings,andthemaximumnumberofcellsreadilycountedbyeye).Athigherconcentrations,theaveragecolonysizewassmallerthanat12-5Itg/ml,althoughtheproportionofcoloniescontainingmorethan60cellswasstillhigherat50/-tg/mlthaninuntreatedcontrols.Colonysizedeter-minationswith2othergliomacelllinesconfirmedthispattern.Similarresultswereobtainedwithdexamethasone,whichalsogavemaximumstimulationat12*5/ag/ml(i-..25JLm).Higherconcentrationswerelesseffectiveinstimulatingclonalgrowththantheequivalentconcentrationsofbetamethasone,suiggestingthatdexa-methasonemaybemoretoxic.SpecificityCulturesderivedfromnormalbrain,clonedunderthesameconditionsasastrocytoma-derivedcultures,showedaWWControl=i0.e%3O~56./.7--I1--60L-A--i-jL-LL..ULL-J--L-LL-J==iLL-1-iL-L-.JULL-i-jL-.LL-A--LIiL-LL-LL-AL-L-jJ--L-.L.--IL-II444HLRI5.6%.mIny/.3.A--.81F--'2.vLF7LF!--"I.--...1.1-1-0125IM255001-25IM255001-2512525500P2512525500VZ126255001-0"14v."I.--w-awvrd;;,-0--.r-mm STEROIDSANDCULTUREDASTROCYTOAIAuca,._iUcnc07Z10987654320(°20Uno10.OFic(.5.ColonysizedistributionofACHcellstireate(dwithdexamethasoneThenumberofcellspercolonyinatleast50colonieswascountedateachconcentration.Itwa

6 snotpossibletoobtainanaccuratecellcounta
snotpossibletoobtainanaccuratecellcountabove60cellspercolony.The%colonieswithmorethan60cells(6ormoredloublings)isshowninthebottomright-handfigure.decreaseincloningefficiencyandcolonysizeatallconcentrationsofbothbeta-methasoneanddexamethasone(Fig.6).Alineofminimaldeviationhepatomacells(MDH),whichhavebeenshowntorespondtodexamethasoneinthislaboratory(Sommerville,personalcom-munication),platedwithandwithoutdexamethasone,gaveacloningefficiencyof230oincontrolsbutnodiscerniblecoloniesatallin125,ug/mlor125,ug/mldexamethasone.Anacromegalicpituitarycellculture,ofsimilaragetotheastrocytomaculturesused,gaveacloningefficiencyof39.400with12-5,jtg/mldexamethasoneand16.9%without.However,thecolonysizedis-~~~~~a...AIC.lb.01.3612255001.36122550,ug/mLFICG.6.Effectofsteroidsonclonalgrowthofcellculturederivedfromnormalbrain.Secondaryculturesweretrypsinizedandinoculatedat200cells/mlin20mlina75-cm2Falconflask.Theflaskswerefixedandstainedafter3weeks.(a)Cloningefficiencyindifferentconcentrationsofbetamethasoine,(b)cloningefficiencyindifferentconcentrationsofdexamethasone,(c)%colonieswith�60cellsinbeta-methasone(d)%colonieswith�60cellsindexamethasone.tributionwasdifferentfromastrocytoma,as500ofthecontrolshadmorethan60cellspercolony,whileonly270oofthetreatedsampleexceeded60cellspercolony,after3weeks'growth.MRC5humandiploidfibroblastsclonedunderdifferentconditions(mediumMCDB104,50mmHEPES,500CO2gasphase,and2%foetalbovineserum)displayedatwo-foldincreaseincloningefficiencyinthepresenceof8,ug/mlbetamethasone.Undertheseconditionscolonysizewasunaffectedbybetamethasoneconcentra-tionsfrom2to16/ag/ml.However,inreducedserumconcentrations,colonysizewasinhibitedbyincreasingdosesofbetamethasone,withmorethanthree-folddiminutioninaveragecolonysize-445c.d.%A.r-i M.GUNERETAL.beingachievedin0-4%serumwith8jtg/mlbetamethasone.DISCUSSIONAproblemencounteredinstudyingtissueculturesderivedfromtumoursisestablishingthecelltypebymorpho-logicalcriteria.Thegeneralproblemofmorphologicalidentificationarisesbecauseofthelackofspecificityoftheap-pearancesofmanycelllinesfromdiversesourcesafteradaptationtoculture(Wein-steinandKornblith,1971).However,thelightmicroscopicconfigurationofastrocytesisgenerallyretainedinshort-termculturethoughitbecomeslessapparentinlatergenerations(Lumsden,1971).Si

7 milarlytheelectronmicroscopicfeaturesofg
milarlytheelectronmicroscopicfeaturesofgliomacellsincultureretainmanyofthefeaturesofneoplasticastro-cytes(WeinsteinandKornblith,1971;Macintyre,PontenandVatter,1972).Theprincipaltypeofcellusedinthepresentstudywasconsideredtobeastrocyticinnature.Thecultures,how-ever,werenotpure,astheoccasionalsmoothmusclecellwasseen,andaproportion(15-20%)ofthecellsmayhavebeenendothelialinorigin(Jaffeetat.,1973;Kawamuraetal.,1974;Hauden-schildetal.,1975),thoughthecharac-teristicrod-shapedcytoplasmicinclusionsfirstdescribedbyWeibelandPalade(1964)werenotseen.Thereisabundantevidencethatdexa-methasonecanhaveadirecteffectontheregulationofcellmetabolismbyenzymeinduction(e.g.Levinson,TomkinsandStellwagen,1971),withregulationprobablyoccurringatthetranscriptionallevel.Littleevidenceexists,however,forageneralizedcytostaticaction.Someauthorshaveshownthatratembryocells(Wrightetal.,1969)andhumanglio-blastoma(Mealeyetat.,1971)aresensitivetohighlevelsofsteroids,andconcentra-tionsof50-100,g/mlhydrocortisone,pred-nisolone,anddexamethasoneproducedcytotoxicityasmeasuredbyreducedmonolayergrowthandcytologicaldamage.Assumingamaximuminvivodosageof24mgorallyevery6h,itisunlikelythattheplasmalevelsinanindividualexceed5-10jug/ml.Theten-foldexcessnecessaryforcytotoxicitywouldbediffi-culttoachieveunlessactiveconcentrationorbindingoccurs,resultinginanunevendistributionbetweentissueandplasmacompartments.Theresultsobtainedhereconfirmpreviousobservationsthatsteroidsmaybecomecytotoxicathigherthanphysio-logicallevels,butnotattheconcentra-tionsnormallyanticipatedduringtreat-ment,particularlyviaoraladministration,theroutecommonlyemployedfortreat-mentofpatientswithbraintumours.BallardandTomkins(1969)andothers(Iype,personalcommunication)haveshownthatcelladhesionisenhancedinthepresenceofdexamethasone,possiblyviaamodificationincellsurfaceglyco-proteins.Thismayhavecontributedtotheincreaseinplatingefficiencyandreductionincellmigrationobservedduringclonalanalysis.Itisimportanttonotethattheobservedincreasesincloningefficiencywereaccompaniedbymarkedchangesinthesizedistributionofthecolonies,implyinganincreaseincellproliferation.Preliminaryresultswithahumanpituitarytumourculturehaveshownadoublingincloningeffi-ciencyindexamethasonebutconsiderablereductionincolonysize.Similarly,MRC5fibroblastshadahighercloning

8 efficiencyinbetamethasonebutformedsmalle
efficiencyinbetamethasonebutformedsmallercol-onies.Thisdemonstratesthatstimulationofplatingefficiencydoesnotnecessarilyleadtoanincreaseinclonalgrowth.Itispossiblethatincreasedcellattachment,alteringtheviablecellcon-centration,mayinfluenceaconditioningofthemedium.However,conditioningisusuallyperformedatmuchhighercellconcentrations,andfeeder-layereffectsarelostbelow10,000cells/ml(Mac-phersonandBryden,1971).Evenatthehighestcloningefficiencyinthepresentseriesofexperiments(26%,WLYplusbetamethasone)theviablecellcon-centrationwouldonlybeapproximately446 STEROIDSANDCULTUREDASTROCYTOMA44712cells/mlafterplatingandwouldnotreach10,000cells/mluntilabout10generations(about1000cells/colony),bywhichtimetheeffectisalreadyapparent.Furtherattemptstoinvestigatethespecificityoftheresponsetodexametha-sonebytreatingratminimaldeviationhepatomacells,whichareknowntorespondtodexamethasone,MRC5fibro-blasts,andcellculturesderivedfromhumanbrain,confirmedthatthestimula-tionofclonalgrowthisspecifictoastro-cytomacultures,althoughtheeffectoncloningefficiencymaybemoregeneral.Furthermore,thedifferenceintheresponsebetweennormalbrain-derivedculturesandthosederivedfromastrocytomaimpliesthatthecellsculturedfromastrocytomaarequalitativelydifferentfromtheendothelial-likecellsobservedinnormalbraincultures.Stimulationofcellproliferationin-dicatedherehasseriousimplicationsforinvivoadministration.However,itshouldbeemphasizedthattheseobservationsaremadeincellcultureatverylowcelldensitiesandsofarnoattempthasbeenmadetoconfirmthisinvivo.Atpresent,thesesteroidshaveundoubtedclinicaladvantagesinthemanagementofpatientswithbraintumoursandtheproperrelevanceofthesefindingsmustawaitfurtherinvitroandinvivoobservations.ThisworkwassupportedbygrantsfromtheMedicalResearchCouncilandtheCancerResearchCampaign,thePeelMedi-calTrustandtheHormoneResearchTrust.DrFreshneyisgratefultoDrR.G.HamandDrW.McKeehanforhelpandadviceduringavisittotheirlaboratoryinBoulder,Colorado.ThesupportoftheYamagiwaYoshidaMemorialRe-searchTrustisgratefullyacknowledgedformakingthisvisitpossible.REFERENCESBALLA,RD,P.L.&To-mKINs,G.Al.(1969)Dexa-methasoneandCellAdhesion.Vature,Lond.,224,344.BECKER,D.P.,YOUNG,H.F.&VRaEs,J.K.(1975)MonitoringinPatientswithBrainTumours.Clin.Neurosurq.,22,364.FREI,E.,III(1972)CombinationCancerTherapy.CancerRes.

9 ,32,2593.FRESHNEY,R.I.(1972)TumourCellsD
,32,2593.FRESHNEY,R.I.(1972)TumourCellsDisaggregatedinCollagenase.Lancet,ii,488.GITTIN,P.J.(1975)CorticosteroidTherapyinPatientswithCerebralTumors:Benefits,Mechan-isms,Problems,Practicalities.SeminarsinOnco-logy,2,49.HAUDENSCHILD,C.C.,COTRAN,R.S.,GIMBRONE,AI.A.&FOLKMAN,J.(1975)FineStructureofVascularEndotheliuminCulture.J.Ultra-struct.Res.,50,22.JAFFE,E.A.,NACHMAN,R.L.,BECKER,C.G.&MINICK,C.R.(1973)CultureofHumanEndo-thelialCellsDerivedfromUmbilicalVeins.IdentificationbyMorphologicandImmunologicCriteria.J.clin.Invest.,52,2745.JONES,K.W.(1974)ChromosomesandMalignancy.Nature,Lond.,252,525.KAWAMIITRA,J.,KAMIJYO,Y.,SUNAGA,T.&NELSON,E.(1974)TubularBodiesinVascularEndothe-liumofaCerebellarNeoplasm.Lab.Invest.,30,358.LEviNsoN,B.B.,TOMKINS,G.M.&STELLWAGEN,R.H.(1971)TheRegulationofTyrosineAmino-transferaseSynthesis.StudtiesinHTCCellswithInhibitorsofRNASynthesis.J.biol.Chem.,246,6297.LUMISDEN,C.E.(1971)TheStudybyTissueCultureofTumoursoftheNervousSystem.InPatho-logyofTumoursoftheNervousSystem.EdsD.S.RtussellandL.J.Rubinstein.London:Arnold.p.334.MACINTYRE,E.H.,PONTfIN,J.&VATTER,A.E.(1972)TheUltrastructureofHumanandMurineAstrocytesandofHumanFibroblastsinCulture.Actapath.mticrobiol.Scand.,Sect.A,80,267.MCKEEHAN,W.L.,MICKEEHAN,K.A.,HAMMOND,S.L.&HAM,R.G.(1976)ImprovedMediumforClonalGrowthofHumanDiploidCellsatLowConcentrationsofSerumProtein.InVitro,inpress.MACPHERSON,I.&BRYDEN,A.(1971)MitomycinCTreatedCellsasFeeders.Exp.CellRes.,69,240.MEALEY,J.,CHEN,T.T.&SCHANZ,G.P.(1971)EffectsofDexamethasoneandMiethylPredni-soloneinCellCulturesofHumanGlioblastoma.J.Neurosurg.,34,324.PITOT,H.C.,PERAINO,C.,MORSE,P.A.,JR.&POTTER,V.R.(1964)HepatomasinTissueCultureComparedwithAdaptingLiverInvivo.NatlCancerIntst.Monogr.,13,229.WEIBEL,E.R.&PALADE,G.E.(1964)NewCyto-plasmicComponentsinArterialEndothelia.J.CellBiol.,23,101.WEINSTEIN,R.S.&KORNBLITH,P.L.(1971)UltrastructtureofaClonedAstrocytomainTissueCulture.Canicer,N.Y.,27,1174.WELLING'TON,J.S.&MooN,H.D.(1961)EffectofHydrocortisoneonHumanCellsinTissueCulture.Proc.Soc.exp.Biol.lMed.,107,556.WELLINGS,S.R.&Moox,H.D.(1961)MorphologicandFunctionalEffectsofHydrocortisoneinTissueCtulture.Lab.Invest.,10,539.WRIGHT,R.L.,SHAUMNIBA,B.&KELLER,J.(1969)TheEffectofGlucocorticosteroidsinGrowthandAletabolismofExperimentalGlialTumours.J.iNeurosurg.,3